GEOLOGICAL SURVEY PROFESSIONAL PAPER 1052

Digitized by Google
 fEB 26 1980
Ill3lili~l1822006529812
lIrl ~I ~~li~i~llllml l'~ll[1 1

University of California. San Diego

Pl.... Note: This item is subject to recall.
Date Due

998

CI3aI(4t91} UCSDLII.

Digitized by Google
Digitized by Google
r"'1f.~~===-~---- . .. ---------
-~

STAR DUNE WITH CURVING ARMS, south of Kuiseb River, Namib Desert, South-West
Africa. Photograph by E. Tad Nichols. (Cover.)

Digitized by Google
Digitized by Google
 A STUDY OF
GLOBAL SAND SEAS

Digitized by Google ....

COMPLEX LINEAR RIDGE with reversing and star dunes superimposed; broad interdunes
on each side. Aerial photograph, taken southwest of the Kuiseb River, Namib Desert, South-
West Africa, by E. Tad Nichols. (Frontispiece.)

Digitized by Google
A Study of
Global Sand Seas
Edwin D. McKee, Editor

GEOLOGICAL SURVEY PROFESSIONAL PAPER 1052

Prepared in cooperation with the

National Aeronautics and Space Administration

UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1979

Digitized by Google
 DSTATESDEP THE INTERIO
CECIL D. ANDRUS, Secretary

GEOLOGICAL SURVEY
H. William tor

Library of Congr ublication Data

Main entry under global sand seas.
(Geological Surve er 1052)
Bibliography.
Includes index.
1. Sand. 2. Sand , Edwin D., 1906. es National Aerona
Space Administration. III. Series: United States Geological Survey Professional Paper 1052.
QE471.2.S88 553'.62 79-1452

For sale by the Superintendent of Documents, U.S. Government Printing Office

VVashin 2
Stock Num 202-3
 FOREWORD
~ NE HUNDRED YEARS AGO, the U.S. Geological Survey had its in-
~ ception in Washington, D.C., when Congress established a unified
geological survey under the Department of the Interior, combining the
duties of and replacing four earlier government surveys. The Survey's
program included "classification of the public land, and examination of
the geological structure, mineral resources, and products of the national
domain. It seems entirely appropriate, therefore, that a professional
It

paper summarizing the progress of investigations and synthesizing our

present state of knowledge in the important field of eolian deposition be
one of those issued to commemorate the Geological Survey's 100th an-
niversary.
Early descriptions and considerations of the characteristics of modern
eolian sand deposits, mostly in the great sand seas of the world, date back
to the 1880's and 1890's, when pioneer geologists and explorers, including
Walther, von Zittel, and Sven Hedin, wrote their classic papers on desert
dunes. One or two decades later, interest in eolian processes greatly in-
creased when wind-formed deposits were recognized in ancient
sandstones in many parts of the world and in rocks of many ages.
Throughout the 20th century, as the science of geology has expanded
and the programs of the U.S. Geological Survey have proliferated to keep
pace, dune studies have had a similar growth. Work was initially con-
centrated mostly on the description of dune forms or morphology and on
analysis of textural features; by midcentury, however, major contribu-
tions had been made to the physics of eolian sand, as exemplified by the
classic work of Bagnold, by detailed studies and interpretations of minor
eolian structures, by statistical analyses of cross-strata dip directions, and
by the development of systems for dune classification. Most recently, in-
terest has been renewed in detailed grain studies, in the study of cross-
strata, and in interpretation of dune patterns by means of aerial photo-
graphs and Landsat imagery.
A major feature of this report on global sand seas is the compilation and
comparison of available data based on many different methods of in-
vestigation. Evidence at one extreme is obtained from the detailed studies
of minute particles and from analysis of individual grains. At the opposite
extreme is evidence obtained from remote sensing, in which dune pat-
terns, recorded from approximately 500 miles in space, are compared from
one sand sea to another. Furthermore, comparisons are made between an-
cient and modern deposits with respect to textures, structures, and other
characteristics. Criteria for recognizing the eolian origin of various an-
cient deposits are discussed. The application of these studies to economic
problems, which is described in one chapter, clearly illustrates the impor-
tance of eolian deposits to our present culture and to human welfare.

H. William Menard
Director, U.S. Geological Survey

VII

Digitized by Google
Digitized by Google
 Contents
[For convenience in cross-referencin& each chapter of this publication has a letter designation)

Pige
A. Introduction to a study of global sand seas, by Edwin D. McKee .................... 1
B. Textural parameters of eolian deposits, by Thomas S. Ahlbrandt ..................... 21
C Differentiation of eolian deposits by discriminant analysis, by R. J. Moiola and A. B.
Spencer..................................................................... 53
D. Red color in dune sand, by Theodore R. Walker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
E. Sedimentary structures in dunes, by Edwin D. McKee, with .two sections on the Lagoa
dune field, Brazil, by Joio J. Bigarella .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
F. Dune forms and wind regime, by Steven G. Fryberger, assisted by Gary Dean. . . . . . . . . 137
G. Laboratory studies of sand pattems resulting from current movements, by Theodore F.
Tyler....................................................................... 171
H. Ancient sandstones considered to be eolian, by Edwin D. McKee, with sections by Joio J.
Bigarella .................................................................... 187
I. Sediments of ancient eolian environments - Reservoir inhomogeneity, by Robert Lupe
and Thomas S. Ahlbrandt ..................................................... 241
J. Morphology and distribution of dunes in sand seas observed by remote sensing, by Carol
S. Breed and Teresa Grow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
K. Regional studies of sand seas using Landsat (ERTS) imagery, by Carol S. Breed, Steven C
Fryberger, Sarah Andrews, Camilla McCauley, Frances Lennartz, Dana Gebel, and Kevin
Horstman .................................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Glossary, prepared by Marguerite Glenn .............................................. 399
Selected references . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 409
Index............................................................................. 423

IX

Digitized by Google
Digitized by Google
 A STUDY OF GLOBAL SAND SEAS

INTRODUCTION TO
A STUDY OF GLOBAL SAND SEAS
Chapter A

By EDWIN D. McKEE

Contents
Page Page
Subject matter .. ......... . ........ . . .. . ... .... .. .. . 3 Wind regimes .... .. . ... ........ .. . . ... .. .. . ..... . . 8
Facets of dune study .. ........... .. . ........ . ..... . 3 Types of dunes ....... . .. ... . . . . ... ... . . .......... . 8
Dune classification ............... . .......... . . . 4 Basis for classifying ... . ..... . . . ............... . 8
Sand-texture studies . .. .. . . . . .... .. ............ . 4 Basic (simple) dune types ............. . . . . . ... . . 8
Physical processes involved .. . .. . .. ............ . 4 Compound dune types .... .. ......... . . .. . .. .. . 13
Statistical studies of mean current directions .... . . 5 Complex dune types ... ..... ..... ......... . .. . . 13
Preservation of surface structures, tracks, and trails 5 Varieties of dunes .... .. .... .. ..... . .... .. ... .. . 13
Analysis of cross-strata ........................ . 5 Interpretation of ancient eolian sandstones ......... . . 15
Application of remote sensing (ERTS) . .. .... . ... . 6 Terminology and geographic names ... . ............. . 15
Data from Skylab 4 ....... .... .. . ..... . .. . ..... . 6 Systems of measurement . ...... ....... ... . ....... . . 17
Studies of dune color ......................... . 7 Acknowledgments ..... .. .. .... ... . . .... . .. . ... . . . . 17

Digitized by Google
2 A STUDY OF GLOBAL SAND SEAS

Illustrations
Page
FIGURE 1. Index map of principal deserts ......................................... 7
2. Photograph of star dune, showing slipfaces and arms, Namib Desert, South-
West Africa ........................................ . . . . . . . . . . . . . . . . 9
3 -5. Block diagrams of dune types:
3. Barchan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4. Barchanoid ridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5. Transverse.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
6. Diagram of sequence of dune types with diminishing sand supply and unidirec-
tional wind ........................................................ 12
7 -12. Block diagrams of dune types:
7. Dome.......................................................... 12
8. Blowout ....................................................... 12
9. Parabolic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
10. linear.......................................................... 13
11. Star............................................................ 13
12. Reversing....................................................... 13
13. Drawings showing examples of compound dunes . . . . . . . . . . . . . . . . . . . . . . . . 14
14. Drawings showing examples of complex dunes.......................... 15
15. Sketches of some distinctive dune varieties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Table
Page
TABLE 1. Terminology of basic (simple) dune types............................... 10

Digitized by Google
 INTRODUCTION 3

Subject Matter mosaics. Although inclusion herein of the actual

mosaics proved to be impractical. the maps derived
1Ir HE BIRTH of the idea that led to this publica- from them do show the distribution and abundance
\tV tion on "Global Sand Seas" dates back to the of various dune types and the relations of these
late 1920s. At that time I was engaged in a study of types to certain associated features. such as
the Coconino Sandstone of Arizona's Grand Can- bedrock. water bodies. and juxtaposed dunes. Fur-
yon. Considerable controversy existed then as to thermore. sand roses included with each of these
whether this sandstone was a subaqueous deposit maps enable the user to draw conclusions on the
or was composed of wind-formed dunes. It became probable relations of wind strength and direction to
apparent that definitive literature was sparse or dune type in a particular area.
lacking on types of dunes. global distribution of Regional studies (chapter K) were a team effort.
these types. the mechanics of their development. Analysis of the Landsat (ERTS) mosaics and map-
the precise nature of their internal structure of ping boundaries of individual dune types were by
cross-stratificiation. and the relation of wind Carol Breed. Synthesis of the rather voluminous
systems to these sand forms. Especially lacking literature and preparation of abstracts covering it
were data on criteria that could confidently be used was by Camilla MacCauley. Actual preparation of
in the recognition of ancient dunes. maps was by Franci Lennartz and later by Sarah
The common denominator in this publication is Andrews. The gathering of data on wind. the
eolian sand bodies. Although the book is concerned calculation of wind roses. and the interpretation of
primarily with desert sand seas. the subject matter their relations to sand bodies were by Steven Fry-
is not restricted to deserts; it includes many berger. assisted by Gary Dean.
references to deposits of coastal sand and to sand
bodies in humid climates. Nor does the book deal Facets of Dune Study
exclusively with dunes. which. according to most
definitions. involve mounds or hills. Many DURING THE COURSE of several decades.
references are made to sand sheets. sand stringers. eolian dune studies have been made in ever-in-
and other types of sand deposits that have no prom- creasing numbers and from many different ap-
inent topographic expression. All sand bodies ac- proaches. As might be expected most of the early
cumulated by the action of wind are discussed. studies were largely descriptive. with observations
Chapters A -J of this publication are primarily focused especially on external form and movement
topical. Chapters cover the grain texture. the color. of sand bodies. Attempts to classify dune types ad-
and the structure of modern dunes and other eolian vanced slowly. but basic forms were early recog-
sands. Special treatment is given to the relation of nized in relation to such factors as wind regime and
wind data to dune interpretation. the evolution of physical barriers.
form in current-deposited sand bodies as deter- Principal approaches to dune study in the United
mined from experimental studies. and the discrimi- States. and probably elsewhere to a considerable
nant analysis technique for differentiating between extent. have been as follows: (1) Classification of
coastal and inland desert sands. This topical part of dune forms and sand patterns. first from ground
the publication also includes an analysis of criteria studies. later from aerial photography. and finally
used in ancient deposits to interpret their eolian from remote sensing; (2) textural studies involving
genesis and a consideration of economic application characteristics of grain surfaces and shapes and
of the principles described. including a discussion comparisons with sand textures representing other
of potentials and problems associated with eolian environments; (3) studies of the physical processes
hydrocarbon reservoirs. The final chapters present involved in wind transport and deposition of sand;
a discussion of the morphology and distribution of (4) statistical studies of cross-strata attitudes or dip
dunes as determined largely from Landsat images. directions and determination of resultant vectors;
Chapter K of the publication is devoted to (5) investigations. both in the laboratory and in the
descriptions of major sand seas based largely on field. of environments favorable to the preservation
thematic maps derived from Landsat (ERTS) of surface structures and animal tracks or trails in

Digitized by Google
4 A STUDY OF GLOBAL SAND SEAS

sand; (6) analysis in three dimensions of eolian paring modern sands of dune, marine, and fluvial
cross-strata including statistical studies of basic types, which differ markedly in various textural
types; (7) quantitative morphology studies made features.
from products of remote sensing in which constant The principal attributes that are involved in tex-
scale allows direct comparisons between sand seas; tural considerations are grain size (mean diameter),
and (8) studies, based primarily on the scanning sorting, sphericity, roundness, pitting, and frosting.
electron microscope, of the factors controlling the Numerous scales and techniques for measuring and
red color that is characteristic of many dunes. recording these properties have been developed
through the years, ranging from rough visual esti-
Dune Classification mates to highly sophisticated measurements and
Most geologists involved in the study of sand from simple tabulations to elaborate histograms and
bodies have been faced with the problem of curves.
describing the type or types of dunes in one area Despite all the studies that have been made and
and of assigning a name - either formalized from the voluminous literature on the subject, results
local terminology, adapted from some other sand from the textural approach in recognizing and
sea, or newly designated. Thus, for the most part, defining eolian deposits have remained highly
local terminologies have proliferated, but formal debatable for two inherent reasons. First, the grain
classifications that apply to many regions have properties that can be described are primarily the
been neglected. Lack of means to make direct com- results of transportation, not deposition. They may
parisons, at scale, between dunes of different represent the effects of intensive movement by
regions has been a major deterrent in this field. water, although deposition was by wind, or the
In North America two principal attempts - both reverse. Second, the grains of a particular wind
in the 1940's - were made to classify dune types. deposit may be and commonly are the product of a
"A Tentative Classification of Sand Dunes: Its Ap- second or third generation, reflecting the properties
plication to Dune History in the Southern High inherited from some earlier sediment. Furthermore,
Plains," by Melton (1940), distinguished into some surface features, such as frosting, may not be
classes those dunes on bare surfaces or loose sand primary; rather, they are the result of diagenesis
and those formed in conflict with vegetation. and, therefore, unrelated to the origin of the
"Dunes of the Western Navajo Country," by Hack deposit.
(1941), indicated three basic controls in develop- The application of textural parameters to eolian
ment; wind regime, available sand, and plant sands is discussed in chapter B by Ahlbrandt. Dis-
growth. The present publication describes a general criminant analysis, which is used to differentiate
classification with two variants, one based on beach, river, and eolian sands, is described by
ground observation of dune forms and internal Moiola and Spencer (chapter C).
structures, the other based largely on dune patterns
recognized on images from remote sensing. Physical Processes Involved
Sand physics, focusing especially on the move-
Sand-Texture Studies ment of sand in air, is critical to an understanding
The texture of eolian sand grains has received of the growth and movement of dunes. This field of
the attention of investigators for a very long time investigation, which necessarily involves both ob-
(Udden, 1898; Dake, 1921; Ries and Conant, 1931; servational data from active dunes and the results
Wentworth, 1932). Interest in textures came about of wind-tunnel experiments on sand transport, was
partly because, on theoretical grounds, wind- the subject of a classic study by Bagnold (1941) en-
abraded sand grains were believed to show certain titled "The Physics of Blown Sand." Bagnold's
features that are rare or absent among waterborne treatise clarified many features of the mechanics of
sand grains and partly because laboratory experi- dune development and gave much impetus to the
ments disclosed major differences, especially in the general study of sand seas.
rate of developing certain textures. Further The investigations of Bagnold constituted pri-
knowledge of textures resulted from studies com- marily a study of the behavior of sand grains in a

Digitized by Google
 INTRODUCTION 5

stream of wind and consisted of "precise measure- detail by numerous geologists. These markings con-
ment and experimental verification." The resulting sist of primary sedimentary structures, varieties of
information explained qualitatively many dune contorted bedding, and the tracks and trails of
forms and, with relatively few modifications, has animals. Such studies are especially important in
stood the test of checking since publication. Addi- the recognition of ancient eolian sands because
tional important contributions to our knowledge of they furnish some compelling evidence of genesis.
the interaction between wind and sand have subse- Minor sedimentary structures that are mod-
quently been described by Sharp (1963, 1966). erately common and that have distinctive features
when formed in windblown sand include ripple
Statistical Studies of marks, rain pits, and slump markings. These struc-
Mean Current Directions tures have been studied both in the field and
through laboratory experiments during and after
The value of measuring cross-strata dip direc- the 1940's (McKee 1945; Poole, 1962; Tanner, 1966)
tions in order to determine regional trends in dune and compared with corresponding structures
movement has been recognized since before the formed under aqueous conditions.
start of the 20th century. This approach has been Studies of penecontemporaneous folding (con-
used in studies both of modern dunes and of an- torted bedding) in eolian sands were conducted in
cient sandstones believed to be eolian in origin. A the laboratory (McKee, Douglass, and Rittenhouse,
model for eolian paleocurrents has been described 1971) and later in the field (McKee and Bigarella,
by Selley (1968, p. 103). 1972). A principal result of these studies is a means
Results of directional vector measurements com- of distinguishing in ancient eolian sandstones
monly have been presented in either of two forms: whether they were formed in wet sand, saturated
(1) a circular histogram or rose diagram, as illus- sand, damp crusty sand, or dry sand. Also, the
trated by the early work of Knight (1929) on the studies supplied a means for distinguishing be-
Casper Formation of Wyoming, U.S.A., and by tween sand accumulated in the zone of tension (up-
Shotton (1937) on the Bunter Mottled Sandstone of per slope) and sand in the compressional zone at
England; and (2) a stereonet as introduced by the base of foresets. Other structural features
Reiche (1938) for analysis of the Coconino served to differentiate between the products of
Sandstone in Arizona. different processes, such as avalanching, loading,
Both methods of illustrating vectors serve well to and surface dragging (McKee, Reynolds, and Baker,
illustrate the spread in wind current directions and 1962a, b).
to focus attention on mean or dominant azimuth Investigations of the conditions under which the
directions. The stereonet has the added advantage tracks and trails of various animals were formed
of showing the degree of dip, which is useful in and preserved were made during the 1940's, both in
differentiating between dune slipfaces and wind- the laboratory (McKee, 1944, 1947) and through
ward-side deposits and in indicating the weight to field observations (Peabody, 1940). These studies
be given various readings. Many statistical in- established criteria for determining the degree of
vestigations of this type, in both modem and an- cohesiveness of track-bearing sand at the time
cient eolian cross-strata, have been made within re- prints were made and for explaining selective
cent years; they not only help in determining the orientation of tracks as commonly observed in
direction or directions of sand transport but also, in eolian cross-strata.
some situations, indicate the type of dune involved.

Preservation of Surface Structures, Analysis of Cross-Strata

Tracks, and Trails In the United States the method of using a direct
approach in determing structure in various types of
The environmental conditions necessary for the dunes started with the trenching of a double
preservation of various markings in eolian sand barchan dune near Leupp, Ariz., in the summer of
deposits have been investigated in considerable 1951 (McKee, 1957). The dune selected was about

Digitized by Google
6 A STUDY OF GLOBAL SAND SEAS

270 feet (80 m) high; dips on the windward slope about 570 miles (915 km) by a multispectral scanner
ranged from 5 to 7. but most dips on the lee side (MSS) aboard the satellite. These images cover
were slightly more than 30. The method used was broad bands or swaths around the globe between lat
to wet the dune by pouring water into it near the 80 N. and lat 80 S.
crest. allowing the water to spread outward and A project designated as "The Morphology, Prove-
downward along cross-strata planes. and then to nance, and Movement of Desert Sand Seas" was
section the dune in selected places by digging developed with the acquisition of ERTS imagery
trenches and pits with a shovel. The result recorded that covered 15 major desert areas around the globe
in three dimensions the structure within the dune. (fig. 1). These images, though representing widely
Methods used in the experiment at Leupp. spaced regions, were all at the same scale. thus per-
though primitive. proved so successful that. with mitting direct comparisons of dune patterns and
minor modifications. they have been used in many meaningful descriptions of form and distribution.
places. and by many geologists since. and a con- The principal investigator on this project was E. D.
siderable fund of information on dune structure has McKee and the coinvestigator was Carol Breed.
resulted. During trenching near Sabha. Libya. in Four to six persons assisted in the program at
1961. a pump was available for lifting water to the various stages.
dune crest (McKee and Tibbitts. 1964); this was far Work was done originally using single-band
easier than the earlier method of transporting water black-and-white positive transparencies and paper
in buckets. At White Sands National Monument. prints at a scale of 1:1,000,000; subsequently, false-
New Mexico. U. S. A.. in 1964 -65 trenches were color ERTS images at the same scale became availa-
dug by bulldozer - a tremendous improvement ble and proved to have many advantages. These
over handshovel work as done elsewhere (McKee. false-color images were assembled into mosiacs of
1966). However. in most places. it has been neces- selected sand seas and formed the base for thematic
sary to use whatever equipment was already on maps portraying patterns of dune types and their
hand at the scene. regional distribution. Through use of these maps,
Many important results from the trenching of surface morphologies could be directly compared.
coastal dunes in Brazil have been obtained in recent The study of sand seas from false-color Landsat
years by Bigarella and associates (Bigarella. Becker. (ERTS) imagery has contributed to an interpretation
and Duarte. 1969; Bigarella. 1972. 1975a, b). In these of the depositional environment, to an under-
dunes internal moisture resulting from the humid standing of dune movement, and to the prediction of
climate has made digging possible without first future sand sea development. Advances were made
pouring water. Likewise, at Killpecker dune field in through the description and comparison of surface
Wyoming, U.S.A., the moisture within some dunes morphologies. For a complete understanding of the
was sufficient to permit trenching at various places sand bodies, however, ground-based studies of cli-
(Ahlbrandt, 1973). mate, field studies of structures determined
Following exposure of the cross-strata pattern. a through trenching, and laboratory studies of texture
record normally is made by photography or by were also necessary. The structural and textural
measuring and sketching. The most exact and most studies contributed to an analysis of factors too
permanent method of recording details of structure. small to establish from space; the meteorological
however, has been to make peels with latex or data, especially wind vectors when superimposed
some other material. on images, helped in understanding the complex
relations between dune forms and wind activity.
Application of Remote Sensing (ERlS)
Data from Skylab 4
With the launching on July 27, 1972, of the un-
manned Earth Resources Technology Satellite During the winter and spring of 1974 when
(ERTS-1), later renamed Landsat-l, a new and Skylab 4 was in orbit a considerable number of
important approach to the investigation of sand photographs were taken with a handheld camera,
seas became available. In an l8-day earth-orbiting by request, of desert features in the major areas
cycle, images were obtained from an altitude of being studied on this project. These photographs

Digitized by Google
 INTRODUCTION 7

PACIRC
OCEAN

A1LAN71C INDIAN
OCEAN OCEAN

TROPIC OF
EXPlANATION

Arid and extremely arid

regions (from Meigs. 1951)

~ Areas mapped from Landsat

~ (ERTS-1) images. this paper

PRINCIPAL DESERT AND SEMIDESERT REGIONS and selected Emirates, Oman, and Yemen; 8, Peski Karakumy and Peski
test areas: 1, Western Sahara, Mauritania and Senegal; 2, North- Kyzylkum, U.S.S.R.; 9, Thar Desert, India and Pakistan; 10,
ern Sahara (Grand Erg Occidental and Grand Erg Oriental), Takla Makan, People's Republic of China; 11, Ala Shan Desert,
Algeria; 3, Northern Sahara, libya; 4, Southern Sahara and People's Republic of China; 12, Great Sandy Desert, Australia;
Sahel, Mauritania, Mali, and Niger; 5, Namib Desert, South- 13, Great Victoria Desert, Australia; 14, Simpson Desert,
West Africa; 6, Kalahari Desert, South-West Africa, Botswana, Australia; 15, Algodones Desert and Gran Desierto, U.S.A and
and South Africa; 7, Rub' al Khali, Saudi Arabia, United Arab Mexico. (Fig. 1.)

were taken from distances much closer than those recognized as a function of iron oxide pigments
of the ERTS images and showed more detail in associated with individual grains of sand and in
many places. Also. they had the advantage of beingvarious stages of oxidation, but how and when the
produced in true color. Sites chosen for taking these
iron has been introduced and when the oxidation
photographs were northern Africa. southern has taken place are controversial problems with no
Africa, Saudi Arabia, India, China, U. S. S. R., and
single. simple answer.
Australia. Recent detailed studies of the red coloring in
dune sands were made by T. R. Walker, using a
Studies of Dune Color scanning electron microscope, as well as a
petrographic microscope, for analysis. Chapter D
A distinctive feature of most sand seas is the describes some of his results, Especially significant
brilliant coloring of the sand - largely red with is the conclusion that at least seven separate and in-
various degrees of intensity. Color has long been dependent factors control the degree (variations in

Digitized by Google
8 A STUDY OF GLOBAL SAND SEAS

intensity) of redness in a sand deposit. This illus- have greatly stimulated efforts to better describe
trates the complexity of the problem. and classify dune forms.
Colors as determined from satellite observation
and handheld cameras and from false-color images Basis for Classifying
of Landsat (ERTS) are useful in locating areas of
color difference within a region and in showing the In the investigation of eolian sands, the
global distribution of various shades and intensities classification recommended is based on two
of particular colors. descriptive attributes - the shape or form of the
sand body and the position and number of slipfaces
(steep lee-side surfaces) (fig. 2; table 1). Generally,
Wind Regimes these two attributes constitute recognizable
features both in closeup ground views and at the
BECAUSE EOLIAN SAND accumulation and
great distances represented by ERTS images. Ex-
dune migration are functions of wind regime, a
ceptions are those sand features that are too small
study of available meteorological data for each sand
to be identified on the Landsat (ERTS) images.
sea was considered to be a necessary part of the
Two slightly different terminologies are used in
global program as designed for ERTS imagery. A
the classification system for this publication. One
comparison of wind regimes with dune types was
version (table 1, col. 3) is derived primarily from an
made for various localities around the world and is
analysis of dune structures and so constitutes both
included in the parts of this Professional Paper
a description of the sand dune forms and a partial
dealing with regional analyses.
record of their genesis. The other version (table 1,
A type of sand rose, as described by Fryberger in
col. 6) of the classification is developed largely
chapter F was developed to show from accumu-
from the patterns of Landsat (ERTS) imagery and is
lated wind data the direction and relative amount
used both in the thematic maps prepared from the
of sand drift at a given locality. By use of a "weight-
ERTS mosaics of the principal sand seas and in
ing equation" the relatively much greater ability of
ERTS-based studies of distribution and
winds of high velocity to cause sand drift is taken
morphology. One version of the classification
into account. Resulting sand roses have been
stresses factors that record the growth and develop-
superimposed on the various regional maps in this
ment of individual dune bodies; the other version
publication to allow direct comparison with associ-
emphasizes the gross features of dune patterns
ated dune types as mapped.
where details cannot be observed.
An examination of sand seas from various ap-
Types of Dunes proaches - including studies of Landsat (ERTS)
images, air photos, ground records, and other
A PRINCIPAL AIM of this investigation has been material- suggests that the number of basic or
to construct an objective and descriptive classifica- simple dune forms is fairly limited (table 1).
tion which segregates into types the various dis- However, a large number of combinations of dune
tinctive dune forms and relates these types to types and an almost infinite number of varieties oc-
specific environments of deposition. Two innova- cur. Dune combinations are of two categories: (1)
tions in methods of study contributed substantially compound dunes made up of two or more dunes of
to the construction of this classification. the same type, coalescing or overlapping, and (2)
The first factor in establishing a sound basis for complex dunes in which two different basic types
classifying eolian sand bodies is improved are combined or superimposed.
knowledge of process of construction, obtained
from direct observation of their structure in pits Basic (Simple) Dune Types
and trenches. The second factor is availability of
data from remote sensing programs, especially in The simple dune types, because they are
the form of Landsat (ERTS) imagery, in which a classified partly on overall form and partly on num-
uniform scale permits direct comparisons between ber of slipfaces, generally reflect factors of the
various parts of the world. These two approaches depositional environments in which they devel-

Digitized by Google
 STAR DUNES, showing three curving arms with steep slipfaces oriented in various direc-
tions. Tsondab Vlei, northeastern part of Namib Desert, South-West Africa. Photograph by
E. Tad Nichols. (fig. 2.)

oped. especially the wind (strength and direction). The genetic relationship between the three dune
the sand supply. vegetation. physical barriers (rock types previously referred to seems to be reasonably
outcrops and water bodies). and distance from well established through the observations of many
source. Various combinations of these factors are geologists and has been described in numerous
responsible for each distinctive dune type and per- papers. These dunes are all characterized by slip-
mit some generalizations concerning their genesis. faces in one direction and therefore represent
unidirectional wind movement. The three types
Probably the best known and commonest basic commonly occur in a definite sequence (fig. 6)
dune forms result from winds having a single domi- downwind as a result of diminishing sand supply.
nant direction and are oriented with their axes at In general. the type represented in a particular area
right angles to wind direction. Such dunes range seems to be a function of the amount of available
from the small crescent-shaped type commonly sand. Barchan dunes clearly result from sparse sand
known as barchan (fig. 3). through parallel rows of and in many areas are scattered over bare rock sur-
coalesced barchans. here referred to as barchanoid faces. Transverse dunes. the other extreme.
ridges (fig. 4). to essentially parallel straight ridges develop where sand is abundant. and they attain
known as transverse dunes (fig. 5). All three types. maximum height where there is a balance between
which may occur in gradational sequence. are strength of wind and available sand.
referred to as "crescentic" on the regional sand- Because most barchanoid-type dunes-
seas maps presented in this publication. Most barchans. barchanoid ridges. and transverse
barchan types are too small to be recognized on dunes - form patterns of parallel wavy lines on
Landsat (ERTS) images. and at that scale the ERTS images. they were referred to collectively as
transverse dunes are difficult to distinguish from "parallel wavy dunes" in the classification of dune
some of the smaller barchanoid ridges and from patterns used during ERTS studies (McKee and
many linear dunes. Breed. 1974a. b) .

Digitized by Google
10 A STUDY OF GLOBAL SAND SEAS

TABLE 1. - Terminology of basic dune types and other eolian deposits

(Examples may show simple or compound dunes; however. they are designated by bale dune type only)

N.me uaed in ground Block dl.gr.m N.me UMd In apece-imlg8ry .nd

Form Number of IltUdy of form, Ilipfece, ~ure numberi .lr'1)hoto 1ItU~ of .-uem .nd
lIipfacel .nd Internel Itructure Chapter A) lTlOfPhoIogy Ch.pt.... J, K)

Sheetlike None, flat

with broad, None Sheet Sheet'
flat surface surface

Thin, elongate None, flat

None Stringer Streak'
strip surface

Circular or For
elliptical mound None 2 Dome Dome-shaped (3)
see
Barchan For Barchan (3)
see

Row of connected
crescents in Barchenoid ridge For detail
see fig. 4
plan view .~

u
i... Berchanoid ridge

Asymmetrical For detail

Transverse ridge
ridge see fig. 5

Circular rim 1 or more For detail (3)

of depression Blowout 4 see fig. B Not recognized

"U" shape in For detail

1 or more Parabolic 4 Parabolic
plan view see fig. 9

Symmetrical For detail

2 Linear (self) Linear
ridge 98 fig. 10

Asymmetrical For detail

ridge 2 Reversing Reversing
see fig. 12

Central peak For detail

with 3 or 3 or more Star Star
more arms see fig. 11

lM.y include moundl .nd other fHtur_ too am.1I to .ppear on L.ndlst imagery.
2lntern.1 Itruc:tur_ may lhow embryo berch.n-type with one IIlpfKe.
38eceu.. of their am.n lize, dome dun_, blowout dun. .nd indlvidu.1 b.rch.n dun. .re difficult to dlltlnguleh on L.ndIIt Imagery.
4[)un. controlled by vegetation.

Digitized by Google
 INTRODUCTION 11

Some sand bodies display no slipfaces. These

bodies are commonly referred to as sheets and
stringers, using the geologic terminology usually
applied to bodies of rock or sediment having such
forms. These sand bodies are not considered to be
dunes, because the term "dune," according to most
definitions, refers to a mound or ridge. A stringer
may be considered to be a variety of sand sheet
having a stringlike shape, and, commonly, though
not exclusively, it develops along the margins or at
the downwind extremity of a sheet. The term
"streak" is used instead of "stringer" on thematic
maps in this publication. Streaks identified on the
basis of Landsat (ERTS) mosaics may, therefore, in-
BARCHAN DUNES. Arrow shows prevailing wind direction. clude barchans and other small dunes which are
(Fig. 3.) not visible on the image.
Dome dunes (fig. 7), which probably consist of
more than one type (as discussed in chapter E on
dune structure), normally show no external slip-
faces and are circular or elliptical in plan view.
However, some dunes of this type internally dis-
close slipfaces dipping in one direction, suggesting
that they began development as barchans, and im-
plying that the dome form is the result of strong
winds that truncated the top and flattened the lee
slopes of a barchan. Many dome dunes of coastal
areas probably have their shapes controlled more
by moisture and vegetation than by strength of
wind, as described later.
A class of dunes in which the forms apparently
are controlled more by partial stabilization from
BARCHANOID RIDGE. Arrow shows prevailing wind direction.
(Fig. 4.) vegetation or moisture or both, than by wind
strength and direction, is represented by the
blowout (fig. 8) and the parabolic dunes (fig. 9).
Such dunes may have slipfaces sloping in one or
many directions depending on what parts of their
rims are free to migrate. They are, to a considerable
extent, features of deflation by wind, but deposition
occurs wherever margins are free to migrate as on
other dunes. The blowout type is normally a circu-
lar bowl, whereas the parabolic develops aU-shape
or V-shape from arms that are anchored and nose
that migrates downwind. A coastal dune of ridge
form which reflects control by vegetation is refer-
red to by some authors as a retention ridge; it may
have small blowouts developed on it.
TRANSVERSE DUNE. Arrow shows prevailing wind direction. Perhaps the most controversial type of dune is
(Fig. 5.) the linear dune (fig. 10) commonly known as the

Digitized by Google
12 A STUDY OF GLOBAL SAND SEAS

ii'
\
.. ....
,.,

--"t .... ~ .'. . ....... .

-.~.
r...,,-..-.: "'1'" '\..
~.''''

... ~;..

., ......

SEQUENCE OF DUNE TYPES with unidirectional wind and (from left to right) diminishing sand supply; transverse.
barchanoid ridge, and barchan. Arrow shows prevailing wind direction. (Fig. 6.)

seif dune in Africa and Saudi Arabia and as the from nearly opposite directions are balanced with
longitudinal dune in many other regions. Whatever respect to strength and duration. Such dunes may
the mechanics of its genesis and number of wind have the general form of transverse ridges, but a
directions involved, it is, in general, a straight ridge
with slipfaces on both sides; groups of these ridges,
as seen in ERTS images or air photos, appear as
parallel straight lines.
Dunes of multiple slipfaces are the result of
winds from several directions; they commonly
have a high central peak and three or more arms ex-
tending radially. Such forms are called star dunes
(fig. 11) and have an infinite variety of shapes.
They mostly grow vertically, rather than migrating
laterally.
Intermediate in character between the star dune
and the transverse ridge is the reversing dune (fig.
12). It characteristically forms where two winds BLOWOUT DUNES. (Fig. 8.)

PARABOLIC DUNES. Arrow shows prevailing wind direction.

DOME DUNES. (Fig. 7.) (Fig. 9.)

Digitized by Google
 INTRODUCTION 13

second slipface. oriented in a direction nearly op- Compound Dune Types

posite that of the primary slipface. develops
periodically. In addition to the basic or simple dune forms just
described. compound forms are abundant in most
sand seas. These consist of two or more of the same
type combined by overlapping or by being superim-
posed (fig. 13) on one another. Examples of com-
pound dunes are (1) barchanoid ridges coalescing.
(2) star dunes coalescing to form an intricate pat-
tern of arms and peaks. (3) little barchan dunes on
the windward flanks of large ones. (4) small
parabolic dunes between the arms of a larger one.
and (5) major ridges covered by many smaller
linear dunes.

LINEAR DUNES. Arrows show probable dominant winds. Complex Dune Types
(Fig. 10.)
Complex dunes in which two different basic
types have coalesced or grown together are repre-
sented in most sand seas by various combinations
(fig. 14). These combinations include linear dunes
in parallel rows with star dunes on their crests;
these are called "chains of stars" and are well
developed in several major sand seas. Complex
dunes also are illustrated by small barchan dunes in
the corridors of linear dunes. by blowout dunes on
transverse dunes. and by large star dunes with
superimposed barchanoid dunes.

Varieties of Dunes
For all dune types - simple. compound. and
STAR DUNES. Arrows show effective wind directions. (Fig. 11.} complex - an almost infinite number of varieties
occurs. The varieties are too numerous to describe
or even classify by name herein. for they probably
represent transitions from one basic type to another
and result from differences in the direction and
strength of wind. the amount of available sand. the
physical obstructions. and other factors that control
dune types. Some varieties that are especially con-
spicuous and have been much discussed in the
literature are barchans with one hom greatly ex-
tended. linear dunes with many branches diverging
from one end suggesting the name feather dunes.
and parabolic dunes that are V-shaped rather than
REVERSING DUNES. Arrows show wind directions. (Fig. 12.) U-shaped (fig. 15).

Digitized by Google .
14 A STUDY OF GLOBAL SAND SEAS

.: ..
,;."
":'
p- , ...

COMPOUND DUNES: A, Barchanoid ridges coalescing. B, Star

dunes coalescing. C, Small barchans on large barchan. D,
Parabolic dunes within a large parabolic dune. E, linear dunes
c on large linear ridge. (Fig. 13.)

Digitized by Google
 INTRODUCTION 15

Interpretation of Ancient
Eolian Sandstones
A MAJOR EFFORT of this investigation was to
establish criteria for the recognition of ancient
eolian-type sandstones (chapter H). Recognition of
wind-deposited sand in ancient sandstones was
recorded nearly 100 years ago by Walther (1888, p.
253) when he stated that a detailed study of the Nu-
bian Sandstone in Sinai and adjacent parts of the
Arabian Desert "has led me to the conviction that
these sandstone deposits are an eolian dune
development * * * and that North Africa contained
deserts of Paleozoic and Mesozoic age."
Subsequently, the eolian origin of certain sand-
stones in the Southwestern United States was
postulated by Huntington (190'7, p. 380) because of
textural and structural features "frequently seen
among modem sand dunes of Persia, Transcaspia,
and Chinese Turkestan." Other early references to . :. : . . . . .~
sandstones believed to have been wind-deposited
are Fourtau (1902), Sherzer and Grabau (1909),
Twenhofel (1926, p. 174, 557), and Cuvillier (1930).
In general, however, compelling evidence for the
recognition of eolian sandstones had to await the
fairly sophisticated analyses of cross-strata charac-
teristics and direction vectors, studies of minor
structures and their preservation, and other ap-
proaches that were introduced about 1937 and later,
as described in chapter E on modern dune struc-
tures.

Terminology and Geographic

Names
TERMS USED in this professional paper are
defined in the glossary. Most definitions are from
"A Desert Glossary," by R. O. Stone (1967), and
from "Glossary of Geology," published by the
American Geological Institute (1972), although
various other glossaries were also consulted. A few
terms introduced for the first time in this paper are

also defined. c
The names of geographic localities outside the
United States are according to the gazetteer for each COMPLEX DUNES: A, Star dunes on linear dune. B, linear dunes
of the countries concerned. The spellings have been with barchans in interdunes. C, Blowouts on transverse dune.
approved by the U. S. Board on Geographic Names (Fig. 14.)

Digitized by Google
16 A STUDY OF GLOBAL SAND SEAS

BARCHAN

LINEAR

SOME DISTINCTIVE VARIETIES of basic dune types. (Fig. 15.)

Digitized by Google
 INTRODUCTION 17

and the U. S. Department of Defense, Defense Map-

ping Agency. International boundaries and geo-
Acknowledgments
graphic names and their spellings do not A VERY LARGE NUMBER of organizations and
necessarily reflect recognition of political status of individuals have contributed ideas, facts, and
an area. materials, as well as time and effort, to assist in the
preparation of this treatise. To all of these we ex-
Systems of Measurement tend our gratitude. Contributions that pertain to
particular chapters are, for the most part,
THIS TREATMENT OF SAND SEAS on a global acknowledged within the appropriate chapters;
scale includes measurements of many kinds and contributions that affect numerous chapters or the
from many sources. Resulting figures should entire Professional Paper are acknowledged here.
preserve as nearly as possible the original measure-
ments as derived, but at the same time should be Geologic and Geographic Data
readily visualized by people of various backgrounds
and training. Thus, where an original measurement For assistance in interpreting local features of
was made in terms of the metric system, it is ex- various sand seas:
pressed in that system, but for the convenience of
some readers, its converted equivalent in the Dr. G. Lennis Berlin, Northern Arizona Univer-
English system follows in parentheses. Likewise, sity, Flagstaff, Arizona, U. S. A.
where an object was measured in inches, feet, or Dr. Arthur L. Bowsher, Arabian American Oil
miles, that measurement as recorded is presented in Co., Dhahran, Saudi Arabia.
the text, succeeded by its metric equivalent in Dr. I. H. Hamrabaev, Institute of Geology and
parentheses. Geophysics, Tashkent, U. S. S. R.
The following English-metric conversion scales Mr. Donald Holm, formerly with Arabian
are given for the convenience of the reader. They American Oil Co., Saudi Arabia.
are for visual comparison only and do not represent Mr. Brad Musick, Office of Arid Lands Studies,
actual units of measurement. Abbreviations shown Tucson, Arizona, U. S. A.
in parentheses are those used in this publication. Dr. Shumil E. Roy, Indian Agricultural
Research Institute, New Delhi, India.
0 0.5 2 3 Inches (in) Dr. M. K. Seely, Desert Ecological Research In-
I I I I
I
I I
I
I I
I
stitute' Walvis Bay, South-West Africa.
o 0.5 1 2 3 4 5 6 Centimetres (em)
Remote Sensing Program
0 5 10 Feet (ft)
, I , I For guidance, support, and assistance in those
I I I I
0 1 2 3 Metres (m) phases of sand sea study based on remote sensing
programs, we are indebeted to the following people
0 5 10 20 30 40 Miles in the United States:
I (mil
I " "I I I
I
I I
I
I
I
National Aeronautics and Space Administra-
0 5 10 20 30 40 Kilometres (km)
tion (NASA):
Comparative English and metric scales of distance. G. Richard Stonesifer, Technical Monitor.
Paul D. Lowman, Scientific Monitor.
Most of the conversions from one system to the EROS Program, Office of U.S. Geological
other were made by Patricia Melrose of the U.S. Survey;
Geological Survey, using a conversion calculator. J.M. DeNoyer, Director.
Except where a high degree of precision was availa- Richard S. Williams, Jr., Geologist.
ble and required by a particular situation, W. Douglas Carter, Geologist.
references throughout the text were converted only Priscilla Woli. EROS Technical Publica-
approximately. tions Office.

Digitized by Google
18 A STUDY OF GLOBAL SAND SEAS

Landsat (ERTS) Office of U.S. Geological Volunteer Assistance for Work on

Survey: Landsat (ERT5-1) Program
W. R. Hemphill, Chief, Resources Work-
ing Group. Many phases of the remote sensing program for
Lawrence C. Rowan, Staff Geologist and Sand Seas Analysis, conducted at Flagstaff,
Coordinator for Remote Sensing. Arizona, U. S. A., were made possible by the volun-
Allan N. Kover, Staff Geologist, Remote teer services of geology students, mostly from the
Sensing. Northern Arizona University. The following per-
Robert D. Regan, Staff Geologist, Remote sons are given special thanks for their contributions
Sensing. to this program:
Skylab, NASA: Marion Blancett: photographic work.
V. R. Wilmarth, Program Scientist. Dana Gebel: Cataloguing imagery acquisitions,
John Kaltenbach, Project Scientist. work on chart of dune terminology, rough
Michael McEwen, Geologist. draft maps.
Kevin Horstman: Measuring dune and inter-
Wind Record and Other Climate Data dune areas with density slicer.
Katherine Nation: Preparing mosaics from
The following persons and institutions assisted in ERTS imagery.
the gathering and interpreting of meteorologic Iris Sanchez: Vegetation studies, dune
records used in compiling wind roses; measurement studies.
Arabian American Oil Co., Dhahran, Saudi
Arabia - Mr. J. P. Mandaville. Photographic and Other Illustrative
Central Arid Zone Research Institute, India-
Dr. H. S. Mann. Materials
Desert Ecological Research Institute, Walvis Photographs for use in illustrating sand bodies of
Bay, South-West Africa - Dr. M. K. Seely. various types and localities were kindly furnished
Indian Meteorological Department, Poona, In- by the following in the United States:
dia- Dr. K. Krishnamurthy. Dr. Stanley Froden, NASA, Ames Research
Meteorological Department, Ministry of Com- Laboratory, Iowa.
munications, Tripoli, Libyan Arab Repub- George Morrison, National Park Service, White
lic- Mr. M. M. Sasy. Sands National Monument, New Mexico.
National Climatic Center, Ashville, N. C., E. Tad Nichols, Tucson, Arizona.
U.S.A. - Dr. Harold L. Crutcher. J. Philip Schafer, Flagstaff, Arizona.
Joseph F. Schreiber, University of Arizona,
Manuscript Review Tucson, Arizona.
For helpful suggestions and criticism, after their John Shelton, La Jolla, California.
technical review of drafts for various chapters we H. T. U. Smith (now deceased), Department of
thank the following: Geology, University of Massachusetts,
John Dingler, University of California at San Amherst, Massachusetts.
Diego, California U. S. A.
Gerald Friedman, Rensselaer Polytechnic In- Editing and Preparation of Text and
stitute, Troy, New York, U. S. A. Illustrations
Dr. A. S. Goudie, Oxford University, England.
Dr. R. L. Maby and colleagues, Arabian Because of the considerable size and numerous
American Oil Co., Dhahran, Saudi Arabia. ramifications involved in this publication, far more
Dr. John Rogers, University of Cape Town, than normal time and effort was required by some
Rondebosch, South Africa. people of the U.S. Geological Survey to bring the
Dr. D. B. Thompson, Manchester University, text and illustrations to their final form. As editors,
England. Patricia Melrose, and Shirley Person did much

Digitized by Google
 INTRODUCTION 19
val uable work on preliminary drafts of the greatly facilitated by the help of the following
manuscript. For work on the completed text during librarians and reference-service personnel here in
early stages, much credit is due Priscilla Patton; for the United States:
later painstaking review and for seeing the Katharine Bartlett, Museum of Northern
manuscript to completion, credit is due Diane Arizona library, Flagstaff, Arizona.
Schnabel; and an excellent job of illustration edit- Elizabeth Behrendt and Helen King, U. S.
ing was done by Sigrid Asher-Bolinder. Final edit- Geological Survey, Denver, Colorado.
ing and preparation for printing was by Carol Hurr; Mary Haymes and Pamela Noaecker, National
graphics are by Arthur Isom. Center for Atmospheric Research, Boulder,
Colorado.
Library and Literature Review Research James Nation, U.S. Geological Survey,
Flagstaff, Arizona.
The extensive program of literature review that Patricia Paylore, University of Arizona library,
was required in support of the ERTS program was Tucson, Arizona.

Digitized by Google
Digitized by Google
 A STUDY OF GLOBAL SAND SEAS

TEXTURAL PARAMETERS OF EOLIAN DEPOSITS

Chapter B

By THOMAS S. AHLBRANDT

Contents
Pille
Summary of conclusions ........................... . 23
Introduction ..................................... . 23
Acknowledgments ................................ . 24
Textural parameters of coastal dune, inland dune, and in-
terdune and serir deposits ....................... . 24
Recognition of depositional environments ........... . 27
Porosity and permeability relationship to textural
parameters ..................................... . 32

21

Digitized by Google
22 A STUDY OF GLOBAL SAND SEAS

Illustrations
FIGURE 16. Map showing general locations of the 506 eolian sand samples studied.....
...
25
17. Photograph of dune and interdune deposits, Namib Desert, South-West Africa 27
18. Histograms of graphical measures for 191 eolian sand samples. . . . . . . . . . . . . 28
19 - 21. Plots of graphical measures of:
19. 291 coastal dune sand samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
20. 175 inland dune sand samples ................................. 30
21. 40 interdune and serir sand samples ............................ 30
22. Histograms showing textural contrasts among dune and interdune sand .... 31
23. Photograph showing large, complex linear dunes which derive sand from
coastal deposits, Namib Desert, South-West Africa ..................... 32
24. Diagram showing relationship between initial porosity and sorting ......... 32
25. Diagram showing the interrelationship of mean grain size and sorting on initial
permeability for wet, packed sand .................................... 33
26. Map and columnar sections of cores and plug samples taken for porosity and
permeability determinations, White Sands National Monument, New Mex-
ico, U.S.A. ......................................................... 34
27. Diagram showing porosity-permeability distribution of dune and interdune
sediments, White Sands National Monument, New Mexico, U.S.A. ....... 35

Tables Page
TABLE 2. Comparison of textural parameters for selected dune fields . . . . . . . . . . . . . . . . 26
3. Porosity and permeability measurements in alluvial, littoral, and coastal dune
sand.............................................................. 33
4. Porosity and permeability measurements in eolian subenvironments: Coastal
dune, inland dune, and interdune sand ............................... 33
5. Graphical measures of sand samples from coastal dunes used as a data base 36
6. Graphical measures of sand samples from inland dunes used as a data base. 44
7. Graphical measures of sand samples from interdune and serir deposits used as
a data base ........................................................ 50

Digitized by Google
 TEXTURAL PARAMETERS OF EOLIAN DEPOSITS 23

Summary of Conclusions tural parameters, such as excellent sorting, positive

skewness, mean grain size, or rounded and frosted
grains uniquely define eolian deposits. In contrast,
f?r SUITE OF 506 eolian sand samples was
.a. divided into 291 coastal dune samples, 175 in- Udden (1914), Shepard and Young (1961), Schlee,
Uchupi, and Trumbull (1964), Moiola and Weiser
land dune samples, and 40 interdune and serir sam-
(1968), Bigarella (1972), Ahlbrandt (1974b), and
ples. Textural parameters for each sample are
others consider these parameters to be considerably
listed. These three subenvironments of eolian
less diagnostic of eolian deposits.
deposits have different textural parameters: (1)
In this chapter and in chapter C, particle-size
well-sorted to very well sorted fine coastal dune
diameter is measured in the phi (4)) grade scale
sand. (2) moderately sorted to well-sorted fine to
followed by the equivalent in millimetres (mm)
medium inland dune sand, and (3) poorly sorted in-
(Wentworth grade scale). The progression is as
terdune and serir sand. The dune-sand samples, in-
follows:
cluding those from coastal and inland dunes, range
in mean grain size from -0.684> (1.6 millimetres) to ... -2 f/J =4 mm
3.4 4> (0.1 millimetres). Most interdune and serir -1 f/J =2 mm
samples are bimodal in the sand fraction and also o f/J =1 mm
have higher silt and clay contents compared with 1 f/J =0.5 mm
adjacent dune-sand samples. Textural contrasts oc- 2 f/J =0.25 mm
cur among samples from different positions on a 3 f/J =0.125 mm
4 f/J =0.0625 mm ...
given dune type, and among samples from dunes
formed under different wind regimes.
In the absence of sedimentary structures as prim- Grains of sand size -1.0 to 4.04> (2.0-0.062 mm)
ary evidence of eolian deposits, textural criteria compose most sand sea deposits (Cooke and War-
become important. Alluvial, littoral, and eolian ren, 1973, p. 255 -267); thus, the sand fraction of
sands differ in their porosity and permeability, eolian deposits is discussed in this chapter. Eolian
which are functions of mean grain size, sorting, and silts (loess), eolian clays, clay coatings on grains
packing arrangements. Impermeable or low per- (Walker, chapter D, this paper) and sand size ag-
meability deposits, such as interdune or fluvial gregates of silt and clay are not discussed.
deposits adjacent to more permeable deposits, par- Deficiencies of certain sizes of clastic material
ticularly well-sorted coastal dune sand, may form a are considered to be eolian indicators by Udden
barrier to fluid migration -oil or water. Therefore, (1914) and Kuenen (1960). Deficiencies of eolian
differentiation of depositional environments in material larger than 0.04>(1 mm) and smaller than
eolianites can enhance fluid recovery operations. 3.3 4> (0.1 mm) were noted by Kuenen (1960), and
Skewness and kurtosis values for eolian deposits deficiencies of clastics smaller than 4.34> (0.05 mm)
are highly variable and were not found to be diag- were observed by Bagnold (1941). In general. defi-
nostic. Sophisticated approaches, such as discrimi- ciencies of clastic particles in the very coarse sand
nant analysis or moment measures, rather than to granule range, 0.0 to -2.04> (1-4 mm), and the
graphical measures of textural parameters may be a very fine sand to silt range, 3.0 to 4.04> (0.125 -0.062
means to differentiate alluvial, littoral, and eolian mm), have been observed in a number of deposi-
sands on a textural basis as discussed by Moiola and tional environments (Shea, 1974). On the basis of
Spencer (chapter C, this paper). analyses of 11,212 samples, Shea (1974) concluded
that these clastic deficiencies do not exist.
Although clastic particles larger than very coarse
Introduction sand, < -0.68 4> (1.6 mm), and smaller than very fine
sand, >3.44>(0.1 mm). are absent in the dune-sand
THE DEGREE to which textural parameters may samples used in this study, the interdune samples
be used to identify eolian deposits is highly con- from the sample suite contain material that ranges
troversial. Folk and Ward (1957), Mason and Folk in size from granule to clay. Because of the
(1958), Friedman (1961), Folk (1962), Mabesoone problems with the deficiencies concept discussed
(1963), Visher (1971), and others believe that tex- by Shea (1974) and the variety of material that oc-

Digitized by Google
24 A STUDY OF GLOBAL SAND SEAS

curs in the interdune and serir samples, this chapter sum (White Sands National Monument, New Mex-
will focus on the clastic material in the samples. ico), olivene (Hawaii), quartzose and feldspathic
Several other types of textural studies will be sands.
briefly discussed. Upturned plates of quartz The textural parameters, derived from quarter
(chevrons) formed during the frosting process can phi sieve data. of mean grain size (Mz) , inclusive
be detected with the scanning electron microscope graphic standard deviation (u,) , inclusive graphic
and, thus, is a potential indicator of eolian pro- skewness (Sk,), and transgraphic kurtosis (K'd as
cesses (Kuenen, 1960; Margolis and Krinsley, 1971; defined by Folk (l968b) will be used in this chapter.
Rogers and Tankard, 1974). In indurated rocks,
defrosting by the diagenetic effects of pressure Acknowledgments
solution and grain alteration reduce the reliability
of this indicator (Kuenen and Perdok, 1962; SAMPLES WERE SIEVED on quarter phi inter-
Margolis and Krinsley, 1971). Despite these vals by R. J. Moiola and A. B. Spencer, both of
problems, Le Ribault (1974) suggested that original Mobil Research and Development Corp., Dallas,
surficial textures may persist in the geologic record. Texas, and their data are at the end of this chapter.
Differential settling velocities of light and heavy Edwin D. McKee of the U.S. Geological Survey pro-
minerals have been used to document eolianites vided 191 sand samples for this study. J. J. Bigarella
and their traction, saltation, and suspension compo- of the Geological Institute, Federal University of
nents (Friedman, 1961; Hand, 1967; Steidtmann, Parana, Brazil, and L. R. Martins of the Ceco-In-
1974). Provided that all grain sizes are available and stituto de Geaciencias-UFRGS, Porto Alegre, Brazil,
stable heavy minerals are used, this technique provided the data for 315 samples from Brazilian
should be diagnostic because the relative sizes of coastal dunes. Textural data from the literature
hydraulically equivalent light and heavy grains were obtained by Camilla McCauley of the U.S.
deposited in air are different from those deposited Geological Survey. I thank them all.
in water.
The purpose of this discussion is to present and
synthesize data for a broad-based collection of Textural Parameters of Costal
eolian deposits throughout the world and the
graphical measures of their textural parameters, Dune; Inland Dune, and
and to provide impetus for future textural studies Interdune and Serir Deposits
of eolian deposits. Data used in this chapter are not
the result of systematic worldwide sampling net- A COMPARISON of published information of
works to prove or disprove a particular hypothesis, mean grain size and sorting values for selected
such as I. G. Wilson's (1973) hypothesis of gran- dune sands (as opposed to interdune) from five
ulometric control of eolian bedforms. There are few continents indicates that the dune sands are
systematic textural studies of entire dune fields; generally in the fine sand range (0.125-0.25 mm)
however. the overall large and varied nature of the and are moderately well sorted (u, values from
population studied in this report (506 samples) may 0.50 -0.71, Trask values from 1.4 -2.0). (See table
lead to significant conclusions. This population 2.) These data represent a variable number of sam-
comprises 191 sand samples primarily from inland ples, a multitude of sampling and laboratory tech-
dune fields and 315 samples mostly from coastal niques, many different dune types. and different
dunes in Brazil (fig. 16). The forms sampled include locations of samples on the dunes. Because of the
star. seif. dome, transverse. barchan, parabolic, general confusion in the literature regarding tex-
hanging (falling), climbing, produne. and coastal tures of eolian deposits, generalizations based upon
dunes. Moreover, the samples came from various textures as discussed in the literature seem inap-
positions on the dunes, such as the crest, windward propriate. The textural parameters for the 506 sand
and leeward or slipface. from ripples. sand ridges, samples used in this chapter and by Moiola and
interdune and serir deposits. The composition of Spencer (chapter C, this paper) were all derived
the samples includes carbonate sands (Bahama Is- from quarter phi sieve data, thus allowing textural
lands) and clastic sands; the latter including gyp- comparisons.

Digitized by Google
 TEXTURAL PARAMETERS OF EOLIAN DEPOSITS 25

1~r---------~p,~~~~~T-------~--t-~r-------~------~~~--------~1~

60"

ASIA..iJ '0.
f?"

40"

..
PACIFC

",

0"
INDIAN
OCPAN OCPAN

OCPAN ..
40"

. ~.

GENERAL LOCATIONS of the 506 eolian sand samples studied. (fig. 16.)

The majority of the 191 samples collected by Ed- these categories. The statistical differentiation,
win D. McKee throughout the world represent in- using discriminant analysis, of coastal dune sand
land dune and interdune environments (fig. 17 and samples from inland dune samples, and of inland
frontispiece), and demonstrate the maximum range dune sand samples from interdune samples is pre-
of textural parameters for eolian deposits. The sented by Moiola and Spencer (chapter C, this
dominant mean grain sizes for these samples are in paper). The textural contrasts among coastal, in-
the medium and fine sand ranges (fig. 18). Sorting land, and interdune samples are illustrated by plots
values show a wide range with a considerable por- of their textural parameters.
tion occurring in the poor sorting category (fig. 18). Comparison of 291 coastal dune sand samples
The histogram of the skewness values shows that with 175 inland dune sand samples shows several
the majority of the samples are positively skewed, differences. The coastal dunes are composed almost
although at least 20 percent of the samples are totally of very well sorted fine sand (fig. 19).
negatively skewed (fig. 18). The kurtosis histogram Furthermore, they are symmetrical to positively
shows considerable scatter (fig. 18). (fine) skewed and are platykurtic to mesokurtic
Data for another 315 samples, mostly from (fig. 19). Inland dune sand samples, by contrast,
Brazilian coastal dunes, were combined with show a much greater range in mean grain size and
McKee's data and then divided into three major sorting values (fig. 20). Skewness values show wide
categories (where n is number of samples): (1) scatter, from positive to negative, and the kurtosis
coastal dune sand (n =291), (2) inland dune sand values tend to be mostly platykurtic (fig. 20).
(n=175), and (3) interdune and serir sand (n=40). The most significant textural differences occur
The data at the end of this chapter (tables 5 -7) are between dune sands and interdune and serir sands.
presented according to these categories and are The mean grain size of 40 interdune and serir sand
grouped alphabetically by geographic area within samples is highly variable, due to their poor sorting

Digitized by Google
26 A STUDY OF GLOBAL SAND SEAS
TABLE 2. - Q)mpari80n of textural parameters of SOlId from selected dune fields and sand 880S
(Leaden (.) Indicate no datil)
Selected_ Mean gnin liR Sorting Reference
(millimetrel)

Grand Ers Occidental. Grand Ers 1Q.10- 0.30 WilBOn (1973. p. 104).
Oriental. Algeria.
Mauritania. Mali ................ 1Q.25- 0.40 Tricart and Brochu (1955. p.
161).
IJbya. Niger. Chad. . . . . . . . . . . . . . . 0.20 Warren (19'12. p. 42) .
.... .. do....................... 0.34 0.55 Hoque (1975. p. 395).
Kalahari Deaert. South Africa ..... 0.21 Goudie. Allchin. and Hedse
(1973. p. 248).
Namib Desert. South-West Africa .. 0.23 Do.

Ala Shan Desert. China. . . . . . . . . . . 40.19 Yu and others (1962. p. 276).

Talda Makan Desert. China ....... 40.17 Chu and others (1962. p. 23).
Thar Desert. India ............... 1Q.10- 0.15 Goudie. Allchin. and Hedse
(1973. p. 248).
Rub' al Khali .. .. .. .. .. .. .. .. .. .. 0.25 0.66 Ahlbrandt (this chapter).
Saudi Arabia (Oman) ............ 0.19 1Q.22 - 0.54 Glennie (19'10. p. 62).
Peski Kyzylkum. U.S.S.R. . . . . . . . . . 1Q.13- 0.25 Yakubov and Bespalova (1961.
p.654).
NOItb "-ric:a ad South ADmica
Algodones dunes. Califomia. 0.28 11.22-3.47 Beheiry (1967. p. 29).
U.S.A. 0.23 1.13 Norris (1966. p. 296).
1Q.09 - 0.30 Van de Kamp (1973. p. 848).
Great Sand Dunes National 10.73 (reversing dunesl)
Monument. Colorado. U.S.A. s 0.24-1.51 1Q.15-1.25 Ahlbrandt (this chapter).
Killpecker dunes. Wyoming. 1Q.13- 0.43 (transverse
U.S.A. dunese) . . . . . . . . . . . . 1Q.16- 0.93 Ahlbrandt (19'14b. p. 43).
Navajo Indian Reservation. 40.20 10.52 Ahlbrandt (this chapter).
Arizona. U.S.A.
30.15-1.10 3 0.35~.96 Do.
White Sands National Monument. 0.25- 0.55 (transverse
New Mexico. US.A. dunesl) . . . . . . . . . . . . 0.55 McKee and Moiola (1975. p.
61); Ahlbrandt (this
chapter).
Gran Desierto. Mexico ........... 1Q.13- 0.25 1.27 Merriam (1969. p. 531).
Coastal dunes. Brazil. . . . . . . . . . . . . 1Q.15- 0.24 1Q.25 - 0.49 Bigarella. Alessi. Becker. and
Duarte (1969. p. 75-76).

Great Sandy Desert ............ . 0.28 (sheet deposits') 0.53 Bettenay (1962. p. 15).
0.53 (parabolic dunese) Do.
Great Victoria Desert .......... . 0.50 (crestse) 1.40 Wopfner and Twidale (1967.
1Q.26- 0.30 p.132).
Simpson Desert ............... . 0.16 0.43 (crest..) Folk (19'11a. p. 27).
1Q.14~.19 0.57 (flanks') Do.
1Textural parametel'l u defined by Folk and Ward (1957). Values calculeted from data in ori8inal manuacript.
I Textural puametel'l u defined by Trask (1931). Avel'888 of values for this parameter.
Range of values for this parameter. Details of sample location or dune typa(.} 88IDpled. when known.

(fig. 21). The bimodality of interdune and serir toward negative skewness with decreasing grain
sands accounts for the poor sorting values of these size (fig. 21), a phenomenon also observed in inland
deposits (fig. 22), as has been noted by authors in dunes of Wyoming, using settling tube data
deserts around the world (Beheiry, 1967; Folk, (Ahlbrandt, 1974b). Kurtosis values show con-
1968a; Warren, 1972; Cooke and Warren, 1973). siderable scatter but are predominantly platykurtic
Skewness values generally show a tendency (fig. 21).

Digitized by Google
 TEXTURAL PARAMETERS OF EOLIAN DEPOSITS 27

Recognition of Depositional have widely differing textural parameters. they

reflect their source which is usually alluvial or lit-
Environments toral sand.
Inland dunes have variable mean grain size and
Traditionally. three depositional environments sorting values. because their source materials are
have been differentiated on a textural basis- normally from alluvial fan. glaciofluviatile.
alluvial. littoral. and eolian. Their delineation based fluviatile. lacustrine. or playa deposits.
solely on graphical measures of textural parameters The textural contrasts among inland dunes are il-
has not been satisfactory, as summarized by Fried- lustrated by comparing samples from different
man (1973. p. 626. 627): "Hence after many years of dune types and by comparing samples from the
empirical measurements. experimentation and base and crest of a given dune type (fig. 22). Tex-
compilation of statistical parameters based on size- tural contrasts occur both among dunes formed in
frequency distribution. textural parameters as clues different wind regimes (fig. 22) and among dunes
to depositional environment have been a source of formed in a given wind regime. For example. in in-
disenchantment. .. land dune fields with predominantly unidirectional
Although the three eolian subenvironments. wind regimes. both McKee (1966) and Ahlbrandt
coastal dune. inland dune. and interdune and serir. (1974b. 1975) observed improved sorting and finer

DUNE AND INTERDUNE DEPOSITS in the northwestern Namib Desert, South-West Africa. The star dune in the background is ap-
proximately 300 feet (100 m) high. Photograph by E. Tad Nichols. (Fig. 17.)
28 A STUDY OF GLOBAL SAND SEAS

~~--------------------------~

~40
Z
w
U
II:
~30

~-
z20
w
@
IE 10

0.0 0.4 2.0

Very
coarse
sand
Coarse
sand
Medium
sand
I
I
Fine
sand
Very
fine
sand
Very
_II ~ ~I
>
1i
'i 1i= !
t
- Poor
0
~ ~

GRAPHIC MEAN (Mz) SORTING (,,~

Finest 3.13 Mean 1.83 Best "I 0.11 Mean 0.73

Coarsest -0.88 "Mean 0.67 Poorest "I 1.88 Median 0.64
Medien 1.93

~ -
!Z40 t-
w
u ~

II:
~ 30 e-
~
>"
u
zw 20 e-
::J
0 roo--
w
IE 10 e-

0 I II
-0.8 -0.4 0.0 0.2 0.4 0.6 0.8
.W .l=! .s.! .l=!
Strongly I..~ i~ iiE 'i Strongly Very 1:: 1:: 1::
.2
~$ ~'f
.~ Ii
J11 >j
coarse fine Ej
platy- -'! -0
skewed 18! ~ II- ill skewed kurtic -i
li_
~3
SKEWNESS (SK~ TRANSGRAPHIC KURTOSIS (K'G)
Most positively s"-<l 0.73 Mean 0.13 Most peaked 0.73 Mean 0.49
Most negatively skewed -0.76 "Mean 0.29 Least peaked 0.33 " Mean 0.07
Median 0.13 Median O.~

GRAPHICAL MEASURES for 191 eolian sand samples. (Fig. 18.)

mean grain size downwind from the source of sand White Sands National Monument. New Mexico.
in a sequence of dome. transverse. barchan. and and also in the Killpecker dune field in south-
parabolic dunes. respectively. Dome dunes. which western Wyoming. The textural differences among
occur on the upwind margins of dune fields. are dune types are apparently not related to composi-
noticeably coarser grained and more poorly sorted tional differences as gypsum sand composes the
than are downwind dune types in the dunes at dunes at White Sands National Monument (McKee.

Digitized by Google
 TEXTURAL PARAMETERS OF EOLIAN DEPOSITS 29

2.000 I I I

1.BOOr- -
Poor
l 1.200 r- -

"~z Moderate 0.800 -

0
III Moderately well
0.400 -
., ..
,
-
Well "" .::.::.=.:.:
-. :-. .::-:=-
Very well 1 - ::-~"I
0.000
1 1

2.000 I I I

1.400- -
Strongly
fine
skewed -
0.800-
..
Fine skewed 0.200
Symmetrical
- . ":"==--
.. ~
~.
-
Coarse skewed
0.400 r- -
Strongly
coarse
skewed I 1 1 I
1.000

0.750
~ I I

~
leptokurtic
III 0.650 - -
~ I--------i
Mesokurtic 0.550 I- -
~ I--------i
u
Platykurtic -
!
~ Very
0.450 -

0.350 -

~~___ 0.2~.04>
platykurtic
I I 1 I
0.0 1.0 2.0 3.0 4.0
2.0mm 1.0 0.50 0.25 0.125 0.083

I ~:~ I
sand sand
I
Coarse Medium
sand
Fine
sand
I I ~i~': I
sand
MEAN GRAIN SIZE. Mz

GRAPHICAL MEASURES of 291 coastal dune sand samples. (Fig. 19.)

1966), whereas quartzose and feldspathic sand that is, winds flow from nearly opposite directions
compose the Killpecker dunes (Ahlbrandt, 1974b). at different times of the year (Merk, 1960). Streams
In reversing and multidirectional wind regimes, emerge from the Sangre de Cristo Range onto
sand accumulates in dunes that may have little net alluvial fans east of the dune field providing poorly
lateral migration. Reversing dunes and star dunes sorted coarse sand to the dunes. Periodically, wind
are examples of dunes that grow vertically rather flow reverses and unconsolidated alluvial and
than migrate laterally. (See fig. 17.) The clastic lacustrine or playa sediments in the San Luis Valley
material within these dunes represents a combina- to the west of the dune field provide well-sorted
tion of available source material rather than the fine sand (E. D. McKee, oral commun., 19'75). The
segregation and winnowing of source material resultant eolian deposits are a combination of the
downwind observed in dunes formed under two sands (fig. 22C).
unidirectional wind regimes. (Compare figs. 22C Thus, considering the textural heterogeneity of
and D with fig. 22A.) For example, the Great Sand inland dune sands, it is understandable why few
Dunes in Colorado have reversing wind flow- textural studies using graphical measures can uni-

Digitized by Google
30 A STUDY OF GLOBAL SAND SEAS
.-----2.000.------r,--".----.--
,----,-----, 2.000 , , , ,
1.600~ - 1.600- -

~
Poor
~
...:::
Poor
- ,

CI
1.2OO~
, "
-
~
1.200- . -
Z f--------l
!5~ Moderate 0.600 ~
.e:
.... ....\ ...~.: .. -
~ Moderate 0.800-
.. -
.
(J) Moderately well .... :.... .::--'"
~ i:'::' _
!5
(J)
Moderately well
Well 0.400 I- . ~::.:~:. -: ::: ...... - Well 0.400-
, .
Very well Very well
I I I I I I I I
0.000 0.000
2.000 , , , I
2.000 , , , ,
Strongly 1.400 ~ - Strongly 1.400- -
fine fine
skewed skewed
0.800 ~ - (J)
0.800- ,
, , Ifl
- --- ...... Z
, . , .., . :... -
~ Coarse
- Fine skewed 0.200-
Fine skewed 0.200 I-
Svmmetrical .:.':~:~:':-?~t2-:::~
., .:
._ .:, Svmmetrical
Coarse skewed (J) skewed
Strongly -0.400 ~ - Strongly 0.400- .. -

, .. ,
coarse coarse
skewed skewed I
1.000 I I I I 1.000

~ 0.750 , , , , , 0.750 , ,.
~ Leptokurtic
Leptokurtic
sa 0.650 I- - 0.650-

~ .- _.. . . " Mesokurtic 0.550- -

..-:" ,',-,'..
Mesokunic 0.550 f- , -
a :'~' ~:..::::.:
..
"-:. .-:,':
..
~ . . :":! w"':: " Platykunic 0.450- ' ,
J: Platykunic 0.450 ~ -
., , . -
~
o
'
.....
(J)
Very 0.350 I- - Very 0.350- '. -

~'-____
platykunic platykunic
I I I I 0.250 I I I I
0.250'-----"-----'---~---'-----'
-1.01/> 0.0 1.0 2.0 3.0 4.0 -1.01/> 0.0 1.0 2.0 3.0 4.0
2.0mm 1.0 0.50 0.25 0.125 0.063 2.0mm 1.0 0.50 0.25 0.125 0.063

I ~:~
sand
I Coarse
sand
IMedium
sand
Fine I i~Z I
I sand sand I ~:~e
sand
I Coarse
sand
IMedium
sand
I Fine
sand I!: I
MEAN GRAIN SIZE, Mz MEAN GRAIN SIZE, Mz

GRAPHICAL MEASURES of 175 inland dune sand samples. GRAPHICAL MEASURES of 40 interdune and serir sand samples.
(Fig. 20.) (Fig. 21.)

quely differentiate inland dune sand from its 23), Much recent detailed. work with graphical
source (Moiola and Weiser, 1968). The composi- measures of textural parameters attempting to
tional similarity of eolian sand and its fluvial sand differentiate coastal dune sand from littoral sand
source, the Colorado River delta, in the north- has met with limited success (Bigarella, Aless(,
western part of the Sonoran Desert has been shown Becker, and Duarte, 1969; Bigarella, 1972; Friedman,
by Merriam (1969). However, Friedman (1973, p, 1966,1973; Martins. 1967). Bigarella, AlessI, Becker,
627) suggested. that moment measures may be more and Duarte (1969, p. 16) found that beach and dune
discriminative, such as the use of sorting of inland sands in Brazil could not be distinguished using
dune and river sand versus swash-zone sand or mean grain size and sorting when the source is a
mean cubed deviation of inland dune versus river fine sand; however, med.ium- to coarse-grained
sand. sources showed dune sands to be better sorted and
Coastal dunes are generally composed of well- finer grained. than beach deposits, Bigarella (1972,
sorted. fine sand derived from beach deposits (fig. p, 14) questioned the significance of positive skew-

Digitized by Google
 TEXTURAL PARAMETERS OF EOLIAN DEPOSITS 31

35 54.0 PERCENT

Crest
30

25 Base Crest

20

15

10
I-
Z 5
~
a:
~ Q2~-~1~~0~~1~~2~3~~4~ -2 -1 -2 -1 0 1 2 3 4 -2 -1 012
~ A BARCHAN DUNE B C REVERSING DUNE D STAR DUNE
i:iz White Sands National
Monument. N. Max.
Libya Greet Sand Dunes. Colo. Saudi Arabia
Crest 626 (717) Crest 609 (645) Crest 601 (984)
g
w Crest 657 (1072)
Base 656 (1071) Base 590 (725)
DUNE DEPOSITS
Base 563 (643) Bese 648 (987)
a:
... 30 59.3 PERCENT
!i:
52
~ 25

20

15

10

o 1 2 -2 -1 -2 -1 012 4
E INTERDUNE G H INTERDUNE
White Sands National Libya Algeria Tunisia
Monument. N. Mex.
4137 (828) 519 (745) 4018 (26) 1621 (693)
INTERDUNE OR SERIR DEPOSITS

HISTOGRAMS SHOWING TEXTURAL CONTRASTS among dune and interdune samples. Values along abscissa are in whole phi;
histogram class intervals are quarter phi. Numbers below each histogram are the computer index number (outside parentheses) and
McKee index number (inside parentheses) of samples listed in tables 5-7. Silt- and clay-size fraction to the right of 4 f/>. (Fig. 22.)

ness as an eolian indicator because some dune and littoral sands. Frosted grains, as viewed with
sands have negative skewness values. Although the scanning electron microscope, have also been
dune sand is usually more rounded than beach used as an eolian indicator; however, that techni-
sand, Bigarella (1972, p. 14) found the most depen- que may be limited by the degree of diagenesis
dable eolian indicator to be light- and heavy- grains have undergone. Graphical measures are
mineral separations. certainly less discriminating and of more limited
use than these latter techniques. Sorting and mean
In the absence of sedimentary structures as prim- grain size values do have application as will be
ary evidence of eolian deposits, other criteria must shown. The application of skewness and kurtosis
be used. Discriminant analysis (Moiola and are, at best, tenuous because they show wide scat-
Spencer, chapter C), certain moment measures, and ter in eolian deposits (figs. 18 -21) and because
perhaps light- and heavy-mineral separations may skewness is dependent on mean grain size
be able to differentiate eolian sands from alluvial (Ahlbrandt, 1975; this chapter, fig. 21).

Digitized by Google
LARGE COMPLEX LINEAR DUNES which derive sand from adjacent coastal deposits. Slipfaces on the dunes face inland at the time
the photograph was taken. Sandfisch Bay, western part of Namib Desert, South-West Africa. Photograph by E. Tad Nichols. (Fig.
23.)

E.----,-----r----~----,-----,_--_.

Porosity and Permeability

Relationship to Textural !Zw 40

Parameters u
a:
~
~ 35
THE MEAN GRAIN SIZE. sorting. and packing
arrangement of grains affect porosity and per- ~
I/)
oa:
meability in alluvial. littoral. and eolian sands.
~ 30
Porosity in modem sediments decreases with
progressively more poorly sorted material regard-
less of grain size (fig. 24). Figure 24 indicates that 25L---~-----L-----L----~----L---~

the initial porosity (from 28 to 43 percent) is as Extremely Very Well Moderate Poor Very
well well poor
much as 15 percent lower in very poorly sorted
SORTING. TRASK SYSTEM
sand than in extremely well sorted sand. The cor-
responding pore-volume loss for a given grain-size RElATIONSHIP BETWEEN initial porosity and sorting. The curve
range is 35 percent. This range of values demon- was drawn for wet, packed sand from data by Beard and Weyl
strates the difference between very well sorted (1973). (Fig. 24.)

Digitized by Google
 TEXTIJRAL PARAMETERS OF EOLIAN DEPOSITS 33
have directional penneabilities that are parallel to
the long dimension of the deposit. The hetero-
geneity of sediments and more irregular packing ar-
Coarse rangement of grains in the river bar sand "packets"
(Pryor. 1973. p.187) resulted in lower effective per-
meabilities for them as compared with beach and
100
Medium
dune sands (table 3). Effective permeability
measures the flow across bedding units or
II)
w
"packets" of contrasting penneability.
U
a:
c(
0 Fine TABLE 3. - Porosity and permeability measurements in
~ alluvial, littoral, and coastal dune sand
~ 10 (Modified from Pryor. 1973. p. 182. n. number of samples: O. dareies: mO.
~ millidareies)
iii
c( Very fine Type of sand
w
~ Alluvial Littoral Coutal dune
a: (point bar)
w
A. (nz 348) (n= 184)
Average porosity .. (percent) .. 41 49 49
Range of porosity .. (percent) .. 17-52 39-56 42-54
1.0 Average permeability ....... . 93D 68D 54D
Range of permeability ...... . 4mD- 3.6- 5-104D
SOOD 166D

The contrast in textures among well-sorted to

very well sorted coastal dune sands. moderately
0.1'----'----'-----'-----'------l._..lo....----.J sorted to well-sorted inland dune sands. and poorly
Extremely Very Well Moderate Poor Very sorted interdune and serir sands should show
well well poor
SORTING. TRASK SYSTEM
progressively lower porosity and penneability. In
order to test the inland dune-interdune porosity
INTERRELATIONSHIP OF MEAN GRAIN SIZE (in miJIimetres) and penneability relationship. cores from four drill
and sorting on initial permeability for wet, packed sand from holes in the White Sands National Monument. New
data by Beard and Weyl (1973). (fig. 25.) Mexico. U.S.A . were studied. The differences in
sorting for the dune and interdune environments at
dune sand and poorly to very poorly sorted inter- White Sands and elsewhere are shown in figure 22.
dune and serir sands. Penneability decreases with The cores penetrate dune and interdune deposits.
poorer sorting and finer mean grain size (fig. 25) both of which are weakly cemented owing to the
but is also affected by packing arrangements (Rit- gypsiferous composition of the sand. The 14 sam-
tenhouse. 1971; Pryor 1973). Less symmetric pack- ples taken for porosity and penneability analyses
ing - for example. orthorhombic packing (rotated (fig. 26) show a general reduction in penneability
30) - results in greater porosity loss due to solu- for the interdune deposits (fig. 27). The high
tion than do cubicly packed grains (Rittenhouse. porosity yet low penneability of the interdune
1971. p. 82.) Porosity loss due to solution is max-
imized in very poorly sorted. extremely well sorted TABLE 4. - Porosity and permeability measurements in
or very angular sand and minimized in very well eolian subenvironments: Coastal dune, inland dune, and inter-
rounded. well-sorted sand (Rittenhouse. 1971. p. dune sand
86). The well-rounded and well-sorted character of (n. number of samples: O. dareies: mO. millidareies)
many eolian deposits favor them to have higher in- Type of sand
itial and diagenetically altered porosities (consider- Coastal dune Inland dune Interdune
(n = 184)1 (n = 6) (n=6)
ing solution only) than do alluvial. and perhaps lit- Average porosity .. (percent).. 49 26 31
toral. deposits. Range of porosity .. (percent) .. 42- 54 23- 30 25-38
Pryor (1973. p. 162) found that within a dune Range Average permeability. . . . . . . . 54 D 137 mD 120mD
of permeability ....... 5 - 104 D 28- 33-
sand body there is little variability in porosity and 450mD 270mD
penneability. whereas river bar and beach sands Pryor. 1973. p. 182.
1

Digitized by Google
34 A STUDY OF GLOBAL SA....) S S

DR ING ITE UM
0- 1 2 3 4 106"15'
-
j U r------------ ..
1 t---+to ! WHITE
I
j iJ
I.
~ ~ I\ ~
SANDS

I~ "
]
NA ON

20 ~ ~ I 1
~l
25-
] ry MONUMENT

~
~

35
~ h
V- .:. ] z i
i i ii
i
I
5 KIL
5
I
ETR
ES

~ _I
4O-:.= -~ I EXPLANATION
r-;-,
o
cati
illing
of
te nd E Play '<
M tain
se
45'~ U
t"= ~
50
TO 51.0 TO 510 u
LA ION
-=t===r
D ~

F-
Dune
dep .
Interdune
epo' t~
059
f===.3 :H="-
Flu PIa
deposits deposits
TO 65.0
D No
T
Total
eco dep

Plug taken for porosity and
rme ity rmi on

LOCATIONS OF CORES and depth of plug samples taken for porosity and permeability determinations, White Sands National
nu nt, W ico .5. ore scr ion nd difi ind ma om cK nd iol 97 .60 5). g.2

nil 70r It
 TEXTURAL PARAMETERS OF EOLIAN DEPOSITS 35

deposits suggests that much of the porosity is in-

effective and probably is incorporated into the fine-
grained matrix.
It appears that the textural differences in eolian
subenvironments also bear significantly on initial
porosity and permeability (table 4).
~ 100
The more poorly sorted interdune deposits may
form effective permeability barriers after burial
owing to greater grain-to-grain contacts which in-
crease porosity loss by pressure solution and ce-
mentation. Tightly cemented, poorly sorted inter-
o Dune dune deposits separate porous well-sorted dune
Interdune
sandstones in the Tensleep and Weber Sandstones
1~O ~ ~ U ~ ~ ~ ~ ~ ~ ~ (Pennsylvanian and Permian) in the Wertz and
POROSITY. IN PERCENT Brady oil fields, Wyoming, U.S.A., forming local
barriers to oil migration. The porosity and per-
POROSITY-PERMEABILITY DISTRIBUTION of dune and inter-
dune sediments from cores, White Sands National Monument,
meability contrasts in the two deposits must relate
New Mexico, U.S.A Analyses by Core Laboratories, Inc., in part to the textural differences of the dune-inter-
Denver, Colorado, U.S.A Location and description of cores are dune sediments that compose part of the Tensleep
shown in figure 26. (fig. 27.) and Weber Sandstones (chapter I).

Digitized by Google
 TABLE 5. - Graphical measures of sand samples from coastal dunes used as a data base ~

Locality Mean grain size Median grain Sort- Skew- Kur-

Computer McKee's size ---.!!!L ness t08is
Principal county Specific area Index No. Index No. mm mm Ski Kb
P,
or area
'" '"
Bahama Islands Bimini 1622 580 1.916 0.2661 1.870 0.2736 0.495 0.221 0.536
531 580 1.186 0.4383 1.200 0.4353 0.425 0.020 0.486
1619 586 1.710 0.3057 1.800 0.2872 0.704 0.137 0.499
Bermuda Elbow Beach 1556 22 2.360 0.1948 2.280 0.2059 0.537 0.272 0.544
Brazil Garopaba Sirin (State of Santa 1510 402 2.630 0.1615 2.610 0.1639 0.299 0.025 0.445
Catarina)
1511 403 2.550 0.1708 2.550 0.1708 0.258 0.014 0.459
1512 404 2.440 0.1843 2.400 0.1895 0.356 0.128 0.468
1513 405 2.480 0.1792 2.480 0.1792 0.300 0.002 0.452
1514 406 2.330 0.1989 2.270 0.2073 0.363 0.166 0.460 >
1515

1516
407

408
2.240

2.660
0.2117

0.1582
2.120 0.2300

2.650 0.1593 0.233

0.309 0.456

0.117
0.503

0.513
~-<
1517 409 2.630 0.1615 2.60q 0.1649 0.313 0.166 0.449 0
"1:1
Brazil Island of Santa Catarina 772 7 1.800 0.2872 1.810 0.2852 0.255 0.248 0.597
773
774
6
5
2.200
1.900
0.2176
0.2679
2.203 0.2172 0.318
1.920 0.2643 0.302
0.053
0.182
0.514
0.525 6
775
776
4
3
1.920
2.050
0.2643
0.2415
1.926 0.2632 0.211
2.076 0.2372 0.223
0.127
0.171
0.559
0.534
~
777 2 2.100 0.2333 2.140 0.2269 0.234 0.261 0.521 ~
778 1 2.060 0.2398 2.096 0.2339 0.195 0.368 0.557
~
779
780
781
4
4
4
2.190
2.100
2.020
0.2192
0.2333
0.2466
2.203
2.130
2.030
0.2172
0.2285
0.2449
0.232
0.252
0.119
0.1""
0.184
0.223
0.510
0.575
0.561
~
782 3 2.120 0.2300 2.173 0.2217 0.190 0.487 0.516

783 3 2.100 0.2333 2.133 0.2280 0.294 0.132 0.530

784 3 2.100 0.2333 2.126 0.2291 0.244 0.263 0.583
785 2 2.290 0.2045 2.296 0.2036 0.176 0.130 0.546
786 2 2.250 0.2102 2.276 0.2065 0.265 0.195 0.528
0 787 2 2.300 0.2031 2.310 0.2017 0.285 0.157 0.494
5"
;:::;:
N" 788 1 2.200 0.2176 2.230 0.2132 0.174 0.308 0.531
CD
0. 789 1 2.200 0.2176 2.216 0.2152 0.185 0.263 0.571
rr 790 1 2.200 0.2176 2.210 0.2161 0.193 0.233 0.550
'<
791 8 2.200 0.2176 2.230 0.2132 0.178 0.230 0.520
C) 792 9 2.300 0.2031 2.220 0.2146 0.184 0.261 0.530
0 793 10 2.250 0.2102 2.210 0.2161 0.183 0.250 0.524

~
794 11 2.300 0.2031 2.210 0.2161 0.178 0.260 0.525
795 12 2.280 0.2059 2.200 0.2176 0.177 0.250 0.530
~ 796 13 2.250 0.2102 2.210 0.2161 0.176 0.230 0.550
rv 797 14 2.300 0.2031 2.250 0.2102 0.178 0.260 0.573
 798 15 2.290 0.2045 2.250 0.2102 0.180 0.261 0.571
799 16 2.300 0.2031 2.210 0.2161 0.183 0.230 0.570
Brazil Laguna (State of Santa Catarina) 1500 241 2.620 0.1627 2.700 0.1539 0.276 0.420 0.532
1501 242 2.590 0.1661 2.600 0.1649 0.195 0.001 0.558
1502 243 2.450 0.1830 2.450 0.1830 0.328 0.068 0.452
1503 244 2.520 0.1743 2.550 0.1708 0.276 0.146 0.411
1504 245 2.640 0.1604 2.700 0.1539 0.243 0.436 0.564
1505 246 2.530 0.1731 2.580 0.1672 0.271 0.207 0.432
1506 247 2.800 0.1436 2.800 0.1436 0.211 0.023 0.464
1507 248 2.660 0.1582 2.700 0.1539 0.223 0.308 0.524
1508 249 2.670 0.1571 2.700 0.1539 0.248 0.169 0.560
1509 250 2.600 0.1649 2.610 0.1638 0.237 0.129 0.534
Brazil Morro Convento. (late Clube) 1538 434 2.320 0.2003 2.400 0.1895 0.387 0.045 0.460

1539 435 2.550 0.1708 2.550 0.1708 0.326 0.050 0.450

1540 436 2.530 0.1731 2.500 0.1768 0.337 0.131 o-i
0.498

~>
1541 437 2.600 0.1649 2.590 0.1661 0.266 0.077 0.490
1542 438 2.540 0.1719 2.530 0.1731 0.296 0.027 0.411
1543 439 2.520 0.1743 2.500 0.1768 0.299 0.148 0.424
t"'"
1544 440 2.540 0.1719 2.540 0.1719 0.344 0.047 0.467 'tJ
1545 441 2.550 0.1708 2.490 0.1780 0.331 0.235 0.460 >
::c
1546 442 2.490 0.1780 2.450 0.1830 0.351 0.158 0.425 >
Brazil Planicie Co.teira (State of Rio
1548
663
444
1
2.440
2.800
0.1843
0.1436
2.450
2.833
0.1830
0.1406
0.329
0.206
0.007
0.267
0.475
0.524
i
Grande do Sol).
;i
::c
en
664 3 2.900 0.1340 2.903 0.1340 0.125 0.148 0.662
665 4 2.600 0.1649 2.603 0.1649 0.251 0.089 0.513 0
"1:1
666 8 2.760 0.1476 2.770 0.1466 0.209 0.124 0.677
667 10 2.800 0.1436 2.886 0.1350 0.331 0.509 0.616 ~
t"'"
668 11 2.890 0.1350 2.840 0.1397 0.297 0.233 0.509

669 14 2.600 0.1649 2.603 0.1649 0.181 0.021 0.547

~
0

~
670 15 2.500 0.1768 2.500 0.1768 0.221 ND 0.587
671 16 2.590 0.1661 2.593 0.1661 0.186 0.013 0.535
672 17 2.500 0.1768 2.526 0.1731 0.130 0.208 0.505
673 18 1.590 0.3322 1.596 0.3299 0.203 0.040 0.537

674 9 2.800 0.1436 2.806 0.1426 0.288 0.044 0.461

675 19 2.560 0.1696 2.580 0.1672 0.212 0.120 0.500
676 20 2.600 0.1649 2.600 0.1649 0.097 0.045 0.529
0 677 21 2.570 0.1684 2.570 0.1684 0.170 0.069 0.542
5" 678 24 2.550 0.1708 2.550 0.1708 0.760 ND 0.487
;:::;:
N"
CD
0.. 679 26 2.600 0.1649 2.610 0.1638 0.178 0.071 0.494
rr 680 28 2.500 0.1768 2.513 0.1756 0.180 0.129 0.543
'< 681 2.580 0.1672
30 2.580 0.1672 0.167 0.053 0.530
C) 682
683
34
36
2.500
2.320
0.1768
0.2003
2.513
2.360
0.1756
0.1948
0.200
0.284
0.156
0.217
0.519
0.505
0 (0)

~
"1

~
rv
 Col)
TABLE 5. - Graphical measures of sand samples from coastal dunes used as a data base - Continued co

Locality Mean grain size Median grain Sort, Skew- Kur-

Computer McKee's size ---..!!!L ness tosis
Principal county
or area
Specific area Index No. Index No.
mm
mm P, Sk, KG

Brazll--Con. Planicie C08teira (State of Rio 684 38 2.700 0.1539 2.680 0.1560 0.185 0.079 0.528
Grande do Sol)--Con. 685 39 2.250 0.2102 2.253 0.2102 0.143 0.018 0.518
686 40 1.050 0.4841 1.093 0.4697 0.471 0.139 0.581
687 41 2.390 0.1908 2.413 0.1882 0.279 0.158 0.520
688 42 2.570 0.1684 2.590 0.1661 0.216 0.198 0.557

689 43 2.480 0.1792 2.486 0.1780 0.251 0.098 0.520

690 45 2.520 0.1743 2.530 0.1731 0.254 0.047 0.511
691 46 2.750 0.1487 2.750 0.1487 0.206 ND 0.601
692 44 2.500 0.1768 2.510 0.1736 0.253 0.086 0.512 >
693

694
695
47

48
49
2.720

2.500
2.510
0.1518

0.1768
0.1756
2.760

2.516
2.530
0.1476

0.1743
0.1731
0.196

0.249
0.238
0.208

0.134
0.113
0.505
0.482
0.484
a
~
696 50 2.480 0.1792 2.476 0.1792 0.230 0.024 0.524 ~
697 51 2.600 0.1649 2.636 0.1604 0.290 0.119 0.493
698

699
700
53

57
59
2.500

2.490
2.480
0.1768

0.1780
0.1792
2.503

2.506
2.466
0.1768

0.1756
0.1805
0.303

0.288
0.291
0.148

0.107
0.020
0.534

0.520
0.560

til
701 62 2.580 0.1672 2.596 0.1649 0.176 0.112 0.510 >
702
703
63
65
2.460
2.500
0.1817 2.483 0.1792 0.280 0.154 0.477 ~
0.1768 2.530 0.1731 0.245 0.200 0.520 til
tz:I
704 67 2.580 0.1672 2.530 0.1731 0.298 0.107 0.471 ~
705 70 2.490 0.1780 2.503 0.1768 0.237 0.060 0.563
706 71 2.410 0.1882 2.420 0.1869 0.271 0.322 0.477
707 72 2.600 0.1649 2.600 0.1649 0.206 ND 0.550
708 73 2.690 0.1550 2.636 0.1604 0.213 0.286 0.541

0
709 78 2.510 0.1756 2.503 0.1768 0.276 0.018 0.521
2' 710 79 2.520 0.1743 2.533 0.1731 0.191 0.100 0.515
;:::;: 711 80 2.400 0.1895 2.420 0.1869 0.243 0.196 0.533
N'
CD 712 81 2.390 0.1908 2.403 0.1895 0.249 0.189 0.590
0..
713 83 2.410 0.1882 2.453 0.1830 0.271 0.258 0.511
rr
'<
714 87 2.370 0.1934 2.391 0.1908 0.293 0.214 0.494
C) 715 89 2.600 0.1649 2.610 0.1638 0.256 0.111 0.599
716 93 2.390 0.1908 2.393 0.1908 0.270 0.077 0.487
0 717 97 2.520 0.1743 2.546 0.1708 0.257 0.172 0.482

~
718 99 2.500 0.1768 2.520 0.1743 0.296 0.148 0.613

~
rv
 719 102 2.480 0.1792 2.506 0.1756 0.266 0.154 0.520
720 109 2.220 0.2146 2.306 0.2017 0.263 0.417 0.450
721 112 2.390 0.1908 2.366 0.1934 0.240 0.024 0.509
722 113 2.310 0.2017 2.343 0.1975 0.261 0.355 0.542
723 114 2.280. 0.2059 2.283 0.2059 0.258 0.269 0.469

724 115 2.420 0.1869 2.466 0.1805 0.289 0.220 0.487

725 116 2.360 0.1948 2.386 0.1908 0.273 0.144 0.460
726 117 2.390 0.1908 2.401 0.1895 0.235 0.099 0.453
727 123 2.510 0.1756 2.540 0.1719 0.247 0.225 0.559
728 124 2.320 0.2003 2.393 0.1908 0.311 0.351 0.493
729 125 2.600 0.1749 2.603 0.1649 0.225 0.028 0.491
730 126 2.600 0.1749 2.630 0.1615 0.276 0.149 0.558
731 129 2.500 0.1768 2.516 0.1756 0.282 0.123 0.505
732 132 2.600 0.1649 2.596 0.1661 0.229 0.011 0.596
733 134 2.500 0.1768 2.510 0.1756 0.275 0.094 0.473 o-i
tzj

734
735
138
140
2.610
2.600
0.1638
0.1649
2.650
2.606
0.1593
0.1649
0.180
0.241
0.894
0.056
0.529
0.534
~::c
736 141 2.600 0.1649 2.566 0.1696 0.236 0.175 0.616 ~
737 142 2.400 0.1895 2.420 0.1869 0.273 0.134 0.507 'a
738 143 2.410 0.1882 2.436 0.1856 0.249 0.177 0.497 >
::c
>
739 144 2.450 0.1830 2.483 0.1792 0.294 0.153 0.505 3:
740 145 2.400 0.1895 2.406 0.1895 0.213 0.125 0.551 ~
tzj
741 147 1.290 0.4090 1.466 0.3635 0.791 0.262 0.410 ::c
742 149 2.400 0.1895 2.403 0.1895 0.215 0.088 0.561 en
743 153 2.500 0.1768 2.523 0.1743 0.264 0.123 0.481 0
'"r.I
tzj
744 155 2.610 0.1638 2.623 0.1627 0.247 0.058 0.501
745 156 2.060 0.2398 2.066 0.2398 0.486 0.049 0.527 ~
746
747
157
158
2.690
2.630
0.1550
0.1615
2.703
2.646
0.1539
0.1604
0.141
0.147
0.291
0.252
0.578
0.608
~
0
748 159 2.600 0.1649 2.626 0.1627 0.178 0.240 0.571 tzj

~
0.127 0.143 0.630
~
749 160 2.690 0.1559 2.693 0.1559
750 161 2.510 0.1756 2.536 0.1731 0.158 0.251 0.585
751 162 2.650 0.1593 2.680 0.1560 0.206 0.240 0.541
752 163 2.480 0.1792 2.520 0.1743 0.237 0.245 0.491
753 164 2.580 0.1672 2.576 0.1684 0.157 0.114 0.632

754 165 2.510 0.1756 2.543 0.1719 0.239 0.211 0.514

0
5. 755 166 2.580 0.1672 2.576 0.1684 0.197 0.034 0.545
;:::;: 756 167 2.500 0.1768 2.516 0.1756 0.193 0.213 0.566
N 757 168 2.490 0.1780 2.496 0.1780 0.221 0.162 0.609
CD
Q. 758 169 2.500 0.1768 2.500 0.1768 0.214 0.166 0.551
rr
'<
759 170 2.490 0.1780 2.470 0.1805 0.281 0.069 0.489
C) 760
761
171
172
2.610
2.550
0.1638
0.1708
2.626
2.566
0.1627 0.189
0.1696 0.225
0.166
0.128
0.548
0.505
0 762 173 2.600 0.1649 2.606 0.1649 0.212 0.086 0.500 (0)

~
763 174 2.490 0.1780 2.533 0.1731 0.240 0.260 0.515 CO

~
rv
 BLE 5o-Gra s of sand sampl 1 dunes used as Continued ~
Locality Mean Median grain Kur-
Computer McKee's size ~ ness tosis
Principal county
or area
Specific area IndexNoo IndexN,oo
mm
mm P, Sk, KG

Brazll--Con. Planicie Costeira (State of Rio 764 175 2.600 0.1649 2.636 0.1615 0.228 0.183 0.488
Grande do Sol 765 176 2.600 o 1649 2 626 0.1627 o 259 o 164 0.472
177 2.600 630 0.1615 9 0.522
178 2.550 556 0.1708 0 0.489
179 2.600 633 0.1615 6 0.544

180 2.500 513 0.1756 3 0.482

181 2.510 567 0.1696 5 0.481
771 182 2.550 0.1708 2.570 0.1684 0.256 0.117 0.483
1567 280 2.693 0.1550 2.700 0.1539 0.201 0.097 0.554
)-
1568 3.10 2.693 0.1550 2.690 0.1550 0.215 0.058 0.523
en
1569
1570
311
311
2.506
2.640
0.1768
0.1604
2.600
2.610
0.1649
0.1638
0.278
0.239
0.360
0.199
0.527
0.542

-<
312 2.493 510 0.1756 5 0.483 0"Jj
313 2.616 600 0.1649 8 0.466
315 2.660 650 0.1593 8 0.534 C"l
t"'"
0
316
317
2.643
2.706
630
700
0.1615
0.1539
7 0.590
0 0.502 ~
en
1576 318 2.573 0.1684 2.550 0.1708 0.283 0.185 0.473
1577
1578
320
321
2.693
2.500
0.1550
0.1768
2.690
2.500
0.1550
0.1768
0.226
0.300
0.029 0.510
0.095 0.473 ~
en
1579
1580
322
324
2.710
2.496
0.1528
0.1780
2.700
2.430
0.1539
0.1856
0.318
0.236
0.022
0.429
0.476
0.483
~
1581 341 2.696 o 1550 2 700 0.1539 0208 o 083 0.534
342 2.700 700 0.1539 0.609
343 2.706 700 0.1539 8 0.534

344 2.753 750 0.1487 9 0.582

345 2.623 600 0.1649 6 0.539
346 2.640 620 0.1627 0 0.514
-J
D 1587 347 2.606 0.1649 2.600 0.1649 0.204 0.090 0.551
~
1588 348 2.500 0.1768 2.500 0.1768 0.280 0.023 0.456
!
2.573 0.1684 2.550 0.1708 0.311 0.128 0.518
:J
--;
1589
1590
1591
350
351
352
2.553
2.653
0.1708
0.1593
2.550
2.680
0.1708
0.1560
0.275
0.264
0.035 0.493
0.352 0.438
353 2.636 640 0.1604 6 0.559
356 2.450 450 0.1830 0.473
 1594 362 2.740 0.1497 2.700 0.1539 0.198 0.342 0.551
1595 364 2.806 0.1436 2.800 0.1436 0.198 0.114 0.547
1596 365 2.643 0.1604 2.700 0.1539 0.253 0.346 0.586
1597 366 2.726 0.1518 2.700 0.1539 0.218 0.215 0.603
1598 367 2.660 0.1582 2.620 0.1627 0.246 0.169 0.595
1599 368 2.700 0.1539 2.710 0.1528 0.210 0.218 0.561
1600 368 2.703 0.1539 2.700 0.1539 0.213 0.006 0.611
1601 369 2.683 0.1560 2.700 0.1539 0.205 0.590 0.615
1602 370 2.686 0.1560 2.680 0.1560 0.218 0.008 0.457
1603 371 2.706 0.1539 2.680 0.1560 0.260 0.111 0.581

1604 386 2.703 0.1539 2.690 0.1550 0.219 0.119 0.713

1605 387 2.653 0.1593 2.650 0.1593 0.250 0.039 0.521
1606 388 2.676 0.1571 2.680 0.1560 0.228 0.014 0.530
1607 389 2.656 0.1593 2.630 0.1615 0.229 0.184 0.518 ;1
1608 390 2.510 0.1756 2.500 0.1768 0.215 0.069 0.523
1609
1610
391
392
2.616
2.700
0.1638
0.1593
2.600
2.700
0.1649
0.1593
0.211
0.206
0.115 0.504
0.071 0.524 ~
:::a
1611
1612
393
398
2.576
2.500
0.1684
0.1768
2.680
2.500
0.1672
0.1768
0.318
0.277
0.381 0.470
ND 0.450 ~
't7
0.1895 2.350 0.1961 0.276 0.342 0.497
~
1613 399 2.400

1624 303 2.670 0.1571 2.740 0.1497 0.303 0.351 0.423 3:

0.073 0.495
1625
1626
304
305
2.710
2.730
0.1528
0.1507
2.700
2.700
0.1539
0.1539
0.257
0.267 0.094 0.501 ~
tzj
1627 306 2.680 0.1560 2.670 0.1571 0.220 0.131 0.505 :::a
en
1628 307 2.660 0.1582 2.660 0.1582 0.246 0.050 0.505 0
"r.I
1629
1630
308
309
2.690
2.690
0.1550
0.1550
2.680
2.690
0.1560
0.1550
0.215
0.213
0.084 0.513
0.071 0.541
~
t"'"
1518
1519
414
415
2.830
2.550
0.1406
0.1708
2.810
2.500
0.1426
0.1768
0.186
0.361
0.145 0.501
0.060 0.591 ~
1520 416 2.680 0.1560 2.620 0.1627 0.239 0.302 0.489 C

1521 417 2.670 0.1571 2.650 0.1593 0.255 0.087 0.478 ~5!3
1522 418 2.720 0.1518 2.700 0.1539 0.255 0.094 0.492 @
1523 419 2.620 0.1627 2.560 0.1696 0.391 0.160 0.589
1524 420 2.490 0.1780 2.450 0.1830 0.485 0.060 0.540
1525 421 2.630 0.1615 2.600 0.1649 0.282 0.081 0.630

1526 422 2.580 0.1672 2.600 0.1649 0.482 0.244 0.661

0 1527 423 2.570 0.1684 2.560 0.1696 0.309 0.047 0.472
5" 0.284 0.495
;:::;: 1528 424 2.600 0.1649 2.550 0.1708 0.287
N" 1529 425 2.660 0.1582 2.660 0.1582 0.329 0.028 0.460
CD
0.. 1530 426 2.580 0.1672 2.570 0.1684 0.300 0.028 0.452
rr
'<
1531 427 2.620 0.1627 2.600 0.1649 0.296 0.079 0.450
C) 1532 428 2.720 0.1518 2.700
2.420
0.1539
0.1869
0.221
0.323
0.157
0.130
0.488
0.819
1533 429 2.440 0.1843
0 1534 430 2.640 0.1604 2.630 0.1615 0.218 0.017 0.589
....
~
2.810 0.1426 2.800 0.1436 0.166 0.124 0.559 ~
1535 431

~
rv
 TABLE 5. - Graphical measures of sand samples from coastal dunes used as a data base - Continued ".
N

Locality Mean srain size Median grain Sort- Skew- Kur-

Computer McKee's size ~ ness tosis
Principe) county
or area
Specific area Index No. Index No.
mm
mm P, Sk, Kb

1536 432 2.410 0.1882 2.420 0.1869 0.313 0.012 0.384

1537 433 2.450 0.1830 2.420 0.1869 0.330 0.161 0.452
Brazil State of R10 Grande do Sul 500 14 2.603 0.1649 2.600 0.1649 0.181 0.021 0.547
501 23 2.556 0.1696 2.550 0.1708 0.170 0.056 0.642
502 17 2.526 0.1731 2.500 0.1708 0.130 0.208 0.505
503 28 2.513 0.1756 2.500 0.1708 0.180 0.129 0.543
504 34 2.513 0.1756 2.500 0.1708 0.200 0.156 0.519
505 19 2.580 0.1672 2.560 0.1696 0.212 0.120 0.500

506 21 2.570 0.1684 2.570 0.1684 0.170 0.069 0.542

507 15 2.500 0.1708 2.500 0.1708 0.221 tID 0.578
508 18 1.596 0.3299 1.590 0.3322 0.203 0.040 0.537 >
509 16 2.593 0.1661 2.590 0.1661 0.186 0.013 0.535
510 20 2.600 0.1649

2.580 0.1672
2.600 0.1649 0.097 0.045 0.529
~-<
511 30 2.580 0.1672 0.167 0.053 0.530 0
512 10 2.886 0.1350 2.800 0.1436 0.331 0.509 0.616 "r.I
513
514
8
3
2.770 0.1466
2.903 0.1340
2.760
2.900
0.1476
0.1340
0.209
0.125
0.124
0.148
0.677
0.662
~
515 11 2.840 0.1397 2.890 0.1350 0.297 0.233 0.509 ~
~
516 2.603
517
4
9
2.600
2.806
0.1649
0.1426 2.800
0.1649
0.1436
0.251
0.288
0.039
0.044
0.513
0.461 ~
Z
518 1 2.833 0.1406 2.800 0.1436 0.206 0.267 0.524 C
rJ'J
533 51 2.636 0.1604 2.600 0.1649 0.290 0.119 0.493 ~
534 56 2.403 0.1895 2.400 0.1895 0.215 0.088 0.561 >
rJ'J

535 42 2.590 0.1661 2.570 0.1684 0.216 0.198 0.557

536 41 2.413 0.1882 2.390 0.1908 0.279 0.158 0.520
537 43 2.486 0.1780 2.480 0.1792 0.251 0.098 0.520
538 46 2.750 0.1487 2.750 0.1487 0.206 HD 0.601
0
539 45 2.530 0.1731 2.520 0.1743 0.254 0.047 0.511
5"
;:::;: 0.524
N" 540 50 2.476 0.1792 2.480 0.1792 0.230 0.024
CD 541 49 2.530 0.1731 2.510 0.1756 0.238 0.113 0.484
0..
rr
542 47 2.760 0.1476 2.720 0.1518 0.196 0.207 0.505
'< 543 39 2.530 0.1731 2.250 0.2102 0.143 0.018 0.518
544 44 2.510 0.1756 2.500 0.1768 0.253 0.086 0.512
C)
26 2.610 0.1638 2.600 0.1649 0.178 0.071 0.494
0 545
546 24 2.550 0.1708 2.550 0.1708 0.760 HD 0.487

~
~
547
548
552
36
586
38
2.,360
1.393
2.680
0.1948
0.3896
0.1560
2.320
1.310
2.700
0.2003
0.4033
0.1539
0.284
0.525
0.185
0.217 0.505
0.178 1.147
0.079 0.528
rv
 California (U.S.A.) Pi8mo Beach 530 548 3.353 0.0981 3.350 0.0981 0.388 0.007 0.484
635 548 2.730 0.1507 2.710 0.1528 0.355 0.580 0.494
Canada Northumberland Coast 1620 623 2.233 0.2132 2.200 0.2176 0.826 0.132 0.464
Libya Ghat 569 623 1.823 0.2832 1.820 0.2832 0.233 0.014 0.513
Mexico Cholla Bay 1560 31 1.833 0.2813 1.800 0.2872 0.617 0.082 0.549
New Jersey (U.S.A.) Long Beach Island 1565 659 1.586 0.3345 1.600 0.3299 0.386 0.064 0.515
610 658 1.680 0.3121 1.680 0.3121 0.365 0.008 0.476
522 659 1.613 0.3276 1.680 0.3121 0.366 0.285 0.495
613 660 1.560 0.3392 1.530 0.:3463 0.400 0.122 0.514

~F:
~
~
~

~
I
0
5"
2:
N
CD
0..
rr
'<

C)
0
~
".
Col)

~
rv
 TABLE 6. - Graphical measures of sand samples from inland dunes used as a data base :t
[Localities are listed alphabetically by country. and specific area where known. The computer index number and McKee's index number
are also listed. Specific sampling information relating to dune type or position on a dune is listed under "remarks". ND indicates
no data]

Locality Mean grain size Median grain Sort- Skew- Kurt-

size ing ness osis

Principal country or Specific area Computer McKee's phi mm phi mm rho 1 Skl """K"G Remarks
area Index No. Index
No.

Inland dunes

Texas (U.S.A.) Padre Island 628 962 2.480 0.1792 2.800 0.1436 0.270 0.l41 0.500
627 965 2.690 0.1550 2.680 0.1560 0.310 0.076 0.421 >
~-<
629 967 2.350 0.1961 2.360 0.1948 0.470 0.683 0.685
Algeria Bou Saade 4019 27 2.026 0.2455 ND ND 0.549 0.247 0.536 Crest of
dune. 0
Ouarsla 1558 28 2.190 0.2192 2.100 0.2333 0.504 0.311 0.531 Seif dune. "1:1
slipface. C)
t""
Arizona(U.S.A.) Coal Mountain 1559 29 1.563 0.3392 1.500 0.3536 0.409 0.321 0.612
Carrizo Mountains
Grand Canyon
4173
1616
1134
566
3.012
3.056
0.1240
0.1207
ND
3.210
ND
0.1081
0.474
0.613
0.116
0.357
0.481
0.540
~
t""
566 566 2.703 0.1539 2.690 0.1550 0.545 0.131 0.529 rJ'J
>
Z
Grand Falls 4028 37 1.195 0.4368 ND ND 0.638 0.291 0.533 0
Hotevilla 605 570 1.900 0.2679 1.900 0.2679 0.366 ND 0.511 ~
Leupp 1551 9 2.686 0.1560 2.700 0.1539 0.530 0.020 0.526 Barchan >
rJ'J
dune. swale
between
crests.
1552 10 2.296 0.2045 2.250 0.2102 0.427 0.220 0.527 Barchan
dune. crest.
1553 11 2.403 0.1895 2.380 0.1921 0.516 0.136 0.528 Do.
4010 12 2.457 0.1821 ND ND 0.516 0.091 0.507 Do.
0 4011 13 2.402 0.1892 ND ND 0.573 0.023 0.519 Barchan
2'
;:::;: dune.
N' windward
CD
0.. slope.
rr 1554 14 2.233 0.2132 2.200 0.2176 0.608 0.090 0.521 Do.
'<
1555 16 2.336 0.1989 2.280 0.2059 0.433 0.193 0.515 Barchan
C) tloenkopi 634 569 2.670 0.1571 2.610 0.1638 0.480 0.230
dune. crest.
0.468 Hanging
0 dune.

~
slipface.
Page 558 568 2.036 0.2432 1.980 0.2535 0.611 0.156 0.500 Seif dune.
~ windward
rv slope.
 1617 568 2.273 0.2073 2.250 0.2102 0.753 0.023 0.432 Do.
Red Lake 584 654 2.293 0.2045 2.300 0.2031 0.615 0.004 0.512 Slipface of
dune. at
2o-ft
height.
White Sands National 4022 30 2.675 0.1566 ND ND 0.363 0.056 0.516
Monument.
Tonalea 1561 33 2.300 0.2031 2.200 0.2176 0.961 0.1610.427 Crest of
dune.
4025 34 2.612 0.1636 ND NO 0.654 -0.006 0.515 Windward
slope of
dune.
4016 23 2.698 0.1541 NO NO 0.731 0.056 0.438 Crest of
dune.
Yuma 1562 35 2.490 0.1780 2.380 0.1921 0.6~7 0.296 0.520 Do. --i

~
~.I0" 36 2.363 0.1948 2.300 0.2031 1.021 0.080 0.427
Australia Betoot 4071 699 2.175 0.2214 NO NO 0.364 0.364 0.552 Seif dune.
Canoing Basin ND 0.574 0.299 0.513 Do.
:::a
4041 625 2.058 0.2401 ND
4042 626 2.558 0.1698 ND NO 0.518 0.192 0.511 Do. ?!"1:1
4043 627 1.665 0.3153 NO NO 0.450 0.359 0.572 Do.
4044 628 1.746 0.2981 NO NO 1.119 0.422 0.450 Do. ~
4045 629 2.402 0.1892 NO NO 0.540 0.267 0.519 Do. >
4046
4047
630
631
1.611
2.236
0.3274
0.2123
NO
NO
ND
ND
0.450
0.738
0.383
0.018
0.560
0.500
Do.
Do. ~
:::a
en
4048 632 1.979 0.2537 NO NO 0.474 0.317 0.579 Do.
4049 633 1.882 0.2713 NO NO 0.264 0.134 0.557 Do. 0
"1:1
Monkira 578 698 2.340 0.1975 2.300 0.2031 0.354 0.243 0.531 Seif dune. tzj
crest. 0
t""
Belgium Ooatende 560 620 2.433 0.1856 2.440 0.1843 0.249 0.081 0.504 :;
California(U.S.A.) Fallon 568 556 1.813 0.2852 1.800 0.2872 0.283 0.200 0.602 Transverse Z
dune. 300 C

~
ft high.
base of
slipface.
528 556 1.810 0.2852 1.790 0.2892 0.303 0.196 0.563 Do. ~
Twentynine Palms 562 653 2.320 0.2003 2.310 0.2017 0.479 0.009 0.513 Transverse
dune.
slipface.
Colorado(U.S.A.) Fort Morgan 4167 1092 1.947 0.2594 NO ND 0.566 0.033 0.510 Climbing
0 dune.
5"
;:::;: Great Sand Dunes National 606 18 1.810 0.2852 1.800 0.2872 0.364 0.134 0.475 Reversing
N" Monument. dune.
CD
0.. 4015 19 1.939 0.2608 ND NO 0.321 0.035 0.531 Reversing
rr dune. high
'<
on slipface.
C) 572 21 -0.146 1.1096 0.250 1.1892 0.825 0.081 0.727 Reversing
dune, base.
0
~
~
en
rv
 TABL Graphi sures of amples . land d d as ad se-Co ed ~
--- ---
Locality
--- --- --- ---
eangrain
---
Media
---
Sort-
---
w-
---
Computer McKee's size __ ~ ness tosis
Remarks
Principal country Specific area Index No. Index No. mm mm P, Ski Kb

Co
-- or -
(U.S.A.
area ---
Great
---
unes N
---
1 5
---
643
'" ---
86 1.
'"
---
0.600
---
0.14
---
485 o.
---
C Mon -Con. 5 644 33 o. 0.800 0.24 212 o. versin
dune. 300
ft above
base.
609 645 2 090 0.2339 2.100 o 2333 0.262 o 050 0.530 R versing
une. c
41 1211 00 O. ND 0.48 157 O. pple
rough.
4177 1212 1.977 0.2540 ND ND 0.433 0.043 0.522 Do.
)-
4097 723 1.455 0.3648 ND ND 1.077 -0.072 0.447 Do.

~-<
4178 1213 1.892 0.2694 ND ND 0.498 -0.025 0.533 Do.
41 1214 38 o. ND 0.50 006 o.
41 1215 17 o. ND 0.44 009 o. a
4181 1216 0.453 0.7305 ND ND 1.029 0.535 0.380 Ripple "Z:f
crest. Cl
t""'
4182 1217 ND 1.0000 ND ND 0.484 0.292 0.585 Do. a
tz:l
41 1218 72 o. ND 1.02 106 O.
F:
Lamar 15 24 60 O. 1.400 0.70 199 o. ndward rJ)
lope. )-
Egypt Ayn Amur.El Kharea 4131 768 2.040 0.2432 ND ND 1.148 -0.090 0.398 Barchan
dune.
~
rJ)
A1 Wahat al Kharijah
Ayn Amu Kharea
4132
41
769
770
-0.112
09
1.0807
o.
ND
ND
ND 0.769
0.93
0.493
584
0.521
O. imbing ~
rJ)
une.
Wadi Ao a. Sina 41 766 85 o. ND 0.44 040 o.
Germany Kremberg-Berlin 1566 692 1.403 0.3789 1.380 0.3842 0.569 0.108 0.499
Greece Macedonia 577 696 3.203 0.1088 3.200 0.1088 0.395 0.047 0.517
India Hooghly River 651 1020 3.000 0.1233 3.020 0.1250 0.290 0.798 0.536
6 21 00 o. 3.220 0.38 079 o.
6 1022 80 o. 3.390 0.55 155 o.
Iran Alwaz 4162 1065 3.011 0.1241 ND ND 0.396 0.048 0.480 Produne(?).
Israel Tel Aviv 633 686 2.503 0.1768 2.500 0.1768 0.260 0.056 0.447
c
'<
Jordan Ram Fort 4128 765 1.434 0.3701 ND ND 0.599 0.206 0.518 Climbing
dune.
Ha
Li
S.A. ) Kilaue
Fezzan
.......
....... 6
6
17
714
30
20
O.
o.
1.620
2.320
0.48
0.40
143
003
O.
o. 1 dun
5 ft
( high. crest.
581 715 1.260 0.4176 1.000 0.5000 0.630 0.613 0.518 Do.
0- 636 716 2.420 0.1869 2.340 0.1975 0.520 0.223 0.519 Do.
 626 717 2.010 0.2483 2.000 0.2500 0.340 0.1170.541 Se1f dune,
45 ft high,
high on
slipface.
587 711 2.330 0.1989 2.140 0.2269 0.534 0.515 0.531 Do.
632 712 2.500 0.1768 2.500 0.1768 0.394 0.204 0.530 Do.
618 713 2.320 0.2003 2.400 0.1895 0.450 0.260 0.479 Do.
588 718 2.176 0.2207 2.150 0.2253 0.729 0.121 0.621 Seif dune,
45 ft high,
middle of
Slipface.
589 719 1.696 0.3078 1.700 0.3078 0.807 0.221 0.402 Do.
520 720 1.963 0.3099 2.180 0.2207 1.103 0.696 0.413 Do.
4094 721 2.097 0.2337 liD liD 0.700 0.013 0.499 Do.
617 721 2.070 0.2382 2.050 0.2415 0.880 0.030 0.487 Do.
;i
615 722 1.950 0.2588 1.950 0.2588
~~
0.458 0.116 0.512 Do.
638 724 1.810 0.2852 1.750 0.2973 0.650 0.194 0.479 Do.
4099 724 1.581 0.3343 liD lID 0.646 0.303 0.519 Do.
555 701 2.280 0.2059 2.250 0.2102 0.436 0.136 0.559 Do.
."
4080 708 2.003 0.2495 liD liD 0.592 0.069 0.516 Do. >
640
621
709
710
2.040
2.360
0.2432
0.1948
2.000 0.2500 0.740
2.380 0.1921 0.190
0.138 0.435 Do.
0.029 0.425 Do.
">
~
Libya--Con. Fezzan--Con. 590 725 2.060 0.2398 2.100 0.2333 0.861 0.266 0.491 Seif dune, ~
tz:I
45 ft high,
base of
slipface.
"
rJl
0
"Ij
4101 725 1.929 0.2626 liD liD 0.819 0.012 0.448 Do.
~
619
580
602
726
704
705
2.100
1.766
1.380
0.2333
0.2932
0.3842
2.050
1.600
1.800
0.2415
0.3299
0.2872
0.879
0.825
0.935
0.075 0.441 Do.
0.321 0.405 Do.
1.497 0.445 Do.
-~
t""

620
570
567
706
707
700
1.260
2.290
1.953
0.4175
0.2045
0.2588
1.080
2.510
2.080
0.4730 1.270
0.1756 1.024
0.2365 0.753
0.334
0.259
0.173
0.489
0.421
0.445
Do.
Do.
Do.
~~
521 723 2.133 0.2285 2.250 0.2102 0.849 0.201 0.542 Seif dune,
45 ft high,
swale, fill
on slipface.
4097 723 1.455 0.3648 ND liD 1.077 -0.072 0.447
0
Sabha 611 662 1.620 0.3253 1.600 0.3299 0.340 0.086 0.521 Seif dune,
5" Slipface.
;:::;:
N"
604 663 1.910 0.2661 1.810 0.2852 0.357 0.288 0.517 Seif dune,
CD windward
0.. slope.
D"
'< 612 665 1.820 0.2832 1.480 0.3585 0.820 0.540 0.456 Se!! dune,
75 ft high,
C) crest.
579 666 0.596 0.6598 0.500 0.7071 0.635 0.445 0.584 Seif dune,
0 base.
I!S
~
529 668 2.383 0.1921 2.800 0.1436 1.344 0.384 0.468 Seif dune.

~
rv
 ABLE6.-Gr es 0 f sand sam d dunes used a Qmtinued t
lity grain size M gram Sort- Skew-
Computer McKee's size ~ ness tosia
Remarks
Principal country
or area
Specific area Index No. Index No.
mm
mm P, Ski KG

Sedalea 608 673 2.070 0.2382 2.000 0.2500 0.890 0.281 0.459 Seif dune,
slipface.
614 0.2624 1. 0.576 0.14
ree Mountain 4065 0.1689 0.560 0.01 dune.
Harshall I jatto Islet 582 0.2535 1. 0.318 0.07
Mexico ihuahua 1550 0.1661 2. 0.365 0.07 t of
e.
H1ssouri(U verly 4004 0.1139 0.403 -0.03
New Hexico(U.S.A.) Santa Rosa 4003 4 1.357 0.3904 ND ND 0.832 0.234 0.519
White Sands National 4023 32 1.047 0.4840 ND ND 0.301 0.001 0.545
Monument.
586 825 1.636 0.3209 1.600 0.3299 0.308 0.187 0.476 Crest of

574 826 1.763 0.2952 1.700 0.3078 0.495 0.262
dune.
0.520 Windward
slope of
~-<
e. 0'"Ij
623 0.3368 1. 0.409 0.38 C'l
527 0.4175 1. 0.765 0.33 dune. t""'
565 0.3439 1. 0.467 0.02 0
III
575 0.3164 1. 0.442 0.26 bolic
t""'
e. til
532 850 1.400 0.3789 1.400 0.3789 0.930 0.038 0.551 Vicinity of
Z
star C
dunes(?).
603 851 1.610 0.3276 1.600 0.3299 0.900 0.041 0.476 Parabolic
dune.
~
til
549 971 1.270 0.4147 1.220 0.4293 0.083 0.008 0.532 Dome dune,
r base.
658 0.4147 1. 0.830 0.00
659 0.2238 2. 0.777 0.09
550 0.2238 2. 0.777 0.09
660 0.3869 1. 0.823 0.10
661 0.2643 1. 0.677 0.50
-.J
D
~ 645 975 1.100 0.4665 1.000 0.5000 0.780 0.240 0.497 Do.
! 644 976 1.603 0.3299 1.580 0.3345 0.890 0.076 0.514 Do.
643 977 2.700 0.1539 2.650 0.1593 0.720 0.130 0.512 Do.
:J
--;
592
591
977
979
1.490
1.766
0.3560
0.2932
1.320
1.720
0.4005
0.3035
0.944
0.515
0.326
0.306
0.501
0.604
Do.
Parabolic
e, near
st.
642 0.1416 1. 0.440 0.42
637 0.3276 1. 0.590 0.42 han
e, crest.
 662 981 1.700 0.2952 1.760 0.3078 0.292 0.0390.530 Transverse
dune, crest.
654 1069 1.310 0.3938 1.330 0.4033 0.255 0.190 0.507 Parabolic
dune, base
of nose.
655 1070 1.400 0.37ll 1.430 0.3789 0.350 0.221 0.530 Parabolic
dune, crest.
656 1071 1.630 0.4ll8 1.280 0.3231 0.470 0.294 0.520 Barchan
dune, base
of slipface.
657 1072 1.720 0.2932 1.770 0.3035 0.350 0.310 0.644 Barchan
dune, crest.
4168 1099 1.719 0.3038 ND ND 0.482 0.315 0.594 Do.
1623 llOO 1.566 0.3392 1.520 0.3487 0.429 0.221 0.564 Parabolic
dune, crest. o-i
557 0 1.663 0.3164 1.680 0.3121 0.633 0.055 0.514 tz:I

~
4187 0 1.462 0.3630 ND ND 0.495 0.427 0.532 Barchan
dune.
4188 0 2.255 0.2095 NO ND 0.870 0.164 0.463 Do. ::G
4189 0 2.278 0.2062 NO ND 0.858 0.180 0~464 Do. i!=
4190 0
652
2.005
2.966
0.2491
0.1285
ND
3.000
ND
0.1250
0.717
0.452
0.170 0.493 Do.
0.088 0.489 Parabolic
;g
Nevada (U.S.A.) Winnemucca 1564 ::G
dune, crest. >
630 652 2.870 0.1368 2.900 0.1340 0.380 0.005 0.526 Parabolic
dune.
~
Peru Loma de Corvina 4185 1226 2.133 0.2280 ND ND 0.524 0.054 0.518 ~
Otums 4033 249 -0.667 1.5878 ND ND 0.153 0.585 0.591 Ripples. ~
Paehaeamac 4184 1224 1.314 0.4022 ND ND 0.645 0.398 0.573 0
"r.I
0.320 0.051 0.487 Star dune,
Saudi Arabia Irq as Subay 601 984 2.670 0.1571 2.680 0.1560
crest. ~
t""
649
646
988
985
2.130
1.430
0.2285
0.37ll
1.910
1.040
0.2661
0.4863
0.800
1.180
0.450 0.470 Do.
0.450 0.415 Star dune,
s:Z
base. 0
tz:I
647 986 2.040 0.2432 1.950 0.2588 0.660 0.266 0.547 Do.
648
650
987
989
1.170
2.330
0.4444
0.1989
1.120
2.300
0.4601
0.2031
0.670
0.540
0.141 0.497 Do.
0.555 0.507 Star dune,
middle of
-ri!~
slipface.
600 993 -0.376 1.3014 0.910 1.8790 0.495 1.144 0.464 Star dune,
bench on
slipface.
0 Texas(U.S.A.) MIdland 4006 8 2.583 0.1669 ND ND 0.533 0.203 0.525
5" Monahans 4174 ll56 2.115 0.2308 NO ND 0.271 -0.008 0.486 Base of
;:::;:
N" slipface.
CD
0..
Oraoge 4001 1 1.584 0.3336 ND ND 0.511 -0.096 0.507
rr Signal Peak 551 3 2.640 0.1604 2.650 0.1593 0.193 0.057 0.520
'< 1549 3 2.083 0.2365 2.030 0.2449 0.448 0.190 0.521
C) Tunisia Bir Sultan 554 693 3.386 0.0954 3.360 0.0974 0.225 0.226 0.566 Seif dune,
middle of
0 Utah(U.S.A.) Great Salt Lake 4175 1180 1.650 0.3186 ND ND 0.422 0.116 0.536
slipface.

~
~
4186 1234 1.629 0.3233 ND ND 0.471 0.l41 0.500 ~

rv
 TABLE 6. - Graphical measures of sand samples from inland dunes used as a data base - Continued g

Locality Mean grain lize Median grain Sort- Skew- Kur-

Computer McKee'l lize ~ naa tOlia
Remarks
Principal country Specific area Index No. Index No. mm mm P, Sk, K'c
or area '" '"
ICauab 624 846 2.420 0.1869 2.400 0.1895 0.240 0.193 0.502 Base of
slipface.
631 847 2.590 0.1661 2.560 0.1696 0.390 0.202 0.487 Crest of
dune, 20 ft
high.
571 848 2.476 0.1792 2.390 0.1908 0.356 0.400 0.485 Windward
slope of
dune.
625 848 1.440 0.3686 1.390 0.3816 0.287 0.332 0.514 Do.
Lund 1614 243 3.276 0.1037 3.280 0.1029 0.468 0.019 0.575 Do.
1615 245 2.976 0.1276 2.980 0.1267 0.492 0.019 0.593 Slipface of >
~
dune.
Monument Valley 4172 1106 2.533 0.1728 ND ND 0.727 -0.035 0.495 Hanging
dune. -<
Mount Carmel Junction 1618 571 2.316 0.2017 2.2'0 0.2102 0.555 0.238 0.495 Produne(?), 0
slipface. "13

San Rafael Swell

583
4170
571
1101
1.993
2.394
0.2517
0.1903
1.950 0.2588
ND ND
0.283
0.644
0.272
0.444
0.545 Do.
0.521 Barchan
dune, crest. "
S
=
4171 1102 2.390 0.1908 ND ND 0.650 0.476 0.530 Barchan ~t/)
dune, base >
of slipface.
~
t/)

Interdune and serb sands ~

t/)

TABLE 7. - Graphical measures of sand samples from interdune and serir deposits used as a data base

o [Localities are listed alphabetically by country, and specific area where known. The computer index number and McKee's index number
5" are also listed. ND indicates no data)
;:::;:
N"
CD
0. Locality Mean grain size Median grain Sort- Skew- Kurt-
rr
'<

C)
o
Principal country or
area
Specific area Computer McKee's
Index No. Index
No.
phi DIll phi
size

.. ing
rh;-I
ness

SkI
osis
"lPG

~
~
Algeria
Egypt
Ouarsla
Ayn Amur
4018
4130
26
767
1.935
-0.683
0.2615
1.6055
ND
ND
ND
ND
1.560 -0.471
0.179 0.702
0.404
0.668
rv Jordan Ram Fort 594 685 1.403 0.3789 1.200 0.4353 1.177 0.318 0.489
 Lybia rezzan 622 702 2.190 0.2192 2.200 0.2176 0.812 0.015 0.507
639 703 2.350 0.1961 2.490 0.1780 0.710 0.240 0.492

556 727 2.370 0.1934 2.350 0.1961 0.971 0.006 0.422

616 728 1.570 0.3368 1.290 0.4090 1.180 0.316 0.421
4106 729 1.754 0.2965 ND ND 0.913 0.117 0.439
4107 729 1.578 0.3349 ND ND 0.936 0.425 0.462
595 730 2.333 0.1989 3.000 0.1250 1.630 0.208 0.371

4109 730 2.279 0.2060 ND ND 1.569 -0.758 0.334

597 731 2.213 0.2161 2.000 0.2500 1.456 0.169 0.360
4111 731 1.684 0.3112 ND ND 1.392 0.541 0.361
561 732 2.003 0.2500 1.700 0.3078 1.420 0.261 0.366
573 733 1.839 0.2793 1.400 0.3789 1.253 0.435 0.382

585 734 1.450 0.3660 0.700 0.6156 1.348 0.860 0.441 ~

~
4115 734 1.155 0.4491 NO ND 1.170 0.725 0.731
523 735 2.886 0.1350 3.680 0.1560 1.305 0.641 0.486
4117
564
735
736
2.412
1.563
0.1879
0.3392
ND
1.400
ND
0.3789
1.408
0.848
-0.565
0.087
0.362
0.346 ";..
t""

524 737 2.670 0.1571

0.4263
3.400
0.350
0.0947
0.7846
1.386
1.616
0.667
0.540
0.356
0.668
~
~
598 738 1.230
4121 739 2.468 0.1807 ND ND 1.397 -0.748 0.363
4122
4123
739
740
2.044
1.677
0.2425
0.3127
NO
ND
ND
ND
1.472 -0.294
1.553 0.067
0.343
0.377 ~
~
4124 743 0.961 0.5137 ND ND 1.464 0.273 0.459 0
~
599 744 1.560 0.3392 0.800 0.5743 1.671 0.497 0.333
519 745 2.450 0.1830 2.850 0.1387 1.390 0.369 0.342 ~
Ghat
576
559
746
675
2.000
2.006
0.2500
0.2483
1.900
2.620
0.2679
0.1627
1.417
1.636
0.082
0.448
0.351
0.368 -
t""
;..
Z
C
664 1.553 0.3415 1.550 0.3415 0.940 0.146 0.505

~
Sabha 553
525 667 2.300 0.2031 2.950 0.1294 1.588 0.515 0.434
Sedales 4061 672 1.769 0.2934 NO ND 1.431 -0.080 0.462
Three Hountains 596 680 2.236 0.2117 2.000 0.2500 1.164 0.246 0.391
526 681 2.210 0.2161 2.150 0.2253 1.113 0.068 0.370

New Hexico(U.S.A.) White Sands National Honument 4137 828 2.047 0.2420 NO ND 1.035 0.126 0.510
4191 0 2.702 0.1537 ND ND 0.820 -0.113 0.460
Peru Pampe Corre Viento, Lomitas 4031 247 -0.093 1.0666 ND ND 0.694 0.701 0.378
0 4032 248 0.787 0.5795 NO ND 1.411 0.251 0.387
5" 0.941 0.056 0.405
;:::;: Tunisia Bir Sultan 1621 693 2.416 0.1882 2.350 0.1961
N"
CD
0..
rr
'<

C)
0
~
....en
~
rv
Digitized by Google
 A STUDY OF GLOBAL SAND SEAS

DIFFERENTIATION OF EOLIAN
DEPOSITS BY DISCRIMINANT
ANALYSIS
ChapterC

By R. J. MOIOLAt and A. B. SPENCERt

Contents
Page
Summary of conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Introduction ...................................... 55
Discriminant analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Application ....................................... 55
Results ........................................... 57

, Mobil Research and Development Corp., P.O. Box 900, Dallas, Texas 75221.

53

Digitized by Google
54 A STUDY OF GLOBAL SAND SEAS

Illustrations
FIGURE 28. Discriminant frequency plots and cumulative distribution curves of beach,
river, coastal dune, and inland dune samples .......................... 56
29. Discriminant frequency plot of 152 inland dune and 291 coastal dune samples 57
30. Discriminant frequency plot of 31 interdune and 38 inland dune samples. . . 56

Digitized by Google
 DIFFERENTIATION OF EOLIAN DEPOSITS BY DISCRIMINANT ANALYSIS 55

Summary of Conclusions parameters could not differentiate beach and

coastal dune sands and could only poorly differen-
t?r LTHOUGH IT WAS NECESSARY to use tex- tiate river and inland dune sands.
,a. tural parameters (mean, standard deviation, Recent studies by Sahu (1964) , Sevon (1966) ,
skewness, kurtosis) -which in themselves are an Greenwood (1969), Tillman (1971), Moiola,
abridgment of the grain-size distribution - as input Spencer, and McKee (1973), and Moiola, Spencer,
variables, inland and coastal eolian deposits were and Weiser (1974) suggest that discriminant
effectively differentiated by discriminant analysis. analysis is probably the most effective grain-size
Previous studies have demonstrated that the weight technique for differentiating modem sand bodies.
fractions of quarter phi grain-size analyses are the In this study, the discriminant technique is used to
preferred input for discriminant analysis (Moiola, differentiate inland dune, coastal dune, and inter-
Spencer, and Weiser, 1974). Therefore, if quarter dune deposits.
phi data were available, even better separation of
these deposits could be expected. This conclusion is Discriminant Analysis
supported by the excellent separation obtained for
the inland dune and interdune deposits of Fezzan, DISCRIMINANT ANALYSIS is a classifying
Libya, using quarter phi grain-size data. technique that differentiates distinct populations
These results confirm the contention of Moiola, by using a number of independent attributes. It was
Spencer, and McKee (1973) and Moiola, Spencer, developed by Fisher (1936) to solve a taxonomic
and Weiser (1974) that discriminant analysis is problem. The method uses a classifying function to
perhaps the most effective method for differentiat- assign samples individually to one of two or more
ing modem eolian sands by their textural charac- populations. A priori knowledge that the popula-
teristics. tions differ is a prerequisite for its use. The tech-
nique defines the coefficients of a function, which
Introduction allows samples described by a number of indepen-
dent variables to be projected onto a line so that the
VARIOUS GRAIN-SIZE TECHNIQUES are distance between projected populations is max-
available for differentiating sand bodies (pettijohn, imized relative to the scatter within each projected
Potter, and Siever, 1972). These techniques are pri- population.
marily attempts to classify and assign sets of The detailed mathematical treatment of the tech-
unknown samples to one of two or more popula- nique is not presented here but can be found in a
tions. Mathematically, they have become progres- number of standard references, such as Wilks
sively more sophisticated, beginning with simple (1962), Miller and Kahn (1962), and Krumbein and
visual comparison of histograms, to plotting tex- Graybill (1965). Its application to grain-size data
tural parameters, to graphical analysis of cumu- has been amplified by Moiola, Spencer, and Weiser
lative curves, and finally to such techniques as dis- (1974). The independent variables can be the
criminant analysis. weight fractions of each size class, textural
Perhaps the most popular approach to date has parameters, or other measurable attributes. Both
been the use of textural parameters plotted as weight fractions of quarter phi grain-size analyses
variables on a two-dimensional coordinate system and textural parameters calculated from sieve data
(Mason and Folk, 1958; Friedman, 1961, 1967). Un- according to the scheme of Folk and Ward (1957)
fortunately, this use of textural parameters has met are used in this study.
with only partial success (Shepard and Young,
1961; Schlee, Uchupi, and Trumbull, 1964; Gees,
1965; Sevon, 1966; Moiola and Weiser, 1968). For
Application
example, Moiola and Weiser (1968) demonstrated THE EFFECTIVENESS of linear discriminant
that certain combinations of textural parameters, analysis for differentiating modem beach, river, in-
such as mean diameter and skewness, can be used land dune, and coastal dune sands has been
to differentiate certain sand deposits, but textural demonstrated by Moiola, Spencer, and Weiser

Digitized by Google
56 A STUDY OF GLOBAL SAND SEAS

100 T""-----:01"""'"----~,.-.,
t""'"MlXiml;lm /'
80 I separetion I
.... I 97 percent ,
~ 60
l I'
ffi
A-
40 : ti,'
I ~~I
20 I \(\'<J."/
I ,,'"
O+-~~~~~~-.-'-'r-r-'-'
o 20 40 60 80 100 120 20 40 60 80 100 120
CUMULATIVE DISTRIBUTION CURVE CUMULATIVE DISTRIBUTION CURVE
Beach mean,

Jhir:
Beech meen ij
6 If II" i i '" i i i .I~U" ~ I iii , i I

I
40 I

.I'.~nl
80
I
120
'
?" I ~4f>~ :r.,'
I I n 'l,it,.~'F
1. I I

~ Inlend dune meen Coutal dune meen

FREQUENCY PLOT FROM WHOLE PHI DATA FREQUENCY PLOT FROM WHOLE PHI DATA

A B

100
Meximum
, 100 ,....---------:::::;:111_-~--.,
I'
iseperetion ,I

!zw
80

60
I 93 percent I
I Gl
....Z
80

60 I
1/
a,
u I :if
ic'l ~ -I
I
~ : I,
ffi
A-
40

20
:
I
,I II:
~
40

20
i Meximl;lm I J!!,
separetlOn : c3j

0
I
I. ,
,
" O+-~~~-"--~~~-r~~.-,....~
93 percent-1 I

0 20 40 60 eo .100 120 o 20 40 60 80 100 120

CUMULATIVE DISTRIBUTION CURVE CUMULATIVE DISTRIBUTION CURVE

., -
Beach mean.

, 4,
I4'0:, ,
II I I ... ,.,.~
.Inlend dune meen

I."
"
r
o,
i iii
80
i i
I~O 40 I I ' 80 ' lie>
I I , II ~I tll\l~ , I ' , . ." , '
River mean
L Coastel dune mean
FREQUENCY PLOT FROM WHOLE PHI DATA FREQUENCY PLOT FROM WHOLE PHI DATA

c D

100

80 I
, ",'
80
.... I
Z
w 60 ~I !z 60
u
II:
I .J'I ~
~..,
W
A-
40 I
~ 40
t-,lMeximum
20 20
.../' seperetion
., 92 percent
0
0 20 40 60 80 100 120 20 40 60 80 100 120
CUMULATIVE DISTRIBUTION CURVE CUMULATIVE DISTRIBUTION CURVE
..
,J~ U Ii
River meen

elF .1.
o' , '40 \ ~ , 120 40':1180'
iii_ I ' I~O
,
'I' .""IPI~ 1'1
Inland dune meen J
' "
, , ' I

Coestel
I 'I'
,1I1I11;r:
du~e meen
~

FREQUENCY PLOT FROM WHOLE PHI DATA FREQUENCY PLOT FROM WHOLE PHI DATA

E F
DISCRIMINANT FREQUENCY plots and cumulative distribution curves of beach, river, coastal dune, and inland dune
samples (modified from Moiola, Spencer, and Weiser, 1974). Scale of 0 -120 is arbitrary. (Fig. 28.)

Digitized by Google
 DIFFERENTIATION OF EOUAN DEPOSITS BY DISCRIMINANT ANALYSIS 57
(1974). In their study, using whole phi analyses of samples from Fezzan, Libya, for which quarter phi
sets of 30 samples from each environment, the weight fractions were available, was used to deter-
differentiation obtained between pairs of sets was mine whether associated inland dune and inter-
no less than 111 percent (fig. 28). They also dune deposits could be differentiated.
demonstrated that linear discriminant analysis
using whole phi data is at least as effective as the
best combination of textural parameters based on Results
quarter phi data in differentiating sands from these
BIVARIANT TEXTURAL PARAMETER plots
environments. In addition, the technique suc-
have not proven particularly effective in separating
cessfully differentiated beach and coastal sands, as
environments in general. The parameters them-
well as river and inland dune sands, whereas tex-
selves, when used as descriptive variables in dis-
tural parameter plots did not.
criminant analysis, have been found to be sensitive
In this study various eolian deposits are differen-
descriptors (Sevon, 1966). Results of the present
tiated. A selected suite of 152 inland dune, 291
study support this contention.
coastal dune, and 38 interdune samples, compiled
by Ahlbrandt (chapter B), is the data base. The dis- Using Folk and Ward's (1957) approximations of
criminant function used is of the following form: the mean, standard deviation, skewness, and kur-
tosis of the grain -size distribution as input for linear
discriminant analysis, sets of inland dune and
coastal dune samples can be easily differentiated
where z is the transformed sample value; C 1, C I (fig. 29). The population of projected coastal dune
cn are the coefficients of the discriminant function; samples tightly clusters around its mean at 80 on
x l' XI Xn are the descriptive variables; and b pro- the arbitrarily chosen scale of 0 -120. Although the
vides scaling so that the means of the two projected population of projected inland dunes is loosely dis-
populations fall at 40 and 80 on an arbitrarily persed about its mean at 40, maximum separation
chosen scale of 0 -120. (91 percent) occurs at approximately 69 on the
Because weight fractions of each size class, scale, with only 24 coastal dune samples and 13 in-
which are ordinarily used, were not available for land dune samples from the total population of 443
the set of coastal dunes, it was necessary to find falling outside their fields. Comparison of the rela-
alternative variables for discriminant analysis. Tex- tive dispersions of the coastal and inland dune
tural parameters were selected. A sub-set of eolian populations suggests that the textural character of

Maximum separation
of 91 percent

r'
COASTAL DUNE

o 10 20 30 40 50 60 70 SO 90 100 110 120

I , I I I

"I f ',III, "f jl'll '1'1'1 11 ' 'lfljl'II'jljljljljlj /111 'III' j I",', "f'l
J I I I I I

INLAND DUNE

DISCRIMINANT FREQUENCY plot of 152 inland dune and 291 coastal dune samples. Inland dune samples indicated below
abscissa; coastal dune samples indicated above abscissa. Maximum separation of 91 percent occurs at 69 on the arbitrarily chosen
scale of 0-120.(Fig. 29.)

Digitized by Google
58 AS OFGLO ANDS

Maxi ration

~
IN NE nt
r ,.........., ,,, I
, I
, , ~ , I
,
0 10 20 30 40 50 60 70 80 100
,, , , , , , ,, , , , , , , , , , , , ,I ,
~I~I~I il~
I ,I ~ ~Ii , I ~ I
~ ~ ~ ~ ~I '
I
I I I

DISCRIMINANT FREQUENCY plot of 31 interdune and 38 inland dune samples. Interdune samples indicated below abscissa; inland
du s m les indicated above abscissa Maximu ation of 97 ent occ t 65 on the b't ily chos I of 0 -120
(Fi

inland dune sands is more variable than that of dune, is shown in figure 30. Of the 69 samples in
co t I dunes. See Ahlbr dt chapt B This the combined set of dune and interd sampl
m ect th varied e of in eolian Fezza ya, onl interd mple 0
de and ne sam ut of I outsi
The effectiveness of the linear discriminant tech- their respective fields. Thus, f17 percent of the sam-
nique in differentiating the intimately associated ples are successfully differentiated with maximum
eo . ubenvi nts, s dune inter- ation ing at the sea

:"VU I.J]
~cy
Digitized by Google
Digitized by Google
 A STUDY OF GLOBAL SAND SEAS

RED COLOR IN DUNE SAND

Chapter D

By THEODORE R. WALKER2

Contents
Page
Summary of conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Introduction ...................................... 63
Acknowledgments ............................. 64
Regional setting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Moisture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Winds and eolian transport ..................... 65
Bedrock...................................... 65
Soils.......................................... 65
Distribution of red sands ....................... 67
Characteristics of the pigment ....................... 68
Sources of iron .................................... 69
Origin of clay coatings on eolian sand grains . . . . . . . . . . 69
Conditions favoring formation of iron oxide pigment . . . 78
Factors affecting regional and local variations in redness 79

University of Colorado, Boulder, Colorado 80309.

61

Digitized by Google
 A STUDY OF GLOBAL SAND SEAS
62

Illustrations
FIGURE 31. Gemini XI space photograph of libya...................................... 64
32. Generalized geologic map of northwestern libya. . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
33. SEM photomicrographs of quartz grain coated by film of reddish-yellow pigment,
and X-ray traces showing elemental composition of analyzed areas ......... 70
34. SEM photomicrographs of quartz grain with red pigment concentrated in indenta-
tions, and X-ray traces showing elemental composition of analyzed areas . . . . 71
35. Photograph of etched hornblende grain from modem dune . . . . . . . . . . . . . . . . . . 72
36. Photograph of etched augite grain from modern dune . . . . . . . . . . . . . . . . . . . . . . . 72
37. SEM photomicrographs of matrix-rich sand from weakly stabilized reddish dune,
northern coastal region, libya, and X-ray traces showing elemental composition
of matrix clay. . . . . .. . . . . . .. . . . . . .. . . .. . . .. . . . . . . . . . . . . . . . .. . . . . .. . . . . . 74
38. SEM photomicrographs of quartz grain with reddish-brown pigment in indenta-
tions, and X-ray traces showing elemental composition of pigmenting material 75
39. SEM photomicrographs of quartz grain collected from pink modern dune, and X-
ray traces showing elemental composition of the pigmental material in fine and
very fine sand ........................................................ 76
40. SEM photomicrographs of quartz grain collected from red dune, and X-ray traces
showing elemental composition of pigmenting material of surface clay and iron
oxide................................................................ 77

TABLE
Table
8. Oay mineral analyses ....................................................
...
67

Digitized by Google
 RED COLOR IN DUNE SAND 63

Summary of Conclusions Each step in the sequence of events leading to the

accumulation of the stain and its gradual reddening
rlrZ OLIAN DEPOSITS in western Libya provide can be documented in deposits that are forming in
~ an outstanding example of modem dune the desert today. This important fact indicates that
sands that have reddened with age and with dis- the reddening is a product of the modem desert en-
tance of transport. The pigment in the dunes is vironment, not the product of an alleged earlier
chemically precipitated iron oxide, which is moist climate.
derived in part from alteration of ferromagnesian
silicate minerals and in part from iron-bearing clay Introduction
minerals that adhere to the surfaces of sand grains.
The clay minerals of the surface coatings are A COMMONLY OBSERVED feature of deserts is
derived mainly from airborne dust mechanically in- that dune sands have a reddish hue and become in-
filtrated into the dunes and into other surficial creasingly redder with distance of transport and
deposits from which the dune sands are derived. with increasing age (Norris, 1969). The origin and
To the extent that iron-bearing minerals are not climatic significance of the pigment in such dunes
in equilibrium with meteoric water, they react with has been a subject of considerable controversy
dew, rain, and interstitial water, and release iron, among geologists for many decades. What is the
which, owing to the oxidizing and alkaline chemi- nature of the pigmenting material and how and
cal environment, precipitates on the surfaces of the when is it accumulated on the grains? Does the pig-
sand grains as yellowish ferric hydrate (limonite). ment reflect the reworking of sands which
Subsequently, as these grains are abraded during developed their color during earlier, much more
eolian transport, the limonite and limonite-stained moist climates than those existing now (as for ex-
clay are removed from exposed surfaces of coarser ample, during Pleistocene pluvial intervals), or is it
grains, but these materials are retained in indenta- the product of processes that are going on at present
tions where they are protected from abrasion. The in deserts? Is the intensity of redness a reliable
fine grains are abraded much more slowly than the measure of the age of dunes, or are factors other
coarse ones and therefore tend to have more inden- than age equally important in the reddening?
tations in which to preserve the pigment. Thus, the Hopefully, the evidence presented in this chapter
finer grains usually are more heavily stained than will help to answer these questions.
are the coarser grains. Conditions that favor the ac- Excellent examples of dune sands that show red-
quisition of additional iron oxide and clay coatings dening in the direction of transport are the ~apra'
on the surface of grains are likely to occur Awhirl and ~ra' Marziiq of western Libya (fig.
whenever the grains are at rest. 31). In the au\h~r's opinion, these and other dune
The light-yellow ferric hydrate is metastable, sands in Libya convincingly document successive
and, upon aging under the chemical conditions ex- stages in the development of red pigment. and they
isting in this desert region, it ultimately converts to are believed to form models that illustrate the proc-
hematite. The aging of ferric hydrate is accom- esses of dune-reddening active today in hot desert
panied by reddening; hence, the sands become red- regions throughout the world.
der with increasing age and distance of transport. The evidence presented here supports the con-
Although aging is prerequisite to the reddening of clusions of Norris (1969) that much of the iron in
the dune sands, the time interval involved is also the pigment is derived internally from the dunes
affected by such factors as the amount of moisture and that processes leading to the reddening are ac-
available, availability of iron, and amount and type tive in modern deserts. The evidence also supports
of clay minerals retained on grains. The effects of the conclusion of Folk (1969) that airborne dust
these factors are regionally variable; therefore, plays an important role in the reddening of dunes.
although the dunes redden with increasing age, the The dunes studied by Folk. in the Simpson Desert
intensity of redness is not a reliable measure of the of Australia. however. are considered by him to
age of the deposits, and it is not a reliable criterion have attained redness from dust that initially was
for correlation. red and was derived from red lateritic soils

Digitized by Google
 64 A STUDY OF GLOBAL SAND SEAS

GEMINI XI SPACE PHOTOGRAPH of libya. Northeastward view from above eastern Algeria. Prevailing wind is from northeast (up-
per right). Note marked increase in redness of ~~ra' Awbiriand a~ra' Marzuq in direction of sand transport. (Fig. 31.)

developed under a former, more humid climate. In Acknowledgments

contrast, evidence presented here indicates that
dust involved in the reddening of dunes in the
Sabara initially was not red, but has become red in This research was supported by National Science
the existing arid climate. Foundation Grant GA -20738 and the University of

Digitized by Google
 RED COLOR IN DUNE SAND 65

Colorado Council on Research and Creative Work. ~ri' Marziiq) are composed of sand derived prin-
John C. Harms (Marathon Oil Co.) and James R. cip811y from sources lying northeastward toward
Steidtmann (University of Wyoming) critically the Gulf of Sidra and in northeastern Libya (figs. 31,
reviewed the manuscript. A. J. Crone provided 32). In addition, the ~ri' Awbiri, particularly its
valuable laboratory assistance. All of this aid is northern arm, probalJly receives a significant con-
gratefully acknowledged. tribution of sand blown southwestward from the
coastal steppe region across the Al l;Iamidahal
tIamra plateau (figs. 31, 32).
Regional Setting
Moisture Bedrock
Bedrock throughout Libya consists dominantly of
WITH THE EXCEPTION of a narrow belt of arid non-red sedimentary rocks of Paleozoic to Tertiary
steppe climate along the Mediterranean coast, age overlain locally by Tertiary volcanic rocks (fig.
western Libya is mostly a region of severe desert 32). All geologic periods are represented by the
(fig. 32). Average annual rainfall along the coast sedimentary rocks, but those rocks of Cretaceous
westward from the Gulf of Sidra is about 200 mm (8 and Tertiary age are the most important insofar as
in.), and the amount diminishes markedly within a the present discussion is concerned. They are
short distance inland. Along the southern margin of areally dominant, and they provide the major
the Gulf of Sidra the desert begins almost at the bedrock sources of the eolian deposits. The Cre-
coastline (fig. 32). The central part of the country taceous and Tertiary sedimentary rocks consist
lies in the driest part of the Sahara, where, at timesmainly of marine limestones and shales but locally
several years may pass with no rainfall whatever. contain interbedded continental deposits of
However, nearly all areas - even those in the sandstone and shale (Conant and Goudarzi, 1964).
driest parts of the desert - receive some rainfall The Cretaceous and Tertiary sandstones are
(Furst, 1970) , and even seemingly negligible mostly orthoquartzites and probably provide a ma-
amounts become important geologically because
jor source of the quartz sand which dominates the
over long periods they can provide enough
dunes. Cretaceous and Tertiary limestone, shales,
moisture to cause mineral alterations that result in
and volcanic rocks provide sources of sand-size
reddening. In addition, diurnal temperature
rock fragments, but grains of these lithologies in
changes commonly are extreme in the Sahara, and
most places constitute only a very small percentage
they lead to frequent occurrences of heavy dew.
of the dune sands.
Although apparently insignificant, such small Tertiary volcanic rocks are the probable major
amounts of moisture, when replenished over and sources of unstable ferromagnesian silicate
over again, may play an important role in the red- minerals, such as augite, hornblende, and epidote,
dening processes described herein. The importance which occur in the heavy-mineral fractions of dune
of dew as an agent of mineral alteration in deserts sands at all the sampled localities. In addition,
has been stressed by many previous writers (among shales, sandstones, and, to a lesser degree,
them, Engel and Sharp, 1958; Kuenen and Perdok, limestones are the ultimate sources of airborne
1962; Margolis and Krinsley, 1971). dust, which, as will be shown in the subsequent
discussion, is a significant factor in the develop-
ment of red pigment in the dune sands. The
Winds and Eolian Transport bedrock formations include some red beds of
Mesozoic age, but these beds are of limited occur-
Throughout most of Libya prevailing winds blow rence and are relatively unimportant as sources of
from the northeast except near the coast, where pigment in the dune sands.
westerlies prevail during the winter months. The
major direction of eolian transport into the interior Soils
of the desert, therefore, is southwestward. Accord-
ingly, the major sand seas that dom~~ate th~ ~and In most places bedrock is overlain by a venee~ of
scape in southwestern Libya (~ra AWhirl and surficial deposits, which are mainly light red eohan

Digitized by Google
66 A STUDY OF GLOBAL SAND SEAS

15 20"

MEDITERRANEAN SEA

GUU=OF
SIDRA Ts

Os

EXPLANAnON
Geology
Os Ouaternary surficial deposits
Ts Tertiary sedimentary rocks
Tv Tertiary volcanic rocks
Mrs Mesozoic sedimentary rocks

25 Pis Paleozoic sedimentary rocks

per Precambrian crystalline rocks
Sample types
Ma Modern alluvium
Md Modern dune sand
Qd Quaternary(?) dune sand
Qa Quaternary(?) alluvium
B Bedrock
S Soil
67 Sample number
Ts
o 100 KlOMETRES
~ ',' " " 100 MILES

GENERALIZED GEOLOGIC MAP of western Libya, showing sample localities. Geology generalized from Conant and Goudarzi
(1964). (Fig. 32.)

sands. Where migration of these deposits has been great soil group known as "Red Desert Soils" and,
temporarily halted either by vegetation, as in many as evidenced by the following characteristics, they
places in the coastal region, or by the development attest to long-term regional aridity.
of protective gravel pavements, as on deflation sur- The soils are calcareous throughout their
faces in the interior of the desert, and on the upland profiles, and in places near the coast they contain
surfaces of dissected alluvial fans, weakly prominent concentrations of carbonate deposits
developed but strikingly reddish (2.5 YR to 5 YR, (caliches) which commonly are a metre (3 ft) or
Munsell Soil Color Chart, Munsell Color Co., 1954) more thick. Somewhat farther inland, where aridity
soils have been formed. These soils belong to the is greater (as in the area beginning about 75 km (47

Digitized by Google
 RED COLOR IN DUNE SAND 67

mil south of Misritah and extending to Qaryat Abu TABLE 8 - Continued

Nujaym), calicl~e zones are composed mainly of
gypsum. Gypsum caliches also occur nearby in the Sample Clay mineral I Sample Clay mineral

desert regions of Tunisia (Page, 1972). In most Solla

places the soils also contain sand- and pebble-sized 5 I. K. ml-im 63 P. K. i

6 I. K. ml-im 65 p. K. I. ml-im
rock fragments of limestone, and in the coastal 9 I. K. ml-im 76 K. I. ml-im
region they typically contain grains of carbonate 13 I. K. ml-im 77 K. I. ml-im
shell debris. lab I.K. ml-im 80 I. K. ml-im
22 I.K. ml-im 81 P. K. i. ml-im
23 I.K. ml-im 82 p. K. i. ml-im
The occurrence of carbonate and gypsum caliche, 28 I.K. ml-im. p 88 K. p. i. ml-im
and the presence of carbonate clasts, imply an ab- 29 p. K. i. ml-im 89 I. K. ml-im
sence of significant leaching during soil formation. 30 p. K. i. ml-im 92 I. K. ml-im
Moreover, soils in the more arid regions commonly 35 P. I. K. ml-im 95 I. K. ml-im
37 P. I. K. ml-im 97 I. K. ml-im
contain palygorskite as a major constituent of the 38 p. K. I. ml-im 99 I. K. ml-im
clay fraction (table 8). This clay mineral requires an
alkaline environment for development and sur- 52 I. K. ml-im 107 I. K. ml-im
59 I. K. ml-!m. ~J 112 I. K. ml-im
vival. The importance of all these characteristics 61 I. K. ml-lm. 113 I. K. ml-im
can hardly be over-emphasized because they indi- 62 I. K. ml-im
Modem alluvium
cate that the soils formed under nonleaching,
4 I. K. ml-im 49 P. I. K. ml-im
alkaline conditions. The nature of the soils provides 7 I. K. ml-im 50 I. K. ML-IM
evidence that for at least as long as the age of these 17 I. K. ml-im 51 I. K. ML-IM
soils, which presumably extends from early 18a I. K. ml-im 54 I. K. ml-im
Pleistocene to Holocene, the climate in this part of 20 I. K. ml-im 57 I. K. ML-IM
the desert has not been significantly more moist 21 I. K. ml-im 66 I. K. ml-im
25 I. K. ml-im 66 K. I. ml-im
than it is at present. Accordingly, the charac-
teristics of the soils provide strong evidence that 31 I. K. ml-im. p 87 I. K. ML-IM
32 I. K. ml-im 93 ML-IM. I. K
the associated reddish pigment, is the product of an 36 I. K. ml-im. (p) 94 I. K. ML-IM
arid climate. 41 I. K. ml-im 98 I. K. ML-IM
Modem dune I8IId
8 K. I. ml-im 100 I. K. ml-im
33 I. P. K. ml-im 105 K. I. ml-im
TABLE 8. - Clay minerals identified in samples analyzed for 44 I. K. ml-im 106 K. I. ml-im
55 I. K. ml-im 672 K. I. ML-IM
this study 71 K. i. ml-im. (p) 674 K. I. ML-IM
ISample localities are shown in rJ8Ule 31. Clay minerals are listed in order of QuaternBl'Y(?) dune I8IId
ilecre8s1ng abundance. Capital letter symbols indicate major constituent; lower 15 I. K. ml-im 111 I. K. ml-im
cae letter symbols indicate minor constituent; parentheses. trace amount. K.
kaolinite; I. illite; ML1M mixed.layer illite montmorillonite; P. palygonkite) 19 I. K. ml-im 114 I. K. ml-im
69 P. K. I. ML-IM
QuaternBl'Y(?) alluvium
Sample Clay minerala Age and rock type
Bedrock
27 K. I. p. ml-im 84 K. I. ml-im
83 K. I. ML-IM 101 I. K. ml-im
1 I. mIim Triassic shale
2 K. I. ml-im Cretaceous shale
10 I. K. ML-IM 00.
11 K. I. ml-im 00. Distribution of Red Sands
26 ML-IM. I. K Miocene shale

53 I. K. ml-im Tertiary clay Eolian sands that show pronounced reddening in

58 K. ML-IM. i Cretaceous shale the direction of transport occur in two parts of
61 K. i. ml-im Cretaceous sandstone
68 K. P. I. ml-im Tertiary sandstone western Libya. One is in the belt of arid steppe cli-
72 K. ml-im Cretaceous claystone mate that extends inland for several tens of
73 K. I. ml-im Cretaceous sandstone kilometres along the Mediterranean coast, west of
74 K. I. ml-im Cretaceous shale the Gulf of Sidra (fig. 32). Here, sand is blown land-
78 K. i. ml-im Cretaceous sandstone ward from the beach and piled into dunes which.
65 K. i Cretaceous co~lomerate
104 I. K. mI-im Tertiary(?) mu stone through time. migrate further inland. The beach

Digitized by Google
68 A STUDY OF GLOBAL SAND SEAS

sands and dunes adjacent to the beach are not even teristically is the reddest. Moreover, fine and very
slightly red, but the dunes become progressively fine grains commonly are more heavily stained and
redder inland, and within a few kilometres of the redder than are coarser grains.
coast they have reddish hues ranging from 5YR to a
maximum redness of about 2.5YR, values from 5 to The mineralogy, texture, and chemical composi-
7, and chromas from 4 to 8, as determined by com- tion of both the thin films and thicker concentra-
parisons with the Munsell Soil Color Chart (Mun- tions of pigment were determined by combining X-
sell Color Co., 1954). The most commonly occun1ng ray-diffraction techniques with scanning electron
colors are designated as "reddish yellow" and microscope (SEM) and energy-dispersive X-ray
"yellowish red" on that chart. analysis. Representative samples of dunes showing
different degrees of redness were treated in the
following manner. Between 10 and 20 grams of
A second region of red dunes is that occupied by sand were split from the bulk sample, and this por-
the sand seas called a!lri' Awhirr and ~ri' tion was washed and treated in an ultrasonic
Marziiq in the intensely arid interior of the oesert cleaner until enough of the pigmenting material
(figs. 31, 32). Most of the sand in these two dune was available for X-ray analysis. Untreated por-
fields has been transported from sources that lie tions of the bulk sample were examined under a
several hundred kilometres to the northeast in the binocular microscope; several grains which showed
very arid coastal region south of the Gulf of Sidra. typical staining characteristics were selected from
These sands, like those in the northern coastal each sample, and the grains were mounted on SEM
region of western Libya, redden progressively in
stubs.
the direction of transport, but, because of the
greater aridity of this region, the distance of After mounting the grains on stubs - but prior to
transport required to achieve a comparable degree coating them with a conductive film of gold and
of redness is much greater here. paladium - each grain was again examined under
the binocular microscope, and a sketch was drawn,
The full range of color change in the interior showing the distribution of concentrations of the
deposits was not studied because those sands were pigmenting material. These sketches were used
not traced to their source area. However, in the area during the SEM examination as maps for locating
between Waddan and Sabhii (figs. 31, 32), which areas on the grains that would be particularly suita-
lies at the upwind end of the sampled area, the col- ble for detailed study. All of the mounted grains
or of the sands range from about 8YR 7/4 (pink) to from each sample were then examined using a
about 6.5YR 6/6 (reddish yellow), whereas near Al Cambridge Stereoscan SEM.S One or more repre-
'Uwaynat, which lies near the downwind end of the sentative grains from each was photographed at
southern arm of the $&bra Awhiri (fig. 32) the color magnifications ranging from about x 200 to about x
is about 4YR 5/8 (ye110wish red). The distance 11,000. Finally, using a Kevex X-ray energy
across which this amount of reddening occurs is spectrometer attached to the SEM, semiquantitative
nearly 700 km (435 mil, or about 10 times the dis- chemical analyses of the pigmenting material were
tance necessary to achieve approximately the same made at locations that could be identified on the
degree of redness in the northern coastal region of photomicrographs. Representative results are il-
western Libya. lustrated in figures 33 and 34 and 37 - 40.

Pigment in the thin films apparently is composed

Characteristics of the Pigment of amorphous ferric hydrate ("limonite") because,
although energy-dispersive X-ray analyses show
WHEN SAMPLES of the stained dune sands that they contain iron (fig. 33, trace 2), they show
'rom both of the regions studied are examined no evidence of crystallinity when viewed under the
..lnder a binocular microscope, the pigment is seen SEM at magnifications of x 50,000 and higher.
to occur, in part, as extremely thin films which
generally form complete coatings on grains and in
part as thicker concentrations which lie within in-
dentations on grains, where the material has been and> Trade names and company names are included for the benefit of the reader
do not imply endorsement of the product or company by the U.s. Geological
protected from abrasion. The latter material charac- Survey.

Digitized by Google
 RED COLOR IN DUNE SAND 69
Moreover, few X-ray-diffraction analyses of con- meteoric water, they provide potentially important
centrates of the pigmenting material reveal the pre- sources of iron in the pigment. That at least some of
sence of crystalline iron oxide minerals. Some of the more unstable of these minerals, such as augite
the red material concentrated in indentations on and hornblende, have undergone alteration is indi-
grains is composed entirely of essentially pure iron cated by the fact that grains of these materials com-
oxide (fig. 34), but most commonly it is composed monly show evidence of pitting by dissolution (fig.
of iron oxide stained platelets of clay minerals, the 35), and they commonly display cockscomblike ter-
composition of which is confirmed by energy-dis- minations blunted by abrasion (fig. 36).
persive X-ray analyses. Such analyses reveal that
most of the platelets are aluminum silicates (figs. Regional comparisons of the heavy-mineral frac-
38 - 40), although some platelets are composed of tions show a noticeable increase in the degree of
pure silica (fig. 33). The clay-mineral and silica- alteration of both augite and hornblende with dis-
platelets occur in aggregates that have clastic tex- tance of travel. For example, near the upwind end
tures similar to those displayed by clay of the ~ri' Awbiri, augite - the more unstable
mechanically infiltrated into sand (Crone, 1974, of these two silicates - typically is only weakly
1975; Walker and Crone, 1974; Walker, 1976). For etched, and hornblende is essentially unaltered. In
reasons explained later, such platelets are in- contrast, near AI 'Uwaynit, at the downwind end
terpreted by the author to be dominantly of that of that sand sea where dunes are the reddest, augite
origin. is nearly absent in the sands, and hornblende has
X-ray analyses of concentrates of clayey mater- become noticeably etched (fig. 35). These relation-
ial, removed from grains by using an ultrasonic' ships support Norris' (1969) contention that grains
cleaner, show that throughout Libya the clay of detrital iron-bearing minerals are important
minerals in the coatings are mixtures of dominantly sources of iron for staining dune sands, even in
kaolinite and illite, with subordinate amounts of very dry deserts.
mixed-layer illite-montmorillonite (table 8). X-ray
analyses also show that only the reddest dune Energy-dispersive X-ray analyses of clay coat-
sands, such as those occurring near Al 'Uwaynat, at ings on representative sand grains in samples col-
the downwind end of the allri' AwhirT (samples lected from dunes throughout the region, show that
672 and 674), contain detectable hematite in the the clays characteristically are iron-bearing. Some
grain coatings. In the author's opinion this hematite of the iron occurs as limonitic coatings on the clay,
represents iron oxide, which initially was precipi- but some also occurs within the lattices of the clay
tated as ferric hydrate and which upon aging either minerals because energy-dispersive X-ray analyses
in place or during transport has converted to show that iron is present in unstained clay. In-
hematite in the manner discussed by Berner (1969) asmuch as these clays are not authigenic, they
and by Langmuir (1971). probably are not in complete chemical equilibrium
with meteoric water, and, hence, upon being
repeatedly moistened by surface and interstitial
Sources of Iron waters, they likely release additional iron. The im-
THE SOURCE OF IRON in the pigment cannot portance of iron-bearing clay minerals as potential
be determined unequivocally from available data. sources of iron for iron oxide pigment in other
However, the fact that the redness of the dunes in- desert sediments has been stressed previously
creases with age and distance of transport, coupled (Walker and Honea, 1969; Walker, 1976).
with the likelihood that source areas of the sands
have not changed, suggests that the iron is derived
internally from the dunes. Two likely sources are Origin of Clay Coatings on
the iron-bearing accessory minerals within the
sands and the clay minerals in the clay coatings. Eolian Sand Grains
Studies of heavy-mineral concentrates show that
grains of iron-bearing minerals, such as augite, INASMUCH AS RED pigment commonly is
hornblende, epidote, ilmenite, and magnetite, are associated with clay that partially coats desert sand
present in dunes throughout the area examined. grains, the origin of the coatings and the circum-
Because these minerals are not in equilibrium with stances that allow them to be preserved on grains

Digitized by Google
70 A STUDY OF GLOBAL SAND SEAS

SCANNING ElECTRON MICROSCOPE (SEM) photo-

micrographs (A ~ of a coarse rounded quartz grain which,
when viewed under the binocular microscope, was completely
coated by a film of reddish-yellow pigment (5YR6I6 on Munsell
Soil Color Chart, Munsell Color Co., 1954); however, the grain
showed no discernible evidence of concentrations of pigment
in indentations on the surface. Rectangle in A is area of B; that in
B is area of C. D, Energy-dispersive X-ray traces showing the ele-
mental composition of the analyzed areas outlined in C (traces 1
and 2). The photomicrographs reveal that the shallow indenta-
tion on the grain contains scattered platelets that were not visi-
ble under the binocular microscope. The elemental data indi-
cate that the platelets are composed of silica and a small amount
of iron (trace 1 in q The platelets that have a clastic texture and
appear to be sitting on the surface of the grain (for example, the
analyzed platelet) are interpreted to be silica dust acquired dur-
ing transport and derived from eolian abrasion of quartz sand
grains. Some of the more firmly attached platelets may be chips
which are spalling from the surface as proposed by Margolis and
Krinsley (1971). The iron is interpreted to be iron oxide which
coats the entire surface of the grain (trace 2 in q, including the
platelets (trace 1 in q . Gold and paladium recorded in the
traces are contained in the conductive coating applied for SEM
examination. (Fig. 33.)

SI Sample 674

Digitized by Google
 RED COLOR IN DUNE SAND 71
SCANNING ElECTRON MICROSCOPE (SEM) photo-
micrographs (A ~ of a medium-sized subangular quartz grain
which contains abundant red (4YR 5/6 on Munsell Soil Color
Chart, Munsell Color Co., 1954) pigment concentrated in inden-
tations on the grain surface. Rectangle in A is area of B; that in B
is area of C. D, Energy-dispersive X-ray traces showing the ele-
mental composition of the pigmenting material in the analyzed
area outlined within rectangle in C. The photomicrographs
reveal that the pigmenting material is composed of anhedral
particles that are oriented parallel to the grain surface. The ele-
mental data indicate that the particles are composed almost en-
tirely of iron oxide. The particles are interpreted to be a poorly
crystalline iron oxide mineral, possibly hematite. Gold and
paladium recorded in the traces are contained in the conductive
coating applied for SEM examination. (Fig. 34.)

I ";,
'- . ':~'~,.~.~'

J. .:.~)~.,:.~.V~.,~<>,::;..~.
, r." " .'"il.. . . , ':,
;~,' .....
, ~.;
.J/.. '.
.(
.;:i<t;. . . . .~
'~
. "~'" ~. '~ "...,.
f4.~"";"~~. "ri~
.' ...~. -, ~.: .:i..',.~...-.t:'. '..1t

.
< ... , .
~ ~';..
\. _

e-
.. - "

..-,.... . "
)I.

.... ~~~
".,
'. - . " .

\."
....
,

.......
, ,..-.\/,~.;-..
,
..

.
-~
". . ....
'.'
:."
_ ".

.... , "'~'O':" .. _<~"

, ';,~. ....... '. ~~
'.,.,

,...
,.",:'":' . . . I .
'."

.. . . ..f .~
'

., '>'..(0 ;,
.. '''~.,
t ... . " : . . . . . ._ "
.
,:\.,";' - " ....
'

'~,' ~'~'
~1'....
, .. ,'.
, (
'i-.' . . .
.'
~'~~.'
"

-t' ....",

Sample 674

Digitized by Google
72 A STUDY OF GLOBAL SAND SEAS

despite long distances of eolian transport needs ex-

planation. In part, the coatings may be inherited
from matrix clays in sandstones and other bedrock
lithologies from which the sand grains are derived.
However, three lines of evidence, when considered
together, indicate that clay coatings commonly con-
sist of airborne dust emplaced on sand grains by
mechanical infiltration of dust particles into the
dunes or into other surficial deposits from which
sand grains of the dunes are derived. This method
of acquiring clay coatings is similar to that proposed
by Folk (1969) to account for the formation of clay
coatings on sand grains in Australian dunes.
The types of evidence are as follows:
1. Clay coatings are not present on sand grains of
the beach from which coastal dune sands are
derived, but they are common on grains in
the dunes located a short distance inland
100 #'Ill
from the beach. Thus, clay coatings must be
acquired as the sand migrates inland.
2. Clay coatings on sand grains of dunes through-
Sample 672 out the region characteristically are com-
ETCHED HORNBLENDE GRAIN from
modern dune near AI 'Uwaynat, libya. (Fig.
posed of aggregates of tiny platelets (figs.
35.) 37 -40) which display clastic textures that are
similar to those produced experimentally by
infiltrating clay-bearing water into sand
(Crone, 1974, 1975). These textures, in tum,
are similar to those produced naturally by
clay which has been infiltrated into sand by
influent seepage of surface water (Walker
and Crone, 1974; Crone, 1975; Walker, 1976).
Such similarities indicate that the clay is
composed of mechanically infiltrated parti-
cles, and influent rainwater is the most logi-
cal agent to accomplish the infiltration.
3. The mineralogy of the clay in the coatings is
similar to that expected of the airborne dust,
suggesting that dust is the probable source of
the clay. This latter point needs further ex-
planation.
Time limitations precluded the establishment of
stations in Libya for the systematic collection of air-
borne dust samples. The mineralogical composition
of airborne dust was predicted, however, by analyz-
ing the clay fraction of representative samples of
100 J.IIO
the major types of potential source materials. The
Sample 33 types and number of samples analyzed were
ETCHED AUGITE GRAIN from modern bedrock (15), Quaternary(?) alluvium (4). Quatern-
dune near As Saddidah, libya. (Fig. 36.) ary(?) dune sand (5), soils (33), modem alluvium

Digitized by Google
 RED COLOR IN DUNE SAND 73
(22). These are compared with the clay in coatings Presumably, part of the clay-size platelets of
on sand grains in the modern dunes. silica that occur on some sand grains (fig. 33) also
The locations of analyzed samples are shown in are derived from airborne dust, but they consist of
fIgure 32. Results of these analyses (table 8) show particles which have been abraded from quartz
that essentially all types of materials which are ex- grains during eolian transport, rather than eroded
posed to the wind and which could provide sources from the bedrock. Some of the silica platelets may
of clay contain the same varieties of clay minerals also represent chips that are in the process of spall-
in roughly the ~e proportions. In all types of ing from the surface of the grains (fIg. 33C) in the
source materials - if palygorskite, a likely manner proposed by Margolis and Krinsley (1971,
authigenic clay. is exluded - kaolinite and illite are p.3397).
the predominant clay minerals, almost without ex- The mechanism for producing clay coatings on
ception, and mixed-layer illite-montmorillonite is sand grains by mechanical infiltration probably
the principal subordinate clay mineral. The same operates throughout the desert, but it is most effec-
varieties of clay minerals in roughly the same pro- tive in areas where rainfall is greatest because, in
portions occur in the coatings on sand grains in the such areas, water penetrates more frequently and
modern dunes (table 8). These mineralogical deeper into the deposits. Consequently, the amount
similarities and the clastic texture of the clay of clay infiltrated into the surfIcial deposits is
strongly suggest that the clay coatings are com- greatest and the rate of formation of clay coatings
posed of infiltrated dust derived ultimately from on the included sand grains is most rapid in the
the bedrock. region of relatively high rainfall near the coast. The
To what extent clay minerals, other than characteristic occurrence of mechanically infIl-
palygorskite, have formed by processes of surface trated clay matrix in the soils and temporarily
weathering is not known, but in view of the striking stabilized dunes in the coastal region (fIg. 37) sup-
similarities between the clay mineralogy of the ports this interpretation.
soils and that of the bedrock (table 8), it seems Although the prevailing winds are onshore from
likely that clays in the soils also are derived mainly the Mediterranean Sea, the airborne dust even in
from the bedrock, not from weathering. Indeed, all the coastal region is derived principally from North
available evidence leads to the conclusion that African sources, not from sources across the sea.
there has been no signifIcant authigenesis of clay The winter westerlies and other winds that diverge
minerals except for the palygorskite. from the prevailing direction have ample energy to
The preceding interpretation readily explains the erode dust from local sources. The transportation of
origin of kaolinite which occurs as a dominant clay dust from the central Sahara northward into Europe
mineral in the alkaline desert soils, as it cannot and east-northeastward into the Middle East is well
reasonably be expected to have formed in place documented (Yaalon and Ganor, 1973).
there. Furthermore, SEM studies of the soils show The infiltrated clayey matrix tends to bind
that clay in the argilic "B" horizons charac- together the framework grains of initially uncon-
teristically has a clastic texture similar to that solidated surficial deposits; thus, the clay helps, at
shown in figure 37, and this texture is indicative of least temporarily, to anchor the deposits. However,
mechanically infiltrated material. The importance owing in part to the friable nature of the surficial
of infiltrated dust on the formation of desert soils deposits and in part to the paucity of vegetation to
has been stressed previously by Yaalon and Ganor anchor them, their stabilization in this dry climate
(1973). The conclusion is reached therefore, that all is only temporary. Subsequent erosion releases the
the clay in all types of the surfIcial deposits has clay-coated sand grains to eolian transport.
been mechanically infiltrated into the sediments. As clay-coated grains are blown about by the
This may have been accomplished either by in- wind, the clay is removed from exposed surfaces of
fIltration of airborne dust by occasional rain, as the grains by abrasion, but it tends to remain in in-
with dunes and soils, or by infiltration of clay sus- dentations, where it is protected from abrasion.
pended in the water of intermittent influent Such clay-coated indentations are common on
streams, as with alluvial deposits. grains of all sizes but are most abundant on small

Digitized by Google
74 A STUDY OF GLOBAL SAND SEAS

SCANNING elECTRON MICROSCOPE (SEM) photo-

micrographs (A ~ of matrix-rich sand collected 2 metres
below the surface of a weakly stabilized red dune in the north-
ern coastal region near Mi~ratih. The color of the clay as deter-
mined under a binocular microscope is reddish yellow (5YR 6/6
on Munsell Soil Color Chart, Munsell Color Co., 1954). Rec-
tangle in A is area of B; that in B is area of C. 0, Energy-dispersive
X-ray traces showing the elemental composition of the matrix
clay in the analyzed area outlined within rectangle in C. The
photomicrographs reveal that the matrix is composed of an ag-
gregate of clay-size platelets (q. The elemental data indicate
that the platelets are composed mainly of aluminum silicates
(clay minerals) and iron. The matrix is interpreted to be com-
posed of airborne dust that has been mechanically infiltrated
into the dune sand and has become reddened by iron oxide
stain subsequent to infiltration. Gold and paladium recorded in
the trace are contained in the conductive coating applied for
SEM examination. (Fig. 37.)

Sample 114

Digitized by Google
RED COWR IN DUNE SAND 75
SCANNING ElECTRON MICROSCOPE (SEM) photo-
micrographs (t\-C) of a fine-grained subrounded quartz grain
which contains abundant light-reddish-brown (5YR 6/4 on
Munsell Soil Color Chart, Munsell Color Co., 1954) pigment
concentrated in indentations. Rectangle in t\ is area of 8; that in
8 is area of C. D, Energy-dispersive X-ray traces showing the ele-
mental composition of the pigmenting material in the analyzed
area within rectangle C. The photomicrographs reveal that the
pigment is composed of aggregates of clay-size particles. The
elemental data indicate that the particles are composed mainly
of aluminum silicates (clay minerals) and iron. The sample from
which this grain was selected was collected from the surface of
a modern red dune about 100 km (62 mil inland from the coast
near Tripoli. The grain is typical of clay-coated sand grains
which comprise a major portion of the dune. The sand in the
dune has been reworked from nearby matrix-rich soils and dune
sands that have characteristics similar to those shown in figure
37. The pigmenting material illustrated here is interpreted as
mechanically infiltrated matrix clay which is analogous to but
coarser than that shown in figure 37 and which during eolian
transport has been retained in indentations on grains where it is
protected from abrasion. Gold and paladium recorded in the
trace are contained in the conductive coating applied for SEM
examination. (Fig. 38.)

51 Sample 8

Digitized by Google
76 A STUDY OF GLOBAL SAND SEAS

SCANNING ElECTRON MICROSCOPE (SEM) photo-

micrographs (A ~ of a fine-grained subangular quartz grain
collected from a pink (8YR 7/4 on Munsell Soil Color Chart,
Munsell Color Co., 1954) modern dune near Qaryat Abu Nu-
jaym. Rectangle in A is area of B; that in B is area of C. 0, Energy-
dispersive X-ray traces showing the elemental composition of
the pigmenting material in the analyzed area within rectangle in
C. This grain is typical of fine and very fine sand in dunes
throughout the interior of the desert. Note low degree of round-
ness and abundance of protected indentations containing con-
centrations of clay platelets analogous to those illustrated in
figure 38. Gold and paladium recorded in the trace are con-
tained in the conductive coating applied for SEM examination.
(Fig. 39.)

Sample 44

Digitized by Google
RED COLOR IN DUNE SAND 77
SCANNING ELECTRON MICROSCOPE (SEM) photo-
micrographs (A -C) of a medium-grained rounded quartz grain
collected from a red (3YR 5/6 on Munsell Soil Color Chart, Mun-
sell Color Co., 1954) dune near AI 'Uwaynat. Rectangle in A is
area of B; that in B is area of C. D, Energy-dispersive X-ray trace
showing the elemental composition of the pigmenting material
in the analyzed area within rectangle C. The photomicrographs,
coupled with the elemental data, show that despite the rounded
character of the grain, iron oxide stained clay platelets are abun-
dant in shallow indentations. Such concentrations of clay and
iron oxide are common on the surfaces of rounded and well-
rounded sand grains throughout the desert. Gold and paladium
recorded in the trace are contained in the conductive coating
applied for SEM examination (Fig. 40.)

Sample 674

Digitized by Google
78 A STUDY OF GLOBAL SAND SEAS

graitls, such as those in the fine and very fine sand The processes that introduce clay presumably
sizes. These small grains show little evidence of operate over and over again as the eolian sands
abrasion despite long distances of eolian transport. migrate and the dune fields accumulate. Therefore,
Characteristically, they tend to have angular to deep infiltration of the dust into the deposit is not
subangular shapes and to retain their original sur- required for the formation of coatings, and the coat-
face irregularities. Such observations support the ings are acquired even where rainfall is negligible.
experimental data of Kuenen (1960), who found
that eolian abrasion decreases rapidly with
decrease in grain size and becomes insignificant for Conditions Favoring
sizes below 150 pm. A typical grain of fine sand is Formation of Iron Oxide
shown in figure 39. A typical coarser grain
(medium sand) is shown in figure 40. Pigment
The development of coatings of infiltrated clay
on sand grains of the Libyan dunes seems to be par- NO DATA ARE AVAILABLE concerning
ticularly favored by the readily available source of chemistry of the water in contact with the sand
clay in the sedimentary rocks that dominate the grains (for example, dew at the ground surface,
bedrock source areas. Erosion of these rocks by pellicular films of water on grains within the
wind, rain, and ephemeral streams provides an vadose zone, and ground water below the water ta-
unending source of clay for airborne dust. In addi- ble), but the water can be assumed with confidence
tion, the developing of the coatings is enhanced by to be oxidizing and alkaline and, therefore, to lie in
prolonged residence of the sand grains in surficial the Eh-pH stability field of ferric hydrates and
deposits in the coastal steppe area, where condi- hematite (Garrels and Christ, 1965, fig. 7.6). The
tions are especially favorable for the mechanical in- paucity of vegetation - indeed, the complete lack
filtration of clay. Sand grains that bypass this zone of it throughout most of the desert -and the lack of
or migrate across it quickly are likely to reach the any other reducing agents provide assurance that
desert interior without having acquired the clay meteoric water, upon infiltration into the sand, re-
coatings, and acquisition of the coatings is much tains dissolved oxygen obtained from the at-
slower once the grains reach the very arid interior. mosphere. Accordingly, an oxidizing environment
Unstained grains occur in all of the active dunes seems a certainty.
that were investigated in this study, both in the Moreover, the widespread occurrence of carbon-
coastal region and in the interior of the desert, but ate, gypsum, and palygorskite in surficial deposits
in both regions they are progressively less common of the desert indicates an alkaline weathering en-
with increased distance of transport. This distribu- vironment. Therefore, iron released by alteration of
tion suggests that the coatings are acquired during any of the above-mentioned iron-bearing minerals,
migration of the dunes, even in the intensely arid including the clay minerals in clay coatings, pre-
regions. cipitates as iron oxide, probably initially as
The mechanisms proposed by Folk (1969) to ex- amorphous ferric hydrate and (or) finely crystalline
plain the acquisition of clay coating on sand grains goethite, both of which are metastable and convert
in the Simpson Desert seems thoroughly adequate to hematite upon aging (Berner, 1969; Langmuir,
to accomplish the task here. That is, during calm 1971).
periods between wind storms, airborne dust settles
onto the dune surfaces where it adheres to sand Stated another way, hot desert climates, such as
grains, particularly if moisture from dew or rain is those in north Africa, are ideally suited for the for-
present. Moreover, during periods of rainfall addi- mation and preservation of iron oxide pigments. In-
tional dust is washed from the atmosphere and car- deed, precursor oxides of hematite and, ultimately,
ried into the dunes by influent rain, where hematite should be expected to occur as stains in
ultimately it also settles onto the surfaces of sand desert deposits wherever and whenever moisture,
grains. even in seemingly negligible amounts, comes into

Digitized by Google
 RED COLOR IN DUNE SAND 79
contact with any kind of iron-bearing minerals that 2. Amount 01 clay retained on sand grains. -
are not in equilibrium with the water. Inasmuch as pigment in the dunes is carried in part
by clay coatings on sand grains, dunes tend to
become redder with the increase in the amount of
Factors Affecting Regional and clay that is acquired and retained on individual
grains. Hence, a significant factor in determining
Local Variations in Redness the redness of dunes is the percentage of sand
grains that have clay coatings, and also the percen-
REDDENING ACCOMPANIES aging of the in- tage of the surface area of these grains that is
itially precipitated precursor oxides of hematite; coated.
consequently, dunes tend to redden with increasing The modem dune from which the sand grain in
age and distance of transport. If aging of the ferric figure 38 was collected is strikingly red because it is
hydrates were the only factor involved in redden- composed mostly of grains which, like the one in
ing, the intensity of redness of dunes would be a the photograph, retain extensive coatings of stained
measure of their age, and their color probably could clay. Sand in this dune was derived, in part, from
be used as a basis for correlation. The evidence that nearby red soils of the desert type described earlier,
has been presented in this chapter, however, shows and in part from older weakly stabilized red dunes,
that redness also is affected by a variety of other both of which contain abundant infiltrated clay
factors which are not uniform everywhere. Conse- matrix similar to that in the example shown in
quently, local or regional variations in numerous figure 37. Owing to the relatively short distance of
environmental factors can be expected to cause transport, a large amount of clay is retained on the
variations in the color of the dunes. Following is a grains. Analogous red dunes of modern or near-
list summarizing these factors, many of them inter- modern age are common in the coastal region of
related, and their effects on the color. northern Libya.
1. AvaiJabiJity 01 unstable iron-bearing 3. Amount 01 moisture avaiJable.-Water
minerals in source rocks.-linstable iron-bearing plays several important roles that enhance the red-
minerals, particularly ferromagnesian silicates, dening process: (a) It provides the medium for
such as augite and hornblende, are perhaps the chemical hydrolysis of iron-bearing minerals (for
most important sources of iron for pigment in dune example, ferromagnesian silicates and clay
sands. If other conditions are equal, dune sands minerals in the grain coatings) and is responsible
derived from source rocks rich in these minerals for the release of iron, which precipiates as oxide
probably will redden faster than sands derived from stain; (b) it provides a mechanism for mechanical
rocks deficient in these minerals. This point was infiltration of airborne clay into dune sands and
stressed by Norris (1969). into other surficial deposits from which dune sands
The effect of differences in source-rock are derived, thereby increasing the amount of clay
mineralogy may explain why the ~ra' Awbarl available for coating and staining sand grains; (c) it
and ~4r8' Marziiq are strikingly redaer than the assists in the growth of vegetation, which in turn
alira' Rabyanah and Sarii'Tibasti (fig. 31), despite tends to stabilize dunes and helps protect the desert
the fact that sand in the ~ra' Rabyanah and Sarir soils from erosion, thus keeping grains longer in the
Tibasti has travelled at ieast as far as that in the environment where conditions are most favorable
other deposits. The ~~ra' Awbari and ~~ra for reddening. Reddening of dune sands is much
Marziiq lie downwind from volcanic rocks exposed faster in the coastal region than in the interior of
in the highlands of AI Hariij al Asway and Jabal as the desert because of the increased amount of
Sawda' (figs. 31, 32), and both deposits almost cer- moisture along the coast. The moisture may be rain,
tainly receive contributions of sand from these dew, or ground water.
rocks. The ~ra' Rabyiinah and Sarir Tibasti, on 4. Time.-Time is required for the alteration of
the other liand, have no comparable source of both the ferromagnesian silicate minerals and the
unstable iron-bearing minerals exposed upwind. clay minerals in grain coatings. Together, they pro-

Digitized by Google
 80 A STUDY OF GLOBAL SAND SEAS

vide the major sources of iron in the pigment. 6. Distance 01 transport.-This factor has the
Because these alterations are continuous and same effect as time. Increased distance of travel
because reddening normally accompanies aging of normally increases the time available for altera-
iron oxides, dunes throughout western Libya have tions that cause reddening; hence, reddening in-
tended to become increasingly red with age. The creases with distance of travel. One might expect
time required to achieve a specific degree of red- the pigment to be removed from sand grains by ero-
ness, however, is decreased where the amount of sion during prolonged eolian transport, but this is
moisture is increased, so reddening is faster in the not the case, as demonstrated by the evidence pres-
steppe region along the coast than in the intensely ented already.
arid interior of the desert. Two opposing and unequal processes act upon
5. Grain size and grain shape.-During eolian grains during their transport: (1) abrasion when the
transport coarse grains are abraded at a faster rate grains are in transit, and (2) accumulation of pig-
and to a greater degree than are fine grains; hence, ment when the grains are at rest. Abrasion is much
they become more rounded and the indentations more effective with coarse grains than with fine,
caused by original irregularities in grain shape tend because coarse grains, owing to greater mass,
to be reduced in size and number. The coatings of abrade more rapidly. Nevertheless, pigment charac-
iron oxide and iron oxide stained clay, which can teristically is preserved even on coarse grains in
persist only if protected from abrasion, are pits, where it is protected from abrasion, and it
therefore less common on the surfaces of coarse commonly occurs as films that envelop coarse
sand than on fine sand. The finer grains, on the grains. Accumulation of pigment, on the other
other hand, because they abrade very slowly, or not hand, is the dominant process with finer grains
at all (Kuenen, 1960), tend to retain original ir- because, owing to their small mass, they abrade
regularities, and iron oxide stain once acquired slowly, if at all.
tends to persist. Fine sand grains in dunes, Fine and very fine grains, which are the predomi-
therefore, normally are coated with more clay, con- nant sizes in the dunes studied in this investigation,
tain more pigment, and are redder than are the tend to accumulate more pigment on their surfaces
coarse grains. when they are at rest than they lose by abrasion
The effect of grain size on color is striking in when they are in transit. The accumulation of pig-
places where modem dunes show two distinctly ment in this manner, coupled with the fact that it
different colors of sand. As an illustration, dunes reddens upon aging, causes the dunes to become in-
located near As Saddadah (fig. 32) were examined creasingly redder with distance of transport. As is
by the author after a day of particularly strong true of time, the distance required to achieve a
winds. Surfaces of the slipfaces on the dunes were specific degree of redness is shorter if the amount
covered by a veneer of dominantly medium sand of moisture is increased. Consequently, in the in-
which had a color of pink to reddish yellow (7.5YR tensely arid interior of the desert, the distance re-
7/5, Munsell Soil Color Chart, Munsell Color Co., quired to achieve a certain degree of redness is
1954) and had been deposited by the strong winds
many times greater than it is in the moister coastal
region.
of the previous day. Exposed in places on these
slipfaces, beneath the veneer of medium sand, were 7. Types 01 clay minerals contained in coat-
distinctly redder (4YR 6/8), fine to very fine sands ings on sand grains.-Iron is carried by clay
that had been deposited earlier by winds of lesser minerals as a constituent of the clay-mineral lattice,
velocity. A likely analogous example has been il- as well as in coatings of iron oxides on the surfaces
lustrated by FUrst (1970, pI. 5) in a color photograph of clay particles (Carroll, 1958). Such iron can be
that clearly shows the dual color of the sand in released by weathering processes if the clay
some dunes in the Sahri' Awbiri. However, the minerals are not in equilibrium with the meteoric
two different colors are attributed by FUrst to water with which they come in contact. Because
different ages of the sands rather than to different the Eh and pH of interstitial moisture are high in
grains sizes. Other analogous relationships between the desert environment, iron released by such
color and grain size in dune sands have been recog- weathering should precipitate almost immediately
nized by Folk (1976), in the Simpson Desert in and increase the amount of iron oxide on the sur-
Australia. face of clay particles. The amount of iron released

Digitized by Google
 RED COLOR IN DUNE SAND 81
by such weathering depends in part on the type of concentrations of such grains. If such dunes occur
clay minerals that are present in the coatings. there. they have only local significance. Grains of
because clays are not equally capable of carrying sand which have been reworked from the Mesozoic
iron as a constituent of the crystal lattice (Carroll. red beds usually can be distinguished readily from
1958. table 1). grains which have developed red pigment in the
Among the clays occurring as coatings on sand manner described earlier. For example. the
grains in the dunes of western Libya. iron is likely reworked grains normally are much redder than the
to be more abundant in illite and mixed-layer illite- pigments forming today. They have hues ranging
montmorillonite than in kaolinite. Grains with from lOR to 7.5R on the Munsell Soil Color Chart
coatings containing high percentages of illite and (Munsell Color Co . 1954) compared with hues of
(or) mixed-layer illite-montmorillonite. therefore. only 5YR to about 3YR in the younger pigments.
probably have greater potential for increasing red- Most reworked sand includes rock fragments of
ness faster than do those with low percentages of red sandstone. red siltstone. and (or) red shale. all
these minerals. The effect of such differences. of which contain matrix with dark-red color. which
however. may be insignificant in the sands studied attests to the derivation of the grains from red beds.
in this investigation because the mixtures of clay Moreover. the red pigment in reworked quartz
minerals in the coatings are virtually the same grains commonly penetrates the interior of the
throughout the region. In other parts of the Sahara grains along incipient fractures and. hence. cannot
or in other deserts. differences in the types of clay easily be removed completely when the grains are
minerals occurring in clay coatings might conceiva- treated with reducing agents. In contrast. pigments
bly make significant differences in the rate at that are forming today are limited to the surfaces of
which reddening occurs. grains. and they are easily and quickly removed by
8. Percentage 01 s8Dd grains derived from reducing agents. Although they are considered
older red beds.-Red beds of Mesozoic age crop out unimportant here. red sands that are reworked
locally in western Libya and red sand grains that from older red beds may play important roles. at
apparently have been reworked from these deposits least locally. in producing red dunes elsewhere.
commonly can be identified in the dunes. These Such dunes. in contrast to those discussed here.
grains. in places. contribute in a minor way to the should show a decrease in redness with distance of
color of the dunes. but the author saw no red dunes transport away from the source of the sand because
in Libya that obviously owed their color mainly to of dilution by non-red sands from other sources.

Digitized by Google
Digitized by Google
 '
.:X),;:::>/jL:'}.';-O: .;' .
. ... <;;:i~W, <::;;i:'"'<.~:'

,'" >:JH\~~,;;~~~;~~;~#"

A STUDY OF GLOBAL SAND SEAS

SEDIMENTARY STRUCTURES IN DUNES

Chapter E

By EDWIN D. McKEE

With sections on THE LAGOA DUNE FIELD, BRAZIL

By JDAD J. BIGARELLA4

'Geological Institute. Federal University of Parana. Brazil

Contents
Paae Page
Summary of conclusions ........................... . 87 Structural features at lagoa dune field, Brazil,
by loao I. BigarelJa .......................... . 114
Introduction ..................................... . 87 Parabolic dunes .............................. . 114
Sedimentary structures common to most dunes ...... . ~9 Method of study .......................... . 114
Cross-stratification ........................ . 114
Sedimentary structures characteristic of certain dune
Nose area ................................ . 114
types ...................................... . 89
Crest area ................................ . 118
Barchan and barchanoid ridge .................. . 89
93
Arms .................................... . 119
Transverse ................................... .
Deformation ............................. . 121
Blowout ..................................... . 94 122
Interdune areas among parabolic dunes ......... .
Parabolic ..................................... . 94
Interdune areas among reversing dunes .......... . 123
Retention ridges .............................. . 96 124
Dissipation of dunes, lagoa, Brazil, by 1010 I. BigarelJa .
Dome ....................................... . 98
General features and environment of dune dissipa-
Linear ....................................... . 102
tion ....................................... . 124
Reversing .................................... . 103
Processes, structures, and soil formation ......... . 127
Star ......................................... . 106
Dunes ................................... . 127
Sedimentary structures of interdune areas ............ . 107 Interdune areas ........................... . 130
Deformational structures in dunes .................. . 111 Deformation results of dune dissipation ......... . 130

Digitized by Google
84 A STUDY OF GLOBAL SAND SEAS

Illustrations
FIGURE 41. Photographs of sand avalanching down dune slipface, White Sands National
Monument, New Mexico, U.S.A. ..................................... 88
42. Cross sections of barchan dune showing structure ....................... 90
43- 45. White Sands National Monument, New Mexico, U.S.A.:
43. Photograph showing series of steeply dipping cross-strata forming
tabular-planar sets in barchanoid ridge. . . . . . . . . . . . . . . . . . . . . . . . . 91
44. Cross sections of barchanoid ridge dune . .. . . . . . . . . . . . . .. .. . . . . . . 91
45. Photograph showing beveled remnant of barchan dunes in interdune
area ....................................................... 91
46. Cross sections and plan view of barchanoid ridge dune, Killpecker dune field,
Wyoming, U.S.A. ................................................... 92
47. Photograph showing barchanoid ridge dunes and interdunes, White Sands Na-
tional Monument, New Mexico, U.S.A................................. 92
48. Cross sections of two transverse dunes in the United States ............. ,. 93
49. Photographs of blowout dunes on shore of Lake Michigan and in Califomia,
U.S.A. ............................................................ 94
50. Photographs showing parabolic dunes, White Sands National Monument, New
Mexico, U.S.A ..................................................... 95
51. Sketched sections through a parabolic dune, White Sands National Monument,
New Mexico, U.S.A. ................................................. 96
52. Sketch of trench wall on nose of parabolic dune, Killpecker dune field, Wyom-
ing, U.S.A ......................................................... 97
53. Cross sections of retention dune showing details of stratification, Jardim 510
Pedro, Praia do Leste, Paran', Brazil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
54. Stereonets for three sections showing angle and direction of dip on foresets of
retention ridge, Guairamar, Paran~, Brazil .............................. 98
55. Photograph showing trench through dome dune, White Sands National Monu-
ment, New Mexico, U.S.A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
56. Cross section of dome dune, White Sands National Monument, New Mexico,
U.S.A ............................................................. 100
57. Cross sections of dome dune, Killpecker dune field, Wyoming, U.S.A. ...... 101
58. Cross sections of coastal dome dune showing characteristic stratification, Praia
do Leste, Parana, Brazil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 102
59. Aerial photograph of seif or linear dunes, north of SebhI, Libya. . . . . . . . . . .. 102
60. Sketch of hypothetical section, showing structure of seif dune. . . . . . . . . . . .. 103
61. Cross section of seif dune near Sebha, Libya, based on series of test pits. . .. 104
62. Aerial photographs of reversing dunes, Great Sand Dunes National Monument,
Colorado, U.S.A. ................................................... 105
63. Drawing showing structure in reversing dune, Great Sand Dunes National
Monument, Colorado, U.S.A. ........................................ 105
64. Photograph of trench in dune parallel to dominant wind direction, White Sands
National Monument, New Mexico, U.S.A., illustrating effects of wind reversal
during winter season. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 106
65. Aerial photograph of star dunes, southeast of Zalim, Saudi Arabia . . . . . . . . .. 107
66. Photograph showing steeply dipping cross-strata in trench at crest of high star
dune, southeast of Zalim, Saudi Arabia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 107
67. Photograph showing star dunes along ridges of linear dunes, Gobabeb, South-
West Africa ........................................................ 108
68- 72. Photographs at White Sands National Monument, New Mexico, U.S.A., show-
ing:
68. Interdune area ............................................... 109
69. Beveled windward surface of transverse dune with truncation of lee-
side foresets ................................................ 110
70. Beveled remnant of barchan dunes in interdune area. . . . . . . . . . . . .. 110
71. Beveled remnant of transverse dune in interdune area . . . . . . . . . . . .. 110
72. Horizontal and irregularly bedded strata in interdune area between
barchanoid ridges ........................................... 111

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 85

FIGURE 73. Block diagrams and photograph of interdune structures between seif dunes,
...
northeast of Sebhli, Libya ........................................... . 111
74. Photograph showing interdune surface among star dunes, Namib Desert,
South-West Africa ................................................. . 112
75. Sketches showing principal types of deformation structures in avalanche
deposits of dunes ................................................. . 113
76. Map showing location of Lagoa dune field, IIha de Santa Catarina, Brazil ... . 115
77 - 86. Parabolic dune, Lagoa dune field, IIha de Santa Catarina, Brazil:
77. Cross sections, plan view, and profiles .......................... . 116
78. Rose diagrams of cross-strata dip directions representing principal parts
of parabolic dune and interdune area ......................... . 119
79. Rose diagrams of attitude of cross-strata deposited in an earlier dune
crest and in the nose of a parabolic dune ..................... . 123
80. Outcropping laminae exposed on beveled surface in interior of
parabolic dune ............................................. . 124
81. Deformational structures observed in eastern arm of dune ........ . 125
82. Photographs of internal structures .............................. . 126
83. Diagrams showing stratification in pits along east-west line in inter-
dune area between arms of dune ............................ . 128
84. Diagrams showing stratification in pits along north-south line of inter-
dune area between arms of dune ............................ . 129
85. Photographs showing structures of interdune area between arms of
large dune ................................................. . 129
86. Rose diagram showing cross-strata dip directions, interdune area .. . 130
87. Diagrams showing reversing dune and interdune structures, Lagoa dune field,
IIha de Santa Catarina, Brazil ........................................ . 131
88. Map showing dissipation ramp leading down from Joaquina Hill to Lagoa dune
field, IIha de Santa Catarina, Brazil ................................... . 132
89. Sketches showing contorted wavy pattern separating individual sand-flow
layers developed during dissipation processes ........................ . 133
90. Schematic cross sections illustrating features resulting from dissipation condi-
tions ............................................................. . 134

Tables
TABLE 9. Classification of principal dune types according to number of slipfaces and to
form.............................................................. 89
10-13. Various data for a parabolic dune, Lagoa dune field, Brazil:
10. Frequency of dip-direction intervals for cross-strata ............... 120
11. Directions and degree of dip for cross-strata. . . . . . . . . . . . . . . . . . . . . . 120
12. Attitude of cross-strata sets in trench F .......................... 121
13. Attitude of cross-strata sets in trench G . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Digitized by Google
Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 87

downward concavity of laminae, the abundance of

Summary of Conclusions scour troughs, and the horizontality of apparent
dips normal to wind direction.
~EDIMENTARY STRUCTURES in dunes con
sist of cross-stratification and of minor surface
features, such as ripple marks and rain pits. Pene- Introduction
contemporaneous deformational features, includ-
ing slump marks, also are present in a wide variety THE PRINCIPAL FACTORS that control dune
of forms. Interdune deposits differ from dunes in development and are therefore responsible for dune
composition and texture, as well as in structure. form are an adequate source of sand and wind of
This chapter is devoted largely to the description sufficient strength and duration to transport and
and analysis of the cross-strata that are charac- deposit it. The geomorphic form assumed by an ac-
teristic of the various dune types. These structures cumulating sand mass is determined primarily by
provide the record of dune development, the means the direction and strength of the responsible wind
of recognizing and identifying dune types in the but may be influenced locally by barriers or deflect-
subsurface and in ancient rocks, and important con- ing factors, such as bedrock, vegetation, water, or
trols of fluid migration and accumulation. other sand masses. The internal structure or cross-
The study of dune cross-strata is still in a pioneer stratification within each dune is a record of its
state and much more work must be done before depositional history.
meaningful generalizations can be made for many Development of dune structures, as described by
varieties of dunes. Most of the work done so far has Bagnold (1941, p. 37), involves two principal proc-
consisted of a direct approach in which trenches or esses: (1) transport of sand particles up the gentle
pits have been dug in dunes, exposing three-dimen- windward (stoss) slope of the dune by saltation, or
sional views of cross-strata. Additional studies bouncing, of grains; and (2) avalanching of sand
have consisted of making statistical analyses of down the slipface on the steep lee side (fig. 41)
foreset dip directions, the vectors of which furnish either by slumping (mass movement) or by grain
information on the direction or directions of sand flowage. Where wind is strong, some sand com-
transport and the type of dune represented. monly is carried outward from the dune crest and
A structural classification based in part on the settles directly from suspension.
number of slipfaces, as represented by the principal Sedimentary structures in eolian dunes com-
directions of high-angle foreset planes, is useful. monly consist of sets of medium- to large-scale
Thus, dunes having slip faces inclined in one direc- foresets that dip downwind and are separated by
tion, two directions, or many directions are recog- bounding surfaces that are horizontal or inclined at
nized, in addition to sheet sands and stringers and low angles. An exception to this basic pattern of
interdune deposits that have no slipfaces. Interdune cross-stratification develops where the available
deposits typically have flat or irregular stratifica- sand load exceeds the amount of sand that the wind
tion. can move up a dune slope. thereby causing deposi-
Descriptions of dune cross-strata are concerned tion of low-angle strata dipping to windward. Other
primarily with the bounding surfaces of sets and exceptions develop when the wind is unusually
the foreset planes within these sets. The cross- strong or changes direction. thus introducing partial
strata are mostly of two basic types - tabular erosion of earlier deposits to form either troughs or
planar and wedge planar. Trough-type cross-strata beveled surfaces that are subsequently buried and
are rare in those dunes that have been examined in- form part of the dune record.
ternally. Although there are many exceptions, Distinctive dune forms have been correlated with
high-angle and large-scale foreset planes are typical genetic factors by many students of sand seas, and
of dune structures in general. The spread and pat- numerous classifications of dune patterns (chapter
tern of dip directions of foresets give valuble infor- J. tables 26 -32) have resulted. A common method
mation on dune types. Some types of dunes also can of subdividing dune forms is to recognize the
be differentiated by such features as the upward or characteristic shapes attributed to the direction or

Digitized by Google
88 A STUDY OF GLOBAL SAND SEAS

AVALANCHING SAND BY SLUMPING and grain flowage down dune delineated by top of avalanching sand. Ripple length in
slipface of dunes, White Sands National Monument, New Mex- lower left is 4 inches (10 cm). C, Series of slumps partially bury-
ico, U.S.A. A, Small blocks of weakly cohesive sand formed ing ripple-marked slipface. Knife handle is 6 inches (15 cm)
from damp crust in upper part of slump masses. Length of base long. 0, Scoop surface where slump mass broke loose near top
of ripple is 6 inches (15 cm). 8, Curving crest of barchanoid of slipface on transverse dune. (Fig. 41.)

directions of the dominant wind or winds involved dunes with three. four. or more arms extending out-
(table 9) . Those dunes that are formed by a prevail- ward from a central high point.
ing wind (or a single dominant wind) include types Although all eolian dunes are primarily con-
commonly referred to as barchan. barchanoid ridge. trolled by wind activity and therefore share many
transverse. dome-shaped. and some parabolic. structural features. some distinctive and probably
Dunes attributed to bidirectional winds may form unique structures characterize each type. The value
linear ridges. locally known as longitudinal dunes of recognizing such distinguishing features should
or seif dunes. where the winds converge at low be apparent. for they make possible the interpreta-
angles. or they may form reversing dunes if winds tion of genesis and an understanding of subsurface
blow from nearly opposite directions. Where winds trends in various sand bodies. Whether the sand
are shifting around the compass they form star has advanced in a constant direction to form a sheet

Digitized by Google
 SEDIMENTARY STRUCTURFS IN DUNES 89
TABLE 9. - Classification of principal dune types according to form and to nlD'lber of
sJipfaces
Number of lliplac:el
None None
(no mound) (mound) One Two Multiple

Sheet. Dome.! Barchan. Linear (seifor Star.

Stringer. Barchan ridge. longitudinal). Blowout.
Climbing. Transverse ridge. Reversing. Parabolic.'
Falling. Retention ridge.
I JntenW atruc:ture may allow burled . . . . . .
I May have only one a1ipf_ developed.

DUNE TYPES
Simple: Dunes in the forms shown above.
Varieties: Dunes having differences in details of topographic form.
Compound: Dunes in which two or more of the same type are superimposed.
Complex: Dunes having two or more types combined.

or has built mostly vertically into a thick lens; at high angles (not uncommonly 30 -34) and
whether it has advanced in one or in more than one represent original angles of repose.
direction; and whether the accretion of additional 2. Sets of tabular-planar cross-strata that in verti-
sediment has been rapid or slow. therefore result- cal sequences in high dunes tend to be
ing in relatively pure sand or in much-contami- progressively thinned from the base upward;
nated sand - these are questions that may be sloping laminae within these sets tend to
answered by structural analysis. repeat the pattern of foresets displayed in the
Dunes that have been studied in detail from the thick basal set. but the angle of dip may be
standpoint of structure are relatively few. but lower (figs. 43.46.48).
trenching and dissection have been undertaken in 3. Bounding planes between individual sets of
the United States at White Sands National Monu- cross-strata that mostly are horizontal or dip
ment in New Mexico. at Great Sand Dunes National leeward at low angles; in high dunes addi-
Monument in Colorado. and at Killpecker dune tional bounding surfaces commonly form in
field in Wyoming. Similar studies have been made downwind parts. with dips at moderately
in Libya. in Saudi Arabia. and at several localities in high angles (20 -28). and these planes trun-
Brazil. Some investigations have been made in soli- cate foresets that dip at still higher angles
tary dunes. others in selected parts of large com- (28 -34; fig. 48).
posite dunes. where extrapolation was necessary to Various other structures apparently are charac-
reveal the complete structure. Many more studies teristic of only one or two dune types. Such
of certain dune types in various geographical situa- features will be discussed in the following pages
tions are needed before generalizations can be under each of the principal dune types in which
made with complete confidence. but some features they occur.
seem very definitely established already.
Sedimentary Structures
Sedimentary Structures Characteristic of
Common to Most Dunes Certain Dune Types
THE FOLLOWING STRUCTURES. on the basis Barchan and Barchanoid Ridge
of our limited present knowledge. are considered to
be common to most types of eolian dunes:
1. Sets of medium- to large-scale cross-strata that THE BARCHAN DUNE. one of the best known
typically consist of foresets dipping leeward and most elemental forms. is also probably the best-

Digitized by Google
90 A STUDY OF GLOBAL SAND SEAS

were covered by a veneer of laminae dipping up-

wind at angles of 5 _7 (fig. 42). On the wings,
variations from the normal trend of stratification
occurred as changes in direction and amount of dip.
The wings contained strata dipping with relatively
A B low angles, resulting from growth in locations of
low wind stability and at extremities of the slipface
surface.
~31 At White Sands National Monument, New Mex-
31 ico, one segment of a barchanoid ridge dune was
trenched and analyzed (McKee, 1966, p. 39). The
c D crescent-shaped mound studied was about 170 feet
(50 m) long, 290 feet (90 m) wide, and approx-
imately 27 feet (9 m) high. Trenching by a
bulldozer, in directions both parallel to and normal
to the prevailing wind, exposed to view cross-
lamination of the dune in three dimensions.
.
Wind direction The pattern of the barchanoid ridge structure as
recorded in the main trench consisted upwind of a
c D series of nearly flat-lying tabular-planar sets of
A -124ft- cross-strata, each about 3 or 4 feet (p.9 -1.2 m) thick
(38m)
and containing foresets that dipped downwind at
CROSS SECTIONS OF BARCHAN DUNE near Leupp, Coconino 26 -34 (figs. 43, 44). In the downwind part of the
County, Arizona, U.S.A., taken through crest in direction of trench, the bounding surfaces of each set changed
dominant wind. Modified from McKee (1957). A, 12 feet (3.7m)
from base, 72 feet (22m) from crest; S, 24 feet (7.3m) from base,
from nearly horizontal to steeply dipping (20 -28)
60 feet (18.3m) from crest; C, 36 feet (11.0m) from base, 48 feet and the foresets between them attained dips up to
(14.6m) from crest; 0, 48 feet (14.6m) from base, 36 feet 34 and lengths of 40 feet (12 m). A few low-angle
(11.0m) from crest; E, crest, on east side. 100 feet =30.5 metres. (2 ~O) windward-dipping laminae capped the
Scale of sections A-O=1 X2 feet. (Fig. 42.) dune, probably formed as windward-side deposits
at times when the wind was not competent to carry
recorded dune in terms of sedimentary structure. its entire load over the crest. Shallow, small-scale
The typical crescent shape, with horns pointing lenses of sand, scattered among sets of windward-
downwind, commonly develops as a single, isolated dipping cross-strata, apparently represent scour-
sand deposit; groups of barchans that coalesce to and-fill deposits of shifting winds.
form barchanoid ridges in a parallel wavy pattern Crossbedding, as seen in the walls of the trench,
are also common. at right angles to wind direction (fig. 448), was
A small isolated barchan dune located a few materially different from that exposed in the main
miles north of Leupp Trading Post, in eastern trench. A section cut straight across the crescent
Coconino County, Arizona, was dissected and its exposed, near its middle, crossbed sets and their
structures were studied in detail in 1951 (McKee, bounding planes that appear nearly horizontal. On
1957, p. 1721). The dune was a double crescent with the flanks both cross-strata and bounding surfaces
a height of about 11 feet (8 m) and a width of 270 dip outward at low angles as a result of curvature of
feet (80 m). Five test pits - three located from base the horns or wings; individual cross-strata sets form
to crest on the windward side and one on the for- wedges that taper toward dune margins, and
ward part of each horn - provided good insight foresets show apparent dips of 12 -23. Small
into the structure of a typical barchan. trough structures, probably formed as blowouts
Laminae in the middle part of the Arizona dune during fluctuations of wind regime as the dune was
consisted of long, even layers dipping consistently forming, were conspicuous near the dune center.
31 -32 downwind. In the lower part of the wind- Horizontal sections cut above the barchan dune
ward slope, the truncated tops of these laminae floor showed the curving lines formed by trunca-

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 91

SERIES OF STEEPLY DIPPING CROSS-STRATA forming tabular-

planar sets in barchanoid ridge dune. Strata tangential with basal
bounding planes. White Sands National Monument, New Mex-
ico, U.S.A. Photograph by E. Tad Nichols. (Fig. 43.)
BEVELED REMNANT OF BARCHAN DUNES in interdune area,
White Sands National Monument, New Mexico, U.S.A. From
McKee (1966). (Fig. 45.)

15

10

--- ---Base of dune

EXPlANAnON
10 20 JJfHT
r-~~I-'---+II---'I~' - - Bounding surface of cross-strata set
6 8 METRES
- - Selected cross-stratum SE

Base of dune

CROSS-STRATA IN BARCHANOID RIDGE DUNE, White Sands National Monument, New Mexico, U.S.A. A, Section parallel to
dominant wind direction; B, section normal to dominant wind direction. Modified from McKee (1966). Lines queried where struc-
ture was obscure in field. (Fig. 44.)

tion of laminae that had formed on the slipfaces of virtually the same sedimentary structures (fig. 46)
the dune. The overall pattern was one of festoons, as those at White Sands. A trench on the lee side
with one arc or crescent overlapping another (fig. revealed foresets dipping at high angles (26 -34)
45). to lee and a bounding surface dipping 4 in the same
Barchanoid ridge dunes of the Killpecker field in direction. Deformational structures consisted of
western Wyoming (Ahlbrandt, 1973, p. 38) showed fadeout laminae and overthrusts.

Digitized by Google
92 A STUDY OF GLOBAL SAND SEAS

w E

w E

FEET
EXPlANATION
Type of send

DRne

EJeoarse
D Red oxidized

!'- ./',/.1 M8gnet~e r~

o 200 FEET
f-I----II 22 Apparent dip.
o 61 METRES indegrea

BARCHANOID RIDGE DUNE, showing stratification to domi-

nant wind direction (arrow), Killpecker dune field, Wyoming,
U.S.A. A, Cross section of lee side, trench 1; B, cross section of
upwind side, trench 2; C, plan view showing trench locations.
Modified from Ahlbrandt (1973), 1 foot =0.3 m. (Fig. 46.)

A second trench in the Killpecker barchanoid

ridge, dug on the upwind side, exposed four sets of
cross-strata, all dipping leeward. In two sets the
dips were high angle (34), in a third they were BARCHANOID RIDGE DUNES and interdunes, White Sands
medium angle (18 -22 and, in a thin but persis-
0 0 ),
National Monument, New Mexico, U.S.A. Photograph by Hollo-
tent unit, they were low angle (8 at one end but
0)
man Air Force Base. (Fig. 47.)
steepened downwind to the angle of repose. The
low-angle strata contained small lenses with scour- sand. Such contrasts in texture apparently resulted
and-fill structures. from poor sorting in the source sand and the prox-
One feature of the barchanoid ridge in the imity of this source. In terms of textural parameters
Killpecker dune field, not recorded in other (chapter B) , interior desert dunes commonly are
barchans, was an alternation of coarse and fine more poorly sorted than coastal dunes because of
laminae and a filling of lenses with relative coarse normal differences in source material.

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 93
FEET
SW 5 NE 40

30

.........
WInd~ edge of dune 81 440 fII8I;
no 10PH18 81 340 filet ---_A--
~o,
......~ 10
......:
~ ....
A o 10I 20, 1I, FEET MAIN TRENCH, NORTH WALL -----_ ~::.:..:..:::::.:::..::~~~~:....:::..~
,
o
NW SE FEET

[~
EXPLANATION
Type of ..nd

SDE TRENCH, EAST WALL

E
B
W

m............. llliiil E
--Bounding
IUrface
- - Selected

32
eros.
ItnItUm
Apparent dip,
indeg .....

D
CROSS SECTIONS OF TWO TRANSVERSE DUNES in the United States. A,8,
White Sands National Monument, New Mexico. From McKee (1966, fig. 7). C,O,
Killpecker dune field, Wyoming. Modified from Ahlbrandt (1973). A, Parallel to
dominant wind direction. 8, Normal to dominant wind direction. C, Parallel to
dominant wind direction on windward side. 0, Parallel to dominant wind direc-
tion on lee side. 1 foot~>.3 metre. (Fig. 48.)

Transverse the smaller. much more irregular barchanoid types.

A typical example selected for sectioning and study
Transverse dunes resemble barchan dunes in was about 40 feet (12 m) high. 400 feet (120 m)
being a product of unidirectional winds. with steep thick. and had a nearly straight crest extending for
slipfaces oriented normal to wind direction. The more than 800 feet (240 m) at right angles to the
transverse dune apparently evolves into the wind direction. The general straightness of this
barchan downwind with an intervening intermedi- crestline was the feature that especially dis-
ate form. here referred to as a barchanoid ridge. tinguished this dune type from adjacent barchanoid
The barchanoid ridges consist of parallel rows of ridges and was reflected in the internal structures
crescents or barchanlike forms that have combined of transverse dunes.
to form parallel wavy sand ridges (fig. 47). Inter- Crossbedding in the main trench at the White
dune areas. forming long corridors. occur between Sands transverse dune. cut parallel to the dominant
pairs of ridges. The barchanoid ridge and transverse wind direction (fig. 48). consisted largely of tabu-
types of dunes. although similar in general form. lar-planar sets of steeply dipping (30 -34) strata
differ in internal structure. within bounding planes that ranged from nearly
Transverse dunes at White Sands National horizontal to moderately high angle (20 -26)
Monument. New Mexico. are the highest dunes in downwind. Foresets in the lower part of the dune
the area and form long parallel ridges. upwind from were very large. some strata extending for 55 feet

Digitized by Google
94 A STUDY OF GLOBAL SAND SEAS

(16 m) downdip and terminating as tangential does not seem to have been recorded as yet. Judging
curves at the dune base. Crossbed sets in the upper from the characteristic shapes and distribution of
part were relatively thin. mostly 2 ~ feet (0.6 -1 .0 blowouts. one might expect a series of scour-and-
m) thick. with nearly parallel bounding planes. fill structures to develop; however. this is yet to be
Perhaps the most diagnostic feature of the determined by trenching techniques.
transverse dune structure was the great extent of
horizontal. or nearly horizontal. parallel laminae
formed by the apparent dip of strata as seen in cross
sections cut normal to wind direction (fig. 48).
Some cross-strata in these sections dipped at low
angles. and a few sets appeared as cut-and-fill (fes-
toon) structures with axes parallel to wind direc-
tion. but high-angle foresets were absent.
A transverse dune at Killpecker dune field.
Wyoming (fig. 48). which was trenched and
analyzed by Ahlbrandt (1973. p. 36). showed most
of the same features. Like the White Sands dunes. it
was located downwind from dome dunes and to
windward of barchans. It was formed largely of
tabular sets of strata dipping steeply (29 ~4)
0

downwind and bounding surfaces. that is. slipfaces.

dipping as much as 210 . Water saturation in parts of
the dune had made the sand cohesive. and over-
thrusts had developed. In dry areas fadeout laminae
were abundant. Another feature of the dune was
the presence of lenticular zones of coarse sand
among laminae of fine sand. probably the result of
scouring and later infilling among strata with low-
angle dips on the windward side.

Blowout
In many coastal areas and others where sand
becomes stabilized by vegetation cover. moisture
BLOWOUT DUNES in the United States. A, South shore of lake
content. or both. a common dune form is referred to Michigan; photograph by R l. Gutschick. B, East of Salton Sea,
as the "blowout" and. in general. appears like a California; photograph by John S. Shelton. (Fig. 49.)
crater (fig. 49) . Blowouts differ widely in size and
shape partly because their rims are anchored in so
many places that even in regions of a dominant Parabolic
wind direction. prevailing winds are deflected in
many directions. and local sand resistance to defla- A parabolic dune is U-shaped or V-shaped and
tion and transport is extremely variable. represents a type of blowout in which the middle
A circular blowout dune may evolve into a U- part has moved forward with respect to the sides. or
shaped parabolic form where a prevailing wind can arms (fig. 50) . This kind of dune occurs in many
cause the advance of an unstable section of the rim. parts of the world but is especially prevalent in
In this manner. the nose of a parabolic dune may coastal areas and along margins of desert regions
develop and migration of this dune type can be in- where vegetation has obtained a foothold. A dis-
itiated. The internal structure of parabolic dunes tinctive feature is the anchoring of the dune arms
has been examined and is described on succeeding by plant growth. which causes the entire dune to be
pages. but the stratification of normal blowouts relatively stable. The middle. or blowout. part of

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 95

AERIAL VIEWS OF PARABOLIC DUNES at White Sands National Monument, New Mexico, U.S.A A, Compare prevailing southwest
winds (in knots) shown by wind rose with orientation of dunes. Photograph by Holloman Air Force Base. B, Vegetation anchoring
arms of dunes. Orientation of dunes same as in A. Individual dunes, including the dragging anns, average about 700 feet (215 m)
long in this area. (Fig. 50.)

the sand mass commonly migrates slowly, forming Bigarella, 1972, p. 672); these compound dunes are
a rounded nose that points downwind. very complicated structurally.
A parabolic dune was trenched at White Sands
Parabolic dunes vary greatly in size. In the Thar
National Monument, New Mexico (fig. 51). In its
Desert of India and Pakistan, they commonly are
very large, individual dunes extending several lower part this dune shows cross-strata with high-
miles downwind, the U-shapes readily discernible angle foresets as in barchanoid ridge and other
in Landsat (ERTS) images. Elsewhere, for example, dune types of the area, but in the upper part,
at Cabo Frio, on the Brazilian coast (McKee and especially on the upwind side, many of the strata
Bigarella, 1972, p. 672), parabolic dunes are only a are very low angle or are nearly horizontal and in
few hundred feet long, and among the Indiana thin sets. Deposition extended from a nearly flat
dunes on the shore of Lake Michigan, they range in top, over the crest, and down the slipface. Its dip-
length "from a few tens of feet to several ping laminae, except near the top, are in sets mostly
2 -5 feet (1-2 m) thick of tabular-planar or wedge-
hundreds" (Cressey, 1928, p. 6).
planar form. A few scattered sets of symmetrically
Because parabolic dunes are largely the product filled trough cross-strata - about 25 -30 feet (7 ~
of unidirectional winds, although secondary move- m) wide and 3 -4 feet (1 m) deep - also occur near
ments may be to both sides, their structures include the top.
many features common to all dunes that are domi- The White Sands parabolic dunes exhibit two
nated by one major wind direction. Their steep structural features seemingly unique to this dune
depositional faces (slipfaces), which form on the type: (1) many foresets are concave downward (fig.
dune nose, have foresets consistently dipping 51), probably resulting from crosswinds that under-
downwind; however, strata in the arms dip normal cut and oversteepened the bases of foresets on the
to the dune axis, both right and left. Various com- dune nose, and (2) organic accumulations that
binations of parabolic dunes occur, such as series locally warped strata, result from root growth,
with one dune inserted inside another (common especially along the bedding surfaces.
among dunes of the Thar Desert, India-Pakistan) The parabolic dune examined at White Sands
and reversing types as at Lagoa, Brazil (McKee and showed an unusually wide directional spread of

Digitized by Google
96 A STUDY OF GLOBAL SAND SEAS

Lagoa dune field in Santa Catarina. Brazil. has dis-

closed many additional features discussed later in
06~ this chapter.
SW 'l:-f}o~ NE
Approximate
Intaraection with aide ~~~ ____r-'_A=tpproc...:o:.:..;;:":.:.u:.:.;:e_
Parabolic dunes. whether inland or coastal. seem
top of dun
to characteristically migrate very slowly (McKee
and Douglass. 1971. p. D113). In most places they
are anchored by vegetation and in coastal areas.
A especially. they are partly stabilized by moisture. In
Brazilian retention dunes. which are in a similar
coastal environment. "the supply of sand from the
source area [beach] seems to be large enough to
overcharge the gentle slope of the dune's surface
* * * but the winds are not so intense nor the sand
so dry that the complete removal of the previously
deposited beds can be accomplished * * * [so] a new
set of strata just barely covers the former one"
(Bigarella. Becker. and Duarte. 1969. p. 38).
The partly stabilized coastal dunes of Brazil also
show virtually concordant attitudes between slip-
EXPLANATION face sets. The difference in angle between strata of
c. - - Bounding surface various sets is small despite disturbances from
~
15 FEET
slumping along slipface surfaces. according to
:0 Are. Hluatr.t~~ -- se=~m croll- Bigarella. Becker. and Duarte (1969. p. 38).
~"tj06" 22 Apparent dip,
O ~ in degr_ Dunes of normal parabolic shape are well
SE ~ developed at the high altitude and under the cool
~ch NW

~
dry climatic conditions of western Wyoming. Their
structures have been examined by Ahlbrandt (1973.
p. 39 -41). One dune of this type is described as
o 5 10 15 2D FEET
having foresets dipping 24 -34. commonly
D I I I I I
26 -30. Internal structures are disturbed by plant
VARIOUSLY ORIENTED SECTIONS through a parabolic dune, roots. and laminae of the dune nose exhibit minor
White Sands National Monument, New Mexico, U.S.A.
Modified from McKee (1966, fig. 9). A, Parallel to dominant
folding. A bulging of the dune nose is reflected in
wind; B, C, parallel to wind direction, showing high-angle the convex-upward form of the laminae as seen in
foresets; D, normal to dominant wind. All measurements are in the test trench (fig. 52). Thus. the Wyoming dune
feet. 1 foot =0.3 metre. (Fig. 51.) displays the same principal structures observed at
White Sands and in the Brazilian coastal dunes.
crossbedding dips (McKee. 1966. table IV) - a dis-
tinctive feature of this dune type. Seventy-three
Retention Ridges
readings of true dip in foresets showed a spread of Retention ridges. common features of many
200. and 161 readings of truncated foresets in adja- coastal dune areas. characteristically form long nar-
cent interdune areas showed a spread of 210. row dunes parallel to the shore on the landward
Barchanoid ridge. transverse. and other dunes up- margins of backshore beach deposits. They develop
wind at White Sands showed much smaller where dune sand accumulates against vegetation
spreads. ranging from 60 to 140. barriers. forming a nearly stabilized type of dune
Structures of parabolic dunes on the coastal plain that may gradually be enveloped by trees and other
of Brazil at Cabo Frio (McKee and Bigarella. 1972. p. plants. This type of dune ridge is common along the
672. fig. 4d) resemble those at White Sands. Recent west coast of the United States and in that area has
extensive trenching. by Bigarella. of a large been referred to by Cooper (1958. p. 55; 1967. p. 22)
parabolic dune with smaller ones inside at the as "precipitation ridge."

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 97
W E structural characteristics of parabolic dunes. "Con-
FEET Plant vex-upward strata * * * are rather frequent in the
dunes studied." and distinctive types of distortion
on many of the slipface surfaces were common
features (Bigarella. Becker. and Duarte. 1969. p.
37. 45).
sand A distinctive feature of Brazilian ridge dunes
zone
(beach dune ridges) with high moisture content is
that "both the topset and foreset are made up by the
same continuous set of strata. and not by different
truncated sets" (Bigarella. Becker. and Duarte. 1969.
p. 36). Probably the abundance of sand supply and
the slow rate of movement (because of internal
dampness) are responsible for this feature. suggest-
ing that it results from the humid climate. rather
than the type of wind. Sinuous sand layers. of
medium scale. that occur in lower parts of the
Brazilian retention ridges (fig. 53C) are similar to
strata in small dome dunes that commonly occur to
seaward. Above the wavy basal beds in the ridges
examined were tabular-planar sets of cross-strata.
In some of these sets the bounding surfaces dipped
~
at low angles. and in others. at moderate to high

\ angles (fig. 53). The tabular-planar sets included

both nonerosional (simple) and erosional types;
few trough-type cross-strata were associated.
Some cross-strata in the retention ridge dunes of
Brazil dip downwind at very high angles (34 -39)
probably owing to high humidity at the time of
deposition (Bigarella. 1972. p. 21). In sections nor-
mal to prevailing wind direction. however. most of
100
I
150
I
200
,
250 FEET
, the stratification is horizontal or nearly horizontal.
except across protuberances in which strata dip at
low angles toward the outer margins. giving a
NORTH WALL OF A TRENCH on the nose of parabolic dune, strongly curving appearance in section. Sets of
Killpecker dune field, Wyoming, U.S.A Strata are concave cross-strata tended to be realtively thick (some
downward in lower two sets. From Ahlbrandt (1975, figs. 8, 14).
=
Measurements in feet. 1 foot 0.3 metre. (Fig. 52.)
more than 3 feet (1 m and to have relatively high
angles of dip near the dune base. but in the upper
part they were thinner (0.32 -0.98 foot (0.1-0.3 m))
Primary structures in retention ridges of Brazil and had lower angles of dip.
were studied in detail by Bigarella. Becker. and
Duarte (1969. p. 28 -29) and by Bigarella (1972. p. The degree and direction of dip from several
20-21). Trenches were dug and cross sections ex- dune sections of the Guairamar dune are illustrated
amined at Jardim Sio Pedro (fig. 53). Guairamar, by stereonets (fig. 54). The spread of high-angle
and p&rto Novo on the coast of Paran'- in Brazil. (>30) foreset dip directions at Jardim sao Pedro is
These dune ridges formed under conditions of high more than 170 (Bigarella, Becker. and Duarte.
humidity and considerable vegetation and included 1969) which is very much greater than the spread of
many "blowout" features. They apparently ad- barchanoid and transverse dunes at White Sands.
vanced very slowly and contained most of the New Mexico, in the United States.

Digitized by Google
98 A STUDY OF GLOBAL SAND SEAS

METRES A
3

2
SCALE
FOR
A.B.C

RETENTION DUNE CROSS SECTIONS showing details of stratification, Jardim sao Pedro, Praia do leste, Paran~, Brazil. A, Section
parallel to prevailing wind; B, section normal to prevailing wind at linguoid protuberance where crest and slipface are curved; C,
section normal to prevailing wind at middle part of dune. Modified from Bigarella, Becker, and Duarte (1969).1 metre= 3.28 feet.
(Fig. 53.)
-39'
N
~~
Dome
Dome dunes develop in the upwind parts of some
dune fields, where winds are sufficiently strong to
bevel and, thus, effectively retard a normal upward
growth of dune crests. Three localities where dunes
of this type are characteristically developed and
where they have been trenched in order to deter-
mine the primary structure pattern are (1) White
s N s Sands National Monument, New Mexico, V.S.A., a
desert environment, (2) the Killpecker dune field,
Wyoming, V.S.A., a high-altitude, cold climate, and
(3) along the southern coast of Brazil, a warm,
humid environment.
The dome dunes at White Sands National Monu-
ment, New Mexico (described in detail by McKee,
1966, p. 26, 27), are round or oval in ground plan,
have rounded tops, and do not display steep
s avalanche faces on their lee sides. One dune - 490
STEREONETS FOR THREE SECTIONS showing angle (distance
feet (150 m) wide, 460 feet (140 m) long, and 20 feet
from center) and direction (azimuth) of dip on foresets of reten- (6 m) high - considered to be typical of that area,
tion ridge, Guairamar, Parana, Brazil. Modified from Bigarella, was selected for trenching and study. Its texture
Becker, and Duarte (1969). (Fig. 54.) showed poorer sorting and a higher percentage of

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 99

laminae, capping the rounded dune top, formed

topset beds that connected with dipping foresets
downwind. Both features distinguish dome dune
structures from those of other dune types in the
area.
Dome dunes in the Killpecker dune field of
western Wyoming, like those at White Sands, occur
farthest upwind of all dunes, and are developed
under the full force of approaching sand storms.
Three trenches dug by Ahlbrandt (1973, p. 32) in a
Killpecker dune revealed structures generally simi-
lar to those at White Sands. In sections parallel to
wind movement, strata dipped consistently down-
wind at the angle of repose (about 34) to form
tabular-planar cross-strata, except on the lee side,
where fractures (fig. 57) were developed from
incipient slipface movement. These fractures
defined slump blocks which locally were responsi-
ble for great oversteepening of beds, some of which
dipped as much as 55. The steepness of dips on
such beds is attributed to the accumulation of snow
TRENCH THROUGH DOME DUNE, White Sands National layers within the deposit with subsequent melting
Monument, New Mexico, U.S.A Section parallel to dominant
and collapse of overlying cross-strata.
wind from left. Scale is in inches. 1 inch = 2.54 centimetres. (Fig.
55.) Other features of the dissected dome dune in the
Killpecker field were (1) beds dipping 1 -40 to
coarse grains than did the other types of dunes lo- windward in the upper 1.5 feet (0.5 m) on the up-
cated farther downwind. wind side, apparently resulting from an overabun-
Structures in the upwind part of the White Sands dance of sand supply, and (2) slump structures on
dome dune consisted largely of tabular-planar the downwind side consisting of many normal
cross-strata (fig. 55) with foresets dipping down- faults and a few fadeout laminae, typical of
wind at high angles (28 -33) . In the downwind avalanches in many dune sands (McKee, Douglass,
part, foresets dipped somewhat less (14 -27), and Rittenhouse, 1971, p. 370 -372).
perhaps because much of the sand there had been Dome dunes on the coasts of Paranti. and Santa
deposited from suspension rather than mainly from Catarina in Brazil occur as miniature foredunes,
slumping (fig. 56A). In a trench normal to the domi- seaward of retention ridges that parallel the coast.
nant wind direction, most of the stratification ap- They resemble dome dunes of the desert areas of
peared to be horizontal or to have very low dip the Western United States in that they also form in
toward the dune margins (fig. 56B). Thus, in basic unobstructed upwind locations and develop with
structure, this dune type resembles other unidirec- oval or circular shapes and with relatively flat tops.
tional wind deposits at White Sands. The dome They differ from the desert dome dunes in their
dunes probably began as isolated barchans, if shape smaller size - typically less than 1 metre (3 ft)
and structure can be used as criteria. high and 12 -14 m (40 -45 ft) in diameter - and in
Two features of the dome dune seem to be dis- the type of structures and large spread of dip direc-
tinctive and are either absent or not well developed tions of their stratification. The differences proba-
in other types of dunes in the area. First, cut-and- bly result mainly from contrasts in depositional en-
fill structures - probably the result of trough- vironment. In coastal Brazil, the domes are formed
forming deflation by strong, unchecked winds, under humid conditions and on a surface
followed by sand deposition under weaker roughened by many grasses and small shrubs.
winds - formed a series of sand-filled scours Structures in Brazilian dome dunes have been ex-
parallel to wind direction. Second, in the same sec- amined and described by Bigarella and Popp (1966,
tion a layer of horizontal or nearly horizontal p. 143 -148) at Barra do SuI, by Bigarella, Becker,

Digitized by Google
 MAIN TRENCH, NORTH WALL Projection of .ide trench ....
------to\~rc;l\ 8
w.11o on mein tr.nch wall
'-----.,
,D -------l
"'---- -t;
-::: Iff'
--'-
w ~I Windward surface
';.Gl

~
'-, ---<::: ~-"
Z;" '" "-~uo-'
<7<"-' "'" - ~ ~ ~
-.
r----:-
'-'"
~,-'''::: ~ ~ :::::::........
f trench
Concealed"
~'::o-...-
-'"-- ~

41)
.....

e of tre
141)
B_ 7
30 50 60 70 BO 00 100 110 120 130 150 160 170 1111 100
A
Iswale
FEET
Lee surface dune surface

15j
10 ~
'~5~8~2~1
.~.
-.. =-._
~ -== ___ 7 --..!...
~='=~E=.1====~/=-L",ee=,u::-rfa-...ce
~=
~
~--=--
5
=:::::::--
~i~
~e:::-._
~-~ -..
---
--.::.:;
:--------..
--=======
~
~
" - ; --- .:.,~"
---
====
~
~
--
~
---
==--
=::::::::
"l::: -
~
=--===
~
~
)~
===
--=:::::::::
~
=
- ~
~.:::::;;-::::
~9-"
~
~
-7_
:.;=
s=::::
<::::::;;::::
ee surfa
~
E

o I 'Concealed sas;of trench - - - B... of trench Concealed? ..... Ba.. 01 trench

1!Il ZIll ZIO 120 Z30 Z4I) Z50 Z60 Z10 ZBO Z!Kl 300 310 320 330 340 350 3BO 370 JIll 3!Kl 400 FET
I J I J I I I I I I ! I I ! I I I I I I I I

con d

~
6
SIDE TRENCH, WEST WALL Lower 4 feet of wall cont.in. en

~
5 numerous horizont.1 .nd w.vv. N
locally diacordant I.minae
Non
Comm garippl ddition lar aurfac _-3= ::=L w.1I
r -_____________iP_p_l_e_s__ esurfa ~~ ~ ~ _____

'I ~ :':chl en
f':
r Dune surface of _of 0
~Throwout ramp surface __ ==---::-. .- ~~ -------~t 7 '-=-':::;2 7~.. _ . ""
i}.YY-"~ -3~-:...3~-- - - -3- --= ___ _ _ _ '--;:,--~ 1- " ~ ----

f ~ ======' B. Inch....... - --, ~ of trene

C ed a.. 0 41) 10 FEET
100 80 160 50 130 20 IIII 70 I I
t I I J I I I

EXPlANATlON
ing su . Select ross-st 21 rent d degr_

CROSS SEOIONS OF OOME DUNE White Sands National Monument New Mexico, U.S A. A) pa II I to do nt wi d d'rectio d(S) al to 'nant
wind ection odifi om M (1966 ,6). meas ents feet feet met lines ueri ere s ure is cure.
v (Fig. 56.)
,
~
 SEDIMENTARY STRUCTURES IN DUNES 101
w E

FEET
B

w E

.. .. : .... ...: ............ '. ' . ' . ' .' .. '

.4 ' :' .: .. ...-: ':':" ':'::'.....: .... '~?':.:.:':::'.':: '.':.,,:': .::... .'
o 1 2 3 4
FEET
A
EXPLANATION

~ Selected cross-strata in fine sand

t.J.)j Medium sand

b:}~:~r::J Interlayered medium and fine sand

~ Fracture
B
- - Erosion surface prevailing
~ n
wind directlO
_ _ _ Surface of planation
o 100 FEET

~
'L _---',

Selected cross-strata

DOME DUNE, Killpecker dune field, Wyoming, U.S.A. A, Stratification on steep lee side in section parallel to wind direction. Note
fractures defining slump blocks caused by snowmelt and collapse of strata. B, Section near crest showing lee-dipping strata overlain
by horizontal strata. C, Section upwind, showing low-angle capping beds dipping upwind. 0, Ground plan of dome dune, showing
location of sections A, B, and C. Modified from Ahlbrandt (1973). 1 foot =0.3 metre. (Fig. 57.)

and Duarte (1969. p. 17) at Praia do Leste. and by and lower parts of desert domes. are lacking. and
Bigarella (1972. p. 24 -25) at Barra do SuI. They con- the characteristic sets of low-angle strata. dipping
sist. in general. of numerous sets of superimposed. in many directions. apparently represent wind-
curving strata in which bounding planes and many ward-side deposits. topsets. and also foresets; few
laminae are convex upward (fig. 58). Uniformly avalanche deposits are recognized. In brief. the
dipping. high-angle foresets. such as form the cores domes consist of blanket deposits resulting from a

Digitized by Google
102 A STUDY OF GLOBAL SAND SEAS

NW
EXPlANATION

~ Bounding surface
--=-~ --:-
0.5
- -::- ~~_ Selected cro....strata

2 3 5 6
A
NE SW
SW NE 1.5

.--
-- -
r.:::::---__
.--. --

li-iliilli~~~--~~iI~~~~iiiiliil
1 =---::-

0.5

0~~~~~~~~~~~~~~~~~1
o c 2 3 4 5

CROSS SECTIONS OF COASTAL DOME DUNE, Praia do leste, Parana, Brazil, showing (A) stratification in section parallel to domi-
nant wind direction, and (8, q stratification in sections normal to dominant wind directions. Modified from Bigarella (1972). All
measurements are in metres. 1 metre = 3.28 feet. (Fig. 58.)

large source of material and a wide spread of winds

that alternately deposit and erode sand on an ir-
regular surface.
Brazilian coastal dome dunes are characterized
by strata with "dip directions * * * distributed com-
pletely around the compass" (Bigarella, 1972, p. 25).
This is partly the result of surface irregularities
from vegetation and partly from retarding effects of
moisture. Both upwind dips and nearly horizontal
topset beds are also developed in domes of desert
sands, but on a larger scale. They occur as blankets
above normal foresets of tabular-planar sets. In
both coastal and desert environments a large sand
source is available, but in the humid environment
with abundant vegetation, accumulation of sand is AERIAL VIEW OF SElF, or linear, dunes, north of Sebha, libya.
Dune ridges are about 50 feet (15 m) high. Distance between
irregular, in time and direction, and associated dunes ranges from 1/4 to 1/l mile (0.4 to 0.8 km). (Fig. 59.)
blowout features are common.
may be distinguished, according to Price (1950, p.
Linear 462), which doubtless is one reason for the current
One of the most common but least understood confusion regarding its genesis. Differences in
types of desert dune is the long straight ridge herein details of shape likewise occur in various sand seas,
called a linear dune (fig. 59). It is commonly refer- and many varieties are recorded, including the
red to as a seif dune in North Africa and Saudi feather-type in Saudi Arabia, the zigzag type in
Arabia, as a sand ridge in Australia, and as a Libya, and the converging, nonparallel type with
longitudinal dune in many other places; its pattern, junctures in Australia.
as observed on Landsat (ERTS) images (McKee and An environment favorable to development of
Breed, 1976, p. 81) has been termed "parallel linear dunes seems to be one in which there is high
straight." At least three subtypes of this dune form wind velocity (Glennie, 1970, p. 95; Bagnold, 1941,

Digitized by Google
 SEDIMENTARY STRUC11JRES IN DUNES 103
p.171; Twidale, 1972, p. 82). As pointed out by Bag- Probably the first reconstruction of linear dune
nold (1941, p. 178), "a strong sand-laden wind with structure was that by Bagnold (1941, p. 242) in
uniform drift over a uniform rough surface has which he showed a nearly symmetrical, hypotheti-
transverse instability so sand tends to deposit in cal section based on "the few tests already made by
longitudinal strips." According to Glennie (1970, p. the author." This section (fig. 80) shows the "ar-
95), "The higher wind velocity, the larger the seif rangement of accretion and encroachment
dune and the greater the interdune spacing." deposits," with the middle and upper parts of the
dune composed of avalanched sand, and the low
The origin of linear dunes has long been a matter
outer margins, or plinth, on both sides, formed of
of controversy; many geologists (Bagnold, 1941, p.
accretion deposits. The section shows growth of a
223; McKee and Tibbits, 1964; Twidale, 1972, p. 103;
perfectly regular type, but, as Bagnold pointed out,
and others) believe them to be the product of two-
successive stages in nature doubtless are very ir-
directional winds, but others attribute them to a
regular. In any event, deposition probably alter-
unidirectional or prevailing wind (Dubief, 1952;
nates from side to side by avalanching down the
Glennie, 1970). Even among the geologists who
steep slipfaces.
adhere to one or the other concept of directional
control by wind, there are many differences in the
mechanisms by which this dune form is attained.
Some linear dunes are explained as forming from
the trailing arms of parabolic dunes that have
become separated from dune noses (Dubief, 1952;
Capot-Rey, 1943; King, 1960; and others); other EXPLANATION
linear dunes are shown to have formed by elonga-
tion of one hom of a barchan by winds oblique to
D Avalanche deposit D Accretion deposit

the prevailing transport direction (Bagnold, 1941, p.

223). An explanation that calls for the development HYPOTHETICAL SECTION, showing structure of seif dune.
Modified from Bagnold (1941, fig. 83). Note foreset dips in op-
of wind cells in interdune areas through pressure posite directions. (Fig. 60.)
gradients along the dune sides was proposed by
Glennie (1970, p. 90), and experiments with run- Confirmation of the validity of the basic structure
ning a roller across a layer of grease suggested to pattern as determined by Bagnold is found in a
Folk (1971b, p. 3464) a theory of helicoidal air flow. series of trenches made by McKee and Tibbits
Recent work on Australian dunes led Twidale (1964, p. 17) across a "seif dune" near Sabha, Libya,
(1972, p. 103) to the conclusion that "longitudinal and by laboratory experiments conducted in
dunes are being initiated from leeside mounds bor- Denver, Colorado, U.S.A., in which a small linear
dering playas and alluvial plains" and that "these dune was developed. In the section across the Li-
mounds cause changes in the airflow in a bi-direc- byan dune (fig. 61), high-angle cross-strata
tional regime." (26-34), presumably slipfaces, show at many
Regardless of the mechanism by which linear places in the upper half of the dune and on both
dunes are formed, their structures make a distinc- sides; low-angle (4 -14) cross-strata occur on the
tive pattern very different from those of barchan- lower flanks of the dune and probably represent ac-
oid-type dunes. In linear dunes that have been ex- cretion beds.
amined thus far, slipfaces develop on both sides of
the crest, and foreset dip directions form two Reversing
clusters of points in nearly opposite directions,
rather than a single grouping. Unfortunately, few The existence in central Asia of dunes formed as
detailed studies based on trenching have yet been a result of reversing winds was noted by Sven
made of these dunes, so examples in different sand Hedin in 1896 (p. 267 -268). Likewise, dunes of this
seas that appear similar in size and form may not type were described and discussed by Cornish in
actually be alike structurally. 1897 (p. 286). Neither man had examined or

Digitized by Google
104 A STUDY OF GLOBAL SAND SEAS

o __
......
5 FEET
----"~

Scale for inset boxes

FEET

~
60

45
~'>-" 17-23

30 7

15

0 6
I~ WOffET ------------------------~~~
S EXPLANATION N
Structurele.. sand ~ Near-vertital laminae 13 Apparent dip.
~ between slip planes In degrees

CROSS SECTION OF SElF DUNE near Sebha, libya, based on series of test pits. From McKee and Tibbitts (1964, p. 17). All measure-
ments are in feet. 1 foot =0.3 metre. (Fig. 61.)

analyzed the internal structures of the dunes, but dunes, a considerable number of low-angle laminae
Cornish's sketches (1897, fig. 7) clearly indicate that probably formed as windward-slope deposits,
that steep slipfaces had developed on opposite sides have been buried in the lower parts during rever-
of a sand hill with re-forming of the dune top dur- sals of wind direction. Contorted structures com-
ing a wing reversal. Another early reference to monly result from slumping during the frequent
reversing dunes is by Bagnold (1941, p. 242), who reversals.
stated, in discussing dunes in North Africa:
"Reverse slip-faces formed during bouts of wind Reversing dunes about 30 feet (9 m) high occur in
from opposite directions can be distinguished at Antarctica, where approximately 30 barchans of
once by direction of the sloping shear-planes." reversing type were reported from Victoria Valley
Reversing dunes at Great Sand Dunes National by Lindsay (1973, p. 1799). His descriptions of these
Monument in southern Colorado in the United dunes indicate that they are much like reversing
States were examined by Merk (1960, p. 127) with dunes elsewhere; however, they have frozen cores.
regard to internal structures. From excavation of a Slopes developed on windward sides are said to dip
few test pits, Merk showed that the bases and lower an average of about 6 in their basal parts, but are
parts of the large dunes indicated deposition by a much steeper (approximately 15) in the upper
prevailing southwesterly wind, whereas the upper parts; the zone of change apparently coinciding
parts commonly re-formed with steep slipfaces with the lower limit of surface moisture. Avalanche
toward the southwest, during periods of easterly slopes are reported to range from 29.5 to 33, which
storm winds. Thus, in a single dune of barchanoid is normal for most dunes elsewhere; this steepness
or transverse type (fig. 62), two groups of steeply readily differentiates the slipfaces from other sur-
dipping foresets, facing nearly opposite directions, faces. Below the slipfaces are relatively low-angle
were developed (fig. 63). Further, in many large (14 -16) slopes referred to as "aprons."

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 105

Primary w i n d -
- Reversing wind
STRUCTURE IN REVERSING DUNE, parallel to dominant wind,
Great Sand Dunes National Monument, Colorado, U.S.A. Based
on photograph by Merk (1960). Primary wind from left to right;
reversing wind from right to left. Set A, A', primary dune foreset
laminae; set B, modified dune windward laminae; set C, pri~ary
dune foreset laminae; set 0, modified dune windward laminae.
(Fig. 63.)

seasonal shifts in wind direction, was dissected and

examined by McKee and Bigarella (1972. p. 677). In
this dune. steep foresets, formed on slipfaces that
dip in two nearly opposite directions, are con-
spicuous. Both stairstep folds and thrusts, charac-
REVERSING DUNES, Great Sand Dunes National Monument, teristic of avalanches in wet. strongly cohesive
near Alamosa, Colorado, U.S.A. Higher dunes are as m~ch as sand. are abundant (McKee and Bigarella. 1972. fig.
700 feet (210 m). Aerial photographs by J. S. Shelton. (Fig. 62.) 3).
The elongate ridges of the Kelso dunes in the Mo-
The attitude of the cross-strata and the structure jave Desert of California probably have received
patterns of the reversing dunes in the Kelso dune more detailed study than any other reversing
field, California. U.S.A.. as determined by Sharp dunes. They lie in a valley nearly surrounded by
(1966. p. 1062) illustrate several distinctive desert ranges but are exposed on the northwest
features: (1) High-angle foresets of lee-side side. from which additional windblown sand is in-
deposits are relatively scarce. suggesting that the troduced. During an interval of 15 years. measure-
"steeply dipping {>30 slipface deposits are rarely
0}
ments were made and data kept at 10 stations
preserved within these dunes." {2} Lee-side bed- {Sharp. 1966}; sand movements were recorded, and
ding which underlies a "veneer of windward changes both in dune form and in the facing direc-
deposits" dipped leeward 10 -15 in many pits. in- tion of dunes were documented. Although no
dicating that most of these strata are from the lower trenching was done. dune movements were
part of foresets. (3) "Bedding relationships are more measured precisely with respect to 10 series of fix-
complex and thickness of units greater in the ed poles. and the nature of dune structures was
crestal areas" presumably because of frequent determined by digging a large number of pits, 4-5
reversals in asymmetry and "back-and-forth move- feet {1.0-1.5 m} deep, at various locations.
ment of the crest."
At Lagoa. Ilha de Santa Catarina, on the southern The most frequent winds in the Kelso dune field
coast of Brazil, a large reversing dune. caused by are from the west. but winds from other quadrants

Digitized by Google
106 A STUDY OF GLOBAL SAND SEAS

are believed (Sharp, 1966, p. 1049) to have higher control (especially mountain ranges) of wind
velocity and, therefore, to be nearly equal in sand- movements seems to dictate the type of dune pat-
moving capacity. In any event, a summary of many tern developed. In brief, most, if not all, of the
measurements shows that the dune ridges moved reversing dunes cited are in areas where the wind
northeast at the slow rate of only 12 feet (3.6 m) in pattern is locally but strongly affected by physical
12 years but that during that time "the crest shifted barriers.
back and forth within a zone about 30 feet (9 m)
wide." Thus, "scouring and truncation of dune Star
crests occur when a sharp crest is subjected to a
powerful reverse wind" (Sharp, 1966, p. 1050) . Star dunes (fig. 65), apparently the result of wind
At White Sands National Monument, New Mex- from several directions, are among the most com-
ico, U.S.A., where dunes are mostly the result of plex of dunes, from the standpoint of primary struc-
unidirectional winds, cross-strata resulting from ture. Two representative examples of such dunes,
reversal in wind direction were preserved in some 100 feet (30 m) and 35 feet (11 m) high, in central
cross sections (fig. 64) parallel to the dominant Saudi Arabia, near Zalim, were examined and
wind direction (McKee and Douglass, 1971, p. 0114, described (McKee, 1966, p. 65) .
fig. 3). Weather records show that in this area the The larger star dune (fig. 65), with four arms or
principal wind direction normally shifts from ridges extending out from a high central crest, indi-
southwest to north for a short period each winter cates by its structure a shifting in the direction of
season, and this shift is recorded in the dune struc- wind movement. Each arm has a slipface on one
ture. side, with characteristic high-angle dips (3r ~3)
To summarize the relatively scant data available, and a more gentle slope (17 -21 on the other (fig.
0 0
)

reversing dunes are characteristically developed 66) . Thus, a seasonal rotating wind direction proba-
under many different climatic regimes. They occur bly was necessary to initially establish the positions
in arid, low-altitude areas (Kelso dune fields, of these arms. Confirmation of a shift in direction of
California, U.S.A.) ; in humid, low-altitude areas wind was found in cross sections exposed in
(Santa Catarina, Brazil); cold, high-altitude areas trenches at the dune crest. These sections showed
(Great Sand Dunes National Monument, Colorado, cross-laminae dipping at high angles in three prin-
U.S.A.); and cold, low-altitude areas (Antarctica). cipal directions which coincided with directions of
Because their development seems to be basically a seasonal wind movement in the area.
function of contrasting wind directions, regardless The smaller star dune that was trenched and
of other environmental factors, the topographic studied near Zalim, Saudi Arabia, had three arms

SECTION IN A TRENCH parallel to dominant wind direction, White Sands National Monument, New Mexico, U.S.A, illustrating
effects of wind reversal during winter season. Scale is in inches. 1 inch = 2.54 centimeters. (Fig. 64.)

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 107
and was structurally more complex and difficult to
explain genetically than the large one. However, it
also showed a complicated mixture of structures
dipping in several directions (McKee, 1966, p. 68)
and almost certainly was the result of alternation
among several wind regimes.
Star dunes occur in many parts of the world and
under a variety of conditions. They are especially
numerous in the Rub' al Khali of Saudi Arabia and
in the Grand Erg Oriental of Algeria but also are
well developed in the Gran Desierto of Sonora,
Mexico. In most places they appear as individual or
separate mounds randomly scattered across a sand
surface, but in other places two or more seem to
have coalesced to form compound dunes. In several
sand deserts (fig. 67), they are distributed along the
crests of linear dunes, forming parallel rows of
stars, sometimes referred to as "strings of beads."
These are complex dunes, composed of two distinc-
tive types.
STAR DUNES, southeast of Zalim, Saudi Arabia. Crests are So far as is known, neither complex nor com-
70-100 feet (20- 30 m) high. Aerial photograph by E. Tad pound star dunes have ever been trenched for
Nichols. (Fig. 65.) structure analysis. Observations on the orientation
of slipfaces suggest, however, that, as the product
of multiwind development, their structures are
complicated and show a wide spread in dip direc-
tions. Furthermore, the great heights attained by
many star dunes indicate a dominant vertical,
rather than horizontal, growth.
The rows of compound dunes suggest two
generations of growth - the earlier one with a
bimodal-wind environment, the later one with
multiple-wind environment. The overall structure
pattern represented in sand deserts of this type can
only be determined with certainty when direct ob-
servations from trenching have been made. En-
tirely unrecorded are the structural characteristics
of interdune areas either in fields of scattered star
dunes or between parallel ridges of compound
linear-star dunes.

Sedimentary Structures
of Interdune Areas
TRENCH AT CREST of 100-foot-high (30.5-m-high) star dune,
southeast of Zalim, Saudi Arabia. Wedge-planar cross-strata dip INTERDUNE AREAS, which are the spaces be-
steeply in three directions. Scale is in inches. 1 inch =2.54 cen- tween dunes, differ greatly in size and shape. de-
timetres. Photograph by E. Tad Nichols. (Fig. 66.) pending largely on the types of dune that are separ-

Digitized by Google
STAR DUNES ALONG RIDGES of linear dunes, Namib Desert, 6 miles (10 km) south of Kuiseb River at Gobabeb, South-West
Africa. Photograph by E. Tad Nichols. (Fig. 67.)

ated by them. Study of deposits that accumulate Regardless of the migration rate. the proportion of
within interdune areas of barchanoid types show clean. well-sorted eolian sand to the impure. con-
the sediment to be mostly (1) dust or silt-size parti- taminated sediment of the interdunes is great. and
cles that settle out of the atmosphere. (2) organic the result is thick sand layers of crossbedded sets.
matter largely derived from vegetation. and (3) separated by thin parting beds of mixed sediment
limited amounts of sand that have strayed from ad- with flat-lying or irregular strata. Such relations
jacent dunes (fig. 68). In the White Sands area that have been established through a drilling progra'm
has been studied in detail these deposits are very for the White Sands dune field. New Mexico
thin. ranging in thickness from 1 to 4 feet (0.3 to 1.3 (McKee and Moiola. 1975). where three or four se-
m). depending largely on how long the surface has quences of dune sand are separated by a corres-
been exposed before burial by advancing dune for- ponding number of interdune deposits.
sets. In some dune fields - for example. in parts of
When dunes migrate across a dune field with a Algeria and Australia - the sand supply is sparse
unidirectional wind. the sand surface of the dune and forms isolated dunes on a bedrock or other non-
ahead of each interdune is constantly being beveled sand surface. In southern Algeria. sand dunes rest
through truncation of its cross-strata (fig. 69). and. on a pediplane cutting Precambrian metamorphic
at the same time. the windward side of the inter- rocks. Under such circumstances. sand is very tran-
dune is being buried and preserved. If dune migra- sient and migrates until some sediment trap causes
tion is very rapid. interdune deposits will be it to accumulate as a large body. In areas of con-
especially thin or even lacking. If migration is very siderable sand accumulation. the interdune surface
slow. an appreciable amount of sediment may ac- may show beveled tops of earlier dunes that have
cumulate on the interdune surface before burial. recently been planed off. Such surfaces define on

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 109

the same dune field. where sand movement is

much more rapid. the thickness probably is propor-
tionately less. Interdune surfaces in many such
places are represented merely by horizontal bed-
ding planes.
The Killpecker dune field in Wyoming. also
formed mainly by unidirectional winds. has been
examined through trenching. and the nature of in-
terdune structures has been determined
(Ahlbrandt. 1973. p. 41. figs. 20. 66. 67) . Trenches in
this field showed mostly horizontal laminae dis-
rupted in places by root tubules and consisting in
part of marl developed in interdunal ponds. The
deposits examined were formed in the upwind part
of the field under the influence of very strong
winds; the sand included some of the coarsest and
most poorly sorted detritus in the entire dune field.
Structures of interdune areas among the coastal
dunes of Brazil. like the just-described North
American desert types. consist mostly of horizontal
laminae that rest upon truncated cross-strata like
those of nearby active dunes (Bigarella. 1972. p. 27) .
The beveled foresets are the remnants of earlier
dunes that migrated across the area; they formed an
overlapping pattern as "the result either of changes
B in wind direction or of irregular superposition of
the curved foreset strata." Coastal-type interdunes
INTERDUNE AREA showing miniature stringers and sheets of in Brazil are further discussed by Bigarella later in
flat-bedded sand (wind-shadow deposits) advancing across in- _ this chapter.
terdune. White Sands National Monument, New Mexico, U.S.A. Interdune surfaces in areas of parallel-straight or
From McKee and Moiola (1975, fig. 6). (Fig. 68.) linear dunes. commonly referred to as seif or
longitudinal. represent a depositional environment
horizontal planes the curving arcs of barchans (fig. very different from that among transverse-type
70). the long. relatively straight crest lines of dunes. Between linear-type dunes. the interdune
transverse ridge dunes (fig. 71). or other patterns surfaces commonly are many times wider than the
representing stages in dune migration. Where these sand ridges that enclose them (fig. 59). Further-
interdune surfaces have been exposed for an ap- more. these surfaces between the long. straight
preciable time. the relict structures are obliterated dunes extend in a direction generally parallel to the
by accumulating sediment. prevailing wind or winds. They have an open ac-
Interdune deposits at White Sands National cessway upwind and few. if any. barriers down-
Monument. New Mexico. U.S.A.. were examined in wind.
vertical sections by trenching in areas contiguous Detailed studies of a well-developed interdune
both to barchanoid ridges and to transverse dunes deposit between linear (seif) dunes in the Fezzan of
(McKee and Moiola. 1975. p. 61) . Structures in Libya were made in 1962 (McKee and Tibbits. 1964.
these sections consisted of thin. flat or irregular p. 13). These studies indicated that effective winds
beds of silts and sands with scattered fragments of moving down the length of a flat interdune surface
plants and with laminae contorted by root growth deposited sand as flat-lying laminae to a thickness
(fig. 72) . The interdune deposits examined at the of at least 4 feet (1 m). as shown in test pits (fig. 73).
surface. and also in cores. range in thickness from 1 No cross-strata were noted. The sand as determined
to 4 feet (0.3 to 1.3 m). A few kilometres upwind in by analysis was poorly sorted. ranging in size from

Digitized by Google
110 A STUDY OF GLOBAL SAND SEAS

BEVELED WINDWARD SURFAa of transverse dune showing truncation of lee-side foresets, White Sands National Monument,
New Mexico, U.S.A. From McKee and Moiola (1975, fig. 68). (Fig. 69.)

BEVELED REMNANT OF TRANSVERSE DUNE in interdune area,

White Sands National Monument, New Mexico, U.S.A From
McKee (1966). (Fig. 71 .)
BEVELED REMNANT OF BARCHAN DUNES in interdune area,
White Sands National Monument, New Mexico, U.S.A. Three
truncated sets of cross-strata are visible. (Fig. 70.)

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 111

Sandy clay
Lag gravels

White
sandy clay

Quartz
pebblesJO

A
0
Unexcavated
s~
HORIZONTAL AND IRREGULARLY BEDDED STRATA in trench,
interdune area between barchanoid ridges, White Sands Na-
tional Monument, New Mexico, U.S.A. Scale is in inches. 1 inch
=2.54 centimetres. From McKee and Moiola (1975, fig. 6). (Fig.
E
"N W
In:hes
60
Structureless
sand

72.) E:t
Sand strata
very fine to coarse, in contrast to well-sorted sand
in nearby dunes. Along the margins of the linear
dunes, which have slipfaces on both sides dipping
fir#!
Gravel. sand.
at high angles (as much as 35), interdune deposits and clay
consisted partly of low-angle strata (10-13), dip-
ping toward the dunes. and partly of very fine B 0
10
e,
structureless sand. No information is available on
55 In:hes ----i Charcoal

the thickness of these interdune deposits.

The nature of interdune deposits between star
dunes (fig. 74). which are the product of multidirec-
tional wind regimes, has not been investigated, to
the author's knowledge. Such dunes mainly build
vertically, rather than migrating laterally. After at-
taining a maximum peak height permitted by the
threshold velocity, they must spread outward from
the base, so it seems probable that the interdunes
develop few of the features characteristic of
troughs (sediment traps) associated with the active
migration of transverse-type dunes. The structural
character of interdunes in fields of star dunes is yet
to be determined.

Deformational
BLOCK DIAGRAMS (A, 8) OF INTERDUNE STRUCTURES and
Structures in Dunes photograph of test pit between seif dunes, northeast of Sebhl,
libya. A, pit 14; 8, pit 12. From McKee and Tibbitts (1964. fig. Ba.
IN ADDITION TO sedimentary structures c). C, Test pit in center of interdune area, shows horizontal
formed by deposition, numerous penecontem- stratification. Steel tape extends 4 feet (1 m) down. Photograph
poraneous structures, caused by deformation of by Gordon C. Tibbitts, Jr. (Fig. 73.)
strata during or immediately following deposition,
are characteristic of various dune types. Principal

Digitized by Google
INTERDUNE SURFACE AMONG STAR DUNES near Tsondab Vlei, Namib Desert, South-West Africa. Photograph by E. Tad Nichols.
(Fig. 74.)

forms of deformation in dune cross-stratification classification was based partly on experimental

are folds. faults. and brecciated strata of which work in the U.S. Geological Survey Sedimentary
there are many varieties. Structures Laboratory in Denver. Colorado. and
partly on field investigations at White Sands Na-
Penecontemporaneous deformation is caused in
tional Monument. New Mexico. Subsequent field
dunes primarily by avalanching, either through
slumping (mass movement) or by grain flowage; studies. testing the validity of this classification on
less frequently it results from loading. plant coastal dunes of Brazil. were made by McKee and
growth. scour-and-fill processes. and sand drag. Bigarella (1972).
Principal types of deformation structures formed
Virtually all forms are considered to be contorted
bedding. and most are very irregular. especially in avalanche deposits are referred to as rotated
when compared to similar-appearing structures structures. warps or gentle folds. drag folds and
flames. high-angle asymmetrical folds. overturned
formed in competent rock masses.
folds, overthrusts. break-aparts. breccias. and
A classification of deformational structures in fadeout laminae (fig. 75). Some of these structures
dunes. with suggested terminology for various are characteristic of tensional stresses. common
forms that have been recognized, was proposed by near the tops of dune slipfaces; others are prin-
McKee. Douglass. and Rittenhouse (1971). This cipally the result of compressional forces that

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 113

develop near the bottoms of dune foresets. Some tional structures are break-aparts and breccias, ro-
are common in dry sand, others develop mostly in tated plates and blocks, and high-angle asymmetri-
wetted sand, saturated sand, or sand with damp cal folds. An analysis of these trends as determined
crusts. in laboratory experiments and in a comparison of
desert sands in New Mexico and coastal sands in
Moisture greatly affects the type of contorted Brazil was tabulated by McKee, Douglass, and Rit-
structure developed within a sand body because it tenhouse (1971, table 5).
largely controls the degree of cohesion of sand par-
ticles. In sand that is dry, cohesion of grains is In many dune sands, both coastal and desert,
slight, and the chief deformational structures are moisture may form wet crusts, as from early morn-
fadeout laminae, drag folds and flames, and warps ing dew. In such deposits, certain layers have been
or gentle folds. In sand that has been wetted, con- dampened or wetted shortly after deposition; so
siderable cohesion develops, and common deforma- wet and dry layers alternate.

'" ....
.........
.............. -... ..... ..... .. ..... .....- .- .......
..- ..-::.- ..'
~.'

........ .
'. ". ". ..... ' . "

". .' .' ... ...... -,~:.. '" ,..."

..-
.. " ,,I e
"... p-
.,'
'. '. '.
".
. .. ........' '"......'.. .....r"
... ,;:'f'
..- ..................:..... .... ....I'
'.
' ~N..
.......'
....... ' .'
.. ... Jo " '
.1- ,/'
...,..,..."
:f .. - .'"
.'
~... ....
r" '
". .,-'

..... ..
.,'
' .'
.'
......s.r"'" ..'
A
.' .'
B c
. ... .... .. ....
..... .... .....,........ \ , -.
..... ..... .. "-" ..... ~. ........
....~, "\'" ~ ..... ......... ..................
....".

-.... ...... .. ........ . '1

............

..-........... '"
- ......j '.. . '....... . .....
..... ..... -....::. ...<:;.~~;~~ . .::. . . .. .... 1'-' ...........
...
_"
I! . . . . . . .. : " .

~ .........,
". .... .....,.. '":- ~ .......-
.,. ::'-....':.. ;",.-
J ...... . ....
-"'"
'-, .....~.~, -1

~. . :. : :.::\::.~:.~;~;,f{:~;::~;~:::.::::::::::: ......... ......... ._..

t'-. ..............
'"'~._''''''' . - ...,..~~
. . .......... ... . -c" ..... ~:-........ ................ .., "t.
...............::: t~
'\:-..,.

.... .. " ' ....

"
........
1--..
.... .........

.."...... ......... ' .. ...
~_
........
...........---.. ...........~ ............
... .. " ..
..... .. . . . . . . . 0
-

' I,.
D E F

.....
......
........
~..... . ...
.~, , ' ...
.....,
, ' ' Co
.... ....;..... .w,..:......
I.
..'..."
..
,..
'L......... -tot.......\ ... ....... .
'\.
.... "~ '
,~
~
~ ~
.~
""-.
"""'" ....... ...!:....
'......'. . . "" '-'..... .....
.... :..
.......
I,
r ,
". .....
...... "~r.,,-
....
G H I
PRINCIPAL TYPES OF DEFORMATIONAL STRUCTURES in avalanche deposits of dunes: A, Rotated structures; 8, warps or gentle
folds; C, flame structure; D, drag fold; E, high-angle asymmetrical folds; F, overturned folds; G, overthrust; H, breakaparts; I, breccias.
Each block represents an area approximately 6X 4 inches (15 x 10 cm). (Fig. 75.)

Digitized by Google
114 A STUDY OF GLOBAL SAND SEAS

Structural Features at tive part of the parabolic dune, trench I on the slip-
face, trench H on the windward side behind the
Lagoa Dune Field, Brazil crest, and trenches C and D oriented oblique to the
wind direction. Trench E and pits Et' E2 , and E3 are
By JOAO J. BIGARELLA alined roughly parallel to the wind direction and
constitute a southern continuation of scarp B, ex-
Parabolic Dunes tending down the windward slope into the inter-
dune area. Trenches F and G, and pits 1 -7 are
THE LAGOA DUNE FIELD is in the east-central
oriented normal both to the dune's arms and to the
part of Ilha de Santa Catarina (State of Santa
dominant wind direction. These pits are located in
Catarina, Brazil). The dune field is bounded on the
the interdune area and dissect some of the small
east by crystalline hills (mostly granite), on the
satellite parabolic dunes.
south by the Campeche Beach (locally referred to as
Joaquina or Lagoa Beach), and on the west and Cross-Stratification
north by the Lagoa da Conceiclo lagoon (fig. 76).
Prevailing winds in this area blow from the Cross-strata in the parabolic dune are medium
northeast, but short-term storm winds come from scale in the crest area but large scale in the nose and
the south. The influence of the prevailing winds on the arms. Many laminae dip at high angles
the dune field is decreased considerably by granitic (29 -34) in the nose and also in the arms where
rock hills located upwind. The prevailing winds are they are normal to the dominant wind direction and
deflected by the mountains to the west of the Lagoa to the dune axis. Both along the crest and in some
da Conceicao and reach the dune area as northerly parts of the arms, the bounding surfaces between
winds. Because the dune field is located in a sets of cross-strata are nearly horizontal.
shadow area, protected from the northeasterly pre- Elsewhere, as in the nose and parts of the arms,
vailing winds, the deflected northerly winds cannot they dip at angles ranging from low to high. As
counteract the action of sand transport from the verified for other dune types (McKee, 1966;
southern short-term storm winds. Thus, the main Bigarella, Becker, and Duarte, 1969), individual sets
eolian sand movement is to the north (Bigarella, of cross-strata tend to be thinner and the laminae
1972). flatter near the top than at the bottom of the dune
The Joaquina Beach constitutes a main source of (fig. 77).
eolian sand. However, the amount of dry sand Cross-strata measurements for all trenches,
available for eolian transport on the beach is too scarps, and pits in the various parts of the dune are
small to counteract the deflation caused by the summarized in tables 10 and 11, and in figure 78.
strength of the southern storm winds in the Measurements in each part are controlled by its
southern part of the dune field. Therefore great local morphological conditions (fig. 77).
numbers of blowout features, including parabolic
dunes, are well developed. In the northern part of Nose Area
the dune field, a series of reversing dunes has
In the nose, high-angle foresets are characteristic
formed from the interaction of opposing winds.
features in the walls of trenches C and I. These
They are slowly moving northward.
trenches were oriented oblique and parallel, respec-
tively, to the southern storm winds (fig. 77). Trench
Method of Study
C, excavated in the slipface, was oriented N. 52 W.
The present study is of a compound parabolic at an angle of about 46 to the direction of the
dune located in the southern part of the Lagoa dune southern storm winds, but approximately in the
field about 480 metres (1,575 ft) from the beach. average cross-strata dip direction as calculated
Most of the internal structures to be recorded and from nine measurements. In trench I, excavated
analyzed were exposed by digging a series of parallel to the direction of the storm winds, five
trenches and pits, and by cleaning two erosion measurements of cross-strata were made, and their
scarps cut by deflation near the dune nose (fig. 77). dips were N. 2 -15 W. The average dip, N. 6 W.,
Trenches C, D, I, and H are located in the most ac- coincides with the main direction of the southern

Digitized by Google
 ED "EN RY TR TUl IN Ur.. 1
, ..... '

EX NA N

n sand

[{:.\j Parabolic dunes 40 metres (131 tt) high

E\)/'J Parabolic dunes 20 metres (65 tt) high

~ Sandy ramps-Oeveloped by dissipation of eolian

sands from earlier dunes
U Low plaln--Most y sandy with few swampy areas

r:~"J mb ncr. lline sam -M yg e

nda as wh app mat

OME

IlE

L CAT NO AG D EF D, de nta tari Sta of ta rin raz Mod ed mB rell 97 fig.

(Fig. 76.)

l
116 A STUDY OF GLOBAL SAND SEAS

EXPLANAnON

- F o rward edge of dun '

Dune crest e slipface

A?!~
1(t\:Zt:~~~ Satellite dune
I "'J it,

Trench
10'10'1011 ' " ' ' Scarp
x Pit

~
~~
~
erosion surface 1(~eS)
Cross-strata (thin '
eavybO,unded
line) by

o~--=F==-+-
Scarp A

S Ttench F

~ ~
~~
Ttench I

2
w

CROSSBSECTI
A-A' ' ONS AND PLAN VI 1,,.... G
, -8, and C -C' sh ow surface
EW profil es across
of paraboli c dunes
d unestudied
and intIn
'erdthe
uneLago a dWI
areas une
'th field ' IIh a of
locations de pits
Santa d trenches
an Ca tanna,
' Brazil show'Ing
, All' measure-

Digitized by Google
 IML. AR TRU __ URL DL..JS 11

N
r A

:~ ~/\j
00 =::::: W - ~
- ~
- ..L..-
~ 100

N S
I!

W
C
:r:. ~
00 m ~
- I
t;1
j
1
E
C
5

\
\.. 1 J
~~~./_~-~
'Q
100 1~

NNW

TtHCt D
N

TtHCt E

scarps, trenc es, and pits. Groups of lines numbered 1 - 5 are approximate locations 0 outcropping aminae. oss s ions
ments in metres. 1 metre=3.28 feet. Modified from Bigarella (1975b, figs. 4, 5). (Fig. 77.)

y- -~ )-- ....
118 A STUDY OF GLOBAL SAND SEAS

storm winds (S. 6 E.) and with the axis orientation posed - both windward strata and downwind
of blowout dunes inside the Lagoa dune field. In strata (fig. 77). Sets of cross-strata are prominent in
trench I, the average cross-strata dip direction coin- erosion scarps A and B, and bounding surfaces dip
cides with the dip direction of the slipface surface. slightly to moderately windward. Some sets of
The rose diagrams of figure 79 illustrate cross- strata dip downwind at low angles, other sets are
strata dip directions from various trenches dug in approximately horizontal.
the slipface and crest areas. The slipface and the The whole crest seems to result from the coales-
crest just behind it have a similar cross-strata at- cence of several low mounds of sand. Rose
titude, but differ in the average of the cross-strata diagrams in figure 79 illustrate the peripherally dip-
dip angles, which are 10 -13 less in the crest beds. ping attitudes of cross-strata deposited in the wind-
ward and downwind parts of the former dune crest.
Crest Area High-angle dips (15 -30) occur among lee-side
sets of cross-strata on the west side of erosion scarp
In the present crest area, trenches 0 and H con- A (fig. 77). On this scarp, the average dip of cross-
stitute southward continuations of trenches C and I strata is 12.8; many angles of dip are between
in the slipface area (fig. 77). In trenches D and H, 3 -11 The resultant southeasterly vector and the
0.
internal structures consist of two main parts: (1) a low-angle strata indicate a predominance of wind-
thick set of cross-strata with high-angle foresets, ward-deposited strata (fig. 77), although about 20
deposited as the nose of this parabolic dune at an percent of the strata dip peripherally in all direc-
earlier stage, and (2) thinner cross-strata sets with tions around the compass, suggesting a very gentle
low- to moderate-angle dips deposited on the crest dome structure.
(fig. 77). Erosion scarp B, on the opposite side of the defla-
An erosion surface truncates the lower set in tion corridor from scarp A, exposed strata dipping
both trenches. On the west side of trench D (fig. mostly less than 10 and averaging 8.4. The strata
77), an early set of crest-strata dipping 6 -11
dip consistently southwestward, suggesting that
toward N. 37 W. is overlain by high-angle foresets, this part of the former crest of the parabolic dune
mostly dipping 29 -35 (average 28) to N. 58 W. had a dome shape.
This succession suggests a former slipface on the Trench E extends southward from erosion scarp
western flank of the nose approximately where B (fig. 77) and has in its upper part strata that dip
trench C is located (fig. 77). The truncated slipface south and belong to the former dome-shaped crest.
is overlain by a sequence of strata dipping at low Internal structures indicate the presence of crest-
angles (1 -11 0) toward the eastern quadrant N. 55 type cross-strata, of which the average dip is 12.9.
E.-S. 52 E. The lower part of the trench is composed of low-
High-angle cross-strata in the lower part of angle strata dipping downwind 8 -16, whereas the
trench H (fig. 77) dip N. 13 E., whereas those from upper part is composed of windward-dipping strata
the upper part (crest beds) dip N. 27 E. (four with dips of 16 -18. The sequence of erosion scarp
measurements). These cross-strata of the crest con- B overlies strata of trench E. Trench E strata, in
tinue northward to form strata of the slipface. In turn, overlie sets of high-angle cross-strata exposed
the area of transition between the crest and the in pits El' Ez' and E3 in the present interdune area.
slipface, strata are convex upward. Both parts of The high-angle (25 -33) strata in these pits are
trench H have dip directions which diverge by lower parts of slipface deposits of a parabolic dune
19 -33, indicating that the former slipface dip nose which migrated northward and was truncated
direction was more north-northeastward than the by deflation. A new crest was later developed
slipface of trench I, which is alined with the wind (lower part of trench E deposits) but was eroded on
direction. its windward side, which formed the deposit down-
Structures representing an earlier position of the wind represented in the erosion scarp B sequence.
crest are exposed scarps A and B. The former crests The crossbedding pattern exposed at the surface,
were cut in half by a deflation corridor and a new upwind from the crest, at the transitional edge of
crest was built downwind. In the erosion scarps,5 the interdune area, is illustrated in figure 80 (out-
about 2.5 metres (8.2 ft) of sand sequences were ex- cropping laminae in fig. 77). The trace of high-angle

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 119
cross-strata represents deposits from the nose and average dip direction for both sides is within a
lower parts of the arms of this parabolic dune. spread of 33. The resultant vector N. 13 E., is 18
Cross-stratification is periodically exposed by to the east of the main direction of the southern
deflation in this area. The three main sets of nose storm winds.
deposits downwind suggest the way in which the
crest and nose split and develop as wide and very Arms
low moundlike features. Basic structures in the arms of the parabolic dune
Rose diagrams (fig. 78) show the attitude of are illustrated in trenches F and G, which are nor-
cross-strata that are exposed behind (upwind of)
Erosion scarps refer to surfaces along the sides of deflated or "blowout" areas
the crest. Measurements were made on both the left which have been smoothed by machete. Trench walls refer to elongate excava-
(west) and right (east) sides of the dune axis. The tions of much longer dimensions than the pila.

I
ENTIRE DUNE

SCARP A; SCARPS (NOSE SCARP B;

TRENCHES C,D AND CREST) A, B; TRENCH E
TRENCHES C, D, E

TRENCH F COMPOSITE OF PITS TRENCH G

~7, E" E2 E3 (INTERDUNE)

ROSE DIAGRAMS OF CROSS-STRATA DIP DIRECTIONS recorded for scarps, trenches, and pits representing principal parts of
parabolic dune and interdune area. Values are in percent. Modified from Bigarella (1975b, fig. 6). (Fig. 78.)

Digitized by Google
 I-l
TABLE 10. - Frequency of dip-direction intervals for cross-strata in trenches, scarps, pits, and surface of a parabolic dune, Lagoa dune
field, &azil ~
IModifUld from Biprella (t97Sb. f.,. 8). Locations shown in fill 77)
Scarpa and trenches Pits Surface
Interval Entire
(degrees) A B C D E F G Total t 2 3 4 S 8 7 Et Ez Ea Total West East Total dune
0-30 1 ... ... ... 1 ... ... 2 ... 5 2 1 2 1 1 1 ... ... 13 27 19 46 61
30-60 2 ... ... 1 ... ... 1 4 1 ... 1 2 5 ... 2 ... 1 ... 12 4 22 26 42
60-90 1 ... ... ... ... 1 25 27 4 ... ... 1 7 ... 4 ... ... ... 16 ... 1 1 44
90-120 1 ... ... 1 ... 1 12 15 ... ... ... ... ... '" 2 ... ... ... 2 ... ... ... 17
120-150 6 ... ... 2 ... 1 ... 9 1 ... 1 ... ... ... ... ... ... ... 2 ... 1 1 12
150-160 10 ... ... ... ... 3 ... 13 1 ...
... ... ... ... ... ... ... ... 1 ... ... 14
160-210 3 ... ... ... 2 ... ... 5 ... ...
... ... ... ... ... ... ... ... ... ...
'"
... ... 5
210-240 1 3 ... ... ... ... ... 4 1 ... ... ...
... ... ... ... ... ... 1 ... ... ... 5
240-270 1 10 ... ... ... 4 5 20 4 ... ... 1 ... ... ... ... ... 2 7 ... ... ... 27
270-300 1 1 2 6 ... 10 9 29 5 2 ... ... 1 1 ... ... 1 1 11 ... ... ... 40
300-330 ... ... 7 11 1 4 2 25 1 1 3 1 ... 2 ... 2 3 5 18 8 ... 8 51 >
330-360 ... ... ... ... 5 5 ... ... ... ...
~-<
1 2 6 1 10 1 2 2 15 21 2 23 48

TABLE 11. - Direction and degree of dip for cross-strata in trenches, scarps, pits, and surface of a parabolic dune, Lagoa dune
field, &azil
~
Locality
(Modified from Blprella (t975b. fia. 8). LocatioDII shown in fia. 77)
Number of Mean dip ConaiIttlDcy Maximum Averaae
~
~
meBlurements direction' ratio" dip dip
Scarps A, B, trenches C, D, E ............. 84 N. 63 W. 277 0.32 35" 16.1"
Trench F ............................... 25 N. 75 W. 285 .51 28" 14.1
Trench G ............................... 54 N. 74 E. 74 .43 34 22.2
Total ............................. 163 N. 53 W. 307 .13 35 17.8
Total pits ......................... 98 N. 5W. 355 0.15 37 17.4
Surface - East side ..................... 45 N. 31" E. 31" 0.95 34 25.5"
o Surface - West side ..................... 60 N. 3W. 357 .84 34" 23.6"
5" Total ............................. 105 N.13 E. 13 0.44 34 24.4"
;:::;:
N" Entire dune ....................... 386 N. 5W. 355 0.42 37 19.6
CD
0.. The frequency of dllHIirection clata from which the mean dip direction was derived are shown in table 10.
rr a Consistency ratio is obtained by dlvidlll8 the length of the vector resultant of all observatioDII at a Biven locality by the 8um of the IIIII8lhs of individual vectora. It
'<
is a weighted _ure of desree of groupins of vecton. It rBJI88S from t to 0.
C)
o
~
~
rv
 SEDIMENTARY STRUCTURES IN DUNES 121

mal to the direction of the southern storm winds Low-angle cross-strata commonly are deposited
(fig. 77). The cross-strata sets dip toward both the on the windward sides of dunes. but in the arm of
eastern and the western quadrants. almost at right this parabolic dune some of the low-angle dipping
angles. or oblique to those on the dune nose. strata seem to have been deposited on the inner
The profile across the arms of the parabolic dune side of the arm at almost right angles to the wind
may be divided into two parts - the inner and the direction.
outer. The inner side. less steep than the outer. The opposing attitudes of cross-strata in trenches
commonly is an area of deflation. The steeper. F and G are illustrated in a rose diagram (fig. 78;
outer profile has a slipface and is an area of deposi- tables 12. 13).
tion.
In trench G. on the right arm of the parabolic TABLE 12. - Attitudes of 25 cross-strata sets in trench F of
dune. the lower part of the sequence on the left side parabolic dune. Lagoa dune field. Brazil
of the trench is composed of westerly dipping strata (Modified from Blprella (t97Sb. table t))
averaging 15.2 (fig. 77). The highest dips are Set Dip Amount Set Dip Amount
19 -22 and the lowest. 9 -13. similar to the mid- No. direction (0) of dip (") No. direction (0) of dip (0)
dle part of trench F below the crest. where dip 1 290 25 14 270 6
2 310 20 15 305 14
angles are also 9 -13 (fig. 77). These cross-strata 3 312 18 16 310 8
are not typical slipface deposits like those shown on 4 332 12 17 325 16
the right (eastern) side of trench G (fig. 77). The 5 315 25 18 310 9
left (western) part of trench G may represent a 6 295 12 19 302 12
7 111 18 20 302 13
former outer part of the left arm of an earlier 8 98 3 21 276 11
migrating parabolic dune. subsequently truncated
by the present dune. 9 6 5 22 280 13
10 136 25 23 279 16
On the east side of trench G (fig. 77). foresets are 11 165 10 24 310 15
mostly medium- to large-scale. The cross-strata dip 12 172 11 25 285 28
13 165 8
eastward. Forty percent of them have high-angle Dip direction: For true north. subtract 12.
dips (30 ~4); most of the others dip only 21 -29.
Bounding surfaces also have moderate dip angles.
The average dip of cross-strata in trench G is 22.2. Deformation
High-angle strata from the outer part of the arm
dissected by trench G seem to overlie low-angle in- Penecontemporaneous deformation. caused by
terdune cross-strata. enriched with humate com- slumping of steep foresets. is common in most
pounds. In some places. an incipient paleo A coastal dune structures (fig. 81). This gravity
horizon is developed (nearly vertical lines. on east slumping occurs during or soon after deposition.
edge of trench G fig. 77). Deformational structures also occur in
The internal structure of trench F is far more moderately to gently dipping strata. Deformed
complex than that of trench G. The right (east) side laminae may occur between undeformed beds (fig.
of trench F (fig 77) is composed in the lower part of 82A) or may affect adjoining sets of strata. The
truncated cross-strata dipping westward at moder- structures produced by deformation seem to be
ate angles (18 -25). They represent deposits of the chaotic - folds. faults. and breccias which consist
left arm of a former parabolic dune which migrated of broken. rolled. and crinkled masses of sand (fig.
across the area. Between this sequence and the one 81). Some contorted structures result from the ac-
forming the crest are nearly flat lying strata (3 -50) tivities of organisms (fig. 82C).
which probably represent interdune deposits. The response to stresses and the type of struc-
Above these interdune strata. the structures of the tures produced during deformation are controlled
arm can be traced across a series of erosion sur- largely by the amount of moisture present in the
faces. each series truncating slipface strata. up to sand (McKee. Douglass. and Rittenhouse. 1971;
the present position of the crest. Bounding surfaces Bigarella. Becker. and Duarte. 1969; McKee and
in trench F dip moderately westward. Bigarella. 1972).

Digitized by Google
122 A STUDY OF GLOBAL SAND SEAS

TABLE 13. - Attitudes of 54 cross-strata sets in trench G of may form in low parts when the phreatic water
parabolic dune, Lagoa dune field, Brazil level rises. A typical vegetation cover of grasses
(Modified from Blprella (111'l5b. table 2))
and shrubs grows in the interdune environment
Set Dip AmouDt Set Dip AmoUDt
No. direction (0) of dip (0) No, direction (0) of dip (0) when the sand is wet, and many small plants.
1 274 20 28 95 27 especially Gramineae and Cyperaceae germinate
2 331 9 29 94 21 there. Humic compounds that originate from decay
3 290 22 30 102 21
4 291 14 31 95 27 of the organic material stain the sandy substratum
5 310 10 32 87 23 and are responsible for the brown color of both
6 278 17 33 95 30 ground water and surface water.
7 305 14 34 95 31 The buried upper surface of the brown structure-
8 332 12 35 107 29
9 288 18 38 97 32 less sand is irregular. Above this surface are se-
10 284 20 37 103 34 quences of cross-stratified eolian sand deposited
11 296 15 38 103 32 either by migration of the main parabolic dune or
12 279 16 39 110 27 by growth of small satellite parabolic dunes (fig.
13 283 19 40 106 37 85).
14 275 14 41 108 30
15 286 13 42 97 34
The attitudes of 98 cross-strata were measured in
16 279 10 43 97 33 the 10 pits. About 42 percent had dip angles greater
17 77 22 44 102 32
18 99 23 45 108 28 than 20, whereas 58 percent had dip angles less
19 110 3 46 97 32 than 15. The maximum dip recorded was 37, and
20 84 12 47 100 29 the average dip was 17. The crossbedding pattern
21 91 21 48 97 30 from the interdune area documents the migration
22 91 27 49 96 27
23 89 16 50 102 33 or movement of former dune bodies across this
area. The traces of several parabolic dunes, both
24 85 26 51 92 20
25 91 31 52 99 8 nose and arms, can be recognized. A rose diagram
26 88 25 53 75 12 (fig. 86) shows most strata in this area dipping
27 102 18 54 84 16 toward the northern quadrants, with an average dip
Dip direction: For true DOrth. subtract 12,
direction of N. 5 W.
Substantial changes in dune morphology along
the Brazilian coast have been caused by climatic
Interdune Areas Among Parabolic change (Bigarella, 1965, 1972). Recent action of con-
Dunes centrated rainfall has greatly modified the original
sand body at Lagoa - the eolian sands were spread
In an interdune area. between the arms of a sur- by floodwaters like a blanket, decreasing in thick-
veyed parabolic dune (fig. 77), two lines of pits ness downstream.
were dug. The east-west line (pits 1-7) provided Climatic changes in the past provided conditions
information concerning the interdune area between under which extensive dune development alter-
the trenches G and F on the trailing arms of the nated with dune modification. Early deposits of
parabolic dune. The north-south line (pits Et -Ea) eolian sands have become mixed with colluvial
forms a southward continuation of trench E. The material producing an unsorted sandy, muddy sedi-
pits are rectangular in cross section with sides ap- ment. Mixing took place during colluvium move-
proximately 1 -2 metres (3 -6 ft) long and depths of ment caused by climatic change.
1 -2 metres (3 -6 ft; figs. 83, 84). The dunes subjected to heavy rainfall became
Characteristic of the interdune area are the ap- saturated with water. This rainfall caused wet slid-
parently structureless sands, enriched with humic ing, sand flows, gullying, and the formation of
compounds derived from incipient soil formation. many alluvial fans that coalesced to form dissipa-
In some places, paleosol A horizons alternate with tion ramps (fig. 76). Many faults were developed in
stratified sand (pit 6, fig. 83). the sand during avalanching. Sand-flow and rain-
The interdune area frequently becomes wet, wash deposits mostly form wavy or lens-shaped
sometimes swampy, after rains, and small ponds layers. and many of them are imbricated. Some

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 123
deposits are largely derived from alteration of the much drier than the corresponding areas among
dunes; others are mixtures of colluvium and parabolic dunes. Some ephemeral ponds are pres-
modified eolian sands. In some places, clay was ent, but most of the area is covered with Gramineae
deposited in ephemeral interdune ponds. Times of and Cyperaceae, and some parts of it consist of sand
soil formation alternated with times of erosion. In plains without vegetation. In the unvegetated area,
the same manner, dissipation deposits are intercal- four pits and two trenches were dug to expose the
ated with deposits having normal dune cross- internal structures of an interdune (fig. 87). Trench
stratification. VU shows a sequence of northward-dipping trun-
cated foresets deposited by short-term storm winds
Interdune Areas Among Reversing from the south. In trenches VUI, XVII. XVIII. and
XIX, high-angle foresets are covered by flat-lying
Dunes beds, dipping less than 1. Where these strata ap-
At Lagoa dune field the depositional environ- proach the slipface of the next dune upwind.
ment of interdune areas among reversing dunes is however. the dip is as much as 10. In trench IX the

TRENCHES

Number of measurements (nml. dip direction (ddl. consistency

ratio (erl. and average dip (adl data for trenchas and scarps

ad
nm dd er (degreasl
Trench C 9 N.55W. 0.99 32.6
D 23 N. 46" W. .66 19.8
I 5 N.07"W. 1.00 29.8
A B H 6 N. 22" E. .99 19.5
SCARPS E 10 N.22"W. .61 12.9
Scarp A 28 5.25" E. .57 12.8
B 14 S.7~W. .98 8.4
N S

CROSS SECTION

ROSE DIAGRAMS OF ATTITUDES of cross-strata deposited in an earlier dune crest (trenches 0, H, E, and scarps A, B) and in the
nose (trenches C, I) of a parabolic dune (fig. 77), lagoa dune field, Brazil. Values are in percent. Generalized cross section through
dune is parallel to dominant wind direction and shows location of trenches and scarps. Table gives statistical data. Modified from
Bigarella (197Sb, fig. 7). (Fig. 79.)

Digitized by Google
124 A STUDY OF GLOBAL SAND SEAS

oI 5
I
10 METRES
I

OUTCROPPING LAMINAE EXPOSED on beveled surface in interior of parabolic dune. Units 1- 5 show details of traces of laminae
as exposed by natural truncation in interdune area (fl8. n). Direction of strike and dip shown by symbols. Scale is approximate.
Modified from Bigarella (1975b, fig. 6). (Fig. 80.)

high-angle foresets are covered by interdune bedding patterns. The processes seem to be cyclic
deposits which range in dip from low angle to and occur only under certain environmental condi-
almost horizontal, and above these is another set of tions; accordingly, dissipation layers alternate with
high-angle dune foresets. dune bedding and are recurrent in the stratigraphic
section. The older dunes in the dune field have lost
Dissipation of Dunes, their original shape by dissipation. Some still show
either a hummocky morphology or a rounded
Lagoa, Brazil topography. Others have been changed to tongue-
shaped layers that are not intercalated among other
By IOAO I. BIGARELLA dune deposits.
Four principal sequences of eolian sands separ-
General Features and Environment of ated by paleosols (A horizons) occur as a dissipa-
Dune Dissipation tion ramp on the slopes of the granite hill (Joaquina
Hill) just east of the dune field (fig. 88; study area
ALONG THE BRAZILIAN coastal plain flash in fig. 76). The contacts between the various sand
floods have changed both internal pattern and sequences are irregular. owing to either erosion or
morphology of the original dunes and beach-ridge the hilly topography of the dunes.
deposits. This reworking (by dissipation) modifies The oldest sequence. which is reddish brown.
or destroys original dune structures. and it con- rests on a clayey lag deposit. composed of granite
centrates colloidal materials in a wavy, irregular and quartz pebbles. which. in turn. overlies
pattern known as dissipation structure. weathered granite. The eolian sand above the
In an area within the Lagoa dune field on the nha gravel pavement originally consisted of fine
de Santa Catarina, Brazil (fig. 76), dissipation struc- windblown grains. This sand subsequently became
tures in parabolic dunes were studied. Flash flood- mixed with granules (primarily quartz and
ing in this area forms the large sandy ramps and the feldspar). clays. and silts from the weathered gra-
small-scale structures that interrupt normal dune nitic mantle.

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 125

SCARP A TRENCH F

~-----------------------------------v~------------------------------------
TRENCH G

DEFORMATIONAL STRUcrURES OBSERVED in parabolic dune, lagoa dune field, Brazil. All measurements are in centimetres. 1
centimetre=0.39 inch. (Fig. 81.)

Digitized by Google
126 A STUDY OF GLOBAL SAND SEAS

E 1m

STRuaURES IN A PARABOLIC DUNE, lagoa dune field, IIha 0 , High-angle strata set dipping eastward in trench showing
de Santa Catarina, Brazil. A - 0, Structures in trench G, eastern string-and-nail grid for recording structures. , F, Structures in
arm of parabolic dune. A, Contorted bedding in strata dipping eastern arm of parabolic dune. , Former crest area exposed by
toward the interdune area between the arms. 8, Strata in trench erosion (deflation), showing downwind strata dipping to the
showing a shift of the dune arm. The lower set dips toward the left, upwind strata dipping to the right and horizontal to
interdune area between the dune arms, and the upper set dips subhorizontal strata corresponding to crest deposits. F, Former
outward. C, Collapse structure caused by sand flow filling near- crest area showing transition from crest-area strata into foreset
by burrowing crab tube. Strata are dipping eastward in trench. strata. (Fig. 82.)

Digitized by Google
 SEDIMENTARY STRUCTURES IN D~ 127
Textural changes are numerous among the dis- The new internal structure commonly is charac-
sipation sands at Joaquina Hill. A mixing of tex- terized by an irregular, wavy pattern accented by
tures occurred when the dune sands slid differences in concentrations of colloidal material
downslope, destroying original structures. Moving composed predominantly of clays, hydrous iron ox-
sands contributed detrital material from the un- ides, and humic compounds (fig. 89). Previously
derlying weathered mantle. Mud and sand flows these features generally were ascribed to soil-form-
containing granules, from upslope, mixed with and ing processes; now, they are known to be the result
covered sliding eolian sands. This formed the final of reworking by dissipation processes. The wavy
sediment layer of an unsorted light red sandy gra- features are especially prominent because they are
nule deposit. outlined with colloidal material along preferential
An erosional unconformity, with local paleosol lines. The concentration of colloidal material
(A horizon) development, separates each sequence follows irregular directions that do not correspond
of eolian sand. Former dunes in the two intermedi- to any previous stratification planes or dune struc-
ate sequences also underwent mass movement, tures. At Joaquina Hill, the colloidal material is
causing structural and compositional changes simi- composed predominantly of clay minerals and hy-
lar to those previously discussed. As a result, each drous iron oxides, as well as minor amounts of
successive sequence has fewer granules. humic compounds. Apparently, colloidal material
tends to concentrate preferentially along structural
patterns that developed during the dissipation of
Processes, Structures, and Soil former dunes.
Formation Flash floods cause the mineral colloidal material,
such as from a weathered granitic mantle, to mix
Dunes with the eolian sands. The amount of mineral mat-
ter reinforcing the structures is variable - in most
The original eolian deposits of the Lagoa dune places, less than 5 or 10 percent. After every sand
field consist of reversing parabolic and captation flow, the colloidal material either concentrates at
dunes, such as stringers and hanging dunes. Sedi- the uppermost section of the sedimentary layer as a
mentary structures differentiate dune sands from "peel" or filters down into lower layers. In the
beach sands of the neighboring source area and the lower layers, colloidal material tends to settle along
patterns of eolian cross-strata allow differentiation structural lines, which makes the strata more
of dune types. readily visible. Structures later become vague as the
At times large parts of dunes or entire dunes un- colloidal enrichment progresses. With an increasing
dergo subsequent structural changes, causing dis- content of colloidal material, the structures lose
appearance of the characteristic eolian cross-strata identity, and the sediments develop a uniform,
patterns. The original structures are replaced by brown color.
much less clearly defined patterns or by structure-
less deposits - that is, dissipation structures. Close to the source of weathered granite mantle,
Structures produced by reworking of eolian the usually thin colloidal peel increases in thick-
sands are not as easily visible as those of original ness and becomes a colluvial layer. Such layers
alternate with reworked eolian sands. The develop-
dunes, but they constitute new sedimentary struc-
tures - that is, dissipation structures. Most of these ment of a colluvial layer over a reworked eolian
sand produces an irregular surface with tongues
structures resemble contorted bedding, but they are
produced by heavy density flow or by loading. and other features that indicate inclusion of the
Some resemble cut-and-fill structures; others sand in the colluvial layer. The overlying colluvial
follow shearing planes and small intraformational layer produces contorted structures in the underly-
faults produced by sand flowage. Tube-shaped ing sediments.
features, reflecting root growth or the burrowing of In surface accumulations deposition of colloidal
animals, such as crabs, also occur in these deposits. material may be contemporaneous with deposition
Small circular or irregular patches may suggest of sand layers. Downward migration of colloidal
root growth. material is a penecontemporaneous feature. The in-

Digitized by Google
128 A STUDY OF GLOBAL SAND SEAS

II II

j~~~~~~~~~~6~ Set
I
2
3
Dip'
30
5 Strike
3
264
274
64
4 37 96
5 24 280
6 8 316
7 6 182
A 1--_ _- 8
9
18 90
5 264
~:.M:;=:.:.:,;.::.:.::.:""",=:..:.;.c:;==""", l~ l~ ~~~
12 13 246
13 32 282
14 15 289
15 28 86
r-"r------:----"'I<"~=::::~-_, l~ ~~ 2r,
18 8 307

4 I

Set Dip Strike"

1m
I 8 341
2 9 93
3 0
4 8 38
5 4 271
6 8 85
7 8 71
3 --------;:;:rr------------------- Set Dip' Strike"
~", : ~ 14 354

~'
7 349
,~ 3 14 344
4 8 334
5 9 351
6 9 49
7 15 39
8 20 327
9 13
10 10
133
6 6 ~=::;;:;;;:%7:::::::="-:::~::::::::J Set DiP' Strike
I 14 348
II II 30 2 8 27
12 25 328 3 24 321
4 30 327
5 10 289

II

7 I
Set DiP' Strike"
I 3 21
2 30 63 Set DiP' Strike
3 28 69 I 12 129
4 II 91 2 12 85
5 28 303 3 12 61
6 27 79 4 12 94
7 26 79 5 6 199
8 34 68 6 16 77
9 10 64 7 5 89
10 25 76 8 7 13
II 24 67 9 13 48
12 22 59

--::=.;::;:Jo-::::...........::......~--..:::......-.::......:~~..::.......::..........J 1:
13 10
~~
34
~; I

OJ
EXPLANATION

Sand strata set

Structureless sand impregnated with humic compounds;

locally includes incipient paleosol (A horizon)

Paleosol (A horizon)

1m
L -_ _ _ _ _---',

III
STRATIFICATION IN PITS 1-7 along east-west line in the interdune area between the arms of a parabolic dune in the Lagoa dune
field, IIha de Santa Catarina, Brazil. Dip and strike are given for each numbered cross-strata set within a pit. Pit locations are shown
in figure 77. Modified from Bigarella (1975b, figs. 14-16). (Fig. 83.)

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 129

~s=:::::~~~~=:::~ Set Dip Strike"

1 28 332
2 23 01
3 28 356
4 31 21
5 26 335

in
U~,
1/1

III

~r=:;===~J~J======1 Set
4
1 Dip"
15 Strike
225327
286
3 23 340
4 24 331
5 31 342
6 11 346
7 13 43

..;r----:::,,...-..:.:.."",..----:,.",..-, Set Dip" Strike

1 16 332
2 25 325
3 25 321
4 33 297
-----L___"""" -:o:~~!l!'JIl~:.:::J ~ ~~
7 22
m
314
8 25 312

~.
III

1m
~ _ _ _-.l1

STRATIFICATION IN PITS E1, E2, and E) along north-south line

of the interdune area between the arms of a parabolic dune,
lagoa dune field, IIha de Santa Catarina, Brazil. Dip and strike
are given for each numbered cross-strata set within a pit. light
stippled areas are nonstratified sand. Pit locations shown in
figure 77. Modified from Bigarella (197Sb, fig. 17). (Fig. 84.)
STRUCTURES OF INTERDUNE AREA between the arms of large
parabolic dune at lagoa dune field, IIha de Santa Catarina,
Brazil. A, Collapse structure of sand filling tube of burrowing
animal; B, set of high-angle cross-strata of earlier dune between
sets of interdune sub horizontal strata; C, high-angle faults in
foresets of earlier dune preserved in interdune area. (Fig. 85.)

Digitized by Google
130 A STUDY OF GLOBAL SAND SEAS

are locally very clear in some of the trenches

studied.
The highest concentration of organic matter is
found in paleosol (A horizon). Humic compounds
derived from the organic soil horizon and from the
decay of plant debris migrate downward through
the sand profile into the zone of saturation below
the water table. causing the sand to become brown.
In migrating. the humic compounds tend to impreg-
nate first on the shear planes and sedimentary
structures produced by the dissipation processes;
the resulting near-black stain makes the structures
easily visible. As the impregnation progresses
laterally. the structures tend to be less visible and
disappear into a dark-brown sandy area. In turn,
ROSE DIAGRAM OF CROSS-STRATA DIP DIRECTIONS, based the fluctuation of the phreatic water level. rich in
on 98 measurements; interdune area, Lagoa dune field, IIha de humic colloids. is responsible for additional impreg-
Santa Catarina, Brazil. (Fig. 86.)
nation of the organically cemented dark-brown
troduced material follows and reinforces the prim- crusts of layers.
ary structures developed by reworking of the eolian
sands (dissipation of former dunes). In this way. Interdune Areas
the apparently structureless pattern that originated
during the dissipation process becomes more evi- In the eastern part of the Lagoa dune field. not far
dent. from the granite hill. several trenches were dug in
In addition to the dissipation action of water. an area of parabolic dunes. The sediment in these
these dunes are subjected to soil-forming processes trenches contained dissipation structures repre-
that create soil horizons. The red and brown colors senting drier climate episodes. during which there
of the apparently structureless eolian sands are was a regional loss of vegetation and only sporadic
generally attributed to soil formation; however. at but intense rainfall. Flash floods had washed
the Lagoa dune field. most dunes investigated weathered material both from the side of the gra-
showed little red color to have been developed in nite hill and from the dune into the interdune area.
situ by such processes. The amount of heavy In the trench wall. sand- and granule-size stream
minerals present in the dunes is not sufficient to deposits and thin clayey beds deposited in small
provide the colloidal materials (clay and iron com- ponds were exposed. These deposits alternate with
pounds) for staining the quartz grains under normal sets of well-stratified dune sand. Thus. the dissipa-
soil-forming conditions. tion structures represent times when the flash
A surficial soil profile is not well developed on floods dominated in the development of dunes.
the dunes. Thus. the red and brown materials are
not commonly a result of downward migration of Deformation Results of Dune
colloidal material from soil formation. Large Dissipation
amounts of foreign materials. mostly colluvial.
have been introduced in the dune field by mud and The response to stresses and the types of struc-
(or) sand flows. which probably originated during ture produced during deformation depend on the
drier climates at times of regionally concentrated amount of moisture present in the sand (McKee.
rainfall. Douglass. and Rittenhouse. 1971). The cyclic se-
Foreign material obviously was introduced into quences at the Lagoa dune field were developed
the dune field. and the migration of this material under varying moisture conditions. Some units con-
definitely results from the action of percolating sist of typical eolian dune sediments; others are
waters and soil formation. Evidences of migration water-reworked dune deposits; some are mixed

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 131

N s
METRES
4

OO=~~~---==
2

8 12 METRES
A Interdu;'e area Interdu"ne area

~
N a _ _---II1m
~I
5
1m
'--_ _--11 5

a1_ _ _ _..Jlm 5

w ~ _ _...JI
O
.5mE B
= ==
8
c
=

EXPLANATION

_ _ Bounding plane

~ Cross-strata

28 Apparent dip. in
degrees

I~!~~~tlons ~f
~~~!~SING ~UNE AND INTERDUNE STRU
in d OS trenches; 8, trenches in ~~r~~~'
Lagoa dune field, IIha de Santa Catarin .
, 0 vert,cal exae8""Uon. Mo<Med from .8 dun.,; C, 'renehe> ;n ;n'eroune .re.. D" .",,1. A, Dune profile
Igarella (197Sb, fig. 13). (Fig. 87.) . Ip of cross-strata is shown

Digitized by Google
132 A STUDY OF GLOBAL SAND SEAS

Oune area

~.~
I LAGOA DUNE FIELD

EXPLANATION

Dsand Stabilized para- Precambrian crystalline

bolic dune basement; mostly granite

D Dissipation ramp 1 1 Surveyed para- - - - - Boundary; dashed

L-.-l bolic dune where approximate

DISSIPATION RAMP lEADING DOWN from Joaquina Hill to lagoa dune field, IIha de Santa Catarina, Brazil. Hachures in-
dicate inward slope. Map is study area of figure 76. Modified from Bigarella (1975b, fig. 2). (Fig. 88.)

Digitized by Google
 SEDIMENTARY STRUCTURES IN DUNES 133
674). Dissipation conditions at Lagoa, however,
were established by an increased amount of water.
Periodic mass movements in these dunes (fig. 90),
such as "wet avalanches," develop faults connected
with systems of shear planes, brecciated sand
masses, and folds. Downward movement of the
saturated sediments differs from normal avalanch-
ing of dry or damp sand. In a more advanced stage,
the action of sand flow and rainwash may spread
across the whole sand body producing its dissipa-
o 5 an o
.... tion, and the resulting structures display a wavy
~
o and lenticular pattern.
The avalanching of sand on the lee side of a dune
produces shear planes in which heavy minerals are
concentrated by downward migration through the
voids between grains. Locally, avalanches in wet
reworked dune sand resulted in wavy and lens-
shaped structures. This movement may have been
started in wet sand by undercutting at the base of
the deposit. The avalanching of wet sand is a mass
movement that develops more than one shear plane
(fault) parallel or subparallel to the main move-
CONTORTED WAVY PATTERN IN DUNE SAND formed by col- ment. A secondary system of shear planes com-
loidal material, mostly clay, separating individual sand-flow monly forms roughly perpendicular to the main
layers developed during dissipation processes. Cross-sections faults.
of dune normal to dip at lagoa dune field, IIha de Santa
Catarina, Brazil. (Fig. 89.)
Normal faults and breccias are common to
avalanching in wet sand (McKee and Bigarella,
1972, p. 674). The faulting due to avalanching is
with sediments derived from the reworking of col- caused by heavy rainfall that soaks the dune.
luvial material from the nearby granite hills. Associated interdune sediments indicate the pre-
Under conditions normally present in an arid cli- sence of small ephemeral streams that may under-
mate, the coastal dunes of Lagoa would have cut the dune base to start sand movement.
formed primarily by the accumulation of sands that Scour-and-fill structures may form as part of a
avalanched down the steep lee faces. The most single process in which the troughs are scoured by
common structures from such avalanching, besides flowing water and subsequently filled with sand.
cross-stratification, are shear planes, thrust faults, Scouring and filling seem to be related to rainfall
high-angle asymmetrical folds, overturned folds, patterns and to the sediment carried in small
breccias, normal faults, fadeout laminae, and inter- ephemeral streams that flow in interdunal areas
tonguing sand lenses (McKee and Bigarella, 1972, p. during the dissipation process.

Digitized by Google
134 TUDY GLO SAND AS

----

o
I
25 em
I

SCH IC C SEa S iIIus . g fea such earin es, f brecc' san avy . ation ures
and e san ers th suit f dissip can s. Bla ots, c urn ; f, fa ws, w sand; gray
whitish sand; bs, brown sand; psA, paleosol A, enriched with organic matter. Hachures indicate predune surface. Modified from
Bigarella (1975b, fig. 20). 1 centimetre = 0.39 inch. (Fig. 90.)
Digitized by Google
Digitized by Google
 A STUDY OF GLOBAL SAND SEAS

DUNE FORMS AND WIND REGIME

ChapterF

By STEVEN G. FRYBERGER,
assisted by GARY DEAN

Contents
Page Page
Summary of conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . .. 141 Dune forms and associated wind environments ....... 150
Introduction ...................................... 142 Summary of observed wind environments of the
Acknowledgments ............................. 142 various dune types. . . . . . . . . . . . . . . . . . . . . . . . . .. 150
Methods of study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 143 Barchanoid (transverse) dunes. . . . . . . . . . . . . . . . . .. 151
Types of available wind data .................... 143 Subtypes and size of barchanoid dunes studied 151
Limitations of wind data. . . . . . . . . . . . . . . . . . . . . . .. 143 Summary of observed wind environments of
Evaluation of wind data ........................ 145 barchanoid dunes. . . . . . . . . . . . . . . . . . . . . . .. 152
Selection of a weighting equation. . . . . . . . . . .. 145 Examples of wind environments of barchanoid
Calculation of drift potentials. . . . . . . . . . . . . . .. 146 dunes.................................. 152
Calculation and plotting of sand roses ...... " 147 White Sands National Monument, New
Additional computations. . . . . . . . . . . . . . . . . . .. 148 Mexico, U.S.A. . . . . . . . . . . . . . . . . . . . . . .. 152
Classification of wind environments ................. 148 Namib Desert, South-West Africa ........ 153
Direction of surface winds .. . . . . . . . . . . . . . . . . . . .. 148 Cherchen Desert, China ................ 154
Energy of surface winds ........................ 150 Ala Shan Desert, China ................. 155

137

Digitized by Google
138

Dune forms and associated wind environments - Continued

...
A STUDY OF GLOBAL SAND SEAS

Dune forms and associated wind environments -

PaBe
Continued
Linear dunes .................................. 155 Linear dunes - Continued
Summary of observed wind environments of Examples of wind environments - Continued
linear dunes . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 155 Erg Hammami region, Mauritania, and Erg
Bilma region, Niger ................... 164
Examples of wind environments of linear dunes 156 Star dunes .................................... 165
Southwest Kalahari Desert, South Africa .. 156 Summary of observed wind environments of star
Western Simpson Desert, Australia . . . . . .. 157 dunes .................................. 165
Ramlat Zallaf sand sea, Libya. . . . . . . . . . .. 161 Examples of wind environments of star dunes. 165
Northwestem Kalahari Desert, South-West Eastern Grand Erg Oriental, Algeria-Libya . 165
Africa, and Great Victoria Desert, Grand Erg Occidental, Algeria ........... 165
Australia ............................ 163 Western Grand Erg Oriental, Algeria ..... ,' 166

Illustrations
Page
FIGURE 91. Photographic copies of N summaries of surface winds from which estimates of
relative sand drift were made ...................................... 144
92. Graph showing sand transport rate versus shear velocity of air motion ...... 145
93. Sketch showing sand rose used in this chapter ........................... 147
94. Graph showing drift potential versus rate of sand drift for two threshold drag
velocities ........................................................ 148
95. Wind rose for Juraid Island, Arabia, showing relative numbers of wind observa-
tions from 360 of the compass .................................... 149
96. Sand roses showing five commonly occurring directional distributions of effec-
tive wind regimes ................................................ 149
97. Sand roses depicting typical high-energy wind regimes of three basic dune
types ............................................................ 151
98. Graph showing drift potential versus RDPIDP for 14 weather stations near
barchanoid dunes ................................................ 152
99. Landsat imagery of barchanoid and parabolic dunes at White Sands National
Monument, New Mexico, U.S.A., and sand roses for Holloman Air Force
Base downwind of the dune field .................................. 154
100. Landsat imagery of barchanoid dunes along the coast of northern Namib
Desert, South-West Africa ......................................... 155
101. Landsat imagery of barchanoid dunes in the eastern Cherchen Desert, China,
and sand roses for Charkhlik (Nochiang), China ...................... 156
102. Wind rose for maximum winds and table of average maximum monthly winds
at Charkhlik (Nochiang), China, 1956--f11 ........................... 156
103. Landsat imagery of barchanoid dunes in the westem Ala Shan Desert near Mao-
Mu (Ting-hsin), China, and sand roses for Mao-Mu ................... 157
104. Graph showing drift potential versus RDPIDP for 13 weather stations near linear
dunes ........................... :. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 158
105. Landsat false-color imagery of linear dunes and sand streaks in the south-
western Kalahari Desert near Upington, South Africa, and sand roses for
Upington ........................................................ 159
106. Wind rose for maximum winds and table of average maximum monthly winds
at Upington, South Africa, 1956--f17 ................................. 160
107. Landsat imagery of linear dunes in the southwestern Simpson Desert near Ood-
nadatta, Australia, and sand roses for Oodnadatta .................... 160
108. Landsat imagery of linear dunes in the Ramlat Zallif sand sea near-Sabhi, Libya,
and sand roses for Sabha . . . . . . . . . . . . . . . . .. ....................... 162
109. Wind rose for maximum winds and table of average maximum monthly winds
at Sabhll, Libya, 1949 and 1961 --fl3 ................................. 162
110. Landsat imagery of presumed linear dunes in the Peski Karakumy near Ekidze,
U.s.S.R., and annual sand rose for Ekidze . . . . . . . . . . . . . . . . . . . . . . . . . . .. 163

Digitized by Google
 ...
DUNE FORMS AND WIND REGIME 139

FIGURE 111. landsat imagery of linear dunes in the northwestern Kalahari Desert near
Mariental, South-West Africa, and annual sand rose for Mariental ...... 164
112. landsat imagery of linear dunes in the Great Victoria Desert near Giles,
Australia, and annual sand rose for Giles ............................ 164
113. Wind rose for maximum winds and table of average maximum monthly winds
at Mariental, South-West Africa, 1960--67 ............................ 165
114. landsat imagery of linear dunes in the Erg Hammami region, northern
Mauritania ....................................................... 165
115. landsat imagery of linear dunes in the Erg Bilma near Bilma, Niger, and sand
roses for Bilma ................................................... 166
116. Graph showing drift potential versus RDPIDP for five weather stations near star
dunes ........................................................... 167
117. landsat imagery of isolated star dunes in the Grand Erg Oriental near
Ghudamis, libya, and sand roses for Ghudamis ...................... 167
118. landsat imagery of star dunes along an escarpment in the Grand Erg Occidental
near Beni Abbes, Algeria, and sand roses for Beni Abbes .............. 166
119. landsat imagery of alined star dunes in the Grand Erg Oriental near Hassi
Messaoud, Algeria, and sand roses for Hassi Messaoud ............... 168

Tables
Page
TABLE 14. Derivation of weighting factors for relative rate of sand transport by substitu-
tion of average wind velocities into the generalized lettau equation .... 147
15. Computation, from N summary, of vector unit total from the west-northwest,
Yuma, Arizona, U.S.A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 147
16. Average monthly and annual drift potentials for 13 desert regions, based on
data from selected stations ........................................ 150
17. localities used to evaluate the wind environments of barchanoid dunes .... 153
18. localities used to evaluate the wind environments of linear dunes ......... 158
19. localities used to evaluate the wind environments of star dunes ........... 167

Digitized by Google
Digitized by Google
 DUNE FORMS AND WIND REGIME 141

Summary of Conclusions The average annual amount of wind energy in

the deserts studied ranges from approximately 80 to
1Ir HE COMPARISON OF WIND REGIMES to 489 Vu. On the basis of rough groupings of deserts
\UI dune fonns on a worldwide basis has become in tenns of wind energy, the wind energy of desert
possible through two types of data. The first is re- environments is classified as follows: low energy,
cently available Landsat imagery for deserts of the OP less than 200 VU; intennediate energy, OP,
entire world, some parts of which are very difficult 200-399 VU; high energy, OP,400 VU or greater.
of access. The second is reliable surface wind infor- Sand roses are circular histograms representing
mation for these regions, available from the Na- the potential effects of winds from the directions of
tional Climatic Center, Asheville, North Carolina, the compass, as evaluated from the modified Lettau
U.S.A. In this report the morphologies of dunes equation. Effective (sand-moving) winds tend to
both from well-studied and from relatively less occur in directional groupings or modes, which are
studied parts of the world were compared with reflected in sand roses. The directional classifica-
local and regional winds. Characteristic wind en- tion of effective wind regimes as represented on
vironments were discerned for barchanoid, linear, sand roses is based on the relationships of modes
and star dunes. and includes the categories herein designated as
By making several simplifying assumptions, sur- "narrow unimodal," "wide unimodal," "acute
face winds were analyzed for direction and relative bimodal," "obtuse bimodal," and "complex."
amount of sand migration (drift), using the Lettau8
equation for sand drift: Barchanoid dunes, including barchans, domes,
and transverse ridge dunes, have similar wind en-
q. g/ C"p=V*z(V* -Yr*). vironments. Wind distributions associated with
these dunes generally are narrow unimodal or wide
This fonnula was modified to: unimodal but may be acute bimodal if one mode is
Q ex: VZ(V-Yr). t, much stronger than the other. Barchanoid dunes in
environments with high drift potentials seem to be
Where Q is a proportionate amount of sand drift, V associated with a higher ROPIDP than those in en-
is average wind velocity at 10-metre (33-ft) height, vironments with low drift potentials.
Yr is impact threshold wind velocity, and t is time Barchanoid dunes are observed in present-day
wind blew, expressed as a percentage in a wind wind environments of low, intennediate, and high
summary. energy (drift potential). Large fields of barchanoid
dunes occur in continental settings and in restricted
Values derived from this fonnula by substitution high-energy settings along coasts. Of all the dune
reflect the sand-moving capacity of the wind for the types observed, barchanoid dunes are associated
time period of the wind summary and are herein with the least variability in effective wind direction
known as drift potentials (OP) , which are for a given drift potential.
numerically expressed in vector units (VU). The
Linear dunes are associated with a variety of
direction and magnitude of the vector resultants of
effective wind environments - from low to high
drift potentials from the 16 compass directions are
energy. Effective wind distributions range from
herein known as the resultant drift direction (ROO)
wide unimodal through complex. The wind en-
and the resultant drift potential (ROP) , respec-
vironments of most linear dunes have a greater
tively. An index of the directional variability of the
directional variability of effective winds than do
wind is the ratio of the resultant drift potential to
those of barchanoid dunes for a given drift poten-
the drift potential, herein known as ROPIDP. The
tial. Simple linear dunes are a nearly exclusive type
greater the directional variability of the effective
over wide regions of the Peski Karakumy and the
winds at a station, the lower its associated ROPIDP
Kalahari, Simpson, Great Sandy, and Great Victoria
will be.
Deserts. The unvegetated dunes in the southwest
Kalahari and Simpson Deserts are associated with
K. and H. Lettau. Meteorology Department. University of Wisconsin. intennediate to high-energy bimodal or wide
Madison. Wisconsin: unpUblished data. 1975. Equation reproduced by pennission
of the authors. unimodal wind regimes. Partially vegetated dunes

Digitized by Google
142 A STUDY OF GLOBAL SAND SEAS

in the northwest Kalahari and Great Victoria Following the suggestion of Bagnold (1941, p. 184),
Deserts are associated with low-energy bimodal or winds were evaluated and defined in terms of their
complex wind regimes. potential sand-moving effectiveness through the
Simple linear dunes can occur in the same high- use of a suitable weighting equation applied to a
energy wind environment as compound linear standardized set of data.
dunes. This suggests that factors other than wind, Parts of this chapter on barchanoid, linear, and
such as available sand and the age of the dunes, star dunes are designed to give a summary of the
may determine the type of linear dune observed at author's observations, based on the included tables
a particular high-energy locality. Compound linear and on material in chapter K. In addition, examples
dunes are most frequently observed in intermedi- are given of observed wind environments of certain
ate- to high-energy wind environments, with wide dune types. These examples were chosen first to il-
unimodal or bimodal distributions of effective lustrate the most common wind environments in
winds. which the representative dune types were ob-
Star dunes have been observed in complex wind served, and second to illustrate problems requiring
regimes of low, intermediate, and high energy. The further research. Finally, examples were chosen,
complex wind regimes associated with star dunes where possible, to compare the wind environments
can differ in several respects from those associated of similar dune types in different regions. Examples
with linear dunes. Unlike linear dunes, star dunes were selected from localities where the author had
have been observed in a complex wind environ- most confidence that the winds measured at nearby
ment of high drift potential, in which complex wind meteorological stations were representative of the
distributions prevail during the windiest months, regional wind environment.
and in which the dominant modes are directly op- The classification of effective wind environ-
posed. ments used in this paper will enable the reader to
Star dunes in a high-energy wind regime near compare one effective wind environment with
Ghudimis, Libya, are isolated mounds, whereas another but is not intended to be a formal system.
star dunes in a low-energy wind regime near Beni Although the sand rose concept (discussed later)
Abbes, Algeria, seem to be connected by sinuous seems useful in delimiting the gross directional and
arms. Some star dunes are developed in linear pat- energy properties of effective wind regimes, the
terns atop "sand sheets," linear dunes, or possibly more important problem of the effect on
crescentic dunes. Wind regimes associated with morphology of rare, high-speed winds versus winds
alined star dunes in the two examples available of intermediate strength remains to be solved.
have characteristics intermediate between dune An important caution should be kept in mind by
types in terms of RDPIDP. the reader in regard to the observations in this
report. Although the available surface wind data
provide a good picture of present-day effective
Introduction winds, the maximum period of record seldom ex-
ceeds 15 years. In contrast, many of the observed
The following conclusions are theoretical dunes are very large and, therefore, possibly are
because the author was not able to visit many of the very old. Thus, some of the large dunes observed,
localities studied from Landsat imagery during the particularly in regions of low or intermediate drift
course of the investigation. Early in the study, potential, may in part have developed in response
however, it was observed that a single dune type to older, different wind regimes. The most reliable
commonly occurred throughout wide stretches of comparisons, therefore, are between dunes and
desert. For example, barchanoid dunes characterize wind environments with high drift potentials.
the eastern Takla Makan Desert of China; linear
dunes, the Erg MaktEHr of Mauritania; and star Acknowledgments
dunes, the Grand Erg Oriental of Algeria. Thus, if
the winds ofthe various regions could be defined, it Technical assistance and criticism for this
seemed possible to obt~in insight into the wind chapter were generously given by Robert P. Sharp,
regimes responsible for creating major dune types. California Institute of Technology; Heinz Lettau,

Digitized by Google
 DUNE FORMS AND WIND REGIME 143
University of Wisconsin; Douglas L. Inman, Scripps observer bias (Ratner, 1950, p. 185), is the tendency
Institute of Oceanography; Harold L. Crutcher, Na- for weather observers to record wind occurrences
tional Climatic Center; Harry Van Loon, National from the prime directions rather than from inter-
Center for Atmospheric Research; and Errol L. mediate directions of the compass. For example,
Montgomery, Northern Arizona University. when an observer is uncertain, he commonly
records wind as coming from the northeast rather
than from the east-northeast.
Methods of Study A second type of inaccuracy occurs during data
gathering because of deviations from standard ob-
Types of Wind Data Available serving conditions as specified in the W.M.O.
"Guide to Meteorological Practices." One common
SAND DRIFT REFERS to the process of sand occurrence is the mounting of an anemometer at a
movement across the desert as a result of surface height other than the standard 10 m (33 ft) specified
winds. Most estimates of sand drift given in this by the World Meteorological Organization. For ex-
report were made from surface wind tabulations ample, the anemometer height at El Golea, Algeria,
known as "N summaries." These were prepared by was 4 m (13 ft) during 1949-56, 22 m (72 ft) during
the Environmental Technical Applications Center 1956-60, and 7 m (23 ft) from 1960-73. Addi-
of the U.S. Air Force. They are stored at the Na- tionally, a station may be sheltered from the wind
tional Climatic Center, Asheville, North Carolina, by nearby trees, buildings, or high ground.
U.S.A., and are available by station name or by
World Meteorological Organization (W.M.O.) num- The mounting of anemometers at nonstandard
ber. Each W.M.O. station summary normally con- heights may have slightly affected some calcula-
tains both monthly and annual data. Wind speed is tions of drift potential, because threshold drag
recorded in knots to the nearest 10 of direction at velocity assumed in calculations is based on wind
3- to 6-hour intervals. The period of record for sta- velocity at a 10-m height. If anemometers are
tions used in this chapter averages 10 years. mounted lower than the standard 10 m, calculated
drift potential will be slightly less than the true drift
Two N summary formats are available (fig. 91). potential which would result from calculations
The first format (fig. 91A) is more useful because based on a 10-m height. This is because wind
wind velocities are divided into 9 or 11 categories, velocities are lower near the ground; thus, the
whereas the second format (fig. 918) has only 5 theoretical threshold velocity will be exceeded less
velocity categories. About 100 summaries using the often.
first format were analyzed. The linear-regression
Inaccuracy also arises during summarization of
technique was used to estimate relative sand drift
data, and it usually develops in two principal ways.
from 34 summaries in the second format.
First, inaccuracy enters a summary when data is
Detailed wind data, other than the N summaries,
condensed from 36 to 16 compass directions. This is
were also obtained from some government offices,
known as procedure error (Wallington, 1968, p.
desert research organizations, corporations, and
293). The result is to create an apparent increase in
libraries. These data were reduced to a form similar
observations from the prime compass directions at
to that shown in figure 91A, then evaluated by the
same methods. the expense of the intermediate directions. Second,
summarizing of observations results in a coarsening
of the resolution of the data in terms of velocity,
Limitations of Wind Data direction, and percent occurrence. Most percen-
tages on N summaries are expressed to the nearest
The surface-wind data used in this report 0.1 percent (fig. 91A). Depending on the number of
generally was of good quality. However, methods observations, however, single occurrences may be
of gathering data and summarization processes both represented in a summary as more than 0.1 percent.
introduced some systematic inaccuracies. For example, an easterly maximum wind of 17
Inaccuracies introduced during data gathering knots was recorded during October at T'ieh-kan-li-
occur primarily in two ways. The first, known as k'o, China, for which period only 137 observations

Digitized by Google
 A STUDY SAND SEAS

DATA PROCESSI NG DI PERCE OF WIND SURFACE WINDS

ETAC/USAF ED
AIR WEATHER SERVIC
(FRO TlONS)

23195 YUMA ARIZONA lAP 48-71 ALL

STATION STATION NAME YEARS MONTH

ALL WEATHER ALL

CLASS HOURS IL.S.T .

CONDITION

SPEED
(KNTS)
OIA.
I J 1116 17 2 J 34 40 ., . "
MEAN
WIND
SPEED

N 1.0 1.8 .4 9.7 8.0

NNE 1.0 .7 .1 7.7 6.6
NE .8 2.1 1.1 .2 .0 .0 .0 4.3 5.8
ENE .5 1.1 .4 .1 .0 .0 2.2 5.7
E .7 1.0 .3 .1 .0 .0 .0 2.2 5.3
ESE .4 .8 .4 .1 .0 .0 1.7 5.8
SE .6 1.4 1.9 1.6 .3 .0 .0 5.8 8.9
SSE .5 1.9 3.0 2.6 .4 .0 .0 8.5 9.5
S .8 2.8 3.2 1.6 .1 .0 8.4 7.8
ssw .8 1.7 1.8 .7 .0 .0 .0 4.9 7.1
sw .8 .3 .0 5.6 6.2
wsw .6 .4 .0 5.7 6.5
W .7 .9 .2 7.3 7.5
WNW .6 1.3 .3 .0 0 6.9 8.2
NW .8 .5 .1 4.3 8.9
NNW .8 .9 .1 5.1 7.8
VARBL
CALM
><
10.8 33.8 29.6
>< .><
13.8 2.1 .3
::::..1>< :><
.0 .0 .0
9.6
100.0 8.7

TOTAL NUMBER OF OBSERVATIONS

180705

AIR WEATHER QUENCV OF SURFACE W SUMMARV.,

CLIMATIC CEN

OATA PAOCES 1059 E 2OJ6 R

VEARS 49.50.5 .81.62.63

DIRECTION

N
NNE
NE
ENE
5.8
2.6
2.1
.6
.4
WIND SPEED GROU
07-16
10.9
3.8
3.2
.9
.5
17-27
2.1
1.0
1.1
.1
.4
.0

.1
GTR40
18.8
7.4
6.5
1.6
1.4
-184
180
40
35
MEAN WIND
SPEEO/KTS
10.2
10.0
10.5
9.0
13.7
ESE .2 .2 .1 .1 .5 13 15.8
SE .2 .8 .3 .2 1.5 37 16.8
SSE 1.2 1.4 3 36 87 16.5
S 5.1 4.0 .1 15.2
SSW 2.2 1.2 .0 14.2
SW 3.6 1.9 .2 15.4
WSW 3.6 .9 .0 11.3
W 4.6 1.8 10.2
WNW 1.3 .1 7.9
NW 1.3 .1 8.0
NNW 1.0 2.0 .1 3.1 78 9.2
VARIABLE
CALM 8.2 203
TOTALS 27.3 45.2 16.5 2.4 .4 100.0 2478 10.8
MAXIMUM WIND SW 50 KTS

8
TWOTY MARIES contain d data from wh i relative
sand dri , 16 directional 11 velocity categ ectional
categori y categories. (Fi

:"VU I.J] _ "-'

 DUNE FORMS AND WIND REGIME 145
are available. This single observation is represented Additionally.
as 0.7 percent of all observations. It is questionable
whether this occurrence represents a group of C" =C' (3/3*)n ,
winds which blew 0.7 percent of the time. or 5.2
hours of the 744 hours. in October. where
Corrections were not made for the limitations C' = universal constant for sand (about 6.7).
just described because it was not known when and 3 = grain diameter of sand moved.
where they occurred. Furthermore. techniques 3* = 0.25 mm (standard diameter). and
available for correcting observer bias and procedure n = empirical constant ~0.5.
error (Wallington. 1968. p. 296; Ratner. 1950. p. 186)
could not be applied uniformly to the different All units are c.g.s. (centimetre-gram-second).
types of wind summaries used in this study. More
detailed future work involving the collection and 1.00 ...---r--r---r-----,,-----,,---r------r1 IIXl
summarization of surface-wind data would be im-
proved by taking the factors previously discussed 700
into account. However. these factors probably did
not affect the data enough to detract from the 600
general conclusions of this study.

Evaluation of Wind Data

BAGNOlD .t1J.
Selection of a Weighting Equation qg/p=3.33 'r\//
~/
,~'"
A number of equations are available to compute ,~"
,'0 0
rates of sand drift if the shear velocities are known. ",
~ o.~ 200
The most useful formulas. including those of Bag- " KAWAMURA
~
"/00
nold (1941). Kawamura (1951). and Zingg (1953) ,,/ ~q9/p-3.1 ('r+30)2 ('r-30)
. " ,"" v 100
were tested by Belly (1964. p. 3 -5) with data from
/"
his wind-tunnel studies. All formulas were found
O~~~/~_L_ _L_-L_-L_~-~O
by Belly to describe the data well when suitably ~ ~ ~ ~
evaluated. A formula recently suggested by K. and SHEAR VELOCITY.V". IN CM/SEC
H. Lettau (written commun . 1975) produces a
RATE OF SAND TRANSPORT (grams moving per second across
theoretical curve that agrees very well with Belly's
a crosswind distance of 1 cm (0.40 in.)), q, versus shear velocity,
work for O.44-mm-diameter sand (fig. 92). For this V* , of air motion, according to theoretical curves by Bagnold
reason. the Lettau equation for the rate of sand drift (1941), Kawamura (1951), and Heinz Lettau (1975, unpublished
was chosen in this study as the basis for the weight- data) in comparison with wind tunnel measurements (circles)
ing equation for sand movement. although other by Belly (1964, p. 17). Reproduced by permission of the
formulas probably would have given similar authors, K. and H. Lettau, Meteorology Department, University
of Wisconsin, Madison, Wisconsin, U.S.A (fig. 92.)
results. Lettau's equation is as follows:
Surface conditions. in addition to shear velocity.
qg = V*2 (V* - V*) (1) control the rate of sand drift. Four important factors
C"p t
are mean grain diameter of the sand (Bagnold. 1941.
where p. 67; Belly. 1964. p. 13). the degree of surface
q =rate of sand drift; roughness (Chepil and Woodruff. 1963. p. 240; Bag-
g =gravitational constant; nold. 1941, p. 71). the amount and kind of vegeta-
C" = empirical constant based on grain size; tive cover (Chepil and Woodruff. 1963. p. 221). and
p =density of air; the amount of moisture in the sand (Belly. 1964. p.
V* =shear velocity; and III 21). These parameters could not be evaluated for
V: = impact threshold shear velocity. or the many of the localities studied. For this reason,
minimum shear velocity required to localities were compared using relative quantities
keep sand in saltation. of potential sand drift.

Digitized by Google
146 A STUDY OF GLOBAL SAND SEAS

The Lettau equation for the rate of sand transport average diameter of many desert dune sands
was generalized as follows: (Ahlbrandt, chapter B, fig. 21). The assumed sur-
face might not serve to predict actual rates of sand
drift in areas with very large dunes. but it is useful
(2) when comparing one area to another in terms of
available wind energy.
Drag velocity is proportional to wind velocity for
a given height (Belly, 1964, p. 18). Therefore, as a A threshold wind velocity must be determined in
first approximation, wind velocities at a 10-m (33- order to use a weighting formula. For a sand surface
ft) height (the standard W.M.O. anemometer of 0.30-mm average diameter quartz sand, the sur-
height) may be substituted for drag velocities. The face roughness factor (Z') as determined by Belly
Lettau equation then becomes: (1964, p. 11- 12) during sand driving was 0.3048 cm
(0.118 in.). The threshold wind velocity at height Z'
(3) (Vt) was 274 cm/s (107 in./s) and V: was 16 cm/s (6
in./s). V; may be extrapolated to a 10-m (33-ft)
where height (the height at which most of the wind data
V = wind velocity of at 10-m (33-ft) height. were collected) using the equation (Bagnold, 1941,
and p.104):
Vt = impact threshold wind velocity at 10 m
(33 ft) (minimum velocity at 10 m (33 V(lO m) = 5.75 V:log ~ + V; (4)
ft) to keep sand in saltation}.
When this is done, a value of 11.6 knots is obtained
This relationship produces a number which ex- for V"~
presses the relative amount of sand potentially
moved by the wind during the time it is presumed This value indicates that for the conditions
to blow. When the factor of time is added to equa- described, threshold wind velocity as measured at a
tion 3 (thereby creating equation 5), the resulting 10-m (33-ft) height should be within the 11-16 knot
number is here referred to as the drift potential velocity category on N summaries, such as that
(OP) and is a measure of the relative amount of po- shown in figure 91A. For this study a value of 12
tential sand drift at a station for a stated period of knots was chosen for V'(lOm)'
time. For convenience, the units of drift potential The assumption was made that a wind speed and
are here called vector units (VU) because wind direction component occurred in nature for an
velocities are treated as vectors. The detailed amount of time proportional to its percentage in the
method of computation of drift potentials is summary. With a few exceptions, such as that pre-
described next. viously described for T'ieh-kan-li-k'o, China, this
assumption seems reasonable because periods of
record are long, observations are taken at different
Calculation of Drift Potentials
times of day and night, and each annual summary
Assumptions Required to Apply the Weighting averages more than 1,000 observations.
Equation
In order to use a weighting equation to determine Derivation of Weighting factors
the effects of surface wind, the condition of the sur- and Application to Wind Data
face over which the wind blows must be assumed.
For our purposes, this surface is assumed to consist Weighting factors as used here are numbers
of loose quartz sand grains with an average which represent the relative rates at which winds
diameter of 0.25 -0.30 mm. The surface is further of differing average velocities can move sand.
assumed to be without bedforms larger than rip- These numbers are derived by substitution of
ples, to be dry, and to be without vegetation. Simi- values for wind velocity (average wind speed of a
lar surfaces have been used for most wind tunnel velocity category) into the weighting equation of
studies of sand drift, and 0.25 -0.30 mm is the Lettau (equation 3) as shown in table 14.

Digitized by Google
 DUNE FORMS AND WIND REGIME 147
TABLE 14. - Derivation of weighting factors for relative rate Calculation and Plotting of Sand Roses
of sand transport by substitution of average wind velocities
A sand rose is a circular histogram which repre-
into the generalized Lettau equation (equation 3 in text)
sents potential sand drift from the 16 directions of
(The value of V' (V - VI) is divided by 100 to reduce weighting factors to smaller
sizes for convenience in plotting sand roses. as described in text. Velocities ex-
the compass (fig. 93). The arms of a sand rose are
ceeding 40 knots are rare and are not computed)
N summary Mean velocity
proportional in length to the potential sand drift
Weighting
velocity of winds in from a given direction as computed in vector units.
factor
category (knots) category V V' (V VI) VO(V - Vtl/l00
Thus, a sand rose expresses graphically both the
11-16 laS 182.3 1.5 2.7
17 -21 19.0 361.0 7.0 25.3 amount of potential sand drift (drift potential) and
22-27 24.5 600.3 12.5 75.0 its directional variability.
28-33 30.5 930.3 18.5 172.1
34-40 37.0 1,369.0 25.0 342.3 These diagrams differ from paleocurrent roses
constructed from crossbedding dip directions,
because the arms point toward the direction from
The weighting factors represent rates of sand which sediment moved, that is "into the wind."
transport, and the percentages of wind occurrence Sand roses are based on surface wind only and thus
in the summaries represent the length of time dur- reflect only potential sand transport. The pattern
ing which the winds blew. Therefore, suggested by the sand rose may be considerably
modified by local conditions.
Q a: yz (V - V,) t ,
Arm
N
or Q a: (weighting factor) t,

where
(5)

Reduction factor
t
(not shown on
monthly sand
t = time wind blew, expressed as a percen- roses)
ROD
tage on N summary; and
ATAR- Station
Q = annual rate of sand drift.
To evaluate the relative amount of sand drift TYPE OF SAND ROSE USED in this chapter. Components are as
follows: Arm (vector-unit totals, plotted in millimetres), propor-
which potentially occurs at a station, the weighting tional in length to potential amount of sand drift from a given
factor derived from equation 3 for each velocity direction toward center circle. Reduction factor, number by
category is multiplied by the percentage occurrence which vector-unit total of each sand rose arm was divided so
of wind in that category for all 16 directions of the the longest arm would plot at <50 mm (2 in.). Of' (drift poten-
summary, and the results are totaled. This com- tial, in vector units), measure of relative sand-moving capability
of wind; derived from reduction of surface-wind data through a
putation is shown for a single direction in table 15. weighting equation (equation 3 in text). ROO (resultant drift
Calculations of this sort are tedious; therefore, a direction, in vector units), net trend of sand drift. Station, name
programmable calculator was used for much of the of meteorological station at which wind data was recorded. (Fig.
work. The most convenient method of expression 93.)
of the results is the sand rose, construction of which
is described later. Sand roses are plotted with a Hewlett-Packard
9810 A calculator and 9862 A plotter.? The vector
TABLE 15. - Q)mputation, from N summary, of vector unit unit totals from each direction are plotted in
total from the west-northwest, Yuma, Arizona, U.S.A.
(The drift potential at the station is the sum of the vector unit totals computed in
millimetres. When the vector unit total for any
the same manner from each of the 16 compass directions. Total vector units direction exceeds 50, all arms of that sand rose are
from west-northwest equals 18.6. Wind data shown in figure 91A) divided by two until the longest arm of the sand
Velocity category (knots) rose plots at less than 50 mm (2 in.) in length. The
11-16 17-21 22-27 26-33 34-40 number by which the arms are divided is known as
Weighting factor 2.7 25.3 75.0 172.0 342.3
Percent
occurrence ... 1.3 .3 .1 0 0 , Use of a specific brand name does not necessarily constitute endorsement of
Vector units .... 3.5 7.6 7.5 0 0 the product by the u.s. Geological Survey.

Digitized by Google
148 A STUDY OF GLOBAL SAND SEAS

the reduction factor and is shown in the center cir- detailed (fig. 91A) summaries for each computa-
cle of the sand rose (fig. 93). tion.
Vector unit totals from the different directions The ratio of the resultant drift potential to the
may be vectorially resolved to a single resultant. drift potential- the ROP/DP- was computed for
The direction of this computed resultant is herein all stations in the study. The RDP/DP is an index of
referred to as the resultant drift direction (ROO). as the directional variability of the wind. Where the
shown on figure 93. The ROO expresses the net wind usually comes from the same direction. the
trend of sand drift. or the direction in which sand ROP/DP approaches unity. In contrast. where the
would tend to drift under the influence of winds wind comes from many directions. the ROP/DP ap-
from the different directions. The magnitude of the proaches zero because the resultant drift potential
resultant drift direction may be found from the is very low.
same data. using the Pythagorean theorem. and is Although this study required only qualitative
herein referred to as the resultant drift potential estimates of rate of sand drift. quantitative rates
(ROP). The ROP expresses. in vector units. the net can be predicted using equation 1. The relationship
sand transport potential when winds from various of drift potential to annual rate of sand drift pre-
directions interact. dicted by this equation for two presumed threshold
drag velocities is shown in figure 94.
Additional Computations
Drift potentials were estimated for a number of
stations in India. Libya. and China. for which only
Classification of Wind
those N summaries in the less detailed format (fig. Environments
91B) were available. Percentages in each velocity
category of the figure 918 summary were appor- Direction of Surface Winds
tioned into two smaller velocity categories of the MOST SETS OF SURFACE WIND observations.
figure 91A summary. based on the ratios of adja- such as those used in this study. exhibit groupings.
cent percentages established by linear regression or distributions. in terms of both direction and
analysis. using data from 36 randomly selected speed (for example. fig. 91A. B). Some sets of sur-
face-wind-direction distributions may be described
50
a: as elliptical. or may be complicated because of mix-
~ ed land and sea breezes. through mixtures of
>
%40 seasonal flows. or for other reasons (Crutcher and
b Baer. 1962. p. 522). For example. the 380-point
! directional wind rose for Juraid Island. Arabia. (fIg.
~
i 30 95) suggests four groups of winds; a group from the
~
west. northwest. northeast. and southeast.
However. most directional observations at Juraid
~
a:
Q 20 Island are encompassed within groups of winds
Q
Z
from the west and northeast which make an obtuse
;a angle with each other. Sand roses plotted from sur-
u..
0 10 face wind data reflect such directional groupings.
Although surface wind distributions. such as that
i for Juraid Island. can be very complex in detail. ex-
11<...---1--'-00-----'200'----300--'----400'----500--'---------'600 perience indicates that. as a first approximation.
DRIFT POTENTIAl. IN veCTOR UNITS five relationships of directional distributions occur
frequently. These relationships form the basis for
DRIFT POTENTIAL (ANNUAl) VERSUS rate of sand drift (m 3/yr) the classification of directional charcteristics of
across a 1-metre (3.3-ft) section normal to the drift direction for
two threshold drag velocities, according to the Lettau equation
effective winds in terms of sand roses shown in
(equation 1 in text) in which ~ = impact threshold shear figure 96. This scheme was adopted in order to
velocity. line A, ~ = 16 cm/s (0.5 ft/s) (Belly, 1964, p. 11); lineB, simplify the discussion of wind regimes and associ-
~ = 19 cm/s (0.58 ft/s) (Bagnold, 1941, p. 60). (Fig. 94.) ated dune forms that follows.

Digitized by Google
 DUNE FORMS AND WIND REGIME 149
N 2. Wide unimodal. - any other directional dis-
380" tribution with a singls peak or mode.
3. Acute bimodal. - a distribution with two
modes, in which the peak directions of the
distributions (longest arms on the sand rose)
of the two modes form an acute angle (here
arbitrarily including also the right angle, 90).
4. Obtuse bimodal. - a distribution with two
modes, in which the peak directions of the
W 27(1' F=-iiiilil =---------i 90" E two modes form an obtuse angle.
5. Complex. - any distribution with more than
two modes, or with poorly defined modes.
The 16 point wind-direction data commonly
will not clearly show more than three modes.
Modes observed on sand roses are not considered
to be significant for purposes of classifying a wind
regime if the modal direction and the two adjacent
180" categories constitute less than about 15 percent of
S
the drift potential at the station. This simplifies the
WIND ROSE FOR JURAID ISLAND, ARABIA (Iat 27 .... 2'03" N., classification of wind regimes by focusing on the
long 4957'23" E.), showing relative numbers of wind observa- dominant modes that are controlled by large
tions for 360 of the compass. Lengths of arms of the rose are
proportional to the number of times wind came from a given
pressure systems. All sand roses tend to reflect pro-
direction. Winds from the west and northeast constitute the cedure and observer bias. Extreme bias results in a
dominant groups. Based on wind records during March "sawtooth" pattern of arms (example, fig. 111B).
10-April 10, 1971, observations taken at 15-minute intervals. This pattern can sometimes make unimodal or
(Data from Arabian American Oil Co.). (Fig. 95.) bimodal wind regimes seem complex.
The RDPIDP which is a measure of the direc-
The five commonly occurring wind regimes (fig. tional variability of the wind, is arbitrarily
96) are: classified as follows: 0.0 to less than 0.3, low; 0.3 to
1. Narrow unimodal. -90 percent or more of the less than 0.8, intermediate; 0.8 or greater, high.
drift potential at a station falls within two ad- Many low ratios (RDPIDP) are associated with
jacent directional categories, or within a 45 complex or obtuse bimodal wind regimes, inter-
arc of the compass. mediate ratios with obtuse bimodal or acute bimo-

c D
Area ....... Walvlablllli, South- Bahrain, Arabia Badanah. Arabia Timimoun, Algeria Ghudlmis, Libva
west Africa
DP .. ........... 518 540 528 246 658
RDP ... 448 435 327 46 46
ROPIDP . ............ 0.88 0.81 0.62 0.19 0.07

FIVE COMMONLY OCCURING RELATIONSHIPS of modes on sand roses. A, Narrow unimodal; 8, wide unimodal; C, acute bimo-
dal, D, obtuse bimodal, (a special example in which the modes are almost exactly opposed); and E, complex.LF(drift potential) and
RLF (resultant drift potential), in vector units. Arrows indicate resultant drift direction (ROO). (Fig. 96.)

Digitized by Google
150 A STUDY OF GLOBAL SAND SEAS

dal wind regimes, and high ratios with wide and the wind-energy structure of various desert
narrow unimodal wind regimes. regions, with regard to sand movement only, are
discussed in detail in chapter K.
Energy of Surface Winds Further, the relative wind energies (in terms of
sand movement) shown in table 16 are rough ap-
Drift potentials, which are measures of the
proximations. In fact, most desert regions are
energy of surface winds in terms of sand move-
strongly zoned in terms of wind energy. For exam-
ment, are classified according to rough groupings of
ple, high drift potentials occur in Northern
average annual drift potential for the desert regions
Mauritania; intermediate to low DP's occur farther
shown in table 16. The deserts in China (127 VU; 81
south toward the senegal River. (See chapter K, fig.
VU) and India (82 VU) and the Kalahari Desert of
South-West Africa (191 VU) constitute a low- 197).
energy group. The northern Saudi Arabian (489
VU) and Libyan (431 VU) deserts are a high -energy Dune Forms and Associated
group. All other deserts studied are in the inter-
mediate-energy group. On the basis of this classifi- Wind Environments
cation, low-energy wind environments have drift Summary of Observed Wind
potentials less than 200 VU; intermediate-energy
wind environments have drift potentials of
Environments of the Various Dune
200-399 VU; and high-energy wind environments Types
have drift potentials of at least 400 Vu. This group-
ing of regions by wind energy in terms of potential THE METHODS OF EVALUATING wind
sand movement applies only to the generally arid regimes discussed under "Methods of Study" in
regions surveyed during this study. Other regions this chapter were used to compare wind environ-
probably have different average drift potentials ments with different dune types as observed on
(perhaps very much higher than desert regions, Landsat imagery and on aerial photographs. The
many of which are known to be relatively calm). dunes and associated wind regimes will be dis-
However, within the regions studied, the cussed in order of presumed increasing complexity
differences in relative sand-moving power of wind of dune type - that is, barchanoid, linear and star.
are related to specific weather patterns. Aspects of A synthesis of the more detailed discussions of

TABLE 16. - Average monthly and annual drift potentials for 13 desert regions, based on data from selected stations
Ie. drift potentials estimated; leaders (...). no data)
Number Annual
of drift
Desert region stations Ian. Feb. Mar. Apr. May lune luly Aug. Sept. Oct. Nov. Dec. potential
Hlp-eaeray wiad eavlroameats
Saudi Arabia and Kuwait (An
Nafiid, north) ............... 10 35 39 52 54 51 66 49 33 20 18 16 25 489
Libya (central, west)* .......... 7 40 42 48 64 51 41 20 18 24 24 22 37 431
Intermedlate-eaeraY wiad envlroameats
Australia (Simpson, south) ..... 1 43 40 27 17 13 10 18 26 52 56 46 43 391
Mauritania ................... 10 45 49 45 38 33 40 26 19 20 20 19 30 384
U.S.S.R. (Peski Karakumy, Peski
Kyzylkum) ................. 15 39 41 43 43 33 25 22 21 23 23 24 29 366
Algeria ...................... 21 21 27 37 48 32 27 18 13 15 16 16 23 293
South-West Africa (Namib) .... 5 8 2 6 17 13 50 19 22 27 44 17 12 237
Saudi Arabia (Rub' al Khah,
north) ...................... 1 23 28 53 32 20 30 1 7 7 201
Low-eaeray wiad envlroameats
South-West Africa (Kalahari) ... 7 14 11 8 10 9 11 18 24 26 26 17 18 191
Mali (Sahel, Niger River) ....... 8 9 12 14 12 19 22 15 9 10 5 5 7 139
China ~Gobi)* ................ 5 9 11 16 23 20 11 7 5 5 5 7 8 127
India ( ajasthan, Thar)* ....... 7 2 2 5 5 10 21 19 9 5 2 1 1 82
China (Takla Makan)* ......... 11 3 2 9 16 16 9 9 5 4 5 2 1 81

Digitized by Google
 DUNE FORMS AND WIND REGIME 151

each dune type and wind regime is given in figure ery (chapter n,
including the barchan, barchanoid
f1l for comparative purposes. Figure f1l demon- ridge, and transverse ridge of the structural
strates the greater directional variability of effec- classification (chapter A) are included in the
tive winds associated with star dunes in contrast to following discussion. Unique wind environments
the less variable wind regimes commonly associ- for each of the subtypes, as defined in chapter A,
ated with barchanoid dunes (in high-energy wind could not be distinguished with the regional data
environments). available. These dunes commonly occur in the
Narrow unimodal to wide unimodal winds are same dune field (McKee, 1966, p. 4, fig. 1;
associated with transverse dunes, such as Ahlbrandt, lf1l5, p. 61). Dome dunes and parabolic
barchanoid ridges. Bimodal winds are commonly dunes may be found in association with barchans.
associated with linear dunes, and complex wind barchanoid, and other transverse-type ridges.
distributions, with star dunes. Dome dunes of the type studied by McKee at
White Sands National Monument, New Mexico,
Barchanoid (Transverse Dunes) U.S.A. (1966, p. 26), are not discussed in this
chapter because they could not be identified by the
Subtypes and Sizes of Barchanoid Dunes author on Landsat imagery.
Studied
Small barchanoid dunes occur both singly and in
Dunes referred to as crescentic in the environments where star dunes (Glennie, 1970, p.
morphological classification used for Landsat imag- 87, figs. 70, 71) or linear dunes (Glennie, lf1l0, p. 94,

A
I DP=518
ANNUAL
DP=9
FEB.
I ,
DP=30
APR.
DP=39
JUNE
DP=62
AUG.
DP=123
OCT.
DP=31
DEC.

t
B Dp.,661 DP=BB DP=53 DP=59 DP=46 DP=42 DP=52
ANNUAL FEB. APR. JUNE AUG. OCT. DEC.

DP=662 DP=50 DP=135 DP=87 DP=24 DP=31 DP=39

ANNUAL FEB. APR. JUNE AUG. OCT. DEC.

TYPICAL HIGH-ENERGY WIND REGIMES of three basic dune types. Annual and bimonthly sand roses depict distribution of effec-
tive winds at each station. A, Narrow unimodal; barchanoid dunes near Pelican Point, South-West Africa. 8, Bimodal; linear dunes
near Fort-Gouraud, Mauritania. C, Complex; star dunes near Ghudiimis, Libya. Number in center circle of each rose is reduction fac-
tor. DP (drift potential, in vector units) is given for each rose. Arrows indicate resultant drift direction (ROD). (Fig. 97.)

Digitized by Google
152 A STUDY OF GLOBAL SAND SEAS
3000
fig. 75} are the dominant type. In the latter places
they apparently develop in response to a seasonal
2500
component of a wind regime (Glennie, 1970, p. 92).
Often, relatively small dunes, such as those ~2000
:;:)
developed in response to a seasonal wind, are not a::
visible on Landsat imagery. With a few exceptions g 1500

(for which adequate field observations are availa- ~ 1000

ble), dunes discussed in this paper are large enough ~
to be clearly defined from Landsat imagery. These
dunes apparently have developed in response to the z
~ 800

annual wind regime, of which the annual sand rose

is an approximate measure.
~ 600
High
~ 400
iii:
0 Intermediate
Summary of Observed Wind Environments of 200

Barchanoid Dunes
Low

0
The wind environments of barchanoid dunes are 0 0.8 0.9 1.0

evaluated with data from the 14 stations listed in ta-

ble 17, and from regional analyses in chapter K. DRIFT POTENTIAL (DP) VERSUS RDPIDP for 14 stations near
Drift potentials range from low (51 VU, Pa-Ch'u, barchanoid dunes, which are represented by dots in the figure.
China) to extremely high (estimated 2,823 VU, All dunes except those near EI Centro, California, U.S.A., and
Pomona, South-West Africa). The data in table 17 Pomona, South-West Africa, were large enough to be observed
on Landsat imagery. Although based on few stations, the dis-
indicate that the barchanoid dune type is not tribution of points indicates that barchanoid (transverse) dunes
restricted to either a low- or high-energy wind en- occur in wind environments with less directional variability
vironment, assuming that dunes observed on the than those of linear or star dunes. (Compare with figs. 104 and
Landsat imagery are adjusted to present-day wind 116.) Barchanoid dunes in environments with high drift poten-
regimes. ActiveB barchanoid dunes seen on Landsat tials (>399 VU) are associated with higher RDPlDI's (less direc-
tional variability) than are barchanoid dunes in environments
imagery seldom occur in a wind environment with
with low drift potentials ( <200 VU). Based on data from table
an RDPIDP lower than 0.50 (fig. 98). They are 17. (Fig. 98.)
associated with less directional variability of effec-
tive winds where they occur in environments of
Examples of Wind Environments of
high energy (fig. 98). Barchanoid dunes develop in
wind regimes with less directional variability than Barchanoid Dunes
those of linear dunes or star dunes for a given White Sands National Monument, New Mexico, U.S.A.
energy environment. (Compare fig. 98 with figs.
104 and U6.) Annual wind regimes associated with Barchanoid dunes, dome dunes, and parabolic
barchanoid dunes may be unimodal or bimodal, dunes can develop in low-energy wind environ-
although in bimodal regimes, one mode normally is ments having considerable directional variability of
much stronger than the other. Extensive fields of effective winds. These three types of dunes occur
barchanoid dunes occur along coastlines swept by at White Sands National Monument, New Mexico
oceanic tradewinds or by monsoon winds, in both (McKee, 1966, p. 4). The annual sand rose for
high- and low-energy settings. In addition, Holloman Air Force Base, located downwind from
barchanoid dunes are the. dominant type in some the dunes at White Sands, has an obtuse bimodal
continental regions such as the Takla Makan and wind distribution (fig. 99). Effective winds from the
Ala Shan Deserts in China. north-northwest and from the southeast occur dur-
ing much of the year. However, these winds have
less potential effect than tlo the southwesterly
'Dunes observed on Landsat imagery are considered to be active where theyoc- winds, which are strongest during April (fig. 99C).
cur in regions of less than 100 mm (3.93 in.) averase annual rainfall. if vesetation The drift potential at Holloman Air Force Base is
(which appears red on false-color imagery) is not apparent. if they have distinct
slipfaces and sharp outlines. or if evidence of movement is available from aerial 149 VU-a low value compared with most other
phot08raphs or published reports. desert regions around the world (table 16).

Digitized by Google
 DUNE FORMS AND WIND REGIME 153
The alinement of parabolic dunes visible in of the Namib Desert. South-West Mrica. near
figures 99A and 99B is parallel to the resultant drift Pelican Point (figs. 97 A. 1(0). Although barchan
direction at Holloman Air Force Base. The gross dunes at this locality are not visible on Landsat
alinement of barchanoid dunes shown in figures (ERTS) imagery. they were observed along the
99A and 99B is roughly at right angles to the resul- coast by M. K. Seely (oral commun . 1975). This
tant drift direction. Individual slipfaces within the coastal region is swept by onshore winds derived
dune field may be oriented in various directions from the southeast tradewinds of the South Atlan-
during the year (McKee. 1966. p. 14) in response to tic Ocean. which are strongest during October in
the effective winds from several directions (fig. the region of Pelican Point. Although the energy of
99C). effective wind - and. thus. the drift potential
Namib Desert Coast, South-West Africa along the coastline - varies greatly during the year
(chapter K. fig. 232). effective winds in the region
Some barchanoid dunes occur in wind environ- as shown in figure 97 A are mostly from the south or
ments with a high drift potential and a narrow southwest and result in a narrow unimodal wind
unimodal wind distribution. For example. distribution. Barchanoid dunes occur in other
transverse ridge. barchanoid ridge. and barchan coastal settings as well. particularly along shores
dunes occur in such an environment along the coast swept by oceanic tradewinds or by monsoonal

TABLE 17. - Localities used to evaluate the wind environments of barchanoid dunes
lOP. drift potential; HOP. resultant drift potential)
Wind regime
Nearest representative Station classification.
Dune coordinates station coordinates Other using
Dune form(s) (lat, long)' (distance from dunes in km) (lat, long) data OP RDP/OP sand rose
Barchan .................... 3150' N., El Centro, California, U.S.A. 3249'N., Yes 525 0.75 Narrow unimodal.
1155O'W. (40). 1154O'W.
Near station ...... Pomona, South-West Africa 27"09' S., No '2,823 .9'1 Do.
(5-10). 1515' E.
Barchanoid ridge ............ 3820' N., Cherchen (Charchan), China 38"08' N., No 85 .60 Wide unimodal.
8530' E. (22). 8532' E.
3710' N., Khotan, China (56) ........... 37V1'N .. No 33 .70 Do.
7930' E. 7955' E.-
39"OO'N., Charkhlik (Nochiang), China' 3840' N., No 185 .79 Do.
8730' E. (40). 88"03' E.
3800'N., So-ch'e (Yarkand), China (SO). 3825' N. No 42 .69 Unclassified.-
78"00' E. 7715' E.-
3930' N., Pa-ch'u, China (30) ........... 3946'N., No 51 .51 Obtuse bimodal.
79"00' E. 7820' E.-
28"OO'N .. Hi'il, Arabia (40) ............. 2730' N., No (7) .78 Wide unimodal.
4140' E. 4140' E.-
Barchan, barchanoid ridge, Near station ...... Holloman Air Force Base, New 3218' N., Yes 149 .37 Obtuse bimodal.
parabolic. Mexico, U.S.A.' (5). 10655' E.
Barchanoid ridge, barchan .... . ..do ............ Walvisbaai, South-West Africa' 2253' S., Yes 518 .86 Narrow unimodal.
(0). 1426' E.
Compound and simple 4020' N., T'ieh-kan-li-k'o, China (20) ..... 4039' N., Yes SO .54 Wide unimodal.
barchanoid ridges. 8730' E. 8742' E.
Compound barchanoid ridge, 4015' N., 100 - Mao-mu (Ting-hsin), China' 4020' N .. No 140 .57 Obtuse bimodal.
and barchanoid ridge. 10030' E. (15). 9945' E.
Compound barchanoid ridge .. 18SO' N., Tidjikdja, Mauritania (30) ...... 1833' N., No 133 .64 Wide unimodal.
113O'W. 1126'W.
Small barchanoid ridges, also Near station ...... Bilma, Niger (5 -10) .......... 1841' N., No 948 .74 Narrow unimodal.
linear dunes. 1255' E.
, 1Cl-km-radius circle centered on this point.
"Yes" indicates airphoto or other data used to confirm identification of dune type.
, Estimated from Beaufort data, one year of record by John Rogers (written commun., 1975).
Coordinates approximate.
Discusaed as example in text.
Annual wind summary not available.
7 Not computed.

Digitized by Google
154 A STUDY OF GLOBAL SAND SEAS

C
~ ~
DP=149 DP=13 DP=30 DP=15 DP=7 DP=6 Dp..,8
ANNUAl FEB. APR. JUNE AUG. OCT. DEC.

LANDSAT IMAGERY OF WHITE SANDS National Monument, included for comparative purposes to illustrate the resolution ot
New Mexico, U.S.A. (A, 8), and sand roses for Holloman Air the bulk-processed Landsat imagery used in this study. C, An-
Force Base, New Mexico, near the downwind end of the dune nual sand rose and six monthly sand roses for Holloman Air
field (q. A, Produced from computer-compatible tape at the Force Base, illustrating the directional variability of effective
Center of Astrogeology, Flagstaff, Arizona; scale approximately winds near the dunes. Annual wind regime is obtuse bimodal.
1:290,000. Parabolic and barchanoid dunes are clearly visible. 8, Principal effective winds occur in April (DP = 30 VV) from the
Produced by bulk-processing facility at Sioux Falls, South southwest. Number in center circle of rose is reduction factor.
Dakota. At this scale (1 :1,000,(00) barchanoid dunes are barely DP (drift potential, in vector units) is given for each rose. Arrows
visible, but parabolic dunes can still be clearly seen. A and 8 are indicate resultant drift direction (ROO). (Fig. 99.)

winds, such as the Wahiba Sands of Arabia and the Most effective winds at Charkhlik (Nochiang),
sands of the Thar Desert of eastern Pakistan and about 40 km (25 mil southeast of the dunes shown
western India. The associated environments can in figure lOlA, are weak (estimated DPat Nochiang
have either high or low drift potentials. is 185 VU) but very steady in direction during the
year (fig. 1018). Thus, they do not provide a ready
Cherchen Desert, China
explanation for the different orientations of the
Several groups of barchanoid dunes may occur transverse dunes. Possibly one set of dunes is older
within a single dune field, each group characterized than the other and is inactive; however, no indica-
by a different size and orientation. In the eastern tions were noted from Landsat imagery, such as
Cherchen Desert, China, three sets of barchanoid vegetation, that any of the dunes are stabilized.
dunes have orientations differing from 45 to 90 Another possible explanation for several orienta-
degrees, depending on locality (fig. lOlA). Effective tions is that the dunes which are alined approx-
surface winds in the Cherchen Desert, and in the imately north-northeast to south-southwest have
eastern Takla Makan basin tend to be weak but developed transverse to the very strongest winds
very steady in direction (chapter K, fig. 206). that occur at the station. These winds are from the

Digitized by Google
 DUNE FORMS AND WIND REGIME 155
morphology from those of the Cherchen Desert to
the west. In the Ala Shan region smaller transverse
dunes grade westward into larger compound
barchanoid ridge dunes. In contrast, the transverse
dunes of the Cherchen Desert have developed as
continuous ridges with relatively continuous slip-
faces (fig. lOlA). Large discontinuous slipfaces oc-
cur on the compound dunes of the Ala Shan region.
These slipfaces are arranged to form a secondary
alinement at an angle of approximately 22 to the
annual resultant drift direction at Mao-Mu (Ting-
hsin) , the closest available wind station. Surface
wind data are not presently available for the in-
terior of the Ala Shan Desert. Regional analysis
(chapter K) suggests that winds in the Ala Shan
Desert are more variable in direction than those in
the Takla Makan Desert. A spring sand-moving
season occurs at both Charkhlik (Nochiang) and
Mao-Mu (Ting-hsin). But winds during April,
which is the month of highest drift potential at both
stations, are about twice as effective at Charkhlik
(Nochiang) as they are at Mao-Mu (Ting-hsin). The
LANDSAT IMAGERY OF THE NORTHERN PART of the Namib winds at Charkhlik (Nochiang) are also more direc-
Desert, South-West Africa, showing barchanoid dunes along the tionally steady during this time (compare April
coastal margin of the desert. The wind regime which has pro-
sand roses for Charkhlik (Nochiang), fig. 101B, and
duced the dunes is probably restricted to the coastal region .
This assumption is based on the distribution of the dunes and
Mao-Mu (Ting-hsin), fig. 103B).
other evidence (chapter K). Large linear dunes occur inland. Simple barchanoid ridge dunes near the station of
Streaks north of the Kuiseb River, oriented northeast-southwest, Mao-Mu (Ting-hsin) occur in a wind environment
are alined roughly with the resultant drift directions at the inland
similar to that of White Sands National Monument,
stations of Zwartbank, Rooi Bank, and Gobabeb during June
and July, the windiest months at these stations. Sand roses for
New Mexico, U.S.A., previously described. The
Pelican Point are shown in figure 97 A. (Fig. 100.) sand roses for each station are similar (compare an-
nual sand roses, figs. 99C, 103B), and the estimated
drift potential at Mao-Mu (Ting-hsin) (140 VU) is
east-northeast for 8 months of the year (figs. lOlA, close to the drift potential at White Sands (Hollo-
102). man Air Force Base, 149 VU).
Occasional extreme gusts or winds, not sustained
long, may have important morphological effects in
such regions as the Cherchen Desert, where winds Linear Dunes
of intermediate strength occur infrequently and Summary of Observed Wind Environments of
where the region as a whole has a low drift poten- Linear Dunes
tial. A more detailed analysis of this problem with
respect to the Cherchen Desert region, shown in The active linear dunes observed on Landsat im-
figure lOlA, is precluded because of limitations in agery occupy a wide range of wind environments in
available wind data. This example is presented, terms of both energy and directional variability.
however, to illustrate the development in the same Drift potentials associated with linear dunes range
field of very large transverse dunes with orienta- from 45 VU (estimated) in an interdune corridor at
tions differing as much as 90. Narabeb, South-West Africa, to 948 VU at Bilma,
Niger (table 18). The wind environment of a linear
Ala Shan Desert, China dune seems to have a greater directional variability
Barchanoid dunes of the western Ala Shan of effective winds (lower RDPIDp) than does the
Desert of China (fig. 103A) differ greatly in wind environment of a barchanoid dune, for a

Digitized by Google
156 A STUDY OF GLOBAL SAND SEAS

LANDSAT IMAGERY OF THE EASTERN CHERCHEN DESERT,

China showing (A) dune field with three sets of transverse
dunes. Two large sets of approximately equal size strike roughly
northeast and northwest; a smaller set also strikes northwest.
One set of dunes may be oriented transverse to strongest winds
in the region indicated by arrow, and other sets may be
transverse to more frequently occurring winds of lesser strength.
Resultant of latter winds corresponds roughly to estimated an-
nual resultant drift direction (ROD, indicated by arrow) at
Nochiang. 8, Annual and six monthly sand roses for Charkhlik
(Nochiang), China, show that most of the effective winds at the
station, and in the eastern Cherchen Desert, are from the north-
east quadrant throughout the year. Number in center circle of
rose is reduction factor. DP (drift potential, in vector units) is
given for each rose. Arrows indicate resultant drift direction
(ROO). (Fig. 101.)

o
B DP=3 DP=44 DP=15 DP=15 DP=12 DP=1
ANNUAL FEB. APR. JUNE AUG. OCT. DEC.

given drift potential (fig. 104). Also. a linear dune complex wind regimes. Junctures are least common
may require less directional variability of effective at localities with high-energy wide unimodal wind
winds with increasing energy of wind environment regimes.
(fig. 104). Wind distributions associated with active
linear dunes range from wide unimodal through Examples of Wind Environments of Linear
complex. Linear dunes are usually alined parallel to Dunes
the resultant drift direction of effective winds in Southwest Kalahari Desert, South Africa
the surrounding environment. This is true whether
the associated wind distribution is wide unimodal. Simple linear dunes are the nearly exclusive type
bimodal. or complex. The alinement is closest at visible on Landsat imagery of the Kalahari Desert of
localities where wind data are most reliable and South Africa (chapter K). Active linear dunes and
dunes most clearly active. as at Fort-Gouraud. sand sheets occur near Upington. in the south-
Mauritania; Upington. South Africa; and Sabha. western part of the Kalahari Desert (fig. 105A). The
Libya. Junctures ("tuning fork" or "Y") are a com- linear dunes and the sand sheets are alined with the
mon feature of linear dunes and were observed at resultant drift direction of a high-energy obtuse
most localities. The junctures usually open into the bimodal wind regime (fig. 1058). Both the north-
effective wind (against the resultant drift direction) westerly and southwesterly components of the
and are most common at localities with low-energy wind regime are persistent throughout the year.

Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.

Direction from which

ENE. ENE. WSW. ENE. NNE . W. ENE. ENE . ENE . NE . ENE . ENE .
maximum winds came

Speed (knots) 35 27 27 35 35 31 27 31 31 31 27 19

WIND ROSE FOR MAXIMUM WINDS and table of average maximum monthly winds at Charkhlik (Nochiang), China, for 1956 -61.
Each tick mark on a wind-rose arm represents a maximum gust from the direction indicated by the arm. (Fig. 102.)

Digitized by Google
 DUNE FORMS AND WIND REGIME 157

t ;p --t( ~ ~ ~ # .
B ~140 ~15 DP-27 ~6 OP-4 ~9 DP-13
ANNUAL FEB. APR. JUNE AUG. OCT. DEC.

MOSAIC OF TWO LANDSAT IMAGES (A) of the western Ala direction (ROO). B, Annual and six monthly sand roses for Mao-
Shan Desert, China. Small barchanoid ridge dunes grade east- Mu (Ting-hsin), China, illustrating directional variability of low-
ward from the region of Mao-Mu (Ting-hsin) into larger com- energy effective winds near the westernmost part of the Ala
pound barchanoid ridge dunes. Barchanoid dunes in this sand Shan Desert. Number in center circle of rose is reduction factor.
sea are peaked, unlike those in the Cherchen Desert (fig. 101A). DP (drift potential, in vector units) is shown for each rose. (Fig.
Arrow on mosaic, and those on roses indicate resultant drift 103.)

The distribution in occurrence of maximum winds wide unimodal distribution of effective winds and
follows closely the pattern of effective winds of all an intermediate drift potential (fig. 107B). The
velocities suggested by the annual sand rose (fig. northerly group (mode) of effective winds is here
lOO). At this locality the resultant drift direction neglected; it has 12 percent of the total DP at the
varies little throughout the year (fig. 105B). station and occurs during the period of weakest
winds.
Western Simpson Desert near Oodnadatta, Australia
Instead of two distinct and relatively persistent
Simple linear dunes are an exclusive type components of the wind regime as at Upington,
throughout much of the Simpson, Great Victoria, several components exist at Oodnadatta. These
and Great Sandy Deserts of Australia. The driest components change in relative intensity during the
and windiest region of the Simpson Desert may be year, resulting in the wide unimodal annual dis-
characterized by a markedly different wind regime tribution dominated by effective winds from the
from that of the southwestern Kalahari Desert just south. The systematic shift of effective winds from
discussed, although the dunes in both regions are of roughly southerly to southeasterly in January, to
the same type (fig. 107 A). The wind regime at Ood- southwesterly in July, and back during the rest of
nadatta, Australia (located approximately 110 km the year results in a gradual swing of the resultant
(68 mil from the dunes shown in fig. 107 A}, has a drift direction. It moves from a northwestward-

Digitized by Google
158 A STUDY OF GLOBAL SAND SEAS

TABLE 18. - Localities used to evaluate the wind environments of linear dunes
IDP. drift potential; RDP. resultant drift potential)
Wlndresime
Neueat repr.entatlve Station c_iflcatlon.
Dune coordinates Itation coordinates Other usill8
Dune fonn(s) (lat. 10118)' (distance from dunes In lun) (lat. Ions) data' DP RDPIDP sandroee
Simple linear ............... 24"37' S. Mariental. South-Weat Africa' 2437'S. Yea so 0.32 Complex.
18"10' E. (25). 1758'E.
25"14'S. Gilea. Australia' (25) ........ . 25"02'S. No 101 .23 00.
128"13'E. 128"18'E.
2733' S.. OodnadaHa. Australia' (110) .. 2733' S. No 388 .48 Wide unimodal.
135" 27' E. 13527' E.
Same as station Ekidze. U.S.S.R.' (0) ......... 41"02' N. No 111 .19 Complex.
5748' E.
Near station .. Sabhi. Libya' (5) ............ 27"01' N. Yea 268 .20 00.
14"26'E.
Compound and simple linear. Near station .. Bilma. Nigel" (5) . . . . . . . . . . . .. 1841' N. No 948 .74 Wide unimodal.
1255'E.
22"30' N. Fort-Gouraud. Mauritania' 2241'N . No 681 .ff! Acute bimodal.
1216'W. (40- SO). 1242'W.
20"50' N. Atar. Mauritania (40) ....... . 20"31' N. No 443 .64 Wide unimodal.
133O'W. 13"04'W.
1930' N.. Akjoujt. Mauritania (20) . . . . .. 1945' N. No 352 .68 00.
1410' W. 14"22' W.
Complex linear. . . . . . . . . . . . .. Same as station Narabeb. South-Weat Africa (in 23"52' S. Yea 45 .24 Unclassified.s
interdune corridor) (0). 1452' E.
Near station .. EI Colea. Algeria (0) ......... 30"30' N. No 270 .26 Complex.
02"52' E.
Simple linear; also linear "sand 2820' S.. Upington, South Africa' 2824' S.. No 520 .68 Obtuse bimodal.
sheets." 2126' E. (10 - 25). 21 16' E.
Compound linear; also linear Near station .. Ash Shiiriqah (Sharjah). Arabia 25"21' N.. No 280 .60 Acute bimodal.
"sand sheets." _________ (5- 10)_._____________ ~~23' E. ______________
, to-km-radius circle centered on this point.
"Yes"lndicatea airphoto or other data used to confinn identification of dune type.
I Discussed 88 example in text.

Coordinates approximate.
Annual wind summary not available.

3000

2500

~
z 2000
:l

ga:: 1500

~
1000
DRIFT POTENTIAL (DP) VERSUS RDPIDP for 13 stations near ~
linear dunes visible on Landsat imagery. Linear dunes repre- .J
~ 800
sented by triangles. Data from figure 98 also shown, barchanoid !z

~
dunes represented by dots. Dot within triangle denotes a 600
locality near which both dune types are well developed. The
distribution of points indicates that wind environments of linear Ii;
a::
400

dunes have greater directional variability than do wind environ-
ments of barchanoid dunes for a particular drift potential. Linear
c
200

dunes in wind environments with high drift potentials are
associated with less directional variability of effective winds

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
than are linear dunes in wind environments with low drift po-
tentials. (Fig. 104.) RDPIDP

Digitized by Google
 DUNE FORMS AND WIND REGIME 159

DP=520 DP=28 DP=29 DP=40 DP=8Il DP=45 DP=48

ANNUAL FEB. APR. JUNE AUG. OCT. DEC.

LANDSAT FALSE-COLOR IMAGERY (A) of part of the south- the author. These streaks were extended sandy areas with
western Kalahari Desert near Upington, South Africa. Red sand barchanoid dunes.) Green vegetation along the Orange River
is yellow on Landsat false-color imagery. Linear dunes with shows as red on Landsat false-color imagery. Clouds are white.
sharp crests are visible near Upington and farther east near the Arrow indicates annual resultant drift direction (ROD) at
Orange River (upper right comer of image). Wider linear streaks Upington. 8, Annual and six monthly sand roses for Upington,
which lie across drainages and hills are marked by lines parallel South Africa, illustrating an obtuse bimodal wind regime. The
to streaks. Although linear in gross aspect, they may contain resultant drift direction (arrows) remains steady and approx-
transverse bedforms too small to be seen on Landsat imagery. imately parallel to the orientation of linear dunes in the region
(Similar streaks are visible on Landsat imagery of northern throughout the year. Number shown in center circle of rose is
Arizona north of the Little Colorado River near Cameron, reduction factor. DP (drift potential, in vector units) is given for
Arizona, in the United States and were examined in the field by each rose (Fig. 105.)

Digitized by Google
160 A STUDY OF GLOBAL SAND SEAS

tJ Jan. Feb. Mar. Apr. Mav June JulV Aug. Sept. Oct. Nov. Dec.

Direction from which

NNW. NW. N. WSW. WSW. N. N. N. NNE . WSW. NNE. W.
maximum winds came

Speed (knots) 33 40 35 31 32 29 36 37 52 35 33 37

WIND ROSE FOR MAXIMUM WINDS and table of average max- Upington closely reflect the obtuse bimodal distribution of all
imum monthly winds at Upington, South Africa, for 1956- 67. effective winds shown in the annual sand rose in figure 105B.
Each tick mark on a wind rose arm represents a maximum gust (Fig. 106.)
from the direction indicated by the arm. Strongest winds at

LANDSAT IMAGERY OF SOUTHWESTERN SIMPSON DESERT, displayed by annual sand rose results from interaction of winds
Australia (A). Straight simple linear dunes are the only type visi- between the southwest and southeast during the year. Arrows
ble on Landsat imagery throughout much of this desert and indicate resultant drift direction. Number in center circle of rose
along its margins. B, Annual and twelve monthly sand roses for is reduction factor. DP (drift potential, in vector units) is given
Oodnadatta, Australia, approximately 110 km (68 mi) from the for each rose. (Fig. 107.)
dunes in A. Wide unimodal wind distribution of effective winds

Digitized by Google
 DUNE FORMS AND WIND REGIME 161

DP=43 DP=40 DP=27 DP-17 >Po=13 >Po=10

JAN. FEB. MAR. APR. MAY JUNE

DP=388
ANNUAL

B DP=46 Dp..43
DP-18 DP=26 DP=52 DP=56
JULY AUG. SEPT. OCT. NOV. DEC.

directed ROD in January, to an east-northeastward- plex pattern (fig. 109). The trend of the dunes near
dire~ted ROD in July, and back. Another region Sabha: is exactly parallel to the annual resultant
characterized by a regular swing of resultant drift drift direction at the station (fig. 108B).
direction back and forth is western Mauritania The parallelism of dune trend and effective wind
(chapter K). This region is also characterized by direction at Sabha suggests that each component of
linear dunes, although both wind regimes and a complex wind regime may have a proportionate
dunes differ in detail from those near Oodnadatta. effect on the growth of dunes and their resulting
orientation. On the other hand, the orientation of
lamlat Zallif sand sea near Sabhi, Libya, and
Pesld Karakumy, U. S. S. R. the dunes is also parallel to the resultant drift direc-
tion during May, the month of highest drift poten-
Some simple linear dunes occur in regions of tialat the station. Therefore, it seems equally possi-
complex wind distributions of low or intermediate ble that most of the growth of the dunes occurs
energy. Long, relatively narrow linear dunes, from May until September, the time of relatively
which are very gently curving on Landsat imagery steady easterly winds. Complex monthly wind
(fig. l08A), occur west of Sabhi, Libya, in a com- regimes during the remainder of the year might be
plex intermediate-energy wind regime (fig. 108B). incapable of obliterating the linear form of the
These dunes, in fact, have a zigzag pattern in plan dunes established during the May to September
when viewed from the ground or on aerial photo- season.
graphs (McKee and Tibbitts, 1964, p. 14-15, plates An example of linear dunes, probably active, in a
I, II). Wind regimes from month to month are com- complex wind regime of low energy occurs in the
plex, with the exception of May through August northern Peski Karakumy in the region of Ekidze,
when easterly effective winds are most important. U.S.S.R. (fig. 110A). Near Ekidze, features that
The occurrence of maximum winds suggests a com- seem to be linear dunes - because they exhibit the

Digitized by Google
162 A STUDY OF GLOBAL SAND SEAS

LANDSAT IMAGERY OF THE RAMLAT ZALLAF SAND SEA near

Sabha, libya (A). linear dunes are gently curving, thin streaks
extending from west-northwest to east-southeast across the im-
age. B, Annual and twelve monthly sand roses for Sabha, libya,
illustrating the complex pattern of effective surface winds.
Dunes shown in A are alined with the annual resultant drift
direction at the station and with the monthly resultant drift
direction during May, when the highest drift potentials occur.
Winds are relatively steady from May through August, during
which about 35 percent of the average annual drift potential oc-
curs. Number inside center circle is reduction factor. Arrows are
resultant drift direction (ROO). DP (drift potential, in vector
units) is given for each rose. (Fig. 108.)

f
DP=17 DP=21 DP=38 DP=30 DP=54 DP=26
~~~_ _ JAN. FEB. MAR. APR. MAY JUNE

DP=268
ANNUAL

B
DP=8 DP=7 DP=11 DP=24 DP=20 DP=13
JULY AUG. SEPT. OCT. NOV. OEC.

N Jan. Feb. Mar. Apr. June Aug. Sept. Oct. Nov. Dec.
Mav JulV

~
t Direction from which
maximum winds came
NNW. NNW. SW. ESE. SE . S. W. NE. WNW. S. SSW. SE.

Speed (knots) 24 24 27 24 30 22 25 19 26 31 24 23

WIND ROSE FOR MAXIMUM WINDS and table of average maximum monthly winds at Sabha, libya, for 1949 and 1961 -63. Each
tick mark on a wind rose arm represents a maximum gust from the direction indicated by the arm. Strongest winds at Sabhi reflect
the complex distribution of all effective winds shown on the annual sand rose in figure 1088. (Fig. 109.)

Digitized by Google
 DUNE FORMS AND WIND REGIME 163

the distribution of maximum winds (fig. 113), this

sand rose suggests a wide unimodal or perhaps
acute bimodal distribution of effective winds. The
wind regime at Giles is complex but composed pri-
marily of wide modes from the north-northeast and
south.
The degree to which the observed morphologies
are a result of present day winds is difficult to
determine from available data. For example,
vegetation on dunes in the Sahel and southwestern
N
Mauritania seems to be associated with climatic

t DP=lll
RDP= 21
RDPIDP=O.19
change which has led to the immobilization of the
dunes (chapter K), although not all the observed
dunes in these regions are out of adjustment, or
alinement, with present-day winds.
In addition to vegetation, which tends to reduce
sand movement, the wind regimes at Mariental and
B Giles have low drift potentials. For example, the
lANDSAT IMAGERY OF PRESUMED LINEAR DUNES (A) in the drift potential of 50 VU at Mariental indicates
Peski Karakumy near Ekidze, U.S.S.R. linear dunes have sharp negligible sand drift if the impact threshold shear
crests and junctures which open upwind (against the resultant velocity (v;') is assumed to be 19 cmts (0.58 ftls) in
drift direction, ROO, indicated by arrow). B, Annual sand rose for
Ekidze, illustrating a complex wind regime dominated by a
figure 94. On the other hand, dunes near Mariental
mode from the east and a mode from the west. linear dunes and possibly near Giles are only partially vegetated,
near Ekidze are alined with the annual ROO (arrow) at the sta- indicating that they may be at least partly active.
tion but exhibit more waviness in plan than do linear dunes
The compatibility of the wind regimes with dune
near Upington, South Africa, and Oodnadatta, Australia, which
types at Giles and Mariental, and the continuity of
have higher energy, bimodal effective wind distributions. DP,
dune trends from partially vegetated to unvege-
drift potential; RDP, resultant drift potential; both in vector units.
(Fig. 110.) tated areas across considerable distances in both the
Kalahari and Simpson Deserts (chapter K) indicate
windward-facing tuning-fork junctures charac- that the dune morphologies are at least partially the
teristic of linear dunes - are perfectly alined with result of present-day winds. In addition, the degree
the resultant drift direction at Ekidze (arrows, figs. of waviness in dune plan in the unvegetated and
110A,B). The complex wind regime at the station is the vegetated regions of the Kalahari Desert and in
dominated by two opposing modes of effective the Australian deserts corresponds roughly to the
wind from roughly the west and the east. directional variability of the wind regimes (com-
pare dunes near Giles and Ekidze with those near
Northwestern Kalahari Desert, South-West Africa, Oodnadatta and Upington).
and Great Victoria Desert, Australia
Simple linear dunes are partially vegetated in Even if present-day low-energy wind regimes
wide regions of the Kalahari, Great Victoria, and have not created the dunes near Mariental and
Great Sandy Deserts. Near Mariental. South-West Giles, they may still serve to maintain the form of
Africa (fig. 111A), and Giles, Australia (fig. 112A), the dunes or have insufficient strength to obliterate
partially vegetated linear dunes are associated with them. More specific knowledge of the degree to
low-energy wind regimes (figs. 111B, 112B) com- which the dunes observed near Mariental and Giles
patible with the observed dune types. are the product of present-day winds would facili-
The sand rose for Mariental reflects a high degree tate interpretation of the paleoclimate of these
of observer or procedure bias. Nevertheless, as does regions.

Digitized by Google
164 A STUDY OF GLOBAL SAND SEAS

1~'

N
DP=50

t RDP=lS
RDPIDP-O.30
N

t
DP=101

LANDSAT IMAGERY OF THE NORTHWESTERN KALAHARI

DESERT (A) near Mariental, South-West Africa. Partially vege-
tated, straight linear dunes with junctures opening upwind, or B
against the resultant drift direction. Resultant drift direction at LANDSAT IMAGERY OF LINEAR DUNES IN THE GREAT VIC-
Mariental indicated by arrow. 8, Annual sand rose for Mariental TORIA DESERT (A) near Giles, Australia. These dunes are more
shows extreme observer or procedure bias (few observations wavy in plan than those near Mariental, South-West Africa (fig.
between the prime points of the compass) but suggests wide 111A). 8, Annual sand rose for Giles illustrates not only the
unimodal, or perhaps acute bimodal, wind distribution. OP, drift complex nature of effective winds but also the difficulties of
potential; ROP, resultant drift potential; both in vector units. Ar- classifying directional characteristics of effective wind regimes.
row indicates resultant drift direction (ROO). (Fig. 111.) The subsidiary groups of the distribution (from approximately
the northwest and east-southeast, as bracketed) each comprise
more than 15 percent of the annual drift potential at the station.
Erg Hammaml Region, Mauritania, and Erg Bllma Most of the effective winds at the station, however, occur in
Region, Niger groups from the north-northwest and south. In this example,
because maximum wind data are unavailable, the classification
Both compound and simple linear dunes occur in may be either complex or obtuse bimodal. The sand rose illus-
regions characterized by acute bimodal. or wide trates well, however, the greater directional variability of effec-
unimodal wind distributions and a high drift poten- tive winds at Giles than at Mariental (fig. 1118). OP, drift poten-
tial. Much of Mauritania northwest of a line from tial, in vector units. Arrow indicates resultant drift direction
Akjoujt through Atar is characterized by acute (ROO). (Fig. 112.)
bimodal winds. and by compound and simple linear
dunes which extend subparallel to the resultant
which is strongest in midsummer (chapter K). The
drift direction of present-day winds (fig. 114).
group of winds from the southwest may represent
The long. straight. very large linear dunes near
the effect of anticyclonic circulation from a winter-
Fort-Gouraud. Mauritania (fig. 114). seem to have
time high-pressure ridge over the northern Sahara.
developed in an acute bimodal high-energy wind
regime in which the two major components are Simple linear dunes and compound linear dunes.
mainly in different seasons (fig. 97B). The more similar in appearance to those near Fort-Gouraud
dominant group of winds. from north through occur near Bilma. Niger (fig. 115A). in a region
northeast. is active from spring through fall. This characterized by a wide unimodal high -energy
group results from anticyclonic tradewind circula- wind regime (fig. 115B). In this region. simple
tion from the North Atlantic high-pressure cell. linear dunes and compound linear dunes occur in

Digitized by Google
 DUNE FORMS AND WIND REGIME 165
N

t Direction from which

maximum winds came
Jan.

NW.
Feb.

NE.
Mar.

N.
Apr.

SE.
May

NW.
June

N.
July

NE.
Aug.

N.
Sept.

N.
Oct.

N.
Nov.

NW.
Dec:.

E.

Speed (knots) 25 35 20 25 23 23 23 25 30 30 25 25

WIND ROSE FOR MAXIMUM WINDS and table of average monthly maximum winds at Mariental, South-West Africa, for 1%0-67.
Each tick mark on a wind rose arm represents a maximum gust from the direction indicated by the arm. Strongest winds at Mariental
support the classification of effective winds in figure 1118 as either wide unimodal or acute bimodal. (Fig. 113.)

with complex wind regimes (table 19). Drift poten-

tials range from low (73 VU at lllizi, Algeria) to
high (658 VU at Ghudiimis, Libya), and RDPIDPs
range from intermediate to low. Isolated star dunes
are associated with RDPIDP of less than 0.20, and
star dunes in chains or atop linear features are
associated with higher RDPIDP (fig. 116).

Examples of Wind Environments of Star

Dunes
Eastern G rand Erg Oriental, Algeria-Libya

Isolated star dunes occur in the Grand Erg Orien-

tal near Ghudiimis, Libya (fig. 117) in a high-energy
complex wind regime (fig. 9'7q. This association
supports the suggestion of McKee (1966, p. 68) that
star dunes develop as a result of effective winds
from several directions.
The complex wind regime at Ghudamis differs
from the complex wind regime associated with
LANDSAT IMAGERY OF LINEAR DUNES in the Erg Hammami linear sand dunes at Sabhii (fig. 108B) in several sig-
region, northern Mauritania. Long, large compound linear dunes
in this region have developed in an acute bimodal high-energy
nificant ways. First, the drift potential at Ghudamis
wind regime (fig. 978). (Fig. 114.) is 2.5 times as great as at Sabha; second, complex
wind distributions prevail during the months of
two different areas. This evidence indicates that highest drift potential at Ghudiimis and become
both simple linear and compound linear dunes can relatively narrow only during the seasons of low
develop in high-energy wind regimes. Perhaps drift potential from December through February. At
available sand supply and age of the dunes, as well Sabha, the reverse situation occurs. Finally,
as wind regime, may be important factors that although two dominant modes may be recognized
determine whether simple linear or compound in the annual sand roses for both Ghudamis and
linear dunes will form at a given locality. Sabha, the dominant modes are perfectly opposed
at Ghudamis but are at approximately a right angle
Star Dunes at Sahba.
Summary of Observed Wind Environments of Grand Erg Occidental, Algeria
Star Dunes
Some star dunes occur in low-energy complex
Star dunes observed in this study are in areas wind regimes. Star dunes occur in such an environ-

Digitized by Google
166 A STUDY OF GLOBAL SAND SEAS

ment along and atop an escarpment near Beni composed of sinuous ridges rather than having
Abbes, Algeria (fig. 118A). The drift potential at formed isolated mounds, 88 at Ghudiimis.
Beni Abbes is low (185 VU), and effective winds
Western Grand Erg Oriental, Algeria
during the month of highest drift potential are com-
plex, as at Ghudamis, Libya (fig. 118B). Star dunes Star dunes have been observed on Landsat imag-
near Beni Abbes, visible in an aerial photograph ery atop other sand features, many of which have a
(Alimen and others, 1957, plate III), seem to be linear aspect in plan view. These linear features on

B DP=948 DP=96 DP=119 DP=71 DP=25 DP=89 DP=55

ANNUAL FEB. APR. JUNE AUG. OCT. DEC.

LANDSAT IMAGERY OF LINEAR DUNES (A) in the Erg Bilma. Both simple and compound linear dunes occur near Bilma in the
same regional wind environment. B, Annual and six monthly sand roses for Bilma, Niger, illustrate a wide unimodal high-energy
wind regime near the dunes. Arrow indicates resultant drift direction (ROO). Number in center circle of rose is reduction factor. DP
(drift potential, in vector units) is given for each rose. (Fig. 115.)

Digitized by Google
 DUNE FORMS AND WIND REGIME 167
TABLE 19. - Localities used to evaluate the wind environments of star dunes
lOP. drift potential; HOP. resultant drift potential)
Wind regime
Nearest representative Station cl8llification.
Dune coordinates atation coordinates Other using
Dune fonu(a) (tat. long)' (diatance from dunes in km) (lat. long) data' OP HOPI OP sand rose
Staf .... . .. .... ........ . .... Near station .. Ghudamis. Ubya' (::::5) ....... 300 08'N . Yes 658 0.09 Complex.
0930' E.
Do. . . .. .. Beni Abbes. Algeria' (0) ...... 300 08'N . Yes 184 .11 Do.
0210'W.
Do. . ..... Illizi. Algeria (0) .... . .. . . . ... 2630' N. No 73 .07 Do.
0829' E.
2622'S.. Aus. South-West Africa (23) 2642'S . .. No (S) (S) Wind rose only.
1614' E. 1614' E.-
Staf dunes in chains Of on 3134' N. Hassi Messaoud. Algeria' (20) . 3140' N . No 200 .31 Complex.
lineaf "sheets." 06"15' E. 06"09' E.
Staf dunes atop lineaf Of Near station . . Bordj Omaf Driss (Fort 28"06' N . No 112 .35 Do.
barchanoid features. Flatters) . Algeria (0) . 0650' E.
10-km-radiua circle centered on thia point.
, "Yes" indicates airphoto or other data used to confirm identification of
dune types.
Discussed as example in text.
Station coordinates approximate. which the star dunes have developed may. in some
Drift potential unknown. places. represent older barchanoid or linear dunes.
or extended linear "sand sheets." Star dunes near
Hassi Messaoud. Algeria. have developed in chains
atop extended linear sand features of indeterminate
identity (fig. 119A). The annual effective wind
regime at Hassi Messaoud is difficult to classify but
2500 is either obtuse bimodal or complex. Effective wind
CIl

~2000
::::I

~1500
~ 1000
~
J
~ 800
!Z
~ 600*

~ High

.
400
~ Hassi Messaoud. Intermediate
200

00
*
* *'
0.2
;/.

0.3
Bordj Omar Oriss

0.4 0.5

...

0.8
--
Low

0.9 1.0
0.1
ROPIOP

DRIFT POTENTIAL (~versus RDPlDPfor five stations near star

dunes, which are represented by stars. Data from figures 98 and
104 also shown; barchanoid dunes represented by dots, linear
dunes represented by triangles. Dot within triangle indicates
locality near which both barchanoid and linear types are well
developed, Isolated star dunes were observed only in wind en-
vironments with low RDPIDP. Star dunes on broadly linear sand
"sheets," or atop other linear features were observed in inter- LANDSAT IMAGERY OF ISOLATED STAR DUNES in the Grand
mediate environments in terms of RDPIDP. Examples in Algeria Erg Oriental near Ghudamis, libya. Annual and six monthly
are dunes near Hassi Messaoud and Bord; Omar Driss (Fort Flat- sand roses for Ghudamis illustrating complex high-energy wind
ters) . (Fig. 116.) regime are shown in figure 97C. (Fig. 117.)

Digitized by Google
168 A STUDY OF GLOBAL SAND SEAS

LANDSAT IMAGERY OF STAR DUNES (A) that occur along and

atop an escarpment at the southern margin of the Grand Erg Oc-
cidental near Beni Abbes, Algeria. These star dunes are not as
large as those near Ghudamis, libya. B, Annual and six monthly
sand roses for Beni Abbes. Annual effective wind regime is com-
plex, low energy. Effective wind regime is complex during April,
the month of highest drift potential. Arrows indicate resultant
drift direction (ROO). Number in center circle of rose is reduc-
tion factor. DP (drift potential, in vector units) is given for each
rose. (Fig. 118.)

!k--
B
DP=185
ANNUAl
~
DP=24
FEB.
DP=20
APR.
)f-
DP=13
JUNE
"Q-
DP=5
AUG.
~
DP=7
OCT.
+ DP=7
DEC.

LANDSAT IMAGERY OF AliNED STAR DUNES (A) atop linear

sand sheets near Hassi Messaoud, Grand Erg Oriental, Algeria.
Exact nature of the sheets is unknown. B, Annual and six
monthly sand roses for Hassi Messaoud, Algeria, illustrating
complex (or possibly obtuse bimodal) intermediate-energy
effective wind regime at the station. The present-day wind en-
vironment of the region seems intermediate between those of
star and linear dunes (figs. 104, 116). The dune pattern may
have developed from present-day winds or may represent the
result of two or more stages of development involving different
wind regimes. Arrows indicate resultant drift direction (ROO), in
vector units. Number in center circle of rose is reduction factor.
DP (drift potential, in vector units) is given for each rose.
(Fig. 119.)

DP=200 DP=23 DP=44 DP=20 DP=7 DP=12 DP=32

ANNUAL FEB. APR. JUNE AUG. OCT. DEC.

Digitized by Google
 DUNE FORMS AND WIND REGIME 169

distributions from month to month tend to be bimo- ters). Algeria. There. star dunes are atop what seem
dal or complex (fig. 119B). This wind regime differs from Landsat imagery (chapter K) to be barchanoid
from other wind regimes associated with star dunes dunes. If the sand bodies upon which the star dunes
because the resultant drift direction at the station is have developed were originally barchanoid dunes.
relatively steady (toward the southeast) during the a climatic change is indicated because the observed
year. This wind environment seems to be inter- environments of star and barchanoid dune types
mediate between that associated with linear dunes differ widely (fig. 116). Present-day winds at Bordj
and that with star dunes. in terms of RDPIDP (fig. Omar Driss are intermediate between those form-
116). ing linear dunes and those forming star dunes. in
An anomaly is presented by dunes in the region terms of RDPIDP.
of Erg Irarrarene near Bordj Omar Driss (Fort Flat-

Digitized by Google
Digitized by Google
 c. -. ... .~

. "'
l"'~... ":'~-:-.~~~'~ :~,
... .~ .... ~

. ,,~Ir~~:: : .. ~:.: ~:~\~;,'~ \. .

-_...

A STUDY OF GLOBAL SAND SEAS

LABORATORY STUDIES OF SAND

PATTERNS RESULTING FROM
CURRENT MOVEMENTS
CbapterG

By THEODORE F. TYLER

Contents

Summary of conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . ..
...
173 Present study - Continued
Introduction ...................................... 173 Wave-tank experiments - Continued
Background information ............................ 173 Experiment 1 ................................ 177
Hierarchy of bed roughness .. . . . . . . . . . . . . . . . . . .. 173 Experiments 2 and 3 ......................... 180
Oscillation ripple .............................. 173 Ripple development patterns . . . . . . . . . . . . . . . . . . 182
Present study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 174 Comparison of aqueous wave ripples (aqueous dunes)
Equipment and procedure ...................... 174 and eolian dunes ............................ 182
Wave-tank experiments. . . . . . . . . . . . . . . . . . . . . . . .. 177 Effects of barriers on ripple development ......... 183

171

Digitized by Google
172

Illustrations
FIGURE 120. Sketch of generalized ripple profile, showing gently sloping stoss side and
steeply sloping lee side ........................................... . 174
121. Sketches of bedform patterns of large-scale ripples (aqueous dunes) near the
Bahama Islands ................................................. . 175
122. Generalized diagram of wave tank used in laboratory experiments ......... . 176
123. Graph showing starting sand profile in wave tank, experiments 1, 2, and 3 .. 176
124. Sketch showing plan view of ripple patterns for experiment 1 after 10,30, and
50 minutes of wave action ....................................... . 178
125. Sketch of plan view of ripple crests, showing process of ripple branching .. . 178
126. Graphs showing ripple height and wavelength after 10 and 50 minutes of wave
action .......................................................... . 179
127. Graphs showing stoss- and lee-side horizontal dimensions after 10 and 50
minutes of wave action .......................................... . 179
128-131. Sketches showing:
128. Plan view of ripple patterns for experiment 2 after 10 and 50 minutes of
wave action ............................................... . 180
129. Plan view of ripple patterns for experiment 3 after 10 and 50 minutes of
wave action ............................................... . 181
130. Sequence of small-scale ripples changing with time and increase in flow
velocity and (or) water depth ................................ . 182
128 -131. Sketches - Continued
131. Sequence of small-scale ripples changing with increase in flow velocity
and (or) water depth ....................................... . 182
132. Photographs, Landsat imagery, and sketches comparing plan views of aqueous
and eolian bedforms ............................................. . 184
133. Sequential photographs showing the effects of a barrier on ripple development
134. Sequential photographs showing bedforms developed in a wind-tunnel experi-
ment ........................................................... . 184
135. Landsat imagery of questionable barchan dunes developing into barchanoid
ridges, Arabian Peninsula ......................................... . 185
136. Landsat imagery of development of dunes across a geomorphically controlled
area, Arabian Peninsula ........................................... . 185

Tables
TABLE 20. Ripple dimensions and ripple symmetry index for experiment 1 after 10,30 and
50 minutes ...................................................... 177
21. Ripple dimensions and ripple symmetry index for experiment 2 after 10 and 50
minutes ......................................................... 181
22. Ripple dimensions and ripple symmetry index for experiment 3 after 10 and 50
minutes ......................................................... 182
23. Bedforms oriented transverse to wave or wind direction .................. 183

Digitized by Google
 LABORATORY STUDIFS OF SAND PATTERNS RFSULTING FROM CURRENT MOVEMENTS 173

Summary of Conclusions tank and procedures used for this study, (3) ripple
pattems developed during three laboratory experi-
1Ir HIS STUDY IS A COMPARISON of laborato- ments and a comparison of aqueous dunes with
\UI"ry-generated small-scale aqueous ripples, eolian dune development, and (4) effects of barriers
natural large-scale ripples (aqueous dunes), and on bedform development in both aqueous and
eolian dunes to show similarities in their develop- eolian environments as evidenced by use of photo-
mental sequences and resulting bedform pattems. graphs of laboratory experiments, oblique aerial
Small-scale aqueous ripple development on smooth photographs, and Landsat (ERTS) imagery.
sloping sand beds and ripple development in the lee
of barriers placed on smooth-leveled sand beds
were studied in a laboratory wave channel by Background Information
means of tracings and still photography. These trac-
ings and photographs were compared with oblique BEDFORMS OCCUR on the surface of non-
aerial photographs and Landsat (ERTS) imagery to cohesive sediments in response to fluid motion. It is
determine similarities between aqueous and eolian extremely difficult to determine the various
bedform pattems. The various bedforms show parameters and mechanisms and the intricate inter-
remarkable similarities although developed under relations which control bedform development in a
different environmental conditions. natural situation. Under laboratory conditions one
parameter at a time may be varied to study its effect
Bedforms can be categorized on the basis of their on bedform development and to provide insight
developmental sequences, as follows: (1) Isolated into this parameter's effect on bedform pattems ob-
crescentic forms (isolated barchan), (2) coalescing
served in the field.
crescentic form (coalescing barchan), (3) parallel-
sinuous form (barchanoid ridge), and (4) parallel- Hierarchy of Bed Roughness
straight form (transverse ridge).
A critical, or threshold, velocity, which is a func-
Introduction tion of grain size, must be reached to initiate motion
in noncohesive sands. Before onset of grain motion
THIS STUDY WAS UNDERTAKEN in the U.S. on a flat bed, only grains themselves contribute to
Geological Survey sedimentary structures laborato- bed roughness. A second-order, and possibly a
ry to determine the transition pattems of small- third-order, of bed roughness may develop once
scale oscillation ripples on gently sloping surfaces, grain movement starts. This hierarchy of bed
and on level surfaces when barriers were present to roughness is: (1) individual grains, (2) small-scale
disrupt the normal wave motion. Tracings were ripples, and (3) dunes. In general, the individual
made as the ripples developed under normal wave grain is the basic building block for this hierarchy
action, and a tracing was made every 10 minutes for of bed roughness, and for the development of a
a 50- to 6O-minute period. Then the ripples were sedimentary structure. Bedform development is
classified according to crest shape. Comparisons also dependent on other factors, such as the
were made between the bedforms of these small- availability of sand, the water or wind velocity,
scale aqueous ripples and those of the large-scale water depth, and time.
ripples (aqueous dunes) developed under natural
conditions. The aqueous forms were then com- Oscillation Ripple
pared with large-scale eolian dunes, by using aerial A ripple is defined as any deviation from a flat
photographs and Landsat (ERTS) imagery. bed that resembles or suggests a ripple of water and
Emphasis was placed on the similarities between that is formed on the bedding surface of a sedimen-
their plan views and transition pattems. tary deposit. Aqueous ripples vary considerably in
The following discussion falls into four catego- size and shape. Figure 120 shows a ripple profile at
ries: (1) hierarchy of bed roughness, (2) the wave right angles to wave propagation and gives the

Digitized by Google
174 A STUDY OF GLOBAL SAND SEAS

Large-scale ripples (aqueous dunes) are not as

,
DIRECTION OF WAVE PROPOGATION

r--
Trough
Stoss side
Crest
------il~-- Lee side --~
well described in the literature as are small-scale
ripples, but some good examples can be found. A
development of bedform patterns at the south-
eastern edge of the Tongue of the Ocean, Bahama
Islands, in water 3- 4 fathoms (5.5-7.3 m) deep is
shown by Newell and Rigby (1957, pI. 10, fig. 1).
Crescentic (barchanoid) forms occur as solitary
~-----Wavelength -----~~
forms which coalesce and eventually form parallel-
sinuous and parallel-straight forms (fig. 121A). The
GENERALIZED RIPPLE PROFILE, showing gently sloping stoss
horns of the simple crescentic forms are 50- 200
side and steeply sloping lee side. (Fig. 120.) feet (15- 61 m) wide, and they coalesce to form
great ripples (aqueous dunes) with wavelengths of
10- 30 feet (3- 9 m). Parallel-straight to parallel-
terms commonly used to describe a ripple. In sinuous transverse ripples (dunes) are developed in
general, a ripple whose crest becomes more planar the vicinity of the Northwest Channel Light,
than those shown in figure 120 is defined as having Bahama Islands (fig. 121B). These forms have
a platform crest. Two ratios are useful in charac- wavelengths of approximately 300 feet (91 m).
terizing a ripple:
1. Ripple symmetry index (RSI): Oscillation megaripples and certain eolian dunes
show similarities in their external profile and
Horizontal projection of stoss side developmental patterns when viewed in plan.
RSI= Thus, figure 120 and the associated terms can also
Horizontal projection of lee side
be used to describe an eolian dune oriented
For a symmetrical ripple formed under transverse to the wind direction. The lee- and stoss-
oscillatory flow conditions (wave ripple), this side dimensions and height of an eolian dune range
ratio should be equal to 1.0; for asymmetrical from a few feet to tens of feet in magnitude, very
wave ripples this ratio ranges from 1.1 to 3.8 similar to the megaripples (aqueous dunes) at the
(Reineck and Singh, 1973, p. 25). Tongue of the Ocean previously described.
2. Ripple index (Rl):
Present Study
RI = Wavelength A DESCRIPTION FOLLOWS of the wave tank
Height and of the procedures used for experiments con-
For both symmetrical and asymmetrical wave ducted in the laboratory. Ripple development on
ripples the ripple index ranges from 5 to in- gently sloping surfaces and the transition patterns
finity, with most wavelengths being 6- 8 of these ripples are compared for three experi-
(Dingler, 1975, p. 95; Reineck and Singh, 1973, ments.
p. 25). For unidirectional current ripples, the
RI is always more than 5. with most values Equipment and Procedure
falling between 8-15 (Reineck and Singh,
1973. p. 29). A 46-foot-long by 2-foot-deep by 1.5-foot-wide
Small-scale oscillation ripples are defined as hav- (14-m-long by 0.6-m-deep by 0.5-m-wide) wave
ing wavelengths ranging from 0.4 to 78.8 inches (1 tank was used for all experiments (fig. 122). Dis-
to 200 cm) and heights from 0.1 to 9.1 inches (0.3 to tances along the wave tank were measured from
23 cm) (Reineck and Singh. 1973. p. 24- 25). Small- the downwave end of the tank. The tank has a
scale unidirectional current ripples are defined as piston-type paddle at the upwave end and a silica
having wavelengths of 1.6 to 23.6 inches (4 to 60 sand beach with a median grain size of 0.319 mm at
cm) and heights of 0.1 to 2.4 inches (0.3 to 6 cm) the downwave end. The study area was the interval
(Reineck and Singh. 1973, p. 29). between 15 and 30 feet (4.6 and 9.1 m). This area in-

Digitized by Google
 LABORATORY STUDIES OF SAND PATTERNS RESULTING FROM CURRENT MOVEMENTS 175

BEDFORM PATTERNS OF LARGE-SCALE RIPPLES (aqueous dunes) in the Bahama Islands. Sketched from photographs in Newell
and Rigby (1957, pI. 10, fig. 1, and fig. 2) A, Southeastern edge of the Tongue of the Ocean. The water is 3 - 4 fathoms (5.5 - 7.3 m)
deep in the area. B, Vicinity of the Northwest Channel light. (Fig. 121.)

Digitized by Google
176 A STIJDY OF GLOBAL SAND SEAS

DIRECTION OF WAVE PROPOGATION

Back of tank

Water level

1 10

GENERALIZED DIAGRAM OF THE WAVE TANK used in the laboratory experiments. Dimensions of the tank are in feet (1
foot =0.31 metres). (Fig. 122.)

eluded the sand bed from the breaker-related bar ,

DIRECTION OF WAVE PROPOGATION

zone upwave to a point where the sand was about 1311.Upw-

0.4 inch (1.0 cm) thick. From 15 to 0 feet (4.6 to 0.0 o
~7.9
Water level
m) the sand-water interphase was affected by the
borelike motion of the water, which occurs after the ~
wave breaks, and the interval was therefore ex- ~ 3.
cluded from this study. I!! ~L--25~""~20:t::::::::::::=---1..L5---1..LO-----L5----.JO
The following parameters were held constant for LENGTH OF BED. IN FEET
the first three experiments: The grain size of the STARTING SAND PROFILE IN WAVE TANK, experiments 1, 2,
sand, D, was 0.319 mm; the water depth, h, at the and 3. 1 inch = 2.54 centimetres; 1 foot = 0.31 metres.
30-foot location (9.1 m) was 8.6 inches (21.8 cm); Measurements for study stop at Xs. (Fig. 123.)
and wave period, T, was 2.5 seconds. Differences in
these parameters, between experiments, reflected 5. Potassium permanganate dye was injected
the initial variations in the sand thickness (fig. 123). into the water to permit visualization of
Another series of experiments were conducted to the flow conditions.
study the effect barriers had on ripple development. 6. Magnetite grains were allowed to settle onto
The sand in the tank from the first experiments was the sediment at various times to indicate
leveled, and the barrier was placed on this leveled surface changes as a function of time.
sand surface. 7. The horizontal projections of the lee- and
The following procedures were used for the ex- stoss-side dimensions were measured
periments: from the ripple tracings.
A. Sloping sand bed without barriers: 8. The rate of ripple migration was calculated
1. The sand was leveled to a predetermined when possible by following the move-
profile. ment of specific ripples over a predeter-
2. Waves were generated for a total time of mined time interval.
50-60 minutes per experiment. 9. The ripple symmetry index (RSI) and the
3. Wave generation was stopped at the end of ripple index (RI) were calculated for
each 10-minute interval, and tracings groups of ripples with similar crest
were made of the bottom (plan view). shapes.
4. Ripple height was measured directly, using 10. Photographs of ripple development were
a metre stick. taken and used for comparative purposes.

Digitized by Google
 LABORATORY STUDIES OF SAND PATTERNS RESULTING FROM CURRENT MOVEMENTS 177

B. Flat sand bed with barriers: pIes (fig. 125, sets B, C, E, F). Those ripples which
1. The sand was leveled and a barrier was did not completely cross the tank (ripple crest 3 of
placed on it. set B) seemed to initiate this branching process.
2. Waves were generated until ripple develop- The segment of the ripple that separated from the
ment began to propagate downwave in a branching bedform migrated downwave and joined
uniform pattern. with the next ripple. This process could be repeated
3. Photographs of ripple development were with the newly branched bedform acting as the
taken and used for comparative purposes. nucleus.
The average ripple wavelength, height, stoss-
Wave-Tank Experiments and lee-side horizontal dimensions, and ripple sym-
To avoid repetition, ripple development is dis- metry index (RSI) were calculated for groups of
cussed in detail for experiment 1 and only briefly similar ripple patterns (table 20).
for experiments 2 and 3. lllustrations, tables, and After 30 minutes of wave action the bar zone was
graphs are used to examine ripple growth from a located at 15.3 feet (4.7 m) (fig. 124). Parallel-
gently sloping sand bed and their developmental straight to slightly parallel-sinuous ripples were
sequences. distinct crests developed throughout the wave tank
(fig. 124).
Experiment 1 One ripple in group E and one in group F were
Oscillatory waves were generated for 60 minutes followed for 20 minutes more (fig. 124) to study
during experiment 1. The initial sand profile is rates of ripple migration. The ripple in group E
shown in figure 123, and the bed in plan at given migrated 0.74 foot (0.23 m) and the ripple in group F
time intervals is shown in figure 124. The wave migrated 0.65 foot (0.20 m). This corresponds to
height, H, ranged from 5 inches (12.7 cm) at 17 feet rates of 0.43 inch (1.1 cm) per minute and 0.39 inch
(5.2 m) to 3.2 inches (8.1 cm) at 24 feet (7.3 m). (1.0 cm) per minute, respectively.
Assuming a horizontal bottom at 24 feet (7.3 m) and The bar was located at 15.6 feet (4.8 m) after 50
using a linear wave theory (Airy, 1845), the max- minutes of wave action. Parallel-sinuous ripples oc-
imum near-bottom fluid velocity, Urn' is calculated curred from 15.6 to 16.4 feet (4.8 to 5 m). Between
to be 10.6 in./s (27 cm/s), where: 16.4 and 17.3 feet (5 and 5.3 m), the ripples did not
extend across the full width of the tank. InI~tead,
Urn = 'If' HIT sin h(M), they terminated approximately 5.9 inches (15 cm)
k = 2 'If'/L is the wave number, and from the front side of the tank (fig. 124). These rip-
L, which is the wavelength, is a function of h and T. ples still had a slightly parallel-sinuous pattern
with platform crests and filled troughs. Crests 10-
Initial ripple development took place during the
first 5 minutes of wave action. Parallel-straight to
slightly parallel-sinuous ripples developed between TABLE 20. - Summary of ripple dimensions (in centimetres)
and ripple symmetry index for experiment 1 after 10, 30, and 50
15.5 and 16 feet (4.7 and 4.9 m) while nonuniformly minutes
spaced crescentic and coalescing crescentic ripples [Lettel'B designating ripple groupe correspond to those in fagure 124)
developed between 16 and 19.5 feet (4.9 and 5.9 m). Ripple Avenage Avenage Avenage
--------
Avenage
-~--- ------
Ripple
Mer 10 minutes (fig. 124) bar development was lo- group wavelength height .t088 lee symmetry
index (RSI)
cated at 15.2 feet (4.6 m), and branching ripples and
After 10 mlnut..
ripples that terminate in mid-tank formed between A ....... 8.8 1.5 5.0 3.8 1.3
15.2 and 18.5 feet (4.6 and 5.6 m). Parallel-sinuous B ....... 8.6 1.5 5.1 3.6 1.4
ripples developed from 18.5 to 20.0 feet (5.6 to 6.1 C ....... 6.7 .9 4.1 2.6 1.6
After mlnut..
m), and ripples which terminated in mid-tank D ....... 9.1 1.4 6.3 3.0 2.1
developed between 20 and 21 feet (6.1 and 6.4 m) E ....... 11.0 1.7 8.3 2.7 3.1
(fig. 124). F ....... 8.8 1.4 6.6 2.8 2.3
G ....... 9.9 1.5 7.0 2.8 2.5
Branching ripple development took place early in After III m1nut..
the experiment, when ripples either joined (fig. H ....... 10.9 1.7 7.4 3.5 2.2
125, sets A, D) or separated to form individual rip- I ........ 10.1 1.8 7.0 3.1 2.3

Digitized by Google
178 A STUDY OF GLOBAL SAND SEAS

DIRECTION OF WAVE PROPOGATION

)

III
GI
'5c
'E
o.....

...J
..
III
GI
:J
C
'E
)l
:..-'
,
~ M
0
a::
I)'
W ..

"
I- GI
~~
W
~
~
,\

III
GI
{
, I
J I (
'5c \ I I I I
'E \ \ I I I
0
It)

I I I I \
lii
.t:
< ( I I I {
I I I
23 22 21 20 19 18 17 16 15
FEET
I I I I I I I I I
7.0 6.7 6.4 6.1 5.8 5.5 5.2 4.9 4.6
METRES
PLAN VIEW OF RIPPLE PATTERNS for experiment 1 after 10, 30, and 50 minutes of wave action . Trough lines dashed. Slanting lines
connecting the same ripple demonstrated ripple migration (distance indicated on line). Letters A -I indicate groups of ripples listed
in table 20. (Fig. 124.)

cated between 17.3 and 19.0 feet (5.3 and 5.8 m)

were platform and had a parallel-straight to slightly
parallel-sinuous pattern. Parallel-straight ripples
with distinct crests developed from 19.0 to 20.0 feet
DIRECTION OF WAVE PROPOGATION
) (5.8 to 6.1 m). Crests from 20.0 to 21.5 feet (6.1 to 6.6
m) were platform. Between 21.7 and 23.5 feet (6.6
and 7.2 m), the ripples did not completely cross the
tank and, therefore, were not included in figure 124
beyond 20.2 feet (6.2 m).
A B c D E F Figures 126 and 127 illustrate the manner in
PROCESS OF RIPPLE BRANCHING. Plan view of ripple crests, which the ripple parameters change as a function of
sets A (earliest time) through F (latest time). (Fig. 125.) time. In figures 126A and B, the RI values of 5, 6,

Digitized by Google
 LABORATORY STUDIFS OF SAND PATTERNS RESULTING FROM CURRENT MOVEMENTS 179

~& __----~----~-----T----~~----~----'

8
2.0

7

1.5
1
POSfTION IN TANK
l
IB 1.0

ri 0.& t:::..._...!...J'--_--L_ _~_ __L.._ _...L.__~
FIont
X.hdI
A &

4

~ 4 1 8 10 12 14 11
!i 2&~---r---r---r---'I'7:RI::-_-::&-r-----' 3 POSITION IN TANK

~ RI-I Ftant

12.0
a::
/ RI-8

7
X.~.. / .
,lxX D
1.5
X

iQ OE---_L____ A
~___L____~___ L_ _ _ _~_ __ L_ _~
POSmoN IN TANK
FIont ~ 0 2 & 1 7 8
D Middle 110r---,----~,,~--~-.-.-----,r----,r--.

I'
It Ba

~----~I----~8----~10~----1L2----~1~4---B~M It
a

WAVELENGTH. IN CENTIMETRES 0 8 f.

GRAPHS SHOWING RIPPLE HEIGHT AND WAVELENGTH after

. .
It. It
(A) 10 minutes and (8) 50 minutes of wave action. Ripple index
Jines (RI = wavelength/height) plotted to reflect their relation to It

the 5 - 8 range of asymmetrical wave ripples. As shown in A, 74 1 l' a

percent of the ripples are in the 5 to 8 range, and in 8, 86 percent
are in the 5 to 8 range. Ripple height and wavelength measured
approximately 10 cm from front and back of wave tank. 1 cen- 5
timetre =0.394 inch. (Fig. 126.)
4

GRAPHS OF STOSS- AND LEE-SIDE horizontal dimensions after 3

(A) 10 minutes and (8) 50 minutes of wave action. Ripple sym- FIont
metry index lines (RSI = Horizontal projection of stoss a Middle
side/Horizontal projection of lee side) plotted to reflect their 2 Bedt
relation to the 1.1-3.8 range of asymmetrical wave ripples. As
shown in A, 80 percent of the ripples are in the 1.1 -3.8 range,
and in 8, 97 percent are in the 1.1 -3.8 range. Stoss- and lee-side
measurements were made approximately 10 cm from the front B
and back of the tank; measurements were made in the middle
of the tank when they could not be made along the front and 2 3 4 5 8
back. 1 centimetre =0.394 inch. (Fig. 127.) HORIZONTAL LEE-5IDE DIMENSIONS. IN CENTIMETRES

Digitized by Google
180 A STUDY OF GLOBAL SAND SEAS

and 8 are indicated by solid lines. After 10 minutes, In experiment 2, irregular crescentic ripples
74 percent of the ripples fell within the 5-8 range, developed throughout the wave tank after approx-
and after 50 minutes 86 percent fell within that imately 3 minutes with the best development oc-
range. Lines of constant RSI (RSI =1.1, 3.8) are curring at about 23 feet (7 m). Parallel-straight to
drawn in fIgures 127A and B to correspond to the slightly parallel-sinuous ripples with a uniform pat-
reported RSI range for asymmetrical wave ripples. tern developed after 10 minutes of wave action (fig.
After 10 and 50 minutes, 80 and 97 percent of the 128; table 21).
ripples, respectively, fell within this range. Ripples In experiment 3, irregular crescentic ripples
which plot outside the indicated ranges fall into one developed after 2 minutes of continuous wave ac-
of four categories: (1) Ripples that are influenced tion. Isolated crescentic ripples coalesced to four
by their position near the bar zone, (2) branching large crescentic ripples after 5 minutes Oat approx-
ripples. (3) washed-out ripples. and (4) ripples at imately 23 feet (7 m). Parallel-straight. branching
the upwave end of the sand bed. The ripples in bedforms. parallel-sinuous. and irregular crescentic
category 4 are not considered to be fully developed. ripples developed after 10 minutes of wave action
(fig. 129; table 22). .
Experiments 2 and 3 After 50 minutes, ripple development was less
The starting sand profiles for experiments 2 and 3 distinct in experiments 2 and 3 (figS. 128. 129; tables
are shown in figure 125. For experiment 2, wave 21.22). The ripples were washed out between ap-
break was located at 17 feet (5.2 m). the wave proximately 17 and 19 feet (5.2 and 5.8 m) with
height. H, was 4.1 inches (10.4 cm) at 19 feet (5.8 m) crests becoming more platform and troughs filling.
and 3.7 inches (9.4 cm) at 23 feet (7 m) (fig. 128). causing the bottom profile to become more planar.
For experiment 3. wave break was located at 17 feet The RI values calculated for experiments 2 and 3
(5.2 m). wave height, H. was 3.9 inches (10 cm) at show that after 10 minutes 90 and 77 percent.
19 feet (5.8 m) (fig. 129). respectively. fell within the 5- 8 range and that

..
11:1
4D
, ,,
I

I
,
:I
c:
DIREcnON OF WAVE PROPOGA nON
'e I
I
I
....
0
I I
~
4D
\ I
~
,I I
I

11:1
4D
'5c:
'e
~
lii
.:::

23 16 15
I I
7.0 6.7 6.4 6.1 5.8 5.5 5.2 4.9 4.6
METRES

PLAN VIEW OF RIPPLE PATTERNS for experiment 2 after 10 and 50 minutes of wave action. Trough line dashed. Letters A - Eindi-
cate groups of ripples listed in table 21. (Fig. 128.)

Digitized by Google
 LABORATORY STIJDIES OF SAND PATI'ERNS RESULTING FROM CURRENT MOVEMENTS 181

DIRECTION OF WAVE PROPOGATION

:8
5I:
'E
-
o

,
/
,
/
/
I
I
I
I
I
22 21 20 19 16
FEET
I I I I I I
6.7 6.4 6.1 5.8 5.5 5.2 4.9
METRES

PLAN VIEW OF RIPPLE PATTERNS for experiment 3 after 10 and 50 minutes of wave action. Trough lines dashed. Letters A- F indi-
cate groups of ripples listed in table 22. (Fig. 129.)

TABLE 21. - Summary of ripple dimensions (in centimetres) TABLE 22. - Summary of ripple dimensions (in centimetres)
and ripple symmetry index for experiment 2 after 10 and 50 and ripple symmetry index for experiment 3 after 10 and 50
minutes minutes
[Lettera designatins ripple groupe correspond to thoae in figure 128) [Lettel'l dealpatins ripple groupe correspond to thOle in figure 129)
Ripple Ave.... Average Average Average Ripple Ripple Ave.... Average Average Average Ripple
group wavelensth height stOlll lee symmetry group wavelensth height stOlll lee aymmetry
index (RS/) index (RS/)
After 10 mlnut.. After 10 mlnut..
A ....... 6.7 1.0 4.0 3.2 1.3 A ....... 9.7 1.5 6.4 3.4 1.9
B ....... 10.9 1.5 6.8 3.3 2.0 B ....... 10.8 1.8 7.2 3.7 2.0
After 110 mlnut.. After III mlnut..
C ....... 12.1 2.1 7.5 4.5 1.7 C ....... 10.2 1.4 6.1 4.1 1.4
D ....... 13.2 1.5 8.3 4.9 1.7 D ....... 18.5 2.5 11.0 7.6 1.5
E ....... 11.5 1.8 7.6 3.9 2.0 E ....... 18.3 1.7 7.5 4.8 1.6
F ....... 8.1 1.4 5.4 2.7 2.0

after 50 minutes 88 percent of the RI values for both cent. respectively. fell within these limits. The
experiments fell within this range. RBI values decrease in the RSI percentages appears to be
calculated for experiments 2 and 3 show that after caused by ripples that are being washed out in the
10 minutes 95 and 100 percent. respectively. fell area between 17 and 19 feet (5.2 and 5.8 m) (figS.
within the 1.1- 3.8 range cited for asymmetrical 128. 129). This washing out resulted in crests
wave ripples and after 50 minutes 90 and 82 per- becoming platform and in troughs filling.

Digitized by Google
182 A STUDY OF GLOBAL SAND SEAS

Ripple Development Patterns Therefore, sand movement was by traction and

Figure 130 shows the transition from one form to saltation on the stoss side and dominantly by
another that oscillatory ripples make with time. suspension and avalanching on the lee side of the
This change in pattern seems to develop as (1) a ripple. With time, this process of sand movement
function of the flow velocity of the water and (or) could change an isolated. crescentic ripple into a
water depth and (2) time. Ripples during experi- coalescing crescentic ripple and a coalescing cres-
mentsl, 2, and 3 appeared to develop in the follow- centic ripple into a parallel-sinuous or parallel-
ing manner: Sand movement was dominantly by straight ripple. If Urn (the maximum near-bottom
traction and saltation in a down wave direction fluid velocity) increased above a certain critical
when the ripples first started to develop. When one velocity, sheet flow instead of oscillatory wave mo-
ripple became more distinct by its height increas- tion would occur near the bottom. With this change
ing, sand movement was affected by the vortex or in Urn toward sheet-flow conditions, ripples become
eddy developed in the lee of the ripple. The circular washed out, and eventually a planar surface devoid
action of the eddy churned up the sediment on the of ripples develops.
lee surface and in the trough in front of the ripple, Unidirectional current ripples undergo a some-
and this churning put the finer sand in suspension. what similar change in ripple patterns as a function
The finer material was held in suspension and of flow velocity and (or) water depth (Allen, 1968,
moved in both a downwave and an upwave direc- p. 93) (fig. 131).
tion due to the oscillation of the passing wave. The
coarser material dropped out of suspension and Comparison of Aqueous Wave
caused avalanching on the lee surface of the ripple.
Ripples (Aqueous Dunes) and
DECREASING DEPTH
INCREASING VELOCITY
) Eolian Dunes
The similarity in crestal patterns of aqueous rip-
ples (dunes) and eolian dunes is striking (table 23;
fig. 132). In an aqueous environment having limited
sand supply, crescentic ripples develop (fig. 132,
aqueous forms; fig. 121). In an eolian environment
6 barchan dunes occur under the same sand-supply
conditions. When sufficient sand is available,
SEQUENCE OF SMAll-SCALE RIPPLES (plan parallel-straight and branching aqueous ripples
view), changing with time and with an increase develop, and they appear to be very similar to
in flow velocity and (or) water depth. Forms
are (1) isolated crescentic, (2) coalescing cres-
eolian barchanoid and transverse ridges.
centic, (3) parallel-sinuous, (4) parallel-straight,
(5) irregular pattern, and (6) bar zone (zone of
ripple destruction). (Fig. 130.) TABLE 23. - Bedforms oriented transverse to wave or
wind direction
DECREASING DEPTH Oscillatory wave ripples Eolian dunes
)
INCREASING VELOCITY Crescentic solitary form Barchan form (solitary)

\1!t11~1l!~
Coalescing crescentic form Barchanoid ridge
(parallel-sinuous form) (coalescing barchanoid
form)
1 2 3

Parallel-straight and Transverse ridge

SEQUENCE OF SMAll-SCALE RIPPLES (plan view)
branching form
changing with an increase in flow velocity and (or)
water depth. Sketched and modified from J. R. l.
Allen (1968, fig. 4.61 a). Forms are (1) parallel-
straight, (2) parallel-sinuous, and (3) linguoidal.
(Fig. 131.)

Digitized by Google
 LABORATORY STUDIES OF SAND PATTERNS RESULTING FROM CURRENT MOVEMENTS 183

AQUEOUS BEDFORM~Laboratory EOLIAN BEDFORM~Sketch and

wave tank examples ERTS examples
1. Crescentic 2. Coalescing crescentic 1. Barchanoid solitary
solitary (parallel-sinuous)

2. Barchanoid ridge
(coalescing barchanoid)

E1183-06194
3. Parallel-straight and branching
3. Transverse ridge

E1128-04253

COMPARISON OF AQUEOUS AND EOLIAN BEDFORMS (plan views). Direction of wave

propogation and of wind is from left to right. White in aqueous bedforms 1 and 2 is bottom
of wave tank exposed by erosion due to limited sand supply. Numbers next to round
photographs refer to Landsat (ERTS) imagery from which examples were taken. (Fig. 132.)

Effects of Barriers on Ripple transverse ripples propagated in a downwave direc-

Development tion from the ball. The asymmetry of the wave-
generated flow prevented upwave ripple formation
Experiments were conducted to study the effects from the ball during this experiment. Initial small-
of barriers on wave ripple development. Barriers scale ripple development is similar in an eolian en-
were placed on a flat sand surface to disrupt the vironment (fig. 134). Comparisons of bedforms can
wave motion, and photographs were taken of the also be made on a much larger scale, as evident on
resultant ripple development. The barriers used Landsat (ERTS) imagery of eolian dunes in the An
were a steel ball 3.5 inches (9.0 cm) in diameter and NafUd Desert, Arabian Peninsula (figs. 135, 136).
clay bricks of various dimensions. Figure 133 il- The images show questionable barchan dunes
lustrates the ball's effect on ripple development. developing into barchanoid ridges, and the develop-
Scour occurred around the edges of the ball, and, ment of dunes across a geomorphically controlled
with time, crescentic, coalescing crescentic, and area.

Digitized by Google
184 A STUDY OF GLOBAL SAND SEAS

SEQUENTIAL PHOTOGRAPHS SHOWING EFFECTS of a barrier and, 8, after 5 minutes. C, Isolated crescentic form developing
on ripple development in a wave tank. Steel ball 3.5 inches (9 after 82 minutes. D, Isolated and coalescing crescentic forms
cm) in diameter. Direction of wave propogation is left to right. developing after 107 minutes. E, Coalescing crescentic and
A, Scour occurring around the edges of the ball after 2 minutes, parallel-straight forms developing after 114 minutes. (Fig. 133.)

SEQUENTIAL PHOTOGRAPHS OF BEDFORMS (1- 6)

developed in a wind-tunnel experiment. Wind direction is from
left to right. Crater is approximately 7 inches (18 cm) in
diameter, modeled in loose sand and placed in the wind tunnel.
Wind velocity is 160 inches per second (420 cm/s). Slight rip-
pling of sand in photographs 4 - 6 is effect of tunnel wall.
Photographs from Greeley and others (1974, p. 487). (Fig. 134.)

Digitized by Google
 LABORATORY STUDIES OF SAND PATIERNS RESULTING FROM CURRENT MOVEMENTS 185

DEVELOPMENT OF DUNES across a geomorphically con-

trolled area, An NafDd Desert, Arabian Peninsula. General
wind direction from northwest. From Landsat (ERTS) imagery
E-1226-06593-7 01 . (Fig. 136.)
QUESTIONABLE BARCHAN DUNES developing into
barchanoid ridges, An NafDd Desert, Arabian Peninsula. General
wind direction from west-northwest. From Landsat (ERTS) im-
agery E-1171-06530-7 01 . (Fig. 135.)

Digitized by Google
Digitized by Google
 ' ....

A STUDY OF GLOBAL SAND SEAS

ANCIENT SANDSTONES
CONSIDERED TO BE EOLIAN
Chapter H
By EDWIN D. McKEE
With sections by JOAO J. BIGARELLA9

Contents

Page Page

Summary of conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . .. 189 lyons Sandstone (Permian), U.S.A. . . . . . . . . . . . . . . . . . . . 217

Introduction ...................................... 189 Casper Formation (Pennsylvanian and Permian), U.S.A.. 221
Distribution and age ............................... 189 Rotliegendes of northwestern Europe (Permian) ....... 225
Characteristics of eolian sandstones . . . . . . . . . . . . . . . . . . 191 lower Mottled Sandstone (Permian?), England ........ 227
Characteristics of interdune record . . . . . . . . . . . . . . . . . .. 193 Frodsham Member of so-called Keuper Sandstone Forma-
Recognition of dune types. . . . . . . . . . . . . . . . . . . . . . . . . . 194 tion of England (Triassic) ......................... 228
Facies associated with dunes . . . . . . . . . . . . . . . . . . . . . . .. 196 Barun Goyot Formation (Upper Cretaceous), Mongolia. 230
Selected examples of eolian sandstones .............. 196 Botucatu and Sambaiba Sandstones of South America
Coconino Sandstone (Permian), U.S.A. ............... 197 (Jurassic and Cretaceous), by Joao J. Bigarella ....... 233
De Chelly Sandstone (Permian), U.S.A. ............... 204 Cave Sandstone and similar sandstones of southern Africa
Navajo Sandstone (Triassic? and Jurassic), U.S.A. ...... 209 (Triassic), by Joao J. Bigarella . . . . . . . . . . . . . . . . . . . . . . 236

187

Digitized by Google
188

Illustrations
Pap
FIGURE 137. Block diagrams of basic forms of cross-strata in eolian-type sandstones 192
138. Photograph of ripple marks in eolian-type Coconino Sandstone showing
rounded crests and high indices .............................. 193
139. Selected stereonets showing distribution of cross-strata dip directions in
typical eolian-type sandstones ................................ 195
140-145. Photographs taken in Grand Canyon National Park, Arizona, U.S.A:
140. Permian Coconino Sandstone forming cliff on west side of Her-
mit Basin .......................................... 198
141. Permian Coconino Sandstone showing sets of high-angle
cross-strata and tangential bases of foresets, Dripping
Springs Trail. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
142. Wedge-planar cross-strata in Permian Coconino Sandstone,
Kaibab Trail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
143. Ripple marks on foreset surface of cross-strata Coconino
Sandstone, Kaibab Trail ............................. 200
144. Slump marks on slipface of cross-strata, Coconino Sandstone,
Grandview Trail .................................... 201
145. Miniature terrace-and-cliff structure formed on foreset surface
of Coconino Sandstone, Bunker Trail ................. 201
146-151. Photographs showing:
146. Contorted bedding in large-scale cross-strata of Coconino
Sandstone near Flagstaff, Arizona . . . . . . . . . . . . . . . . . . . . . 202
147. Footprints of probable reptile preserved on dip slope of
cross bedded Coconino Sandstone, Grand Canyon Na-
tional Park, Arizona .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
148. Uphill tracks of chuckwalla lizard made on foreset slopes of
fine-grained dune sand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
149. Trails of millipedes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
150. large- and medium-scale cross-strata of tabular-planar type in
Permian De Chelly Sandstone, Canyon De Chelly, Arizona 2rTl
151. Interdune deposit between cross-strata sets of eolian dune
type, De Chelly Sandstone, Canyon De Chelly, Arizona 208
152. Isopach-lithofacies map and cross-section showing distribution and
thickness of Navajo, Kayenta, and correlative units. . . . . . . . . . . . . . 210
153 -156. Photographs showing:
153. Ooss-strata in Navajo Sandstone north of Kanab, Utah. . . . . 212
154. Contorted structures in Navajo Sandstone, Kanab Creek area,
Utah.............................................. 213
155. Navajo Sandstone, showing large-scale wedge-planar cross-
strata between tabular-planar cross-strata, Zion National
Park, Utah ........................................ . 216
156. Pits of raindrops in modern sand of lyons texture ......... . 219
157. Block diagram of festoon cross-lamination from sandstone of Casper For-
mation, Wyoming .......................................... . 222
158. Aerial photograph of blowout dunes along south shore of lake Michigan 223
159. Outcropping areas in South America of the Mesozoic Sambaiba and
Botucatu Sandstone ........................................ . 234

Table
TABLE 24. Age, thickness, composition, and postulated environments of units form-
ing lower "Keuper" and underlying sandstones, England ........ 229

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 189
Triassic(?) age. Subsequently, many other
Summary of Conclusions sandstones of that region, having similar textural
and structural characteristics, have been attributed
~NcmNT SANDSTONES of many ages and in to the accumulation of dune sand. Additional kinds
~many parts of the world have been at~ibuted of evidence have been recognized in some of these
to an eolian origin. They are generally recogDlzed by formations, which make their eolian genesis
their primary structures, especially cross-strata reasonably certain, but in others the origin has re-
which characteristically are on a large scale and dip mained debatable and subject to different in-
at high angles. Other diagnostic structures are ripple terpretations.
marks with distinctive orientation and indices, For the most part, the eolian-type sandstones of
unique forms of slump marks, and contorted bed-
the Western United States, over which controversy
ding of types characteristic of dry sand. Commonly,
regarding genesis has continued, are mixtures of
the sand is well sorted and fine to medium grained,
two or more facies in which intertonguing relations
and has surficial markings, such as frosting and pit-
exist or one rock type grades into another. Where
ting. Sharply defined animal tracks and trails are
either fluviatile and eolian facies or marine and
also distinctive features of eolian sandstones;
eolian facies abut, or overlie one another, the diag-
associated units of interdune deposits containing
nostic features of one environment frequently have
nonmarine faunas occur within sequences. Thus,
been used in interpreting both. Furthermore, the
although the genesis of some sandstones of this
surfaces and deposits of interdune areas that nor-
type is open to question, the origin of most seems
mally occur intercalated between sets of dune
sufficiently definite to merit calling them eolian-
sands commonly have not been recognized for what
type sandstones.
they are and have confused interpretations by their
As in modem sand seas, the dominant wind or noneolian features.
winds were largely responsible for development of Major advances in understanding the eolian
individual dune forms preserved in each eolian- development of certain sandstone bodies resulted
type sandstone. Because wind direction commonly from several basic pioneer investigations. Statistical
can be determined from orientation of slipfaces, dip studies on the orientation of cross-strata and their
directions of cross-strata in ancient eolian-type relations to current (wind) directions were initiated
sandstones normally will reveal both the course of in the United States by Knight (1929) in the Casper
the wind and the type of dune represented. Formation of Wyoming and by Reiche (1938) in the
Unidirectional, bidirectional, and possibly multi- Coconino Sandstone of Arizona. Analysis of cross-
directional winds can be identified in many ancient strata patterns and of cross-strata surfaces of ero-
sandstones by the clustering and spread of dip-
sion and deposition in migrating dunes of the
directional measurements in cross-strata. The type
barchanoid and transverse types were elucidated
of resulting dune - barchan, linear, or star - can
by Shotton (1937) for rocks referred to as the
also be recognized by the dip-direction spread and
"Lower Bunter Sandstones" in England. Shotton's
by certain other features of the cross-strata. Coastal
study served to explain basic relations between
dunes can be distinguished from interior desert
wind direction, availability of sand, and dune struc-
dunes in part by the geometry or shape of the sand
ture. Still later, work of Bagnold (1941) on sand
body and in part by their intertonguing or other
movement and dune processes in modem deserts
close association with bordering marine deposits.
clarified many features of wind activity, greatly
stimulating the study of eolian deposits in general
Introduction and providing a better understanding of their
LONG SWEEPING CROSS-STRATA of even- development.
grained sandstone that form the sheer walls of
many canyons in southwestern Utah led Hun-
Distribution and Age
tington in 1907 (p. 381) to ascribe an eolian origin to SANDSTONES OF PROBABLE eolian origin are
these rocks (Navajo Sandstone) of Jurassic and widespread in both time and space throughout the

Digitized by Google
190 A STUDY OF GLOBAL SAND SEAS

world. Cross-strata attributed to deposition by wind Of the just-listed sandstones in the Western
have been reported from rocks of the Precambrian, United States, some clearly represent extensive
Paleozoic, Mesozoic, and Cenozoic Eras; they are dune deposits of interior deserts, as evidenced not
described from North America, South America, only by the great lateral extent of the sandstone
Europe, Asia, Africa, and Australia. Not all these bodies but also by distinctive features of primary
assignments to eolian genesis have stood the test of structure and by faunal traces. To illustrate desert
critical analysis, but, in general, more sandstones dune deposits, the Coconino Sandstone and the
are interpreted as dune deposits today than were a Navajo Sandstone are selected for detailed discus-
few decades ago. In brief, the well-sorted fine- sion later in this chapter. In contrast, coastal-plain
grained sandstones with large-scale cross-strata sand accumulations are illustrated by the Lyons
that commonly are attributed to eolian processes Sandstone (Permian) and the Casper Formation
are numerous in rocks of various ages, and they (Permian and Pennsylvanian), which contain struc-
constitute a considerable amount of the geologic tures typical of a coastal dune environment and in-
column. tertongue with marine strata. The De Chelly
Eolian-type sandstones are especially numerous Sandstone seems to be in a position intermediate
and widely distributed throughout the Western and between the Coconino and various noneolian
Southwestern United States in rocks of late depositional environments.
Paleozoic and early to middle Mesozoic age. They Extensive eolian like deposits are represented in
have been extensively studied by many geologists, South America (Brazil) by the Botucatu and Sam-
although genesis remains unsolved in a number of baiba Sandstones of Mesozoic age and in Africa by
formations. A summary of these sandstones of the Cave and time-equivalent sandstones which are
probable eolian origin has been made for the Col- also of Mesozoic age. Texturally and structurally
orado Plateau by Poole (1962). He grouped them by these formations resemble sandstones of the
age and recognized a total of 15 lithic units in 4 ma- Western United states and are described by
jor age categories. Some of these sandstones com- Bigarella later in this chapter. In Africa also are
prise entire formations; some consist of tongues "feldspathic arenites" of Zambia said to be of Eo-
that extend into other lithologically dissimilar for- Cambrian [late Precambrian] age and attributed by
mations; others are merely parts or facies of a for- Garlick (1969, p. 114) to an eolian origin. In
mation associated with one or more other Australia the Chatsworth Limestone of Cambrian
lithofacies. age, with a thickness of 12 m (36 ft) has been at-
The principal eolian-type formations of the Col- tributed to eolian deposition on the basis of struc-
orado Plateau listed by Poole are as follows: tural features (Shergold and others, 1976, p. 10,
Late Jurassic time 33- 34).
Entrada Sandstone In northwestern Europe the Rotliegendes of Per-
Cow Springs Sandstone mian age has generally been interpreted as a desert
Jurassic and Triassic(?) time deposit that includes considerable eolian sand.
Navajo Sandstone Because of its petroleum potential, it has been ex-
Aztec Sandstone tensively studied in England, the Netherlands, Ger-
Jurassic(?) and Triassic(?) time many, and elsewhere. In England, the "so-called
Nugget Sandstone Lower 'Bunter' Sandstone or Lower Bunter Mottled
Triassic time Sandstone, probably Permian in age" and also the
Wingate Sandstone "Yellow Sands of Durham and parts of the Keuper
Moenkopi Formation Sandstone of probable Permian and Triassic Age"
Permian time are believed to be eolian (D. B. Thompson, written
Coconino Sandstone commun., 1976). For comparative purposes, these
White Rim Sandstone Member of the Cutler stratigraphic units are briefly described in this
Formation chapter.
De Chelly Sandstone The Barun Goyot Formation of Late Cretaceous
Permian and Pennsylvanian time age in central Asia probably represents deposits of a
Weber Sandstone great interior desert. Exceptionally thick and

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE FDUAN 191
widespread interdune sediments occur between contain chert grains. Cross-stratified gypsum sands.
sets of large-scale cross-stratified sandstones con- believed to be wind deposited. occur locally in the
sidered to be eolian. The interdune sediments prob- Triassic Moenkopi Formation of Arizona (McKee.
ably were accumulated both by intermittent 1954. pI. 38).
streams and in temporary ponds and lakes. Unique Textures in the ancient eolianlike sandstones
in these deposits - both dune and interdune - is resemble. in most respects. those in modern dune
their extensive fauna of vertebrate animals. fields; however. many of the same characteristics
The Precambrian Makgabeng Formation of the are also found in modern beach deposits and some
Waterberg Supergroup (northern Transvaal) is con- other deposits. The sand is typically very fine to
sidered (Meinster and Tickell. 1976. p. 191) "of medium. well sorted. rounded. and frosted. In many
aeolian origin and shows many of the charac- sandstones. these grains represent reworked
teristics of transverse dunes." material (second or third generation); hence. the
textural properties may represent inherited
Characteristics of Eolian features and bear little relation to length or manner
of transport in the present regime. Furthermore.
Sandstones frosting may be the result of postdepositional
ANCIENT SANDSTONES believed to be of chemical alteration.
eolian origin are mostly light colored (white. buff. In some ancient eolian sandstones the texture
yellow. and pale red). highly cross-stratified. even has been greatly modified by quartz overgrowths or
grained. and cliff-forming. The criteria by which by iron oxide coatings on the grains. or both. Com-
their eolian genesis may be determined are of two monly. there is intergranular silica. calcite. or iron
types: (1) features that are distinctive of modern oxide. but matrix is rare because silt and finer parti-
eolian deposits and therefore may be considered cles forming dust were winnowed out by the wind
diagnostic. or (2) features that are compatible with at the time of deposition. Coarse grains. except
an eolian environment but furnish no proof of wind those developed as lag on the horizontal interdune
deposition. deposits. are also lacking. Most lag deposits that
The geometry of dune sandstone bodies probably have been described. as in the Lyons and Casper
depends largely on the general environment of Formations. are coarse-grained sandstones. but gra-
deposition - whether an inland desert basin or a nules may occur on some interdune surfaces.
coastal strip. Some ancient sand bodies that are especially near the margins of a dune field.
very extensive. such as the Wingate. Navajo. and
Coconino Sandstones of western North America. Whereas composition reflects the source of a
are either tabular- or wedge-shaped on a huge scale sediment and texture is largely a function of
and definitely represent sinking basins. Others. transportation. primary structures furnish the
such as the Lyons Sandstone and the Casper For- record of manner of deposition. The most common
mation. farther east. occur as relatively narrow and most distinctive structure of eolian deposits is
sheets that apparently bordered strandlines and cross-stratification. Other characteristic structures
~eveloped along the margins of regressing seas.
preserved in some eolian-type sandstones are ripple
Boundaries of eolian sandstones mostly are marks. rain pits. and slump marks. All these are
abrupt. for these sandstones normally interfinger diagnostic features of dune deposits. especially
with. rather than grade into. adjacent facies. Exam- when used as part of a critical evaluation of all
ples are numerous of contacts either with marine available data (Selley. 1970. p. 52).
deposits or with fluviatile and lacustrine deposits. Cross-strata of dune-type sandstones are charac-
Contrasts in lithology usually make the contacts teristically. though not entirely. large to medium
easily recognizable. scale and high angled. They commonly attain
Most ancient dune-type sandstones are com- lengths of more than 100 feet (30 m) and dip
posed almost entirely of quartz grains. and many 20 -30. These dips are somewhat less than the
are classed as orthoquartzitic. Some. like the De angle of repose for modern dune-sand slipfaces.
Chelly. Lyons. and Casper. contain small percen- which is commonly 33 or 34. but the difference is
tages of feldspar (much authigenic). and others attributed to subsequent compaction (Glennie.

Digitized by Google
192 A STUDY OF GLOBAL SAND SEAS

1972, p. 1058; Walker and Harms, 1972, p. 280) or to tion vector opposite that of the regional current
differential preservation with higher, steeper parts direction. (See description under Coconino
removed (Poole, 1962, p. 148). Low-angle cross- Sandstone in this chapter.)
strata typical of windward-side deposition are occa- Cross-stratification is mostly of two basic
sionally preserved on modern dunes and have been types - tabular planar and wedge planar (fig. 137).
recognized in ancient sandstones by their dip-direc- In the tabular-planar type, lower bounding surfaces
are believed to represent interdune deposits or ero-
sion surfaces; dipping foresets above form tangents
at their bases as in most modem barchanoid dunes.
In the wedge-planar type, abutting sets of cross-
strata probably reflect shifting winds within a
single dune sequence. Simple lenticular types are
recorded, but the trough type is rare, with the nota-
ble exception of the remarkable "festoon struc-
tures" in sandstone of the Casper Formation. These
structures have not been established unequivocally
as eolian but may represent "blowout" dunes.
Ripple marks, not uncommonly preserved in
A many of the eolian-type sandstones, are distinctive
(McKee, 1945, p. 318). They invariably have high
ripple indices (above 15, mostly far above), in con-
trast to the low indices of water-formed ripple
marks (Tanner, 1966). In addition, most of them are
oriented with their parallel troughs and crests pass-
ing up and down the steeply dipping foresets of
cross-strata - a feature not recorded for ripple
marks under water - and their crests generally are
well rounded (fig. 138).
Adhesion ripples, formed where dry sand is
blown onto moist surfaces and fixed by surface ten-
sion, are recorded (Glennie, 1972, p. 1061) at many
B horizons in the Rotliegendes of northwest Europe
and are viewed as evidence of interdune surfaces.
Raindrop craters with raised rims and reoriented
basins formed on the dip slopes of high-angle cross-
strata have been recorded from the Coconino
Sandstone (McKee, 1934b, p. 102) and the Lyons
Sandstone (Walker and Harms, 1972, p. 282). These
pits are typical of modern raindrop craters in dry
sand and are considered to be good evidence of
eolian deposition.
Various forms of avalanche or slump marks are
preserved on cross-strata surfaces in several eolian
type sandstones (McKee, 1945, p. 321), and con-
torted strata are locally prominent, although, in
c general, such structures are surprisingly uncom-
BASIC FORMS OF CROSS-STRATA recognized
mon. In the Coconino Sandstone the "miniature
in eolian-type sandstones. A, Tabular-planar, step" variety seems to be good evidence of slump-
abundant; 8, wedge-planar, abundant; C, ing at the angle of repose when the sand was wet
trough type, rare. (Fig. 137.) (Reiche, 1938, p. 918), and other varieties suggest

. . Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE FDUAN 193
dry sand movement (McKee, Douglass, and Rit- covered, showing clearly that the sand in the area
tenhouse, 1971, p. 368). was continental (Brady, 1935; Camp, 1936). Many
dinosaur eggs and some dinosaur skeletons have
been found in dune-type cross-strata of the Barun
Goyot Formation in Mongolia (Gradzinski and
Jerzykiewicz, 1974b, p. 140).

Characteristics of
Interdune Record
COMPARED TO THE AMOUNT of attention
that has been given dune deposits, the record of in-
terdunes has been largely neglected and is as yet
only poorly understood. Interdune areas differ
greatly from place to place in modem dune fields -
RIPPLE MARKS with high ripple indices and rounded crests in ranging from long narrow corridors between
slab of eolian-type sandstone, Coconino Sandstone, Kaibab
Trail, Grand Canyon National Park, Arizona, U.S.A. (Fig. 138.)
barchanoid ridges to small rounded or elliptical
depressions in "blowout" areas and from very
wide, flat surfaces between linear dunes to the typi-
Some of the most conclusive evidence of eolian cal radially sloping surfaces within parabolic dunes.
origin for cross-stratified sandstones is the presence In the geologic record some interdune deposits
of clearly preserved footprints of four-footed can be recognized by texture and structure, but
animals on steep dip slopes. Experimental work has many interdunes are recorded merely as erosion
shown that tracks with such excellent preservation surfaces within vertical sequences of dune sands.
and with orientation uphill only must have been Whether or not characteristic sediments are accum-
formed on dry sand at or near the angle of repose. ulated and preserved seems to be at least partly a
These tracks could not have been formed under function of time for detritus to accumulate before
water (Peabody, in McKee, 1947, p. 27) or in either being buried by migrating dunes. The process in-
wet or damp-crust sand (McKee, 1947, p. 25). volved in the preservation of interdunes has been
Tracks of this type are common and widespread in described for barchanoid and transverse ridges of
the Coconino Sandstone and are also present in the the White Sands dune field, New Mexico (McKee
De Chelly and Lyons Sandstones. and Moiola, 1975).
Selective preservation of certain parts of dune
Some interdune areas within major sand seas
sands doubtless is responsible for absence in the
develop in an environment controlled by standing
geologic record of many features that occur in
bodies of water, either at the surface or as ground
modern dunes. Upwind features, for instance, are
water directly below. The term "interdune
mostly destroyed by erosion on that side of a dune,
sabkhas" was applied to such environments by
and the steeply dipping cross-strata at the top of
Glennie (1972, p. 1061); he described the deposits
each set commonly are nearly or entirely removed
that accumulate in them as consisting of "small, ir-
by planation prior to deposition of the overlying set.
regular. and more or less horizontally bedded adhe-
Thus, the normally abundant, horizontally oriented
sion ripples" of sand grains that "range from fine to
ripple marks, as well as many tracks and trails on
coarse and generally are poorly sorted." Sediments
low-angle surfaces, are very rarely preserved in
with similar textures and structures, preserved in
eolian sandstones. the Permian Rotliegendes of northwest Europe,
Vertebrate fossil remains are rarely found in the were interpreted by Glennie as interdune deposits.
highly permeable cross-stratified sand of eolian "Steeply dipping eolian sandstones," separated by 9
deposits, presumably because conditions for preser- inches of nearly horizontal adhesion-rippled
vation were poor. In the Navajo Sandstone, two ar- sandstone are shown (Glennie, 1972, fig. 14) in a
ticulated skeletons of dinosaurs have been dis- drill-core photograph.

Digitized by Google
194 A STUDY OF GLOBAL SAND SEAS

Excellent examples of strata believed to have Recognition

been formed as interdune deposits, without adhe-
sion ripples, are preserved in the type De Chelly of Dune Types
Sandstone of Canyon De Chelly, Arizona, where
sets of high-angle large-scale cross-strata of eolian THE RECOGNITION OF various basic types of
type are separated by extensive thin units of dunes in ancient sandstones is possible where cer-
horizontal strata composed of poorly sorted red- tain diagnostic features are well developed and ac-
brown silt and sand (McKee, 1934a). cessible for study. Those differences that result
from changes in wind direction, amount of availa-
In many eolian-type sandstones, interdunes are ble sand, and the presence or absence of moisture in
represented by few or no deposits. However, their the sand at the time of deposition are commonly
positions in vertical sections are marked ~y sur- reflected in the form of distinctive structures in the
faces of beveling that form the nearly honzontal sandstones.
bounding planes of adjoining cross-strata sets. Amount of spread and distribution of cross-strata
Many such surfaces are developed in the Coconino, dip directions are perhaps the most satisfactory
Navajo, and similar sandstones of the Southwestern features for differentiating types of dunes in an-
United States. They are prominent planes and are cient rocks. Study of structures in modern analogs
more irregular than the diagonal erosion surfaces has furnished some useful and reliable generaliza-
that separate wedge cross sets, probably because tions. For example, unidirectional winds that are
they represent greater time spans. In some responsible for barchans, barchanoid ridges, and
sandstones, such as the Lyons Sandstone and the transverse ridges normally have croBS-strata dip
Casper Formation, the truncation planes are gently directions restricted to a single small arc, about one
curving, rather than flat, and have coarse grains in- quadrant or less (fig. 139). This type of distribution
terpreted as lag scattered on their surfaces. Injec- (A, B) is illustrated by rose diagrams or stereonets
tion structures in some deposits, like the Lyons for the Navajo Sandstone, Coconino Sandstone, and
Sandstone, have been recorded. numerous others in the Colorado Plateau of the
In various ancient eolian-type sandstones, the Southwestern United States.
bounding surfaces overlying large-scale croBS-strata In linear dunes, commonly called seif or
commonly preserve the truncated upper traces of longitudinal, the distribution of cross-strata dip
those strata. The resulting dune pattern may then directions (chapter E, fig. 61) is very different from
etch out on weathering to show on a horizontal that in transverse dunes. Even though no general
plane the character of cross-strata sets. Such struc- agreement on exact cause of linear dunes has yet
tures are observed on many modern interdune sur- been reached, their pattern of cross-strata was pre-
faces before their burial by fresh sand (chapter E,
dicted (Shotton,1956, p. 460) and has been recorded
figs. 45, 70, 71). Likewise, cross-strata sets have
in trenches (McKee and Tibbitts, 1964, fig. 6). It is
been recorded in ancient eolian-type sandstones,
represented on a stereonet by two separate clusters
including the Lyons Sandstone of Colorado (Walker
of dip attitudes almost opposite one another (Glen-
and Harms, 1972, p. 284) and the Navajo and
nie, 1972, fig. 13b). Ancient eolian sandstones
Coconino Sandstones of Arizona.
representing linear (seif) dunes have been recog-
Coarse sand or granules formed as lag are in-
nized in the Permian Rotliegendes of the North Sea
troduced in interdune areas by the process of creep
area by Glennie (1972, p. 1061) and probably occur
resulting from saltation (Bagnold, 1941, p. 102).
in many other formations.
Probably, they are largely restricted to upwind.
areas (McKee and Tibbitts, 1964, p. 13) or to inter- A third distinctive pattern represented by the
dunes near source materials as along coasts because spread in cr088-strata dip directions is that of
of the slow particle movement involved. Most ex- parabolic and other "blowout" -type dunes. These
amples in ancient dune sandstones containing lag dunes occur wherever vegetation or moisture, or
seem to be from coastal dune areas. both, tend to stabilize or anchor the sand but are

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 195
especially characteristic of coastal areas. The the characteristic spread of cross-strata orientation
feature of the spread in dip directions is its width. (fig. 139).
encompassing an arc of 180 or more. some being as
0 Reversing dunes. in which nearly opposite wind
great as 270 as shown by Bigarella (chapter E. figs.
0 directions periodically affect the sand body in alter-
78. 79; tables 10. 11). Among ancient rocks. the nate direction and so develop two opposed slip-
Casper Formation of Wyoming (Knight. 1929. p. 65. faces. form distinctive patterns of dip directions.
66; this chapter. fig. 139) and the Lyons Sandstone Such dunes occur in various modern sand seas
of Colorado (Walker and Harms. 1972. p. 281) show (among them. Great Sand Dunes National Monu-

N N N

S
A B B'
PERMIAN COCONINO SANDSTONE TRIASSIC(7) AND JURASSIC NAVAJO SANDSTONE. MIDDlE UNrr

N N N

c S
c' D
PENNSYLVANIAN AND PERMIAN .CASPER FORMATION PERMIAN DE CHELLY SANDSTONE

SELECTED STEREONETS, SHOWING DISTRIBUTION OF CROSS-STRATA dip directions (dots) in typical eolian-type sandstones,
Westem United States. Azimuth of dot represents direction of dip; distance of dot from center indicates degree of dip. Dashed line
enclosing dots represents foreset direction considered the result of downwind<urrent deposition. A, Bunker Trail, Grand Canyon
National Park, Arizona (Reiche, 1938). B, Central Buckhom Wash, Utah. 8', San Rafael River, Utah (Kiersch, 1950). C, Red Buttes
area, Wyoming. C, Sand Creek area, Wyoming (adapted from histograms by Knight, 1929). 0, White House Trail, Canyon De Chelly,
Arizona (Reiche, 1938). (Fig. 139.)

Digitized by Google
196 A STUDY OF GLOBAL SAND SEAS

ment. Colorado. U.S.A.. and Florinopolis. Brazil) Deposits of desert lakes and playas. which
and should be expected in ancient eolian characteristically are shallow and ephemeral. con-
sandstones but as yet have not been recognized or tain low-energy textures and structures except
recorded. where they border high-relief areas that contribute
Additional structures that may be helpful in alluvial sediments. Most desert lakes and ponds ac-
determining the type of dune represented by an an- cumulate salt and clay in horizontal beds. various
cient eolian deposit are the uncommon low-angle amounts of windblown sand and. in some places.
dips that sometimes are formed on upwind surfaces anhydrite or other salts. They typically develop
of transverse ridge dunes and that contrast with mudcracks and mud curls during dry times when
normal high-angle cross-strata of the lee side. water is absent. Sandstone dikes commonly are
These structures were recorded in the Lower Mot- associated with these deposits. In inland sebkhas or
tled Sandstone by Shotton (1937. fig. 2) and in the sinking basins. where the water table rises and
Coconino by Reiche (1938. fig. 1. p. 909). Concave periodically reaches the interdune surface, charac-
downward foresets are distinctive features of many teristic sedimentary features are adhesion ripples
parabolic dunes. formed on the foresets of the dune and irregular or horizontal beds. commonly accom-
nose which is vulnerable to undercutting and panied by crusts of salt or gypsum (Glennie. 1972,
steepening near its base by crosswinds (McKee. p.l053).
1966. p. 51; Ahlbrandt. 1973. p. 39). Many dune fields in both humid and arid regions
Other features distinctive of different dune form adjacent to the sea and so are a part of the
types. discussed elsewhere in this chapter. are the coastal environment. Whether the wind is offshore
slump marks of wet and dry sand. the contrasts in or onshore. the beach, bar. or shallow-water marine
typical contorted structures. and differences in sands may intertongue with dune sands and pro-
associated interdune surfaces. vide a complex of environments containing typical
dune structures proximate to marine deposits. The
Facies Associated With Dunes Permian Lyons Sandstone of Colorado and the Per-
A major difficulty in the recognition of ancient mian and Pennsylvanian Casper Formation of
eolian deposits in the geologic record is their inti- Wyoming are probable examples of coastal dUne
mate association with rocks of other facies and the deposits that are adjacent to marine sediments.
temptation to use the criteria preserved in one for
interpreting the other. Interdune deposits that com-
Selected Examples of Eolian
monly involve features of an aqueous environment Sandstones
have been discussed. but in addition. all eolian This discussion of ancient eolian sandstones
deposits must have on their borders various facies seeks to describe the characteristics of sandstones
that are fluviatile. lacustrine. or marine and that in- believed to have been formed by wind action, to in-
tertongue with and. under some conditions. grade dicate those attributes considered evidence of
into them. eolian origin, and to show methods of distinguish-
Fluviatile deposits in desert areas can generally ing between the different basic dune types. Exam-
be categorized as river channel. flood plain. or ples of formations thought to have been formed by
alluvial fan. The river channels in desert areas wind activity are also given. From 30 or more
commonly are dry streambeds. referred to as ar- sandstones known to qualify as dunelike types, 12
royos or washes in the Southwestern United States examples were chosen for analysis.
and as wadis in north Africa; their deposits consist The sandstones discussed herein (I) ~ forma-
primarily of high-energy sediments but may be in- tions that have been studied in detail, (2) represent
termixed with eolian or ponded materials. The principal types of deposition and illustrate distipc-
flood-plain sediments are mostly overbank deposits tive features of various dune forms, and (3) are
and are largely low-energy varieties that contrast geographically widely distributed. Because the data
with those of the channels. Alluvial fans containing used came from many different sources and repre-
much coarse detritus are common on the margins of sent various degrees of detailed research. the con-
most dune areas that are bordered by mountainous clusions presented here differ considerably from
terrain. area to area in both scope and dependability.

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 197

The sandstone formations believed to be exam- fine grained, nearly pure quartz sandstone with
ples of interior desert dune-field deposits that were siliceous cement, has a massive appearance, and is
selected for analysis are the Coconino Sandstone of cross-stratified on a large scale. Cross-strata mostly
Arizona, U.S.A., of Permian age, and the Lower dip at high angles (fig. 141). Commonly, crossbed
Mottled Sandstone of England of Permian (1) age; sets form large truncated wedges, and numerous in-
the Navajo Sandstone of Arizona and Utah, U.S.A., dividual beds extend as much as 60 or 80 feet (18 or
the BotucatU Sandstone of Brazil, and the Cave 24 m), curving to form a tangent with the underly-
Sandstone of southern Africa, all of Mesozoic age. ing bounding plane.
Sandstones illustrating a coastal dune environment Although the Coconino characteristically is
are the Lyons Sandstone of Colorado and the white, it locally has a brownish cast, as in Marble
Casper Formation (eolian origin uncertain) of Canyon, or is bright red, as near Flagstaff. Most of
Wyoming, both in the United States and of late the sloping bedding surfaces are extremely smooth
Paleozoic age. and even, but some contain abundant ripple marks,
The Frodsham Member of the so-c~lled Keuper clearly impressed fossil footprints, and other minor
Sandstone of the Permian-Triassic Cheshire Basin structures. A general uniformity of texture and
in England, though very limited in extent, illus- structure is characteristic.
trates the dome type of dune, and the discussion
presents a method of recognizing it through struc- Distribution and Thickness
tural characteristics. The De Chelly Sandstone of The Coconino Sandstone forms a great sedimen-
Arizona typically shows well-developed interdune tary wedge across most of northern Arizona. Along
deposits between sands interpreted as transverse the entire Arizona-Utah boundary, it is either very
dunes, and it illustrates complex relations along its thin or absent, ranging in thickness from 40 feet (12
margins between dune and other continental facies. m) at Lees Ferry to 60 feet (18 m) at Hacks Canyon,
Finally, evidence from the Rotliegendes of the farther west, and 50 feet (15 m) at Grand Gulch
North Sea area is reviewed because it is believed to mine, near the Nevada State line. To the south 100
include examples of seif dunes as well as transverse miles (160 km) or more, it is more than 500 feet (150
types and because typical sebkha structures are m) thick along the Mogollon Rim of central Arizona
recognized in its interdune strata. and attains a maximum thickness of more than
1,000 feet (300 m) near the town of Pine. The
Coconino Sandstone Coconino Sandstone covers about 32,000 square
miles (51,500 kmz).
(Permian), U.S.A.
Formation Boundaries
THE PERMIAN COCONINO SANDSTONE of The white sandstone of the Coconino rests upon
northern Arizona is considered by many geologists a flat surface of red mudstone of the Hermit Shale
to represent the deposits of a large sand sea. It con- across nearly its entire area of deposition.
sists of large-scale cross-stratified quartz sandstone Generally, the contact is sharp and even, but in
(fig. 140), flat-bedded interdune deposits are sparse. some places the surface of the red Hermit is
The general lack of interdunes, with only horizon- trenched with large shrinkage cracks, probably the
tal bedding planes occurring between sets result of desiccation, which are filled with white
of cross-strata, suggests an abundant sand source sandstone of Coconino type that apparently filtered
with resulting rapid dune migration and burial of down from above.
interdune spaces. The large scale of the foresets
suggests the growth of high dunes. The upper boundary of the Coconino in Grand
Canyon sections is marked by sandstone of the
The Coconino Sandstone forms a conspicuous Toroweap Formation (Rawson and Turner, 1974).
cliff in the upper walls of Grand Canyon, along the Steeply dipping crossbeds directly below are trun-
Mogollon Rim to the south and east, and in many cated everywhere to form a flat surface at the con-
other parts of the region, where it occurs as a pro- tact. Quartz grains probably derived from the
minent white to buff layer, commonly several Coconino form the basal unit of the Toroweap and
hundred feet thick. The Coconino is a uniformly suggest that the sand of the Coconino had not yet

Digitized by Google
198 A STUDY OF GLOBAL SAND SEAS

PERMIAN COCONINO SANDSTONE, forming cliff on west side of Hermit Basin, Grand Canyon National Park,
Arizona, U.S.A. Large-scale tabular-planar cross-strata are prominently displayed. (Fig. 140.)

PERMIAN COCONINO SANDSTONE, showing sets of high-angle cross-strata and tangential bases (arrow) of foresets,
Dripping Springs Trail, Grand Canyon National Park, Arizona, U.S.A. (Fig. 141 .)

Digitized by Google
 ANCIENT SANDSTONFS CONSIDERED TO BE EOUAN 199
consolidated when the formation's top surface was
beveled and quartz sand began to cover it.
Texture
The Coconino Sandstone is composed domi-
nantly of well-sorted fine-grained (0.125- 0.25
mm) sandstone. Nearly all the grains are rounded to
subangular. and many are pitted or frosted. The
sandstone consists of white quartz with a few scat-
tered feldspar grains and traces of heavy minerals
(McKee. 1934b. p. 91). As evident from the pale col-
or. the grains are free of iron oxides in most
localities.
Uniformity of grain size and various textural
features of the Coconino sand grains have been WEDGE-PLANAR CROSS-STRATA, a common structure of the
described in detail by McKee (1934b p. 90-96). Permian Coconino Sandstone, Kaibab Trail, Grand Canyon Na-
Because sorting and texture are functions of tional Park, Arizona, U.S.A. (Fig. 142.)
transportation rather than deposition. and. further.
may reflect processes of an earlier generation. they directions of dip and expressed numerically
cannot be considered to be proof of eolian origin. measures of the consistency in direction of in-
Grain characteristics of the Coconino are. however. dividual dips. This study was the earliest in
typical in all respects of many modem dune sands. America to employ a rigid statistical approach and
has been followed by numerous similar investiga-
Primary Structures tions made to determine mean current directions
Stratification within various sedimentary environments. This
The most conspicuous primary structure quantitative and statistical method of study has
preserved in the Coconino Sandstone is the large- become an important tool for the analysis of cross-
scale cross-stratification. It is composed of long strata.
sweeping layers. many of them 30- 40 feet (9-12 A distinctive feature of the stratification in both
m) long. and some of them as long as 80 feet (24 m). dipping and horizontal strata of the Coconino
They dip mostly at 25 - 30. but a few reach a max- Sandstone is a characteristic splitting to form flat
imum of about 34 (McKee. 1940. p. 823). In general. bedding surfaces. The detachment of one laminae
the dip is southward. but the spread of readings is from another in sandstone of such uniform com-
relatively great. ranging between SW. and SE. position suggests either marked differences in grain
(Reiche. 1938. p. 908). size or differences in cohesion between individual
Cross-stratification of the Coconino is uniform laminae probably resulting from variations in wet-
throughout wide areas. forming large irregular ness or dryness of the sand surface at time of burial.
wedges truncated by overlying layers (fig. 142). Because analyses show few differences in grain size
Thin laminae form parallel curves. concave up- between laminae. the easy detachment along
ward. that are tangential with underlying bounding planes of stratification seems to reflect a lack of
planes (McKee. 1934b. p. 98). Contrasting with the cohesion in the original deposit (McKee. 1934b. p.
common high-angle cross-strata are some low- 100).
angle 15) cross-strata that dip in the opposite Ripple Marks
direction. They occur in limited numbers in a few Although not abundant throughout the forma-
localities (McKee. 1940. p. 823). Horizontal laminae tion. ripple marks are locally numerous and distinc-
constitute the basal 3- 6 feet (1- 2 m) of the forma- tive (fig. 143). They all seem to be of one general
tion at many localities. form - low. wide. and asymmetrical with ripple
A stereographic polar net for representation of indexes (ratio of wavelength to amplitude) greater
crossbedding attitudes was first used by Reiche than 15 (McKee. 1934b. p. 101). Most crests and
(1938) in connection with analyses of Coconino troughs are oriented parallel to the direction of dip
cross-strata. His analyses yielded weighted average of the foreset slopes.

Digitized by Google
200 A STUDY OF GLOBAL SAND SEAS

avalanching. constitute nearly all the sediment

preserved in ancient dune sandstones. In barchan
dunes or barchanoid ridges. lee-side ripple marks
are largely restricted to the sides or wings where
crosswinds are effective (McKee. 1934b. p. 101).
Second. most ripple marks formed in dry sand are
unlikely to be preserved on any surface unless the
sand later develops cohesion through moisture
received as dew or light rain. These special condi-
tions required for the preservation of ripple marks
in dune sand place a strong limitation on the places
and times favorable for permanent preservation
(McKee. 1945. p. 315 - 316).
Raindrop Impressions
Raindrop impressions in sand form small cra-
terlike pits. some of which have been recognized in
RIPPLE MARKS ON FORESET SURFACE of cross-strata, the Coconino Sandstone at several localities
Coconino Sandstone, Kaibab Trail, Grand Canyon National (McKee: 1934b. p. 102). These pits are especially
Park, Arizona, U.S.A. (Fig. 143.) significant in interpreting genesis. for they give
positive evidence of subaerial deposition. They
A detailed analysis made in 1945 of ripple marks differ in several respects from rain pits in mud.
in the Coconino. in which about 50 sets were ex- such as have frequently been recorded in the
amined. shows compelling evidence. on the basis of literature.
five characteristics. that the ripple marks were Rain pits in fine sand form thin. detached layers.
wind formed (McKee. 1945. p. 317- 320). The diag- or shells. roughly circular but with irregular
nostic features are as follows: margins resulting from cohesion of grains by the in-
1. Large ripple index (lowest was 17; most were troduction of moisture. These detached shells of
considerably higher). sand. when formed on sloping surfaces like those of
2. Ripple crests were definitely rounded in all the Coconino. have become reoriented with respect
specimens. to the lamination plane. and each circular pit tends
3. A concentration of coarser sand grains occurred to face upward. or vertically. thus raising the
on or about the crests in most specimens. downslope rim (McKee. 1945. p. 323).
4. All specimens in situ had crests trending up and Rain pits on sloping surfaces of fine dry sand are
down surfaces of sloping laminae; none was readily reproduced in the laboratory; they are very
normal to dip. distinctive and are not likely to be confused with
5. Crests and troughs were straight and parallel. craters having other origin. "Only brief showers
with little change in direction where exposed can leave such a record. for on the surface of a dune
for distances of 3 feet or more. that is thoroughly wetted. the shells lose
The apparent discrepancy between the abun- coherence" (McKee. 1945. p. 324. fig. 7). This struc-
dance of ripple marks on modern dunes and the ture therefore may be considered an additional in-
general scarcity and localized distribution of this dication of dry-sand origin.
structure in the Coconino may be attributed to two
factors related to preservation. First. a large ma- Slump Marks
jority of ripple marks in modem dunes are formed
on the gentle windward slopes of saltation. but in Slump marks indicate mass movement or
ancient dunes, because of differential preservation, avalanching. In sand dunes they are commonly
most windward-side deposits have not been formed when lee-side slopes near the angle of
preserved. The steep slipfaces of lee sides, where repose are oversteepened. One form, as described
few ripple marks form because deposition is by by Bagnold (1937, p. 433) , is the result of "suc-

Digitized by Google
 ANCIENT SANDSTONFS CONSIDERED TO BE EOUAN 201
cessive discontinuous jerks with miniature land-
slides." Another (McKee. 1945. p. 321) is a series of
variable and irregular lines roughly parallel to the
direction of slope and marking the border of the
sand mass that has slumped (fig. 144). These and
other types of dry-sand slump marks have been
recognized on surfaces of the Coconino (McKee.
1945. figs. 5. 6).

"MINIATURE TERRACE-AND-CLIFF STRUCTURE" (Reiche,

1938, p. 918) or "stair-step fault structure" (McKee, Douglass,
and Rittenhouse, 1971, fig. 8b) formed on foreset surface of
Coconino Sandstone, Bunker Trail, Grand Canyon National
Park, Arizona, U.S.A. White scale is 6 inches (15 cm) long. (Fig.
145.)

ly understandable otherwise." This structure con-

sists of "a succession of abrupt rises or steps of
about one-tenth inch each. ascending the slope"
(McKee. 1934b. p. 103); otherwise. the surface of
the slab is smooth and regular. The genesis of this
feature is attributed by Reiche to the local disrup-
tion and breaking down. with rapid uphill reces-
sion. of successive layers of recently wetted sand
on the lee side of a dune. This structure has been
SLUMP MARKS ON SLIPFACE of cross-strata, Coconino noted in the Frodsham Member of the so-called
Sandstone, Grandview Trail, Grand Canyon National Park, Keuper Sandstone Formation in England (D. B.
Arizona, U.S.A. Scale in inches; 1 inch equals 2.54 cm. (Fig. 144.) Thompson. written commun . 1975) and has been
reproduced experimentally in the U.S. Geological
The nature of slump marks and the differences Survey sedimentary structures laboratory in
between those formed in dry sand. damp sand. and Denver (McKee. Douglass. and Rittenhouse. 1971.
wetted sand have been studied extensively. both in fig. 8b) .
the laboratory and in the field (McKee. 1945. p. 320;
McKee. Douglass. and Rittenhouse. 1971. p. Contorted Bedding
363-375; McKee and Bigarella. 1972. p. 673- 677). Penecontemporaneously deformed cross-strata
Those described from the Coconino Sandstone are not common in the Coconino Sandstone.
seem to give further strong evidence of the forma- although they occur in a few places on a large scale.
tion's eolian origin. A spectacular example is near Doney Crater. north-
east of Flagstaff. Arizona (fig. 146). where dipping
Miscellaneous foresets many feet in length have been folded
A minor structure of the Coconino. referred to as locally. but the cross-strata above and below are
"miniature terrace-and-cliff structure" (fig. 145). undisturbed. The sand apparently had some
was said by Reiche (1938. p. 918) to have "ready ex- coherence at the time of contortion. but whether
planation as a dune feature. but * * * [to be] scarce- the strata were wet or dry is not clear. for this type

Digitized by Google
 202 A STUDY OF GLOBAL SAND SEAS

CONTORTED BEDDING IN LARGE-SCALE CROSS-STRATA of

Coconino Sandstone about 2 miles (3.2 km) northeast of Doney
Crater, near Flagstaff, Arizona, U.S.A. (Fig. 146.)

of fold has been developed under both conditions

(McKee. Douglass. and Rittenhouse. 1971. p. 372).

Tracks and Trails

The fauna of the Coconino Sandstone consists
exclusively of the "footprints of vertebrate animals
and the trails of worms and other invertebrates"
(McKee. 1934b. p. 106). The invertebrates were not
classified by Gilmore (1926. p. 33). but they were
compared by Brady (1939. p. 32) with modern
arthropods. All the vertebrate tracks were formed
FOOTPRINTS OF VERTEBRATE animal, probably reptile,
by quadrupeds (fig. 147); however. they vary
preserved on dip slope (about 30) of crossbedded Coconino
widely in size and character. Fifteen genera and 22 Sandstone. Tracks point uphill. Grand Canyon National Park,
species were named and described by Gilmore Arizona, U.S.A. Photograph by U.S. National Park Service. About
(1926. 1927. 1928). one-fourth natural size. (Fig. 147.)
No skeletal parts of the vertebrate animals
responsible for Coconino footprints have yet been
found. According to Schuchert (1918. p. 350). this An additional attribute of the Coconino
lack of organic remains is "probably due to the footprints. important to their interpretation. is the
originally loose and repeatedly reworked sands. a fact that these trackways "with only three excep-
most unfavorable habitat for animals." tions out of many hundreds examined. were all
The tracks have sharp definition. even margins. leading uphill" (McKee. 1944. p. 62). Explanation of
and general excellence of preservation. The signifi- this distinctive feature was obtained through a
cance of these features in interpreting the environ- series of controlled experiments made in 1943 by
ment of deposition was clearly stated by Matthes the writer (McKee. 1944. 1947). On loose sand
(in McKee. 1934b. p. 99). as follows: "the footprints derived from the Coconino and deposited at high
have not the soft. vague outlines which tracks made angles (25 - 28) comparable to that in crossbed-
in loose sand under water commonly have * * * ding of the Coconino. a chuckwalla lizard was en-
even the deep impressions made by the claws of the ticed to walk both up and down the surfaces of dry
animals being well preserved." loose sand. damp (sprayed) sand with a crust. and

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 203

UPHILL TRACKS OF CHUCKWALLA LIZARD made on foreset slopes at 28 of fine dune sand. A, Dry sand. 8, Damp sand with
crust. C, Wet sand. Experiments made in laboratory, Museum of Northern Arizona. Flagstaff, Arizona, U.S.A. (Fig. 148.)

dunes. Structures are typical of barchan or

wet sand (fig. 148). Only in dry sand. uphill. could
barchanoid-ridge and transverse types. In some
footprints be formed similar in size and clarity to
those of the Coconino. Damp and wet sands gave places. the curving shapes of small barchan types
very different results. and tracks made downhill are recognizable. but mostly the cross-stratification
were consistently destroyed by avalanching. is on a very large scale with individual lamination
Experiments made with modem millipedes lend planes extending downdip for 50-150 feet (15 - 30
m) or more. The planes have sweeping curves in
further support to the belief that the Coconino was
deposited as dry sand (McKee. 1944. p. 70) . The their lower parts to form tangents with the underly-
trails formed by these invertebrates consist of ing bounding surfaces (figs. 141. 142). They clearly
either two or three small furrows between low are the result of unidirectional winds. as shown by
crossbedding vector analyses (fig. 139).
parallel ridges when made on low-angle slopes (less
than 12) of dry sand (fig. 149). The uphill trails Although most of the minor structures and all
contained two furrows. and the downhill. three. the animal tracks and trails probably were formed
These animals were too light to leave trails in wet
in dry sand. their permanent preservation required.
or damp sand and were unable to effectively climb first. prompt surface dampening by dew or other
slopes steeper than about 20. The similarity is moisture. Dune surfaces with sand crusts ap-
strong between millipede trails formed under ex- parently formed in this manner are recorded from
perimental conditions and invertebrate markings inmany desert areas of today and explain cohesion of
certain low-angle beds of the Coconino. surface grains that preserve surface structures
(McKee, 1934b, p. 100; Glennie, 1972, p. 1049).
Environment of Deposition
Selective preservation of lee-slope deposits is a
The Coconino Sandstone is believed to have second controlling factor in the protection of sur-
been deposited largely as dry sand in the form of face features in dunes. Mostly, the gentle wind-

Digitized by Google
204 A STUDY OF GLOBAL SAND SEAS

TRAILS OF MILLIPEDES. A, In Coconino Sandstone, low-angle foresets near base, Grand Canyon National Park. 8, In
modem dry dune sand on 18 0 slope. Trails made in experimental trough at Museum of Northern Arizona, Flagstaff,
Arizona, U.S.A. Scale in inches; 1 inch equals 2.54 cm. (Fig. 149.)

ward slopes of dunes are continually eroded. but 4. Ripple marks which are common on surfaces of
downwind sides are constantly being buried by high-angle crossbedding suggest eolian deposi-
avalanching sand and protected. Thus. ripple marks tion both by their high indexes (above 15) and
and other structures upwind are seldom preserved. by their orientation with axes parallel to dip
yet the relatively rare ripple marks on the slipfaces slopes.
and the tracks of animals on such slopes have. if 5. The local preservation of a distinctive type of
buried promptly. a good chance of permanent rain pit. Such pits illustrate the cohesion of
preservation. sand grains with added moisture and a reorien-
tation of the crater axes with respect to bed-
Summary of Criteria for Eolian Origin ding slopes.
6. Successions of miniature rises or steps ascending
The basis for considering the Coconino
dip slopes of crossbeds.
Sandstone to be of eolian origin involves numerous
7. The preservation in fine sand of reptile
criteria. some of which are distinctive of an eolian
footprints and probable millipede trails with
environment and others merely compatible with
sharp definition and clear impression.
but not diagnostic of it. No single type of evidence
8. The consistent orientation of reptilian tracks up
seems entirely conclusive. but. together. the
various features present very strong evidence. The (not down) the steep foreset slopes.
principal criteria of dune deposition are as follows:
1. The extent and homogeneity of the sand body. De Chelly Sandstone
2. The tabular-planar and wedge-planar type and
large scale of cross-stratification. The common
(Permian), U.S.A.
high-angle deposits are interpreted as slipfaces THE DE CHELLY SANDSTONE of northeastern
on the lee sides of dunes. and the relatively Arizona is generally considered to be the lateral
rare low-angle cross-strata that dip toward the equivalent of the Coconino Sandstone to the west
opposite quadrant apparently represent (McKee. 1934a. p. 231; Reiche. 1938. p. 924; Peirce.
deposits of windward slopes. 1963. p. 137). although Baars (1962) showed it as
3. Slump marks of several varieties preserved on older in his correlations. Like the Coconino. the De
the steeply dipping surfaces of lee-side Chelly is believed to be an eolian deposit. at least in
deposits. These are distinctive of dry sand part. but is much more complex and includes
avalanching. several distinctive non eolian members and facies.

Digitized by Google
 ANCIENT SANDSTONFS CONSmERED TO BE EOUAN 205
Contrasts in sedimentary structures, grain-size dis- Peirce (1963, p. 59), this facies may be represented
tribution, and composition distinguish these other in neighboring localities, such as Nazlini and Bonito
facies and indicate various noneolian depositional Canyons, although largely lacking the charac-
environments. teristic sets of horizontal beds. If all areas men-
Although the contrasting facies just referred to tioned are included in this facies, it has a north-
have been recognized as discrete parts of the De south extent of 60 miles (95 km) and definitely
Chelly by most geologists (among them, Read, 1951; merges into normal Coconino-like deposits across a
Allen and Balk, 1954; Peirce, 1963) who have wide belt.
studied them, the following discussion concentrates Composition
on the two facies that are considered to be eolian.
Facies that occupy the eastern margin of the forma- Compositionally, the De Chelly Sandstone
ranges from nearly pure quartz sandstone or ortho-
tion almost certainly represent various subaqueous
quartzite to beds containing enough feldspar to be
environments. Farther west in the type locality at
classified as subarkoses. Specks of what seem to be
Canyon De Chelly, most of the sandstone consists
kaolin are common throughout parts of the forma-
of alternating units believed to be dune and inter-
tion and may have been developed by the break-
dune deposits. To the south and the northwest,
down of feldspars. They are characteristic of the
pure sandstone considered to be eolian closely
crossbedded sandstones in Canyon De Chelly and
resembles the Coconino and may indeed be an
canyons to the south (McKee, 1934a, p. 224), and in
eastern extension of the main body of that forma-
tion. subsurface samples from the Four Corners area to
the north (McKee, Oriel, and others, 1967).
Distribution and Thickness Most of the quartz grains are white, but sparse
Rocks commonly assigned to the De Chelly red quartz grains characteristically are scattered
Sandstone form a relatively narrow belt extending through much of the formation (Gregory, 1917, p.
about 125 miles (200 km) from north to south in 33; McKee, 1934a, p. 225). They probably result
northeastern Arizona. The formation is about 800 from ferric material that coats the grains or fills sur-
feet (240 m) thick at its type locality at Canyon De face irregularities. Thin-section study revealed that
Chelly but is progressively thinner both north and they have developed partly before and partly after
south. At Bonito Canyon, Hunter's Point, and Oak quartz overgrowths (Peirce, 1963, p. 115).
Springs Cliffs, between 25 and 40 miles (40 and 65 In the thin flat-bedded silty sandstones of proba-
km) to the southeast, its thickness ranges from 640 bly subaqueous or interdune origin, mica is
feet (200 m) to less than 500 feet (150 m).1t thins to widespread, but it seems to be rare or absent in the
approximately 250 feet (75 m) at the south end of large-scale cross-stratified sandstones. Heavy
the Defiance Plateau (Peirce, 1963, p. 16). North- minerals are rare in all these strata.
ward from Canyon De Chelly, it terminates in the Texture
subsurface, probably somewhere near the Arizona-
Utah border. Analyses of 125 samples showed the De Chelly
(Peirce, 1963, p. 93) to consist principally of fine
The eastern margin of the De Chelly is obscure sand with a modal size between 0.125 and 0.177
both because of facies complications and because of mm, a median diameter of 0.164 mm, and a sorting
the lack of surface exposures. In a westward direc- coefficient (Trask) of 1.25 mm. Therefore, the for-
tion, the De Chelly is represented by a facies with mation as a whole is considered to be a rather
large-scale cross-strata essentially identical to those uniform "well-sorted fine-grained sandstone."
of the Coconino. Although the size analyses of the De Chelly, as
A facies consisting of alternating sets of horizon- averaged for all its sandstone facies, reveal a
tal beds and of large-scale cross-strata is, so far as is general overall similarity, perhaps more significant
known, limited to the upper, or major, cliff member are comparisons made by Peirce (1963, p. 94) of
(570 ft (175 m) thick) of the De Chelly in Canyon grains from each of the various structural types
De Chelly and its environs. The area of outcrop is (facies) of sandstone. In the upper, or main,
about 20 miles (30 km) square, but, as suggested by sandstone member in Canyon De Chelly, sand

Digitized by Google
206 A STUDY OF GLOBAL SAND SEAS

grains of the large-scale cross-strata that are sent interdune deposits. The large-scale cross-strata
believed to be dune sands "average 0.170 mm in constitute a major part of the total deposit and oc-
median diameter. [which is] slightly coarser than cur in thick layers. in terms of feet or metres.
those of the formation as a whole" (Peirce. 1963. whereas the horizontal strata form thin sets. a few
fig. 8. D). Horizontally stratified silty sandstones. of which are more than 3 feet (more than 1 m) thick
here interpreted as interdune deposits. that form and which laterally may be represented by bedding
thin zones repeated throughout this member. have (parting) planes. These units of alternating
a wider range in grain size. with more coarse grains. crossbeds and horizontal beds were noted by
as well as a higher silt-clay content (Peirce. 1963. McKee (1934a. p. 222. 224) and described in detail
figs. lOB. 10E). by Peirce (1963. p. 51--67).
A comparison of sand from the large-scale The large-scale cross-stratified sandstones that
(eolian) cross-strata of the upper member at Can- form the main cliff of De Chelly Sandstone in Can-
yon De Chelly with sand from the corresponding yon De Chelly. similar in structure to the Coconino
unit at six localities to the south and southeast Sandstone and to the Coconino-like sandstones
shows similar size properties in all sections (Peirce. near Ganado to the south and in Monument Valley
1963. table 2). as follows: to the northwest. have been considered extensions
Coarse sand: Very little or none. (facies) of the De Chelly (Gregory. 1917; Baker.
Silt and clay: Less than 10 percent. 1936). The large-scale cross-stratified sandstones
Median grain form two principal structural types - tabular
size (mm): 0.149- 0.186. planar and wedge planar. In the tabular-planar
Sorting index (Trask): 1.19-1.32. type. bounding planes both above and below the
foresets are nearly horizontal and truncate the un-
Comparing the eolian-type sandstone of the up-
derlying crossbeds (fig. 150).
per member with sandstones that Peirce (1963. ta-
ble 2) considered of aqueous origin in the lower In the wedge-planar type. oblique erosion sur-
member at Bonito Canyon to the southeast discloses faces separate sets of cross-strata. The tabular-
the contrast shown in the following unnumbered planar type is interpreted as representing an inter-
table. dune surface on which no deposits accumulated
(McKee and Moiola. 1975); the wedge-planar type
Coarse Median Sorting is developed within a migrating dune of barchanoid
sand Silt-clay grain size index
(percent) (percent) (nun) (Trask) or transverse type and is common in the dunes at
Canyon De White Sands National Monument. New Mexico
Chelly ...... Trace 5 0.170 1.21
Bonito Canyon 0.4 9 .126 1.50 (McKee. 1966. p. 59).
Sand grains in the De Chelly. regardless of facies. The thin horizontal silty sandstone beds that
are nearly all rounded or subrounded. Commonly. characteristically alternate with large cross-strata
they are pitted or grooved. and many are frosted. rest on flat erosion surfaces that truncate cross-
This frosting. however. is caused by "the light scat- strata sets (fig. 151). Followed laterally. some of the
tering effect of myriads of minute quartz crystals. horizontal beds are seen to thin and wedge out. but
or incipient silicification" (Peirce. 1963. p. 98). the underlying erosion surface continues as a
bounding plane between sets of cross-strata above
and below. A detailed description of these struc-
Primary Structures tures (Peirce. 1963. p. 56- 59) suggests that
Stratification although the flat-bedded silty deposits occupy only
The upper unit or member of the De Chelly restricted areas on truncation planes. they occur as
Sandstone in Canyon De Chelly contains two alter- sedimentary structural features throughout an area
nating and contrasting types of stratification that of many square miles.
together distinguish it from all other facies in the The distribution. composition. and structure of
formation. These cross-strata consist of (1) a large- the thin lensing flat beds that occur between large-
scale high-angle type. interpreted as eolian. and (2) scale cross strata. strongly resemble the interdune
a thin. virtually horizontal type believed to repre- deposits now developing in the dune field at White

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 207

LARGE-SCALE AND MEDIUM-SCALE CROSS-STRATA of tabular-planar type in Permian De Chelly Sandstone, Canyon
De Chelly, Arizona, U.S.A (Fig. 150.)

Sands National Monument, New Mexico (McKee them constitutes an eastern extension of the
and Moiola, 1975). If the White Sands analog is Coconino or at least a recurrence of its eolian facies.
valid, the thickness of a flat-bedded deposit (inter-
dune) may be considered a function of the time of Tracks and Trails
exposure before the next dune migrates over and
Footprints of several species of vertebrate
buries it. Likewise, the horizontal bedding planes
animals and trails of invertebrates, similar to forms
lacking deposits are the record of surfaces that were
represented in the Coconino Sandstone of Grand
buried rapidly by advancing dunes.
Canyon, occur in the large-scale cross-stratified
sandstone member of the De Chelly at Nazlini Can-
Ripple Marks yon (McKee, 1934a, p. 227) and at Kinlichee
Ripple marks in several parts of the Monument (McKee, 1934a, p. 232) south of Canyon De Chelly.
Valley area are preserved on cross-stratified They are also reported from western Monument
sandstones assigned to the De Chelly (Peirce, 1963, Valley in crossbedded sandstone assigned to the De
pIs. 25, 26). They are described as "having axes Chelly (Peirce, 1963, p. 26). All these tracks and
aligned up and down dip" and are therefore con- trails lead up the steep dip slopes of the cross-strata,
sidered to be wind-formed. They resemble closely as do those in the Grand Canyon.
the abundant ripple marks described from the The significance of these footprints in the history
Coconino Sandstone of Grand Canyon and give of the sandstone is threefold. First, they suggest a
further indication that the sandstone containing direct correlation between the Coconino Sandstone

Digitized by Google
 208 A STUDY OF GLOBAL SAND SEAS

HORIZONTALLY BEDDED SILTY SANDSTONE, interdune deposit (arrow) between cross-strata sets of eolian dune
type. Permian De Chelly Sandstone, Canyon De Chelly, Arizona, U.S.A (Fig. 151.)

and the De Chelly Sandstone and furnish evidence in that direction. Between cross-bedding sets, in
of similar environments of deposition. Second, the many places, are thin sets of horizontal beds
sharpness of the imprints and the general manner believed to have developed as interdune deposits.
of preservation indicate that the imprints were The other two members in Canyon De Chelly are
formed in dry sand. Third, their preferred orienta- probably of subaqueous origin. The thin (45 ft,
tion, up the bedding slope only (never down), is 14m) middle member, composed of micaceous
evidence of having been formed in dune sand near siltstone and silty sandstone, is considered to repre-
the angle of repose. (See the preceding discussion sent a relatively low-energy deposit and may have
of the Coconino Sandstone.) formed on the margin of saline basins to the south.
The lower member, formed of cross-stratified sand,
Environment of Deposition is believed by Peirce (1963, p. 145) to be "a high
energy fluvial" deposit accumulated on "an aggrad-
Of the three members that form the De Chelly ing delta."
Sandstone in its type locality, only the upper is con-
sidered to be dominantly of eolian origin. It forms Two additional members (or facies) in the De
the major cliff in the walls of Canyon De Chelly Chelly Sandstone occur in the eastern part of the
where it is nearly 600 feet (180 m) thick. The large- Defiance uplift (Peirce, 1963). These members un-
scale cross-strata in this unit show a persistent derlie the cliff-forming upper member and are con-
southwestward dip and are interpreted as indicat- sidered by Peirce (1963, p 148-149) to be low-
ing a prevailing wind that caused dunes to migrate energy lagoon or tidal flat deposits overlying "near-

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 209

shore marine deposits," together with some cross- in a regional environment of aridity. In brief, con-
stratified sandstone, possibly eolian. siderable evidence is available to suggest that the
The foregoing interpretations suggest that in the Navajo represents an extensive sand sea of a great
western part of the region the De Chelly Sandstone interior desert, mostly or entirely Lower Jurassic
represents stream deposits followed by extensive (Lias) time.
dune deposits but that, in a parallel belt to the east, Obviously, making generalizations concerning
near-shore marine deposits preceded the develop- the character and genesis of this huge sand body is
ment of eolian strata. difficult. Observations made in one area may not
apply entirely to another area, yet many of the dis-
Summary of Criteria for Eolian Origin tinctive features are common to most parts. The
local exceptions, doubtless related to subenviron-
Those facies of the De Chelly Sandstone believea
ments, have prompted 'Various different interpreta-
to be composed of windblown deposits are of two
tions. The really remarkable feature of the Navajo
principal types. One facies resembles the Coconino
is the high degree of uniformity of the sandstone
in its uniformity of grain size and structure and in
and its general consistency throughout the region.
large-scale high-angle cross-stratification. Charac-
teristics of the cross-strata indicate deposition as
barchans or barchanoid ridges. The orientation of Distribution and Thickness
ripple marks and the state of preservation of rep The Navajo Sandstone as currently recognized
tilian tracks are further evidence of eolian origin; (McKee, Oriel, Swanson, and others, 1956, table 1)
these features are common to both the Coconino extends across much of Utah, northeastern
and the western part of the De Chelly. Arizona, and southwestern Colorado. It is correl-
The second facies of probable eolian origin in ated with, and almost certainly is a continuation of,
the De Chelly consists of an alternation of (1) the Aztec Sandstone (Triassic? and Jurassic) to the
large-scale high-angle cross-strata forming tabular- west in southern Nevada and California, and of the
planar and wedge-planar sets and (2) very thin Nugget Sandstone (Triassic? and Jurassic?) to the
sets of horizontally bedded silty sandstone. From north and east in northern Utah, southwestern
the standpoint of both texture and structure, these Wyoming, southeastern Idaho, and northwestern
two types of sandstone compare well with the Colorado (Baker and others, 1936). As shown by
alternating dune and interdune deposits of a isopachs in figure 152, this great cross-stratified
modern dune field developed under conditions of a sandstone body extends approximately 600 miles
prevailing wind. (965 km) from north to south and at least 250 miles
(400 km) from east to west. These measurements
Navajo Sandstone (Triassic' approximate the present extent of the sandstone;
and Jurassic), U.S.A. the original sand sea doubtless extended much
farther but has been removed by Cenozoic erosion,
THE NAVAJO SANDSTONE has been selected as in northwestern Arizona and eastern Nevada.
to illustrate eolian-type sandstone because of its The Navajo Sandstone is more than 1,000 feet
broad geographic extent and great thickness and (300 m) thick in parts of northeastern Arizona and
because of the large-scale and excellent develop- more than 2,000 feet (600 m) thick in southwestern
ment of its primary structures, permitting it to be Utah (fig. 152); its correlative, the Aztec Sandstone,
analyzed across a wide area and observed in rela- is 3,000 feet (900 m) thick in the Mohave Desert of
tion to other facies. Preservation in the Navajo California. Along its eastern and southeastern
Sandstone of small dinosaurs and other forms of margins in Colorado and northern Arizona the
continental life makes it unlikely that this sand- Navajo forms a wedge and thins to a vanishing
stone represents accumulations of sand waves in point; elsewhere its margins are less definite.
deep marine waters of the North Sea type, at least Along the southern margins of the Navajo
in areas where those animals are present. Various Sandstone, prominent tongues of large-scale cross-
features of associated facies indicate development stratified sandstone extend southward and south-

Digitized by Google
210 A STUDY OF GLOBAL SAND SEAS

116" 114" 112" 1100 108"

44

IDAHO

42"

: -;'~7
. 2000
" :.
'.::

NEW MEXICO

A R z o N A

ISOPACH-LITHOFACIES MAP AND CROSS-SECTION showing distribution and thickness of Triassic(?) and Jurassic Navajo
Sandstone, Triassic(?) Kayenta Formation, and correlative units. (Modified from McKee, Oriel, and others, 1956, pI. 4). 100 feet
equals 30.5 metres. (Fig. 152.)

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 211

westward into the dominantly flat-bedded sand-

stones, siltstones, and limestones of the Kayenta
Formation (Harshbarger and others, 1957, p. 22; R.
Navejo, Kayenta, and correlative ..ndetone. F. Wilson, 1958, p.116). Some of these tongues have
DuM. indlt:ete .,... of 60",. , " . , . appreciable thickness and have been named by
workers in various parts of the region - for exam-
ple, the Triassic (1) Lamb Point (R. F. Wilson, 1958)
and Shurtz Sandstone Tongues of the Navajo

.-
Area where Navejo, Kayenta, and correlative ..ndltOnel ab-
Ient within the region of their distribution (Averitt and others, 1955). A correlation of named
tongues of the Navajo is included in table 1 of
Hatchell (1967, p. 25).
Outcrop or well where Navajo, Kayenta, and correlative
..ndltone. are pre lent The Kayenta Formation consists of sedimentary
Thit:Ilnea, In feet rocks believed to be fluvial and probably represents
the deposits of southwestward-flowing streams.
I:> Thus, the tongues of the Navajo are interpreted
Outcrop or well where Navajo, Kayenta, and correlative (Harshbarger and others, 1957, p. 1) as windblown
..ndltone. are abient
Older roca .'" wain by younger Ju","it;. CreteceGu" or (Iocelly) sands from the north or northwest that "gradually
Tertiery roca overwhelmed" the streambeds.
---500---
I.opech in area. of good control Composition and Texture
Thit:laltla, in feet
Studies in several widely separated areas indi-
-----500-'-- cate that nearly everywhere the Navajo Sandstone
I.opech in area. of poor control
Thit:Jrnflsl, in feet; queried whetfI NPfIC/IIIIy questioneble. consists largely of fine-grained quartz sand, with
Dotted wile", inferred
very minor amounts of feldspar and a few other ac-
-'---500-'-'-'- cessory minerals. It apparently is homogeneous and
I.opach .howing inferred former thickne of Navajo, well sorted throughout, and in central Utah the
Kayenta, and correlative ..ndltone. where they are known Navajo is described as "monotonously uniform" in
to be eroded
Thicltn.", In feet texture (Kiersch, 1950, p. 927). The quartz particles
have been reported as medium- to fine-grained
with an average diameter of 0.21 mm in north-
eastern Arizona (Harshbarger and others, 1957, p.
~ 19) but as fine-grained to very fine grained near
Ii Navajo Mountain on the Utah -Arizona border
~ (Hatchell, 1967, p. 30). Sand grains are uniformly
18 about 0.2 mm in diameter (fine grained) in the
2A _ northern part of the San Rafael Swell area, Utah
Croubedded
_
.. ' .... ...;,..; .. ;; (Kiersch, 1950, p. 927) .
In the Navajo Sandstone, most quartz grains are
~ rounded or subrounded, and relatively few parti-
cles are subangular or angular (Kiersch, 1950,
HorimnteIIC8Ie x 1.5 p.927; Harshbarger and others, 1957, p. 22;
VertlCllllClle eoggerlted
~ Hatchell, 1967, p. 30). Among samples from the
o 100 MlES northeastern Arizona area examined by
I ! I I I Harshbarger, Repenning, and Irwin, about 25 per-
o 200 KI.OMETRfS cent were determined to be rounded. Frosting has
SECTION OF NAVAJO, KAYENTA, AND CORRELATIVE been noted on quartz grains in many areas, and
SANDSTONES ALONG LINE A-A'
angular overgrowths of silica, forming ortho-
(Fig. 152. - Continued.) quartzite, are reported (Kiersch, 1950, p. 927). Ce-
ment is mainly silica but locally iron oxide or
calcite (Hatchell, 1967, p. 30; Kiersch, 1950, p. 927).

Digitized by Google
212 A STUDY OF GLOBAL SAND SEAS

Horizontally interbedded units locally among the 1967. p. 30). The prevalence of high-angle dips.
m888ive cosets of sandstone throughout the region greater than 20. in most areas. is noted by many
are thin layers and lenses of aphanitic limestone geologists. among them R. F. Wilson. (1958. p. 148.
and siltstone (Baker and others. 1936. p. 52; table 6).
Harshbarger and others. 1957. p. 19). that probably Ripple Marks
represent interdune deposits. Wind-faceted peb- Ripple marks. though uncommon in the Navajo
bles. ranging in diameter up to more than 2 inches
Sandstone. occur locally in widely scattered parts
(5 cm). locally occur between units of cross-
of the formation. They have been recorded from the
stratified sandstone in the San Rafael Swell and
east-central Utah area by Kiersch (1950. p. 935) and
Green River Desert areas of east-central Utah from the Arizona-Utah border by Hatchell (1967.
(Baker and others. 1936. pis. 13. 14). Small pebbles
p. 31). both of whom considered them to be wind-
and coarse sand grains of quartz and chalcedony oc- formed types because of orientation with crests
cur along some planes of cross-strata in Buckhorn parallel to the dip slopes of crossbeds. Ripple marks
Wash in east-central Utah (Kiersch. 1950. p. 927). have also been reported from southwestern Utah
and cl888ified as eolian on the basis of ripple indices
Primary Structures (Tanner. 1964. p. 432).
Stratification
Primary structures of the Navajo Sandstone of
southern Utah consist "of large wedge-shaped
planar sets of large-scale. high-angle. tangential
cross-strata. The bounding surfaces of individual
sets of cross-strata are planar sets that dip gently in
a direction opposite to that of the steeply dipping
cross-strata" (R. F. Wilson. 1958. p. 112. 141. 145).
The formation contains both tabular-planar and
wedge-planar cross-strata. These structures are
characteristic of the formation in the Zion Canyon
area and in the Kanab area farther east (fig. 153).
In the Navajo of northeastern Arizona. a domi-
nance of large-scale wedge-planar and tabular-
planar cross-strata has been recorded both by
Harshbarger. Repenning. and Irwin (1957. p. 19)
and by Hatchell (1967. p. 30). Simple wedge struc-
tures (without beveled tops) have also been
recorded there by Hatchell and in east-central Utah
by Kiersch (1950. p. 930). An absence of trough-
type cross-strata was noted by R. F. Wilson (1958.
p.147) and by Hatchell (1967. p. 32). although. ac-
cording to Wilson. trough structures - some of
them large scale - are common in sandstones of
the underlying Kayenta Formation of probable
fluviatile origin.
In southwestern Utah much of the Navajo cross-
stratification is large scale. commonly exceeding 50
feet (15 m) in length and forming sweeping arcs.
concave upward. The thickness of sets in east-
central Utah is as much as 15 feet (4.5 m) (Kiersch. CROSS-STRATA IN NAVAJO SANDSTONE, north of Kanab,
1950. p. 930) and in the Navajo Mountain area of Utah, U.S.A. A, Wedge-planar type. 8, Tabular-planar type.
Utah and Arizona is up to 100 feet (33 m) (Hatchell. Photographs by E. Tad Nichols. (Fig. 153.)

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 213
The relative scarcity of ripple marks in the Nava- red to by most authors as slump structures. In east-
jo. as compared with their great abundance in most central Utah contorted zones have been described
modem dune fields. is probably a function of selec- by Kiersch (1950. p. 941) as 35-50 feet (10-15 m)
tive preservation. as in the Coconino (McKee. thick. occurring at many widely scattered localities
19Mb. p. 101). Only the relatively few ripple marks in the upper middle part of the formation. In south-
that form from crosswinds on the lee-side slipface western Utah. zones of contorted strata (R. F.
have a chance of burial and preservation. Further- Wilson. 1958. p. 142) range in thickness from a few
more. even though wind ripples form in dry sand. feet to 20 feet (6 m) and extend laterally as much as
permanent preservation requires some degree of 100 feet (30 m). In northeastern Arizona. they occur
cohesion. usually contributed by dew or light rain. a in foresets "in many places" (Harshbarger and
requirement further limiting the chances of ripple- others, 1957; Hatchell, 1967. p. 31).
mark survival (McKee. 1945. p. 324). Irregular folds, some of them recumbent, in the
contorted zones of the Navajo have generally been
Contorted Bedding attributed to oversteepening on the lee sides of
Zones of contorted bedding in which irregularly dunes or to the weight of overlying sediment (R. F.
folded strata occur between sets of undisturbed Wilson, 1958, p. 142). These processes can be dis-
strata are fairly numerous in various parts of the tinguished by the type of fold that has developed
Navajo Sandstone (fig. 154). They have been refer- (McKee. Reynolds. and Baker, 1962a); however, the
environmental implications which are recognized

CONTORTED STRUCTURES IN NAVAJO SANDSTONE, Kanab Creek area, Utah, U.S.A Photograph by E. Tad
Nichols, (Fig. 154.)

Digitized by Google
214 A STUDY OF GLOBAL SAND SEAS

by the degree of sharpness or "fuzziness" in result- terdune deposits: "sand-filled desiccation cracks in
ing fold structures are more important. These limestone," "a bed of nodular-weathering silty red
features are largely a function of the cohesiveness sandstone," "The occurrence of excellent drei-
of the sand at the time of deformation and serve to kanter," and "local thin lenses of dense un-
differentiate between saturated. wet. damp. and fossiliferous gray limestone from a few feet to
dry sand (McKee. Douglass. and Rittenhouse. 1971. several miles in diameter" interpreted as the pro-
table 1). Folds resembling many in the Navajo duct of "ephemeral water-filled basins" and
Sandstone are recorded in modem dune deposits. "formed through algal or purely chemical actions."
especially where the sand has been wetted (McKee The limestones contain mud cracks.
and Bigarella. 1972. p. 674).
Fauna
Interdune Deposits F088ils are uncommon in the Navajo Sandstone.
Although the term "interdune" has not been the total recorded forms consisting of two small
mentioned in any of the many publications dealing bipedal dinosaurs (Brady, 1935; Camp, 1936). some
with the Navajo Sandstone. probably many of the dinosaur tracks, a few ostracodes, and some other
references to thin horizontal beds. siltstone or crustaceans (Harshbarger and others, 1957, p. 1).
limestone lenses. and other distinctive deposits These animals probably inhabited interdune areas,
among the massive units of cross-stratified as evidenced by the dinosaur tracks near Navajo
sandstones actually refer to interdune deposits. If Canyon. Arizona (Gregory. 1917), that are pre-
so. then interdune deposits are well represented in served in thin lenses of limestone probably precipi-
the Navajo by a variety of sedimentary bodies hav- tated in ponds by chemical processes. The
ing in common thinness. limited lateral extent. and ostracodes likewise are from limestone lenses
either horizontal or uneven stratification. (Harshbarger and others. 1957) and presumably
In the Lamb Point Tongue of the Navajo in lived in shallow ephemeral water bodies between
central southern Utah are "scattered thin lenses of dunes.
flat-bedded to laminated, pale reddish brown Probably more important as evidence of the
siltstone and claystone" (R. F. Wilson. 1958. p. 141). sandstone genesis than the animals themselves is
The lenses are not more than 2 or 3 feet (0.5-1 m) their manner of preservation. For example. the
thick and they pinch out along strike. Farther west skeleton of the small dinosaur recorded by Camp
in Zion Canyon. Utah. R. F. Wilson (1958. p. 150) (1936. p. 39) from northern Arizona was buried on a
referred to Navajo cross-strata that alternate with surface parallel to the dipping "planes of cro88bed-
flat-lying beds or with beds that have low-angle ding of the Navajo sandstone at that point" with the
dips. The claystone and siltstone lenticular bodies axis of the body horizontal and the right hind foot
probably formed as interdunes. according to pressed into the sand. The creature seems to have
modern analogs (McKee and Moiola. 1975). been covered "by shifting sand before death. or
Locally. Navajo interbeds include extensive soon after" (Camp. 1936. p. 41).
lenses of thin-bedded limestone. presumably The uniform sand that constitutes the Navajo.
formed in ephemeral ponds. Among the sandstones Nugget. and Aztec Sandstones across a vast area is
of east-central Utah are "limy and argillaceous generally interpreted as an eolian accumulation
beds" (Kiersch. 1950. p. 927). In northeastern with a source to the north or northwest
Arizona, many lenticular beds of cherty limestone (Harshbarger and others. 1957. p. 3). Studies of
occur within the Navajo (Harshbarger and others. wind directions as determined from cross-strata
1957, p. 19). and thin discontinuous beds of gray dips in these sandstones have been made by
limestone. calcareous siltstone, sandstone. and numerous geologists in many parts of the region
limestone conglomerate form platy or flaggy units (for example. Poole. 1962. fig. 163.1). Cross-strata
between sandstones (Hatchell. 1967. p. 2). orientations. indicated by arrows on Poole's map.
Numerous distinctive structures and other show dominant north and northeast winds for the
features in the Navajo of eastern Utah probably are Nugget and Navajo Sandstones in northeastern
attributable (Baker and others. 1936. p. 12.52) to in- Utah and northwestern Colorado and for the Aztec

Digitized by Google
 ANCIENT SANDSTONFS CONSIDERED TO BE EOUAN 215
Sandstone in southern Nevada, but show northwest bly was beginning to transgress across it from the
winds for the Navajo Sandstone in most of north. This is indicated at the upper contact of the
southeastern Utah and adjacent Arizona. "The con- Navajo in east-central Utah, where inundation by
sistency of cross-strata dip bearings in a southerly the sea is evidenced by reworking of the sand. leav-
direction suggests [for that period] a broad belt of ing "swash marks, beach cusps and other water
strong and persistent northerly winds * * *." (Poole, marks" (Kiersch, 1950, p. 927). The gradational
1962, p. 0150). sediments locally at the Navajo-Carmel contact in-
The sources of the sand have been postulated by dicate that there was no appreciable hiatus but vir-
various geologists; but no one source sufficiently tually continuous, uninterrupted sedimentation.
great to have furnished the entire volume repre- Not all geologists agree with the concept of an
sented by these sandstones can readily be eolian origin of the Navajo Sandstone. Largely from
visualized. A major source in western Nevada examination of a thin stratigraphic section near
where mountain uplift probably occurred at that Woodside, Utah, and textural analyses of six sand
time is suggested by conglomerates in the adjacent samples, Visher (1971, p. 1423) postulated a marine
region (Harshbarger and others, 1957, p. 3). In high-energy channel sequence. From field mea-
northeastern Arizona, where sands of the Navajo surements and textural studies at seven sections in
apparently advanced eastward and eventually southern Utah, Marzolf (1989) postulated a
covered the fluvial Kayenta Formation, "Kayenta marginal marine environment, rather than an arid
streams [are believed to have been] the primary interior desert.
source of material for the Navajo dunes * * *."
(Harshbarger and others, 1957, p. 25). A source far Summary of Criteria for Eolian Origin
to the north, certainly for the Nugget sandstone,
also has been suggested. A synthesis of criteria for considering the Navajo
The depositional environment of the Navajo and Sandstone to be eolian was prepared 40 years ago
its correlative sandstones has generally been recog- by Baker, Dane, and Reeside (1936, p. 52) when
nized as a great interior desert basin that sank gra- they suggested five kinds of evidence - three posi-
dually as the several thousand feet of sand accumu- tive and two negative. Positive evidence consisted
lated. Aridity is evident from such features as the of (1) the abundance and "gigantic scale" of the
wind-faceted pebbles in east-central Utah and cross-strata, (2) the well-rounded grains, mostly
desiccation cracks in limestone beds of interdune well sorted, but with coarse grains scattered locally
deposits, and it can readily be inferred from the along bedding planes, and (3) the truncation of
great size and thickness of the sand sea. That some cross-strata sets by other sets, dipping "at all angles
dunes were very high is shown by the large scale of without system," that is, tangential foresets. Nega-
individual foresets and by the thickness of some tive evidence was (1) a lack of numerous regular
cross-strata sets. (horizontal) bedding planes and (2) the general
scarcity of silt, ripple marks, mudcracks, mud-
Dunes of the Navajo Sandstone are believed to
pellet conglomerates, and other evidences of water
have been transverse or barchanoid types, accord-
ing to analyses of the cross-strata (R. F. Wilson, action.
1958. p. 150) and the orientation and average dips of As indicated elsewhere in this chapter, textural
cross-strata. The cross-strata show a very small arc features and the various types of negative evidence,
of spread, especially when compared with the though compatible with a dune environment, can-
spread of underlying fluvial deposits (R. F. Wilson, not be considered incontrovertible proof. However,
1958, p. 148, table 6). Between sets of dune sands in the discovery of wind-faceted quartz pebbles in
many places are thin beds or lenses thought to be what may be considered interdune deposits of the
interdune deposits that consist of flat-bedded Navajo in San Rafael Swell and Green River Desert
siltstones, claystones, and limestones (R. F. Wilson, (Baker and others, 1936, pIs. 13, 14) probably con-
1958, p. 141) and probably made up the desert floor. stitutes important additional evidence. Still other
While the Navajo Sandstone was forming in the evidence has been accumulated in support of an
southern part of the region, the Carmel Sea proba- eolian genesis during later years.

Digitized by Google
216 A STUDY OF GLOBAL SAND SEAS

Detailed analysis by R. F. Wilson (1958. p.143) of 3. High-angle (>20) dips of foresets.

cross-strata in both the Navajo Sandstone and the 4. Cross-strata dip directions with consistent orien-
Lamb Point Tongue of the Navajo in southern Utah tation and narrow range of vectors.
clarified several points concerning their structure. 5. Cross-strata largely tangential and concave up-
After comparison with analogs in modem barchan- ward.
and transverse-type dunes. the following features The numerous zones of penecontemporaneous
seem to be common to both modem and ancient: contorted bedding with sharp outlines and irregular
1. Type of cross-strata -largely tabular and wedge fold patterns in the Navajo are considered evidence
planar. no large trough types (fig. 155). of sand that has become wet but not saturated. and
2. Numerous large-scale foresets. 50 feet or more those zones are comparable to structures in some
(15 m) long. modem dune fields.

NAVAJO SANDSTONE, showing large-scale wedge-planar cross-strata (middle) between tabular-planar cross-strata (above and
below), Zion National Park, Utah, U.S.A Photograph by E. Tad Nichols. (Fig. 155.)

Digitized by Google
 ANCIENT SANDSTONFS CONSIDERED TO BE EOUAN 217
The vast extent and considerable thickness of Distribution and Thickness
well sorted fine sand with uniform character
throughout rarely occur except in eolian environ- The Lyons Sandstone forms a narrow foothills
ments. A possible exception is in areas of large sub- belt. extending about 160 miles (250 km) from its
marine sand waves that are said Oordan, 1962, northernmost outcrop near the Wyoming - Col-
p. 840) to resemble desert dunes in general form; orado line to its southernmost outcrop. south of
however, nothing is yet known of their internal Colorado Springs (Hubert. 1960. p. 35). Colorado. Its
structure. In any event, the association of articul- thickness. as determined in a series of measured
ated skeletons of terrestrial dinosaurs in some parts sections along the outcrop belt between Boulder
of the Navajo precludes the possibility of those and Colorado Springs. ranges from 150 to 680 feet
parts having formed under marine conditions. (45- 200 m) (Hubert. 1960. p. 25). At its north end.
Eolian-type ripple marks in the Navajo seem the formation wedges out between tongues of the
good evidence of a dune environment. Numerous Permian Satanka Formation. To the south. it
lenses and thin beds of limestone and siltstone be- wedges out between red Paleozoic sandstones and
tween sets of cross-strata in the Navajo are believed siltstones (W. O. Thompson. 1949).
to represent deposits of interdunes. Eastward in the subsurface of eastern Colorado.
the Lyons is represented by various facies. probably
Lyons Sandstone (Permian), marine. of sandstone. siltstone and evaporites. as
shown on subsurface maps by Maher (1954) and
U.S.A. Broin (1957). East of Colorado Springs. the Lyons
THE LYONS SANDSTONE of Permian age in seems to maintain its identity for nearly 40 miles
Colorado was selected for description in this (65 km) before grading into other facies (Maher.
chapter for two principal reasons. First, outcrops at 1954).
its type locality near Lyons, Colorado, offer ex- The following discussion of the Lyons Sandstone
cellent opportunities for detailed examination and concerns mainly that part which consists of large-
careful analysis (Walker and Harms, 1972). Second, scale high-angle cross-strata. believed to be of
it apparently represents deposits of a coastal dune eolian origin. at the type locality and in a few other
area in relation to other facies for which considera- limited areas along the outcrop belt. Intimately re-
ble environmental data are available. lated noneoUan facies of the formation are of in-
In many respects, the Lyons Sandstone closely terest primarily because they furnished supporting
resembles the Coconino Sandstone of Arizona and evidence concerning the depositional environment
many of the features considered to be evidence of and paleogeographic setting. The eolian-type
the Coconino's eolian origin also occur in the deposits in this formation are only known over a
Lyons. The nature of cross-strata. the uniformity of relatively few miles along the north-south belt and
fine sand. the presence of high-index ripple marks. are very different in extent from other wind-
rain pits. and distinctive kinds of slump marks. and formed sandstones. such as the Coconino and the
the presence of nearly identical tracks of vertebrate Navajo.
animals show a generally similar depositional en-
vironment for both formations. In certain other Composition and Texture
features. the two differ markedly - the size and Along its outcrop belt the Lyons Sandstone con-
geometry of the entire cross-stratified sand body. sists of two principal rock types - "arkoses (24
the character of erosion planes considered to be in- percent) and feldspathic quartzites (75 percent)"
terdunal surfaces. and the abundance and distribu- (Hubert. 1960. p. 13. 1(0). The arkoses are mainly
tion pattern of relatively coarse sand concentrates. conglomerates and pebbly sandstones; intraforma-
The question presented here is not whether the tional conglomerates. lenticular dolomitic quartz-
Lyons Sandstone of the type locality is eolian - ites. and micaceous siltstone and shales constitute
this seems to be established on rather firm grounds about 1 percent. The arkoses and many of the
- rather. what type of dune is represented and feldspathic quartzites indicate by their textures.
why was its depositional environment somewhat structures. and other features noneolian deposi-
different from that of the Coconino. tional environments. The eolian facies consist of a

Digitized by Google
218 A STUDY OF GLOBAL SAND SEAS

relatively uniform fine-grained feldspathic sand- The presence of relatively small subsets of cross-
stone which has conveniently been delimited by strata within the major sets of high-angle cross-
Walker and Harms (1972. p. 279) as the "flagstone strata is indicated by the traces of laminae exposed
beds." on the truncated surfaces that form bounding
Texturally. sandstones of the high-angle large- planes between sets. The nature of these subsets
scale cross-strata in the Lyons Sandstone. as repre- was recorded and analyzed in detail by Walker and
sented at the type locality. are characteristically Harms (1972. p. 281). who believed them to repre-
fine grained and well sorted. Pebbles. granules and sent a record of nonuniform accretion along a
beds of fine detritus. such as silt and clay, are ab- planar surface resulting from wind drift. Such
sent (Walker and Harms. 1972. p. 284). The grains lateral shifting of cross-strata to form subsets ap-
include both quartz and feldspar which mostly are parently is common in many modem dunes of
tightly cemented by quartz; much of the feldspar various types. and can be seen both on the floors of
has been altered by replacement. dune trenches and on interdune surfaces before
burial (McKee, 1966. pI. VIlA). Shifts in wind direc-
Sand grains of medium or coarse size are con- tion also result in wedge-planar sets in many dunes.
centrated on erosion surfaces that truncate sets of (See chapter E. "Sedimentary Structures in
cross-strata throughout the formation. Relatively Dunes.")
coarse grains uniformly scattered over these sur- Cross-strata in the Lyons. as described. resemble
faces are interpreted as lag deposits of interdunes structures in many modem dunes but are dupli-
(Walker and Harms. 1972. p. 284). Such lag deposits cated in relatively few other environments. The
occur on serirs and interdune surfaces of various maximum dip angles of the foresets are commonly
modem dune fields (McKee and Tibbitts. 1964. pI. 4- 5 less than the angles of repose or maximum
IB). dips (33- 34) of many modem dunes. In ancient
The relatively clay-free clean sand that com- dune deposits. a maximum angle of dip consistently
poses the dune-type sandstone of the Lyons has lower than in modem dry sand has been attributed
been called variously a first-generation and a sec- to compaction (Glennie. 1972. p. 1061; Walker and
ond-generation sand. Its relative purity has been at- Harms. 1972. p. 280) and to removal of unstable
tributed to the abrasion and winnowing of arkosic grains by intrastratal alteration (Walker and Harms.
material shed from the ancestral Front Range high- 1972. p. 280). An alternative explanation is
land (Maher. 1954. p. 2237). and to accumulation differential preservation in cross-strata sets. where
along a "Permian strand by waves and currents that the upper parts of foresets that are relatively higher
carried in clean sands probably derived from pre- angled commonly are removed by erosion and
viously existing sediments" (W. O. Thompson. planation (Poole. 1962. p. Dl48; Walker and Harms.
1954). 1972. p. 280).

Primary Structures Miscellaneous

Stratification Several minor structures occur in many places in
the ubiquitous cross-stratification of the eolian-
The Lyons Sandstone. as exposed at the type type facies of the Lyons Sandstone (Walker and
locality near Lyons. Colorado. is characterized by Harms. 1972. p. 279). These structures include
large-scale high-angle cross-strata. These cross- asymmetrical ripple marks invariably oriented up
strata are mostly of the tabular-planar type with in- and down the dip surfaces of foresets and with very
dividual sets as much as 42 feet (13 m) thick and low crests that are parallel and straight for long dis-
with bounding planes described as "scoop-like ero- tances. These ripple marks have the high indices of
sional surfaces" (Walker and Harms. 1972. p. 279. wind ripples. Other structures are raindrop pits
281). Many sets of strata extend laterally "for with typical ring shapes and raised rims (fig. 158)
hundreds of yards" with little change in character; and slump marks (called avalanche structures by
laminae within them are straight or gently curved Walker and Harms). including the distinctive "step
in cross section. type" that is characteristic of many slipface

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 219
1. Experiments with living reptiles on steep slopes
of fine sand indicate that such tracks could
readily be formed in dry sand but not in wet or
damp sand (McKee. 1944. p. 68; 1947. p. 27).
2. Experiments on the forming of tracks by living
salamanders showed the virtual impossibility
for such animals to leave diagnostic trackways
on a sand surface under water (peabody. in
McKee. 1947. p. 27).
3. In extensive observations of modem amphibian
trackways on mudflats. no examples of the
clear outlines of individual footprints
preserved under water were found. even
where some trackway patterns remained
(Peabody. in McKee. 1947. p. 27).
The environmental conditions under which
tracks in dry sand might be permanently preserved
were discussed under "Coconino Sandstone. tt
PITS OF RAINDROPS in modem sand of lyons Sandstone tex-
ture. One inch equals 2.54 cm. (Fig. 156.) Environment of Deposition
avalanches. where the surface is slightly cohesive Two aspects of the regional environment in
(McKee. Douglass. and Rittenhouse. 1971. p. 369). which the Lyons Sandstone was formed are
These minor structures seem to be similar in all generally agreed upon. First. the formation was
essential respects to the ripple marks. raindrop deposited in a long narrow coastal area extending
pits. and slump marks that are recorded from the from north to south with a seaway to the east. Sec-
Coconino. De Chelly. and some other eolian-type ond. it was flanked to the west by a granitic terrane
sandstones. Their probable modes of preservation referred to as the ancestral Front Range highlands,
and their interpretation were discussed in the sec- and this upland contributed arkosic detritus to the
tion entitled "Coconino Sandstone. tt It is sufficient formation. Details concerning the subenvironments
to say that these structures all give strong support- in which various facies of the formation developed
ing evidence for deposition in an eolian environ- have been interpreted in many ways and ideas
ment. about their genesis have differed greatly.

Tracks and Trails Various investigators have attributed the sedi-

mentary deposits that compose the Lyons Sand-
The footprints of vertebrate animals. probably stone to accumulation (1) by streams, (2) on
reptilian. and the tracks of scorpionlike inverte- beaches and bars. (3) by the wind. and (4) through
brates preserved on the steeply dipping foresets of combinations of these factors. Deposition by
the Lyons Sandstone were described many years streams, at least in part, is advocated by Maher
ago (Henderson. 1924). These clear and sharp (1954), Vail (1917), W. O. Thompson (1949), and
tracks. which preserve delicate details. including Weimer and Land (1972) . Littoral sedimentation,
the claw marks of the reptiles. probably are identi- primarily beach and bar accumulation, is referred
cal to forms reported from the Coconino and De to by W. O. Thompson (1949). Maher (1954), and
Chelly Sandstones. The manner of their preserva- Hubert (1960). Eolian deposition, involving the
tion and the conditions under which they ap- highly cross-stratified facies, is advocated by Vail
parently were preserved indicate deposition in dry (1917), Tieje (1923), and Walker and Harms (1972).
dune sand. Most of the early workers gave little specific evi-
The validity of the preceding conclusions is indi- dence to support their conclusions, seemingly
cated by the following: reached largely on the basis of paleogeographic and

Digitized by Google
220 A STUDY OF GLOBAL SAND SEAS

other general considerations. Eolian deposition They differ even more from the very broad, flat in-
now seems well established for major parts of the terdunes between linear (seif) dunes, which com-
Lyons. but numerous other facies, both coastal and monly have horizontal strata (McKee and Tibbitts,
alluvial, probably occur in the formation. 1964, fig. 8) accumulated on their surfaces. (See
The kind of dune that is represented by the section entitled "Characteristics of Interdune
large-scale cross-strata of the Lyons Sandstone at Record," this chapter.)
and near its type locality has caused much Another aspect of the Lyons dune deposits, lend-
speculation. Several types of pertinent evidence ing support to interpretation of their origin as
have been discussed, including analysis of cross- parabolic types, is their geographic position.
strata dip directions, descriptive information on General agreement among concerned geologists,
the nature and extent of interdune surfaces. and based on a considerable amount of evidence. indi-
conclusions based on paleogeographic and en- cates that the Lyons Sandstone was formed in a
vironmental considerations. coastal environment. so a variety of coastal dune is
Systematic directional data based on dip readings inferred. If such an inference is correct. then the
have not been obtained for the eolian-type unit of evidence already cited - blowout-type interdunes.
the Lyons Sandstone. The sets of large-scale cross- divergent wind directions. large-scale cross-strata.
strata at its type locality were reported by Walker and common wind ripple marks - all seem com-
and Harms (1972, p. 281) "to dip in a broad arc from patible with the parabolic dune interpretation.
southeast to west." As further stated by them, "the
highly variable local dip directions show that accre- Summary of Criteria for Eolian Origin
tion occurred along [rock] faces of diverse orienta-
tion." Statistically sound data are needed to firmly The following features indicate the eolian origin
establish the amount of spread; nevertheless, at of the highly cross-stratified sandstone facies of the
least suggestive is the considerable width of spread Lyons Sandstone:
in dip directions. This spread, as described, forms a 1. A composition of clean fine quartzose and
pattern similar to that of coastal parabolic dunes in feldspathic sand. well sorted. subangular to
Brazil determined by Bigarella (this chapter). The well rounded. and commonly frosted. Such
pattern in the Lyons seems very different from that features. although not incontrovertible proof
shown to be typical of barchanoid and transverse of genesis. are compatible with the texture of
dunes (Glennie, 1972, p. 1061) with their narrow many modem dune deposits.
spread (clustering of points). It contrasts even more 2. Thin layers of scattered but rather uniformly
with the nearly opposite double clusters of points spaced. coarse grains resting on surfaces of
on linear (seif. longitudinal) dunes illustrated by truncation within fine cross-stratified sand.
Glennie (1972, fig. 13) and by McKee and Tibbitts The coarse detritus is interpreted as lag
(1964, fig. 6). material.
3. Major primary structures consist of large-scale
Compelling evidence suggests that the wind-
deposited sands of the Lyons represent parabolic high-angle cross-strata in tabular-planar sets.
dunes. not barchanoid- or linear-type dunes. Inter- Such structures are typical of most dunes
dune surfaces between sets of high-angle cross- formed under conditions of unidirectional
strata are "scoop-like erosional surfaces" that "in- winds but rarely occur in other environments.
tersect along gently curved nicklines" (Walker and 4. Curving scoop-shaped surfaces. exposing the
Harms. 1972. p. 281). These surfaces are broadly beveled tops of cross-strata that represent sets
curved and exposed the beveled tops of earlier of earlier development. are characteristic
cross-strata sets. as is typical of most modem inter- features of interdunes developing in a
dunes (McKee and Moiola, 1975). They represent a parabolic-dune environment.
blowout phenomenon, which marks an early stage 5. Numerous asymmetrical ripple marks oriented
of coastal parabolic dunes. and differ from inter- parallel to the dip directions of foresets which
dunes among barchans and transverse ridges in that have high ripple indices characteristic of wind-
they are relatively small and have curving surfaces. formed ripple marks.

Digitized by Google
 ANCIENT SANDSTONFS CONSIDERED TO BE EOIJAN 221
6. Surfaces of some strata have slump features, ments which may produce large festoons are dis-
both the small steplike variety and the cussed on following pages.
dendritic type that are commonly formed from
avalanching of damp, cohesive sand. Surfaces Distribution and Thickness
also have raindrop pits.
7. Locally, the sharp, well-preserved footprints of The Casper Formation. including both lime-
four-footed animals, probably reptiles, occur stones and sandstones. bounds much of the south-
on the steep foresets. This type of preserva- ern part of the Laramie Basin and flanks the
tion, characteristic of a dry-sand environment, Laramie Range in southern Wyoming. an area 75
is not possible under water. miles (120 km) wide and 100 miles (160 km) long.
The cross-stratified sandstones, believed to be
eolian. occupy the entire thickness of the formation
Casper Formation on its southeastern margin encompassing at least
1.000 square miles (1.600 kml) (Knight. 1929. p. 56).
(Pennsylvanian and Permian),
The Casper is more than 700 feet (200 m) thick
U.S.A. east of the Laramie Range; near the center of the
THE GENESIS OF the cross-stratified sandstone Laramie Basin about 550 feet (165 km) thick. and in
of the Casper Formation in the southern part of the the Sand Creek area to the south (where the cross-
Laramie Basin, Wyoming, has long been a subject of stratified sandstones have been most studied) 120
dispute. Whether this formation is marine or eolian feet (36 m) thick (Steidtmann. 1974). Its facies ap-
has been difficult to demonstrate largely because parently are lateral extensions of still other and
composition and texture of the sandstone could very different facies represented in the Permian
represent either environment, and the size and type and Pennsylvanian Tensleep Sandstone to the
of cross-strata and contorted beds that are its most northwest. It may also be at least partly contem-
distinctive structures have not yet been demonstra- poraneous with the Permian and Peimsylvanian
ted to occur in any modem deposits. arkosic Fountain Formation to the south, which is
Fortunately, for the understanding of the origin considered to be formed largely of coalescing
of the sandstone facies in the Casper, an excellent alluvial fans (Hubert. 1960).
detailed description and analysis of its structure
was made many years ago by Knight (1929). This Composition and Texture
pioneer work, a model for much subsequent work
in the study of cross-stratification, called attention The cross-stratified sandstone of the Casper is
to the remarkable and baffling type of structure fine grained and well sorted. Exceptions to this
which Knight termed "festoon cross-lamination." uniformity are laminae at the bottoms of cross-
(See discussion of "Festoon Structures" later in this strata troughs that commonly contain evenly dis-
section on the Casper Formation.) Today, it is com- tributed. scattered coarse grains of quartz and
monly referred to as large-scale trough cross- feldspar. These grains may be lag material (Steidt-
stratification. mann, 1974. p. 1838); if so. the laminae probably
represent interdune surfaces developed in an en-
Festoon structures that are extensive and large,
vironment of blowout dunes.
like those in the trough-type cross-strata of the
Casper, do not seem to have good analogs in any Sandstone of the Casper is classed as calcareous
modem sand deposits that have been examined by subarkose or orthoquartzite. according to Steidt-
trenching; nevertheless, the Casper is believed by mann (1974. p. 1838). In some areas it is inter-
many geologists to represent a dune environment, stratified with beds of marine aphanitic limestone
on the basis of other data. Accordingly, the Casper, that attain a maximum total thickness of 210 feet
as a unique form, is described herein. Trough struc- (64 m) on the west flank of the Laramie Range.
tures of festoon type in some modem noneolian Limestones are absent in the southern part of the
deposits are described. and certain dune environ- area (Knight. 1929. p. 53).

Digitized by Google
222 A STUDY OF GLOBAL SAND SEAS

Primary Structures festoon Structures

Stratification The term "festoon cross-lamination" as origi-
nally used by Knight (1929) for cross-strata of the
The sandstone facies of the Casper Formation
Permian and Pennsylvanian Casper Formation was
throughout a broad area characteristically consists
applied to a type of trough cross-stratification filled
of large trough cross-stratification, commonly
either symmetrically or asymmetrically and having
termed "festoon lamination." This type of cross-
a large range in size.
strata consists of a repetition of two elements -
erosion surfaces that form plunging troughs and The term "festoon-type cross-bedding" has been
thin concave-upward laminae that fill the troughs used by Tanner (1965, p. 584; 1966) for a structure
and conform largely to the shape of the trough apparently very different from the original "festoon
floors (fig. 157). lamination" of Knight. Tanner probably referred to
The erosion troughs in the Casper differ greatly the wedge-planar sets in the Jurassic Entrada
in size, ranging in width from 5 to 1,000 feet (1.5 to Sandstone, which he stated "form additional evi-
300 m), in length from 25 feet to a few thousand dence * * * of a wind origin." These sediment-filled
feet (8 m to a few hundred metres), and in depth troughs are superimposed one above another and
from a few feet to 100 feet (30 m) (Knight, 1929, overlap to form the characteristic festoon pattern.
p. 59). The surface of each trough is cut across at The genesis of this structure was attributed by
one end by one or more overlying troughs, and so Knight to marine processes. His evidence did not
each trough has "the general shape of a long, consist of characteristics of the cross-strata,
shallow, scoop-like depression." The average in- however, and his chief reason was the close asso-
clination of trough surfaces is 10-15. ciation of sandstones of the Casper with other
Most of the laminae that fill the erosion troughs facies that were of undoubted marine origin. Since
form symmetrical sets, only a few being asymmetri- that pioneer study, several geologists have con-
cal. The angles of inclination of the strata as deter- tended that an eolian origin better explained
mined by Knight (1929, p. 84) from about 800 read- various features of the sandstone, especially the
ings were 15-25 and averaged approximately 20. minor structures.
Very few sets dipped toward the northeast quad- Today, the origin of the festoon-type cross-
rant; nearly all dip directions fell within the 220 stratification in the Casper still remains an enigma.
arc from N. 35 W. southwestward to S. 75 E. Probably the greatest deterrent to a widespread ac-
(Knight, 1929, fig. 30). ceptance of the eolian origin has been the fact that
no modern analogs displaying all the Casper's
characteristics have been recorded. Point bars in
major river systems seem to be the closest modern
analog.
Festoon structures in river point bars were first
reported and illustrated by Frazier and Osanik
(1961) from the Mississippi River Delta. Sections
transverse to current direction revealed well-
developed festoons forming superimposed semicir-
cular troughs filled with concave-upward strata
comparable to those of the Casper. Principal
differences were lenses of gravel and other coarse
detritus, local accumulations of plant debris, and a
lack of any very large channels {up to 1,000 ft (300
m) wide and 100 ft (30 m) deep), such as occur in
FESTOON CROSS-LAMINATION, as described by Knight the Casper. Other examples of festoon structures in
(1929), from sandstone of Casper Formation, Wyoming, U.S.A. river sediments were recorded by Hamblin (1961),
This structure consists of overlapping, symmetrically filled Harms, MacKenzie, and McCubbin (1963) , Harms
troughs. (Fig. 157.) and Fahnestock (1965), and Williams (1968), an in-

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 223

dication that they are fairly common in a fluvial en-

vironment.
How fluviatile large-scale trough cross-strata
formed - by a single process of scour-and-fill re-
lated to large sand waves, or by two stages involv-
ing the filling through megaripples of previously
cut troughs - has been controversial. Whatever the
process, the resulting structures differ from the
Casper-type structures in being considerably small-
er, shallower, and less symmetrically filled. Trough
structures preserved in dry river channels in South
Australia are described (Williams, 1968, p. 138) as
having relatively coarse texture - that is, coarse to
very coarse sand with some granules and pebbles; BLOWOUT DUNES ALONG SOUTH SHORE of Lake Michigan,
locally, they include concentrations of sticks and U.S.A. Photograph by R. C. Gutschick. (Fig. 158.)
leaves.
In modem eolian dune deposits, large trough Knight (1929, p. 27) to asymmetrical filling with
structures seem to be rare, and those that are sand. This process forms slopes above the angle of
recorded are mostly shallow and not sufficiently repose with a resulting creep down the over-
overlapping to form festoon structures. Scattered steepened side and a tendency toward overturning
channels were found at White Sands National the compressional folds. This process was believed
Monument, New Mexico, in trenches parallel to the by Knight to have required saturated sand; dry sand
dune crest of a transverse dune (McKee, 1966, fig . dune deposits were not considered favorable.
7), but few were found in other types of dunes in Recent experiments on contorted bedding
that area. Similarly, in the coastal dune ridges of (McKee, Douglass, and Rittenhouse, 1971, p. 373)
Brazil examined by Bigarella, Becker, and Duarte showed, however, that a high degree of cohesion is
(1969) uncommon trough structures are shallow necessary in forming structures of this type and
and lack well-developed festoon patterns. that such cohesion occurs only in wetted sand
In modern dune areas, festoon structure (such as occurs in dunes that have been exposed to
analogous to that of the Casper is most likely considerable rainfall or other moisture) and not in
among coastal dunes where "blowouts" are abun- saturated sand. Coastal dunes of Brazil have been
dant. Trenching, followed by detailed analysis of found to contain many contorted structures
structures, apparently has not been attempted in (Bigarella, Becker, and Duarte, 1969; McKee and
this environment, but the shape, abundance, and Bigarella, 1972, p. 673), although most of these are
size of blowouts in some regions (fig. 158) mayac- much smaller than those of the Casper Formation.
count for large festoon structures.
Ripple Marks
Contorted Bedding Low wide ripple marks with parallel crests and
Numerous sets of trough cross-strata in the troughs, oriented up and down the dip slopes of
Casper Sandstone are crumpled and contorted on a cross-strata. are relatively numerous in the Casper.
huge scale, with folds having amplitudes up to 25 These ripple marks are basically like those in the
feet (8 m) locally. This contortion of bedding ap- Coconino Sandstone of Arizona which were
parently developed during or immediately after studied many years ago and described by the author
deposition of the sand; examples are numerous of (McKee. 1934b. p. 100; McKee. 1945. p. 323) . Similar
highly deformed strata in one set that is directly ad- ripple marks have since been recorded from the De
jacent to or overlying uncontorted sets (Knight, Chelly Sandstone. the Lyons Sandstone. and
1929, p. 74) . various others thought to be of eolian origin and
The deformation of strata within the Casper discussed in this chapter.
troughs, resulting in large, locally developed folds Low-amplitude ripple marks in the Casper For-
oriented in many directions, was attributed by mation were described by Steidtmann (1974. p.

Digitized by Google
224 A STUDY OF GLOBAL SAND SEAS

1837). as follows: "The ripple crests are relatively preservation could scarcely have occurred either in
straight. parallel. slightly asymmetric. and spaced saturated sand or in dry loose sand (Hanley and
5.5 to 9.0 cm apart. Their heights are from 1.5 to 3 others. 1971. p. 1067); a moist cohesive sand. possi-
mm; therefore. the ripple indices are 38 to 30." Both bly in a coastal environment. would have been
orientation of the ripples and the ripple indices are most favorable. Development and burial of these
typical of eolian forms. trails probably were like those attributed to the in-
The ancient ripple marks just described resemble vertebrate tracks of the Coconino (McKee. 1934b. p.
structures developed on the lee faces of various 108. pI. 13a).
kinds of modem dunes formed when winds normal
to the prevailing wind are active. Modem lee-side Environment of Deposition
ripple marks have been observed on both The broad regional environment in which the
barchanoid ridge and transverse ridge dunes at Casper Formation was deposited is readily deter-
White Sands National Monument. New Mexico. mined. Criteria that seem convincing are available
(Chapter E. figs. 41C, 41D). on reversing dunes at to permit an interpretation. and geologists have
Great Sand Dunes. Colorado. on transverse dunes reached general agreement on paleogeographic
at Padre Island. Texas. and many others. controls (Knight. 1929; Steidtmann. 1974). Ap-
parently. a mountain mass of some relief - the an-
Interdune Deposits cestral Front Range - extended from northwest to
Thin lenticular limestones that occur "along low- southeast across south-central Wyoming. Parallel to
relief truncation surfaces within the sandstone and northeast of those highlands was the margin of
units" of the Casper Formation were reported by a seaway. Between the mountains and the sea was a
Hanley and Steidtmann (1973). These limestones. broad coastal plain across which rivers transported
of local distribution. do not contain any marine much arkosic sediment. Evidence for the sea is fur-
macrofossils. are described as "ostracodal. peloidal. nished by limestones of the Casper containing
microsparites" and contain shrinkage cracks. gas many kinds of marine fossils (Miller and Thomas.
cavities. and burrow traces. 1936; Thompson and Thomas. 1953).
From the description. these limestone lenses are The environment of deposition of the trough-
interpreted herein as interdune deposits. rather type cross-stratified sandstones in the Casper
than as tongues of normal marine or of lagoonal clearly is related to the environment of the inter-
limestone. They are believed (Hanley and bedded or closely associated marine limestones.
Steidtmann. 1973. p. 428) to have been "precipi- Whether the sandstone is marine or eolian remains
tated in small. shallow ponds in which environ- uncertain. mainly because no known analog in
mental conditions were harsh because of fluctua- modern sediments completely matches the sedi-
tion in water temperature. depth. and salinity." mentary structures observed in the Casper. Despite
some evidence of dune deposition. an interpreta-
Trace Fossils tion of the remarkably large-scale trough structure
Two varieties of trace fossils in the Casper For- cannot be considered to be secure until a modern
mation have been recorded (Hanley and others. analog is found.
1971). One variety consists of ribbonlike trails on
the upper bounding surfaces of cross-strata that fill Summary of Criteria for Eolian Origin
troughs of the Casper; the narrow trails intersect
one another but do not branch. The second variety Probably the most compelling evidence that
consists of tubular internal traces. Animals respon- sandstones of the Casper were formed through
sible for the first variety of trail have variously deposition by the wind is found in the relatively
been assumed to be insects ("beetles"). arthropods. numerous eolian-type ripple marks with crests
and annelids (Hanley and others. 1971. p. 1067); oriented parallel to the dip of foresets and with high
however. the main importance of these trails is that ripple indices.
they indicate the probable depositional environ- Further suggestion of eolian origin is furnished
ment of the enclosing trough-type structures. Their by surfaces of scattered coarse sand at the bottoms

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 225

of cross-strata troughs. These deposits are in- western Germany, from Scotland and southern
terpreted as lag sand of interdune areas; thus, the Norway, across England and the Netherlands to
troughs may be considered blowout phenomena, central Germany (Kent and Walmsley, 1970). This
such as are common in many coastal dune fields Permian desert probably approached "in scale the
and the fillings are blowout-type dunes. Supporting area of the present Sahara, especially if one con-
this concept is the very wide spread of cross-strata siders its possible former extension into North
dip directions, characteristic of some coastal dune America," according to Glennie (1972, p. 1069),
areas, such as the Lagoa dune field, Brazil (chap- who has suggested that it may originally have been
ter E). "between the paleolatitudes of roughly 10 N. and
Results of experiments on settling velocities of 30 S.,.. as indicated by paleomagnetic data. Since
light and heavy minerals made by Steidtmann then, the landmass containing this low-latitude
(1974, p. 1840) are presented by him as corrobora- paleodesert apparently has migrated by continental
tive evidence of an eolian deposition. drift to its present position.
The presence of lenticular limestone beds locally The thickness of the Rotliegendes is about 750
distributed along truncation surfaces of cross- feet (225 m) along its southern margin (Gill, 1967;
stratified sandstones, their ostracodal peloidal tex- Kent and Walmsley, 1970) and more than 5,000 feet
ture, and their associated shrinkage cracks, gas (1,500 m) in the central part of the depositional
cavities, and burrow traces strongly suggest an in- basin in northern Germany (Kent and Walmsley.
terdunal origin. Trace fossils, consisting of rib- 1970). These figures for distribution and thickness
bonlike trails on upper bounding surfaces of cross- refer to the entire formation or desert environment
strata, indicate by their clear preservation that the and include several noneolian facies. The major en-
sand was neither dry nor saturated at the time of vironments or facies as classified by Glennie (1972,
burial but probably was moist and cohesive. fig. 17) are dune, wadi, sebkha, and desert lake.
Various features of texture and structure are Sandstones in the southern North Sea area that are
compatible with the concept of an eolian origin, but attributed to a dune-sand origin have a thickness up
a conclusive test of genesis involving dissection of to 600 feet (180 m) (Glennie, 1972, p. 1048).
dune types on a modern coastal area, where trough-
type structures are exposed, has not yet been made. Texture
Quartz grains in those sandstones of the
Rotliegendes of Northwestern Rotliegendes that are considered to be eolian
generally range in size from fine to medium; locally
Europe (Permian) some coarse grains occur. The finer grains are
THE PERMIAN RED-BED sequence of north- mostly subangular, whereas the coarser ones tend
western Europe, normally referred to in Germany to be subrounded to rounded. "Impact fractures
as the Rotliegendes, represents a great desert area typical of those found in modern aeolian sands
of the past, and so merits discussion here. A careful have been recognized during electron-microscope
analysis of these strata and a comparison with investigation of the lower Permian Penrith
deposits of some modern desert areas were made by Sandstone" (D. B. Smith, 1972, p. 13). The cements
Glennie (1972, p. 1048). The following brief discus- are principally hematite and authigenic clay, and
sion is largely a synthesis of his work as it pertains locally dolomite or anhydrite. Argillaceous matrix
to those sandstones in the Rotliegendes believed to is lacking (Glennie, 1972, p. 1058).
be eolian. Additional information on the Interbedded detrital sediments that comprise
Rotliegendes and its environment of deposition is thin, roughly horizontal sets between crossbedded
derived from Falke (1972) for central and western units of dune type are interpreted by Glennie (1972,
Europe and from Wills (1948,1956) for England. p. 1061) as interdune sebkha deposits that are com-
posed largely of fine to coarse sand. As in many
Distribution and Facies modern interdune deposits, such horizontal beds
Sedimentary rocks of the Rotliegendes extend are characteristically poorly sorted and are silty.
over a wide area from Poland to France and They commonly contain many small irregular

Digitized by Google
 226 A STUDY OF GLOBAL SAND SEAS

adhesion
material. ripples but very little argillaceous def med as small lrreg
. ular mounds or ridges formed
where dry sand is blown onto a nearly horizontal
Primary Structures humid surface and fixed there by surface tension.
Laminae are concave upward.
Stratification
Summary of Criteria for Eolian Origin
Cross-stratified sandstones, interpreted as eolian
typically occur as large-scale planar types with Large sequences of cross-stratified sandstone in
steeply dipping (20 - 27) foresets. Dip directions the Rotliegendes have long been attributed to ac-
of laminae are relatively uniform, dips tend to in- cumulations of sand by eolian processes, because
crease upward, and truncated tops of sets are over- their textural and structural properties are similar
lain by subhorizontal finely laminated sandstones to those of modem dune deposits and because they
(Glennie, 1972, p.l058). In many places these dune- are closely related to surrounding and interbedded
type laminae overlie other sand sequences that are deposits recognized as desert facies. These facies
considered to represent a wadi environment. Ster- include wadis, desert lakes, and sebkhas. Few, if
eographic polar nets prepared by Glennie (1970, any, of the properties described for the
fig. 68) strongly suggest that both barchanoid Rotliegendes can individually be considered proof
(transverse) and linear (seil, longitudinal) dunes are of a dune origin, yet in aggregate they present
represented in the sandstones believed to be eolian. strong evidence. On the other hand, recent
Within barchanoid-type dunes, the dip attitudes of petrographic studies led Pryor (1971a, b) to ques-
loresets are characteristically concentrated in a tion this origin for some of the sandstones and to
single, limited area which is in line with the postulate a genesis in "a shallow marine environ-
ment" where sand was reworked by the sea and
predominant wind direction. In linear-type dunes,
transported southwestward.
however, foreset dip directions are concentrated in
Most eolian sands are characterized by domi-
two widely separated, nearly opposite directions
nantly subrounded frosted well-sorted grains and
(McKee and Tibbitts, 1964, p. 12). Borehole data
by lack of argillaceous matter. Such textural
plotted for the basal Permian "Yellow Sands of
features are common in the Rotliegendes. Struc-
Durham," of Great Britain, show a form that, ac-
tures including well-laminated high-angle cross-
cording to Smith and Francis (1967, p. 97- 98, fig.
strata are typical, but not indisputable, evidence.
18) "strikingly resembles" a seif dune, such as that
More compelling as evidence are the individual se-
developed under two alternating wind regimes
quences that show upward progressions from
postulated by Bagnold (1941).
subhorizontal to steeply inclined laminae and a
The trend and pattern of cross-strata "in several record of interdunes shown by intraformational un-
basins of central and western Britain have shown conformities and by horizontally bedded adhesion
that the sand [was] accumulated in those places in ripples. Additional strong evidence furnished by
barchans" under easterly winds as described by D. cross-strata is their common occurrence as large-
B. Smith (1972, p. 13), but he further states that in scale planar types.
northeastern England "east-north-east trending Statistical studies of dip directions on cross-
longitudinal dunes were present." Cross-strata dip strata, indicating a small spread in trends of deposi-
directions in the Rotliegendes, according to Glennie tional currents, offer evidence of eolian origin and
(1972, p. 1061), suggest that seif dunes predomi- also a means of differentiating between the linear-
nated in Britain and West Germany and that the type dunes (with two nearly opposite clusters of
barchanoid-type was prevalent in other areas (that readings) and the barchanoid-types (with a single,
is, away from the Variscan Mountains). larger grouping of directional points).
Adhesion Ripples Strongly suggesting a desert dune genesis is the
association of large-scale dune-type cross-strata
Interdune deposits on the Rotliegendes are thin with evaporite basin deposits of anhydrite, halite,
zones primarily consisting of "small, irregular, and and red clay above and with probable wadi deposits
more or less horizontally-bedded adhesion ripples" of conglomerate, mudcracked clays, and flat-bed-
(Glennie, 1972, p. 1061). Adhesion ripples are ded sands below.

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 227
cludes some scattered feldspar grains (Shotton,
Lower Mottled Sandstone 1937, p. 535). The dull-red coloring of the rock is at-
(Permianl), England tributed to thin grain coatings of iron oxide, which
is believed to be "the only cementation." The
ONE OF THE FIRST detailed studies of eolian- sandstone has "extraordinary uniformity," as con-
type sandstones, critically analyzed and studied in firmed by analyses of samples from 476 exposures.
detail, was Shotton's 1937 investigation of the The average grain size was determined to be 0.2
"Lower Bunter Sandstone," now commonly refer- mm, with the median diameter ranging from 0.129
red to as Lower Mottled Sandstone, of the Bir- to 0.410 mm; no particles are smaller than 0.05 mm
mingham area in England. His work, which formed or larger than 1 mm (Shotton, 1937, p. 552). The
Po model for many later studies, not only presented fineness of grain and high degree of sorting charac-
evidence for ascribing an eolian origin but also terize the entire sandstone throughout the area ex-
developed criteria for interpreting the type of dune, amined.
believed to be barchan. His report was a penetrating
analysis of the manner in which dune structures
Primary Structures
are developed and the effect of various factors in- Stratification
volved. His statistical study of cross-strata dip A systematic investigation of cross-stratification
directions and discussion of their significance in in the "Lower Bunter Sandstone," undertaken by
terms of regional wind currents and dune migration Shotton and others, enabled Shotton (1937) to make
paved the way for an understanding of the environ- a detailed description and careful analysis of its
ment and history of eolian deposits in many other structures. Cross-strata in the formation are
regions. The classic study by Shotton was recorded as dominantly high angle, ranging bet-
published almost simultaneously with Reiche's ween 20 and 30, and are mostly large scale, form-
pioneer work (1938) along similar lines - analysis ing tabular-planar sets up to nearly 100 feet (30 m)
of the Coconino Sandstone of Arizona. thick. Individual foresets are steepest in their upper
Both because of its historic significance and parts and are truncated by low-angle erosion sur-
because of the excellence of deductive reasoning faces at their tops; they are tangential with the un-
with regard to a basic dune type, here classed as the derlying bounding planes. Between sets of cross-
barchanoid dune ridge, Shotton's work qualifies the strata low-angle erosion surfaces dip in a direction
"Bunter" for discussion in this publication. Nearly opposite that of the cross-strata. These low-angle
all data used herein are derived from Shotton's planes are interpreted as windward-side dune sur-
classic paper; additional important references are faces.
Wills (1948; 1956, figs. 2416) and Shotton (1956). Statistical measurements of cross-strata dip
directions made at 476 exposures and consisting of
Distribution and Thickness 1,142 readings furnished data for determining both
The Lower "Bunter" Mottled sandstone (Shot- current-direction vectors and the amount of spread
ton, 1937, p. 535) "consists of nothing but sand," involved in dip directions of each outcrop area
with no clay beds and no pebbles in the four areas studied. Using the mean direction of dip as repre-
of outcrop, west and northwest of Birmingham, senting the dominant wind direction, a regional
England, where it was studied. These outcrops ex- wind from the east was determined. Calculating the
tended along a belt roughly 25 miles (40 km) north- size of arc represented by all readings in a particu-
south and 10 miles (15 km) wide. A maximum lar area disclosed a considerable curvature in dune
thickness of 850 feet (260 m) for the formation is ridges to be typical. Hence, a barchanoid ridge,
reported (Shotton, 1937, p. 536) from a borehole in rather than a straight transverse-type ridge, is
the area studied. In other parts of the southern postulated for this formation.
Cheshire basin, the same formation is known to be
650 feet (217 m) thick. Environment of Deposition
Composition and Texture The sand desert with inferred barchanoid-type
dunes of the "Bunter" Mottled Sandstone is in-
The Lower "Bunter" Mottled Sandstone is terpreted by Shotton (1937, p. 550) as occupying an
formed almost entirely of quartz sand grains but in- elongate strip west of a highland formed during the

Digitized by Google
228 A STUDY OF GLOBAL SAND SEAS

Hercynian orogeny. This highland "fulfilled the Cheshire. England. were compared with structures
double function of abstracting moisture from the exposed by trenching in modern dunes of New
prevailing east wind. and of providing the material Mexico in the United States (McKee. 1966).
to build up the sand sea." The sandstone study by Thompson represents a
Between the postulated highlands to the east and detailed description of a relatively small but dis-
the dune-type cross-strata of the "Bunter" are peb- tinctive eolian-type sandstone and presents a rather
ble and sand beds. interpreted as weathering convincing case. through cross-strata analysis. for
deposits derived from the uplands and spread by in- the presence of dome-shaped dunes that had not
termittent streams across the lower slopes of the previously been recognized in ancient rocks. The
hills to the edges of the desert. These alluvial method of tracing major bounding erosion surfaces.
deposits. therefore. are considered the principal formed at the base and top of dunes. and of record-
source of the dune sands. Despite some uncertainty ing the patterns of cross-strata sets between these
in correlation of the fluviatile deposits with eolian surfaces. is an effective means of establishing the
strata (Shotton. 1937. p. 352) because no transition diagnostic structural features of various dune types
zone now exists. much evidence supports the rela- (D. B. Thompson. 1969). Furthermore. the relation-
tionship. ship of these dune sands to other closely associated
sedimentary facies illustrates a complex set of
Summary of Criteria for Eolian Origin depositional environments.
The Frodsham Member of the so-called Keuper
Textural features of the Lower "Bunter" Mottled Sandstone Formation. as discussed here. is a
Sandstone (Shotton. 1937. p. 538) are comparable in relatively thin (as great as 55 m or 170 ft) member
all respects with the textures of sands in most of red fine-grained sandstone which forms the QP-
modern dunes but show marked differences from per part of a sequence that contains medium- to
the textures developed in many other depositional coarse-grained sandstones and conglomeratic
environments. Distinctive textural features are the sandstones interbedded with other members com-
fineness of grain. the high degree of sorting. and the posed of soft fine-grained sandstone (table 24; D. B.
absence of both coarse detritus and silt-clay parti- Thompson. 19708. b). The Frodsham Member is
cles. present in about 2.000 kmz (1.240 miZ) of the
Most compelling evidences of eolian origin are Cheshire and East Irish Sea Basins. but the
the size and type of cross-strata represented in the Frodsham [facies (sensu stricto)] with dome-
formation - high-angle large-scale tabular-planar. shaped and transverse dunes apparently is
with long curving surfaces tangential to the bound- restricted to a small area near Frodsham and Run-
ing planes at their bases. The frequency distribu- corn in northwest Cheshire. where D. B. Thompson
tion of the cross-strata dips suggests that the dunes (1969) made his detailed studies.
were of the barchanoid-ridge type. migrating in a
constant direction and accumulating one set upon Composition and Texture
another.
The eolian-type sandstone of the Frodsham
Frodsham Member of Member (D. B. Thompson. 1969. p. 272) consists
dominantly of quartz sand with sparse grains of
So-Called Keuper Sandstone feldspar and little or no mica. The electron
microscope enabled the identification of authigenic
Formation (Triassic), England feldspar (B. Waugh. oral commun. to D. B.
Thompson. 1975). In associated beds believed to be
ONE OF THE MOST SUCCESSFUL attempts to fluviatile and lacustrine. feldspars are much more
interpret ancient deposits of probable eolian origin numerous and mica flakes are common; clay
in terms of specific dune types. is D. B. Thompson's minerals are illite and kaolinite.
(1969) study of the Frodsham Member of the so-
called Keuper Sandstone Formation (Triassic). For The roundness of quartz grains in the "Frodsham
his investigation. cross-strata patterns recorded in facies" is described as "subangular to subrounded"
sandstone of the Frodsham Member in West and grain-size analyses show (D. B. Thompson.

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 229
TABLE 24- Age, thickness, composition, and postulated depositional environments of units formins lDwer "Keuper" and un-
derlying sandstones in West Cheshire, England
[Prepued by D. B. 'I'bo1llplOll. Unlvenity of Keele. 1l17li)
Maximum Depoeltlonal
Ale IJlfonMI formatiana thlclm_ IIIlvlronment
(metres)
Scythian "Keuper" Intertidal and 7f1uvial.
7Anisian Waterstones
Lower Frodsham Soft Sandstone Member 55 Eolian: Sand sea rather

I
"Keuper" than coastal dunes.
Scythian Sandstone Delamere Pebbly Sandstone Member 85 Fluvial low sinuosity.
(as much as Thumaston Soft Sandstone Member 45 Eolian and fluvial.
220 m thick)
Alderley Conslomerate Member 34 (not Fluvial low sinuosity.
always
present)
Possible unconformity
Scythian Bunter 950 Eolian and fluvial;
Upper Mottled fluvial mostly near base.
Sandstone

7Scythian Bunter
Pebble beds

1969, table 3) that in most sections the sand is domi- bottoms of cosets that may include numerous tabu-
nantly fine grained, and in some, mostly medium lar-planar and wedge-planar sets. "Shallow trough
grained. Very few grains in any sections are above structures are present * * * but form only a small
or below these size grades, thus attesting to the high proportion of the structures present;" most of them
degree of sorting in the sandstone. Most grains have occur directly below major erosion surfaces.
coatings of diagenetic red ferric oxide, deposited From the cross-stratification patterns analyzed,
presumably by circulating ground waters (D. B. D. B. Thompson (1969, p. 284) recognized "at least
Thompson, 1969, p. 272). eight vertically and laterally overlapping eolian
The attributes of composition and texture just dunes," six of which were of the dome-shaped type
described are all consistent with an eolian origin, with a tendency toward the transverse type. The
but they reflect genesis and transportation rather beveled tops of cross-strata sets and other structural
than deposition. The presence of lenses and other features as described for dome-shaped dunes at
concentrations of "millet seed" sand has been White Sands (McKee, 1966) suggest that the
viewed by various geologists as evidence of wind degradation of the transverse dunes by very strong
deposition; however, according to D. B. Thompson winds is responsible for features described in the
(1969, p. 270) some entire beds in the so-called "Keuper" dome-shaped dunes.
"Keuper Sandstone Formation," dominated by such Penecontemporaneous contorted structures are
sand, also include large pebbles and clay galls. not uncommon in tabular-planar and wedge-planar
sets of cross-strata in the "Frodsham facies of the
Primary Structures Keuper Sandstone." "Simple concentrically folded
Stratification laminae," several feet thick and overlain by un-
The "Frodsham facies" - the probable eolian disturbed strata, closely resemble the sand-drag
part of the "Keuper Sandstone Formation," as type (McKee, Reynolds, and Baker, 1962b). Other
described by D. B. Thompson (1969, p. 273, 285) - varieties, which are numerous, apparently are like
consists mostly of large sets of cross-strata with types attributed to dry and to damp sand (D. B.
very long laminated foresets, some of which extend Thompson, 1969, p. 285).
downward. and laterally into flat-bedded and rip- Dip-direction vectors, indicating paleocurrent
ple-bedded bottomsets of siltstone and mudstone. summaries for 10 dunes analyzed in the "Frodsham
Major bounding planes of erosion mark tops and facies," were determined by D. B. Thompson (1969,

Digitized by Google
230 A STIJDY OF GLOBAL SAND SEAS

p. 278). The resulting wind roses show a general To separate eolian and noneolian deposits, D. B.
westerly wind movement and a spread expectable Thompson (1969, p. 271) used five criteria, which,
for dunes, such as dome-shaped and transverse "when taken as a whole rule out the possibility of
type, formed under a unidirectional wind regime. fluvial or other environmental interpretations": (1)
the recognition of groups and arrangements of sedi-
Environment of Deposition mentary structures of specific dune types; (2) the
presence of very large sets (up to 13 m) of fine-
The eolian-type sandstone, referred to as the grained cross-stratified sandstone; (3) uniformly
"Frodsham facies" by D. B. Thompson (1969, p. fine grained texture of the sandstone; (4) a general
286), was interpreted by him -largely on the basis absence or virtual absence of feldspar and mica,
of cross-strata analysis - to consist principally of even though these minerals are common in the
dome-shaped dunes, attributed to strong winds, associated fluvial lithofacies; and (5) accompanying
and, to a lesser extent, of transverse ridge dunes concentrations of "millet seed" sand.
from more moderate breezes. Both dune types
result from unidirectional paleowinds that, as The first two and the fourth of Thompson's cri-
determined from cross-strata dips, came from the teria seem to be compelling positive evidence of
east. Fluctuating crosswinds were largely from the eolian deposition; the third and fifth, although com-
southeast and southwest. patible, may also occur in some other environ-
ments. Important additional evidence of dune en-
Various lithofacies associated with the "Frod- vironment is the recognition in the stratification of
sham facies" were recognized and interpreted by
bounding surfaces of erosion, formed through inter-
D. B. Thompson (1969, p. 267): interbedded red dune processes. Furthermore, the interbedding of
mudstones, red-brown shaly mudstones and eolian- and fluviatile-types is strongly suggestive of
siltstones, all considered to be of fluvial and
blown sands overwhelming the surfaces of river
lacustrine origin and possibly formed between flood plains or of ponded (sebkha) areas.
dunes as interdune ponded deposits. Other
deposits, with clay galls, mica concentrates, and Barun Goyot Formation
argillaceous units, are also undoubtedly of suba-
queous origin. Brown shaly flat-bedded and ripple- (Upper Cretaceous),
bedded strata among the cross-stratified sandstones
resemble modern interdune deposits recorded Mongolia
elsewhere.
The "Keuper" waterstones immediately above UPPER CRETACEOUS STRATA called the
the Frodsham Member are partly intertidal, sug- Barun Goyot Formation, formerly the Upper
gesting that the dome-shaped transverse dunes Nemegt Beds, in the northern part of the Gobi
might represent a coastal sand belt. The negative Desert of Mongolia, are reviewed herein because
skew of the grain-size distribution of some of the they include exceptionally extensive interdune
samples suggests derivation from a beach. On the deposits and contain a large and varied nonmarine
other hand, the basinwide areal extent and geome- fauna for a paleodesert environment. The
try of the Frodsham Member suggests the existence sandstone of this formation is considered to be
of a sand sea followed by a basinwide transgres- eolian not only because of characteristic features of
sion. structure and texture attributed to wind deposition
in many other deposits but also because of its
Summary of Criteria for Eolian Origin associated facies and its abundant nonmarine
The part of the "Keuper Sandstone Formation" fauna.
called the "Frodsham facies," which is believed to Continental deposits of Late Cretaceous age in
be composed dominantly of dome-shaped and central Asia have been known for many years and
transverse dunes, is intimately associated with have received considerable fame because of the
facies representing contrasting depositional en- dinosaur and mammalian remains obtained from
vironments. The closeness of several distinct facies them during a series of paleontological expeditions
may lead to misinterpretations and confusion by American, Soviet, and Polish-Mongolian groups
where differentiation is not carefully made. in the 1920's, 1940's, and 1960's, respectively. The

Digitized by Google
 ANCIENT SANDSTONFS CONSIDERED TO BE EOUAN 231
data on sedimentology reviewed in the following weathered are mixed together (Gradzinski and
synthesis of the Barun Goyot Formation are largely Jerzykiewicz, 1974a, p. 253).
from the descriptions and interpretations by
Gradzinski and Jerzykiewicz (1972, 1974a, b). Primary Structures
Stratification
Distribution and Thickness
In the Baron Goyot Formation. five sediment
Dune-type deposits of the Baron Goyot Forma- types, based on differences in both lithology and
tion are widely exposed in the Nemegt Basin, an primary structures. were recognized by Gradzinski
east-trending graben in the Gobi and adjoining and Jerzykiewicz (1974b). Of these types. the
areas of central Asia. The formation has been "mega cross-stratified units" and possibly the
studied in detail in two main places - the Nemegt "massive, 'structureless sandstone'" are probably
locality, to the west, and the Khulsan locality, to the of eOlian origin. The "flat-bedded sandstones,"
east, covering an area of "about 100 sq. km [260 "diversely stratified sandstones." and "alternating
mi2]" (Gradzinski and Jerzykiewicz, 1974b, p. 113). claystones and sandstones" probably are mostly or
The total thickness of the Baron Goyot Forma- entirely interdune deposits.
tion has been given by Gradzinski and Jerzykiewicz The "mega cross-stratified sandstone." as
(19748, p. 252) as "about 100 meters" (300 ft) and described (Gradzinski and Jerzykiewicz. 19748. p.
described as grading upward into "fluvial deposits" 259). seems to correspond in all essential structural
of the Nemegt Formation "about 400 m [1,200 feet] features to the large-scale high-angled cross-strata
thick." The Baron Goyot Formation. however. is of other sandstones considered to be of eolian-type
composed only in part of large-scale cross-strata. that are discussed in this chapter. Its cross-strata
.considered to be eolian. and much of its thickness dip at high angles (generally 25- 30. maximum
comprises interdune deposits of other facies. 36). They are mostly tabular planar. less com-
monly wedge planar. and rarely trough shaped.
Composition and Texture Bounding planes between sets of cross-strata are
The results of grain-size distribution studies by mostly even and horizontal for long distances and
Gradzinski and Jerzykiewicz (19748. p. 263; 1974b. form sets of cross-strata 3-10 m (9- 30 ft) thick.
p. 122. table 1. fig. 5) are summarized in their table mostly greater than 5 m (15 ft). The length of in-
of statistical parameters and in histograms. From dividual cross-strata commonly is considerable-
these analyses, sand of the eolian-type cross-strata several tens of metres (90 ft or more). Dip direc-
is shown to be mainly fine- or medium-grained. tions are relatively constant. with variations less
moderately sorted and slightly skewed on the posi- than 20.
tive side. Because the "mega cross-stratified sandstones"
Studies of sphericity, roundness. and surface tex- grade both laterally and vertically into "massive.
ture were also made of sand grains from each prin- 'structureless' sandstones." some of the latter may
cipal facies. Subrounded and rounded grains pre- also represent wind-deposited sand, but internal
dominated in most samples. although some structures in them have largely been destroyed.
subangular and some well-rounded grains were "Indistinct traces of internal structures are visible
noted (Gradzinski and Jerzykiewicz. 19748. p. 257). in some beds * * * [including] horizontal lamination
Mean values for sphericity ranged from 0.66 to 0.72. and deformational structures of overturned folds"
Surface frosting was recorded for 21- 45 percent of (Gradzinski and Jerzkiewicz, 1974b. p. 134). Thus.
the grains. and both ridges and pitting were not un- slumping or other types of contortion may have
common on sands of the large-scale cross-strata. contributed to eradication of primary structure.
Composition of the sandstone. calculated from Contorted Bedding
eight analyses. was 69- 82 percent quartz. 18- 28
percent feldspar. and 1- 4 percent other minerals. Within sets of large-scale cross-strata -between
including mica and heavy minerals. principally sets that are undeformed - are overturned folds
epidote. The feldspar is mostly fresh but locally and other varieties of contorted and disrupted bed-
weathered. and in some samples fresh and ding indicating a penecontemporaneous origin of

Digitized by Google
232 A STUDY OF GLOBAL SAND SEAS

the structures (Gradzinski and Jerzykiewicz. 19748. faces (Gradzinski and Jerzykiewicz. 1974b. p. 134).
p. 263. fIg. 11a). Deformational structures also are All these features suggest deposition by stream cur-
reported from the "structureless" sandstone facies rents.
suggesting that processes of contortion may be The facies described as "alternating claystones
responsible for the partial or complete obliteration and sandstones" consists of "quiet water" deposits
of primary structures in parts of the Barun Goyot as indicated by bedding that is commonly Oat or
Formation (Gradzinski and Jerzykiewicz. 19748. p. wavy. a scarcity of crossbedding. truncation of un-
274; 1974b. p. 134). derlying eolian-type cross-strata. strong deforma-
Detailed descriptions of the "contorted and dis- tion of mud layers. and deformational features in
rupted" cross-strata in the Barun Goyot Formation sand. such as load casts. disrupted sandstone beds.
are not available; thus. it cannot yet be determined and isolated sandstone blocks.
whether they are structures resulting from loading An additional facies of the Barun Goyot Forma-
in loose sand. of the types described from Brazil tion. referred to as "flat bedded sandstone" by
(McKee and Bigarella. 1972. fig. 4). or whether they Gradzinski and Jerzykiewicz (1974b.p. 125). OCCUIB
were formed by avalanching of sand deposits. as mostly in the upper part of the formation and pro-
demonstrated in laboratory experiments (McKee. bably represents a major change or shift in environ-
Douglass. and Rittenhouse. 1971. p. 372). If the ment following eolian deposition. It consists of
Barun Goyot sands include overturned folds or units a few metres thick composed mostly of
overthrusts. the noncohesive or cohesive nature of sandstone. with some siltstone and sandy
the sand at the time of deformation might be deter- claystone. Bedding is largely horizontal or low
mined. and. from this information. inferences could angle (2-5). The low-angle beds occur between
be made concerning the dryness or wetness of the horizontal sets and commonly flatten laterally. Ero-
environment. sion channels are rare. This facies is interpreted as
playa type (Gradzinski and Jerzykiewicz. 1974b. p.
Interdune Deposits 129-140).
Strata interpreted as interdune deposits are
abundantly represented in the Barun Goyot Forma- Fauna
tion. They consist of facies referred to by No other eolian-type sandstone has produced so
Gradzinski and Jerzykiewicz (1974b. p. 136) as rich and varied a fauna as has the Barun Goyot of
"diversely stratified sandstones" and "alternating central Asia. From the sandstones of this formation
claystones and sandstones." The first of these types "several specimens of dinosaurs. very numerous
occurs almost exclusively in the lower part of the dinosaur eggs. a few crocodiles. several tortoises. a
formation. where it fIlls erosion channels as much pterosaur. about 80 specimens of lizards. about 50
as 3 m (9 or 10 ft) wide. scoured in the large-scale specimens of mammals .* * * were found within a
cross-strata of eolian type. The "alternating few weeks in the Khulsan and Nemegt localities"
claystones and sandstones." in contrast. are mostly (Gradzinski and Jerzykiewicz. 1974b. p. 140).
sheetlike. horizontal or wavy strata that interfInger Most of these fossils were found in nondune
with the eolian-type large-scale cross-strata and. facies - that is. in deposits believed to be inter-
therefore. were formed contemporaneously with dune consisting of various water-deposited sedi-
them. ments. especially those attributed to intermittent
Textural and structural characteristics of the streams and lakes. However. in the large-scale
"diversely stratified sandstones" are (1) the com- cross-strata of dune-type sandstone. a few dinosaur
mon inclusion of small. locally derived pebbles. (2) skeletons and numerous dinosaur eggs have been
the incorporation of a few large foreign pebbles and found. Furthermore. in the massive "structureless"
cobbles and numerous skeletal remains of verte- sandstones "in the immediate neighborhood of
brate animals. (3) poorly sorted sandstones. (4) mega cross-stratified units." nearly complete
common trough-type cross-strata. (5) wedging of dinosaur skeletons and many intact dinosaur eggs
bedding units within short distances. (6) large were found. The distribution of these fossils clearly
variation in the dip directions .of cross-strata. and seems to have been controlled by ecological con-
(7) the filling of erosion pockets and scoured sur- siderations - the wet or moist vegetated interdune

Digitized by Google
 ANCIENT SANDSTONES CONSIDERED TO BE EOUAN 233

areas being far more favorable, both for the living Formation were wind deposited was summarized
fauna and for its preservaiton than the bare dune by Gradzinski and Jerzykiewicz (19748, p. 272;
environment (Gradzinski and Jerzykiewicz, 1974b, 1974b, p. 137) - the considerable thickness of in-
p.141). dividual sets, the high-angle dips of most cross-
strata, the tabular-planar form that predominates
Environment of Deposition among crossbedding sets, and the wide, even form
Interpretation of the depositional environments of major bounding surfaces. Furthermore, the
represented in the Barun Goyot Formation, sup- relatively constant dip direction of cross-strata and
ported by a considerable amount of evidence, was the lateral continuity of sets strongly suggest that
presented by Gradzinski and Jerzykiewicz (1974a, the sandstone is formed of dunes of the transverse
p. 272). The thick units of large-scale cross-strata ridge type developed under the influence of
(termed "mega cross-stratified units") are con- unidirectional winds.
sidered, principally on the basis of texture and Textural features - fine- to medium-grain size,
structure, to have formed as wind-deposited dunes. moderate sorting (good to fair on the Payne scale),
Furthermore, from the direction and narrow spread and common grain-surface features including frost-
of cross-strata dip directions, they are believed to ing and pitting - are all characteristic of most
have resulted from unidirectional, westerly winds modern dune deposits.
and to have been primarily of the transverse-ridge Additional data bearing on the environment of
dune type. deposition are the associated and intertonguing
Interdune deposits of at least two varieties are in- deposits of flat-bedded and wavy-bedded sands and
terspersed within the windblown sands. Some of muds, the thin conglomerate beds, the abundant
these interdune deposits apparently represent a nonmarine fauna (including dinosaur skeletons and
fluvial variety, formed by flooding of interdune eggs), and other features of interdune deposits that
areas; they include both "upper flow regime" sands are interspersed among typical dune-type
(largely horizontal beds) and "lower flow regime" sandstones.
types. These deposits - the "diversely stratified
sandstones" of Gradzinski and Jerzykiewicz (1974b, Botucatu and Sambaiba
p. 139) - include many gravels, channel-fill sedi-
ments, wavy laminae, trough bedding, and other in-
Sandstones of South America
dicators of current action. (Jurassic and Cretaceous)
A second variety of interdune deposits consists By JOAO J. BIGARELLA
of alternating claystones and noncrossbedded ,
sandstones. These strata, deposited synchronously THE BOTUCATU SANDSTONE of Brazil con-
with the eolian sandstones, probably formed in in- stitutes one of the world's largest and most
termittent lakes (Gradzinski and Jerzykiewicz, widespread sequences of sandstone of probable
1974b, p. 136, 138). eolian origin (fig. 159) - an extent of more than 1.3
The relatively great thickness, 2-4 m (6-12 ft), million km2 (800,000 mi2). This sandstone not only
of both fluvial-and lacustrine-type interdune units, occurs throughout the Parani Basin (States of Sio
suggests either that dune migration and burial of in- Paulo and Parani), but also extends beyond this
terdunes was slow or that interdune sedimentation basin's present edge into the States of Mato Grosso
was much more rapid than for other ancient eolian and Minas Gerais. Probably, the Botucatu
sandstones described herein. The clear evidence of Sandstone originally connected with similar
ponding and the abundance of fossil vertebrate re- sandstones of the Parnaiba Basin in northern Brazil
mains suggest considerable vegetation in the inter- (States of Maranhao and Piauf). The present dis-
dune area. tribution of these sandstones is shown in figure 159.
In northern Brazil, the Sambaiba Formation of
Summary of Criteria for Eolian Origin the Parnaiba Basin is considered to be equivalent to
the Botucatu sequence. This formation is composed
Evidence that large-scale high-angle cross-strata of large-scale crossbedded sandstones, alternating
("mega cross-stratified units") of the Baron Goyot with probable subaqueous deposits.

Digitized by Google
234 A STIJDY OF GLOBAL SAND SEAS

o 600 KlOMETRES
I
I I
o 300 MlfS
Scale is approximate

''

INDEX MAP SHOWING OUTLINES OF AREAS

A ANDB

OUTCROPPING AREAS OF SELECTED MESOZOIC eolian sandstones (dar\< stipple) in South America: A, Sambaiba Sandstone; B,
Botucatu Sandstone. (Fig. 159.)

The Botucat6 Sandstone. first described by designated it as the Caiua Sandstone (Washburne.
Gonzaga de Campos (1889). includes all large-scale 1930; Maack. 1941). but others (Almeida. 1954) con-
cross-stratified sandstone units intercalated with or sider it to be a continuation of the Botucat6
underlying the Serra Geral basaltic lava flows. Sandstone.
Another sequence of large-scale cross-stratified
sandstones of probable eolian origin overlies The thickness of the Botucatu Sandstone is
basaltic flows in the States of Mato Grosso. sao generally 10- 20 m (30- 65 ft) but is locally more
Paulo. and Paran' (fig. 159). Some authors have than 30 m (100 ft).

Digitized by Google
 ANCIENT SANDSTONFS CONSIDERED TO BE EOUAN 235
The Botucatli Sandstone was long considered to samples contain three or four textural classes with
be Triassic-Jurassic in age. However, radiometric more than 1 percent frequency. In the Botucatu
age determinations of the associated and intercal- Sandstone three classes predominate in much of the
ated basaltic rocks indicate an average age of formation, but close to source areas in the
120-130 m.y. for the lava flows (Cordani and Van- crystalline shield many of the samples include four
doros, 1967), and so its age is now considered to be textural classes. Sandstones - supposedly
Late Jurassic -Early Cretaceous. fluviatile but intercalated in the dune sequence-
contain as many as seven textural classes.
Facies In the Botucatli paleodesert, the modal class, ac-
The sandstone facies of the Botucatu, with large- cording to recalculated analyses, is mostly fine-
scale cross-stratification considered to be eolian, is grained sand (0.125-0.25 mm) and includes some
in some places interbedded with another facies, the medium-grained sand (0.25- 0.5 mm). The
Piramboia, believed to be fluviatile (Pacheco, 1913). medium-grained sand is mainly near the crystalline
This fluviatile facies generally occurs stratigraph- shield source area; only in a few places is it down-
ically below eolian sandstones and probably repre- wind, toward the central part of the paleodesert.
sents intermittent stream deposition in small basins The average mean diameter of recalculated sam-
without outlet. ples of large-scale cross-stratified Botucatu
Another supposed fluviolacustrine or playa-lake Sandstone decreases southward from 0.209 mm in
sequence, which is interbedded with the large-scale Minas Gerais, to 0.183 mm in sro Paulo and to 0.178
crossbedded eolian sandstones of the BotucaUi, has mm in Parana; it also decreases northward from
been called Santana facies (Almeida and Barbosa, 0.233 mm in Rio Grande do SuI to 0.215 mm in San-
1953, p. 64). "The Santana facies contains a Con- ta Catarina. In both series grain size decreases
chostraca (branchiopod) assemblage consisting of downwind from crystalline areas. The sandstones
Bairdestheria barbosai; Palaeolimnadia petrii; of the northern area of the Botucatu paleodesert are
Euestheria mendesi; Pachecoia acuminata; P. generally finer grained (0.186 mm) than those in
rodriguesi; Candonopsis pyriformis candona?; the southern area (0.218 mm). The northern and
Estheriella sp." southern areas were under the influence of
Sandstones of probable fluviatile origin occur in different paleowind regimes. The wind in the
the lowermost part of the Botucatu sequence and southern area was stronger, as revealed by larger
are interbedded with the large-scale cross-stratified cross-strata, larger sand grains, and larger average
sandstones of probable eolian origin. They crop out mean grain-size diameter.
in many places in the States of Parana, Sao Paulo, The Botucatu Sandstone is dominantly moder-
and Minas Gerais. ately to well sorted. Sorting mostly is progressively
The equivalent of the Botucatu Sandstone is better downwind - that is, away from the
known in Uruguay as the Tacuaremb~ Sandstone crystalline shield areas. Trask sorting coefficients
and in Paraguay as the Misiones Formation. (So) decrease from Minas Gerais (So = 1.44) to Sio
However, not all the strata mapped as Misiones Paulo (So = 1.33) and Parana (So = 1.32). They
Formation (Eckel, 1959) are actually equivalent to decrease also from Rio Grande do SuI (So = 1.46) to
the Botucatu Sandstone. Santa Catarina (So = 1.39). The samples from the
northern area of the Botucatu paleodesert are
Texture slightly better sorted than are those from the
In the Parana Basin of Brazil, for a distance of southern area. As the mean grain-size diameter in-
2,500 km (1,550 mil along the eastern belt of out- creases, the sorting coefficient becomes poorer.
crops, 99 samples of eolian-type Botucatu Ventifacts are scattered along some zones of the
Sandstone were collected for textural studies. Botucatu Sandstone.
Because most samples contained tails of fine- Primary Structures
grained material (silt and clay) introduced by
weathering or by postdepositional impregnation, Stratification
grain-size distribution was recalculated, based on Cross-stratification typical of eolian slipface
quartz grains, to be 0.062 mm. The recalculated deposits is the outstanding sedimentary structure of

Digitized by Google
236 A STUDY OF GLOBAL SAND SEAS

the Botucatu Sandstone and is abundant in almost Crossbedding measurements of the Sambaiba
all outcrops. The size of sets may range from small Formation. which is the Botucatu Sandstone
to large in the same section. Large sets of cross- equivalent in northern Brazil. indicates that pre-
strata with lengths of dipping beds of 30 m (100 ft) vailing winds responsible for sand movement were
or more and heights of about 12-15 m (4O- 50 ft) the southeasterly tradewinds. The Botucatu and
occur in many localities. Sambaiba eolian sandstones were deposited in a
low-latitude paleodesert.
Environment of Deposition
Paleowind circulation in the Botucatu paleo- Summary of Criteria for Eolian Origin
desert was investigated by Almeida (1953. 1954). Criteria responsible for considering major parts
Bigarella and Salamuni (1961). and Bigarella and of the BotucatU and Sambaiba Sandstones to be of
Oliveira (1966). Measurements in the correlative eolian origin consist of sedimentary structures and
Sambaiba Formation of northern Brazil were made grain textures. The most compelling structural evi-
by Bigarella. Montenegro. and Coutinho (unpub. dences are the very large scale and the high angles
data). common to much of the wedge-planar and tabular-
Crossbedding dip directions in the southern part planar cross-stratification. Textural features are en-
of the BotucatU paleodesert (Santa Catarina and Rio tirely compatible with this interpretation.
Grande do SuI in Brazil. and Uruguay) indicate that Throughout a large part of Brazil, the great extent
paleowinds blew from the west and west-south- of moderately sorted to well-sorted sand, which is
west. These winds were deflected toward the north dominantly of fine to medium size, can be ex-
in southern Paraguay. In the northern part of the plained by few environments other than eolian.
Botucatu paleodesert (Mato Grosso. Goie, Minas Additional evidence of genesis is furnished by scat-
Gerais, Slo Paulo, and Parana in Brazil, and part of tered ventifacts.
Paraguay), crossbedding measurements indicate
mean paleowind trends from the north or north-
northeast, representing return tradewinds. In
Cave Sandstone and
Paran~, these northerly paleowinds were then Similar Sandstones of
deflected toward the west.
In the Botucatu paleodesert, southern air masses
Southern Africa (Triassic)
,.,
moved northward, leaving their traces mainly in By JOAO J. BIGARELLA
Paranfl and 810 Paulo. and. less frequently. the
northern air masses penetrated the southern areas; THE SUCCESSION OF GONDWANA BEDS in
so, cold air masses and warmer ones probably con- southern Africa is known as the Karroo System.
flicted. Rainfall was therefore probably greatest The basal beds, named the Dwyka Group, are in-
north of Santa Catarina. This postulate is supported terpreted as glacial and periglacial deposits. The
by deposits in that region of the Piramboia and San- uppermost sedimentary beds are believed to have
tana facies. which seem to represent subaqueous been deposited under arid or semiarid conditions,
environments. Such deposits. which are at the base although much of the sand was deposited in suba-
of, and interbedded with. supposed eolian deposits. queous environments. These uppermost strata are
are common in Parana and 880 Paulo but decrease known by various names - Cave Sandstone in the
in abundance northward. Waterlaid deposits are Republic of South Africa, Botswana, Lesotho, and
rare in the southern part of the Botucatu Swaziland; the Bushveld Sandstone in the central
paleodesert (Bigarella and Salamuni. 1961). Transvaal; Etjo Sandstone in South-West Africa;
Crossbedding measurements in Parana have a and the Forest and Nyamandlovu Sandstones in
very low consistency ratio. probably mostly the Rhodesia. They are mostly Triassic in age but pro-
result of the conflict between opposing wind direc- bably represent a range of about 75 m.y.
tions, westerly and northerly. The area of conflict These sandstones are conformably overlain by or
between these winds extends westward into intercalated in their upper parts with lavas of
Paraguay. various compositions. Sedimentation continued

Digitized by Google
 ANCIENT SANDSTONm) CONSIDERED TO BE EOUAN 237
after the first outpourings of lava. The com- Sandstone. However, in most localities in Lesotho,
paratively thin layers of sandstone between flows Orange Free State, and along the Lebombo Moun-
indicate the absence of a major unconformity. In tains of eastern South Africa and Swaziland, this
places, as in northern Transvaal, the prelava sandstone is not clearly stratified. The unstratified
sandstone surface is very uneven, possibly reflect- layers may represent deposition by flash floods that
ing an uneven dune topography or contem- originated in a semiarid climate. Judged by texture,
poraneous erosion by running water. The main out- these layers probably consist of either mudflows or
crops of the sandstone in Lesotho were studied by sand flows.
Beukes (1970), who divided the sequence into three Probably the Cave Sandstone from Lesotho and
parts: lower and middle massive to thick-bedded the Lebombo Mountains should be renamed
sandstones, and an upper sandstone containing because it and the Bushveld Sandstone are much
large-scale cross-stratification. In Rhodesia the up- older than either the well-cross-stratified eolian
per part is called the Nyamandlovu Sandstone, and sandstone of northern Transvaal near the border
its cross-strata are generally considered to be of with Botswana or the Etjo Sandstone of the Huab
eolian origin. Valley.

Distribution and Thickness The Etjo and Cave Sandstones are characterized
by eolian-type cross-stratification, mostly large to
The Cave Sandstone and equivalents in the
medium wedge-shaped sets, interpreted as products
Republic of South Africa, Lesotho, and Rhodesia
of transverse and barchan dunes. Some convex-up-
conformably overlie a succession of red and purple
ward strata may be related to parabolic dunes or
mudstones, shales, and sandstones. The Cave
other types of eolian bedforms which produce con-
Sandstone attains a maximum thickness of about
vex-upward structures (that is, retention dunes,
350 m (1,150 ft) in the Orange Free State and in
protuberances on transverse dunes).
northern Transvaal, but in most other places, it is
less than half this thickness. In Rhodesia it is Within the same section, the size of the cross-
relatively thin; in South-West Africa it generally is strata sets ranges from small to large. Cross-strata
60-150 m (200- 500 ft) thick and attains about 600 consist of the wedge-planar and tabular-planar
m (2,000 ft) in maximum thickness. types. Dips of the strata very seldom exceed 33; the
maximum dip recorded is 40. The average dips are:
Composition and Texture
Cave Sandstone (around Lesotho) ..... 16.1
The composition of the Cave Sandstone differs Cave Sandstone (northern Transvaal) .. 21.9
considerably from place to place. In the upper Etjo Sandstone ....................... 21.9
cross-stratified part, detrital grains include 60-73
percent quartz and 1.0- 9.3 percent feldspar.
Miscellaneous matrix ranges from 10 to 21 percent Fauna
(Beukes, 1970). The remainder of the rock consists Remains of reptiles, crustaceans, fish, insects,
of rock fragments, silica cement, and heavy and plants indicate an age of Late Triassic or older
minerals. for the Cave Sandstone in South Africa and
The mean grain size of the cross-stratified Rhodesia (Haughton, 1969, p. 398). The remains of
sandstones ranges from 0.170 mm on the north- charred trees just below the first lava flow occur in
western side of Lesotho to 0.060 mm on the the Orange Free State and reptilian tracks are fairly
southeastern side. common in South-West Africa, Lesotho, and the
Primary Structures Orange Free State. Sharp, clear tracks of at least
five different types of reptiles were recently dis-
Stratification covered in the northern Transvaal on dune cross-
Along the Huab Valley in South-West Africa, strata (foresets) having dips of about 27. At the
and also in parts of northern Transvaal, dunelike Kariba Dam in Rhodesia, reptilian remains occur in
structures are very prominent in the Cave sandstone intercalated in lava.

Digitized by Google
238 A STUDY OF GLOBAL SAND SEAS

Environment of Deposition the Cave Sandstone interpretation as eolian de-

pends not only on its characteristic cross-strata pat-
Although the climate probably varied considera-
tern but also on associated environmental features.
bly, the main bulk of the sandstone apparently was
As in most eolian sandstones, its sorting is better
not deposited under extreme desert conditions. In
than for subaqueous sandstones. The Cave is
many places, the presence of water is indicated by
neither associated nor intercalated with sequences
current ripple marks, erosion channels, and lenses of siltstone and claystone, and it does not include
of shale probably deposited in old depressions.
gravel beds.
Summary of Criteria for Eolian Origin Convex-upward cross-strata are believed to be
typical eolian features developed in some types of
Although sandstones that developed in different dunes. Similar types of cross-strata rarely occur in
environments may show similar structural features, subaqueous sandstones.

Digitized by Google
Digitized by Google
Digitized by Google
 A STUDY OF GLOBAL SAND SEAS

SEDIMENTS OF THE ANCIENT EOLIAN

ENVIRONMENT-RESERVOIR
INHOMOGENEITY
Chapter I

By ROBERT LUPE and THOMAS S. AHLBRANDT

Contents
Page

Summary of conclusions .......................... . 243 The model .. . . . . . . . . . . . . . . . . . . . . . . .. ............. 245

Introduction ..................................... . 243 The model and the subsurface situation . . . . . . . . . . . . . . 245
Acknowledgments ................................ . 245 Application. . . .. . . . . .. .. . . .. . . .. . . .. . . . . . . . . . . . . . . 247

Illustrations
Page
FIGURE 160. Schematic diagram and cross section of typical distribution of dune, inter-
dune, and extradune deposits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 244
161. Graph showing porosity-permeability distribution of dune and interdune sedi-
ments from cores, White sands National Monument, New Mexico, U.S.A. 245
162. Graph of porosity versus permeability, showing delineation of three environ-
ments of deposition ................................................ 246
163. Electric logs showing eolian sedimentary structures and depositional environ-
ments ............................................................. 248
164. Cross section of depositional environments interpreted from cores of the
Tensleep Sandstone ................................................ 250
165. Diagram of core permeability and porosity and environment of deposition in
the Weber Sandstone ............................................... 251

241

Digitized by Google
Digitized by Google
 SEDIMENTS OF THE ANCIENT FDUAN ENVIRONMENT-RESERVOIR INHOMOGENEITY 243

Summary of Conclusions Introduction

?12 PPER PALEOZOIC TO MESOZOIC eolian MANY OF THE UPPER PALEOZOIC to
~ blanket sandstones of the Colorado Plateau Mesozoic blanket sandstones of the Colorado
and the Rocky Mountains of Colorado and southern Plateau and the Rocky Mountains of Colorado and
Wyoming are texturally complex. As petroleum southern Wyoming were deposited in eolian and
reservoirs, they commonly have poor performance closely associated noneolian environments. They
histories because they represent a depositional are the reservoir rocks in many oil and gas fields.
system comprised of three closely associated These sandstone units are widespread and of
subenvironments - dune, interdune, and ex- uniform thickness and seem texturally
tradune. Sediments of each subenvironment have homogeneous. There are, however, complex tex-
different textural properties which have resulted tural variations within the units that may affect
from different depositional processes. Compared porosity and permeability and be detrimental to
with interdune and extradune sediments, dune their performance as reservoir rocks. These tex-
sediments are generally more porous and permea- tural differences are related to environments of
ble and may be better quality reservoirs. Extradune deposition, and a conceptual model is proposed to
sediments (a new term) include all deposits adja- explain them. The model will aid the geologist in
cent to a dune field and are mainly subaqueous evaluating porosity and permeability problems and
deposits. Interdune sediments are here restricted to result in significant savings in time and money.
those nondune sediments deposited in the The eolian depositional system contains three
relatively flat areas between dunes. A conceptual subenvironments which are related both in time
model is proposed, in which dune sediments may and by source: dune, interdune, and extradune. The
be enveloped by extradune sediments as the term "dune" is applied in the classic geomorphic
depositional system evolves, resulting in a tex- sense of Bagnold (1941). The term "interdune" is
turally inhomogeneous reservoir that has poor fluid restricted to the description of the relatively flat
migration properties. area between the dunes of a dune complex. The
This model of textural inhomogeneity in eolian new term "extradune" describes the area extending
blanket sandstones was applied to the Weber beyond the dune field, including some areas pre-
Sandstone in the Brady field and its equivalent, the viously referred to as interdune. Interdune sedi-
Tensleep Sandstone, in the Wertz and Lost Soldier ments probably were originally deposited by the
fields, Sweetwater County, Wyoming, U.S.A. Data same wind activity that built the dune (McKee and
were obtained from both outcrop and subsurface Moiola, 1975) and are commonly reworked by
and included environmental interpretation, tex- water or wind (Glennie, 1970). They are commonly
tural analysis, and plotting of the distribution of preserved as the flat-lying sediments at the base of
depositional subenvironments. As expected from the crossbed sets of a dune sequence (fig. 160). Ex-
the model, dune sediments in Brady field were tradune sediments were probably deposited by
markedly different in texture from interdune and processes independent of the dune-building process
extradune sediments. The predicted geometric dis- and involve sediments of diverse environments,
tribution of subenvironments was confirmed in such as alluvial fan, wadi, serir, stream, sebkha,
Lost Soldier and Wertz fields. However, in these playa, lake, beach, and tidal flat.
two fields, secondary cementation and fracturing The new term "extradune" fulfills two purposes:
has obscured the original porosity and permeability (1) It specifically describes a geomorphic area
contrasts. The distribution of porosity and per- which has often been incorrectly termed "inter-
meability, a characteristic depending partly on dune," and (2) it serves as a general term which
depositional processes, could impede fluid migra- could include a variety of subenvironments. The
tion in the reservoir and significantly reduce recov- term can be used in general reference to various
ery of hydrocarbons. subenvironments or used when the identification of

Digitized by Google
244 A STUDY OF GLOBAL SAND SEAS

/---------------'"
'\
, \
I
Playa I
I
I
I
I
I

,
\
\
\
\
\ Water
\. Boundary of extra dune area

A
'----...1- ---------------_.-
A A'

DEPOSITS

Dune Interdune Alluvial fan

E2ZJ
Nearshore
~
Fluvial
~
Playa and sebkha
surface

SCHEMATIC DIAGRAM AND CROSS SEOION of dune, interdune, and extradune deposits. A, Plan view of typical distribution of
dune, interdune, and extradune deposits. B, Cross section illustrating evolution of a dune-extradune system, ending with the
geographic configuration of A. Subenvironments of the system migrate laterally as well as vertically, resulting in isolation of more
porous and permeable dune sediments . (Fig. 160.)

Digitized by Google
 SEDIMENTS OF THE ANCIENT EOUAN ENVIRONMENT-RESERVOIR INHOMOGENEITY 245

specific subenvironments is uncertain but the 1001hr---r---,-.....---.,.-~---r-,....--.---r--"""T'"-.--..,

stratigraphic relation to dune sediments is clear.
o DUNE
Acknowledgments

I
"INTERDUNE

WE PARTICULARLY THANK Amoco Produc-

tion Co. for their support and gracious granting of
..
time and expense. We also thank Champlin ~
Petroleum Co., operator of Brady field, and Moun- ~100
..
I
tain Fuel Supply Co. and Exxon U.S.A., the other
Brady field partners; and Hilliard Oil and Gas Co.
for permission to publish data in this report. Pasco,
Inc., provided data for the Lost Soldier -Wertz field
studies.
The Model
THE DEGREE OF TEXTURAL contrast in dune,
interdune, and extradune sediments reflects DISTRIBUTION OF POROSITY and permeability in dune and
source-material texture, depositional processes, interdune sediments from cores, White Sands National Monu-
ment, New Mexico, U.S.A. (Fig. 161.)
and the spatial relations of subenvironments within
the depositional system. the degree of isolation depending on the original
Dune sediments are relatively better sorted, bet- distance between dune complexes and on the way
ter rounded, and more compositionally mature than in which succeeding depositional systems overlie
are sediments of most other depositional environ- one another. Movement of fluid would be affected.
ments (Bagnold, 1941; Twenhofel, 1945; Kuenen, Furthermore, the porosity and permeability con-
1960; Kukal, 1971; Ahlbrandt, chapter B, this trast may be enhanced by the effects of burial (Rit-
paper). And, because porosity and permeability in tenhouse, 1971).
modem sediments are largely functions of particle
size, shape, and sorting (Griffiths, 1967), dune sand
is generally more porous and more permeable than
are the source, interdune, or extradune sediments.
The Model and the Subsurface
Extradune sediments are predominantly sub- Situation
aqueous and, therefore, are usually more poorly
sorted and less porous and permeable than are dune EXTENDING THE DUNE-EXTRADUNE model
sediments. Additionally, porosity and permeability for eolian blanket sands to the subsurface has po-
in these sediments could be further reduced tentially valuable results for the petroleum indus-
because the topographically lower interdune and try. Two preliminary steps are required - first,
extradune areas are commonly the sites of early ce- recognition of sedimentary environments from con-
mentation (Glennie, 1972). Figure 161 shows thp ventional well-bore data, and, second, recognition
distribution of porosity and permeability in modem of the relations between depositional environments
dune-interdune sediments; figure 162 illustrates the and the distribution of porosity and permeability, as
distribution of porosity and permeability for an- just described. Then, the size and distribution of
cient dune-extradune sediments. various subenvironments must be determined. In
Dune and extradune sediments abut laterally, some places the spatial relations of subenviron-
and successive dune-extradune depositional ments can only be approximated from outcrop and
systems may overlie one another, with dune sands (or) subsurface information. Outcrop data may not
in vertical contact either with other dune sands or resemble the actual subsurface situation, and sub-
with extradune sediments (fig. 160). Dune sand surface data may not be geographically dense
bodies may become isolated from one another, with enough to give a complete picture.

Digitized by Google
246 A STUDY OF GLOBAL SAND SEAS

------ DUNE
- - - INTERDUNE AND EXTRADUNE
(EXCEPT SHOREFACE)
- - - SHOREFACE

o 2 4 6 8 10 12 14 16 18
CORE POROSITY, IN PERCENT

GRAPH OF POROSITY VERSUS PERMEABILITY, showing delineation of three known environments of deposition. Values were ob-
tained each foot from 630 feet (192 m) of core of the Weber Sandstone from three wells in the Brady field, T. 16 N., R. 101 W.,
Sweetwater County, Wyoming, U.S.A. A grid was superimposed over the plotted data, and the number of points within each square
was counted and assigned to the center of each square. Those points are data points for contours (Fig. 162.)

Conversely, subsurface information can provide crossbed sets and the flat-lying interdune and ex-
accurate quantitative data for the model. Routine tradune beds. The resistivity curve of the dual in-
log analyses include porosity and, commonly, per- duction laterolog also responded to the same sedi-
meability determinations. Cores can provide not mentary structures. The resistivity curve res-
only porosity and permeability data but also infor- ponded inversely to porosity differences - the
mation on depositional environments. Electric logs, lower the porosity, the higher the resistivity. In this
if periodically calibrated with core data, can also example, the porosity contrast resulted from
provide information on depositional environment. porosity differences of sediments deposited in
different sedimentary environments - that is,
Figure 163 is an example in which conventional more porous dune sediments versus less porous in-
logs were used to interpret environments of deposi- terdune and extradune sediments. Hence, the most
tion. The dipmeter is the most effective tool for basic suite of logs may not only provide the quan-
describing sedimentary environments (Gilreath titative data for the construction of a model but also
and Maricelli, 1964; Campbell, 1968), but the discriminate subenvironments in blanket eolian
resistivity curve of modem electric logs may also sandstones, if secondary cementation has not
provide the same data. In figure 163 the dipmeter obscured the original porosity distribution. The
log accurately recorded the high-angle dune savings of time and money could be significant.

Digitized by Google
 SEDIMENTS OF mE ANCIENT FDLIAN ENVIRONMENT-RESERVOIR INHOMOGENEITY 247

Application preserved only in Wertz and Brady fields, where

they are isolated by foreshore, shoreface, tidal flat,
A STUDY OF THE PERMIAN and Pennsylva- supratidal, or sebkha extradune sediments, and to a
nian Weber Sandstone in the Brady field and its lesser degree, interdune deposits. Furthermore,
correlative 90 miles (150 km) to the northeast, the dune sandstone was preserved in upward-thinning
Tensleep Sandstone in Lost Soldier and Wertz transgressive and regressive depositional cycles,
fields, Sweetwater County, Wyoming, tested the which are normally bounded by intraformational
application of the dune-extradune model to the pre- unconformities.
diction of reservoir complexities. The Weber and Dune sandstones were recognized by their thick
the Tensleep were selected for three reasons: (1) bedding, their even and relatively continuous
They were considered to be an eolian blanket parallel laminae, which commonly dip at high
sandstone. (2) They had a history elsewhere of re- angles, and their lack of holes from burrowing
quiring an unpredictably large number of closely animals. They are very fine grained to fine-grained,
spaced wells to drain a field. For example, at well-sorted to very well sorted sandstones contain-
Rangely field, Colorado, Chevron Oil Co. is ing frosted grains. Interdune deposits separate dune
developing the field on only 10-acre spacing sandstones and are relatively thin {less than 10 ft (3
because of poor fluid migration (Western Oil m and horizontally bedded. They contain poorly
Reporter, December 10, 1973). The cause of the sorted bimodal sandstones with frosted grains and
poor fluid migration was unknown. The 10-acre evaporite cements, commonly anhydrite. Flaser
spacing required to adequately develop the bedding and adhesions ripples were observed in
Tensleep of Little Buffalo basin field, northwest some interdune sediments.
Wyoming, described by Emmett, Beaver, and Mc- The general criteria used to differentiate ex-
Caleb (1972), was attributed to reservoir tradune subenvironments included many described
heterogeneity. The possibility of similar poor fluid by Campbell (1971). Supratidal and sebkha deposits
migration reservoir performance in the Weber contain rip-up clasts, flaser bedding, adhesion rip-
Sandstone in the Brady field or other deep fields, ples, abundant precipitates{such as anhydrite and
where drilling is costly, made the question an dolomite), birdseye structures, and desiccation
economic one. (3) The active exploration underway features. Tital flat deposits have discontinuous
in 1975 in the Weber in the Brady field and wavy nonparallel laminae and algal structures, all
elsewhere permits further testing of the dune-ex- of which are horizontally bedded and commonly
tradune model and resultant porosity and per- burrowed. Foreshore sandstones contain relatively
meability problems. discontinuous, even-parallel to wavy-parallel
The study of the Brady, Lost Soldier, and Wertz laminations, which dip less than 10 and contain
0

fields was based largely on core examination. Four rare burrows and scour-and-fill structures. They
thousand five hundred feet of core from 45 wells in are moderately to well sorted. Shoreface deposits
Lost Soldier and Wertz fields was described in are extensively burrowed and contain contorted,
detail, and depositional environments were slumped, and discontinuous wavy nonparallel to
differentiated by Reynolds, Ahlbrandt, Fox, and parallel relict laminae with maximum dips of 20 0

Lambert (1975, 1976) and by Reynolds and Fox Symmetrical wave ripples were occasionally ob-
(1976). Six hundred fifty feet of core from three served in the relict laminae. The shoreface deposits
wells in Brady field was examined by Lupe. In addi- are also distinctly finer grained than juxtaposed
tion, core porosity and permeability were measured foreshore deposits.
in the cores, and Weber Sandstone outcrops in the Although the original environments of deposi-
Uinta Mountains, 175 miles (280 km) southwest of tion and textures of the Weber and Tensleep
Brady field, were examined. Sandstones in the Brady and Lost Soldier -Wertz
Depositional environments were interpreted fields were virtually identical, the present porosity
from cores of selected wells in these fields. The and permeability of the two areas are significantly
Weber or Tensleep Sandstone is a dune-extradune different. The Tensleep in the Lost Soldier-Wertz
system composed of mixed nearshore marine and fields has been tightly cemented to an average in-
eolian sediments (fig. 164). Dune sandstones were tergranular porosity of 5 percent, and production is

Digitized by Google
248 A STUDY OF GLOBAL SAND SEAS
DtPMETER READING
Oegr_ of dip
DUAL INDUCTION LATEROLOG
o 20 40 60 80

\~
~
c: c:
j 4300
r-~

1-" '
.,440
......... ....,.
fi
~.f ~
...
I---
.....-Tail shcNvs
direction
of dip

~ -~ ~~ !
-.
4400 - ~

? I ~'
~
c: ~ ...
l'PO' .,.,
!.! .2 ~ I!
en
en
<{
a:
...,
:J
10
E
.f
G) >
4500
-ie
~~

'>
.z:= "-
.."
c: r-
~
....,.,~
~
~
'tl
c:
::l
en )
~ , ....
4600
"
"---", I~'

-~ ~,
G.~
E~ 4700
~
c:l"
.f
/'
~
1--S " I,," .. 10-
AAOCl (L ~~
,

'-..."--,. - J
<Ii
en
Q;
l " .~
0>
0>
::l 4900 - ( - Ir1
z

;::
Q'
en
~
(C - rr-,; t - ~ ,
en
<{ 5000- - ~
a:
...,
:J
~
..
- ~
c
Z
<{
;::
( S ~
~
Ir

.5
-
U
in
en
~
a:
<Ii
en
Q;
I----~ 1----- 5100
l.~
=
.-
....
r;. ~r--

t- O>
0>
IL o.!. ~
::l
Z
'0
t: ~
5200
~=-- -:..,. , , ~ T
,
8. ~ ~'. ....
i
! t
~
. ~ I

...J

\
5300 > ti
~
~

'"U'

~
en ....
0>.,
k
cC c;
-
~
....
,.
,.1-
~ ..
'",
::l-E ~

~
.J: '" ...
u~
5400
~ ...

RECOGNITION OF SEDIMENTARY STRUCTURES and depositional environments from electric logs (Hilliard Oil and Gas Co., Joyce
dual induction laterolog distinguish between dune, interdune, and extradune sediments. Note correlation of porosity with

Digitized by Google
 SEDIMENTS OF THE ANCIENT EOLIAN ENVIRONMENT-RESERVOIR INHOMOGENEITY 249
SEDIMENTARY STRUCTURE ORIENTATION SONIC LOG POROSITY.
DERIVED FROM DI~TER READING IN PERCENT
ENVIRONMENTS OF OEPOSmoN
Borehole position o 5 10 15 20

- l- I-
Not dune-related ~.
.~ ~
~,f
-
,-J Not dune-related

. !.I

Not dune-related
1~ .,::>
..
~
"c:
,..r ..J
///----- ~:...: une "
I/'J

r--i Extradune
Dune
Extradune

r
~J
Not dune-related

Data inconclusive '-= Data inconclusive

~~ _rJ Dune
Interdune
-- ~~ ,J
--
Dune
.-
~~ "7
~
Dune
n eraune .;
I/'J
c

1MII
~~
~ 1---- - -- -
-
-~
/
r-'
~
- r-- - - --
Dune

Dune
- -
II
z"
;::
Q"

~
.,
::>

Data inconclusive
~ Data inconclusive r--- 0z

....", ;::
~
- - - yune
~~
1--"- - f--
-~ Dune
I/'J

1--' .
InterdunelJiTriii
-- .;
(I)

C
I/'J
$
a:
I-
~~ ~. Dune 0>
go
.- I Extradune ...
Z
0
t::
!.
~~ I Dune ...
i
Extradune ~
~

Data inconclusive Data inconclusive

. . . 1r
0>.,
I/'J
I/'J
Not dune-related "1i E $
6~~
1=

Creek 1. SE1/.NW1/. sec. 8, T. 15 N., R. 103 W ., Sweetwater County, Wyoming, U.S.A.). Both the dipmeter and resistivity curve of
sedimentary structures - that is, with depositional environment. (Fig. 163.)

Digitized by Google
250 A STUDY OF GLOBAL SAND SEAS

WEST EAST
LOST SOLDIER OIL FIELD
TR10
C1
WERTZ OIL FIELD

428 27ABC

METRES ~ET
75

200

50

100
25

o 0

TR10
o 2 KILOMETRES C1
T.
I
I I I 3 6 26
N.
o .5 1 Mile ABC

'-1 MILE~

I~\)::]
Dune
~
Interdune
D
Foreshore
DShoreface
__
EXPLANATION

Supmldal Tidal
-----
w- - - ---
------
Lagoon
I
I
Core--Oahed where
and flat not eveil8ble
aebIcha

CROSS SECTION OF DEPOSITIONAL ENVIRONMENTS interpreted from cores of the Tensleep Sandstone in the Lost Soldier and
Wertz oil fields, Sweetwater County, Wyoming, U.S.A. From Reynolds, Ahlbrandt, Fox, and Lambert (1975,1976) and Reynolds
and Fox (1976). The top of the Amsden Formation is used as a horizontal datum. (Fig. 164.)

Digitized by Google
 SEDIMENTS OF THE ANCIENT EOUAN ENVIRONMENT-RESERVOIR INHOMOGENEITY 251

from complexly fractured strata. In Brady field.

however. much original porosity and permeability
PERMEABLfTY. POROSrTY. ENVIRONMENT
OF
remains in the Weber; production is from strata
DEPOSmON having intergranular porosity of a maximum 15-20
ABC
percent.
The combination of the studies of the two fields
13_ +--IIiE:::=-it----..!:::::;---j
creates a nearly complete picture of dune-ex-
tradune geometry. of the original distribution of
4200
porosity and permeability. and of its potential in-
fluence on fluid migration. The close control in Lost
Upper IhcInIr.ce
Soldier -Wertz (fig. 164). combined with the
Lower foreIhore
porosity-permeability contrasts resulting from the
4210 same depositional processes which are preserved in
13.900 -+-----../1::1 I - -...........~ For.hont
IrartIdII the Brady field (fig. 165). provides a subsurface ex-
SupqtlcMI ample of the dune-extradune model. Relatively
For.hontl
r----+ TIdeI flat more porous and permeable dune sand is isolated
by generally less porous and permeable extradune
Dune sediment at Brady field. so fluid migration may be
influenced.
13.950 +---=---l t==E=:::;"'1 From the study of the Weber Sandstone and ex-
4230 amination of the Nugget Sandstone in southern
Wyoming and its correlative. the Navajo Sandstone
in the Colorado Plateau. we conclude that dune-ex-
tradune systems. such as the Weber and Tensleep
4240 Sandstones. present greater reservoir problems
14.l11D -+------=""!I 1--""""'-"" than dune-interdune types. such as the Nugget.
Dune-interdune systems present less of a problem
because (1) interdune sediments are usually thin-
4250 ner than extradune sediments and are therefore
less of a porosity and permeability barrier; and (2)
interdune sediments are commonly lenticular. a
14.tEO +c:=:::::;;;;;;:::=! ""==---l condition which would allow communication be-
42SI tween dune bodies. Furthermore. judged from pro-
duction histories. porosity-permeability problems
in dune-interdune reservoirs apparently are not
Intertidll common.
4270 Effective fluid movement between well bores is
14.100 +-----flI--_~ influenced by the complex permeability barriers
that resulted from dune-extradune depositional
processes. Where fluid movement between well
~==!~vlntertldll
~ ~I bores is critical. as in retrograde reservoirs where
SupqtlcMI reservoir pressure must be maintained. the efficient
r----+-1nbIttIdI1 recovery of hydrocarbons requires knowledge of
Dune
~~~~~~-IntdWMI
14.150...L ,~ the distribution of depositional environments and
A Dune the resultant porosity and permeability charac-
B ShorefIIce
C Interdune and teristics.
other extr.ctWMI

DIAGRAM OF CORE PERMEABILITY and porosity and environ-

ment of deposition in the Weber Sandstone, Champlin
Petroleum Co. Brady Unit 2, NE1f4NE1f4 sec. 2, T. 16 N., R. 101 W.,
Sweetwater County, Wyoming, U.S.A. (Fig. 165.)

Digitized by Google
Digitized by Google
 A STUDY OF GLOBAL SAND SEAS

MORPHOLOGY AND DISTRIBUTION OF

DUNES IN SAND SEAS OBSERVED BY
REMOTE SENSING
ChapterJ

By CAROL S. BREED and TERESA GROW

Contents
Page Page
Summary of conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 linear dunes . . . . . . .. . . . . .. . . ... . .. . . .... . .. .. . . . . . 261
Introduction . . . .. . . . .. .... .. . .. ... . . . .. . . .... . . . . . 257 Crescentic dunes ..... . . . . .. . . . .. .. . . .. . .. . . . . . ... . 265
Classification of eolian sand features based on landsat Star dunes . . . . . . . . . .. . . .. . ... . .. . . . . .. ... .. . .. . .. . 275
imagery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 257 Parabolic dunes . . . .. . . .. . . .. ... . . . . . . .. . ...... . .. . 277
Genetic and local terms .. . . ... . .. . .. . ... ... . . .. 257 Dome-shaped dunes . ... . . . . . . . . .. ... . . .. .... .. . .. . 278
Terms based on remote sensing .. . . . . . .... ... . . . 258 Sand sheets and streaks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Occurrence in sand seas . . . . . . . . . . . . . . . . . . . . . . .. 260 Interdune areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

253

Digitized by Google
254 A STUDY OF GLOBAL SAND SEAS

Illustrations
Pige

FIGURE 166. Sketches showing methods of measuring length, width (or diameter), ana
wavelength of dunes on Landsat imagery and on Skylab and aerial photo-
graphs ............................................................. 258
167. Aerial photographs of common varieties of simple linear dunes ............ 262
168. Aerial and space photographs of common varieties of complex linear dunes 264
169. Landsat imagery sample areas on which linear dune width, length, and
wavelength were measured .......................................... 265
170. Histograms showing frequency distribution of dune width, length, and
wavelength in simple linear dunes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 266
171. Histograms showing frequency distribution of dune wavelength in compound
and complex linear dunes ........................................... 267
172. Scatter diagram showing correlation of mean wavelength and mean width in
linear dunes in relation to dune form ................................. 266
173. Aerial photographs of simple crescentic dunes ........................... 266
174. Aerial and space photographs of compound and complex crescentic dunes. 270
175. Landsat imagery sample areas on which compound and complex crescentic
dune ridge widths, lengths, and wavelengths were measured ............ 272
176. Histograms showing frequency distribution of wavelengths in small crescentic
dune ridges ........................................................ 272
177. Histograms showing frequency distribution of widths, lengths, and
wavelengths in compound crescentic dune ridges . . . . . . . . . . . . . . . . . . . . .. 273
178. Scatter diagrams showing relation of values of crescentic dunes. . . . . . . . . . .. 274
179. Aerial photographs of simple, compound, and complex star dunes ......... 276
180. Landsat imagery sample areas of star dunes, Arabia and western North America 278
181. Histograms showing frequency distribution of diameters and wavelengths of
star dunes ......................................................... 279
182. Scatter diagram showing relation of mean dune diameter to mean dune
wavelength in star dunes ............................................ 279
183. Landsat imagery sample areas of parabolic dunes, Thar Desert, India and
Pakistan, and Colombia-Venezuela border ............................. 280
184. Aerial photograph of complex domes in the Nafud ath Thuwayrat, northern
Saudi Arabia ....................................................... 280
185. Landsat imagery sample areas of dome-shaped dunes. . . . . . . . . . . . . . . . . . . .. 281
186. Histograms showing frequency distribution of diameters and wavelengths of
dome-shaped dunes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 281
187. Landsat imagery showing sand sheets and streaks associated with linear dunes,
Sahara, Mauritania .................................................. 282

Tables
Page

TABLE 25. Examples of types and forms of dunes, using classification based on remote
sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
26 - 30. Chronological lists of terms considered equivalent to classification based on
Landsat imagery:
26. Linear dunes ................................................. 284
27. Crescentic dunes ............................................. 265
28. Star dunes ................................................... 288
29. Parabolic dunes .............................................. 289
30. Dome-shaped dunes .......................................... 290

Digitized by Google
MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 255

Page
TABLE 31 - 35. Chronological lists of local terms for:
31. Complex dunes of a" types ................................... . 290
32. Sand seas ................................................... . 291
33. Patterns of eolian sand deposits in sand seas .................... . 292
34. Sand sheets and streaks ....................................... . 293
35. Interdune areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 295
36. Morphometry of linear dunes in sand seas .............................. 296
37. Similarity of linear dune varieties in widely distant sand seas, shown by com-
parison of ratios derived from measurements .......................... 297
38. Morphometry of crescentic dunes in sand seas .......................... 297
39. Similarity of compound crescentic dunes in widely distant sand seas, shown by
comparison of ratios derived from measurements ...................... 298
40. Morphometry of star dunes in sand seas ................................ 299
41. Similarity of star dunes in widely distant sand seas, shown by comparison of
ratios derived from measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 300
42. Morphometry of dome and dome-shaped dunes . . . . . . . . . . . . . . . . . . . . . . . .. 302
43. Similarity of giant domes and dome-shaped dunes in widely distant sand seas,
shown by comparison of ratios derived from measurements. . . . . . . . . . . .. 302

Digitized by Google
Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 257

and photographs used for sample measurements,

Summary of Conclusions and latitude and longitude of the sample areas, are
~ UNES OF SAND SEAS are classified in this given in the tables at the end of this chapter.
1.Q chapter according to their external shapes and Dune lengths, widths or diameters, and wave-
slipface orientations as observed in planimetric lengths (crest-to-crest distances) were measured
view on Landsat imagery and on Skylab and aerial with an optical comparator on 1:1,000,000 scale
photographs. Types observed (in order of their rela- bulk-processed Landsat transparencies, according
tive abundance) are linear, crescentic, star, to the conventions illustrated in figure 166. The
parabolic, and dome-shaped dunes. Within dune lower limit of resolution for measurement of sand
types, different stages of slipface development are features was 100 metres (328 ft). Dunes measured
used in this chapter to distinguish simple, com- on Landsat imagery commonly are as much as 3,000
pound, and complex forms. Measurements of dune m (3 km, or about 2 mil wide. Some individual
length, width (or diameter), and wavelength have linear dunes are more than 50 km (30 mil long and,
been made on imagery and photographs of sand thus, extend beyond the sample areas. The size of
seas where dune type and form are uniform. most Landsat sample areas is 2,500 kmz (1,550 miZ).
Dune patterns in sand seas seem regular because Data resulting from the measurements are
the measured values for dune length, width, and arithmetic means of various distance measure-
wavelength are fairly evenly distributed about ments and are two-dimensional. Measured values
mean values that are characteristic of each sand are analyzed and compared to show distribution of
sea. Patterns made by dunes of the same type are dune sizes and spacing of dune types in various
similar, despite their occurrence in widely distant sand seas, and to distinguish varieties among dunes
sand seas, because the relationships of mean length, of the same general type. Associations are shown
width (or diameter), and wavelength are similar among variables of length, width (or diameter), and
among dunes of each type, regardless of differences spacing (wavelength) of dunes of certain types,
in size, form, or geographic location. Progression in regardless of their size and geographic location.
dune form, from simple to compound or complex, is
commonly associated with growth to larger size.
Within dune types and forms, varieties are dis- Classification of Eolian Sand
tinguished in this study by assessing degrees of
similarity as shown by correspondence of scale
Features Based on Landsat
ratios derived from their measured mean lengths, Imagery
widths (or diameters), and wavelengths.
Genetic and Local Terms
Introduction MOST CLASSIFICATIONS OF DUNES are
SATELLITE IMAGES OF UNIFORM 1:1,000,000 based on local descriptions or on terms which carry
scale and worldwide coverage provide a base for genetic implications, such as "longitudinal" or
quantitative and qualitative studies of large-scale "transverse." (See note by E. D. McKee, editor. tO)
eolian landforms in widely distributed sand seas. The genetic terms imply a certain mode of origin
Desert surfaces recorded on Landsat imagery of based on slipface orientation relative to the dune
Africa, Asia, Australia, and North and South body, for many dunes generally move and (or)
America were examined for evidence of eolian
sand. Large-scale eolian landforms, mainly dunes, II NOTE - the term "transvene" is perhape an unfortunate one. for it sugests
were measured on approximately 110 selected sam- a genetic connotation and has been used by some geologists to imply a dune with
crest transvene to wind direction. However. it has 10118 been used to designate a
ple areas on Landsat imagery. Some small dunes, sand rielse with slip face on one side only. and. therefore. it is related to barchans
mostly in minor sand seas of North America, were and barchanoid ricJses. differill8 only in overall shape and some features of struc-
ture.ln this publication "transvene" is not used in a lIenetic sense. but the term is
measured in sample areas on aerial and Skylab retained because of tradition and lonll u88lle for a specific type.
photographs. Identification numbers for all images E. D. McKee. editor.

Digitized by Google
258 A STUDY OF GLOBAL SAND SEAS

grow in the direction of avalanching. However, in barchanoid ridge) does not imply that all of the
space imagery where relationships of wind regimes forms within each type category have similar inter-
to dune types in a given locality are not yet proved, nal structures. The classification based on three-
assumptions that certain dune forms are "longi- dimensional internal dune structures is given in
tudinal" or "transverse," in a genetic sense, seem to Chapter A (table 1).
the authors of this chapter inappropriate. Neverthe-
less, a few well-established local terms such as Terms Based on Remote Sensing
"barchans" and "seifs," are retained to describe dis-
tinctive types of dunes recognized in most sand Eolian sand features recognized on Landsat imag-
seas. ery, locally supplemented by Skylab or aerial
Tables 26-30 list, in chronologic order, terms photographs, are dunes, irregularly shaped sand
used by previous authors for dunes classified in this sheets, and elongate sand streaks. A morphological
chapter as linear, crescentic, parabolic, star, or description of eolian sand bodies is primarily a
dome-shaped. Tables 31 -35 provide a similar classification of dune types and forms. The term
chronological listing, by author, of local terms for "dune" is used here for all well-defined eolian sand
eolian landforms herein classified as complex mounds or ridges that exist independently of sur-
dunes of all types, sand seas, patterns of eolian sand rounding topography (Bagnold, 1941, p. 7), whether
deposits in sand seas, sand sheets and streaks, and slipfaces are visible or not. The classification sug-
interdune areas. Each entry in tables 26 -35 shows gested here is based on remote sensing and recogn-
the desert region to which the term applies. These izes types, forms, and varieties of dunes (table 25).
tables are at the end of this chapter. (See note by E. D. McKee, editor. tt )
The similarity in name of very large dunes, such
as the dome-shaped dunes, compound barchans, Dune Types
and compound linear dunes observed in this study Types of dunes as defined here are based on the
of Landsat imagery, with smaller dunes of com- external shape of the dune and the arrangement of
parable shape {simple dome, barchan, and seif or its slipfaces, if any, relative to its shape in plan
view.
Linear dunes are straight or slightly sinuous,
longitudinally symmetrical sand ridges, much
longer than they are wide, that are bounded on both
sides by opposite-facing slipfaces. Linear dunes are
features commonly referred to as "longitudinal"
dunes, "seifs," and "sandridges" (table 26).
Crescentic dunes are mounds or ridges of sand
Mounds Ridges
that are crescent-shaped or composed of crescent-
shaped segments, each generally as wide or wider
A B
than it is long, and each mound or segment
bounded on its concave side by a slipface. The
category includes sand bodies that are referred to
by many geologists as "barchan" and as
"transverse" dunes (table 27). In planimetric view
on space imagery, even large-scale "transverse"
dunes that seem in ground view to be straight

\I NOTE - The classification used here and in the followill8 retlional descrip-
METHODS OF MEASURING ON LANDSAT IMAGERY, Skylab tions (chapter K) is derived largely from studies of Landsat imasery supplemented
and aerial photographs, the length (L), width (W),or diameter locally by a few Skylab and aerial photographs. and represents an essentially two-
dimensional clBBSification. It differs in certain basic respects from the clBBSifica-
(D), and wavelength ().) of the following dune types: A, Linear;
tion based on three-dimensional studies (chapter A). especially structural. used
8, crescentic; C, star; D, parabolic compound form; E, dome- by other authors of this professional paper.
shaped. (Fig. 166). E. D. McKee. editor.

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 259

TABLE 25. - Examples of types and forms of dunes, using classification in this chapter, based on remote sensing
(ERTS (E) .nd Sky"b (SL) numbers.re below phot08raphs. NR indiClites not recognized on Landsat imagery. Phot08raph scales 1:1.000.000. Crescentic
type includes transverse type u used by other .uthors in this publication J

Form
Type
Simple Compound Complex

Linear .
E-1179-00593 E-1167-06310 E-1187-06433 E-1117-06551 S~L2S E-1111~5

.,'
Crescentic
E-1128-04253 E-1227-oS501 E-1183-06194 E-1183-06194 E-1183-06194 E-142~3262

E-1183-06203 E-110~22 E-1183-06194

Star NR

E-1111~5 E-1183-06194

,,-- ~~ .
~ .~ .

Parabolic NR
E-1189-17102 E-1177-14255

Dome-like NR
E-1155-07040 E-1077-04420

ridges appear to curve slightly in crescent-shaped elude sand bodies known locally as "oghurd"
segments, though each segment may be a kilometre dunes, pyramidal dunes, and dune "massifs," sum-
or more wide from horn to horn. marized in table 28. Parabolic dunes are U -shaped
Star dunes are radially symmetrical, pyramidal or V-shaped mounds of sand, trailed by elongated
sand mounds with slipfaces on arms that radiate arms, having slipfaces on the outer. convex sides
from the high center part of the mound. They in- (table 29), and are typically associated with

Digitized by Google
260 A STUDY OF GLOBAL SAND SEAS

vegetation. Dome-shaped dunes are semicircular Dune Varieties

mounds of sand that generally lack surface slip-
faces (table 30). Within the basic classes of type and form are
numerous dune varieties. Varieties of simple and
Dune Forms compound forms of most dune types have been ob-
served in many sand seas. They are compared by
Forms of dunes are defined by the stage of slip- applying a method developed by Strahler (1958. p.
face development and. thus. to some extent by size. 292 -294) for assessing similarity of fluvially pro-
Simple dune forms are mounds or ridges with slip- duced landforms. According to this method of
faces. all of which are similar in size or character. dimensional analysis. landforms are considered to
Simple linear dunes are the main sand features of be similar when dimensionless numbers formed by
the Simpson. Great Sandy. and Great Victoria the ratios of their measured values are closely simi-
Deserts in Australia. and of the Kalahari Desert in lar (Strahler. 1958. p. 279). Certain varieties of sim-
South Africa. Simple barchans. barchanoid ridges. ple and compound dunes are shown by this method
and simple linear and star dunes are the main dune to be common to many widely distant sand seas.
forms of most North American deserts. but they are Common names for these dunes are given in tables
minor. inconspicuous components of most major 26-30. Varieties of complex dunes. however, tend
sand seas observed on Landsat imagery. to be unique to specific localities, so their names
Many dunes in sand seas of Asia and northern usually only have local application. Those names
Africa are extremely large and are characterized by are listed in table 31.
extensive development of secondary sets of slip-
faces of lesser size than those which define the
main mound or ridge. If secondary slipfaces are of Occurrence in Sand Seas
the same kind as those that define the main ridge or Large areas of eolian sand constitute sand seas,
mound. the dune is a compound form; if they are of so-called because the dunes in them tend to be
a different kind. the dune is a complex form. repeated in orderly sequences rather than occurring
Compound dunes are mounds or ridges of sand as scattered solitary features. Sand seas occur from
on which smaller dunes of similar type and slipface above the Arctic Circle in Alaska to below the An-
orientation are superimposed. For example. a com- tarctic Circle in Antarctica, and at all latitudes be-
pound crescentic dune consists of a sand mound or tween. Most major sand seas, however. lie within a
ridge having similar but smaller secondary cres- belt between lats 40 N. and 30 S., in Africa, Asia,
cent-shaped dunes on its gentle slopes. The ar- and Australia. General distribution of the world's
rangement of slipfaces. of similar appearance. gives sand seas is discussed by I. G. Wilson (1971b, 1973).
some dunes of compound form a typically broad. Local names for many sand seas are summarized in
"two-storied" aspect (Kadar. 1966. p. 447). Subsidi- table 32. Descriptive terms used by various workers
ary slipfaces on compound crescentic dunes are to identify orderly patterns of eolian sand deposits
seldom clearly visible on Landsat imagery. but they in sand seas are summarized in table 33.
can be identified by inspection of aerial or Skylab In the following parts of this chapter, dunes are
photographs. or from descriptions based on field described according to type. in order of their rela-
observations. tive abundance. as observed on Landsat imagery of
Complex dunes are combinations of two or more major sand seas. Linear dunes are the main dune
dune types. The most common complex forms ob- type in the Southern Hemisphere deserts of
served on Landsat imagery are linear or crescentic Australia and southern Africa. They occupy about
ridges with superimposed dunes of some other half the area of the Rub' al Khali in southwestern
type. such as star dunes. developed along their Saudi Arabia. and they occur in several sand seas of
crests. Generally. complex forms consist of a main the Sahara of northern Africa. Crescentic dunes are
dune type and a secondary dune type and can be more common than linear dunes in the Northern
classed as mainly linear. mainly crescentic. and so Hemisphere. They occupy large areas of the eastern
forth. However. some complex dunes have approx- Rub' al Khali in southeastern Saudi Arabia, several
imately equal components of different dune types; sand seas in the Sahara. and most sand seas in
where no one dune type is dominant. the dunes are northern China, eastern U.S.S.R.. western Argen-
here considered to be complex unclassified. tina. and western North America.

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 261
Star dunes occupy far less area within sand seas Some linear dunes occur as solitary features
than do either linear or crescentic dunes but, like (chapter 1<, fig. 193) but mostly they form parallel
them, are widely distributed. Major occurrences of ridges separated by sandy, gravelly, or rocky inter-
star dunes are in sand seas of the northern Sahara dune corridors in sand seas. Linear dunes, ap-
in Algeria and Libya, the Gran Desierto of Mexico, parently of simple form, cover most of the
and around the margins of sand seas in northern Australian deserts and the Kalahari Desert of South
China, the Arabian Peninsula, and the Namib Africa, but in many other sand seas linear dunes
Desert of South-West Africa. are mostly compound or complex.
Parabolic dunes of a size visible on Landsat imag- The largest solitary linear dune measured on
ery are rare. Observations on these images are Landsat imagery is a 100-km-Iong (26-mi-Iong), 1-
mostly limited to the Thar Desert of India and km-wide (0.62-mi-wide), complex dune between
Pakistan, and to low-lying areas along the Colom- Atar and Akjoujt, Mauritania, in the northwest
bia-Venezuela border near Buenavista, small dune Sahara (chapter K, fig. 193). The largest linear
fields in the Kalahari Desert of South Africa, and dunes forming parallel ridges in a sand sea are in
White Sands in western North America. Also rare the Qa'amiyat region of the southwestern Rub' al
on Landsat imagery are domes. or dome-shaped Khali, Saudi Arabia, where individual dunes can be
dunes; they are observed in sand seas of the north- traced for 190 km (118 mil and range in width from
ern Arabian Peninsula, northern China, and the 0.30 to 2.7 km (0.2 to 1.7 mil.
Sahara of northwestern Algeria.
Sand sheets and streaks (termed "stringers" by Morphometry
many authors) are abundant in most major sand Dune width, length, and wavelength were
seas, but these features have no common geometric measured on sixty-seven 2,500-km2 (l,550-mi2)
characteristics on Landsat imagery by which they sample areas of Landsat imagery showing linear
can be compared from desert to desert. Some of dunes of simple, compound, and complex form in
them are known to contain small dunes that are northern Africa, southern Africa, Australia, and the
below the resolution of Landsat imagery. Common Arabian Peninsula. Measurements of linear dunes
terms for them are given in table 34. of simple form on the Navajo Indian Reservation,
northern Arizona, U.S.A., were made from high-
Linear Dunes altitude aerial photographs (fig. 1678). Some sam-
ple areas on other continents are shown in figure
General Distribution 169. Histograms (figs. 170, 171) show distribution of
the measured values in sample areas of various
SIMPLE, COMPOUND, AND COMPLEX linear sand seas. In general, measured values for dune
dunes are observed on Landsat imagery and on width, length, and wavelength (crest-to-crest dis-
Skylab and aerial photographs (figs. 167, 168). Sim- tance) tend to be grouped about characteristic
ple linear dunes are single or bifurcated ridges with means (summarized in table 36) that give a regular
narrow crests that may be sharp or subdued but pattern to each sand sea.
which do not have slipfaces of secondary dunes Measured values of dune wavelength for simple
developed upon them (fig. 167). Compound linear linear dunes in sampled areas of the Simpson, Great
dunes are generally much broader linear ridges that Sandy, and Kalahari Deserts, and Arizona sand seas
have relatively smaller linear ridges, each with its are distributed lognormally, unlike the distribution
own slipfaces, on their tops. In the "feathered" of measured values for simple linear dune
variety of compound linear dunes, subsidiary ridges wavelength in the sampled area of the Rub' al Khali
intersect or spread obliquely from the main ridge. (fig. 170), and linear dunes of compound and com-
Linear dunes of compound form in Arabian sand plex forms (fig. 171). The greatly skewed distribu-
seas are shown on aerial photographs by Glennie tions of wavelength 12 indicate that spacing of most
(19'70, fig. 74) and by Holm (1960, fig. 7). Complex
.. Tran8formation8 of the data. for stati8tical treatment. ara considered un-
linear dunes may be narrow or broad linear ridges necessary because the sample mean of 8uch di8tributions is considered to be
that have other dune types, such as star, dome- "mora than 90 percent efficient" a8 an estimate of the population mean. "provided
shaped, and barchan (crescentic) dunes, superim- 1.2.
that the coefficient of variation is less than correspondill8 to a variance of the
10000rithma of about 0.9" (Koch and Link. 1970. p. 219-220). No coemcients of
posed upon them (fig. 168). variation8 (q shown in figure 110 exceed the allowable figure.

Digitized by Google
262 A STUDY OF GLOBAL SAND SEAS

simple linear dunes of the variety common to the Tests were made of measurements within these
Kalahari and to Australian and Arizona deserts is deserts to determine whether statistically signifi-
fairly even and ranges about a mean value charac- cant differences in mean dune width, W, mean
teristic of each area, but that extremely wide spac- dune spacing, X (crest-to-crest distance), and mean
ing of a few individual dunes is common. frequency, T{number of dunes per kilometre), are
Measured values for widths and lengths of the associated with relative distances downwind or
linear dunes are normally distributed, except in the with positions in the open parts of sand seas, as op-
sampled Kalahari and Arizona localities, for posed to positions near topographic barriers.
reasons that are not yet clear. Several different methods were tried.
Size and spacing relationships of linear dunes First, mean wavelengths of dunes in samples
sampled in different sand seas were examined to measured near hills, playas, and streambeds in the
see whether relationships of these variables are Australian and Kalahari sand seas were compared
constant regardless of differences in dune size and with mean wavelengths of dunes in samples
geographic location. A scatter diagram (fig. 172) measured in open areas without such topograhic
shows a linear arithmetic relationship between barriers. Differences significant at the 0.098 level
mean dune width and spacing in groups of linear for the Simpson Desert, 0.047 level for the Great
dunes of all forms and varieties, in 12 sand seas of Sandy Desert, and 0.010 level for the Kalahari
Africa, Asia, Australia, and North America. Sizes of Desert show that mean spacing of dunes in these
dunes represented in figure 172 range from an sampled areas is significantly wider near
average of 43 m (141 ft) wide, on the Navajo Indian
Reservation, Arizona, U.S.A., to an average of 1,480
m (4,860 ft) wide, at the western end of the Rub' al
Khali, Saudi Arabia (table 36).
Regression analysis indicates that wavelength,
which is presumed to be an independent variable,
tends to be about twice the mean width of linear
dunes, regardless of their size. The relationship also
indicates a progression of dune form, from simple
to compound or complex, with increasing size. In-
creasing size of linear dunes thus seems to be
associated with development of subsidiary dunes,
which may have slipfaces of the same or different
orientations than the slipfaces of the main dune.
Growth to great size is believed to be a function of
time (I. G. Wilson, 1971a, p. 182-183), as well as of
sand supply and effective winds.

Relation to Topographic Barriers and

Wind Regimes
In sand seas such as the Australian deserts, the
southwest Kalahari Desert of South Africa, and the
Qa'amiyat region of the western Rub' al Khali,
Saudi Arabia (chapter K), groups of linear dunes COMMON VARIETIES OF SIMPLE LINEAR DUNES (above and
extend for hundreds of kilometres. Large groups of facing page). A, Bifurcating but straight dunes in the northwest
Simpson Desert, Australia. Mean width of the dunes is 0.22 km
linear dunes in the Sahara, on the Arabian Penin- (0.14 mil. Aerial photograph by E. Tad Nichols. 8, Linear dunes
sula, in southern Africa, and in Australia generally on the Navajo Indian Reservation, Arizona, U.S.A. Aerial photo-
are ali ned parallel or subparallel to the resultant graph by National Aeronautics and Space Administration (lSC
direction of effective winds (chapters F and K). 236 28 0073). (Fig. 167.)

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 263

(Fig. 167 - Continued.)

topographic barriers than where such barriers are of the standard deviation to the mean of each sam-
absent. ple). versus distance downwind in the Rub' al Khali
Second. regularity of dune spacing (represented and the Kalahari. Simpson. and Great Sandy
by dispersion of wavelength measurements) rela- Deserts. No correlation was found.
tive to distance downwind in a sand sea was tested Third. the numbers of dunes in six sample areas
by plotting the coefficient of variation. C (the ratio of the Simpson Desert were counted to see whether

Digitized by Google
264 A STUDY OF GLOBAL SAND SEAS

significant changes in dune numbers occur with

distance downwind (measured in a direction
parallel to resultant wind direction). No significant
difference in dune numbers was found either be-
tween adjacent samples or between extreme up-
wind (southeast) and downwind (northwest) sam-
ples in the Simpson Desert.
The negative results of these tests suggest that
simple distance from the upwind end of a sand sea
(and, presumably. distance from the sand supply) is
not an important factor in determining the distribu-
tion patterns of linear dunes.

Identification of Varieties
Ratios derived from measurements of the mean
widths (\IV). lengths (L), and wavelengths (X) of
dunes in each sample area define the geometric
characteristics of the sampled dunes in each sand
sea (table 36). The degree of similarity of simple
linear dunes in different sand seas is shown by the
degree of correspondence of their ratios (table 37).
At least two varieties of simple linear dunes can
be distinguished by the degree of similarity of the
sampled dunes:
1. Long narrow linear dunes of the Australia
"sandridge" variety, common to the Simpson
and Great Sandy Deserts, show excellent to
good correspondence to linear dunes of simi-
lar size in the Kalahari Desert and good to fair
correspondence to much smaller linear dunes
on the Navajo Indian Reservation in Arizona.
2. Short linear dunes of the "seif" variety in the
Rub' al Khali sand sea of Saudi Arabia, east of
Al 'Ubaylah (chapter K, fig. 241) show very
poor to good correspondence to the linear
dunes in the Navajo Indian Reservation of
Arizona in the United States, in southern
Africa (fig. 1691\). and in parts of Australia
(fig. 169B). Comparison of ratios, derived
from measurements, thus seems to support
AERIAL AND SPACE PHOTOGRAPHS OF COMPLEX LINEAR the suggestion by Price (1950, p. 462-463)
DUNES. A, Subsidiary star dunes and crescentic dunes on the that two varieties of linear dunes have been
crests of complex linear ridges in the central Namib Desert, described as seifs - one by Bagnold (1941, p.
South-West Africa. Dunes average 2.2 km (1.4 mil from crest to 222 -229) in the Sahara of Libya. and the
crest. Photograph by H. T. U. Smith. B, Skylab EREP (Earth
Resources Experiment Package) photograph showing linear
other, called sandridges, by Madigan (1946. p.
dunes with barchanoid ridges on their crests in the northwest 45 -63) in the Australian deserts.
Sahara, Algeria, northern Africa. Photograph by National Aero- Comparisons of dimensionless ratios for linear
nautics and Space Administration (SL 3 - 28 365). (Fig. 168.) dunes of compound and complex forms are not

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 265

1B"3O'

B"3O'

o 10 20 .a KM
I I
I
o
I
10
I

LANDSAT IMAGERY SAMPLE AREAS on which measurements of linear width, length, and wavelength were made.
Numbers in parentheses refer to the regions listed in table 36. Each sample is an area 50 km by 50 km, or 2,500 km2
(1,550 mi2) in a sand sea. A, Kalahari Desert, southern Africa; simple dunes (4). 8, Great Sandy Desert, Australia; sim-
ple dunes (3). C, Southwestern Rub' al Khali, Saudi Arabia; compound dunes (8). D, Southwestern Sahara; com-
pound dunes (6). E, Namib Desert, South-West Africa; complex dunes (9). F, Western Rub' al Khali, Saudi Arabia;
complex dunes (12). (Fig. 169.)

given because. as indicated in table 36. lengths of graphs. Simple crescentic dunes are crescent-
these dunes commonly are greater than the sample shaped mounds. commonly referred to as barchan
lengths. dunes. Crescentic ridges. commonly known either
as barchanoid ridges or as transverse ridges. have a
Crescentic Dunes single slipface on each arc. In most places they are
General Distribution relatively small (fig. 173). Curvature of the seg-
ments of some crescentic ridges is extreme. and the
ON THE BASIS OF SLIPFACE DEVELOP- term "barchanoid" has been used (Cooper. 1967. p.
MENT. three basic forms of crescentic dunes- 28) to distinguish this variety from the very slightly
whether mounds or ridges - can be discerned on curved variety of ridge commonly called transverse
Landsat imagery and on aerial and Skylab photo- on the basis of structure (table 27).

Digitized by Google
266 A STIJDY OF GLOBAL SAND SEAS

40.-------, 200
A .-O.IM 1 - 3.65 I - 0.15 B
S-0.0111 S - l.83 S - 0.076 100
....- e-o.3I e - O.50 e - 0.51
... 20 I"- - 110 I-o.t
S-1.01
.-- .-0.22
S-O'-
e-l .12 10,... e-o.

0
.......T lID

a ... ..... ... ... a ..... ..

T

: ! ..s
WIln4
.. .. ;
S 8 N
LENGn4
OIl ~ ;
WAYaENGn4 140
10

110
C
40 F-
I--
-
120
lID lOll 20-
.-0.21 ... 100
S-0.12
10 e-o.a ~
-
.- a '" .. :
140 0 T I
"!
10 ~ 0 0

120 10 WIln4
40
10 l-24.0I
S-13.1i2
100 40 e-ul
20
40
10 20
0

10 0
20
-:
0.

~ .. .....
"! ~
: : .....
-
0 0 0
40
WIln4
; ..... ~
'"
..., ...
N N ~ ...'" ...,.; ... : ...t<i
WAYaENGn4
:;; ; ..... ... ...
II! ,.; ,.;

I-o.lO
20 S-UI 100 . . . . - - - - - - - - - - - , 10
e-o.12 D
I-o.70 .-0.29
0
; .. -...,
... ~ "! ... ...... -
... "!... :;;
~ '" ..;.; ~
..- II!~
10
S-o....
e-o." -- S-O.111
e-o.1i2 -
0 ~ N
'"
10
WAVElfNGllf
eo 20 - r- -
C ;,.-
i-22.7
S-12.31 o
40 e-l.o I

20

0
o
.-
0
; d ~ : ~ ~ ~ ~ ~ ~ ~ : ~ 000 0 0

::: ~ R ~ R ; Iii III

0. 0. 0 0. 0 0
; Iii III WAVELENGn4 LENGn4
LENGn4
40 measured on aerial photographs (A) and landsat transparencies

0j
E "-0.31 I-1.41
S-0.12
e-0.32
S-5M
-OAI S-UI
e-0.41
(8 - f); W, mean dune width; mean dune length, I, mean r;
"'20 dune wavelength; 5, standard deviation; C, coefficient of varia-
tion . Values along abscissa are in kilometres. Numbers in
parentheses refer to regions listed in table 36. A, Navajo Indian
0
; ..
... :;!
WIln4
~
0 .. ..
... ~ 0.
0
~ ~
LENGn4
0.
R
0
.. ..... '"
0 0.

WAVELENGn4
... ...N
0. 0
,.;
Reservation, Arizona, U.S.A. Area sampled is 625 km2 (390 mi2)
(1) . 8, Simpson Desert, Australia. Area sampled is 30,000 km2
(18,640 mi2), (2). C, Great Sandy Desert, Australia. Area sampled
is 35,000 km2 (21,740 mi2) (3). 0, Kalahari Desert, South Africa.
HISTOGRAMS SHOWING FREQUENCY DISTRIBUTION (f) of Area sampled is 15,000 km2 (9,320 mi2), (4). f, Northern Rub' al
dune widths (W), dune lengths (L), and dune wavelengths (X), Khali, Saudi Arabia. Area sampled is 2,500 km2 (1,550 mi2) (5).
in kilometres, in sample areas of linear dunes of simple form, (Fig. 170.)

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 267

0~-r~r--r-4F=9

2.0 2.5 10 3.5 4.0 4.5 1.0 1.5 2.G 2.5 3.0 1.G 1.5 2.0 2.5 3.0
A B C

50

40 i-1.I
I-3.24 S-o"
C-OA2 i-3.21
3D S-1.11 S-o,-
... C-o.31 c-on
20

10

o~~~~--~~-+~~~~
1.0 1.5 2.0 2.5 10 3.5 4.0 4.5 5.0 5.5 0.5 1.0 1.5 2.0 2.5 3.0 3.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
D E F

HISTOGRAMS SHOWING FREQUENCY DISTRIBUTION complex dunes. Area sampled is 2,500 km2 (1,550 mi2)
(f) of dune wavelengths ().), in kilometres, in compound (9). C, Southwestem Sahara, Mauritania; compound
and complex linear dunes measured on landsat dunes. Area sampled is 2,500 km2 (1,550 mil) (6). D,
transparencies: I, mean dune wavelength; 5, standard Northern Sahara, Algeria; complex dunes. Area sampled is
deviation; C, coefficient of variation. Numbers in 2,500 km2 (1,550 mi2) (10). E, Southern Sahara, Niger;
parentheses refer to regions listed in table 36. Values compound dunes. Area sampled is 7,500 km2 (4,660 mi2)
along abscissa are in kilometres. A, Western Rub' al Khali, (7). F, Southern Sahara, Niger; complex dunes. Area
Saudi Arabia; complex dunes. Area sampled is 2,500 km2 sampled is 2,500 km2 (1,550 mi2) (11). (Fig. 171.)
(1,550 mi2) (12). B, Namib Desert, South-West Africa;

Compound crescentic dunes (the "mega- ary transverse ridges on the windward slope. which
barchans" of Norris. 1966. p. 292-296) are generally appears to be typical where the basic dune form
much larger than simple crescentic dunes and are reaches large size." Similar features are shown on
characterized by the presence of numerous subsidi- an aerial photograph of the Algodones Dunes.
ary crescentic dunes with slipfaces oriented the California (fig. 1740). in a close view of that area
same as the main slipfaces but developed sec- by Norris and Norris (1961. pI. 3. fig. 1) and in the
ondarily on the gentle slope of each major dune Peski Karakumy. U.S.S.R . illustrated by Suslov
mound or ridge. Several examples are shown on (1961. fig. 14-3). In Saudi Arabia. compound cres-
aerial photographs in figure 174. Compound cres- centic dune ridges showing pronounced curvature
centic mounds are typified by the Pur-pur dune of of the segments (fig. 1741\) have been described
Peru (Simons. 1956). which is a large solitary (Holm. 1960. p. 1372) as "giant crescentic massifs"
barchan. 0.75-0.85 km (0.46-0.52 mil wide and 2.1 (table 27).
km (1.3 mi) long. with a major slipface forming its Very large compound crescentic dune ridges
steep slope; numerous minor crescents. each with a were observed on Landsat imagery of sand seas at
slipface on its lee side. occur on the windward relatively high latitudes. in the Peski Karakumy
slope. (U.S.S.R.). Takla Makan and Ala Shan Deserts
Compound crescentic dune ridges are more com- (northern China). and in the Nebraska Sand Hills
mon than large solitary barchans and are typified (U.S.A.). Lower latitude sand seas of large com-
by the Nebraska Sand Hills. described by H. T. U. pound crescentic dunes occur in the eastern Rub' al
Smith (1965. p. 564) as "transverse. and probably Khali (Saudi Arabia). several parts of the Sahara
* * * of the compound variety. with minor second- (northern Africa). and in the Algodones and Gran

Digitized by Google
268 A STUDY OF GLOBAL SAND SEAS

3.4
3.2

3.0
2.8

2.6
2.4
2.2
2.0

1.8
1-<
1.6
1.4 .5

1.2

1.0 LINEAR DUNES

0.8 Simple
@ Compound
0.6
o Complex
0.4
0.2
.1
OL-J-~

o
__L-J--L~L-~-L~__~-L~__~-L~
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1o 'f
I---r----""Ti-""""T'j.... METRES
lSOOFEET
W
SCATTER DIAGRAM SHOWING CORRElATION of mean dune
wavelength (X) and mean dune width (W) in linear dunes, in re-
lation to dune form. Measurements in kilometres. The correla-
tion coefficient (r) = 0.93, shows excellent correlation. Num-
bers beside points refer to regions listed in table 36. (Fig. 172.)

o 200 METRES
I
I
o
I
600 FEET
Comb Ridge, near the Arizona-Utah border, U.S.A., at lat 36"51'
N., long 110 0 W. Photograph from Denny, Warren, Dow, and
Dale (1%8, p. 14, fig. Ariz. 3). 8, Barchanoid ridges, White
o Sands, New Mexico, U.S.A. (near lat 32 0 48' N., long 106~6'
I I W .). Photograph used by permission of Holloman Air Force
a 1 MI Base. C, Transverse ridges grading southward into barchanoid
ridges and barchans in the Sahara (near lat 20 0 N., long 11 0 W .).
AERIAL PHOTOGRAPHS OF SIMPLE CRESCENTIC DUNES. A, Photograph from Monod (1958, pI. XXVII, p. 173, used by per-
Transverse ridges with slightly crescentic segments north of mission of the author). (Fig. 173.)

Digitized by Google
 MORPHOLOGY AND DlSTRffilITION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 269

Desierto regions of California and Mexico, respec- coastal crescentic dunes commonly are oriented
tively. The largest dunes of crescentic type that transverse to resultant directions of fairly strong
have been noted on images are in the Takla Makan present-day winds (chapters F and K).
Desert of China (chapter K), where mean width
(horn-to-horn) of the segments is 3.24 km (about 2 Morphometry
mil. Large crescentic dunes, in general, are oriented Dune width (horn-to-horn distance of each cres-
transverse to resultant directions of effective cent or crescentic segment), dune length (distance
winds, but present winds in the regions of these from base of windward slope to base of slipface),
large dunes mostly are of relatively low energies and dune wavelength (crest-to-crest distance) were
(chapter F), compared with energies of present measured (fig. 166) on forty-five 2,500-kmz (1,500-
winds in many other desert regions. miZ) sample areas of Landsat imagery, representing
Complex crescentic dunes include both large and large, compound crescentic dunes in both northern
small mounds or ridges on which subsidiary dunes and southern Africa, the Arabian Peninsula,
of a different type and orientation - most com- Pakistan and India, U.S.S.R., northern China, and
monly star dunes - have formed. Some crescentic western North America. Wavelengths of smaller,
dunes, typified by ridges in the Great Sand Dunes possibly simple crescentic ridges were measured on
National Monument, Colorado (McKee, 1966, p. 61, numerous Landsat imagery samples of the Arabian
pI. E; Denny and others, 1968, fig. lOB) are included Peninsula, Pakistan, China, and South-West Africa.
in this discussion as a variety of complex crescentic Some sample areas of Landsat imagery for most of
dunes because they seem to be combinations of these regions are shown in figure 175. Simple cres-
both crescentic and star dune types. Some dunes in centic (barchanoid) dunes at White Sands, New
the Namib Desert believed to be of the complex Mexico, and compound crescentic dunes in the
crescentic variety are shown on a Skylab photo- Gran Desierto, Mexico, and in the Algodones
graph in figure 174C Dunes, California, were measured on aerial and
Groups of relatively small crescentic mounds Skylab photographs.
seen on Landsat images occur mainly in the Measurements in each sampled sand sea are
southern Sahara and Libyan deserts of northern summarized in table 38, and the frequency distribu-
Africa and in the southeastern Rub' alKhali, Saudi tions of measured values are shown by histograms
Arabia. These barchanlike dunes are 0.20-1.5 km in figures 176 and 177. Measurements of individual
(0.12--0.93 mil wide and 0.25-1.1 km (0.15-1.7 mil dunes show a wide range in many sand seas, but as
long; most are probably compound barchans rather a group, measurements in each sampled area tend
than simple barchans. A typical field of these dunes to be distributed close to the mean values. Fairly
in the western Sahara of Mali is shown on a sample uniform distribution of dune lengths, widths, and
of Landsat imagery (fig. 175A). wavelengths defines the apparent regularity of
Small crescentic dune ridges are characteristic of dune patterns in each sand sea.
many dune areas bordering the sea, such as the Size and spacing relationships of crescentic dune
coastal part of the Namib Desert (chapter K, figs. ridges sampled in different sand seas were ex-
222, 223), the Persian Gulf COBl!.t (eastern Qatar and amined to determine whether these relationships
the area around Sebkhat Mati) and the southern are constant despite differences in dune sizes and
coast of Mauritania near the senegal River (figs. geographic locations. Sizes of dunes represented in
188, 191). Small crescentic dune ridges also occur figure 178 range from crescentic segments with a
inland, where they commonly override much larger mean width of 111 m (365 ft) at White Sands, New
dunes of crescentic type, as in the AI Jiwa' area of Mexico (fig. 173B), to segments with a mean width
the eastern Rub' al Khali (chapter K, fig. 24O) and of 3.24 km (2 mil in the Takla Makan Desert of
on the eastern edge of the Takla Makan sand sea northern China (fig. 175D). Scatter diagrams (fig.
near T'ieh-kan-li-k'o (chapter K, fig. 261A). Size of 178) show linear arithmetic relationships between
the small dunes, however, precludes making the mean widths of crescentic dune segments and
measurements on Landsat imagery, except for their mean wavelengths, between the mean widths
wavelength, so apparent similarities cannot be of crescentic dune segments and their mean
tested by comparisons of their scale ratios. Small lengths, and between the mean lengths of crescen-

Digitized by Google
270 A STUDY OF GLOBAL SAND SEAS

AERIAL AND SPACE PHOTOGRAPHS OF COMPOUND and

complex crescentic dunes (above and facing page): A, Com-
pound crescentic ridges in the eastern Rub' al Khali (AI'Uruq al
Mu'taridah), Saudi Arabia, at approximately lat 23 N., long 54
E. One giant crescentic segment is outlined to show relation to
much smaller, simple barchans that cross the interdune areas
between the ridges. Segments of these ridges average 2.76 km
(about 1.75 mil from horn to horn and are about 3.1 km (0.3 mil
long, from the base of the gentle slope to base of the main slip- dunes nearest the river are about 300 m (1,000 ft) high (Barnard,
face. Photograph by the U.S. Department of Defense. B, Com- 1973, p. 2, 5). Photograph by National Aeronautics and Space
plexcrescentic dune ridges in the southeastern Rub' al Khali, Administration (SL 4 -207 -8073). 0, Compound crescentic
Saudi Arabia, at approximately lat 21 45' N., long 5430' E. Star dune ridges in the Algodones Dunes, California, U.S.A. Major
dunes have formed on the crests of the ridges. Photograph by slipfaces are on the south-facing lee slopes, and numerous sub-
the U.S. Department of Defense. C, Complex crescentic dunes sidiary slipfaces are on the gentler slopes. Photograph by Na-
along the valley of the Tschaubrivier, South-West Africa. The tional Aeronautics and Space Administration. (Fig. 174.)

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 271

tic dune ridges and their mean wavelengths. The varieties of compound crescentic dune ridges can
latter association is particularly pronounced (cor- be distinguished:
relation coefficient = O.90), and this suggests that 1. Crescentic ridges whose segments are much
spacing of crescentic dunes varies directly with wider than they are long, and which are not
dune length, regardless of the size, form, or location markedly curved; these include the Aoukir
of the dunes. Correlation of mean widths of cres- dunes of the southwestern Sahara in
centic dune segments with their mean lengths is Mauritania and the Nebraska Sand Hills,
also strong (r = 0.81) in dunes of all sizes, forms, U.S.A. (figs. 175F, ~. These ridges are known
and locations. which indicates that the correlation by most geologists as "transverse" dunes and
of width and length of crescentic mounds are referred to by that name in structural
(barchans). as recorded by Finkel (1959. p. 628) , studies in this publication.
seems to apply also to segments of crescentic
2. Crescentic ridges whose segments are as wide
ridges. As in linear dunes (fig. 172) progression in
or slightly wider than they are long and
form from simple to compound or complex in cres-
which are markedly curved; these include the
centic dunes seems to be associated with growth to
compound crescentic dunes of the eastern
larger size (fig. 178).
Rub' al Khali, Saudi Arabia, the Takla Makan
and Ala Shan Deserts, China, (figs. 174A;
Identification of Varieties 175B. D, C) and the Thar Desert near Sukkur.
Planimetric similarity among crescentic dune Pakistan (chapter K. fig. 2568).
ridges of compound form is shown by correspon- Dunes of the compound crescentic variety show
dence of scale ratios derived from their measure- very good to excellent correspondence to each
ments (table 39). Differences in degrees of corres- others' scale ratios but mostly poor to fair corres-
pondence of the ratios indicate that two extreme pondence to the ratios of dunes of the variety com-

Digitized by Google
272 A STUDY OF GLOBAL SAND SEAS

; '
.....

.'~\
. ,. / ~ /1:,- , 'll..
'1"I If.i).'~~~~'j,-.,: r;~",'
'. f..-r." , ' , f"l~
- '.~ ,~
. .... ,,,.
.' ." """. , -I .' ,
,

~ .. ""

o 10 20 :II l KM
I
o
, ,

10
I
"20M! '
LANDSAT IMAGERY SAMPLE AREAS ON WHICH COMPOUND in figure 174A. C, Large barchanoid variety of compound cres-
crescentic dune width, length, and wavelength were measured. centic dunes in the central Ala Shan Desert, China (20). 0, Com-
Each sample area is 2,500 km2 (1,500 mil). Numbers in pound crescentic dunes in the northeast Takla Makan Desert,
parentheses refer to regions listed in table 38. A, Large crescentic China (15). f, Compound crescentic dunes in the Nebraska
dunes, probably compound barchans, in the western Sahara, Sand Hills, U.S.A (19). F, Compound crescentic dunes in the
Mali (8). B, Compound crescentic dunes in the eastern Rub' al Aoukir sand sea, western Sahara, Mauritania (21). (Fig. 175.)
Khali, Saudi Arabia (14); these are the dunes seen in closer view
20 ,..-----:::----,
A B i-O.58
S-0.13
c i-o.12I
S-0.18
C-0.22 C-O.25

O'f--+-+--f-~
0.04 0.08 0.12 0.18 0.20 0.2 0.3 0.. 0.5 0.8 0.7 0.8 0.9 1.0 0.2 0.3 0.. 0.5 0.8 0.7 0.8 0.9 1.0

~:kd bh l fECth
i-o.87 i-o... F
S-O.28 S-035

0.2 0 . 0.8 0.8 1.0 1.2 U

C-O.32

1.8
I
1.8
I
2.0 0 0.2 0. 0.8 0.8
i
1.0 1.2
I
U
I
C-G.5'

1.6
I
1.8
I 12.0 0.5 0.7 0.9 1.1 1.3 1.5

HISTOGRAMS SHOWING FREQUENCY DISTRIBUTION (I) of Desert north of Umarkot, Pakistan (3). C, Rub' al Khali, AI Jiwc1'
wavelengths (X), in kilometres, in small crescentic dune ridges area, Saudi Arabia, (5). 0, Coastal Namib Desert between Meob
measured on aerial photographs and Landsat imagery. Numbers Bay and Sandfisch Bay, South-West Africa (4). f, Coastal Arabian
in parentheses refer to regions listed in table 38. X, mean peninsula along Persian Gulf between Qatar and Sebkhat MatT,
wavelength, in kilometres; 5, standard deviation; C, coefficient Saudi Arabia (2). F, Eastern Takla Makan Desert near T'ieh-kan-li-
of variation. A, White Sands, New Mexico, U.S.A. (1). B, Thar k'o, China (6). (Fig. 176.)

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 273

I-2M
S-o..
e-o.20

1.4 1.1 2.2 2.8 3.0 3A 3.1 4.2

~ ~ -O.7 1
_ e - O . 2..
i-3.0J
i
5.5 1.8 2.3 2.7 3.1 3.5 3.8 4.3 ...7 5.1 5.5

_ A-1.2"
I .. . . S-O.33
r-- e-O.27 -

-
0
0.2 OA 0.' 0.1 1.0 1.2 1.4 1.' 0.3 0.7 1.1 1.5 1.8 2.3 2.7 3.1 0.3 0.7 1.1 1.5 1.8 2.3

ern nl
20
l-1.31 IV-2.22 I-2.27

I!
D S-O.29
... ,0 S-0.I1 S-o.&o
e-o.22 e-0.3. e-o.22
0
0.1 0.8 1.0 1.2 1.4 1.' 1.1 2.0 0.9 1.3 1.7 2.1 2.S 2.8 3.3
J
3.7 0.'
f 1.0 1.4 1.1 2.2 2.8 3.0 3."

20
E -1.' -3.1"
S-0.40 S-1.IO
... 10 e-o.33

0
0.7 1.1 1.S 1.8 2.3 1.4 2.2 3.0 3.8 .... SA 0.1 1.2 1.' 2.0 2." 2.8

20

b l
F l-o.8' IV-1.04
... 10 S-O.27 S-O.39
e-o.", e-o.37
0
0.2 0." 0.' 0.8 1.0 1.2 1.4 1.' 0." 0.' 0.1 1.0 1.2
n
1.4 1.'
i
1.1 2.0 OA 0.1 1.2 1.' 2.0 2." 2.8

':riih
IV-1.81 I-1.07
Z'0!J
S-0.3'
e-o.:
S-0.7'
e-0."7
S-OA.-
e-OA1

1
0." 0.' 0.8 1.0 1.2 1.4 1.' 1.8 2.0 2.2 0." 0.8 1.2 1.' 2.0 2." 2.8 3.2 3.' 0.3 0.7 1.1 1.S 1.9 2.3
LENGTH WIDTH WAVB.ENGTH

HISTOGRAMS SHOWING FREQUENCY DISTRIBUTION (I) of China. Area sampled is 5,000 km2 (3,100 mi2) (15). C, Northern
compound crescentic dune widths (W), lengths (L), and Sahara, Algeria and Tunisia. Area sampled is 5,000 km2 (3,100
wavelengths (X) , in kilometres, measured on landsat imagery mil) (10). D, Peski Karakumy near Gory Bukantau, U.S.S.R. Area
and Skylab photographs. W, mean dune width; I, mean dune sampled is 2,500 km2 (1,550 mi2) (16). E, Nebraska Sand Hills,
length; 1:, mean dune wavelength; 5, standard deviation; C, U.S.A Area sampled is 3,000 km2 (1,860 mi2) (19) . F, Gran
coefficient of variation. Numbers in parentheses refer to regions Desierto, Sonora, Mexico. Area sampled is 650 km2 (400 mi2)
listed in table 38. A, Eastern Rub' al Khali, Saudi Arabia. Area (12). G, Algodones Dunes, U.S.A Area sampled is 500 km2 (310
sampled is 10,000 km2 (6,200 mi2) (14). 8, Takla Makan Desert, mil) (13). (Fig. 177.)

Digitized by Google
 A STUDY AND SEAS

i i i i i i i i

2.'2.2 !
2.4

2.0
1.8 ."., nJ
~,,.. ,. 7

j
i i
~ ": ~
II! .-
.- .- .-
i
~ ...
q
i
... .,... ... ... ... .....
i
N
I
.- C'!
I i
~
I
IA
N
i
~
i
N
i i
~
I I
q .- C'!
(WI cwi (WI ::I ~
i
(WI
i
IA
cwi
~
i
(WI ::i ..
I
. I
;
(WI
I

cwi ~
I

~
W

3.0
3.2!
2.8
2.6
2.4
2.2
.' 21 -
.e"
1.0

,,,,,,,,,J
0.1
0.6

DIAGRAMS SH ION OF VALU ear dots refer ed in table 38.

dunes in sand s n landsat image oefficient. A, M h (W) versus m
y aerial photog p . dots represent unes wave ength (A); B, mean u e e g h (I) versus m
identified from aerial photographs and field reports as com- width (W); C, mean dune length (I) versus mean dune
pound ridges. Solid dots represent small simple(?) crescentic wavelength (A). (Fig. 178.)
ridges and large barchans. Measurements in kilometres. Num-

ferred to as ' Somecomp ra, Mexico, ntermediate

ic dunes inc e study, suc een the e rved and e
the Algodon California. U ht varieties, rrespondenc
D). the Pes near Mary rand to dune main varieti
kantau. U.S. the Gran Des very poor to ut they are

:"VU I.J] _ "-'

 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 275

ciently distinct from the others to be identified with dunes that are crescentic in shape but have star
separate names. dunes on their crests (fig. 179D). Toward the south
Only compound dunes, among the crescentic (Gulf of California), star dunes coalesce in east-
forms, were compared by dimensional analysis. west linear chains. The largest star dunes observed
Crescentic dune ridges in several sand seas that are in this study are in the Grand Erg Oriental, Algeria
suspected to be compound cannot be definitely (chapter K, fig. 2108).
classed as such because of poor resolution on Land- Analysis of the distribution of star dunes in rela-
sat imagery and lack of aerial photographs. Such tionship to distribution of effective winds (chapters
regions include the An Nafiid in Saudi Arabia and F and K) suggests that star dunes occur where
parts of the northeastern Sahara of Algeria and effective winds blow with moderate strength from
Tunisia. Dunes of equivalent size in the Gran several directions.
Desierto, Sonora, Mexico, have been determined
from Skylab photographs to be compound. Morphometry
Histograms (fig. 181) show frequency distribu-
tions of star dune diameters and wavelengths in the
Star Dunes sampled sand seas. Measurements in ten 2,500-kmz
General Distribution (l,550-miZ) sample areas, and four smaller areas of
Landsat imagery and Skylab photographs are sum-
SIMPLE, COMPOUND, AND COMPLEX star marized in table 40. As in sand seas of other types
dunes were examined on Landsat imagery and on of dunes, star dune measurements tend to be
Skylab and aerial photographs. Simple star dunes grouped about characteristic mean values, thus
have three or four arms, all with slipfaces of about defining the regular pattern for each sand sea.
the same size (fig. 179A). Compound star dunes However, the scatter diagram (fig. 182) shows only
have both large primary arms and small subsidiary a very weak association between mean diameters
arms with slipfaces (fig. 1798). Complex star dunes of star dunes and their mean wavelengths in the
are combinations of star and other dune types, in sampled sand seas. The weak correlation (r = 0.52)
which the star element predominates. Examples are indicates that spacing of star dunes, unlike that of
star dunes with crescentic curvatures, recorded in linear and crescentic dunes (figs. 172, 178) does not
the southeast Rub' al Khali, Saudi Arabia (fig. vary directly with their size. Differences in the rela-
179C) , and Gran Desierto, Mexico (fig. 179D) , as tionship of dune spacing to dune size in different
well as star dunes in linear chains in the Gran types of dunes may reflect differences in processes
Desierto, Mexico, and the Grand Erg Oriental, of growth. Star dunes, unlike linear and crescentic
Algeria (fig. 180A. D). dunes, grow upward, rather than forward or
Star dunes in simple, compound, and complex laterally, in response to effective winds.
forms are the dominant dunes of the Grand Erg
Oriental, (fig. 1808), but elsewhere they are limited Identification of Varieties
to rather small areas. In the Ala Shan Desert, China,
and the Namib Desert, South-West Mrica (chapter Ratios derived from measurements of mean
K, figs. 264B. 224), star dunes occur around the diameters and mean wavelengths of star dunes in
margins of sand seas, typically near topographic sampled sand seas (table 40) are compared in table
barriers. In southeast Saudi Arabia, Yemen, and 41. Reliability of the comparisons is limited by the
weak association of diameter and wavelength
Oman, small star dunes occur near the mouths of
among star dunes (fig. 182). Also, differences in
dry washes along the southeastern edge of the Rub'
degrees of correspondence of scale ratios support
al Khali, and they grade northward and westward
the suggestion, based on aerial photographs, that
into much larger star dunes (fig. 179C), which in
turn merge with very large southward-migrating three apparently different varieties of star dunes
crescentic dune ridges that have star dunes on their exist:
crests (fig. 1748). In the Gran Desierto, Mexico, 1. Sharp-pointed radiating pyramidal dunes are
simple and compound star dunes occur around the typical of parts of the Gran Desierto, Mexico,
Sierra del Rosario (fig. 180A). To the east (toward the Ala Shan Desert, China, parts of the
Cerro Pinacate), star dunes grade into complex southern Sahara in Niger, and much of the

Digitized by Google
276 A STUDY OF GLOBAL SAND SEAS

--

\ ... . Complex
crescentic
~ dunes

\ . ~{'""\,!_
Sierra d ' I
e Ro
6.L
- ~~-.
rio

Star dunes Compound

... .:..
." ' "
" 10 KM crescentic
. '" - " , dunes
D ,
I."'. :
. /
"

r.
5 MI
\. "
'

115"00'

Digitized by Google
 MORPHOLOGY AND DISTRffiUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 277

AERIAL PHOTOGRAPHS OF SIMPLE, COMPOUND, and com-

plex star dunes (facing page). A, Simple star dunes, Gran Desier-
Parabolic Dunes
to, at approximately lat 32 0 N., long 115 0 W. Arms of these General Distribution
dunes are generally less than 100 m (330 tt) long. Photograph by
TWO FORMS OF PARABOLIC DUNFS are
E. Tad Nichols. 8, Compound star dunes, Gran Desierto, Sonora,
recognized on Landsat imagery and on aerial photo-
Mexico, at approximately lat 32 0 N., long 114"30' W. Diameter
graphs. Simple parabolic dunes are U -shaped or
of these dunes averages 700 m (2,300 tt). One dune is outlined
for contrast with smaller, simple star dunes. C, Complex star
V -shaped and consist of a convex mound with two
dunes, eastern Rub' al Khali, Saudi Arabia, at approximately lat
long trailing arms. Dunes of this kind at White
20 0 45' N., long 54 0 45' E. Photograph by the U.S. Department of
Sands National Monument, New Mexico, U.S.A.
Defense. 0, Star dunes in the Gran Desierto, Sonora, Mexico,
(McKee, 1966, p. 25, 50-51) are shown on an aerial
grade eastward into complex crescentic dunes with stars on
photograph (chapter E, fig. 50).
their crests, and these grade into compound crescentic dunes
near the southern Cerro Pinacate. Photography by National Compound parabolic dunes are U -shaped,
Aeronautics and Space Administration (SL 4 -92 - 358). V-shaped, or rake-shaped and have numerous se-
(Fig. 179.} condary trailing arms. In the Thar Desert (India and
Pakistan, chapter K) rakelike clusters of parabolic
Grand Erg Oriental in Algeria. Similarities of dunes cover about 100,000 kmz (62,000 miZ). Their
dunes in Mexico and in Algeria are illustrated planimetric aspect was sketched (Verstappen,
in low-altitude aerial photographs of a part of 1966&, fig. 3; A. S. Goudie, written commun. to
the Gran Desierto (fig. 1798) and a part of the McKee, 1974) from aerial photographs and maps
Grand Erg Oriental (Capot-Rey and Capot- (chapter K, fig. 254). Their aspect on Landsat imag-
Rey, 1948, pI. 2). ery is shown in figure 183A. The largest parabolic
2. Rounded compact star dunes with arms that do dunes observed on Landsat imagery are near the
not project far from the main mass of each Colombia-Venezuela border at about lat 6 30' N., 0

dune (H. T. U. Smith, 1963, p. 45) are typical long 69 W. (fIg. 1838}. These dunes are U-shaped
0

of the southern margin of the Grand Erg or V -shaped rather than rakelike and have
Oriental, Algeria (chapter K, fig. 2108). Cor- numerous trailing arms as much as 12 km (7 mil
respondence of the southern rounded variety long.
to the sharp-pointed variety of star dunes Parabolic dunes are described from the Mid-
farther north in the Grand Erg Oriental is western United States (Melton, 1940, p.126;28; H.
very poor to fair (table 41). T. U. Smith, 1948; Ahlbrandt, 1974b, p. 45), coastal
3. Sharp-crested dunes with arms greatly elon- Oregon and California (Cooper, 1967, p. 22-25),
gated in a preferred direction may be a third White Sands National Monument, New Mexico
variety. Such dunes are observed in the (McKee, 1966, p. 50-51), localities in coastal Brazil
Namib Desert (chapter K, fig. 224), where (chapter E) and Europe (Kadar, 1966, p. 446).
they grade into complex reversing dunes, in Parabolic dunes in the Thar Desert and at White
the eastern Rub' al Khali, where they also are Sands are alined parallel to the resultant direction
associated with complex dunes (fig. 179C) , of effective winds from the southwest (dune noses
and in the northern Grand Erg Oriental, point to the northeast) (McKee, Breed, and others,
Algeria. These dunes show excellent corres- 1974, figs. 25,5; this publication, chapters F and K
pondence to each other but very poor to fair and figs. 252, 253).
correspondence with dunes of the first two
varieties. Morphometry
Star dunes described by McKee (1966, p. 65-M) Compound parabolic dunes in the sampled part
in the 'Irq as Subay', Saudi Arabia, are different of the Thar Desert (fig. 183A) have an average of
from all these varieties, for they have extremely seven arms, mean lengths of 2.6 km (1.6 mil, and a
long arms that radiate far from the main body of mean width of 2.4 km (1.5 mil. Their length-width
each dune. These dunes are not compared with ratio is 1.1. Compound parabolic dunes in South
other star dunes in table 41. America (fig. 1838} have a length-width ratio of 3.7,

Digitized by Google
278 A STUDY OF GLOBAL SAND SEAS

114"30'

32"00'_

LANDSAT IMAGERY SAMPLE AREAS OF STAR DUNES in Arabia

and western North America. Each area is 2,SOO km2 (1,SSO mi2).
Numbers in parentheses refer to regions listed in table 40. A,
Simple, compound, and complex star dunes, Gran Desierto,
Sonora, Mexico, including the arrays shown in figures 179A and
B (1). B, Compound star dunes in the southern Grand Erg Orien-
tal, Algeria (7). C, Compound and complex star dunes,
southeastern Rub' al Khali, Saudi Arabia (S). Dunes in the north-
ern part of this area are shown in detail in figure 179C. 0, Com-
plex star dunes (chain variety), Grand Erg Occidental, Algeria.
(Fig. 180).

identical to the 3.7 length-width ratio of much

smaller parabolic dunes at the southeast end of the
White Sands dune field. New Mexico. U.S.A.
(chapter E. fig. 50).

Dome-shaped Dunes
General Distribution

MOUNDS OF SAND that generally lack external

slipfaces have been described as dome dunes in the
53"30' Western United States localities. such as White

Digitized by Google
 MORPHOLOGY AND DlSTRffiUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 279

40...------_____-----, HISTOGRAMS SHOWING FREQUENCY DISTRIBUTION (f) of

A D-OJM i-2.08 diameters (0), and wavelengths (A.), in kilometres, of star dunes
30 S-0.24 s-o.s.. measured on Landsat imagery and Skylab photographs. 0, mean
e-G.28 e-G.28
dune diameter; I, mean dune wavelength; 5, standard devia-
10 tion; and C, coefficient of variation. Numbers in parentheses
o 1--+--+-+-+~i""'""I refer to regions listed in table 40. A, Southeastern Rub' al Khali,
31.4 0.8 0.8 1.0 1.2 1.4 1.8 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Saudi Arabia. Area sampled is 12,500 kml (7,770 mil) (5). 8,
1'-0.70 i-1.19 Gran Desierto, Sonora, Mexico. Area sampled is 1,000 kml (620
20 S-0.24 S-O.88
... e-O.35 e-0.74 mil) (1). C, Ala Shan Desert, China. Area sampled is 2,000 kml
10 ......._ _
I:::-~
(1,240 mil) (4).0, Southern Sahara, Niger. Area sampled is 1,000
o I--+--+--f--f===! kml (620 mil) (3). E, Entire Grand Erg Oriental, Algeria. Area

Q
200.2 0.4 0.8 0.8 1.0 1.2 0.0 D.4 0.8 1.2 1.8 2.0 2.4 2.8 3.2 sampled is 10,000 kml (6,200 mil) (6). F, Southern margin ofthe

EJ I I bJ:g~j
D-0.74
... 10 S-0.21 Grand Erg Oriental, Algeria. Area sampled is 2,500 kml (1,550
e-G.28 mil) (7). G, Northern part of the Grand Erg Oriental, Algeria.
o Area sampled is 7,500 km2 (4,660 mil) (8). (Fig. 181.)
300.3 0.5 0.7 0.9 1.1 0.1 0.5 0.9 1.3 1.7 2.1 2.5 2.9 3.3
D 1'-0.81 i-1.O
...
20 S-O.27
e-O.4C
S-o.s..
e-O.54
10
o 1---+--4--fII~---i~
0.1 0.3 0.5 0.7 0.9 1.1 1.3 0.0 D.4 0.8 1.2 1.8 2.0 2.4 2.8 3.2 2.8 r-~-....--r----.-....--r----.-....--r----.---'

... 10
20

o
DIAMETER

,
WAVELENGlH

,
D-o.86
~:~:
i
J
0.2 0.4 0.8 0.8 1.0 1.2 1.4 1.8 1.8 2.0 2.2 2.4 2.8 2.8 3.0
DIAMETER
~.--------------------.
E
50 i-2.07 1-<
S-1.05
40 e-O.11
"'30
20
10
1.2

1.0 ~a---L_---L.._....I-_"'-------L_---L.._....I-_L----L_....L----.J
0.8 0.7 1.1 1.2 1.3 1.4 1.5 1.8 1.7

... 10
1J
o F=~-+--+--f===t'=9
0.8 1.0 1.4 1.8 2.2 2.8 3.0 1.4 1.8 2.2 2.8 3.0 3.4 SCATTER DIAGRAM SHOWING RELATIONSHIP of mean dune
DIAMETER WAVELENGlH
diameter (D) to mean dune wavelength (i) in star dunes
G i-2.04 sampled on landsat imagery and Skylab photographs. All dunes
S-0.81 are believed to be compound or complex forms. All measure-
e-0.4O
ments are in kilometres. I, correlation coefficient. Numbers in
20 .---------....-:-:="1 parentheses refer to regions listed in table 40. -(Fig. 182.)
... 10

... 50 o 1--+--+-+4--l
0.3 0.5 0.7 0.9 1.1 1.3
DIAMETER
40
30
20
10
01---+--4--+~~1--+--+-~-+~-r-~~
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 8.0 8.5 7.0
WAVELENGlH

Digitized by Google
280 A STUDY OF GLOBAL SAND SEAS

01"00'

o 10 20 1I KM
ro-r'L---r-1r......,----I1
~ 1~ 1~ MI
COMPLEX, GIANT DOMES IN THE NAFUD ATH THUWAYRAT,
northern Saudi Arabia, east of Buraydah at about long 44 30' E.,
lat 2645' N. Dark spots in the interdune areas are oases.
LANDSAT IMAGERY SAMPLE AREAS OF Diameters of dome-shaped dunes in this sand sea average
PARABOLIC DUNES of the (A) Thar about 1 km (0.6 mil. The appearance of these dunes on Landsat
Desert, India and Pakistan, and (8) Colom- imagery is shown in chapter K, fig. 244. Aerial photograph by
bia-Venezuela border region. (Fig. 183.) the U.S. Department of Defense. (Fig. 184.)

Sands National Monument, New Mexico (McKee, similar dome-shaped appearance on Landsat imag-
1966, p. 26-27), and the Killpecker dunes, Wyom- ery, which have not been studied on aerial photo-
ing (Ahlbrandt, 1974b, p. 45). Terminology for graphs nor described in previous work, are ob-
dome-shaped dunes is summarized in table 30. served in the northern Grand Erg Occidental,
Small dome-shaped dunes are not visible on land- Algeria (fig. 185A) , and in the northern Takla
sat imagery or on space photographs, but a few Makan Desert, China (fig. 1858).
groups of very large dome-shaped dunes have been In each of the three geographic regions just refer-
observed. In the northern sand seas of Saudi red to, large dome-shaped dunes are located at the
Arabia, such large features are described (Holm, far upwind margins of sand seas. Small dome dunes
1953) as giant domes. are located at the upwind end of both the White
On aerial photographs (fig. 184), the Arabian Sands dune field (McKee, 1966, p. 26-27) and the
dome-shaped dunes are interpreted as complex for Killpecker, Wyoming, dune field (Ahlbrandt,
their rounded crests are covered by rows of cres- 1974b, p. 45). Locations of dome dunes in both these
centic dune ridges. Adjacent domes are separated fields may be related to the presence of fairly strong
by low ridges of indeterminate character. The and unidirectional effective winds, as suggested by
nature of these dome-shaped dunes on Landsat im- analysis of wind patterns (chapters F, K) and by in-
agery is shown in chapter K, figure 244. Dunes of ternal structure.

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 281

lOA I-3.34
... 2010 S-1.18
e-o.36
o F=t--t--"-I--I--"","",
0." 0.8 0.8 1.0 1.2 1." 1.8 1.8 0.51.5 2.5 3.54.5 5.58.57.5 8.5
DIAMETER WAVELENGTH
40.
30 l'-1.21 I-2.18
S-G.28 S-O.
"'20 e-O.23 e-0.32
10
0~~-4~::::::j::::=l
0.8 0.8 1.0 1.2 1.4 1.8 1.8 0.1 1.2 1.8 2.0 2." 2.8 3.2 3.8 ".0 .....
DIAME1fR WAVELENGTH

2O~
C
... 10
D-1 .51
e-0.19
S'"'O.29
~ e-0.2"
S-O.83
I-3.42

o
0.1 1.21.' 2.0 1.5 2.0 2.5 3.0 3.5 ".0 " .5 5.0
DIAMETER WAVELENGTH
.~
HISTOGRAMS SHOWING FREQUENCY DISTRIBUTION (I) of
diameters (D) and wavelengths (X), in kilometres, of dome-
shaped dunes measured on Landsat transparencies. D, mean
dune diameter; X, mean dune wavelength; 5, standard devia-
tion; C, coefficient of variation. Numbers in parentheses refer to
regions listed in table 42. A, Takla Makan Desert, China. Area
sampled is 2,500 km2 (1,550 mi2) (1). 8, Sand seas, northern
Saudi Arabia. Area sampled is 3,000 km2 (1,860 mi2) (3). C,
10 Sahara, Algeria. Area sampled is 2,500 km2 (1,550 mi2) (2). (Fig.
I I I I 186.)
5 10 15 MI

43) of all observed dome-shaped dunes show close

LANDSAT IMAGERY SAMPLE AREAS
SHOWING DOME-SHAPED DUNES. A,
similarity of these features.
Giant domes(?) in the northern part of the
Grand Erg Occidental, Algeria, north of EI
Golea. 8, Giant domes(?) in the northern
Sand Sheets and Streaks
Takla Makan Desert, China. Some of these General Distribution
features have linear "tails" 0.9-7.0 km
(0.6-4.3 mil long. (Fig. 185.) IRREGULAR PATCHES OF EOLIAN SAND
with external shapes definable on Landsat imagery
in terms of the bounding topography, and without
discernible slipfaces, are herein termed "sand
Morphometry sheets." Elongate patches with sharply defined
Measurements of domes and dome-shaped dunes edges as seen on Landsat imagery are herein termed
were made for five sample areas of Landsat imag- "streaks." (These are the sand stringers mentioned
ery, partly shown in figure 185 and summarized in by other geologists in this publication). Other
table 42. Histograms (fig. 186) show that diameters authors' terms for these features are shown in table
and wavelengths of dome-shaped dunes are 34.
broadly distributed about characteristic mean Sand sheets are ubiquitous to nearly all sand seas
values in each sand sea. Samples are too few, studied, but sand streaks are less common and seem
however, to test relationships of mean diameter to be associated mainly with sand seas of linear
and mean wavelength between dome-shaped dunes dunes. Figure 187 shows large eolian sand streaks
of different sizes and in different geographic loca- in the northwest Sahara of Mauritania. Figure 167B
tions. Comparisons of dimensionless ratios (table shows smaller streaks on the Navajo Indian Reser-

Digitized by Google
282 A STUDY OF GLOBAL SAND SEAS

barchanoid dunes, some of which are discernible on

high-altitude aerial photographs.

Interdune Areas
General Distribution
DUNES IN SAND SEAS are commonly separated
from each other by interdune areas whose well-
defined shapes are associated with dunes of certain
types, regardless of scale of the dunes or their
geographic location. For example, interdune areas
in sand seas of linear dune type are generally open
corridors called "couloirs" in the formerly
"French" Sahara, "straate" in the Kalahari and
Namib Deserts of southern Africa, and "sandy
lanes" in Australia. Some local names distinguish
sand-covered interdune corridors from those that
are bedrock.
Names for interdune areas in sand areas of cres-
centic dunes generally reflect the degree of closure
of the segments of these dunes and, thus, suggest
the ease or difficulty of travel through the sand sea.
An early name applied by geographers to fully
enclosed interdune areas in sand seas of crescentic
dune type are "bajirs." Hedin (1905a), used this
term to describe large flat-floored basins between
dunes in the eastern Takla Makan Desert of China
LANDSAT IMAGERY SHOWING SAND SHEETS and streaks (fig. 175D). Equivalent terms in the Arabian and
(stringers) associated with linear dunes in the northwestern Sahara sand seas are "fuljes" and "feidjs." The
Sahara, Mauritania. The area shown is about 34,000 km2 (31,100 term "fulje" was applied (Melton, 1940; and H. T.
mil). Effective winds in this region are from the northeast. (Fig. U. Smith, 1946) to interdune hollows in the Mid-
187.)
western United States.
vation in northern Arizona, U.S.A. Other areas Examples of small-scale interdune areas of the
where streaks are conspicuous are the southern fulje type are shown in aerial photographs of White
Sahara (chapter K, fig. 201) and the northern part of Sands and northwest Sahara (figs. 173B, q; large-
the Arabian Peninsula (chapter K, fig. 200). scale interdune areas of the same type are shown
Analysis of effective winds in these regions (chap- on figures 174A and 175B of the eastern Rub' al
ters F and K) suggests that sand streaks are associ- Khali, Saudi Arabia. Various local names for inter-
ated with strong, dominantly unimodal effective dune areas are listed in table 35.
winds. No measurements were made of sand sheets A density slicer with planimeter can be used to
or streaks. measure relative areas of dune and interdune sur-
Sand streaks (stringers) were noted on orbital faces in some sand seas, such as the Grand Erg
photographs by Verstappen and van Zuidam (1970, Oriental. Algeria, where Landsat transparences
p. 41). Ground investigations were made by S. G. show extreme photographic contrast between dune
Fryberger of such sand bodies in an area northeast and interdune surfaces. The technique is only
of the Little Colorado River, Arizona, U.S.A. Sand- locally successful, however (chapter K), so com-
streak surfaces there had numerous small parisons from desert to desert have not been made.

Digitized by Google
 Tables 26-39
Th_ tables sive chl'Ono\ogica1lists of local tenna for various types of dunes and the morphometry of such dunes
derived from measurements of I8mple _ on Landsat imqery and Skylab and aerial photographs.

Digitized by Google
284 A STIIDY OF GLOBAL SAND SEAS

TABLE 28. - OIrono'osicollfst of tenn. considered equivalent to dIMes c'_fied O8lfneor. bosed on Landsat imagery
._---------------------------------_._-------------------
T.TM (r.ferenC"f" p.~e.
ftlure, plate, and tabl .. nUlibera in parentheses)
... __.. _--------------
AUlh,.r Vpar 51..,1" rid." c-pound rldle
(and r~R illn)

H,.dln. SVt'n 1905. Th ...hold. (235-236). IS_I dune ....... I.tlona ()54-355).
(T.kla llakan o.-."rt. China). Ordinary ridge. ()23; fig. 267). Dun.. -range (357).
---------------------~-------------------------------------------------
Aufr"re. Leon 1928 Lea dun..a lonlitudlnale. (833). 15..... 1 chilnes de sable (833).
(Po.--rly "Preach" Sahara, ChUn. . lonlltudln.le. (833).
IIorth Afrlc.). 1930 do. (220. 227. 229). Dr or bra. (220).
(5..e1 chilne. d. sable (220. 230).
1932 Oblique dunea (42. 44). No tem.
1934 Dune. de conjonctlon. dune. d'lncidence (132. 138). Chilnea .vec v.lue. lonlltudlnal (al e1.) (137; fli. 2).
15_ I clllne vec v._. obliques (137; fli. 2).
chilnea dl.lona1 (1)9).
IS..e1 dune. de conjonctlon et dune. dlncidenc. (132).

"lno1d. R. It.. 1933 51nll" .elf dune. (107-112. 122). lIb.leback rldle (107. 125).
(Llby.n de.ert. IIorth It.frlc.). Slnlle line or ranle of self. (122). lei t. of .elf dune. (l08. 109).
1941 Lonlltudlna1 dune elf (189. 222-225). lIb.leback. (230-2)1; pI. U.).
Self dune chain (224_-::.22...:5..:.)...:._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
lladllan. C. L. 1936 Sandridge. (206. 210-227). JIo teb.
(51....... De rt. Au.tr.lIa). Lonlltudlnal rldlea (224-227).
1938 Sandrldle. (513-515). Do.
Lonlltudlna1 rldl (51).
1946 S.ndrldle. (45-63; pl. 5-8). [5_1 II.nt aandrldl" .nd wha1eback aandrldp.
(flp. 1. 2; pI. 8).
Strallht .nd par.llel ndrldle. (57).
Self (pI. .Iouf) (50). Seit rldl cro lng aand rldle t _11 anal. en
Cre.t or . .If (61). echelon (SO).
1la1ton. F. A. 1940 Lonlltudl""l dune. (113. 121; fll. 9. 10). IS-I 10nlltudlnal dune. (121-l22).
(Mldwe.tern IIDlted St.te.). [5OMI vlnd-llbadov dun (120). Oblique wlndrlft dune.!?) (129. fll. 19. 23).
[ _ I wlndrlft dun (129-130).
[5_1 .ource-bord.rlnl 1. . dune. (122: fll. 10. 13). Oblique lonlltudlnal dun. . (fig. 27).
Mack. J. T. 1941 Lonlltudina1 dune (240-243). 10 ten.
(Nav.jo Indian .... n.tlon.
U.S.A.).

Cepot-"y. "'bert 1945 Sand ridle (400). Dune. of quadrature (par.Uel .If) (401).
(Fo . . . r1y "French" SaiNlr Lonaitudl""l fo ..... dune lonaitudi""le (401). Dune ch.in with .ulf (403: fli. 10).
IIorth Africa). Dune. of Incidence (oblique .If) (401).
1947b 110 te ... Cordone (87).
FUe. d iouf allona'. (89).
Cepot-",. lIIb.rt. .nd 1948 Lonlitudina1 fo ..... dune 10naitudina1. (67). Do.
Cepot-ley. Fr.ncoi.e Slf (pl. Iouf) (63. 66-73). Dos d. ba1.ina (whal.baclte) (63-70).
(ror..rly "Prench" Sahara. alb (pl.. aleb) (73). [ _ I diouf. bra or dr. . (72. 75).
IIorth Afric.).
Cepot-ley. lIIb.rt 1953 110 t .... Do. (145).
(Fo . . .r1, "French" Sabara.
IIorth It.frlca).
"!th. H. T. U. 1946 LonIHudl""l dunea (197). 110 tem.
(Rebruk U.S.A.; 1963 do. (1. 17-18). FI.hbone pattern of lonlitudin.1 dune ridg (18).
Sah.ra. IIorth Africa). Lee dune rldle. (34; fli. 6). c-pound lonlltudina1 dune. (20. 41 . fli. 13).
1965 Loqltudina1 dune. (566-568; pI. 3. 411.). 110 tel'1l.
Li.... r ridle. (557).
1968 Lonaltudina1 dune.; if dune. (43). Do.
Ho1a. D. A. 1953 10. ten. Uruq (ll2).
(Saudi Arabi.n d.Mrt.). 1957 IS_I ai.... lda1 dune. (1746). .....toaoainl line. . cOllJ>le"e. (1746).
1960 do. (1371; fli. 4). Linear cOllJ>le"". (1373).
LI .... r rid ... (1371). Uruq (1373. 1377; fli. 7).
1968 do. (978). Linear cOllJ>l."". (978).
[5_1 at .... ida1 dune. (978).
McKee, I. D. 1957 Loqitudina1 dune (1720-1721). No tera.
(1Ihite Sand., U.S.A.; 1966 do. (16: pI. nl). Do.
Lib,... and Saudi Arabian S.lf dun. . (3. 10. 16. 61; pI. nl).
d ert.).
Coope<. V. S. 1958 Lonaltudi""l dun. . (7. 49, 60-(4). [5_ I oblique rldle dunea with bua1 plinth
(V t cout. IIorth _rica). .nd .ctlve creat. (49).
Self if chain (61).
[5oM1 oblique rid. . dune. (7. 49. 59).
1967 do. (109-ll0). Ito tera.
Lonlitudlna1 e1 ..... t. (26).

Grov A. T. 1958 Linear dune. (526). No tel'1l.

(Southern Sahara. North Afric.; Lonlitudlna1 dune. (526-530).
Kal.h. . I De.ert. lIouth 1960b do. (203) [ - I .1ab dunes (203).
Afrlc.). 1969 do. (195-202. 2ll; til. 3). Do. (195-201. 208).
Linear dune. (195). IS-I qoz (195).
Sand ridle. (195-202. 211; fli. 3).

MDnod. Theodore 1958 Le aUk (pl. 10Ok) (25. 142. 147-150; fil 10(2). Le lib lonlitudln.1e (89).
(Por.erl, "French" S.ura, 33)
IIorth Africa). Le faille 10nlhudine1.: Ie aUk. Ie. cordon. vlfo It.lib: cordon. _jeure. et de .10Ok auperl_M.
(.loii1t) et l'elb (pl.. alib) (141. 147-150). (71-72; fli. 33).
SloGk oblique .ur e1b (141. 148).
Irovn, G. , .. 1960 Lonlltudin.1 or aelf dunes (156-157). No teb.
(Ar.blan de.erts).

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 285

TABLE 26. - ClIronologicallist of tenn. considered equivalent to dWl88 classified a.linear, based on Landsat imagery-
Continued
Tt" .... (rE'feren('f' page. flaure, plate, and tabl. nUlibera in parenthesea)

A Yt"a SllIPle rldll

lorrie, I. S. 196 (618; fil. 1; pI. 10 tera.

CAl1fonla

196 (294-295, 305). Do.

CAl1fonla ).

Povera, t. W.. aDel otbera 1966 [Sa.e) ufiiq; .and. rid, and dUDe cbalu;, . . yf dune. I - I urUq (0100).
(Saudi Arabian de.ert.). (0100)
Habbutt, J. A. 1967 Lonaltudlnal dune. (149). No tem.
(Auatral1a) 1968 do. (143).
IS<>aIe I parallel rlda" (141). ISoaoel parallel rldle. with obllqu. cr t. (141).
Slng1.-cre.ted lonaltudlnal rldae. (144). Oblique cre.ta1 dune. (143).
Short dune. (144-146). ClI.ax dune type (47).
p 196 and ridae. (301; 10 tem.

5 196 une, rid,e, linear Sand levee (2

46) . Wh.leback (25
G 1970 r .e H dune. (89-9 Pletstocene a nd. (9G-93; fla. 74
aerta)

197 94) Alab ridge. (

(Southern Africa). CoIIf>o.1 te fOBe (95).
C1uatered dendritic rldae. -(97; fig. 2.).
Goudie, A. 5 Allchln. Irldlet, 1973 Longitudinal dunes (245-246, 248). No tera.
and Hedae, K. T. M.
(,I"ar Desert. India).
Wllaon. I. G. 1971b Longitudinal dunes (196). Draa. (182, 191.198).
(A1aerlan Ergs). SeH dunes. self dune chain. 086-187; fig. 4).
1972 do. 082-183; pI. III I, C). Cooopound (seH) dune rldae (208).
Longitudinal dune. 093. 201; pI. III C).
itudlnal dunt"8 (19
197 If dune chain. (84 No ten.
-------
C arren, Andrew 197 une. (295, 301. 30 Draa (50, 231
(295, 301, 304, 30
1-304. 309, 317, 3
I Kee, E. D. 197 160-171) . No tenl.

reed, c. s. 197 , flga. 9-3, 9-4). Do.

(All deserts). 1974b Parallel 8trataht or linear dun~8 (665, 668. 614; Ltn.ar and feath~red l1n.ar _Iadune. (614; f11. 3. 7).
fig. 3).
KcKe~, E. D., and othera 1974 do. O. 17. 52-55. 62, 64, 84; fig . 2. 9, 13. Do. 0, 17, 23,47.62,84; fig 2, 7. 13, 17, 20;
(All duerts). 17, 20-22; tabl~ 1). toble 1).
Mclee, E. D., and Breed, C. S. 1976 Poralipi strslght or linear dunee (81; fig. 55). No tef'll.
(All d~.erta).

ClIronologicall sidered equival sifiecJ as cresc Landsat imagery

entic. aa uaed here barchanoid ridle, ea of other leo101 atlonl

e erence Teru (ref~rence pale, laure, plate, aDd table nu.ber 1n parntbe. . . )
Author
Year SI-t>le _nd 5111f>Ie ridge CoIIf>ound ..,und CoIIf>ound rld,_
(and re,lon)

Hedin, Sven 1905a Id 1 crescent. (323; fig. 268). Fu.ed dunes, dune triplets CoIIf>oaite dune (271-273; (SOlIe] dune acc-..ulationa
(T.kl. Hak.n (351-352; fia 294, 295). figs; 238-244). (352-354; fia 296-
De.ert, Caine). Individual dune. (351; fig. Tranaver.e threshold. (316-)20, 298)
293) 361-363)
Barthana or .andrldle. (227,
324, 32
Cre.centi
Chain. of
(320)
Chains of 51 i
fig. 29
CoIIf>ound 8;
fig 30
1905b Cre.cen Chaine of No tera. 10 teB.
(404) (403)
Chain. of coalucinl d......
(416)
Aufr. . . , Leon 1928 Sif (pI. aiouf) or barchans Le. vasue. de ubl~ (833). No tem. No tem.
(rorM'rly "French" (833-835)
Sahara, North 1930 larch.n. (224, 226). No tem. do. Do.
Africa) 1932 No tera. Le. vogue. de ub1e (42). do. Do.
1934 do. do. (134, 137; fla. 2). ZeaJ.ah (pI. ......ul)( 7) o.fin tran.veraa}ea
(133) . (139)
Dun d'i 2)
Le. dune.
286 A STUDY OF GLOBAL SAND SEAS

TABLE 27. - Cluonological list of terms considered equivalent to dunes clOB8ified as crescentic, based on Landsat imagery-
Continued
Reference Ter.. (reference pa.e, f1l'1re, plate, and table D.-.r 10 parnthe... )
Author
n) Y.ar mund a Coapo

A. 1933 109-ll2) No No t. No t .....

rt, 1941 221) Co ..... (188). Coap1e Do.
rica) lIound
L. 1936 Ie dun Tr a (123- No t. 110 t .....
De rt,
ie) 223-225) Tr 27).
1938 larcb.a, crcantle dun do. (513) do. Do.
(514)
1946 do. (50, 56-58). 110 telW. do. Do.

!!alton, r. A. 1940 larcan dune (113-117, 140; Peat. aDd fu1je (fi,. 25). 110 telW. loter ctiDI trannerH
OUdwe.t.m UDitad fi, .1-5, 24, 26). Tranever dune r:l.. (113. d ..... a.ri (fla. 26).
Stat. . ). 117-118; fi,. 7).
Iaolated trauever dun.
rid ea (119' fl 8).

ndian
1941 ) . T
s.
I, 250).
).
110 te 10 tem.

--..
tioa, C
" (250).

IIobert 1945 II 10 tara.

"Preach" S
rth fi,.
Africa). 1947b Barkhan.. (96-97). Jlo tera. 110 tera. Do.
1953 do. (144). Slf (.uif) (144). do. Do.
CApot-Iay, IIobert.Dd 1948 do. (63, 66-69, 71-72). Dun. . trannerae1e. (73). _ , .."p10ida1 dUDe. Do.
Capot-lay Francol _ If, _atodon (73; pl.
(ro ....rly "rr...cb 11.1)
Sahara, lorth Q8rd (71)
Africa)

s.1th B. T. V. 1946 Barcbena (197). Tranevera. dune rid (177). 1to tera. 110 terw.
, V.S.A.: 1953 T 2). do Do.
Borth 1956 R do Giant Ca.p08
(1735)
1963 II. 2). T , 16-17, Coapo (19, Coapound be
40, 19, 39; f
1965 B do 10 tanl.
Coapo rcba-
noid
Coapo
) . Coapoalte be
1968
35, fl,. 3-3).

Bola, D. A. 1953 Barkhan dun. . (106). I-I dbarr (l10). 10 tem. No tem.
(Saudi Arabien 1960 Barchan dune. (1371). lSoae I dbar (1372). do. Do.
d rta> Crcentic dun (1373). Giant creacent. (1371). Giant cre.cent., aunt
cr c.ent ic . . . . if.
(1371-1372).
Creacentic hollow.(?) (1370, Giant creacent Ie dune
1373; fi,. 3). coap1ex (1375; fl,. 5)
1968 Cre.centic dUD. . (975). . . rchana in traonerae ridge 110 tem No tem.
977)
, 978). T 78).

1957 ) . N 10 te 10 tem.
Dd., 1966 25, 39-40, Ba adationa do Do.
Indian Ic) . ranaverae
tion, 9) IID)
Libyan
Borth
Afric.) .

Cooper, W. S. 1958 Barchan. (27-28). Undulatin, dune rid,e (7). HURe barchan with .yate. of No tem.
(Weat coast, lorth Co.pre ed isodlametric Tranaverse ridlea (2S, 27-34, ana a toao8inS transver.e
Merica) . transverae ridge (28). 58) . rld,e. on I it.1 back (28).
Tranllver.e wave. (33-34, 43).
1967 No tena. do. (5). No term. Do.
Transverae ridles (2l t 31,
109) .
IIa fig. I;

1958 T 26-531) . No te No tera.

Sahara, 1960b 2) do Do.
frlca;
I Desert,
1969 fig. 10). "
C do Do.

n Africa).
dore 1958 ). atnl"urs, Elem(> Bourreleta
(For_rly ttFrenC'h" undu1at ions du tYPl" Myreyye majpuree-le8 pspudo-a14b (141) .
Sahara, North (2S, b8, 143, 145, 170; fl~. (25, b8. 141-142; fig.
Africa) 10; pI. 5). 10(2) ).
S@rie de bourrelets
tranSVE'rsaUX (145).
BourTE'lets geants (147).
L okl';. typlqu.s (92-94. 141;
figs. 15-18).

Le (141).
Le Iques,
erie de
e. (143).
 MORPHOLOGY AND DISTRIBlITION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 287
TABLE 27. - Chronological list of terms considered equivalent to dunes classified as crescentic, based on Landsat imagery-
Continued
Reference Terms (rpferenct" page, figure, plate, and table nUllber in parentheaeB)
Au
Year 5 Co~un eo.pound ridge

No 1961 aarchan Transverae eo.plex barchan plex barchan. (60

617)
Very large. Dve ry 1ar8e overlappl
barchans (617 .... ch.n. (617).

No 1966 Barchan No tetw. Glant barchans lant barchan

(AlgandoR Dunea, (292-296) _gabarchanB (292.
Californi. and 296)
Mexico) Crescentic dunea (292).

Kad'r, Laalo 1966 No tena. Crescent dunes or barchans. Two-storied barchans (441). No ter'll.
(Centra. Europe). Twin barchans, transver.al
dunes (438, 447; fiS. 28).
Pover R. W., and 1966 Simple barchan dunes (0100).
--------------------------------------------------
Compound barchan dunes, or Complex bare hans , giant eo.plex barchans g1ant
others aimple rounded tranSverae barchan .. (0100). barchan. (0100).
rldse (0

Pe 1967 Solitary - )01; aarkhan C'ha No tera. ..,lex barchan ridg

flS. I I02(); t or "davana n (295,
300; flS. 101).
Ridge-bareh 1veolar davana. or
alveolar barkhan s
(298, 300; table 3
Stone, R. O. 1967 Barchan; also, barcan, barehan.:'" Clltnbing dune (220). Me,abarchan (232). No ter..
(All deserts). barkan. barkhan (21S). Transverse dune (251).
Crescpnt 1c dune (220).
Mabbutt, J. A. 19b8 Barchans (144). Transvl'rse crescent ic links, No ter... No ter..
(Kalahari Deaprt, transverse elements cres-
southprn Afr1ca, cent ic tranSVerat> e 1e-.ntS,
S1_paon Oeaert, crt.'scentiC' transverse dunes
Auat ra 1 tal . or ridge. (143-144).
Go 1970 Barchans No ter.. o ter'll.

Go 1973 No tenl. Transverse do. 00.

249)

a)

GI 1970 Barchan -88) Barchanoid No ter.. iant barchan co.p 1e

(88; fiS. 72).
deserts) Transverse dune. (95-99). Barchan-like complexes
(98-99; fl,. 78).
Wilson, I. G. 1971b Barchan dunea (185-186; fI,. 4). Transverae dunes (196). Ora. . (182, 191, 198). Dra.s (182, 191, 198),
(AI,erlan Ersa). 1972 do. (181,192,208). Crescentic barchanoid or Transverse draas (191). Transverse draas (191).
transverSe sandy ridge
(180-181).
Transverse ridges (182, 185, Barchanoid draa. (178-179; Ilarchano1d dra.. (178-
190, 199-202). pI. lID, E). 179); pI. lID, E).
Giant cOIIPos1te dunes
(draa. J (181).
1973 Barchan No terti. do. (100; f archan01d draas (10
(fI,.17).

Co 1973 Barchans Transverse Draa-sized bare raa-aized barchana

297, 10; 283, 291 283) (282-28)
flsa.
aarchanoid Mesa-barchans ( ,a-barchano (297).
tranaver
297, 311~ ).
(So .... J Elba. (l09, 32 7) .

Breed, C. S., and 1973 Barchan (163). Crescentic dunea (162). No tel'1l. ~o tenl.
McKee, E. O.
(All deserta).
McKee. E. D., and 1974b No term. Parallel wavy or crescentic Crescentic .egadunes, giant Crescent 1c .egadunes.
Breed, C. S. dune. (665, 668; fisa. 9, creacents. or .egabarchan. glant creacents or
(All deserta). 10) (668; flS. 4,7). or . .gaharchans (668;
(fiS . 4, 7, 10).
1974 Barchan do. (3 Crescentic .ega rescent i(' -.egadunes
81; fI 17, 25; crescents. or giant crescents or
(3, 24, 69, _gabarehans (J, 2
BarC"hanoid 3, 7, 12, 15 74,83; figs. 3, 7
typeo, pe 14, 15, 19).
topograp unE' co.pleox.s ('Ir dr
figs. 3, (28, hQ-70, Rll.
17, 25). r.scf"ntic. fishscal
Transverse (fIS. 2~).
Akle (21,81,84). Transverse. parallE'1 wavy
(b2)
McKee, E. 0 and 1976 No tera. [Some) parallel wavy or cres- Meaabarchans or giant [SollE'1 parallel-wavy or
Breed, C. S centic dunes of the ail.ple- crescenta (82). erE'8CE'nt tC" dune com-
(All deserts). badc type (81-82). plex.s tnc1udinR (so .... )
ftsh8cale typf' and gtant
creacpnt or megabarchan
type (82; fig. S6).
288 A STUDY OF GLOBAL SAND SEAS

TABLE 28. - Chronological Jist of terms considered equivalent to dunes classified as star, based on Landsat imagery

Reference Terms (reference page, figure, plate, and table numbers in parentheaes)
Author Year Simple star Compound star
(and region)
Hedin, Sven 1905a Pyramidal dunes (330, 363). No term.
(Tak1a Kakan Deaert, China).
Aufrere, Leon 1930 No term. Oghroud (223).
(Formerly "French" Sahara, 1932 [Some) rhourds simples (44). Les grands rhourds polypyramides (45).
North Africa). 1934 Co11ines de sable (rhourds) (137, Rhourd po1ypyramide (141).
139-140; figs. 2,4(3.
[Some) rhourds degages (139).
Capot-Rey, Robert 1945 [Some) khurds, guerns (393, 397; fig. 7). [Some) khurds (397; fig. 7).
(Sahare, North Africa). 1947a Oghroud, mastodonte, iguidi (86, 89). Pyramides aux sommets pointus (89).
1953 No term. Do. (145).
Capot-Rey, Robert, and 1948 Qord, ghourd (72-73; pl. 113). Les grandes pyramides (oghroud) 73, 77).
Capot-Rey, Francoise Oghroud (pl. II, 2, no. 3).
(Sahara, North Africa).
Smith, H. T. U. 1946 Rhourd (198). No term.
(Nebraska, U.S.A.; 1963 Embryonic forms of peaked dunes (21, 44; Peaked complexes (21, 43; fig. 15).
Sahara, North Africa). fig. 16). Doma1 dune complexes (22, 45; fig. 17).
Holm, D. A. 1953 Sand peaks or star dunes (112). No term.
(Saudi Arabian deserts). 1960 Pyramidal dunes (1371-1373). Pyramidal dunes (1373).
1968 No term. Do. (977).

Honod, Theodore 1958 [Some) ghords (141). [Some) ghords (141).

(For_r1y "French" Sahara, [Some) chaines ghordiques (20).
North Africa).
Brown, G. F. 1960 Star dunes (159). No term.
(Saudi Arabian deserts).
Powers, R. W., and others 1966 [So_) sand mountains (D100). Giant sigmoidal and pyramidal aand peaks
(Saudi Arabian deserts). (D100).
McKee, E. D. 1966 Star dune (3, 10, 62-68; pl. VIII A, B). No term.
(Irq as Subay' , Saudi Arabia).
Petrov, M. P. 1967 Conical dunes (295, 300-301; figs. 10lc, No term.
(U.S.S.R.) 102(4); table 36).

Stone, R. O. 1967 Rhourd (242). Dune massif (225).

(All deserts). Star dune (249). Guern (230).
Rhourd (242).
Star dune (249).
Glennie, K. W. 1970 No term. Stellate dunes (86-88; figs. 70-71).
(Saudi Arabian deserts).
Wilson, 1. G. 1971a [So_) star-shaped pyramids, rhourds Draa peaks, rhourds (268).
(Algerian Ergs). (270)
Star-shaped pyramids, rhourds (270).
1972 No term. Stellate rosettes (178-179; pI. II C).
Draa peaks (oghourds) (188).
Oghourds or star dunes (190-191).
1973 do. Draa peaks, rhourds (99; fig. 16).

Cooke, R. U., and Warren, Andrew 1973 [Some) sand mountains or rhourds (304). [Some) sand mountains or rhourds (304).
(All deserts).

Breed, C. S., and McKee, E. D. 1973 Star or radial dunes (162; fig. 9). Radial (star) mega dunes (162; figs. 3,
(All deserts) 9)
McKee, E. D., and Breed, C. S. 1974a Star dunes (9-3). Star dunes (9-3; f1gs. 9-1, 9-2).
(All deserts). 1974b Star or radial dunes (665, 668). Star or radial megadunes (668; fig. 5).
McKee, E. D., and others 1974 do. (3, 17, 83-85; figs. 4, 8; Do. 0, 17, 24, 47, 83-85, figs. 4,
(All deserts). table 1). 8,10,12,17,19; table 1).
Dune complexes or draa (28, 82).
McKee, E. D., and Breed, C. S. 1976 Star or radial dunes (81). Star or radial mega dunes (82).
(All deserts)

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 289

TABLE 29. - Chronological list of terms considered equivalent to dunes classified as parabolic, based on Landsat imagery
Ref .. rence T.. rms (referenc .. pag.. , figur .. , plate, and table numbers in parentheses)
Author Year Simpl .. U-shaped dun .. s Compound U-shaped dun .. s
(and r .. gion)
H.. din, Sven 1905a Spoon-shaped dunes(?) (331: fig. 270). No term.
(Takla Makan Desert, China).
Aufrere, Leon 1928 Les dunes paraboliques et I .. s dun .. s en U No term.
(Formerly "French" Sahara, (834-835)
North Africa). 1930 Dunes de mousson, dun .. s .. n rat .. au (228). 00.
1934 L.. s dun .. s paraboliqu.. s .. t I .. s duns en U Les Chaines bouclees (134).
(134)
Madigan, C. L. 193b U-shaped dune, fulje-type dunes (224). No t .. rm.
(Simpson Desert, Australia). 194b Blowouts (57). 00.
Parabolic bowouts or fulj .. s (bl).
Capot-Rey, Robert, and 1948 Dunes paraboliques (75). No term.
Capot-Rey, Francoise
(Formerly "French" Sahara,
North Africa).
McKee, E. D. 1957 Parabolic dune (1720-1721). No term.
(White Sands, U.S.A.). Blowouts (1725).
19b6 do. (9-10).
Parabolic dune (3, 'I-II, 15,25, SO-51). Do.
U-shaped or V-shaped dune (9).
Grove, A. T. 1958 Parabolic dunes (529). No term.
(Southern Sahara, North
Africa: Kalahari Des .. rt, 19b9 Blowouts (199). 00.
Southern Africa).
Cooper, W. S. 1967 Parabola dunes (22-25, Ill). Compound parabola (31).
(West coast, North America). [Some) giant parabolas (24). [Some) giant parabolas (Ill).
Stone, R. O. 19b7 Blowout (217). No term.
(All deserts). Chevron dune (219).
Elongate blowout dune (22~).
Hairpin dune (230).
Parabolic dune (23b).
U-shaped dune (251).
Upsiloidal dune (252).
Melton, F. A. 1940 Blowout or parabolic dunes 033, 126-127). "Nested" elongate-blowout dune .. erit's
(Midwestern United States). (128) .
Elongate-blowout dunes (127-128; figs.
17, 18).
Hack, J. T. 1941 Parabolic dunes of deflation and No t .. rm.
(Navajo Indian Reservation, accumulation (240, 242-243).
U.S.A.)
Smith, H. T. U. 1946 Upsiloidal dune (197-199). No term.
(Nebraska, U.S.A.: 1953 Phytogenic dunes (102). Do.
Sahara, North Africa). 1965 Blowouts (557, 569-570). 00.
1968 do. (33). Elongate U-shaped dunes and subparallel
ridges (44).
Goudie, A., Allchin, Bridget, 1973 Parabolic dunes (248). No term.
and Hedge, K. T. M.
(Thar Desert, India).
Cooke, R. U., and Warren, Andrew 1973 Blowouts, parabolic dune, lunettes (318). Nested parabolic system (JI8).
(All deserts).
Ahlbrandt, T. S. 1974b Parabolic dunes (45). No tE'rm.
(Wyoming, U.S.A.).
McKee, E. D., and Breed, C. S. 1974b Parabolic or U-shaped dunes (~~5, ~68; Parabolic dune array, Thar Desert
(All deserts). figs. 6, 'I). (6b8: fig. ~).
McKee, E. D., and others 1974 do. (3, 24, 53, ~1-62, 83, 85; Do. (24, ~2: figs. 5, III, 25).
(All deserts) figs. 5, 9; table 1). Dune complexes or draa (28, 8~).
McK.. e, E. D., and Breed, C. S. 197b do. (81-82). Parabolic dunes, India (fig. 58).
(All deserts)

Digitized by Google
 A STUDY 0 NDSEAS

30. - Chronolo s considered eq

classified as do sed on Landsat J
erence pale I f ilure,
Author Year table nUlllben In parentheaes)
(and rei ion)
B_gnold. R. A.. 1941 Rounded mound_ (1) (217).
CLlby.n deurt. North Africa}.
Hol_, D. A. 1953 eoae-ahaped dune. (106-112; fiS . 2-8).
(Saudi Arabian de_efta).
McKee, E. D. 1966 Docae-ahaped dune. (l0. lS, 25-27).
(White Sandi. U.S.A,>,
W and others .and .,und. (Dl00).
rabhn duena).
T. S . (45)
U.S.A.).

TABLE 31. - Chronological list of local terms for dunes considered equivalent to those classified as complex, based on Landsat
imagery
Reference
Terms (reference page, figur~ plate, and table numbers in parentheses)
Author Year
(and region)
-----
en 905a [Some] ions (400).
Makan Desert,
Leon 930 [Some] r fluk (220, 2
ly "French" Sa Slassel oghounds (223
h Africa). 932 Cha ins 44) .
934 Les dun 133) .
Chaines avec vagues obliques et pitons (draas) (fig. 2).
Rhourds engages (1939).
Madigan, C. L. 1936 Sandridges with minor crescentic dunes on top (212).
(Simpson Desert, Australia). 1938 Do. (514).
1946 [Some] plinths with varieties of crest formations (49-51).
A. 940 Complex ig. 28) .
tern United St
, Robert 945 Dunes 0 (sif normal to ) (401).
rly "French" S [Some] (404; fig. 11)
th Africa). 947a Do.
y, Robert, and 948 Demkhas ) .
Capot-Rey Francoise
(Formerly "French" Sahara,
North Africa).

Holm, D. A. 1957 Linear belts with V-shaped chevrons; dune complexes; [some] Sigmoidal
(Saudi Arabian deserts). dunes (1746).
1960 Hooked dunes (1371-1372).
Linear 1373)
968 Comp I ex omp lexes (977)
D. 966 Reversi I, pI. II E).
Sand Dunes Na
ment, U.S.A.)
eodore 958 Grands (83) .
(Formerly "French" Sahara, Dunes en rateau, modele transverse sur 'elb; [some] famille des akles;
North Africa). systeme complex (141).
Smith, H. T. U. 1903 Complex longitudinal dunes (20-21; figs. 10, 14).
(Nebraska, U.S.A.: Ridged dune complex (22; fig. 18).
Sahara, North Africa). 1968 Do. (36).
R. W., and oth 1966 [Some] [DIOO].
Arabian deser
. P. CJ67 Alveola s wi th conical
U.S.S.R.) . Longi[u se and interfe forms (297; f
O. 967 Fish ho une (228).
serts) . Sigmoid

:"VU I.J] _ "-'

 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 291

TABLE 31. - Chronological list of local terms for dunes considered equivalent to those classified as complex, based on Landsat
imagery - Continued
Reference
Terms (reference page, figur~ plate, and table numbers in parentheses)
Author Year
(and region)
Mabbutt, J. A. 1968 [Some] ridges with multiple or braiding crests (143; Pl. VI).
(Australia). [Some] dune networks with intersecting trends (146).
Reticulate dunes (148).
Grove, A. T. 1969 [Some] alab dunes (195-201; pl. VIlla).
(Namib Desert, South-West
Africa).
Glennie, K. W. 1970 Seif dunes with barchan-like slipfaces (94-95; fig. 75).
(Saudi Arabian deserts). [Some] dunes of the Uruq al Mu'taridah (96-98; fig. 76).
Wilson, 1. G. 1971b [Some] draas (182, 191, 198).
(Algerian Ergs, North Africa). Lingoid networks with rhomboidal or hexagonal cells (182).
Giant draa ridges (pl. lIlA, F).
Cooke, R. U., and Warren, Andrew 1973 [Some I draa (231-232; fig. 4.27).
(All deserts). [Some] draa ridges (235, 296-297).
Oblique draa patterns (300).
Breed, C. S. , and McKee, E. D. 1973 Linear megadunes (fig. 7).
(Namib Desert).
McKee, E. D., and Breed, C. S. 1974a Linear chains of star dunes (9-3; figs. 9-1, 9-2).
(All deserts). Do. (668; fig. 5).
1974b. Chevron or basketweave dunes (66~; fig. 4).
McKee, E. D. , and others 1974 Star dunes in chains (19; figs. 4, 8; table 1).
(All deserts). Linear type (fig. 7).
Chevron or basketweave dune~ (81; fig. 3; table 1).
Reversing dunes (fig. 10).
McKee, E. D., and Breed, C. S. 1976 Chains of star dunes (82; fig. 54).
(All deserts).

TABLE 32. - Chronological Jist of local terms for various sand seas
Reference
Ter- (reference, paae. flaur., plate, and TerM (reference, pale, fllur., plate, and
Author Y.ar table n..-bel's In parenthea.a) Author Y.ar table n..-ben 1n parenth.a.s)
Cand re&ion) (and re.loD)
Hedin. Sven 1905. "5 " of .. nd (234). 1101'1'1., I. Mo, and IIont_, It. S. 1961 Dune (617).
(Tak" "-ken Desert. China). (Alaodonea Dune., ClUfol'nt.
Aufre!' Leon 1928 Pa ae dunaire, 1_ _ 1' d ble (8)4). and Mextco).
(FoI'Mrly "French" Sahara,
North Africa.
19lO
1934
Erl (22S).
Un ch_p d. dune. (132).
Salth, H. T. U.
(Nebr ka, U.S.A.:
1963
1968
Du. . fi.ld (36. 43;
Do. (29).
fI,. 8, IS).
Lo Mr d abl. (134. 138-139). 'Sahara, North Africa).
laanold A. 1933 Sand ... (110). Meltee, I. D. 1966 Dun.field (10).
(Libyan 4."l't. North Africa). Dun.field (08). (White Sanda, Navajo Indian
MacUaan, C. L. 1936 Mer d. . .bl. (224). . .a.nation, U.S.A Libyan
(St. .aon Deaert, Auetralla). Sandrld,. de rt (206, 208. 21Si fil' 2). d.a.n; Arabian deaern).
1946 Do. (S6). Powera, R. W., and othera 1966 Rlfud (D100).
Capot-ley, Robert 1945 Era (91). (Saudi Arabian Haerta).
('orMr!, "'rench" Sahara, Sea of nd. 1. _1' de ble (393). Stone, R. O. 1967 Erl (227).
IIonh Africa). 1947. Houle de ble (88). (All deaerta). Sand aea U4S)
19S3 Lo . .r d. ubi. (l4S).
Altli (US-146). Goudl A. S. 1969 Dune country (404).
Capot-"y, Iobert and 1948 Houl. d. ubi. (79). (Southern Africa).
Capot. ley, .rancola, Goudie, A. S., Allchin. aridaet. 1973 Er,. (24S; tab 1. 1).
(ForMrl, "'rench" Sahara. and Hedae. It. T. M. Dun. field. (247. 249).
North Africa). (Southern Africa; n.ar Deaert).

lola, D. A. 19S3 "fud 006-112; 11 2-8). Glennie, It. W. 1970 Duneft.ld (84).
(Saudi ArabSan d ern) . 1960 Mafuel (1369-1374). (Saudi Aublan derta).

"-'butt, J. A. 19S5 Sandv.ld (24. 26). WUaon, I. C. 1971. Era, aand-a.a (264-267).
(lalahari DeMrt. aouthern 1967 Sandrld.e de.err. (l4. (Alaerian Ira., North Africa). 1911b Erl. 180. 184. 190-198; ftl. 3),
Africa; St.,.on Du.rt Dun.fi.ld. (173). 1972 Do. (187-188).
AuatraUs). 1968 Do. (140). Sand-ridae desern (187).
1973 Erl (77-IOS).
Cooper, W. S. 19S8 Era (61). Dr.a eras, dune eras (91).
(W.at co t. North -'-erica). 1967 lody of dun (39).
Dun. co.plex (l.2, 44-47, S7-~8; fl,. 1; Cooke, R. U., and Warren, Andrew
(All des.rn).
197) Sand a.a (sO. 229).
Era (229.300.306,312,322-327).
pl. 8-10. 1 ) .
Dun.fi.ld 011-31S. 322).
Grove, A. T. 19S8 Erl (S26-S33).
(Southern Sahara. North Qo. (S31). Ireed. C. Sot and McKee, E. D. 1973 Sand a ere (1960).
Africa) . (All desena).
McKee, E. D., and others 1974 S.nd a (]-850).
IIoftod t Theodore 19S8 Erl (22. SS. 63), (All d.aert.).
(ror.rly "'rench" Sahara, McKee, E. D nd Ir.ed. C. S. 1976 De rt .and (81-82).
IIorth Africa). (All desert a).

Digitized by Google
292 A STUDY OF GLOBAL SAND SEAS

TABLE 33. - Chronological Jist of local terms for patterns of eolian sand deposits in sand seas
Referenct's
Terms (reft'renct' pagt', figure, plate, and table numbers in parentheses)
Author
(and region)
-------------.-----
Hedin, Sven 1905a Network of dunes, bajirs, and thresholds (360, 365; figs. 301-303,
(Takla Makan Desert, China). 317; pI. 52).
Auft'ere, Leon 1928 Les sable en colonne ou en monome (834).
(Formerly "French" Sahara, 1932 Zemloul (44-45).
North Africa). 1934 Dunes espacees et les dunes serrees (133).
Vagues entrecroisees ou reticulees (138-139).
Bagnold, R. A. 1933 Sand sea of parallel seif dune lines (110).
(Libyan desert, North Africa). Belt of barchans (106, 123-124; fig. 3).
1941 Belt of barchans (218-221).
Hultiple seif cbains (231-235).
Hadigan, C. L. 1936 Sand-ridge grid, grid of parallel ridges (211, 218; fig. 11).
(Simpson Desert, Australia).
Helton, F. A. 1940 Array of dunes (118-119).
(Hidwestern United States, Peak and fulje topography; intersecting transverse dune series
U.S.A.) (131-132; fig. 25, 26).
Trains of barchan dunes (figs. 2, 3).
Smith, H. T. U. 1963 Barchan field (30; fig. 2).
(Nebraska, U.S.A.; Trains of barchans (16).
Sahara, North Africa). Random pattern (43; fig. 15).
Parallel, subparallel, divergent, reticulate group patterns (17-18).
1965 Pronounced parallelism like waves of the sea (563).
Ridge and furrow topography (566).

Capot-Rey, Robert 1947a ChAinons dunaires (88).

(Formerly "French" Sahara, Les grandes massifs dunaires (94).
North Africa). Essaims [swarms] de barkhanes (97).
1953 Do. (144).
Capot-Rey, Robert, and 1948 Train [of barchans] (66).
Capot-Rey, Francoise. Hassif d'un aspect alveole (73).
(Formerly "French Sahara,
North Africa).
Cooper, W. S. 1958 Reticulate pattern with rhomboidal meshes (31).
(West coast, North America). Oblique ridge pattern (25, 49-60, 62; fig. 2; pl. 11).
Longitudinal-ridge pattern (62).
Transverse-ridge pattern (25, 27, 58, 64; fig. 1; pl. 7, 11).
1967 Dune belts (23; figs. 2, 3; pl. 8, 9).
Dune sheets (33-40, 58-60; figs. I, 2; pl. 8).
Norris, R. H., and Norris, K. S. 1961 Peak and hollow topography (608).
(Algodones Dunes, California Chains of dunes, chif10nes (617).
and Mexico). Barchan chains (618).
Norris, R. H. 1966 Barchan swarms (294-295).
(Algodones Dunes, California Dune cluster (292).
and Hexico).
Habbutt, J. A. 1967 Parallel ridges (58, pl. 18c).
(Kalahari Desert, southern 1968 Do. (141-143, 146; pI. 1).
Africa; Simpson Desert, Barchan trains (114)
Australia) Widely-spaced parallel ridges (143, 145; pl. 6).
Convergence and deflection of dune ridges (145, 148; pl. 7).
Alined short dunes (146).
Braided, forked, arcuate crests (146).
Arcuate belts of dunes (144).
Dune networks (145-146; pl. 8).
Branching patterns (147).
Honod, Theodore 1958 Regime alAb-sloOk, regime longitudinal typique (59).
(Formerly "French" Sahara, Regime alAb ayOun (20, 77, 80, 141).
North Africa. Syst~mes transverses et syst~mes longitudinaux (143).
Barchans gregaires (141).
Les akles (58, 92-94, 129, 141, 173; pl. 32).
Un akle reticule (143).
Erg en rateau ou en peine (145).

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 293

TABLE 33. - Chronological Jist of local terms for patterns of eoJian sand deposits in sand seas - Continued
---- - ------------------ ----
Referenct's
Terms (reft'rence pa~e, figure, plate, and table numbers in parentheses)
Author
(and region)

Grove. A. T. 1958 Ridges and hollows (526-531)

(Southern Sahara, North Africa, Dunefield (527).
Kalahari Desert, South
Africa).

Petrov, M. P. 1967 Crisscrossing barkban chains (300; table 36).

(U.S.S.R.)
Stone, R. O. 1967 Complex dunes, peak and depression (fulji) topography (220-221)
(All deserts).

Holm, D. A. 1968 Dune complex, dune terrains, dune belts (976).

(Saudi Arabian deserts). Waves of dunes (978).
Goudie, A. S. 1969 Dune network (404).
(Southern Africa).
Goudie, A. S., A1chin, Bridget, 1973 Dendritic, reticulate, and parallel ridges (93, 97, 100; fig. 2).
and Hedge, K. T. M. Clustered dendritic ridges (97).
(Thar Desert, India).
Glennie, K. W. 1970 Dune pattern (101).
(Saudi Arabian deserts).

Wilson, I. G. lQ7la Network of draa ridges (268).

(Algerian Ergs, Nortb Complex dune nets (270).
Africa). lQ7lb Tw~-trend dune networks (186-187; fig. 4).
Bedform patterns (182-183).
IQ72 00. 071-IQ2, 19S-204).
Lunate network, linguoid network with rhomboidal or hexagonal cells (182).
[Some] ~eif dune ridge varieties (194. fig. 4).
Gridir~n. fishsrale. and braided patterns (207-208).
1973 Bedform patterns (78. 97).
Barchan dunes in belts (84).
Gridiron, fishscale, and braided patterns (100: fig. 17).
Oblique left-handed, en echelon chain (102).

Cooke, R. U., and Warren, Andrew 1973 Akl~ (285, 298, 311: fig. 435).
(all deserts). Erg patterns (322-327).
Transverse patterns (2QS).
Rectangular rPticult's: l'omplex pattprns (2Q6).
Dune patterns (229, 283-287, 322-327; figs. 4.25, 4.26).
Draa patterns (297; fig. 4.34).
McKee, E. D., and others 1974 Parallel straight or wavy dune complexes (17, 24, 47, 53, 62, 68;
(All deserts). figs. I, 2, 17, 19, 25; table 1).
Dune patterns (68).
Parabolic dune arrays (fig. 5).
McKee, E. D., and Breed, C. S. 1976 Parallel straight or wavy dune complexps (81-82; figs. 55. 56).
(All deserts)

TABLE 34. - Chronological Jist of local terms for sand sheets and streaks (stringers)
[Streaks are considered equivalent to deposits termed "stringers" in preceding chapters of this pu cation]
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - --------------
Reference Terms (reference page, figure, plate. and tabl~ numb~rs in parentheses)
Author - - - - - - - - - - - - - - - - - - - - - - - - - ------ ------- -------------
Year Sand sheets Sand streaks
(and region)
Bagnold, R. A. 1933 Sand-sheets (105. 11l). Sand shadows and sand drifts (121;
(Libyan Desert, North Africa). fig. J).
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _1_9_4_l_ _5_an_d_s_h_ee_t_s_(_24_3_-_2_4_5}_._____________0_0. (I 89-19_5_}_.____
Madigan, C. L. 1938 Sandy plain (514 No term.
(Simpson Desert, Australia). 1946 Sand sheets (58). 00.
--------------------------------------------------------
Melton, F. A. 1940 No term. Umbracer or wind shadow dunes
(Midwestern, United States). (113, l20).

Digitized by Google
294 A STUDY OF GLOBAL SAND SEAS

TABLE 34. - ChlOnologicallist of local terms for sand sheets and streaks - Continued
[Streaks are considered equivalent to deposits termed "stringers" in preceding chapters of this publication)

Reference Terms (reference page, figure, plate, and table numbers in parentheses)
Author Year Sand sheets Sand streaks
(and region)
Hack, J. T. 1941 No term. Fallin~ dunes (241).
(Navajo Indian Reservation,
U.S.A.)
Smi th, H. T. U. 1946 No term. Wind-shadow dunes (197).
(Nebraska, U.S.A.: 1963 do. Do. (35; fig. 7).
Sahara, North Africa). Sand streamers (32, fig. 4).
Sand drifts (33, fig. 5).
Climbing dunes, falling dunes, sand
glaciers (14-15).

Capot-Rey, Robert and 1948 Nappes de sable, tenere, reg (63). No term.
Capot-Rey, Francoise
(Formerly "French" Sahara,
North Africa).
Holm, D. A. 1953 [Some) zibarr (I 10) No term.
(Saudi Arabian deserts). 19(,0 (Some) zibar (1372). Do.

Mabbutt, J. A. 1955 Flat sand spread (27). No term.

(Kalahari, Desert, southern 1967 Sand plain (149). Do.
Africa). 1'1(,8 do. ()39-l40, 145; pI. 1.5) Do.

Cooper, W. S. 1958 (Some) dune sheets (33-34, M). No term.

(West ~oast, North Ameri~a). 1967 do. (Fig. I: pI. 11) Do.
Grove, A. T. 1958 Sandy drift (526-529). No term.
(Southern Sahara, North Africa:
Kalahari Desert, southern
Afrira).
Norris, R. M. 1966 No term. Sand streamers (294-295, 305).
(Algodones Dunes, California
and Mexi~o).
Powers, R. W. and others 1966 No term. Giant elongate sound mounds (DIOO).
(Saudi Arabian deserts).
Stone, R. O. 1967 Sand sheet (245) Windrift dune, windshadow dune (253).
(All deserts). Falling dune (228).
Sand drift, sand glacier 245).
Source--bordering lee dune (249).
Umbrafon dune (252).
Glennie, K. W. 1'170 Shept sands (106-109; figs. 89-90). No term.
(Saudi Arabian deserts).
Goudip, A. S.
----------.------------------------------------------------------------------
1970 sand shepts No term.
Topographically-indu~ed
(Southern Africa). (95)
Goudie, A. S., All~hin, Bridget, 1973 do. (246). Lee drifts (249).
and Hedgp, K. T. M. Sand drifts, windward drifts (24(" 249). Wind drifts (244, 248-249).
(Thar Desert, India).
Wilson, 1. G. 1971a Sand plain (2(,4). No term.
(Algerian Ergs, North Afri~a). 1972 No tprm. Drifts downwind of scarps (191).
1973 Sand-sheet (103-104: fig. 21). No term.
Sand patches (78).
Ripple ergs ('II).
Coarse-sand ergs (104).
Cooke, R. U., and Warrpn, Andrew 1'173 No term. Lee dunes, trailing dunes, trailing
(All desprfs). sand patches (314-315).
-----------
Sagan, Carl, and others 1971 Streaks (179-215).
(On thp planpt, Mars).
McKpp, E. D., and Brped, C. S.
-----------.------------
1'l74h Sand sheers (,65, 6(,8). Stringers (665,668).
(All dpserts).
McKpp, E. D. and others 1'174 do. (3, 52,61,73,85: figs. 17, Do. (49,61; fig. 20).
(All deserts). 19: tshlp I). Streaks (3, 52, 55, 61, 85: figs. 6,
17:tablel).
McKep, E. D. and Brped, C. S. 1'l76 Sand ~hpets (81-82; fig. 59). Stringers, stringer dunes (81-82:
(All dpsprts). fig. 5'1).

Digitized by Google
 RPHOLOGY ON OF DUN AS OBSERVED SENSING

Chronological Ii s for interdune

Reference
Terms (refpr ure, plate, an rs in parenthe
Author Year
(and region)
Hedin, Sven 1905a Bajirs or depressions (234-251, 311-369).
(Tak1a Makan Desert, China).
Aufrere, Leon 1928 Feidjs [sandy I or gassis [bare I; bahirs (834).
"French" Sahara Do. (22
ica). Vallees eoli
Intervallees (226, 230).
Les s1110ns (220, 225).
Couloirs (13
Haouds (138)
Madigan, C. Gibber or clay ats
(Simpson Desert, Australia). Inter-ridge areas (223).
1938 Inter-ridge spaces; long, narrow clay pans or gibber plains, flats (514).
1946 Sandy lanes (43, 49).
Transverse hollows (52).
Helton, F. A. 1940 Fuljes or enclosed basins (131).
Hid United States

Fuljis or in s (2l4).
ert, North Afr
Narrow areas waves of sand
ian Reservatio
U.S.A.)
Capot-Rey, Robert 1945 Gassi: interdune trough without sand; feij, sand-covered interdune
(Formerly "French" Sahara, trough (395, 401; fig. 4.5).
North Africa). Sebkhas (392).
Interdune passages, couloirs, aftout (404).
1947a Do. (89).
Cuvettes, un oles (145).
bert and Do. (89-
Francois Interdune pa rs, aftout (73
French Sahara,
ica)

U. Fulji (198).
(Nebraska, U.S.A.; 1963 Inter-dune basins (40; fig. 12).
Sahara, North Africa). 1965 Interdune basins, troughs, areas (562).
1968 Narrow grooves, network of corridors, broad flats, depressions (32).
1969 Interdune depressions (17, fig. 14).
Holm, D. A. 1953 Interdune valleys (105).
(Saudi Arabian deserts). Deflation hollows (109).
Interdune ar
re Les interdun 0(2
French" Sahara
ica) .
Interdunes s
Ayn (pI., ay
rdunes rocheux ) .
(pI., t9ar) (8
Hollows (526
hara, North A c, Interdune depressions, interdunal corridors, strate (straat) (199).
Kalahari Desert, southern
Africa).
Norris, R. H., and NorriS, K. S. 1961 Large, flat-floored, sand-free depressions (605, 608-610; figs. 1, 2;
(Algodones Dunes, California pI. 2, 3).
and Mexico).
N rris R H 1966 Intradune ho ions (292-297,
Dunes, Californ
) .
Interdune su s (10, 51-56,
, Navajo India
n, U.S.A.;
ert; Irq as Su
Saudi Arabia).

:"VU I.J] _ "-'

296 A STUDY OF GLOBAL SAND SEAS

TABLE 35 - Chronological list of local terms for interdune areas - Continued

Reference
Terms (refprence page, figure, plate, and table numbers in parentheses)
Author Year
(and region)
Cooper, W. S. 1967 Linear series of hallows between dune waves (34).
(West coast, North America).
Petrov, M. P. 1967 Interbarkhan depressions (304).
(U.S.S.R.) Interridge depressions (307).
Stone, R. O. 1967 Dune valley (225).
(All deserts). Fulji, or Fulje, depression between barchan dunes (228).

Mabbutt, J. A. 1968 Fuljes, interdune corridors (143).

(Kalahari Desert, southern Enclosed or partly enclosed depressions in dune networks (146).
Africa; Simpson Desert,
Australia)
Glennie, K. W. 1970 Inland interdune sehkhas, interdune sebkhas, (60-61, 97-98; fig. 77).
(Saudi Arabian deserts). Interdune deflation plane (77).
Sand-covered inter-dune areas (78).
Interdune areas (90, 107-108).
Feidjes (101).
Goudie, A. S. 1970 Inter-dune corridors (95).
(Southern Africa). Inter-cluster corridors, straats (97).
Wilson, I. G. 1971a Troughs (264)
(Algerian Ergs, North Africa). Sand-free hollows (267).
Dune hollows (270).
1973 00. (92).

Cooke, R. U., and Warren, Andrew 1973 Interdune corridors (322).

(All deserts) Interdraa corridors (327).
Breed, C. S., and McKee, E. D. 1973 Intradune corridors (162).
(All deserts).
McKee, E. D., and others 1974 Interdune surfaces or areas (70-71).
(All deserts)
McKee, E. D., and Breed, C. S. 1976 Interdune areas (81).
(All deserts).

TABLE 36. - Morphometry of linear dunes in sand seas

[W, _an width, L,..an len.,th (*, . . . or all dune. extend beyond ..s.,. of illaC)eh I, _an .8velelMJth (cr t-to-crt distance),
P, -.n frequency (nu.ber of dun per kil~tr.) aero -...ple, noEWll to trend of dun...
All . . . .ur. .nt. 1n kl1~tr )

Are.
lle9ion' t..tltude, lon9itude Dert. or And . . . area ....pled ii ii/X i./X F Fiqure
(lao') No.

Siolple (1) linear dune.

3
4
5
23-2730' S., 131-14030' B.
1930'-2330' S., 12230'-12830' E.
2330'-2545' S., 1730'-1930. It.
2030'-2130' M., 52-5230' B.
Navajo Indian
Arizona, . . . .).rvation
U.S.A. (northern
______________________

SiJlpaon De.ert (Australia) ----------------

Grot Sandy Deoert (Australia) ------------
Kalahari Desert (southern Africa) ---------
Northaastern Rub' al Khal1 (Baudi Arabia)-
_
625
30,000
35,000
15,000
2.500
0.043
.22
.29
.29
.38
3.65
24.0
22.7
26.0
14.34
0.15
.90
.90
.70
1.41
84.9
109.1
78.3
89.7
37.7
0.29
.24
32
.41
.27
24.3
26.67
25.22
37.14
10.17
4.92
1.44
1.06
1.26
.64
-.
168.

169
169A
241

COIIpound 1 inear dun

6 21-22. N., 08-10. W. Sout_stem sahara (Nauritania) ---------- 2,500 0.94 >100.00 1.93 10'.4 0.49 51.81 0.48 169D
7 lS-19- N 13-1430' B. Southam sahara (NiCJ1lr) ------------------- 7,500 1.06 >40. 1.90 37.7 .56 21.0S- .29 _.
8 1630'-1915' N., 45-30'-50 8. Southwestern Rub' al KhaU (Baudi Arabia)- 20,000 1.21 >83. 2.18 68.6 .55 38.n,. .48 169('

C~lex linear dun

9
10
11
12
24-2445' 5., 1445'-1530' B.
2630'-28- N., 00.30'-0230' E.
19-1930' N., 1S-1S030' B.
1845'-1930' N 4515'-46 B.
Naaib Desert (South-Ne.t Africa) ----------
Northern Bahara (AlCJ1lril\)-----------------
Southern sahara (Niqer) -------------------
Ne.tern Rub' al Khali (Baudi Arabia) ------
2,500
2,500
2,500
2,500
O.SS
1.09
1.28
1.~8
27
30.75
65
73
2.20
3.2~
3.28
3.17
30.7
20.2
50.8
49.3
0.40
.34
.39
.~7
12.27
9.A9
19.81
23.03
0.47
.33
.30
.21 -.
1691f
1688

1691'

'Sources of data--(l) Aerial photoqraphs, National Aeronautics and Space Adainiatration. 236-28-0071, 236-28-0072.
Fo11owinq are n_red Landsat i_qes. (2) El315-00140, El227-00264, 111224-00100, B1369-00135, 1369-00141, 111133-00040, 111296-00090, 1367-0025,
111224-00091, (3) 111344-01160, B1127-01111, 111377-00585, 111127-01113, B1415-01091, E11GO-Ol045, 112~-on5J5, 1100-01051, E1125-00591, E1124-00542,
81344-01165, 81413-00584, 111344-01162, (4) 1147-08160, 111416-08085, 111416-08091, 1145-08052, (5) E1184-06255, (6) E1119-10324, (7) &1192-08553,
El193-09012, (8) E1107-06433, E1132-06380, 1187-06435, 1170-06493, (9) E138J-0~270, (10) Blll5-100H, (11) 1192-08553, (12) B1189-06550.

Digitized by Google
 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 297

TABLE 37. - Similarity of linear dune varieties in widely distant sand seas shown by comparison of ratios derived from
measurements of mean lengths.L, mean width W, and mean wavelength,X
(Subscript following name of sand sea indicates first or second of the pair. Subscripts to L, W, and ~ indicate first or second of sand sea pair]

.' L L Degree of
corresponden

5 24.0 0.90
6 12 6. .9 26.67 24 29 Fair
3.65 . 15

Great Sandy 1------------- 22.7 .29 .90

6.22 6.74 6.0 78.3 84.9 25.22 24.3 .32 .29 Good
Northern Arizona 2 -------- 3.65 .043 . 15

Kalahari 1 ---------------- 26.0 .29 .70

7.12 6.74 4.67 89.7 84.9 37.14 24.3 .41 .29 Very poor.
Northern Arizona 2 -------- 3.65 .043 .15

Northern Rub' al Khali l -- 14.34 .38 1.41

R4 9. .9 10.17 24 29 Very poor.
3.65 .15

26.0 .70
o .7 .3 37.14 25 32 Excellent.
G 22.7 .90

26.0 .70
32 .7 .1 37.14 26 24 Good
Slm.pson 2 ----------------- 24.0 22 .90

Simpson l ----------------- 24.0 .22 .90

1.06 .76 1.0 109.1 78.3 26.67 25.22 .24 .32 Excellent.
Great SandY2------------- 22.7 .29 .90

Simpson l ----------------- 24.0 .22 .90

1.67 .58 .64 109.1 37.7 26.67 10.17 .24 .27 Fair.
Northern Rub' al Xhali 2 -- 14.34 .38 1.41

G 22.7 .90
76 .6 .7 25.22 10 27 Good.
ali 2 -- 14.34 1.41

26.0 .70
76 .5 .7 37 .14 10 27 Fair.
1.41

lGe09raphic lOCAtion of Band seas: Simpson and Great Sandy Deserts, Australia; northern Arizona, U.S.A., Kalahari Desert, southern Africa;
Rub' a1 Khal!, Saudi Arabia.
2Standard for degree of correspondence, based on maximum difference in scale ratios: 0.01-0.50" Excellent; 0.51-1.0 -
Gooc!; 1.01-1.50. Fair; 1.51-2.0"" Poor; 2.10 or more - Very poor.

TABLE 38. - Morphometry of crescentic dunes in sand seas

, _an width; ).. _ 1 _asur_nts in k s (---), not deter. ype includes trans
as used by other & licaUonl

Are.. Range of Ranq

Re Deser "_led dune e ii dun Fig
(kII' ) len~ h wave I No

e and probable ala.p r1dqes

32"45' N., lIbite Sancia (U.S.A.)-------- 7.5 0.03-0.185 0.072 0.06-0.215 0.1l12 0.06-0.198 0.1l15 0.65 1.0 0.65 1738
106"15' II.
22-25 N., Persian Gulf (United Arab
5130'-5230' E. a1ratee) ----------------- .44 0.2 -1.8 .71 0.07-2.0 .59 .62 I. 2 .75 None.
26"-26"45' N., Thar Denrt (near u.&rkot,
69-70 E. Pakistan) ----------------- 5,000 0.2 -0.6 .47 0.5 -1.5 .93 0.2 -1.0 .58 .51 1. 6 .81 255
2330'-2430' S NaIlib c:oaat (South-Nest
1430'-1445' E., Africa) ------------------- 270 0.3 -1.3 .68 0.7-1.5 1.12 0.2 -1.9 .87 .61 1.29 .78 223
A N., AI Jiwa.
, E. (Saudi 4,500 0.2 -0. .3 .80 0.2 - 1.27 .94
Eastern 2,000 0.5-1. .5 .94 0.6 - I. 06 .90
E.
western 2,500 0.6-1. .0 1.66 0.7 - I. 51 .76 No
" B.

e cOllpOund crescent nd ridqes)

8 2545' N., llestern Sahara (Mali) ------- 2,500 0.5 -0.8 0.62 0.5 -1.2 0.77 0.80 175A
0130'-02. N.
9 2030' N., South.astern Rub' al lOlal1
55"15'-56" E. (Saudi Arabia) ------------ 2,500 0.3 -1.1 .67 0.85-2.2 1.43 0.5 -3.2 1. 76 .47 0.81 0.38 None.
10 3220'-30. N., Northeastern Sahara
0830'-0930' E. (Algeria and Tunisia) ----- 5,000 0.3 -1. 5 .65 0.5 -3.0 1.43 0.5 -2.0 1.24 .45 1.15 .52 210A
II 2730'-29 N., An NatUd (northern
39-40 E. Saudi Arabia)------------- 5,000 0.5 -2.0 .80 0.6 -2.7 1.41 0.8 -3.3 1.84 .57 .77 .43 None.
298 A STUDY OF GLOBAL SAND SEAS

TABLE 38 - Morphometry of crescentic dunes in sand seas - Continued

[I:, _an 18n1" ii, ..... width, ~, _an _".length. All _ _ t. in kU.-tre.. ~ra (---), not date.-1nad. ere......tic type include. tr.......r . .
type a . . - by other author. in thi. publication)

JIan98 of JIan98 of
lla9ion' Latitude, o...rt aancJ . . . area dune dune ii
lonVitude length width

CCJIIIPQURd creacentlc dune rid,..

3145'-32N. , Gran o..ierto (Sonora,

11330'-114. W. Mexico) ------------------- 650 0.3 -1.5 0.66 0.5 -2.0 1.04 0.5 -2.5 1.38 0.63 0.75 0.48 179D
13 32.5'-3310' N., Allodone. (california,
11445'-115.15' w. U.S.A.) ------------------- 500 0.5 -2.25 .88 0.5 -3.5 1.61 0.4 -2.5 1.07 .55 1.50 .82 17U
14 2123' N., Ai 'urUq a1 Nut t.rl~.
53l0'-SS E. tern Rub' a1 1Cha1i
(Saudi Arabia) ------------ 10,000 1.3 -3.3 2.09 1.6 -4.0 2.76 1.5 -4.0 2.56 .76 1.08 .82 174C,17511
15 3930'-40.30' N., Cherchen sand. , Takla
8645'-8730' E. llal<an (China)------------- 5,000 1.1 -3.4 2.20 2.0 -5.2 3.24 2.0 -5.5 3.0 .68 1.08 .73 175D
16 37-31.5' N., PU Karak"'"Y (near Mary,
6230'-6415' 8. U.5.5.R.)--------------- 2,500 0.8 -1.5 1.16 1.2 -3.5 1.96 1.0 -2.5 1.76 .59 1.11 .66 lIone.
17 4230'-4330' H., "ak1 ltArak"'"Y (near
6230' -65 E. Bukantau, U.S.S.R.)------- 10,000 0.8 -2.0 1.31 1.0 -3.1 2.22 0.85-3.1 2.27 .59 .98 .58 Do.
18 2,30'-2730' II., ....t.rn Thar o. ..rt (n r
6830'-6930' E. Sukkur, Pakt.tan)--------- 2,500 0.75-2.0 1.30 1.0 -2.5 1.50 0.7 -2.5 1.44 .81 1.04 .90 256
19 4130'-4230' N., Nebraaka Sand H111a (U.S.A.) 3,000 0.8 -2.3 1.65 1.5 -5.0 3.14 1.0 -2.5 1. 73 .52 1.81 .95 175lr
101-102 W.
20 40-40.30' N., Pa-tan-chi-lln Sha-.o, Ala
10130'-10230' E. Shan De.ert (China) ------- 2,500 1.25-3.0 2.16 2.0 -4.5 2.88 1.8 -4.25 2.92 .75 .99 .74 175C
21 1730'-19- N., Aoukir .and a.a, 8outh-
09-13 W. western Sahara (Mauritania) 5,000 1.2 -2.1 1.59 2.2 -9.6 4.10 1.0 -2.5 1.71 .39 2.40 .93 175,.

sources of data--(1) Aerial photD9raph. Ro11_ Air Force san. Followinv

are all n_red Land. .t ~.. (2) 81131-06303, El131-06310, (3) 1191-05211,
(4) 1383-08270, (5) E1148-06251, (6) 1146-04251, (7) 1115-04~4h (8) 1115-10083,
(9) 1182-0614h (10) 1109-09313, (11) 111231-07270, E1231-07272, (14) 81183-06194,
(15) 81146-04251, E1146-04254, (16) 111126-05581, (17) E1126-o5563, E1127-o621;
(18) 1192-05267, 111120-05267, (19) E1026-17012, (20) El117-03222; (21) E1227-10335,
E1229-10445. SItYUIII photovraph.. (12) S!A-76-085, (13) 5L4-76-083.

TABLE 39. - Similarity of compound crescentic dunes in widely distant sand seas. shown by comparison of ratios derived from
measurements of mean length.I, mean width. W. and mean wavelength.A:
[Subscript following n...., of sand sea indicates first or .econd of the pair. Subscripts to I:, if, and X indicate first or .econd of sand sea pair]

Deqree of
Pair of sand seas' I: if correspondence 2

Ala Shan 1 ---------------- 2.16 2.BB 2.92

1.03 1.04 1.14 0.75 0.76 0.99 1.08 0.74 0.82 Excellent.
Eastern Rub' a1 Khali 2 --- 2.09 2.76 2.56

TaU. Makan 1 ------------- 2.20 3.24 3.0

1.05 1.18 1.17 .68 .76 1.08 1.08 73 .82 Excellent
Eastern Rub' a1 Khali 2 --- 2.09 2.74 2.56

Takla Makan 1 ------------- 2.20 3.24 3.0

1.02 1.13 1.03 .68 .75 1.08 99 .73 .74 Excellent
Ala Shan 2 ---------------- 2.16 2.B8 2.92

Bukantau 1 ---------------- 1.31 2.22 2.27

1.13 1.13 1.29 .59 .59 .98 1.11 .58 .66 Excellent.
Mary2-------------------- 1.16 1.96 1. 76

Ala Shan 1 ---------------- 2.16 2.88 2.92

1.86 1.47 1.66 .75 .59 .99 1.11 74 .66 Very good
Mary 2 -------------------- 1.16 1.96 1. 76

Takla Makan 1 ------------- 2.20 3.24 3.0

1.90 1.65 1. 70 .68 59 1.08 1.11 .73 .66 Very good
Mary2-------------------- 1.16 1.96 1. 76

Ala Shan 1 ---------------- 2.16 2.88 2.92

1.66 1.92 2.03 .75 .87 .99 1.04 .74 90 Very good
Thar 2 -------------------- 1.30 1.50 1.44

Takla Makan 1 ------------- 2.20 3.24 3.0

1.68 1.46 1. 32 .68 .59 LOB .98 .73 58 Very good
Bukantau 2 ---------------- 1.ll 2.22 2.27

Thar 1-------------------- 1.30 1.50 1.44

1.97 1.44 1.04 .87 .63 1.04 .75 .94 .48 Poor.
Gran Desierto 2 ----------- .66 1.04 1.38

Takla Makan 1 ------------- 2.20 3.24 3.0

3.33 3.12 2.17 .68 .63 1.08 .75 .73 .48 Very poor.
Gran Desierto 2 ----------- .66 1.04 1.38

Digitized by Google
 MORPHOLOGY AND DISTRmUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 299

TABLE 39 - Similarity of compound crescentic dunes in widely distant sand seas, shown by comparison of ratios
derived from measurements of mean length, t, mean width, W, and mean wavelength,'i' - Continued

(Subscript following name of sand sea indicates first or second of the pair. Subscripts to L, ii, and X indicate first or second of sand sea pair)

Degree of
Pair of sand 8e... 1 L ii correspondence 2

Algodonea 1 -------------- .88 1.61 1.07

1.33 1.55 .78 .55 .63 1.50 .75 .82 .48 Pair.
Gran oesierto2 ----------- .66 1.04 1.38

TaUa Makan1 ------------- 2.20 3.24 3.0

2.5 2.01 2.8 .68 55 1.08 1.50 .73 .82 Pair
Algodones 2 --------------- .88 1.61 1.07

Eastern Rub' al Ithali 1--- 2. 09 2.76 2.56

2.37 1.71 2.39 .76 .55 1.08 1.50 .82 .82 Pair
Algodonea 2 -------------- .88 1.61 1.07

Nebraska Sand Hilla 1 ----- 1.65 3.14 1. 73

1.88 2.71 1.62 .52 55 1.81 1.50 .95 .82 Very poor
Algodones 2 --------------- .88 1.16 1.07

Mary1-------------------- 1.16 1.96 1. 76

1.32 1.22 1.64 .59 .55 1.11 1.50 .66 82 Very good
Algodonea 2 --------------- .88 1.61 1.07

Ala Slw>l---------------- 2.16 2.88 2.92

1.31 .92 1.69 .75 .52 .99 1.81 .74 95 Pair
Nebraska Sand Hil1s2 ----- 1. 65 3.14 1.73

Ala Slw>l---------------- 2.16 2.88 2.92

2.46 1. 79 2.73 .75 .55 .99 1.50 .74 .82 Poor
Algodones 2 --------------- 88 1.61 1.07

Thar1-------------------- 1.30 1.50 1.44

1.48 .93 1.35 .87 .55 1.04 1.50 .90 .82 Good.
AlgodoneB 2 --------------- .88 1.61 1.07

Eastern Rub' al IChali 1 --- 2.09 2.76 2.56

1.27 .88 1.48 .76 .52 1.08 1.81 .82 .95 Pair
Nebraska Sand HillB2 ----- 1.65 3.14 1.73

Eastarn Rub' al IChali 1 --- 2.09 2.76 2.56

1.31 .67 1.50 .76 .39 1.08 2.40 .82 .93 Poor.
Aouk.i.r 2 ------------------ 1.59 4.10 1.71

Aouk.&r1------------------ 1.59 4.10 1.71

.96 1.31 .99 .39 .52 2.40 1.81 .93 .95 Very good
Nebr. .ka Sand Hil18 2 ----- 1.65 3.14 1.73

lGeographic location of sand seas: Ala Shan Desert, China, Rub' al Khali, Saudi Arabia: Tak!a MAkan Desert, China; Peski Karuumy, near
Bultantau and Mary, UaS.S.R.; Gran Desierto, Mexico; Thar Desert, India and Pakistan: Algodones Desert and Nebraska Sand Hills, U.S.A. r
Aoukir, Maruitani...
2Standard for degree of correspondence, based on maximum difference in scale ratios: 0.01-0.20. Excellent; 0.21-0.39 111: Very good;
0.40-0.59. Good; 0.60-0.79. Fair; 0.80-0.99. Poor, 1.00 or more = Very poor.

TABLE 40. - Morphometry of star dunes in sand seas

Range of Range of
Latitude, longitude De.ert or sand aea ar dune ii dune Fiqure
diaMter wavelength No.

31"5'-3215' N6, 114-11430' W. Gran o.sierto (Mexico) ---------------------- 1,000 0.25-1. 2 0.70 0.1 -3.0 1.19 0.59 IBOA
2 2550' S., 1540' . Naa1b Deaert (SOuth-weat Africa) ------------ 1,000 0.4 -1.0 .68 O.B -4.5 2.21 31 None
3 1930'-20. N., 16-17. . SOutl)ern sahara (Niger) --------------------- 1,000 0.2 -1.2 .61 0.15-3.0 1.0 .61 Do.
4 3950'-40. II., 9930'-10130' 8. Ala Shan Desert (China) --------------------- 2,000 0.4 -1.0 .74 0.3 -3.2 1.37 .54 2648

5 19-2030' N., 5330'-5430' B. Southe tern Rub' al lChali (Saudi Arabia) --- 12,500 0.5 -1.3 .B4 0.97-2.86 2.06 .41 18QC
6 2930'-32- N., 00,30'-009 E. Entire Grand Erg oriental (Algeria) --------- 10,000 0.4 -3.0 .95 0.8 -6.7 2.07 .46 2108
7 2930'-3015. N., 00,30'-00815' . SOuthern aaZ'gin, Grand Erg Oriental (Algeria) 2,500 0.7 -3.0 1.65 1.5 -3.1 2.17 .76 1808
B 3030'-32 N., 00730'-009. E. Northern Grand Erg Oriental (Alger!a) -------- 7,500 0.4 -1.3 .71 0.8 -6.7 2.04 .35 2108

'SOurces of data are numbered Landsat ilIage., (1) 1159-17451, (2) 13B2-0B214, (3) 1209-08494, (4) 1209-03340, (5) El417-06191, 1147-06202,
1183-06200, 1183-06203, (6) Ell09-09320, 1109-09322, 1110-09374, (7) J!1109-09322, (B) Ell09-09~20, Ell10-09374.

Digitized by Google
 A STUDY AND SEAS

1. - Similarity widely distant wn bycompariso ved from meas

mean diame n wavelength, i
[Subscr 2 name of s icates first of the pair.
Subscripts to D and A indicate first or second of sand sea pair]

Degree of
Pairs of sand seas l D
correspondence 2

sierto ---- 0.70 19

I
0.54 .31 Fair
.68 21

n 1 --------- .74 37
1.06 1.15 .54 .59 Excellent.
Gran Desiert0 2 ------------------ .70 1.19

Eastern Rub' a1 Kha1i 1 ---------- .84 2.06

1.2 1. 73 .41 . 59 Fair
Gran Desiert0 2 ------------------ .70 1.19

rg Oriental .95 07
1. 74 .59 Good
siert0 2 ---- .70 19

Southern Grand Erg or1enta1 1 ---- 1.65 .17

2.36 1.82 .76 .59 Fair.
Gran Desiert0 2 ------------------ .70 1.19

Northern Grand Erg Orienta1 1 ---- .71 2.04

1.01 1.71 .35 .59 Poor .
. 70 19

sierto1 ---- .70 19

1.19 .61 Exce
n Sahara --
2
.61 o
Eastern Rub' a1 Kha1i 1 ---------- .84 2.06
1.14 1.50 .41 .54 Good.
Ala Shan 2----------------------- .74 1.37

Grand Erg Orienta1 l ------------- .95 2.07

1 28 1.51 46 .54 Very ood
n 2 --------- .74 37

rg Oriental .95 07
1.0 .41 Exce
Rub' a1 Kh .84 06

Ala Shan 1 ----------------------- .74 1.37

1.21 1. 37 .54 .61 Very good.
Southern Sahara 2 ---------------- .61 1.0

Eastern Rub' a1 Kha1i ---------- .84 2.06

1.01 .35 Very
n Grand Erg .71 04

n Grand Erg 1.65 17

1.05 .41 Very
Rub' a1 Kh .84 .06
2

:"VU I.J] _ "-'

 MORPHOLOGY AND DISTRIBUTION OF DUNES IN SAND SEAS OBSERVED BY REMOTE SENSING 301

TABLE 41. - Similarity of star dunes in widely distant sand seas, shown by comparison of ratios derived from measurements of
mean diameter, D, and mean wavelengtli, 'i - Continued

[Subscript following name of sand sea indicates first or second of the pair.
Subscripts to 5 and A indicate first or second of sand sea pair]

Degree of
Pairs of sand seas l correspondence 2

Southern Grand Erg Orienta1 1 ---- 1.65 2.17

2.23 1.58 .76 .54 Poor.
Alan Shan 2 ---------------------- .74 1.37

Southern Grand Erg Orienta1 1 ---- 1.65 2.17

2.43 .98 .76 .31 Very poor.
Namib 2 -------------------------- .68 2.21

Eastern Rub' a1 Kha1i 1 ---------- .84 2.06

1.24 .93 .41 31 Good
Namib 2 -------------------------- .68 2.21

Grand Erg Oriental 1 ------------- .95 2.07

1.4 .94 .46 .31 Fair.
Namib 2 -------------------------- .68 2.21

Grand Erg Orienta1 1 ------------- .95 2.07

1.56 2.07 .46 .61 Fair.
Southern Sahara 2 ---------------- .61 1.0

Eastern Rub' a1 Kha1i 1 ---------- .84 2.06

1.38 2.06 .41 .61 Poor .
Southern Sahara 2 ---------------- . 61 1.0

Namib 1 -------------------------- .68 2.21

1.11 2.21 .31 .61 Very poor.
Southern Sahara 2 ---------------- .61 1.0

Northern Grand Erg orienta1 1 ---- .71 2.04

1.16 2.04 .35 .61 Very poor.
Southern Sahara 2 ---------------- .61 1.0

lGeographic location of sand seas: Gran Desierto, Mexico; Namib Desert, South-West Africa;
Ala Shan Desert, China; Rub' a1 Kha1i, Saudi Arabia; Grand Erg Oriental, Algeria; Southern
Sahara, Niger.
2Standard for degree of correspondence, based on maximum difference in scale ratios:
0.01-0.15 = Excellent; 0.16-0.30 - Very good; 0.31-0.45 = Good; 0.46-0.60 = Fair; 0.61-0.75
Poor; 0.76 or more = Very poor. More stringent criteria are set than for dunes of linear and
crescentic types because fewer ratios are compared for dunes of star type.

Digitized by Google
302 A STUDY OF GLOBAL SAND SEAS

TABLE 42. - Morphometry of dome and dome-shaped dunes

[0, mean di_terr ~t _an wavelength, *, approxt-te. All .....urements in kilOlMttrea)

Are.. RAnge of Range o!

Fiqure
Region I wtitude, lonqtitude De.ert or Band a... area oaapled dune ii dUne X ii/X '10.
(laP') diameter wavelength

40-4030' N., 83-8345' E. Taltla Haltan (China) ----------------------------- 2,500 0.6-1.B 1.12 0.B-5.4 3.34 0.34 lB5S
32 3230' N., 002-00230' E. Nor~hern Grand Erg Occidental (Algeria) --------- 1,000 0.9-2.0 1.51 1.6-5.0 3.42 .44 IB5A
2520'-27. N., 44-45- E. Nafuda (northern saudi Arabia) ------------------ 3,000 O.B-1.BO 1.21 1.0-4.4 2.1B .56 244

'SOurces of data are n_red Landsat images. (1) E1131-04424, (2) E11B6-10003, (3) El137-07043.

TABLE 43. - Similarity of domes and dome-shaped dunes in widely distant sand seas, shown by comparison of ratios derived
from mean diameter, D, and mean wavelength'>~
[Subscript followin2 name of sand sea indicates first or second of the pair.
Subscripts to D and X indicate first or second of sand sea pair]

Degree of
Pairs of sand seas 1 D
correspondence 2

Nafuds 1 ------------------------- 1.21 2.1B

LOB 0.65 0.56 0.33 Good.
Northern Takla Makan 2 ----------- 1.12 3.34

Northern Grand Erg Occidental l -- 1.51 3.42

1.35 1.02 .44 .33 Good.
Northern Takla Makan 2 ----------- 1.12 3.34

Northern Grand Erg Occidental l -- 1.51 3.42

1.25 1.57 .44 .55 Good.
Nafuds 2 ------------------------- 1.21 2.1B

lGeographic location of sand seas: Nafuds, Saudi Arabia: Takla Makan Desert, China:
Grand Erg Occidental, Algera.
2Standard for degree of correspondence, based on maximum difference in scale ratios:
0.01-0.15 = Excellent: 0.16-0.30 = Very good: 0.31-0.45 = Good: 0.46-0.60 = Fair: 0.61-
0.75 = Poor: 0.76 or more = Very poor. More stringent criteria are set than for dunes of
linear and crescentic types because fewer ratios are compared for dunes of dome type.

Digitized by Google
Digitized by Google
Digitized by Google
 .. :.. ,::.. ... :.

"
"
....
', .
. ..
'
' .. '

.', ':," .'

A STUDY OF GLOBAL SAND SEAS

REGIONAL STUDIES OF SAND SEAS,

USING LANDSAT (ERTS) IMAGERY
Chapter K
By CAROL S. BREED, STEVEN G. FRYBERGER, SARAH ANDREWS, CAMILLA McCAULEY
FRANCES LENNARTZ, DANA GEBEL, and KEVIN HORSTMAN

Contents
Pase
Method of study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Summary of conclusions, review of previous work, interpretation of Landsat imagery, surface
wind flow and precipitation, and potential sand drift for sand seas in eight desert
regions:
Western Sahara of Mauritania, Senegal, Mali, and Spanish Sahara .................... 310
Sahel and southern Sahara of Mauritania, Mali, and Niger. . . . . . . . . . . . . . . . . . . . . . . . . . . 317
Northern Sahara of Algeria, western Libya, and Tunisia ............................. 326
Namib Desert of South-West Africa .......... ;................................... 346
Kalahari Desert of South-West Africa, South Africa, and Botswana .. . . . . . . . . . . . . . . . . . 355
Arabian sand seas of Saudi Arabia, Oman, Yemen, and adjacent States . . . . . . . . . . . . . . . 362
Thar Desert of India and Pakistan. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Talda Makan and Ala Shan Deserts of the Peoples Republic of China. . . . . . . . . . . . . . . . . 381

305

Digitized by Google
306 A STUDY OF GLOBAL SAND SEAS

Illustrations
p.
FIGURE 188. Distribution and morphology of eolian sand, western Sahara.............. 311
189. Diagram of sand rose used in this chapter. . . . . . .. . . . . . . . . . . . . . . . . . . . . . .. 312
190. Skylab photograph of elongate sand seas, northwest Sahara and Spanish Sahara 312
191. Landsat false-color imagery showing vegetated, large linear dunes in relation to
precipitation, southern Mauritania. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 314
192. Landsat imagery showing compound linear dunes, Makteir sand sea,
Mauritania . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 315
193. Landsat false-color imagery of a solitary linear dune between Atar and Akjoujt,
Mauritania ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 316
194. Landsat imagery showing compound crescentic dunes, Aoukir sand sea,
Mauritania . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 317
195. Map of sea-level mean-pressure contours, Sahara and adjacent regions ..... 318
1%. Graph of variation of rainfall during year, Rosso, Akjoujt, and Fort-Gouraud,
Mauritania ....................................................... 319
197. Map of resultant drift directions, drift potentials, and resultant drift potentials at
weather stations in Mauritania ..................................... 320
198. Sand roses for 6 months for Nouakchott, Mauritania. . . . . . . . . . . . . . . . . .. . .. 321
199. Sand roses for 12 months for Boutilimit, Mauritania ...................... 322
200. Graph of drift potential during the year at Boutilimit, Mauritania ........... 323
201. Map showing distribution and morphology of eolian sand, Sahel and southern
Sahara ........................................................... 324
202. Landsat imagery showing compound linear dunes of the Azaouid sand sea
near Timbuktu ................................................... 326
203. Monthly sand roses for Nioro du Sahel, Mali ............................. 328
204. Graph of variability of rainfall during the year at Gao and Nioro du Sahel, Mali 328
205. Graph of drift potential during the year at representative stations in the Sahel 329
206. Map showing distribution and morphology of eolian sand in Algeria and
southern Tunisia ................................................. 330
207. Map showing distribution and morphology of eolian sand in northwestern
Libya and southeastern Algeria ..................................... 332
208. Landsat imagery showing the southwestern margin of the Grand Erg Occiden-
tal, Algeria ....................................................... 334
209. Landsat imagery showing subparallel complex dunes of linear and star types of
the 'Erg Chech, Algeria ............................................ 336
210. Landsat false-color imagery showing northeastern part of the Grand Erg Orien-
tal, Algeria and Tunisia ............................................ 337
211. Skylab photograph showing star dunes in linear chains, southwestern end of
the Grand Erg Oriental, Algeria .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 339
212. Landsat imagery of the central and western parts of the ~~ri' Awbarl, Libya. 340
213. Skylab photograph of dunes at the west edge of the ~~ra' Marzuq, Libya ... 342
214. Weather maps of two meteorological events which produce southwesterly and
northwesterly winds over the northern Sahara ....................... 343
215. Sand roses for 6 months for Laghouat, Algeria ........................... 344
216. Sand roses for 4 months for Ghardaia, Algeria ............................ 344
217. Sand roses for 6 months for Ouargla, Algeria ............................ 345
218. Sand roses for 6 months for Qabis, Tunisia .............................. 345
219. Sand roses for 6 months for I-n-Salah, Algeria ............................ 346
220. Graph showing yearly drift potential, Laghouat and Ouargla, Algeria . . . . . . .. 346
221. Graph showing yearly drift potential, Ghardaia and I-n-Salah, Algeria ....... 347
222. Map showing distribution and morphology of eolian sand, central Namib
Desert, southern Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 348
223. Landsat imagery of western parts of the main sand sea, central Namib Desert 350
224. Landsat false-color imagery of northeastern parts of main sand sea, central
Namib Desert .................................................... 351
225. Photograph of red sand dunes and tan playa sediments along the interior
margin of the Namib Desert at the Sossus Vlei ....................... 353

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 307

FIGURE 226. Maps of mean sea level isobars and isobars of the 850-millibar-pressure surface,
p.
southem Africa .................................................. . 354
227. Wind roses from central Namib Desert at luderitz, South-West Africa ..... . 354
228. Graph showing drift potential from the northeast quadrant during June and
July, north-central Namib Desert .................................. . 354
229. Sand roses for 11 months for Gobabeb, South-West Africa ............... . 355
230. Sand roses for 8 months for Rooi Bank, South-West Africa ............... . 356
231. Wind rose for Aus, South-West Africa .................................. . 356
232. Graph showing drift potential at Pelican Point, Gobabeb, and Narabeb, South-
West Africa ...................................................... . 357
233. Map showing distribution and morphology of eolian sand, western Kalahari
Desert, southern Africa ........................................... . 358
234. landsat false-color imagery of large linear sand features, invasive sand sheets,
and abandoned drainages of the Molopo River and Kurumanrivier,
southern Kalahari Desert ......................................... . 359
235. Sand roses for 12 months for Ghanzi, Botswana, northeast Kalahari Desert .. 360
236. Graph showing average monthly rainfall versus average monthly drift potential,
Upington, South-West Africa ...................................... . 361
237. Graph showing drift potential at four stations in the Kalahari Desert ....... . 361
238. Graph showing distribution and morphology of eolian sand in Saudi Arabia,
Oman, Yemen, and adjacent States of the southern Arabian Peninsula .. 384
239. Skylab oblique photograph showing red linear dunes, western Rub' al Kahli . 366
240. landsat imagery of crescentic dune ridges, northeastern Rub' al Khali ...... . 387
241. landsat imagery of the north-central part of the Rub' al Khali ............. . 367
242. landsat imagery showing the southern part of the Wahiba Sands, Oman ... . 368
243. Outline map, with sand roses, of the northern Arabian Peninsula ......... . 369
244. landsat false-color imagery of Ad Dahna, Saudi Arabia ................... . 370
245. Map of sea-level mean-pressure contours on the Arabian Peninsula and adja-
cent regions .................................................... . 372
246. Sand roses for 6 months for Badanah, Saudi Arabia ...................... . 373
247. Sand roses for 12 months for Dhahran, Saudi Arabia ..................... . 373
248. Sand roses for 6 months for Ash Shariqah, Saudi Arabia .................. . 374
249. Sand roses for 9 months for AI 'Ubaylah, Saudi Arabia ................... . 374
250. Graph showing drift potential during the year at Badanah, AI Qay~i:imah, and
Dhahran, Saudi Arabia ........................................... . 375
251. Graph showing drift potential during the year at Ash Shariqah and AI 'Ubaylah,
Saudi Arabia .................................................... . 375
252. Map showing distribution and morphology of eolian sand in the Thar Desert of
India and Pakistan ............................................... . 378
253. landsat imagery of compound parabolic dune clusters along the southern
margin of the Thar Desert ........................................ . 379
254. Sketch showing rakelike arrangement of crests of parabolic dune arms compris-
ing compound forms near Jodhpur, India .......................... . 360
255. landsat imagery of vegetated parabolic and linear dunes in the southwestern
part of the Thar Desert ........................................... . 360
256. landsat imagery of linear and crescentic dunes in the western part of the Thar
Desert .......................................................... . 382
257. Graph showing relation of rainfall patterns to effective winds, Jaisalmer and
Barmer, Thar Desert .............................................. . 383
258. Outline map of the Thar Desert and resultant drift directions for June and Janu-
ary ............................................................. . 383
259. Graph showing drift potential at Bhuj and Barmer, Thar Desert ........... . 383
260. Map showing distribution and morphology of eolian sand in the Takla Makan
Desert, Peoples Republic of China ................................. . 384
261. landsat imagery showing compound crescentic dunes and closed elongate in-
terdune areas in the Cherchen Desert, northern Takla Makan Desert, south
of the Tarim River, Peoples Republic of China ...................... . 387

Digitized by Google
308 A STUDY OF GLOBAL SAND SEAS

FIGURE 262. Landsat imagery of the dunes along the Ho-t'ien Ho, westem Takla Makan
Desert........................................................... 388
263. Map showing distribution and morphology of eolian sand in the Ala Shan
Desert, Peoples Republic of China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
264. Landsat imagery of the main mass of sand, Pa-tan-chi-lin Sha-mo, Ala Shan
Desert, Peoples Republic of China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
265. Landsat imagery of dune fields, eastern Ala Shan Desert. . . . . . . . . . . . . . . . . . . 394
266. Maps showing mean sea level pressure distribution over the Takla Makan and
Ala Shan Deserts ................................................. 395
267. Sand roses for 6 months for Khotan, southwestern Takla Makan Desert, Peoples
Republic of China ................................................ 396
268. Sand roses for 6 months for Chi-Ian-t'ai, ellstern Ala Shan Desert, Peoples
Republic of China ................................................ 396
269. Sand roses for 6 months for Kucha, north-central Tarim Basin, Peoples Republic
of China. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
270. Graph showing variation of drift potential during the year at representative sta-
tions within the Takla Makan Desert. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
271. Graph of drift potential at Chi-Ian-t'ai and Mao-mu, Ala Shan Desert, Peoples
Republic of China ................................................ 397

Table
TABLE 44. Planimetric similarity of linear dunes of the Wahiba Sands, Oman, and linear
dunes of the Namib Desert, South-West Africa, shown by comparison of
their scale ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 369

Digitized by Google
 REGIONAL STUDIFS OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 309

Method of Study Niger; (3) northern Sahara of Algeria, western

Libya, and Tunisia; (4) Namib Desert of South-
1Ir HIS CHAPTER IS A SERIES of regional West Africa; (5) Kalahari Desert of South-West
~ studies of selected sand seas and is based on Africa and Botswana; (6) Arabian sand seas of
available literature, surface wind summaries, and Saudi Arabia, Oman, Yemen, and adjacent States;
data obtained through remote sensing. Studies of 8 (7) Thar Desert of India and Pakistan; and (8) Takla
desert regions include sand seas in parts of 17 coun- Makan and Ala Shan Deserts of the Peoples
tries. Mosaics were prepared from Landsat (ERTS) Republic of China.
imagery covering each of the regions, and, from
each mosaic, a map was made of sand bodies and
other features and their distribution. Supplemen-
Definitions
tary information about dunes of each region was The following list of foreign terms used on maps
obtained from Skylab and aerial photographs, and in this chapter are given English definitions. All
from the ground studies of other workers. definitions are adapted from gazetteers prepared by
The Landsat imagery used to map dunes are at a
scale of 1:1,000,000. Many features of relatively
Definition Map on which UIed
small dunes, including details of dune shape and Term
slipface development commonly used in dune Adrar .... Mountains ........... Sahara (western).
classifications by field geologists, are not visible at Berg . . . . .. Mountains . . . . . . . . . .. Namib. Kalahari.
this scale. For this reason, a classification some- Cap ...... Cape. point .......... Sahara (western).
what different from that used in the ground studies Dhar ..... Plateau .............. Sahara (western).
and low-altitude aerial photograph studies treated Djebel .... Hills ................ Sahara (northern).
in chapters A - I was found to be expedient here. Erg ....... Dunes. desert sand Sahara (northern).
For example, the term "crescentic dune" is here region.
used to include all dune types that are generally Hamada .. Plateau. stony plain ... Sahara (northern).
referred to as barchans, barchanoid ridges, or Ho ....... Stream, river . . . . . . . .. Takla Makan. Ala Shan.
Hu ....... Lake ................ Takla Makan. Ala Shan.
transverse dunes. Jabal ..... Hills ................ Rub' al Khali.
Classification of principal dune types according
to form and to number of slipfaces is given in Kum ..... Sand area . . . . . . . . . . .. Takla Makan. Ala Shan.
chapter E, table 9. We developed an expanded ver- Lac . . . .. .. Lake ................ Sahara. Sahel.
sion of this classification (chapter J) for our studies Massif . . .. Mountains . . . . . . . . . .. Sahara.
because dunes seen on Landsat imagery may not be Oued . . . .. Wadi . . . . . . . . . . . . . . .. Sahara. Sahel.
directly comparable to those that have been studied Pan ...... Playa ............... Kalahari.
on the ground. A comparison of the general ter- Plato ..... Plateau .. .. .. .. .. .... Kalahari.
minology for dune classification based on observa- Po ....... Lake ................ Takla Makan. Ala Shan.
tions of Landsat imagery and that based on field Puts .. .... Well, spring .. .. .. .... Kalahari.
studies is given in chapter A, table 1. Rann ..... Marsh . . . . . . . . . . . . . .. Thar.
Originally, plans for this study were to include Sebkha ... Playa. alkali flat ..... Sahara.
large desert regions in Australia, southern U.S.S.R., Sha-mo . .. Desert. sand area ..... Takla Makan. Ala Shan.
and the Sonoran Desert of the United States and Shan ..... Mountains. range . . . .. Takla Makan. Ala Shan.
Mexico. Although Landsat imagery was obtained Shatt ..... Sebkha .............. Sahara (northern).
and mosaics were prepared for these regions, time Tassili .... Mountains ........... Sahara (northern).
limitations precluded compilation of data for use in
'Uriiq . . . .. Sand dunes, dune area. Rub' al Khali.
this chapter.
Vallee .... Wadi .. .. .. .. .. .. .... Sahel.
The areas selected for this study, using Landsat Vlei .. . . .. Pan ................. Namib.
imagery, are as follows: (1) Western Sahara of WadI . . . .. Intermittent or dry Sahara, Rub' al Khali.
stream. stream valley.
Mauritania, Senegal, Mali, and Spanish Sahara; (2) oasis.
Sahel and southern Sahara of Mauritania, Mali, and

Digitized by Google
310 A STUDY OF GLOBAL SAND SEAS

the U.S. Board on Geographic Names. Those defini- Introduction

tions marked by an asterisk (*) are adapted from Several long, sharply defined sand seas (known
Webster's New Intemational Dictionary of the as ergs in northern Africa), extend southwestward
English Language, Second Edition, unabridged, from Algeria across the western Sahara. The
1960, C &: G Merriam Co., Publishers.
morphology of the dunes and the extent of the sand
seas, as determined from Landsat imagery, is
Western Sahara of Mauritania, shown in figure 188. Sand roses representing the
potential sand-moving effect of present-day surface
Senegal, Mali, and Spanish winds, and isohyets (lines of equal precipitation)
Sahara show present climatic factors in relation to sand
distribution (figs. 188, 189). A regional view of the
Summary of Conclusions sand seas is provided by a Skylab photograph
(fig. 190).
Dunes of the western Sahara are limited to two
Few of the eolian sand areas shown in figure 188
main types -linear and crescentic - distributed
are described in published reports. Sand seas along
throughout wide areas. Linear dunes are the most
the east margin have been described by Monod
common type, extending to the coast in both
(1958). Reconnaissance studies of dune areas near
parallel and en echelon forms from hundreds of
the senegal River and northward to about lat 17 N.
kilometres inland. Compound and simple linear
have been made by Tricart and Brochu (1955),
dunes are the most common forms in northern
Leprun (1971), Boulet, Guichard, and Vieillefon
Mauritania. Heavily vegetated compound linear
(1971), and by others. Grove and Warren (1968)
dunes of proportions similar to those in northem
briefly summarized work in the eastem and
Mauritania occur in southern Mauritania, near
southem parts of the region. Useful maps of the
Mederdra and Boutilimit. Crescentic dunes are region are the U.S. Department of Commerce, U.S.
common in the southern part of the desert (Aoukar Air Force Operational Navigation Charts H -1,1 -1,
sand sea), and in smaller fields along the Atlantic K-l, and 1-2; and the U.S. Army Map Service
coast. Sand sheets, generally associated with linear Series 1301, sheets NE 29, NF 29, NF 28, and NE 28;
dunes, occur throughout Mauritania, westem Mali, all maps are at a scale of 1:1,000,000. A regional
Spanish Sahara, and northern Senegal. landform map by Raisz (1952) is at a scale of 1 inch
Rainfall in the westem Sahara ranges from less equals 60 miles (1:3,801,600).
than 100 mm (4 in.) in the northem regions near
Fort-Gouraud to more than 300 mm (12 in.) in the
southern regions near Rosso. Effective winds in the Review of Previous Work
western Sahara are mostly from the north and
northeast and result mainly from anticyclonic cir- Dunes in the area bounded by Boutilimit, Saint-
culation from the Sahara and Azores high-pressure Louis, and Louga (fig. 188) are described (Tricart
cells. and Brochu, 1955, p. 148-149) as belonging to a
Relatively unvegetated linear dunes and sand zone of old red parallel dunes, 40-50 km (25-30
sheets in sand seas of northern Mauritania are in a mil long, about 1 km (0.6 mil wide, and 10-20 m
high-energy wind environment. These sand seas (33 -66 ft) high. Near the coast, interdune corridors
typically are elongate and are alined subparallel to are reported to be about 2 km (1 mil wide and dune
the resultant drift direction of present-day winds. ridges are 20-30 m (66-99 ft) high. Along the
In some places, however, linear dunes are alined Aftotit'es Saheli (fig. 188), old red dunes have been
oblique to trends of sand seas but parallel to direc- truncated by marine erosion and are overlain by re-
tions of some effective winds. Vegetated, or par- cent brown or yellow crescentic coastal dunes (Tri-
tially vegetated, linear and crescentic dunes in cart and Brochu, 1955, p. 148. 150). Sand depths of
southem Mauritania are in present-day intermedi- 20-50 m (60-165 ft) are reported near Mederdra
ate- or low-energy wind environments with con- and Saint-Louis by Tricart and Brochu (1955, p.
siderable rainfall. Large crescentic dunes in the 146-147).
Aoukar sand seas generally occur downwind of The mean diameter of dune sand in the southem
escarpments or other topographic barriers. area is reported to range from 0.25 to 0.40 mm -

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 311

N EXPlANATION

LINEAR DUNES

EIJ
..". . Mostly simple. BOme compound

~ Feathered compound

CRESCENTIC DUNES
~

Bj . "- ",-', "

,~:
"" ,
::

,
Simple barchenoid ridges

Compound barchanoid ridges with

- long. straight segments
ff) Complex crescentic ridges with
linear elements superlmpoeed

oy .
SHEET AND STREAKS

Compound sheets and streaks with

linear elements
BEDROCK OUTCROPS OR BARRIERS

ROCK AND GRAVEL DESERT PlAINS

WITH NO LARGE DUNES
PlAYAS AND COASTAL SEBKHAS

VEGETATED PLAIN

-500- ISOHYET-Average annual rainfall.

In millimetres
BOUNDARY OF SAND SEA

MAP LOCATION

o 200 KlOMETRES
I i! ! I
o 100 MIlES

DISTRIBUTION AND MORPHOLOGY OF EOLIAN SAND of the western Sahara of northern Africa. Map based on Landsat imagery.
Isohyets from Griffiths and Soliman (1972). (Fig. 188.)

Digitized by Goog Ie
312 A STUDY OF GLOBAL SAND SEAS

~
rm
Reduction factor
Resultant drift potential (ROP) --'22
259 1
Drift potential (OP) .......

~ResUltant drift direction (ROD)

TYPE OF SAND ROSE used in this chapter. See chapter F, figure
93 and caption for explanation of parts of rose. (Fig. 169.)

that is. fine to medium grains (Tricart and Brochu.

1955. p. 156-162). Sands of recent active dunes
along the coast and elsewhere are reported to have
a maximum diameter of 0.29 mm. Sand of the older
red dunes is generally better sorted than that of the
recent dunes and ranges in diameter from 0.46 to
0.56 mm (mostly medium grained).
Three groups of immobile dunes are described
(Leprun. 1971) in the Ferlo region of northwest
Senegal. south of Dagena (fig. 188). Differentiation
among the three groups of dunes is on the basis of
soil development. texture. color. and vegetative
cover.
The sequence of development of dunes in
southern Mauritania and northern Senegal. based
on paleontologic and pedologic studies. may be
summarized as follows:
1. The large red simple to compound linear dunes
trending east-northeast in southwestern
Mauritania and northern Senegal are believed
to have formed during an arid period when
sea level was at least 50 m (165 ft) 19wer than
it is at present, about 18.000 years before the
present (Grove and Warren. 1968. p. 196;
Boulet and others. 1971. p. 104. 113).
2. Reddening of the dune soils and gullying
reportedly took place during a pluvial period.
when the ~Mgal and other rivers cut chan-
nels through the linear dune fields to the
Atlantic Ocean (Grove and Warren. 1968. p.
196).
ElONGATE SAND SEAS IN THE NORTHWEST SAHARA of
3. Sea level rose during the Nouakchottian
Mauritania and Spanish Sahara . Skylab photograph
transgression (5.500 to 5.000 years before the Sl4-136-37S6. Bright pink areas are interpreted as sand seas;
present) and the lower senegal River valley grayish-pink areas are gravelly, sandy, and rocky plains without
was flooded as far as Bogue. 250 km (155 mil large dunes; and dark areas include bedrock outcrops and
inland. Beaches between the great linear coastal sebkhas. (Fig. 190.)
dunes were formed at an elevation 4 -6 m
(13 -18 ft) above the present sea level (Grove
and Warren. 1968. p. 196).

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 313

4. The sea regressed, and sand from the beaches westward to the base of bedrock outcrops near Ak-
was blown by winds into the coastal "dunes joujt (fig. 193). The dune (DrA' Malichigdane) is 100
jaunes" of Tricart and Brochu (1955, p. 148). km (62 mil long and about 1 km (0.6 mil wide and is
5. Gullying of old red dunes in the southern part oriented parallel to the resultant drift direction at
of the Aoukar sand sea (fig. 188) is continuing Atar, as indicated by a sand rose on the map (fig.
at present (Daveau, 1965). Gullying of dunes 188). Near Akjoujt, the dune has small crescentic
in the coastal region is interpreted (Tricart elements which give it a complex form. Other
and Brochu, 1955, p. 153 ) as evidence of a linear features in the same vicinity resemble the
transition zone between the arid northern en- solitary linear dune in size and orientation but are
vironment and the more humid southern en- composed of single rows of en echelon short linear
vironment. dunes.
The Aoukir sand sea of southern Mauritania and
Mali (fig. 194) consists mainly of crescentic dune
Interpretation of Landsat Imagery ridges. Dunes in the western part of this sand sea
are compound crescentic. Measurements of simple
Analysis of Landsat imagery shows that two crescentic dunes were made on images of the
main dune types -linear and crescentic - domi- Aoukir sand sea southeast of Tidjikdja. Boundaries
nate the western Sahara. No dunes of star, dome, or of fields of crescentic dunes in the Aoukir sand sea
parabolic type were recognized on Landsat imagery characteristically match the spurs and reentrants of
of this region. Complex dunes are relatively rare. plateau escarpments.
The simple and compound forms of linear and cres- Small fields of minor, probably simple crescentic
centic dunes are similar to those in other sand seas dunes occur along the Atlantic coast near the
that have been examined on Landsat imagery SEnegal River (fig. 188) and among bedrock out-
(chapter n. Sand sheets and streaks are con- crops south and east of Atar.
spicuous features of the western Sahara, especially
in northern Mauritania. Sand sheets dominate the sand seas in the north-
ern part of the western Sahara to a much greater ex-
Linear dunes of simple to compound form occur
tent than in any other deserts observed on Landsat
in elongate sand seas that extend from north-
imagery. North and west of Atar and south of Fort-
western Mali to the coast of Mauritania (figs. 188,
Gouraud are numerous prominent sand sheets
190, 191). The largest single linear dune seen on
associated with rock outcrops and escarpments
Landsat imagery is in central Mauritania.
(chapter J, fig. 187). Linear dunes occur on many of
Measurements of linear dunes near Mederdra these sheets, and some sheets have small groups of
and Boutilimit (fig. 191), made on Landsat imagery,
crescentic dunes. Some sheets are 100 km (62 mil
yield an average dune width of 0.66 km (0.41 mil. long, or longer (figs. 190, 194).
Spacing (wavelength, measured crest to crest) in
this area varies widely, but averages 1.58 km (0.98 Sandstorms which blow southwestward, parallel
mi). In some areas, linear dunes as much as 20 km to the trend of the sand seas, have been observed on
(12 mil long are alined en echelon, oblique to- the Landsat imagery of northern Mauritania, a region
trend of the elongate sand seas. Near Atar, at about characterized by strong winds, sand sheets, and
lat 2rN., long 14W. (fig. 193), trends of the elon- linear dunes.
gate sand seas are N. ~oE., but linear dunes are
oriented N. 40 E.
0
Surface Wind Flow and Precipitation
Compound linear dunes measured on Landsat
images of the MakteIr sand sea, northeastern Surface wind circulation in the western Sahara
Mauritania (fig. 192) commonly are more than 50 (between lat 14 N. and 26 N., and long 6 W. to 19
km (30 mil long, have an average width of 0.97 km W.) is controlled by the Azores and Sahara high-
(0.60 mil, and are very evenly spaced about 1.4 km pressure cells, and the Intertropical Convergence
(0.66 mil from crest to crest. Zone (I.T.C.Z.). During winter, the Azores high is
A solitary linear dune extends from a gap in an nearest the African coast, and the Saharan high is
escarpment south of Atar and crosses a sandy plain strongest (fig. 195A). Outward flow from these

Digitized by Google
314 A STUDY OF GLOBAL SAND SEAS

VEGETATED LARGE LINEAR DUNES shown in relation to precipitation isohyets (in millimetres per year) in southern Mauritania.
Landsat false-color imagery E1069 -10554. Green vegetation appears red, and red dune sands appear yellow on false-color imagery.
(Fig. 191.)

pressure systems results in winds from the north- moved farthest north, stations south of lat 18 N.
east or east over much of the western Sahara. experience southwesterly monsoonal winds.
By June the Sahara high has been replaced by a Rainfall in the western Sahara ranges from less
thermal low in the central Sahara (fig. 195B). At than 50 mm/yr (2 in./yr) at Bir Mogrein (Fort-Trin-
this time the Azores high is strongest and is cen- quet) in the north to more than 300 mm/yr (12
tered farthest from the coast at approximately lat in./yr) at Rosso, in the south (fig. 188). Rain
3000' N., long 3230' W., resulting in more north- showers and thunderstorms during summer are
erly wind flow at most stations. Some southern sta- associated with the I.T.C.Z., which moves north-
tions near the coast (from Boutilimit westward) ex- ward to about lat 16 N. Thus, rainfall and associ-
perience northwesterly winds during this season. ated southwest winds begin earlier and end later at
During July-September, when the I.T.C.Z. has southern stations than at northern stations (fig.

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 315

COMPOUND LINEAR DUNES IN THE MAKTE'iR SAND SEA north of the Er Richat structural dome, Mauritania. landsat imagery
E1103 -10434. A basin filled with eolian sand is bounded on the west by hogback ridges. Numerous sand streaks and small linear
dunes extend downwind (southwestward) from the hogbacks. (Fig. 192.)

196). Areas of active sand as seen on Landsat imag- tions generally toward the west and the southwest
ery (fig. 193) are mostly north of the 2()()-mm (8-in.) in the winter (fig. 197 A) and toward the south-
isohyet shown in figure 188. Considerable vegeta- southwest, south. or southeast during the summer
tion is visible on imagery of dunes south of this line (fig. 1978).
(fig. 191).
The cyclic swing of resultant drift directions dur-
Direction and Amount of Sand Drift ing the year occurs most strongly in the south-
western, heavily vegetated part of the region (fig.
Annual resultant directions of sand drift (RDD's) 198). Stations south of approximately 18 N. ex-
at most stations in Mauritania are toward the south- perience reversals in resultant drift directions dur-
west or south (fig. 188). Annual sand roses for these ing July -September that may be caused by the
stations indicate considerable directional variability southwest monsoon (fig. 199), but the reversal is
of wind during the year(fig. 188). This variability is usually not of sufficient strength or duration to
cyclic and is reflected in seasonal shifts in resultant have much effect on potential sand movement.
drift directions at all stations in the region. These Moreover, the reversal occurs during the time of
shifts follow trends in the surface wind circulation maximum rainfall in the south. which tends to
just described; they produce resultant drift direc- further reduce its effect.

Digitized by Google
316 A STUDY OF GLOBAL SAND SEAS

SOLITARY LINEAR DUNE (ORA' MALICHIGDANE) between Atar and Akjoujt, Mauritania. landsat false-color imagery
E1140-10500. Relatively small elongate fields of crescentic dunes occur south of the linear dune. The intense yellow in the false-
color imagery indicates that the sand is bright red and lacks vegetative cover (which appears red on false-color imagery); contrast
with vegetated (red) linear dunes in figure 191 in southern Mauritania. (Fig. 193.)

Average drift potential of nine stations in the 200). A second potential sand-moving season in
region (excluding Port-~tienne, which has an Mauritania occurs at some stations during winter
anomalously high drift potential) is 394 VU (vector Oanuary and February}, when anticyclonicity in
units, chapter F), an intermediate value compared north Africa is most strongly developed (fig. 200),
to other deserts of the world (chapter F, table 16). If Distribution of the elongate sand seas and linear
Port-Etienne (2,232 VU) is included in the com- dunes of the western Sahara agrees in general with
putations, the average drift potential is 569 VU, and the annual resultant drift direction of sand com-
the region as a whole is considered to be a high- puted from records of present-day winds. The
energy desert. Total drift potentials and ratios of agreement suggests that major changes in wind
the resultant drift potential (RDP) to the total drift directions have not occurred since emplacement of
potential (DP) are highest in the northern areas and the sand seas, thought to be about 18,000 years
lowest in the southern areas (fig. 188), RDPIDP before the present (Grove and Warren, 1968, p.
decreases regularly toward the southwestern part 196). The anomalous trends of some linear dunes
of Mauritania, as a result of the greater directional (oblique to the trends of the sand seas in which
variability of the wind in this region. The strongest they occur) may result from the action of certain
potential sand-moving season at some stations, effective components within the general wind
such as Port-Etienne and Bir Mogre'in (Fort-Trin- regime, in the manner described by I. G. Wilson
quet), occurs in June, when the western Sahara lies (1973, p. 97) and by Warren (1970, 1972) for aline-
in the zone of strong winds between the Azores ments of dunes in northern and southern parts of
high and the thermal low over the Sahara (fig. 195, the Sahara.

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 317

Sahel and Southern Sahara of region. Crescentic ridges there, however, have slip-
faces toward the northwest. Broad sand "streaks"
Mauritania, Mali, and Niger extend from east-northeast to west-southwest
across the entire middle Niger region, approx-
Summary of Conclusions imately parallel to the resultant drift direction com-
puted from present-day winds.
LINEAR DUNES, CRESCENTIC DUNES, and
elongate sand sheets occur in the Sahel and Most of the region studied averages more than
southern Sahara of northern Africa. Most of these 100 mm (4 in.) of rainfall per year. Isohyets show
features are at least partially vegetated. Two sets of that much of the region lies in the Sahelian climatic
dunes, with different orientations, occur along the zone, where average yearly rainfall may be as much
Niger River near Timbuktu. One set of closely as 600 mm (24 in.) per year; the rest of the region
spaced, narrow linear dunes with sharp crests is lies in the sub-Saharan zone, where average annual
alined at N. 67 E., in approximate agreement with rainfall ranges from 100 to 250 mm (4 to 10 in.) per
the resultant drift direction of the northeast trade year (Hills, 1966, fig. 15, table 15).
winds. More widely spaced, eroded ridges of in- Present-day effective winds come mainly from
determinate type form a second set in the same the northeast or southwest throughout the Sahel.
region and are oriented approximately east-west. Drift potentials vary widely from station to station
Compound crescentic dunes in the Aoukir sand but are generally lowest in the south. Highest drift
sea have unusually straight segments with slipfaces potentials at most stations occur in June, but some
toward the west-southwest. Very large barchans stations also have a winter sand-moving season.
with slipfaces toward the southwest or west occur Eolian landform dimensions, orientation, degree of
in the Valle de l' Azaouk in the eastern part of the weathering and erosion, and relationship to un-

COMPOUND CRESCENTIC DUNES IN THE AOUKAR SAND SEA, Mauritania. landsat imagery E1140-10502. This field is directly
south of the area shown in figure 193. (Fig. 194.)

Digitized by Google
318 A STUDY OF GLOBAL SAND SEAS

._.._ .. __ ._ ..... - ..... - . - . _... _._------'

~- ..-~-.. -~-.---~.-_,~-------~-o~----~_

iB \0"
- .._ .. _._. _ ..._ . - - - - . - .-.--- . _ .._ - _.....----_.----.-_._._-_. __._---_.._ - - - - '
SEA-lEVEl MEAN-PRESSURE CONTOURS, in millibars, of the Sahara and adjacent regions in (A) Janu-
ary and (8) June. Average seasonal trends in surface wind flow indicated by arrows. Region of westem
Mauritania included in this study is shown by dark pattern. The position of Intertropical Convergence
Zone (!.T.C.Z.) indicated by pattern is modified from Crutcher and Meserve (1970). (Fig. 195.)

Digitized by Google
 REGIONAL STUDIFS OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 319
(1958). Tricart (1959. 1965). Dresch and Rougerie
(1960) , Daveau (1965). Boulet. Guichard. and
Vieillefon (1971). Leprun (1971), and others. Other
papers (in English) are by Grove (1958). Prescott
and White (1960). Grove and Warren (1968). and
White (1971). A report on eolian landforms ob-
served on Skylab photographs includes a discussion
of the Middle Niger River region (McKee and
100 Breed. 1977). Useful maps of the region are U.S.

I
Department of Commerce. U.S. Air Force Opera-
tional Navigation Charts J-2. J-3. K -1. and K -2;
and U.S. Army Map Service, Series 1301. NF 30. NE
30. and NE 31; all maps are at a scale of 1:1.000.000.

Review of Previous Work

The Aoukar sand sea {"Awkar" of Monod,1958}
is an area of compound crescentic dunes in
southern Mauritania (the western part was de-
scribed earlier in this chapter). (See also fig. 188.)
Dunes in the eastern part of the Aoukar sand sea
are described (Grove and Warren, 1968, p. 197 -198)
as a sea of "massed barchans," "unusually rec-
tilinear transverse ridges" 40-50 m (130-165 ft)
high, oriented north-northwest to south-southeast
(with slipfaces toward the southwest) and covered
with bushes, grasses, and trees. Soil development
and intermittent gullying of the Aoukar dunes is in-
VARIATION IN RAINFALL DURING THE YEAR at Rosso, Akjou-
terpreted (Daveau, 1965; Grove and Warren, 1968,
jt, and Fort-Gouraud, Mauritania. Rainfall is greater and begins
earlier in the summer at the more southern stations. (Fig. 196.)
p. 198) as evidence of climatic oscillations in the
Sahel.
derlying topography - as observed on Landsat im- Dunes in the inland delta region east of the
agery of the sand seas north of the inland delta of Aoukar sand sea, near Timbuktu (fig. 201), are
the Middle Niger River and in the Vallee de described by Tricart (1959, 1965) , by Boulet,
Guichard, and Vieillefon (1971, p. 110-111), and by
l'Azaouk area - are interpreted as evidence of
more than one episode of aridity and eolian ac- Grove and Warren (1968, p. 198-200). The oldest
tivity. described set of dunes consists of ridges 10 -20 m
(about 32 -65 ft) high and several kilometres long
Introduction (Grove and Warren, 1968, p. 199) which formed
locally by eolian winnowing of Niger River sands
Landsat imagery has been used to map and in- (Tricart, 1959. p. 335; 1965, p. 180-181) during a dry
terpret. with the aid of Skylab photographs. the dis- period, probably 20,000-15,000 years before the
tribution and morphology of eolian sand along the present (Boulet and others, 1971, p. 111-112). Red-
southern border of the Sahara (fig. 201). Severe dening of the upper 2 -3 m (7 -10 ft) of the dune
drought and encroachment of eolian sand in that sands, and their erosion by the Niger River, occur-
region are thought by some workers to be caused by red during a more humid period 5,500 -5,000 years
poor agricultural and grazing practices \lut are ~t before the present (Grove and Warren, 1968, p.
tributed by others to changing climate. 199). Formation of "grey-brown and yellow dunes
Published reports (in French) o~ eolian features which are lower and much less continuous but
in the region are by Urvoy (1935. 1942) , Monod sharper than the older ones" (Grove and Warren,

Digitized by Google
320 A STUDY OF GLOBAL SAND SEAS

15" 10"

,, ,,
25"
A B

SPANISH SAHARA I
... J
SPANISH SAHARA I
...J
/' /'
hFort-Gouraud I

,,I : ,,j / t
---------~ ---------~

_51
63

MAUIUI'ANIA

18 13
......21 /1i
11
17
"NouakdIoIl
\ 4
.... BouIImIt
17 ~BouIImIt
22 48 19
58 48

/ / I

15"

o 4OOKlOMETRES o 400 KLOMElRES

I I
I I I I I I
o 200MlES o 200MlES

RESULTANT DRIFT DIRECTIONS (arrows), drift potentials (denominators), and resultant drift potentials (numerators) at weather
stations in Mauritania during (A) February and (8) July. Terms are defined in chapter F. During February, resultant drift directions are
toward the west or southwest over most of the region. During July, resultant drift directions are more toward the south and are
roughly toward the southeast in the southern, vegetated parts of the sand sea near Rosso, Nouakchott, and Boutilimit. (Fig. 197.)

1968, p. 199) occurred later, during an arid period. cised into lateritic plateau surfaces east of the great
Early maps of sand seas in the Middle Niger middle bend of the Niger River, near Menab (fig.
River region between Mopti and Timbuktu (Urvoy, 201). These dunes are assigned to a "pre-
1942, p. 48; Monod, 1958, p. 27; Grove and Warren, Nouakchottian eolian episode," about 7,OOO~,500
1968, p. 199) do not distinguish between two dune years befor the present (Boulet and others, 1971, p.
sets subsequently recognized on Skylab photo- 112-113). Other sets of dunes in the area between
graphs (McKee and Breed, 1977). Relationships ob- Tillal>ery and Bourem are mainly compound linear
served on space photographs and Landsat imagery dunes (Prescott and White, 1960, p. 200-201).
are thus difficult to reconcile with work of previous Encroaching eolian sand sheets in the Vallee de
authors. l'Azaouk area are described as flat, featureless
Crescentic dunes described as "great swellings" plains of drift sand from the Sahara that have filled
several hundred metres wide are reported in the and blocked most river channels (White, 1971, p.
Vallee de I'Azaouk, an abandoned river valley in- 70).

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 321

N changes of color on Landsat imagery indicate that

these dunes are vegetated.
East of the AoukAr sand sea, between Mopti and
Timbuktu (fig. 201), is a field of much eroded dune
ridges of indeterminate type (which appear gray-
ish-pink on imagery) that are oriented nearly east-
west across the course of the Niger River. Braided
OP-15 DiP-1 river courses are adjusted to the pattern of these
FEB. APR.
parallel dune ridges, thus forming a trellislike ar-
rangement in which numerous tributary streams
and narrow lakes occupy interdune spaces (fig. 202
B).

~ .. In contrast to the set of eroded dune ridges are

fields of much narrower, sharper, brighter hued,
and differently oriented linear dunes. The narrow
dunes can be traced on imagery (fig. 202) from com-
OP-17 DiP-I pound linear dunes in the Azaouid sand sea, which
JUNE AUG.
occupies the southern Sahara north, west, and east
of Timbuktu (Monod, 1958), southward into the Lac
Niangay area. These narrower dunes strike west-
southwest across parts of the other dunes near Lac
Niangay and also across some of the trellis-pat-
terned drainage (fig. 202B). Superposition of the
narrow dunes over the east-west set, to which .the
Niger River had become adjusted is interpreted as
OP-I evidence of more than one episode of large-scale
DEC. eolian activity.
SAND ROSES FOR 6 MONTHS FOR NOUAKCHOTT, East of Bourem and Gao in the Vallee de
Mauritania (fig. 188). Resultant drift direction (arrow) swings I'Azaouk region (fig. 201, 202) are numerous in-
cyclically toward the southwest in February, toward the east- dividual crescentic dunes, probably barchans, com-
southeast in August, and then back. Southwest winds of the monly 1 km (0.15 mil or more from hom to hom,
monsoon season do not have much effect at this station
(note August sand roses) because Nouakchott is too far
which are clearly visible against the dark laterite
north (Iat 18"07' N.). Drift potential, in vector units (chapter plateau surfaces recorded on Landsat transparen-
F), given for each month. (Fig. 198.) cies. These are among the largest individual dunes
of crescentic type that the authors have observed
on Landsat imagery and are probably compound
Interpretation of Landsat Imagery barchans (chapter J). To the south and west in the
Vall~e de I'Azaouk region are fields of compound
Measurements of Landsat transparencies of com-
pound crescentic dune ridges in the Aouklr sand crescentic dune ridges similar in form to those of
sea of Mauritania show that these dunes are com- the Aoukar sand sea farther west but with slipfaces
posed of slightly crescentic segments that range in to the northwest, rather than to the southwest.
hom-to-hom width from 2.2 to 9.6 km (1.4-5.9 mil Broad sheets of pale-red sand strike west-south-
(average, 4.1 km or 2.5 mil. The dune ridges are west across the entire Middle Niger region (McKee
spaced an average 1.71 km (1 mil apart (crest-to- and Breed, 1977) from at least long 5 E. to long 10
crest). The Aoukiir dunes have the straightest seg- W. (fig. 201). Numerous sand-free spaces between
ments of any ridges of crescentic (transverse) type the sand "streaks" expose underlying topography,
observed on Landsat imagery of the world's sand including remnants of abandoned drainages.
seas (chapter J, tables 38,39). The Aouklr dunes are Triangular-shaped lakes, such as Lac Faguibine and
red (yellow on false-color imagery). Seasonal Lac Niangay, occupy depressions in some of the

Digitized by Google
322 A STUDY OF GLOBAL SAND SEAS

DP-33 DP-22 DP-32

JAN. FEB. MAR.

DIP-13
AUG.

DIP'"'12 DP-18
NOY. DEC.

SAND ROSES FOR 12 MONTHS FOR BOUTILIMIT, Mauritania. This station experiences a cyclic swing of the resultant drift direc-
tion (arrow) during the year. It also experiences a reversal in effective wind directions during summer. The reversal is a result of
westerly through southwesterly winds which occur when the Intertropical Covergance Zone has moved farthest north. Sand roses
for Fort-Gouraud, which exemplify the pattern of winds in regions farther north, are shown in chapter F, figure 97B. Drift potential,
in vector units (chapter F), is given for each month. (Fig. 199.)

sand-free areas. The sand sheets, like the narrow to the south. As in the western Sahara (fig. 195), an-
linear dunes associated with them, extend into and ticyclonic circulation from the Sahara and Azores
across some of the old eroded dune ridges near the highs results in flow of surface tradewinds from
Niger River (fig. 2028). northeast to southwest across much of the region
during winter, but surface wind flow during sum-
mer may be from the northeast or southwest.
Surface Wind Flow and Precipitation Winds from the southwest (southwest monsoon),
resulting from cyclonic circulation associated with
Climatic zones of the Sahel and southern Sahel in the Intertropical Convergence Zone (I.T.C.Z.), are
southern Mauritania, Senegal, Upper Volta, Mali, strongest during June through August. Because the
Niger, and Chad (Hills, 1966, fig. 15-1) are in- monsoon advances from south to north, stations
fluenced by circulation from the Sahara and Azores along the southern margin of the Sahara are
highs to the north and equatorial trough (I.T.C.Z.) affected most by the southwest winds. Greater fre-

Digitized by Google
 REGIONAL STUDIFS OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 323
quency of the southwest winds at the moPe plex wind distributions at most stations (fig. 203 B).
southern stations is apparent from the annual sand All of the Sahel and southern Sahara region, by
roses shown on figure 201. definition (Hills, 1966, table 15), receives more than
Annual wind distributions at most stations are 100 mm (4 in.) of rainfall per year, as illustrated by
complex, reflecting the interaction of the pressure isohyets (fig. 201). Most of this rain occurs during
systems just described. During winter, when the June -September, the period of the southwest mon-
region is dominated by anticyclonic flow from the soon (fig. 204). Rainfall increases from north to
northeast and east, wind distributions are nar- south, reflecting the greater influence of the mon-
rowed, and winds are steadiest (fig. 203A). During soon at the more southern stations. Effects of the
summer, however, interaction of the southwest increased rainfall are evident on Landsat imagery,
monsoon and the northeast winds results in com- which show an increasing amount of vegetation

A B

A
100 1\ 600
I\
I \
/1\
./ \ I
I \\
_/ \ I \
\I \
V \-BIR MOGRElN
\
\
\
\
\
ATAR \
\
\
\

100

OJ F M A M J J A SON 0 J~~~L-~~7-~~L-~~~~O~~N~D
MONTHS OF THE YEAR

VARIATION IN DRIFT POTENTIAL during the year (A) at Boutilimit, Atar, and Bir Mogrefn (Fort Trinquet) and (8) at Port-Etienne,
Mauritania. Bir Mogrefn and Port-"Etienne experience strong sand-moving seasons in June, when the northern part of the region is in
the zone between the Azores high and the thermal low in the Sahara. Some stations in the north, such as Atar, may, in addition, ex-
perience a sand-moving season in January-February, when anticyclonicity is strongest over the central Sahara (Sahara high). (Fig.
200.)

Digitized by Google
324 A STUDY OF GLOBAL SAND SEAS

o 50 100 150 200 250 l)() KlOMETRES

I II I I I II II I I I I
o 25 50 75 100 125 150 175 MLES

DISTRIBUTION AND MORPHOLOGY OF EOLIAN SAND in the Sahel and southern Sahara of northern Africa (above and facing
page). Map based on Landsat imagery. Isohyets from Griffiths and Soliman (1972) and U.S. Naval Weather Service (1968). (Fig. 201.)

southward from areas near the 1oo-mm (4-in.) winds (chapter P, under methods of study, wind
isohyet toward the areas of greater rainfall. data available and limitations). Sand roses are not
adjusted for varying effects of vegetation at
Direction and Amount of Sand Drift different stations or for presumed dampening
effects of rainfall on sand movement. Potentially
Annual resultant drift directions t3 across most of effective winds from the southwest, shown on the
the region shown in figure 201 are generally west- sand roses for Timbuktu and Gao (fig. 201), blow
ward. Dunes and sand sheets observed on Landsat across heavily vegetated areas and are commonly
imagery are alined with resultant drift directions at associated with rainfall. Thus, the true sand-mov-
several stations; for example. Kidal (off the map to ing effectiveness of the southwest winds, relative
the north). Tillaoory, Mopti, Nioro du Sahel, and to effectiveness of the northeast winds (which blow
Nema. Near other stations, such as Timbuktu and across the barren Sahara), is probably less than im-
Gao, resultant drift directions are neither alined plied by those sand roses.
with nor transverse to the dunes. The discrepancy Compound crescentic dune ridges with north-
may result from the unusual conditions for obser- west-facing slipfaces, in the eastern part of the
vation at these stations. An unusually low drift po-
tential at Timbuktu, for example (fig. 201), suggests
that the station may be sheltered from the northeast " New tenns are defined in chapter F and in the 1I10lll8ry.

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 325

EXPLANATION
LINEAR DUNES

~{J :::::on
CRESCENTIC DUNES

Simple barchanoid ridges

Compound

Barchanoid ridges

Barchan.

SAND SHEETS AND STREAKS

I.....----J
BEDROCK OUTCROP OR BARRIER
I"--~
ROCK OR GRAVEL DESERT PlAIN
I--~ WITH NO LARGE SAND DUNES
MAP LOCATION VEGETATED FLOODPLAIN OR MARSH
...................
--200--ISOHYET-Average annual rainfall, in millimetres

- - - - - BOUNDARY OF SAND SEA

(Fig. 201. -Continued.)

region, may be in agreement with the northward Drift potentials within the region studied vary
resultant drift directions suggested by the sand widely possibly because vegetation in some
roses for Menaka and Gao (fig. 201). The closely localities may shelter stations from sand-moving
spaced linear dunes north of Timbuktu, oriented N. winds. Some stations north of lat 16 N. record
67 E. (fig. 202), are in approximate alinement with moderate or high drift potentials, for example
modem desert tradewinds (northeast winds), the 'Ayoun el 'AtrOllS, (382 VU), Nema (803 VU), and
effects of which are reflected in the sand roses for Gao (388 VU). Drift potentials may be higher in
N~ma and 'Ayoun el 'Atrous farther west. regions north of lat 16 N. than in more southern

Digitized by Google
326 A STUDY OF GLOBAL SAND SEAS

COMPOUND LINEAR DUNES OF THE AZAOUAD SAND SEA (A) near Timbuktu (above). landsat imagery E1096-10043. The
Niger River runs from west to east. B, Two sets of dunes north of Mopti and south of Timbuktu, directly south of the area shown in
A (facing page). landsat imagery E1132-100S2. The Niger River system is adjusted to the presence of large east-west dunes and is
overridden on the east by sets of narrow linear dunes and sand sheets oriented like those to the north in A. Spaces between the
large east-west dunes and some sand-free areas between the sand sheets are occupied by lakes. (Fig. 202.)

regions because the former are exposed longer to Sahara. These sand seas consist mostly of large
strong northeast winds than to the relatively weak compound and complex star and crescentic dunes.
southwest winds. or because more northern regions as in the Grand Erg Occidental. the Grand Erg
are less vegetated. Most stations studied have a po- Oriental of Algeria. and the~' Marzuq of Libya.
tential sand-moving season in June. as shown in They are mostly confined to basins separated by
figure 205. Some stations also have a potentially plateaus and low mountain ranges. Sand seas of
strong winter sand-moving season; for example. linear dunes. such as the Er Raoui. Erg-Iguidi. and
'Ayoun el 'Atrous (fig. 205) . Portions of the Sahel 'Erg Chech. commonly extend for long distances
and southern Sahara. like the Thar Desert of India across rock and gravel plains.
and Pakistan. generally are exposed to highest drift Large isolated star dunes are a common type in
potentials prior to the onset of the rainy season and the northern Sahara. In the central and southern
after an extended period of dryness. parts of the Grand Erg Occidental. star dunes have
mean diameters which range from 0.7 to 1.7 km (0.4
to 1.1 mil and on the downwind margin of the
Northern Sahara of Algeria, Grand Erg Oriental they have maximum diameters
Tunisia, and Libya of 2.4 km (1.5 mil. Large star dunes also occur along
Summary of Conclusions the southern margin of the Grand Erg Occidental
and grade northward into compound crescentic
SAND SEAS CONTAINING distinctive dune dunes in the interior of the sand sea. Star dunes in
types extend across large parts of the northern the ~$ii' AwbarI. Libya. and the Grand Erg Orien-

Digitized by Google
 REGIONAL STUDIPS OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 327

(Fig. 202. -Continued.)

tal, Algeria, are arranged in linear trends which These complex wind regimes reflect the interaction
cross. of winter westerlies, summer northeasterlies, and
Widely spaced (5.7 km, or 3.5 mil compound the southwesterly and northwesterly winds which
linear dunes in the Er Raoui, Erg-Iguidi, and 'Erg result from cyclonic circulation and frontal
Chech extend south-southwestward into northern passages. Annual resultant drift directions are
Mali and Mauritania. Barchanoid ridge dunes form toward the southeast at most stations in the Grand
a network pattern in the northeastern part of the Erg Oriental.
Grand Erg Occidental, mostly in Tunisia along the
border with Algeria and Libya. Unimodal or bimodal high -energy wind regimes
Sand seas in the Ramlat Zalliif and ~a\trii' Awbirl occur in southern Tunisia, a region characterized
of Libya include crescentic dunes, star dunes, and by a network pattern of barchanoid ridges.
linear dunes. Large crescentic dunes, and isolated Wind regimes southwards from lat 30 N. in 0

star dunes with an average horn-to-horn width western Algeria are generally unimodal or bimodal.
Linear and crescentic dunes are common types in
(and spacing) of 2 km (3 mil occur in the ~J;uii'
this region. Resultant drift directions are generally
Marziiq, to the south in Libya. Fields of barchans
toward the southwest in the southern part of the
extend in trains from the ~a' Marzuq southwest-
Grand Erg Occidental and the 'Erg Chech.
ward across the Plateau du Mangueni to the Great
Erg of Bilma in Niger. Complex wind regimes occur in the Ramlat
Most of the northern Sahara receives less than lallaf and ~' Awbar'i of western Libya. Near
100 mm (4 in.) average annual rainfall. Complex in- Ghudamis and Nalut in northwest Libya, annual
termediate- to high-energy wind regimes character- resultant drift directions are toward the northeast.
ize the Grand Erg Occidental and the Grand Erg To the southeast near Sabhii, and farther south near
Oriental of Algeria. Very large star dunes occupy Bordj Omar Driss (Fort Flatters), annual resultant
hundreds of square kilometres in these deserts. drift directions are toward the west or southwest.

Digitized by Google
 328 A STUDY OF GLOBAL SAND SEAS

1
A DP-13
FEB. 220

200

ft,
I \
I ,
I \
I ,
DP-16
B 100 I \
JUNE
i I
I
I ,
\
,
~
MONTHLY SAND ROSES FOR NIORO DU SAHEl, MAU, for I ,
(A) February, during the winter season of steady east-to-north- 80 I ,
I ,
east winds, and for (8) June, when the southwest monsoon in-
teracts with east-to-northeast circulation to produce a complex
I
I \,
wind distribution. Arrowed vector represents resultant drift 80
I ,
direction, in vector units. (Fig. 203.) I ,
I ,
I ,
I ,
On the average, intermediate drift potentials oc- I ,
cur in the deserts of Algeria, whereas high drift po- I \
tentials occur in the deserts of northwest Libya. 20 I
) \
,
Highest drift potentials occur in March or April at I
I \
,
most places in the northern part of the area
described but occur in June at some southern sta-
tions. MONTHS OF THE YEAR

Introduction VARIATION IN RAINFAll DURING AN ENTIRE YEAR at Gao,

Mali, lat 16"15' N. (dashed line) and Nioro du Sahel, Mali,lat
Landsat imagery has been used to map and 15"14' N. (solid line). The southwest monsoon is farthest north
describe the distribution and morphology of eolian in August, resulting in greatest rainfall at that time. Most
sand in several main sand seas of the northern southern stations receive greater rainfall than stations in the
Sahara. The main sand seas in Algeria, Tunisia (fig. north because they are subject to the southwest monsoon for a
206), and western Libya (fig. 20'7) are in structural longer period. (Fig. 204.)
and topographic basins, separated by plateaus and
low mountain ranges. Meteorological data were ob- The main geologic studies of the sand seas in
tained from numerous weather stations in and Algeria and northwestern Libya are by Alimen
around the sand seas. Sand roses on the maps (fig. Doudoux-Fenet, Ferrere, and Palau-Caddoux
206, 207) show potential sand-moving power of sur- (1957), Alimen (1965), Capot-Rey (1943, 1945,
face winds, and isohyets are shown for areas that 1947a, b), and I. G. Wilson (1971a, 1972, 1973).
receive 100 mm (4 in.) or more rainfall per year. Quaternary stratigraphy of a region that includes

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 329

and sand sheets in the northern ~ii' Awbiiri and

~' Marziiq are mostly limited to reconnaissance
papers (Capot-Rey, 1947b) and brief sections within
reports on the geology or hydrogeology of the
region (Conant and Goudarzi, 1967, p. 721-728;
80 Jones, 1971) .
I",
I , Field studies of linear dunes in the Ramlat Zalliif
I ,
I , are reported by McKee and Tibbitts (1964).
I , Features of Libyan sand seas observed on Gemini
I ,

I 80
/"".........
I
I
I

I,
,
,
V-,M811
space photographs are described by Pesce (1968). A
physiographic map of northern Africa by Raisz
(1952) shows the sand features at a scale of 1 inch
'"
" II ,
, equals 60 miles (1:3,601,600). Other useful maps of
the northern Sahara are the U.S. Department of
\ J \
y. /\ \ Commerce, U.S. Air Force Operational Navigation
Charts H-2, H-3, H-4, G-l, G-2, J-3, and J-4;
1,"" , I" V ' , \\ U.S. Army Map Service, Series 1301, NH 30, NG 29,
1
1 \ \ .. - - ,
\
NH 31, NG 31, NG 30, NF 31, NH 32, NH 33, NG 32,
/ . / '.--.,
'A.-1ft" 'AIroQa M8Urlt8n..-'\ \\ NG 33; and U.S. Department of Commerce, U.S. Air
/ \ \ Force World Aeronautical Chart 541. All maps are
\ \ at a scale of 1:1,000,000.
\. \./
,.......,/~---
The Grand Erg Occidental and the IErg
Chech, Algeria
Review of Previous Work
South of the foothills of the Atlas Saharien and
MONTHS OF THE YEAR
north of the Plateau du Tademait (fig. 206) is a
topographic basin containing the Grand Erg Occi-
VARIATION IN DRIFT POTENTIAL during the year at represen- dental. The sand sea, which covers about 103,000
tative stations in the Sahel. All stations have a sand-moving
season in June; 'Ayoun el I Atrous has a sand-moving season in
km2 (64,000 mi2) (Wilson, 1973, p. 86) is bounded on
February (winter). (Fig. 205.) the west by the Oued Saoura, a large intermittent
stream. The sand sea is partly separated from the
the Grand Erg Occidental is summarized by 'Erg Chech (to the southwest) by low mountains,
Chavaillon (1964). and separated from the Grand Erg Oriental (to the
French and Arabic terms for various eolian land- east) by plateaus.
forms are shown in tables 26 -35 in chapter J. Direct Thirty-two samples of sand from large compound
translation of Capot-Rey's nomenclature is compli- star dunes along the west margin of the sand sea,
cated by his usage of the tern "sir' as a curved slip- between Taghit and Kerzaz (fig. 206) were
face on a dune of any type (Capot-Rey, 1945, p. analyzed for grain-size distribution (Alimen and
392-394, 400, fig. 1; Capot-Rey and Capot-Rey, others, 1957). In the large star dunes grain size is
1948, pI. 1 (3, whereas to most authors it means a unimodal, ranging from 0.20 to 0.22 mm. Sands of
linear dune. minor dunes, sand sheets, and interdune surfaces in
Studies of dune forms and processes in central the central part of the sand sea are bimodal, with a
and southern Libya (Bagnold, 1933, 1941) have range in size from 0.18 to 0.26 mm. All dunes of the
largely been in the Sarir Kalanshiyii, at Rami! al Grand Erg Occidental, according to Alimen,
KabIr and ~' Rabyiinah in the Libyan desert, Doudoux-Fenet, Ferrere, and Palau-Caddoux (1957,
and in the Great Western Desert of Egypt east of the p. 108-109, 112, 135-136, 167-169) and Alimen
region studied for this chapter. References to dunes (1965, p. 282), were formed by eolian reworking of

Digitized by Google
 330 A STUDY OF GLOBAL SAND SEAS

o 50 100 150 KLOMETEAS

I I
I I I I I
o 25 50 75 MILES

~
685

4 ~
Ghardlll8 \.

'r-
.'

MAP lOCATiil'l

(:~

/ ~:~~f
.
., '
, ,

2"

DISTRIBUTION AND MORPHOLOGY OF EOLIAN SAND in Algeria and western Tunisia. Map based on Landsat imagery. Isohyets from
Dubief (1952). (Fig. 206.)

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 331

..k~ '''k.: ~
north of
Touggourt
"
....... .
Touggourt
"

EXPLANATION

Compound
Compound, in linear chains
Megabarchans
UNDIFFERENTIATED COMPLEX DUNES
Barchanoid ridges
SIMPLE SHEETS AND STREAKS
Discontinuous barchanoid ridges

B
COMPOUND SHEETS AND STREAKS WITH
LINEAR DUNES Barchanoid ridges with closed inter-
LINEAR ELEMENTS

.' :
dunes (fishllCllle pattem)

B
e.... . ::.: .. .
:: . . .. . .. Simple, short
, BEDROCK OUTCROP OR BARRIER
Complex crescentic ridges with
Compound
8
".: . . >'.<1 : ;:~: Linear elements superimposed ROCK OR GRAVEL DESERT PlAIN WITH
.\ .. : . : r
-.:'/ . ".
Feathered NO LARGE SAND DUNES
-/ .;.' -~. .~ .: ~;

B
.-
~: .... ;";, "
Dome-shaped dunes superimposed PLAYA
Complex with STAR DUNES
v='l Crescentic dunes super- ':..... ' VEGETATED MARSH
~
-.
Simple, and some compound
imposed
~ Star dunes superimposed
B . " '. '.:.0:
Compound, in linear chains - 1 00 - ISOHYET-Average annual rainfall. in
~ . ::Y.~

t;;:;;i~l
CRESCENTIC DUNES
Simple barchanoid ridges
::::
D . .
.',:, -.",
:

.,-;
DOMESHAPED DUNES
Simple and compound -
millimetre.
- - BOUNDARY OF SAND SEA
- INTERMITTENT DRAINAGE-Showing
direction of flow
(Fig. 206. - Continued.)

Digitized by Google
332 A STUDY OF GLOBAL SAND SEAS

13"

\
\

\,,
\

25
C5'

Djanet ~
/~
51

MAP lOCATlCl'I

a 50 100 150 KILOMETRES

I I I
I I
a 25 50 75 MILES

DISTRIBUTION AND MORPHOLOGY OF EOLIAN SAND in sand seas of northwestern Libya and southeastern Algeria. Map based
on landsat imagery. Absence of isohyets indicates that the entire region receives less than 1()() mm annual precipitation (Fig. 207.)

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 333

EXPLANATION

LINEAR DUNES
2B" B. . . ..i... Simple. short

. "," :.".' Mostly Simple. lOme compound

Compound

H Feathered
Two l81li superimposed

27'
td
D
Complex with star dunes superimposed
CRESCENTIC DUNES
Simple barchanoid ridges

. :
Compound

8
( Barchanoid ridges

Discontinuous barchanoid ridges

. ~, "
. ~: .. , ).
Barchans
Complex

Reversing ridges
Ridges with star dunes superimp088d

Ridges with linear elements superimposed

STAR DUNES
Simple. and lOme compound

Compound in sinuous chains

SHEElS AND STREAKS

BEDROCK OUTCROP OR BARRER

ROCK OR GRAVEL DESERT PLAIN WITH NO

LARGE SAND DUNES

D PLAYA

- - - BOUNDARY OF SAND SEA

INTERMITTENT DRAINAG~Showing
direction of flow
'lA'

, ",'

\
(Fig. 207 - Continued.)

Digitized by Goog Ie
334 A STUDY OF GLOBAL SAND SEAS

basal early Quaternary (Mazzerien) sediments, white calcareous silt and containing gastropod
mainly alluvium, and she suggested that sand in the shells, commonly overlie the interdune crusts
big star dunes along the western margin has under- (Capot-Rey, 1945, p. 398).
gone more eolian reworking than has the sand of Sand of major dunes in the Grand Erg Occidental
the interior. Suggested sources of sand in the appears a uniform ocher (All men and others, 1957,
Quaternary sediments are outcrops of Miocene- p. 173 -174) but in bright sunlight appears as
Pliocene sandstones (Alimen and others, 1957, p. various shades of red. Recent minor dunes in the
38-40). sand sea and on Quaternary terraces of the Oued
Sand of the Grand Erg Occidental is said to be Saoura are yellow (Alimen and others, 1957, p.
composed almost exclusively of quartz with sparse 173-174). Distinct color differences of various
tourmaline and zircon. This pure composition was groups of dunes are interpreted (Alimen and
interpreted by Capot-Rey (1945, p. 393) as indicat- others, 1957, p. 175-177, figs. 83, 89) as evidence of
ing a long sedimentary history. A similar conclu- dune formation under various climatic conditons
sion is reached by Alimen, Doudoux-Fenet, Fer- and at different times.
rere, aDd Palau-Caddoux (1957) on the basis of
grain-size distribution of sand and on the variety of Interpretation of Landsat Imagery
sizes, colors, and orientations represented.
Interdune surfaces commonly are silicocal- Imagery of the western part of the Grand Erg Oc-
careous crusts developed on sands similar to those cidental (fig. 208) shows very large compound star
of the dunes but lacking red color (Capot-Rey, dunes along their downwind (western) margins
1945). Remnants of Neolithic bog soils, composed of between Taghit and Kerzaz. Toward the interior of

SOUTHWESTERN MARGIN OF THE GRAND ERG OCCIDENTAL, Algeria. A, Complex crescentic and star dunes of the interior of
the sand sea grading westward into large compound star dunes along the Oued Saoura. landsat imagery E1116 -1 0123. 8, Southem
boundary of the Grand Erg Occidental at Kerzaz and large linear dunes of the Erg er Raoui. landsat imagery E1115 -10071 . The area
shown in 8 is just southeast of the area in A. (Fig. 208.)

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 335

the sand sea, star dunes grade into complex dunes, widely spaced on rock and gravel plains. Dunes
crescentic and star (fig. 206). Measurements of measured on Landsat imagery (fig. 209) have a
Landsat transparencies of star dunes along 50 km mean width of 1.0 km (0.6 mi) and are spaced an
(30 mi) of the Oued Saoura, north of Kerzaz, show average 5.7 km (3.5 mi) apart. Their wide spacing is
that the star dunes have a mean diameter of 0.9 km unusual, compared with that of linear dunes in
(0.5 mi) and are spaced an average of 1.3 km (0.8 other deserts (chapter J, table 36). Dunes of the 'Erg
mi) apart. Among this group of dunes is the Great Chech, which are yellow on false-color imagery are
Dune at Kerzaz, illustrated by Alimen, Doudoux- actually red (chapter J, fig. 168B).
Fenet, Ferrere, and Palau-Caddoux (1957, pI. V, 1), Linear dunes of the 'Erg Chech can be traced on
which is more than 200 m (650 ft) high. imagery (partly shown in fig. 209) southwestward
into sand seas of northwestern Mali and north-
The western margin of the Grand Erg Occidental eastern Mauritania. Extension of the sand several
matches the contours of folded mountain ranges hundred kilometres southwestward, in the form of
(fig. 208), which although of relatively low relief, linear dunes, seems to support I. G. Wilson's (1971a,
exert profound topographic control on the down- p. 196) contention that dunes such as these are
wind boundary of the sand sea. "sand-passing" dunes that transport sand great dis-
Beyond the ranges to the southwest are complex tances from basin to basin across the Sahara.
linear and star dunes of the Erg er Raoui (fig. 208B).
The linear dunes, which have star dunes on their The G rand Erg Oriental, Algeria and
crests, extend southwestward across plains and Tunisia
through gaps in the low-lying ranges to the 'Erg Review of Previous Work
Chech (fig. 209) and Erg Iguidi. The 'Erg Chech
consists of very large southwest-trending com- East of Ghardala and EI Golea and north of the
pound linear and complex linear and star dunes Hamada de Tinrhert lies the Grand Erg Oriental of
01"00'
\
GRAND ERG OCCIDENTAl

(Fig. 208. -Continued.)

Digitized by Google
336 A STUDY OF GLOBAL SAND SEAS

OrJ)' Fieldwork in the Grand Erg Oriental has not been

\ extensive. A ground traverse across the Erg by I. G.
Wilson (1971a) provided sand samples and field ob-
servations for that author's (1973) geomorphic
analysis, based primarily on aerial photographs and
analysis of topographic maps. The regions of high
dunes, which I. G. Wilson (1973, p. 104) called
"draas," contain mixed sand sizes; the rest of the
sand sea, according to him, is composed of sand
having a modal size of about 0.16 mm (fine sand).
Whether I. G. Wilson's figures are entirely based
on the 620 sand samples collected in the northern
part of the Algerian ergs (I. G. Wilson, 1971a, p.
185) or were derived from formulae based on
postulated relationships between size, type, and
grain-size distribution of dune and draa (I. G.
Wilson, 1973, p. 94-97, 101, figs. 18, 19) is not clear.
Wilson stated (1973, p. lOS) that "in the absence of
direct data, the variations of grain size in the ergs
can be roughly inferred from the patterns of draa
size and orientation, and from aerial photographs
by variations in sand color and the distribution of
slipfaces on the dunes." Thus, some workers
believe that coarse sands develop dunes of size and
SUBPARALLEL COMPLEX DUNES of linear and star types of the orientation different from those of fine sands for a
'Erg Chech, Algeria. Landsat imagery E1116-10135. Widely given area, presumably because sand of each grain-
spaced dunes extend southwestward across gravel and rock size is acted upon in various ways by different com-
plains toward the sand seas of Mali and Mauritania. Outline in-
ponents of the wind regime (I. G. Wilson, 1973, p.
dicates part of image on which measurements of dune length,
width, and wavelength were made. (Fig. 209.) 95-96; Warren, 1972).
Sand in the Grand Erg Oriental is believed (I. G.
Wilson, 1973, p. 84-85) to be derived from deflation
southeastern Algeria and western Tunisia. The of alluvium in basins upwind of the sand sea. Esti-
Grand Erg Oriental includes an area of about mated mean deflation rates for the bajada surface
192,000 km2 (119,000 mi2), and is about 70 percent near Biskra, beyond the northern margin of the
sand-covered (I. G. Wilson, 1973, p. 91-92). The Grand Erg Oriental (fig. 206, 219) is 1 -4 m (3 -13 ft)
sand, if spread out, would have a mean thickness of in 2,000 years (I. G. Wilson, 1971a, p. 188). Esti-
26 m (85 ft). The mean height of dune ridges in the mated age of the Grand Erg Oriental is 1,350,000
Grand Erg Oriental is 117 m (385 ft), and they are years, based on an average annual net flow into the
spaced an average of 1.6 km (1 mil apart (I. G. area (by wind) of 6 million tons of sand (I. G.
Wilson, 1973, p. 91 and fig. 7, p. 88). The center of Wilson, 1971a, p. 197).
the Grand Erg Oriental, with about 90 percent sand-
cover, is continuous over about 80,000 km2 (50,000 Interpretation of Landsat Imagery
mi2) (I. G. Wilson, 1973, p. 92). The rock and gravel
basement is nowhere more than a few metres Landsat imagery of the Grand Erg Oriental (fig.
beneath the dunes of the Grand Erg Oriental, even 210) shows that the lowest part of the area is oc-
where sand covers 90 percent of the surface (I. G. cupied by a large salt playa, the ShaH al Jarid. The
Wilson, 1973, p. 92). Perennial grass (Aristido sp.) Grand Erg Oriental lies downwind of the playa, and
and shrubs grow on dune slopes and in interdune is bounded on the south by slopes that rise toward
areas (Capot-Rey, 1947b, p. 185). the Hamada de Tinrhert and Plateau du Tademait

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 337

(fig. 206) . The abrupt southern boundary of the Dunes south of the Shan al Jarld and along the
sand sea seems to be topographically controlled. for border between Algeria and Tunisia are barchanoid
its margin matches the 306-m (l.000-ft) contours of and compound barchanoid ridges in a network pat-
the plateau slope (fig. 210) . The Grand Erg Oriental tern (Monod. 1958; Cooke and Warren. 1973;
tapers off to the north into sand sheets. south of the chapter J. table 33). Segmented sand ridges (yellow)
Shall al JarId. To the east. in Tunisia. large dunes enclose interdune hollows. which appear bluish- or
end in a distinct rampartlike ridge of sand (fig. 210). greenish-gray on the false-color image. Elements of

.
N

Ster
dunea

NORTHEASTERN PART OF THE GRAND ERG ORIENTAL, pattern of barchanoid ridge dunes (yellow) and fully enclosed
Algeria and Tunisia. landsat false-color imagery E1109--()9313. interdune areas (bluish- and greenish-gray) merges with a pat-
Size of crescentic ridges gradually increases southward. To the tern of isolated star dunes. The dark blue area at the top left of
east the dunes, which appear yellow, end in a high, curved the imag~ is a playa, the Shas~ al Jarid. Red spots are vegetated
boundary ridge of sand. To the south and west, the network oases. (Fig. 210.)

Digitized by Google
338 A STUDY OF GLOBAL SAND SEAS

the pattern are progressively larger toward the are measured in a direction parallel to the resultant
south and west (downwind) margins of the sand drift direction in this part of the Sahara. The
sea (fig. 210), where the dune ridges are replaced by differences may have resulted from changes in the
star dunes. structure of wind regime with distance across the
The central part of the sand sea and its southern sand sea.
margin west of Ghudamis consist of isolated star
dunes with a distinctive pattern. These star dunes The ~ahra' AwbarI and Ramlat lallaf,
are not distributed at random, but form at the nodes S Libya
of crossing dune trends, according to I. G. Wilson
(1971b, p. 268; 1973, p. 99, fig. 16). The association Review of Previous Work
of star dunes with crossing dune trends can be South of the AI Hamadah al Hamra' and east of
traced westward into the western part of the Grand the border with Alg~ria is the ~a(u-a' Awbiri and its
Erg Oriental, where the star dunes merge into eastern extension, the Ramlat Zalliif. Sand in the
linear dunes to make complex forms resembling Ramlat Zallaf is composed mainly of rounded
chains (fig. 211). grains with coatings of red or yellow iron oxide
Measurements of the isolated star dunes in the (McKee and Tibbitts, 1964, p. 5). Sand samples
Grand Erg Oriental were made on Landsat imagery were taken from linear dunes, interdune areas, and
(chapter J, fig. 180B, tables 40, 41). Mean diameter sand and gravel plains near Sabha (fig. 207), and
of the dunes ranges from 0.7 km (0.4 mil at the each type of sand deposit has a distinctive texture
northern (upwind) end of the Grand Erg Oriental at (McKee and Tibbitts, 1964, p. 8.-10). Depth of sand
about lat 32 N., long 8 E., to 1.7 km (1 mil at the is not well known, although, for an area in the
southern margin (fig. 211). Mean wavelength southern ~' Awbari, Glennie (1970, enclosure
(crest-to-crest distance) ranges from 0.8-6.7 km 4) showed 120 m (395 ft) of unconsolidated sand
(0.5-4.1 mil at the northern end to 1.5-3.1 km overlying Cretaceous rocks. Farther east McKee
(0.9-1.9 mil at the southern end. I. G. Wilson (1973, and Tibbitts (1964, p. 5) recorded well samples of
p. 92, 98) has computed the cross-sectional areas of eolian sand more than 1,000 ft (305 m) thick.
relatively small star dunes at the north end of the Linear dunes in the ~*a' Awbiri are described
field (about lat 3230' N., long 730' E.) to be as by Glennie (1970, p. 95) as about 100 m (330 ft) high
much as 10,000 m2 (33,000 ft2) for each dune. and 100 km (62 mi) or more long, with surfaces
covered by small barchanlike dunes. The latter
Measurements of the star dune chains in the
represent eolian activity related to present winds,
western part of the Grand Erg Oriental (fig. 211)
whereas the large linear dunes were probably
were made on three sample areas of Landsat imag-
formed by stronger winds of the past, according to
ery. The measurements show that differences in
Glennie, (1970, p. 95).
dune size and spacing occur in a downwind direc-
tion in the sand sea. At the north (upwind) end of Interpretation of Landsat Imagery
the field, the dune chains are spaced an average of
3.1 km (2 mi) apart and are composed of star dunes Landsat imagery of the eastern, central, and
with a mean diameter of 0.9 km (0.5 mil. At the western ~*ii' Awbari (fig. 212) show a definite
southern (downwind) end of the field, the chains relation between topography, sand distribution, and
are spaced an average 6.3 km (4 mil apart and are dune type. Crossing trends of large dune ridges are
composed of star dunes with a mean diameter of 1.6 evident on the southern margins of the sand sea
km (1 mi). Relative areas of dune/interdune sur- (fig. 212); in the center of the field, the ridges merge
faces were measured by Kevin Horstman on Land- into a field of star dunes. Star dunes also occur
sat transparencies, using a density-slicer to exploit around the edge of the Haqtadat Zegher (fig. 212).
the color contrast between the two kinds of sur- Margins of the sand sea coincide approximately
faces. Ratio of dune/interdune areas at the north with contours of the surrounding plateaus. Eleva-
end of the field is 80/20 and at the south end of the tion of the Hmpadat Zegher, although slight (500 m,
field is 40/60. or 1,650 ft; fig. 207) apparently is sufficient to pre-
vent the large dunes to the east at lat 2730' N., long
Differences in dune size, spacing, and relative 13 E., which have an elevation of 300 m (985 ft),
area from north to south in the Grand Erg Oriental from invading it.

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 339

The ~a~ra' Marziiq, Libya and Niger of dune sand from the Grand Erg Occidental and
the Grand Erg Oriental of Algeria. Samples show
Review of Previous Work bimodal distribution of grain sizes, with most of the
The ~ra' Marzuq (fig. 207) occupies 58,000 km2 sand grains larger than 0.35 mm (coarse or larger),
(36.000 mi2) in a structural and topographic basin although sand less than 0.10 mm (very fine) also oc-
bounded on the north, west and southwest by the curs. The bimodal character of the ~$a' Marzuq
Hammadat Marzuq, a 1,450-km-long (900-mi-Iong) sand diminishes from east to west across the sand
cuesta of post-Tasillian-Nubian continental rocks. sea, and at the western margin the sands have a
mainly sandstones (Capot-Rey, 1947a. p. 71. 81; Co- high proportion of fine grains (Capot-Rey, 1947a, p.
nant and Goudarzi. 1967. p. 721-728). Unconsoli- 85).
dated sands 1.200 m (3.940 ft) thick are reported to A relationship between grain size of the sand and
overlie Paleozoic rocks in the ~~ra' Marziiq (Glen- two basic forms of sand topography in the ~a'
nie. 1970. enclosure 4). Marzuq is recorded by Capot-Rey (1947a, p. 86-87).
Sand of the ~$ii' Marzuq is mainly quartz, and, Low, rolling, nearly fixed dunes of the southern
according to Capot-Rey (1947a, p. 85), the median part of the sand sea are composed of coarse sands;
size of the grains is considerably greater than that these are the basal, residual sands of the erg. Above

SOUTHWESTERN END OF THE GRAND ERG ORIENTAL, Algeria, showing 200-km-long (120-mi-long) field of star dunes in linear
chains. Skylab photograph Sl4 -138 -3885. Southern ~.undary of the field corresponds approximately to the 4OO-m (1 ,300-ft) con-
tour of the Hamada de Tinrhert and Plateau du Tademait (fig. 206). Height of large star dunes above the rock- and gravel-covered
interdune surfaces at the southern margin is about 230 m (750 ft). (Fig. 211.)

Digitized by Google
340 A STUDY OF GLOBAL SAND SEAS

the coarse sands - especially in the northern and Interpretation of Landsat Imagery
western parts of the basin - are finer sands which A Skylab photograph (fig. 213) shows major dune
comprise active high dunes with steep slipfaces. forms and their distribution at the west end of the
Interdune surfaces in the ~a1,u'a' Marzuq com- ~a9ra' Marziiq. Along the northwestern part of this
monly have calcareous crusts. No trees are reported sand sea, at about lat 2530' N., long 1215' E., are
from the interdune areas, but grasses and shrubs numerous isolated large complex dunes, which
(Aristida sp., Calligonum sp.t Cornulaca sp.) are have a mean horn-to-horn width of 2.0 km (1.3 mil
common (Capot-Rey, 1947a, p. 100-103). and a mean length of 1.5 km (1 mil and are spaced
Two main types of dunes were recognized during an average 2.0 km (1.3 mil apart. To the south and
a ground traverse by Capot-Rey (1947a, p. 87 -89). east, the isolated dunes merge into compound cres-
In the area north of Marzuq (fig. 207) t compound centic dune ridges. Southwest-trending linear ele-
linear dunes are dominant. These are composed, ac- ments are part of the dune pattern at the south end
cording to Capot:Rey, of dune ridges alined one of the basin.
beside the other, either parallel or oblique to the Still farther south, fields of barchans extend
main axis of the chain. A second main type is the across escarpments of the Plateau du Mangueni,
compound crescentic dune ridge associated with and sand is apparently being blown southwestward
isolated large barchans, reported to be as much as (fig. 213). The crescentic dunes can be traced on
180 m (600 ft) high (Capot-Rey, 1947a, p. 88-89). Landsat imagery several hundred kilometres south-

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 341

westward. from the edge of the ~~ra' Marziig to storms and frontal passages (U.S. Weather Bureau.
the Great Erg of Bilma in Niger. and U.S. Army Air Forces. 1944. p. 7). and these
mainly affect the northern Sahara. Winds resulting
Surface Wind Flow and Precipitation from the presumed effects of cyclones and frontal
passages are shown by dashed arrows in figure 195.
Surface wind flow in the northern Sahara. in-
cluding Algeria and Libya. is controlled by several Except in the highlands and on the northern
major pressure systems. as shown in figure 195. coast. the northern Sahara receives less than 100
During the winter. westerly winds prevail in north- mm (0.4 in.) of rain per year (figs. 206. 207).
ern Algeria. Tunisia. and northern Libya north of Average annual rainfall variability for most sta-
approximately lat 30 N. to the Mediterranean tions is extreme (Griffiths and Soliman. 1972. p.
coast. South of this latitude. outward flow from the f17 -98). Rainfall over most of the region is insuffi-
Azores and Sahara highs results in north to north- cient to support vegetation recognizable on Landsat
east winds. During the summer. a thermal low is imagery.
centered at approximately lat 26 N. and long 03 E. Direction of Sand Drift
and the Sahara high has been replaced by a ridge of
high pressure extending northwestward into the Stations in Algeria from lat 30 N. to the coast (in-
0

Mediterranean Sea. Outward flow of air from the cluding EI Golea. Ouargla. and Gharda'ia) are
Azores high. which is strongest in June. and from characterized by complex wind regimes (fig. 206).
high pressure over the Mediterranean Sea into the These regimes result from the interaction of the
thermal low results in northeast to east tradewinds westerly effective winds of winter. the north-
over the deserts of Algeria and Libya. easterly effective winds of summer. and the south-
Another important element of surface wind cir- westerly effective winds resulting from cyclones
culation results from the passage of cyclones (low- and cold fronts. Seasonal changes in resultant drift
pressure cells) from east to west along the Mediter- directions occur in the northern Sahara in Algeria
ranean Sea and from the passage across the desert at Laghouat (fig. 215). Ghardala (fig. 216). and
of cold fronts (fig. 214). Both events result in strong Ouargla (fig. 217). The complex effective wind dis-
winds from the west and southwest across wide tributions that characterize northern Algeria may
regions of the desert. Northwest winds sometimes account for the occurrence of star dunes. isolated
occur behind a cold front (fig. 214B). Winter. spring. and in linear chains. throughout the Grand Erg
or late autumn are the principal seasons of cyclonic Oriental and parts of the Grand Erg Occidental (fig.

CENTRAl(A) AND WESTERN(B) PARTS OF

THE SAHRA' AWBART, libya. landsat imagery

E110S--09101; E1106--09160. A, South-trend-
ing ridges in the north, southwest-trending
ridges in the east, west-trending ridges in the
south, and star dunes in the center of the sand
sea. B, Star dunes surrounding the dune-free
Hamadat Zegher. (Fig. 212.)

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 343

DUNES AT THE WEST EDGE OF THE ,SAHRA' , MARZUQ, libya northeast, probably due to influence of the nearby
(facing page). Skylab photograph . Red eolian sand fills Atlas Saharien on local winds. Annual resultant
topographically low areas and extends southwestward through
drift directions at the east margin of the Grand Erg
gaps in the t;tammadat Marzuq. line surrounds a field of isol-
ated combinations of star and crescentic complex dunes. (Fig. Oriental near Qabis, Tunisia, are mostly toward the
213.) northeast (fig. 206, 218). However, weak north-
easterly effective winds prevail in this region dur-
206). Star dunes are commonly found in areas with ing the summer, producing a resultant drift direc-
complex wind regimes. tion toward the southwest during that season (figs.
Despite complex effective wind distributions, 218, 197).
many stations north of lat 30 N. in eastern Algeria
0
Wind regimes from lat 30 N. southward in
0

have annual resultant drift directions generally western Algeria (including I-n-Salah, Reggane, and
toward the southeast (fig. 206). Bechar and .Aln Timimoun) are generally not complex (fig. 206).
Sefra, in western Algeria, along the northwest Dune types in the central and southern parts of the
margins of the Grand Erg Occidental (fig. 206) have Grand Erg Occidental are mostly linear and cres-
annual resultant drift directions to the north and centic ridges, which are associated with less

EXPlANAT10N

H Center of hlgh-prtIIUre cell !lobar

0 W.ther Itetlon
L Center of low-prtIIUre cell
d' Poln.. Into wlnd-large barb, 10
@ Celm knott; IITI8I1 barb, 3-6 knots

LNdlng edge of cold front

DIrection of ..nd drift

TWO METEOROLOGICAL EVENTS which produce south- Atmospheric Administration synoptic chart of Northern
westerly and northwesterly winds over large areas of the north- Hemisphere, April 3, 1971, 1200 G.M.T. 8, Passage of a cold
ern Sahara. A, Passage of a low-pressure system (cyclone) front from west to east across the northern desert . Winds in ad-
from west to east to the north of the desert . Coun- vance of the front are from the southwest, and those behind the
terclockwise circulation around the low results in southwest front, from the northwest. From National Oceanographic and
winds from Tamanrasset to NalUt, libya, and in southeast winds Atmospheric Administration synoptic chart of the Northern
at Biskra, northern Algeria. From National Oceanographic and Hemisphere, January 17, 1972, 1200 G.M .T. (Fig. 214.)

Digitized by Google
344 A STUDY OF GLOBAL SAND SEAS

N DP=51 DP=161

N
t FEB. APR.

t DP=33
FEB.

DP=16 DP=9
JUNE AUG. SAND ROSES FOR 4 MONTHS FOR GHARDAi"A, Algeria, be-
tween the Grand Erg Occidental and the Grand Erg Oriental (fig.
2(6). The season of winter westerly winds ends earlier at Ghar-
da'i'a than at Laghouat. The sand-moving potential of winds
(drift potential) is much greater at this station in the spring at
161 VV (vector units) than in winter (December, 44 VV; Febru-
ary,51 VV). (Fig. 216.)

AwbirI (lat 2701' N., long 1406' E.) (chapter F, fig.

115B) and south-southwestward beyond the map
DP=29 region at lat 2908' N., long 1557' E. Surface wind
DEC.
flow in western Libyan sand seas approximates that
SAND ROSES FOR 6 MONTHS FOR lAGHOUAT, north of the shown in figure 195, but data are few (fig. 207). An-
Grand Erg Occidental, Algeria. Westerly effective winds in nual sand roses for Bordj Omar Driss (Fort Flatters,
winter (from December to February) result in east-southeast- lat 2808' N., long 0650' E.), Illizi (lat 2630' N., long
ward resultant drift direction. By June, northwest winds have 0838' E.), and Djanet (lat 2433' N., long 0938' K),
been partially replaced by northeast winds, and resultant drift
direction is toward the west-southwest. Southwest winds are
Algeria (fig. 207), indicate complex wind regimes.
most prevalent in August. Drift potential, in vector units, is All three stations are situated near mountainous
given for each month. (Fig. 215.) terrain which may locally affect wind regimes.

variability of wind direction than are star dunes. Amount of Sand Drift
Resultant drift directions at I-n-Salah, Timimoun,
and Reggane (fig. 206) are toward the southwest, The average drift potential of 21 stations around
approximately parallel to the trend of linear dunes the Algerian sand seas is 239 VU (vector units,
in that region (fig. 209). Monthly resultant drift chapter F, table 16), and of 7 stations around the
directions at I-n-Salah are uniformly to the south- Libyan sand seas is 431 Vu. These are intermediate
west throughout the year (fig. 219) because I-n- and high values, respectively, compared to those of
Salah is south of the zone of winter westerlies and other desert regions around the world (chapter F,
because frontal and cyclonic activity is greatest in table 16). As in other regions, however, a wide
regions farther north. range of values occurs at individual stations, from
Annual resultant drift directions are toward the 51 VU at Djanet (fig. 207), to 571 VU at Biskra (fig.
west-northwest at Sabhi in the eastern ~~ra' 206). In general, northeastern Algeria, southern

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 345

t
N

t OP=19
FEB.
OP=52
APR.
DP-8 OP-3
JUNE AUG.

OP=27
JUNE
~ OP=8
AUG.
~ OP=6
OCT.

SAND ROSES FOR 6 MONTHS FOR QABIS, Tunisia, near the

northeast margin of the Grand Erg Oriental. Winter westerlies at
this Mediterranean coastal station are from the southwest rather
than the northwest, as in the Sahara at GhardaIa (fig. 216) or
Ouargla (fig. 217), and are stronger than spring winds. Summer
northwesterlies are weaker in this region than at stations farther
west, such as Ghardaia and Ouargla. Thus, annual resultant drift
directions at this station and at nearby Remidah, Ghudamis (in
the region of very large isolated star dunes), and NalUt (fig. 2(6)
are toward the northeast. (Fig. 218.)

Tademalt and Hamada de Tinrhert have intennedi-

DP=25
DEC. ate to low drift potentials. Exceptions to this pattern
occur at I-n-Salah (541 VU) and Reggane (555 VU).
SAND ROSES FOR 6 MONTHS FOR OUARGlA, Algeria, at the In Libya. drift potentials decrease southward and
northwest edge of the Grand Erg Oriental, illustrating complex- eastward toward Jiilii (132 VU) and AI Kufrah (81
ity of the wind regime in that sand sea throughout the year. The
complexity results from the interaction of the three dominant
VU). which are located at long 2r34' E. and long
components of northern Sahara wind regimes; winter 2320' E . respectively. near the center of the Sahara
westerlies, summer northeasterlies, and the southwesterlies high-pressure cell (fig. 195).
resulting from frontal passages and cyclones. Despite the varia-
tion of resultant drift direction during the year at laghouat (fig. The season of highest drift potentials in northern
215), Ghardaia (fig. 216), and Ouargla, annual resultant drift Algeria. Tunisia. and northern Libya usually occurs
directions at all of these stations are toward the southeast, ap- in March or April (fig. 220). Many stations in the
proximately parallel to the long dimensions of the chains of star north also have strong wintet sand-moving seasons
dunes southeast of Hassi Messaoud. (Fig. 217.) (fig. 220). The range of drift potential at many
northern stations. such as Ghardaia and Biskra.
Tunisia. and adjacent northwest Libya are in a Algeria. and Ghudiimis. Libya. may be extreme (fig.
region of intennediate to high drift potentials (fig. 221). Farther sound towards I-n-Salah and Djanet.
206). Western Algeria. including the Grand Erg Oc- Algeria. highest drift potentials can occur in June
cidental and areas to the south of the Plateau du and July.

Digitized by Google
346 A STUDY OF GLOBAL SAND SEAS

.r---------------------------------~

50 1\
~ I \
I \
DP=69 z I
::I
I
\--- Ouargla
FEB.
N ~ I
I \
\

t ti
~
~
J
30
I
I
I
I
I \

~ I
,I
~
l-
Z
W
,
I

~ I
I

DP=27 DP=37 Ii: , I

JUNE AUG. ~
0 10

OJ 0

DP=22 VARIATION IN DRIFT POTENTIAL during the year at Laghouat

DEC. (solid line) and Ouargla (dashed line), Algeria. These stations
typify much of the northern Sahara, including southern Tunisia
SAND ROSES FOR 6 MONTHS FOR I-N-SALAH, Algeria, south and northwestern libya. Laghouat experiences high drift poten-
of the Grand Erg Occidental and east of the 'Erg Chech, illustrat- tials in the winter and spring. Ouargla, which is closer to the in-
ing the steadiness of easterly to northeasterly effective winds terior, experiences highest drift potentials in the spring but is too
and resultant drift direction toward the west-southwest and to far south to have equally high drift potentials in the winter. (Fig.
the southwest. This wind regime, in a region of sand sheets, 220.)
contrasts markedly with some more complex wind regimes
Average annual rainfall is less than 100 mIIi (3.9
farther north in Algeria in a region of star and complex dunes.
(See figs. 216-218.) (Fig. 219.) in.) throughout the Namib Desert, except in some of
its extreme eastern parts. Heavy dew along the
Namib Desert of South-West in.) of coast may be the equivalent of as much as 25 mm (1
annual rainfall. High-energy unimodal
Africa winds from the southwest or south occur along the
Atlantic coast in a zone which corresponds roughly
Summary of Conclusions to the region of crescentic dunes. An intermediate-
CRESCENTIC, LINEAR, AND STAR DUNES oc- to low-energy bimodal wind regime may occur in
cur in the Namib Desert in three elongate zones the central zone of linear dunes; a complex low-
parallel to the Atlantic coast. Crescentic dunes with energy wind regime prevails near the star dunes of
an average wavelength of 0.9 km (0.6 mil occupy a the eastern Namib Desert. The observed present-
belt 5-30 km (3.1-18.6 mil wide along the Atlantic day wind regimes of the Namib Desert are roughly
coast between Swakopmund and Luderitz. Com- compatible with the observed dune types in the
pound linear dunes occupy much of the central por- several zones. The rapid decrease in wind energy
tion of the desert, from Ltlderitz northward to the inland from the coast, however, suggests that linear
Kuiseb River. These linear dunes average 2.2 km and star dunes in the interior are considerably less
(1.4 mil in wavelength and 27 km (16.8 mil in active than are crescentic dunes along the coast.
length. Isolated fields of star dunes, reversing
dunes, and complex dunes of indeterminate type Introduction
occur along the east margin of the Namib Desert.
The star dunes have a mean diameter of 0.7 km (0.4 The main sand sea of the central Namib Desert,
mil and are spaced an average of 2.2 km (1.4 mil South-West Africa, is bounded on the north by the
apart. In general, redness of the dune sands seems Kuiseb River, on the west by the Atlantic Ocean, on
to increase inland from the coast. the east by the Great Escarpment (Goudie, 1972, p.

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 347

1.r-------~----------------------~ Landsat imagery has been used to map eolian

sand distribution and morphology in the main sand
sea of the central Namib Desert (fig. 222). In-
terpretations were made with the aid of Skylab
photographs.
100
Principal published reports (in German) about
the Namib sand sea are by Range (1927) and Besler
(1972). A major work (in Afrikaans) on dunes of
the central Namib is by Barnard (1973). Reports (in

~ ,,, I English) include a descriptive geomorphic paper by

Gevers (1936). general geographic treatises by
Z
= I
I Wellington (1955) and by Logan (1960). and two
papers by Goudie (1970. 1972) that give descrip-
I
I
I
I
I
tions of the dunes near Gobabeb. Relationship of
~
,
I
I
dunes to the Tsondab River and Tschaubrivier is
described by Seely and Sandelowsky (1974).
~
\ ~
\
\ /
/ / Comprehensive aerial photography of the Namib
~ Desert is unavailable and maps of the sand sea. U.S.
~

Department of Commerce. U.S. Air Force. Opera-

Ii; tional Navigation Charts P -3. P -4. and Q-4. state
I
that "relief data are unavailable."

Review of Previous Work

Sand of the Namib Desert is thought by most
20 workers {}. Rogers. in Seely and Sandelowsky.
1974. p. 61) to be derived mainly from fluvial sedi-
ments transported to the Namib platform by the
perennial Orange River. in an area off the map (fig.
222) at about lat 28 S. long 17 E. Some sand is also
0 0

brought in by intermittent streams from the Great

Escarpment (fig. 222). Some of the sand near the
MONTHS OF THE YEAR
coast may be of marine origin (Gevers. 1936. p. 71).
The sand sea may be as old as Miocene according to
VARIATION IN DRIFT POTENTIAL during the year at Ghardala Barnard (1973. p. 2). Recent active sands in the
and I-n-Salah, Algeria, illustrating the extreme variation in drift Namib are believed to be derived mostly from
potential which may occur at some stations. Hassi Messaoud,
reworking of Tertiary dunes (Martin. in Seely and
Algeria, and Ghudamis, Libya, also exhibit this variable pattern.
I-n-Salah is far enough south to reflect the summer high in drift Sandelowsky. 1974. p. 61).
potentials which occurs at southern stations. (Fig. 221.) Field data from the Namib sand sea consist
mainly of descriptions of dune areas near the coast
15); it ends at the south in sandy plains north of or accessible from Gobabeb (fig. 222). Sand forms
Luderitz and Aus. The sand sea area is about 34.000 near Gobabeb are at the downwind end of the sand
km2 (21.000 mi2) (Barnard. 1973. p. 2). It rests upon sea near topographic barriers and thus, may not be
the Namib platform. a Tertiary erosional surface representative of those to the south in the main part
cut on schists. quartzites. and granite intrusives. of the sand sea. Linear ridges near Gobabeb are
Regional slope of the Namib platform is about 2.3 described by Goudie (1970. p. 94 -95) as "composite
m/km (65 ft/mi) from the base of the escarpment to forms made up of as many as five subparallel and
the Atlantic Ocean. Numerous intermittent streams sinuous segments which sometimes isolate deep
flow from the escarpment into playas among the hollows within the dune." Interdune valleys near
dunes. and rock outliers protrude through the sand Gobabeb are largely sand free corridors containing
as inselbergs. a few small barchans (Goudie. 1970. p. 95).

Digitized by Google
 348 A STUDY OF GLOBAL SAND SEAS

lE,
. .... .
.... .-'-

.:.: .... ::-.: .

~ ".~ .~ ... :
EXPLANATION
LINEAR DUNES
Simple, short

Mostly simple, some compound

Complex with star dunes super-
~ . . ~ imposed
CRESCENTIC DUNES
Simple barchanoid ridges

Reversing ridges

STAR DUNES
Simple, and some compound
Complex with reversing dunes
superimposed
SHEETS AND STREAKS

BEDROCK OUTCROP OR BARRER

ROCK OR GRAVEL DESERT PlAIN
WITH NO LARGE SAND DUNES
- 2 0 0 - ISOHYET-Average annual rainfall,
in millimetres
- - -- BOUNDARY OF SAND SEA

24

\
\
25

26"

a 50 100 150 KILOMETRES

27' I I
25
I
I
50
I
I
75 MILES
I

MAP LOCATION

Digitized by Google
 REGIONAL STUDIFS OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 349

DISTRIBUTION AND MORPHOLOGY OF EOLIAN SAND in the The main vegetation among the dunes near
central Namib Desert of South-West Africa. Map based on Gobabeb consists of sparse clumps of Stipagrostis
Landsat imagery. Sand roses for Zwartbank, Rooi Bank, and
sabulicola (Goudie, 1970, p. 94). Interdune areas
Gobabeb are each for the month during which highest drift po-
tential occurs and are not directly comparable to annual sand support periodic growth of grass following infre-
roses (annual sand roses for these stations are not available). quent precipitation (M. K. Seely, oral commun.,
Drift potentials, resultant drift potentials, and resultant drift 1975).
directions for Zwartbank and Rooi Bank were estimated from Fresh or brackish water that occurs beneath
December-to-August, and December-to-July data, respectively. dunes near the coast, about 85 km (50 mil west of
An annual sand rose is shown for Pelican Point. Isohyets are
from Grove (1969). (Fig. 222.)
the playa at the end of the Tsondab River (figs. 222,
223), is interpreted as evidence that the Tsondab
River, the Tschaubrivier (fig. 224), and probably
Dune ridges of the central Namib are described
other intermittent streams that flow from the Great
by Logan (1960, p. 136) as having a chaotic and dis-
Escarpment at one time reached the Atlantic but
orderly pattern, the result of disruption of their
that their courses have been truncated by the north-
crests by "blowouts," and partial closure of their in-
ward movement of dunes (Seely and Sandelowsky,
terdune troughs by "transverse" dunes. A threefold
1974). A late Pleistocene age for the earliest oc-
division of dunes is recognized by Barnard (1973, p.
cupation by man of the area near Narabeb, west of
2) in the sand sea of the central Namib. It consists of
the Tsondab Vlei (figs. 222,223), is based on the oc-
(1) "a littoral belt of closely packed transverse
curence of Early Stone Age tools, estimated to be
dunes," (2) an interior zone of "longitudinal" dunes
40,000 -60,000 years old, in an interdune valley
which he believed to have formed by deflation, and
(Seely and Sandelowsky, 1974, p. 64).
(3) an eastern zone of "multicyclic" complex dunes.
Some reported dune heights are as follows: 65 m
(210 ft) for dunes in the coastal belt between Interpretation of Landsat Imalery
Sandfisch Bay and Zwartbank (Stapff, 1887, in Dune types are distributed in three distinct zones
Goudie, 1970, p. 94); 80-100 m (260-380 ft) for across the sand sea - a coastal zone of crescentic
"basically linear dunes" at Gobabeb (Goudie, 1970, dunes, a central zone of compound linear dunes,
p. 94; 1972, p. 24) and 300 m (nearly 1,000 ft) for and an interior zone of star dunes and reversing
"high, sharply crested dunes" near the Sossus Vlei dunes. Local variations of dune patterns within the
along the interior margin of the sand sea (Barnard, zones occur especially near inselbergs and valleys
1973, p. 2,5). of intermittent streams.
Sand of the Namib Desert is generally described
as brown (Goudie, 1970, p. 94) or red (Logan, 1960, Closely spaced crescentic dune ridges occupy a
p. 136). Inland dunes were said by Logan (1960, p. belt 5 -30 km (3 -18 mil wide along the Atlantic
136) to be "brick red," whereas dunes near the coast coast between Swakopmund and Liideritz (figs.
are recorded as "yellow-white." He stated that 222, 223). Measurements of dune width (hom to
"This difference in coloration [between coastal and hom), length, and wavelength (spacing) were made
inland dunes] is believed to be the result of the on Landsat imagery (fig. 223, sample area 2) at 10-
greater age of the inland dunes which has allowed km (6-mi) intervals between Sandfisch Bay and
greater oxidation of the iron components within the Meob Bay. Mean dune width is 1.12 km (0.69 mil,
sand" (Logan, 1960, p. 136). mean length is 0.88 km (0.42 mil and mean dune
The Namib Desert sand is reported to be 90 per- wavelength, 0.87 km (0.54 mi). These dunes are
cent quartz and of uniform size, except that devia- comparable to crescentic dunes in other areas ob-
tions in median diameter and sorting occur near the served on Landsat imagery (chapter J, table 38).
coast (Barnard, 1973, p. 2). Textures of Namib sand Compound linear dunes and complex dunes,
from dunes near Gobabeb and from "sand glaciers" mainly linear dune ridges with small star dunes and
(sand sheets. including climbing dunes) near Ross- reversing dunes developed along their crests
ing Mountain, east of Swakopmund and north of (chapter J, fig. 168), extend from the middle of the
the Swakop River, have been reported by Goudie desert east of Liideritz to an abrupt boundary at the
(1970, p. 96; 1972, p. 24-25, fig. 16). south bank of the Kuiseb River (figs. 222, 223).

Digitized by Google
350 A STIJDY OF GLOBAL SAND SEAS

24000'

WESTERN PARTS OF THE MAIN SAND SEA in the central Namib Desert. landsat imagery E1383 - 08264;
E1383 - 08270. Outlined areas are samples of imagery on which (1) complex linear dunes with superimposed star
dunes were measured and (2) crescentic dune ridges were measured. (Fig. 223.)

The linear dunes of the central zone occur in dis- straight pattern of dunes alined N. 9 E. (fig. 223)
tinct parallel straight patterns (figs. 222, 223, 224). are most pronounced near valleys of the Tsondab
Disruptions of individual dune crests and interdune River and Tschaubrivier (figs. 223. 224).
troughs seem to the authors to create a complexity Measurement of the linear dunes in a 2,500 km2
of dune forms but not a "chaos" as described by (1,550 mi2) sample area in the center of the sand sea
Logan (1960, p. 136). Variations from the generally (fig. 223, sample area 1) yields the following mean

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 351

16"00'
\

o 40 KM
I
I I I
o 20 MI
.1

NORTHEAST MARGIN OF THE MAIN SAND SEA near inselbergs of the Great Escarpment (to the east) in the Namib Desert. landsat
false-color imagery E1202 - 08222. Disruption of the dune pattern around the Sossus Vlei (playa) is shown . (Fig. 224.)

values: dune width, 0.83 km (0.51 mi); dune length, dunes in the interior marginal area contrast greatly
27 km (16 mil; dune wavelength 2.20 km (1.36 mil. with the simple crescentic dunes of the coastal belt
The Namib dunes are compared with other dunes but merge into the linear dunes of the middle zone
of linear type in chapter J (tables 36, 37). (figs. 222, 223, 224) .
Fields of star dunes, reversing dunes, and dunes
of indeterminate (complex?) type occur along the Star dunes north of Aus (fig. 222) range in
eastern margin of the sand sea, near pediplains and diameter from 0.4 km (0.2 mil to 1 km (0.6 mi),
inselbergs of the Great Escarpment (fig. 224) . The mean diameter, 0.7 km (0.4 mil, and are spaced an
sharply delineated boundaries of dune fields (fig. average 2.2 km (1.4 mil apart. The Namib star
224) coincide approximately with the 1,200-m dunes are of medium size, compared with those of
(4,000-ft) contour line shown on a map of the area other deserts (chapter J, tables 40, 41). They have a
(U.S. Department of Commerce, U.S. Air Force preferred east-west orientation (large arms oriented
Operational Navigation Chart Q -4) . Varieties of N. 48 W.).

Digitized by Google
352 A STUDY OF GL(lRAL SAND SEAS

Landsat imagery shows dune-free surfaces on plateau (fig. 226). Counterclockwise circulation
pediments around outliers of the Great Escarpment around the South Atlantic high during the winter
and in valleys of intermittent streams (fig. 224). season results in southerly winds along the Atlantic
Such dune-free areas around topographic barriers coast. These winds may be deflected inland by the
may result from sheetwash by occasional torrential heating of the desert sands and thus become south-
rains along the escarpment. The escarpment is westerly winds. particularly near the coast (Harry
generally parallel to the l00-mm precipitation Van Loon. oral commun., 1974). Strong easterly
isohyet on the map (fig. 222). The eastern margin of winds also occur in the Namib Desert during the
the sand sea is marked by a high rampartlike ridge winter and are locally referred to as the "Berg
of sand (fig. 224). described by Logan (1960. p. 136) Winds." They may be initiated when coastal lows
and similar in aspect on Landsat imagery to bound- appear on the west coast during the summer or
aries of sand seas in Tunisia. as described earlier winter and move around to the south coast
(fig. 210). (Schulze. 1972, p. 506). promoting winds across the
Figure 225 shows dunes in the interior zone at adjacent desert. During the summer. a thermal low
the Sossus VleL Interbedding of white and tan develops in the interior plateau and promotes land-
playa sediments with the red sand of the dunes is ward circulation throughout southern Africa (fig.
common at the ends of the intermittent streams (M. 226). Onshore (southwest) winds in the Namib
K. Seely. written commun . 1975). Reddish hues of Desert are strongest during spring and summer.
eolian sand translate on Landsat false-color imag- The Namib Desert receives less than 100 mm (4
ery as shades of yellow (fig. 224). and the deeper in.) of rain per year. as shown by the isohyets in
the red of the sand. the more intense its yellow tone figure 222. and the variability of this rainfall may be
on the false-color imagery. In general. observations as much as 80 percent from year to year (Schulze.
made from Landsat imagery (fig. 224). Skylab 1972. p. 512. 514). An important source of moisture
photographs (McKee and Breed. 1977). and ground along the coast is the heavy dew associated with
photographs (fig. 225) confirm Logan's (1960. p. frequent fogs. This moisture may amount to as
136) statement that dunes farthest inland are also much as 2.5 cm (1 in.) of equivalent rainfall in a
the reddest. In general. also. large linear dunes in year (Royal Navy and South African Air Force.
the central zone of the Namib sand sea are redder 1944. p. 37).
than are crescentic dunes both to the west and east
of them. Reddest of all are the star dunes and other Direction and Amount of Sand Drift
eolian sand features. mainly complex dunes and
sand sheets. around inselbergs near the eastern Annual resultant drift directions along the Namib
margin (fig. 224). Relative degrees of redness can- coast. and as far as 50 km (31 mil inland. are toward
not be attributed simply to age differences. or to the north or northeast as shown by annual sand
different lengths of time in a subaerial environment rose and wind roses from Alexander Bay (south of
because numerous factors other than time can also the mapped area). Liideritz (fig. 227). and Pelican
control the degree of redness (Walker. chapter D). Point. South Africa (fig. 222). This northward and
northeastward direction of sand drift along the
coast is steady throughout the year. with the excep-
Climatic Regime 14 tion of May-June. when the east winds ("Berg
Southern Africa lies within a belt of high at- Winds") increase in strength. (See monthly sand
mospheric pressure centered at approximately lat roses for Pelican Point. chapter F. fig. 97 A). Resul-
30 S. (Schulze, 1972. p. 503). Surface wind circula-
0
tant drift directions at stations farther inland along
tion over the Namib Desert during the winter is the Kuiseb River. at the north end of the sand sea
controlled by the South Atlantic high. the Indian (fig. 222) seem to be toward the southwest. This ap-
Ocean high, and high pressure over the interior parent change in direction of potential sand move-
ment may result from a decrease in the strength of
the southwest wind from the coast inland. and a
corresponding increase in the strength of the
,. Principal sources of meteorological data are National Climatic Center.
Asheville. North Carolina. U.S.A.. and Desert Ecological Research Unit. easterly winds ("Berg Winds") in the same direc-
Walvisbaai. South-West Africa. tion.

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 353

RED SAN D DUN ES AN D TAN PLAY A SEDIMENTS along the interior margin of the Namib Desert. Ground photograph; view at the
Sossus Vlei. Photograph by M. K. Seely. (Fig. 225.)

Two lines of evidence suggest that at inland sta- west wind is an important sand-moving wind only
tions the potential sand-moving effect of the east during October (fig. 229). At Rooi Bank, only 30 km
winds is greater than that of southwest winds. First, (19 mil from the coast, the "Berg Wind" is also a
during June and July, which is the season of east very important sand-moving wind (fig. 230).
winds, the drift potential of winds from the north- Available data from stations along the Kuiseb
east quadrant (encompasses the "Berg Winds") is River suggest that the resultant drift direction along
highest at inland stations (fig. 228). Second, the the northern interior margin of the sand sea is
high energy of the sea breeze, or southwest wind, roughly southwestward but wind data from
seems to diminish rapidly inland, as shown by the Narabeb, a station within the sand sea, suggest a
decrease in drift potentials from Pelican Point east- northeastward resultant drift direction (fIg. 222).
ward to Gobabeb (fig. 222). At Gobabeb, the south- All the stations except those at the coast may to

Digitized by Google
 354 A STUDY OF GLOBAL SAND SEAS

100r-----------~----------_r----------~

2r!' 3r!'
Pelican Point Rooi Bank Zwertbank Goblbeb
WEST .....' - - - - - - STATIONS - - - - - - - - . . EAST

MONTHLY DRIFT POTENTIAL, in vector units, of winds from

the northeast quadrant ("Berg Winds") during June and July at
four stations from west to east across the northern part of the
Namib Desert. During June, the "Berg Wind" increases in
strength between Gobabeb and Rooi Bank, then decreases
rapidly between Rooi Bank and Pelican Point (fig. 222). South-
west resultant drift directions at Rooi Bank and Zwart bank dur-
ing June (fig. 222) aline with sand streaks north of the Kuiseb
,0 " River, partly shown in figure 223. Asterisks indicate drift poten-
40" _ , 0 ,2 1010 tial of southwest wind at Pelican Point during June and July, for
tB;~~~~~~=:~~IO~~L~~~~~~~ comparison. (Fig. 226.)

some degree be sheltered from effective winds by

MEAN SEA lEVEl ISOBARS (solid lines) and isobars of the 650-
millibar-pressure surface (dashed lines) at 1200 G.M.T. Arrows
high dunes nearby. All wind regimes. however.
indicate general wind directions. A, June (winter in Southern reflect both the southwest and east winds to some
Hemisphere); 8, January (summer in Southern Hemisphere). degree and suggest that in the central part of the
Modified from Schulze (1972, p. 504). (Fig. 226.) desert. wind regimes may be roughly bimodal. with
the two modes approximately equal in effective-
ness. A bimodal wind regime is compatible with
linear dunes (chapter F).
A comparison of the low drift potentials at most
inland stations with those at stations on the coast
ANNUAl. suggests that the dunes in the center and along the
EVENING interior margins of the sand sea are much less ac-
tive than those near the coast. Wind data from Aus
N (fig. 231) suggests that some areas along the eastern
margin of the Namib Desert may have complex

I
WIND ROSES SHOWING STRONG SOUTHERLY WIND
ANNUAl.
MORNING

REGIME at the south end of the sand sea in the central Namib
wind regimes. Aus is about 80 km (48 mil from
fields of star dunes. which are generally in areas
that have variable wind directions (chapter F. fig.
116).
The average drift potential of the Namib Desert is
237 VU, which places it in a moderate-energy range
Desert at liideritz. Number in centers of circles equal percent
calm; white, 14-27 knots (25-50 km/hr); black, 26-40 knots when compared to other deserts of the world
(51 -75 km/hr). (Fig. 227.) (chapter F. table 15). The coastal stations of Alex-

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 355

N These linear dunes average 0.3 km (0.2 mil in width

iA
and 26.0 km (16 mil in length, but many individual
dunes are much longer. A second type of feature
consists of long narrow sand bodies (which may be
DP=2 DP-O DP=1
JAN. FEB. MAR.
very large linear dunes or elongate sand sheets)
that trend east-west in a region of about 15,000 km2
(9,300 mi2) between the Blydeverwachterplato and
the Molopo River valley. These sand features have
an average width of 2.3 km (1.5 mil, and some ex-
tend more than 100 km (62 mil across hills and
DP-8 DP-21 valleys. They swing southward at the Molopo River
MAY JUNE and merge with the sand sea of linear dunes to the
east.
Annual rainfall in the southwestern Kalahari
Desert is less than 200 mm (8 in.) in most places but
increases northward and eastward - from regions
of little vegetation into regions of much greater
vegetation. At present, the Kalahari Desert as a

~ SEPT.
whole has little wind energy available for sand
movement, but drift potentials vary widely within
the desert. High drift potentials occur in the less-
vegetated, southernmost part of the study area.
Low drift potentials occur in the northern parts of
the study area in regions of partially vegetated sim-
ple linear dunes. The linear dunes and (or) sand
DP=O
sheets in the southwestern Kalahari Desert are
NOV. ali ned with the resultant drift directions of present-
day winds.
SAND ROSES FOR 11 MONTHS FOR GOBABEB, South-West Introduction
Africa (location shown in figs. 222, 223). The southwest wind is
strongest during October and is reflected in the long southwest The Kalahari physiographic province
arm of the sand rose. East ("Berg") wind is strongest in June, (Wellington, 1955, p. 52) includes about 1,613,800
July, and August. December sand rose is unavailable. Drift po- km2 (1,002,800 mi2) of lowland sandy desert, but
tential is given for each month. (Fig. 229.) only the southwestern part - in South Africa,
Botswana, and South-West Africa - has dunes
ander Bay and Pelican Point experience a wide recognizable on Landsat imagery. The sand sea of
range in drift potential, with highest values occur- the southwestern Kalahari Desert is in a basin of in-
ring at Pelican Point (122 VU) during the spring ternal drainage centered at Abiekwasputs on the
(fig. 232). Maximum drift potentials at inland sta- Molopo River (fig. 233). The lower parts of this
tions may occur in June or September (fig. 232). river and its tributaries (the Auob, Nossob, and
others) are abandoned drainages that have been
Kalahari Desert of South-West choked with eolian sand for at least 1,000 years
(Lewis, 1936, p. 30, fig. 6). They are thus cut off
Africa, South Africa, and from the perennial Orange River, which marks the
Botswana southern boundary of the Kalahari dune country
and drains surrounding territories.
Summary of Conclusions Published reports of the Kalahari region include
Linear dunes are the most common sand features reconnaissance work (A. W. Rogers, 1934, 1936),
of the western Kalahari Desert. Partially vegetated, field observations of dunes (Lewis, 1936), a general
mostly simple linear dunes trend northwest- geographic treatise (Wellington, 1955), accounts of
southeast and cover about 100,000 km2 (62,000 mi2). geology and geomorphology in the southwestern

Digitized by Google
356 A STUDY OF GLOBAL SAND SEAS

r~
DP=3 DP=O DP=30
JAN. FEB. APR.

SAND ROSES FOR 8 MONTHS FOR ROOI BANK, South-West during much of the year. Data for for August -November are
Africa, approximately 30 km (18 mil from the coast. Although unavailable. Compare with sand roses for Pelican Point for the
this station is very near the coast, the "Berg Wind" (at this sta- same months, chapter F., figure 97 A. Drift potential is given for
tion a northeast wind) is the most important sand-moving wind each month. (Fig. 230.)

NE morphology and distribution (Goudie. 1969. 1970).

Useful maps are the U.S. Department of Commerce.
U.S. Air Force Operational Navigation Charts P-4
and Q-4.
1:::J.----E Landsat imagery supplemented by the Skylab
photographs have been used to map and interpret
eolian sand morphology and distribution in the
southwestern Kalahari Desert. Data from
SE meteorological stations in southern Africa were
used to show relationships of the sand bodies to
s present wind regimes and precipitation patterns
(fig. 233).
WIND ROSE FOR AUS, South-West Africa, near fields of star
dunes in the interior of the Namib Desert. Data were insufficient Review of Previous Work
for computation of sand roses but indicate sand-moving winds
from several directions. 1 cm equals 10 percent occurrence.
About 200 sand samples obtained from dune
light areas, winds 14-27 knots (25-50 km/hr); dark areas,
crests. slopes. and interdune areas near the Aoub
winds 28 -40 knots (51 -75 km/hr). Source of wind data is Royal
Navy and South African Air Force (1944). (Fig. 231.)and Nossob Rivers. and Kurumanrivier in the
southern part of the sand sea are reported (Lewis.
1936) to be composed of partially rounded sand
Kalahari (Mabbutt. 1955) and Botswana (McCon- grains that have a composition of 90 percent quartz
nell. 1959; Boocock and van Straten. 1962). a paper with accessory zircon. garnet. feldspar. ilmenite.
describing landforms in relation to climatic change and tourmaline. Samples from dune crests include
(Grove. 1969). and two papers describing dune fine to medium sand. whereas samples from inter-

Digitized by Google
 REGIONAL STIJDlES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 357

Eolian and fluviatile sediments of the Kalahari

Beds of Tertiary-Pleistocene age occur throughout
most of the Kalahari Desert (Boocock and van
120 Straten. 1962. p. 151-154). Suggested sources for
some of the younger sands of the Kalahari Beds are
the eolian Triassic Cave Sandstone and its
110 equivalents. which crop out in several parts of the
region (Boocock and van Straten. 1962. p. 154).
100
Eolian sands of the Upper Kalahari Beds of late Ter-
tiary-Pleistocene age are also known as red sands.
Plateau Sands. and Kalahari Sands (Mabbutt. 1955.
90 p. 20. 25-27; table 1). and as sands of Kalahari type
(Boocock and van Straten. 1962, p. 151-154).
80
Samples of some of the near-surface eolian sand
in the Kalahari sand sea contain detrital kyanite.
epidote, and staurolite. decreasing in amount to the
southeast (A. W. Rogers. 1934. p. 11; Poldervaart.
1955. p. 106). This has led several workers to con-
clude that the latest redistribution of eolian sand of
Kalahari type in Botswana has been from the north-
west (Boocock and van Straten. 1962. p. 154).
50 Dune sand in the Kalahari Desert is described
(Grove. 1969. p. 199) as yellowish red (5 YR/5/6 to
2.5 YR/4/8; Goddard and others. 1948) without
40
regular differences in hue from dune crests to inter-
dune troughs. Observations in Botswana (Boocock
and van Straten. 1962. p. 153) have indicated that
sands of Kalahari type occur both as dark-reddish-
brown consolidated sand with considerable
/"~Gobebeb
amounts of clay and silt. and as belts of grayish-
\ white less consolidated sand.

,if ,
/ "-
N.I'II~
Unconsolidated eolian sand in the Kalahari

---~
~J.
.. ~ ,"
.......
/',
'Of'
,
'-~,....--
Desert ranges in thickness from 3 to 33 m (10 to 100
ft) (Grove. 1969. p. 195). but the entire sequence of
F
- "'- ...
MA
,,/
M J J AS ON 0 Tertiary-Pleistocene Kalahari Beds composed of
MONTHS OF THE YEAR eolian and fluviatile sediments is 100 m (300 ft) or
more thick. with maximum thicknesses where
VARIATION IN DRIFT POTENTIAL (DP), in vector units, at
these sediments have filled in old drainage chan-
Pelican Point, Gobabeb, and Narabeb in the Namib Desert,
South-West Africa. All stations experience a sand-moving nels (Boocock and van Straten. 1962. p. 152. 154). At
season in June, a result of the east ("Berg") wind. Pelican Point Sekutani (beyond the map area. in the southeastern
experiences highest drift potentials during October, when the part of the basin at about lat 2540' S. long 2420'
southwest wind is strongest. (Fig. 232.) E.) a borehole through a buried channel penetrated
157 m (515 ft) of Kalahari Beds (Boocock and van
dune areas include very fine to very coarse sand Straten. 1962. p. 152). and at Mafeking (east of
(Lewis. 1936. p. 22-23, 26-28). A "few dozen" Tsabong at approximately lat 26 S. long 2540' E.).
samples obtained between Gobabis and the Molopo a borehole penetrated 141 m (463 ft) of "Kalahari
River in the northwestern part of the sand sea con- deposits" (A. W. Rogers. 1934. p. 10).
tained moderately rounded sand (A. W. Rogers. Dune crests reportedly are bare except for grasses
1934. p. 11). (Stipagrostis and Eragrostis sp.). Dune slopes sup-

Digitized by Google
 358 A STUDY OF GLOBAL SAND SEAS

...... . .
~ 1
sane36

'<L~
, 108

EXPLANATION

LINEAR DUNES

Simple, short

Mostly simple, some oompound

SIMPLE PARABOLIC DUNES

SIMPLE SHEETS AND STREAKS

ROCK OR GRAVEL DESERT PLAIN WITH NO LARGE SAND DUNES

VEGETATED PLAINS

BEDROCK OUTCRO PS OR BARR IERS

- -200--- ISOHYET-Average annual rainfall, in millimetres

- - - - - BOUNDARY OF SAND SEA

o 50 100 150 KilOMETRES MAP LOCATION
I I
I I I I I
o 25 50 75 MilES

DISTRIBUTION AND MORPHOLOGY OF EOLIAN SAND in the western Kalahari Desert, southern Africa. Map based on Landsat
imagery. Isohyets from Grove (1969). (Fig. 233.)

port shrubs and trees as much as 3 m (10 ft) tall. and and Acacia haematoxylon) that grow to heights of
interdune areas have trees (mainly Acacia giraffae 5 m (16 ft) (Grove. 1969. p. 199).

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 359

Dunes in the lower Molopo River area above

Abiekwasputs (fig. 233) are described (Wellington. 20":ll'
\
1955. pi. 8A) as longitudinal ridges with
V-junctions opening to the northwest (upwind); tl
they are 10 to 25 m (33 to 82 ft) high and are
spaced about 300 m (1.000 ft) apart (Goudie. 1969,
p. 404). The distribution of these dunes was \ '2.
measured on aerial photographs by Goudie (1969),
who found that their branching is random except
\
where breaks of underlying slope and other
topographic influences occur.

Interpretation of Landsat Imagery

Dunes in the southwest Kalahari Desert are Kururnanrluie r
mainly of linear (longitudinal) types; barchans and ~~. ~0
other varieties described by some workers (Lewis, hi Hekskeenpen 0,0
1936, p. 30; Grove, 1969, p. 199) are mostly too small :~. /' +
to be discerned on Landsat imagery and on Skylab "00,:',:::" -;
photographs. An exception (fig. 233) is a field of11 .' -
parabolic dunes between Reitfontein and the Auob
River, which is evident on Landsat transparencies.
Most of the Kalahari sand sea consists of ,.
southeast-trending linear dunes of apparently sim- P.~
ple form which cover about 100,000 km2 (62,100
mi2). Width, length, and wavelength (distance from
crest to crest) of these dunes were measured on
Landsat transparencies of six 2,500-km2 (1,550-mi2)
sample areas in and around the regions shown in
chapter J, figure 169A. Results are discussed in
chapter J. Simple linear dunes of the Kalahari ob-
served on space imagery average 0.3 km (0.2 mil in
width and 26.0 km (16 mil in length and have an
average wavelength of 0.7 km (0.4 mil. Their aspect
on imagery made during the rainy season implies
that flanks of the dunes are stabilized by vegeta- -.!.

tion.
Very large, irregularly distributed linear sand
features. described previously as "stringer dunes"
(McKee and Breed, 1976. fig. 59), cover 15.000 km2
(9.300 mi2) and trend eastward from the Blydever-
,.,
wachterplato to the Molopo River valley. There,
they merge with the sand sea of simple linear dunes ABANDONED DRAINAGES of the lower Molopo River, Nossob
just described. and the alinement both of dunes and River, and Kurumanrivier (white), playas or pans, (blue and
of other sand features swings southward (fig. 234). white), and dunes and invasive sand sheets (yellow). landsat
The extent of vegetative cover on the large sand false-color imagery E1198-o8000. Elongate sand bodies (sheets
or very large linear dunes) extend un deflected across badlands
features is not known. They are shown as areas of (bluish-green in false-color) in the southwest corner and merge
"Kalahari sand" on some earlier maps (Mabbutt. with linear dunes that continue south from the area shown in
1955. pI. IV). chapter J, figure 169A. (Fig. 234.)

Digitized by Google
360 A STUDY OF GLOBAL SAND SEAS

These large east-trending sand features are ob- wind flow follows a counterclockwise trend. from
served on Landsat imagery (fig. 234) to cross hilly east-northeast winds at Ghanzi to northerly winds
country and abandoned river channels without at Upington (Royal Navy and South African Air
change of course or of shape. which is a common Force. 1944. fig. 2). During the summer. anti-
characteristic of linear dunes. Measurements of- cyclonic circulation is weak. and winds are varia-
these large sand features on Landsat imagery ble. Development of a thermal low over the
(McKee and Breed. 1976. fig. 59) show that they Kalahari Desert during this season promotes wind
range in length from 50 to more than 100 km (31 to flow from the southwest in the region of the desert
63 mil and have an average width of 2.3 km (1.4 near Upington and Keetmanshoop (Schulze. 1965.
mil. If they are linear dunes ridges. they are among p. 235-238). (See also the annual sand roses for
the longest and widest dunes of linear type seen on Keetmanshoop and Upington. fig. 233. which
imagery of sand seas (chapter J. table 36). reflect this southwest component.)

Climatic Regimes Annual rainfall in the Kalahari Desert ranges

from 144 mm (5.5 in.) at Keetmanshoop and 204
The Kalahari Desert is dominated by high at- mm (8 in.) at Upington in the west to more than 400
mospheric pressure in the winter and low pressure mm (16 in.) at Ghanzi in the northeast. Isohyets are
in the summer (fig. 226). During the winter. surface roughly parallel from northwest to southeast. with

1
DP-28 DP=26 DP=lS
JAN. FEB. APR.

DP-14 DP-ll DP-1S DP-3l

MAY JUNE JULY AUG.

l)pz. DP-69 DP=28 DP-30

SEPT. OCT. NOV. DEC.

SAND ROSES FOR 12 MONTHS FOR GHANZI, Botswana, in the northeastern part of the Kalahari Desert. Although monthly wind
distributions tend to be complex, resultant drift directions from month to month are toward the southwest. (Fig. 235.)

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 381
80 .-----r-----r-....----r-,....-..,.-----rK-"....----r-,....---.80

A
,
I
1\
\
\
I \
I \
\
\
\
20

" \\
\
\

~~~~~~A-~~-J~~~-~~~D

MONTHS OF THE YEAR

AVERAGE MONTHLY RAINFAll versus average monthly drift

potential of Upington, South-West Africa. Season of highest
drift potential (June -October) corresponds to season of least
rainfall, which increases the probability of sand movement dur-
ing this period. (Fig. 236.)

rainfall increasing toward the northeast (fig. 233).

The driest season in this region is winter. and the
wettest season is summer. Summer rainfall occurs o
mostly in the form of thunderstorms. These begin
earliest in the northeastern part of the region (near
Ghanzi) and during the summer extend westward VARIATION IN DRIFT POTENTIAL during the year at four sta-
across the desert (Schulze. 1972. p. 515. fig. 3). tions in the Kalahari Desert: Ghanzi and Tsane, Botswana
(northeast and southeast desert, respectively); and Mariental
and Gobabis, South-West Africa (northwest desert). Sand
Direction and Amount of Sand Drift movement at stations with high drift potentials (Ghanzi,
Upington) usually occurs during well-defined seasons of high
Evidence of vegetation on Landsat imagery. and drift potential, with periods of much lower drift potentials the
data from published sources (Grove. 1969; Flint and remainder of the year. (Fig. 237.)
Bond. 1968) confirm that most of the Kalahari
Desert east of the 200-mm (8-in.) isohyet is a this region. thus. cannot be deduced on the basis of
relatively inactive sane sea. However. most of the morphology alone. as can be done in the Sahel
present-day resultant drift directions are parallel or (described earlier in this chapter). where crossed
subparallel to trends of all linear dunes and (or) dune trends clearly indicate two periods of dune
sand streaks visible on Landsat imagery. Dune growth.
alinements seem continuous throughout the Most present-day wind regimes in the Kalahari
southern Kalahari Desert. from the less vegetated Desert are complex (fig. 233). although resultant
southwestern part to the heavily vegetated regions drift directions at the various stations are relatively
of presumably less active dunes in the northwest steady during the year (fig. 235). The region of less
and east. Distinct stages of dune development in vegetated. presumably active sand in the south-

Digitized by Google
362 A STUDY OF GLOBAL SAND SEAS

western part of the desert is characterized by ob- The compound crescentic dunes in AI 'Uriiq al
tuse bimodal winds. as shown by the annual sand Mu'tariejah (northeastern Rub' al Khali) are
rose for Upington (fig. 233). The weaker southwest unusually large. with a mean width (horn to horn)
mode occurs principally during the summer. Addi- of 2.8 km (1.7 mil and a mean length of 2.1 km (1.3
tionally. the effective wind regime at Upington (fig. mil. Presence of a field of smaller south-facing
189) is similar on the basis of energy and directional crescentic ridges at the north edge of AI 'Uruq al
characteristics (chapter F. fig. 105) to present-day Mu'taridah might be interpreted as evidence of at
effective winds associated with linear dunes in least tw'o major episodes of dune building.
other regions. This relationship indicates that linear Compound linear dunes at the western (down-
dunes visible on Landsat imagery of the Upington wind) margin of the Rub' al Khali are the largest
region may have developed in response to winds observed in this study. averaging 1.5 km (1 mil
similar to those occuring in the region today. wide. Compound linear dunes in the southwestern
The average drift potential of nine stations in the Rub' al Khali have mean widths of 0.7 km (0.4 mil
Kalahari Desert is 191 VU. indicating that the desert and lengths of commonly more than 100 km (60
as a whole has relatively little energy available for mil. Their southwest trend is parallel to the annual
sand movement (chapter F. table 16). However. resultant drift direction computed for AI 'Ubaylah.
drift potentials within the desert range widely from Large compound linear dunes. similar in width and
36 VU at Gobabis to 520 VU at Upington. Most sta- length to linear dunes of the Namib Desert. South-
tions have low drift potentials. which. combined West Africa. occur in the Wahiba Sands. Oman.
with the effects of vegetation in weakening effec- Sheets of sand associated with linear. crescentic.
tive winds. may mean that these areas have negligi- and star dunes occur in the Ad Dahna' and the sand
ble amounts of sand drift under present-day condi- seas in the northern part of the Arabian Peninsula.
tions. The important exception to this pattern is at Giant dome-shaped dunes with a mean diameter of
Upington. where the season of minimum rains 1.2 km (0.8 mil fill the NafUd ath Thuwayrit. NafUd
coincides with the season of highest drift potential as Sirr. and Nafud Shuqayyiqah.
(fig. 236). as in India-Pakistan. The time of highest Average annual precipitation is less than 100 mm
drift potential at a station depends upon its position per year (0.39 in./yr) throughout most of the sandy
in the Kalahari Desert. At Upington (southwest regions of the Arabian Peninsula. Surface wind cir-
desert). the highest drift potential is in August (fig. culation over the Arabian Peninsula is controlled in
236); at Gobabis and Mariental (west and northwest the winter by the Sahara High of the African conti-
desert). it is in September (fig. 237); and at Tsane nent and in the summer by a large low-pressure cell
and Ghanzi (eastern desert). it is in September and over the Indian subcontinent. The northern deserts
October. The greatest range of drift potential during of Arabia experience high wind energies. The most
the year is observed at the sites with the highest an- northerly regions. near Badanah. have highest drift
nual drift potentials (Upington. Ghanzi. figs. 236. potentials in the spring (April); farther southeast.
237). near Dhahran. maximum drift potentials occur dur-
ing June. Low to intermediate annual drift poten-
Arabian Sand Seas of Saudi tials may occur in the Rub' al Khali.
Arabia, Oman, Yemen, and Annual resultant drift directions in An NafUd
and Ad Dahna' trend in a broad arc from east
Adjacent States through south. exactly paralleling the trend of the
sand seas. Annual resultant drift directions in the
Summary of Conclusions Rub' al Khali are toward the east near Ash Shari-
Large compound and complex linear. crescentic. qah. toward the north along the eastern margin. and
star. and dome-shaped dunes are observed on Land- toward the southwest in the western part of the
sat imagery of Arabian sand seas. Dunes of the Rub' desert.
al Khali. in the southern part of the Arabian Penin- Unimodal. bimodal. and complex wind distribu-
sula. are distributed according to type into three tions in An NafUd and Ad Dahna' are roughly com-
regions - crescentic dunes in the northeast. star patible with the variety of dune types in the region.
dunes along the eastern and southern margins. and Wind data for the Rub' al Khali are incomplete. but
linear dunes throughout the western half. the unimodal effective wind distribution at AI

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 363

'Ubaylah is compatible with the crescentic dunes 1960, p. 157). The growth of dunes to as much as
observed near the station. 300 m (1.000 ft) high in the Arabian sand seas, such
as those described by Brown, Layne, Goudarzi. and
Introduction MacLean (1963). may require thousands of years (I.
G. Wilson, 1971a, p. 183). A Pleistocene age for the
Sand seas with very large dunes extend over large dunes of the Al 'Uriiq al Mu'tari9ah, in the
795,000 km2 (477,000 mi2) of the Arabian Peninsula eastern Rub' al Khali, is suggested by Glennie
(I. G. Wilson, 1973. table 1). Main sand seas are the (1970, p. 96).
Rub' al Khali and Wahiba Sands in the south and Two forms of crescentic dunes were recognized
the An Nafud. Ad Oahna'. and numerous smaller in the eastern Rub' al Khali (fig. 238) and described
sand seas in the north. Landsat imagery has been by Glennie (1970, p. 88-89) as "predominantly
used to map the distribution and morphology of barchan dunes. both simple and complex," and as
eolian sand in the southern part of the Arabian "giant sand massifs in various forms, often of com-
Peninsula (fig. 238) and to illustrate the varieties of plex barchanlike structure with their axes trans-
dunes and their distribution in the northern sand verse to the dominant winds." The latter dunes are
seas. Climatic data are shown by precipitation termed "giant crescentic massifs" by Holm (1960, p.
isohyets (fig. 238) and sand roses (figs. 238. 243). 1372, fig. 5) and are herein referred to as "com-
Geologic quadrangle maps of the U.S. Geological pound crescentic dunes," using the classification
Survey published between 1956 and 1963 described in chapter J. (See also chapter J. table 27.)
(Bramkamp and others. 1956; Bramkamp and Compound linear dunes of a "hooked" variety
Ramirez. 1958. 1959a. b. 1960. 1963; Bramkamp. are recognized in the north-central Rub' al Khali
Gierhart, and others. 1963; Bramkamp. Ramirez, (Holm, 1960. p. 1371-1372; 1968. fig. 3). Linear
and others, 1963; Ramirez and others, 1963a, b; and dunes 100 m (330 ft) high, 20 -200 km (12 -125 mil
Brown and others, 1963) show the general distribu- long. and 1-2 km (0.6-1.2 mil wide, separated by
tion and morphology of the sand. Names used flat interdune corridors, are described in the
herein conform in spelling to usage on the U.S. Qa'amiyat region of the southwestern Rub' al Khali
Geological Survey map of the Arabian Peninsula by Bunker (1953. p. 428-429) and by Holm (1960. p.
(U.S. Geological Survey, 1963). Two index maps 1373). These dunes. which have subsidiary linear
show distribution of dune areas (Holm. 1960, fig. 1; ridges oblique and parallel to the main ridge (Holm,
1968, fig. 3), and three maps show eolian landforms 1960. fig. 7). are classified as compound linear
in the eastern part of the Rub' al Khali (Glennie, dunes (chapter J. tables 25. 26. 36).
1970, enclosures 1-3). Other useful maps are the Star dunes are recognized in the Ramlat as
U.S. Department of Commerce, U.S. Air Force ,
Sahama at the southeastern margin of the Rub' al
Operational Navigation Charts J- 6, J- 7, H - 5. Khali by Thesiger (in Bagnold. 1951. fig. 4). Bey-
and H -6, at a scale of 1:1,000,000. doun (1966. p. H5). and Glennie (1970. p. 87. figs.
70-71). "Pyramidal" (star) dunes in the Rub' al
The Rub' al Khali Khali and in the northern sand seas are described
Review of Previous Work by Holm (1960. p. 1371. 1373). Star dunes on the
edge of the 'Irq as Subay' east of ~alim (about lat
Much of southern Arabia is a structural basin
2230' N.. long 43 E.) were sectioned and sampled
with its axis plunging gently northeastward into the by McKee (1966).
Persian Gulf (Powers and others, 1966, p. 0106).
The Rub' al Khali sand sea (fig. 238) covers about Interpretation of Landsat Imagery
560,000 km2 (216,000 mi2) of the region (I. G.
Wilson. 1973, p. 86). It extends from the United Landsat imagery and Skylab photographs show
Arab Emirates westward almost 1,500 km (930 mil that dunes in the Rub' al Khali are distributed by
to the foothills of the Yemen Mountains. Sources of type into three principal areas: (1) the northeast. Al
eolian sand in the Rub' al Khali are unknown, 'Uruq al Mu'taridah area. characterized by crescen-
although some authors believe that the sand has tic dunes. (2) th~ eastern and southern margins of
been transported by wind from alluvial (wadi) sedi- the sand sea. characterized by star dunes. and (3)
ments that were eroded from Paleozoic and the western half of the sand sea. consisting mainly
Mesozoic sandstones (Holm. 1960, p. 1370; Brown, of linear dunes. The great extent of linear dunes

Digitized by Google
 364 A STUDY OF GLOBAL SAND SEAS

'SO

50 100 150 KILOMETRES

I I I I I
I
o 25 50 75 MILES

DISTRIBUTION AND MORPHOLOGY OF EOLIAN SAND in Saudi Arabia, Oman, Yemen, and adjacent States of the southem Ara-
higher in elevated

westward - from about long 50 E., to the Jabal are some of the largest crescentic dunes observed in
Tuwayq -AI 'Ari~ escarpment at about long 457' this study (chapter J, tables 38, 39). These crescen-
E., is apparent in the regional oblique view on a tic dunes are composed of curved segments with
Skylab photograph (fig. 239). slipfaces that have a mean width (hom to horn) of
The northeastern Rub' al Khali includes the Al 2.8 km (1.7 mil (fig. 240). The dunes have a mean
'Uriiq al Mu'taridah
, sand sea (fig. 238), in which length of 2.1 km (1 .3 mil and are spaced an average

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 365

EXPlANATION

LINEAR DUNES
N Simple, short

/
Mostly simple, some compound

Compound feathered

Complex
With ster dunes

Feathered, with ster dunes

euperimpoeed
With crescentic dunes
superimposed
CRESCENTIC DUNES
Simple berchenoid ridges
Compound
Berchenoid end short ridges
Berchenoid ridges
Complex crescentic ridges
with ster dunes superimpoeed

8."'
STAR DUNES
/ ,- ,
Simple end some compound
" ', f.

... ~ .~ ,
Complex sters with dome-sheped
.~ ' , ,,~ . ""\ .

m
dunes overlepping
UNDIFFERENTIATED COMPLEX
DUNES

SHEETS AND STREAKS
BEDROCK OUTCROP OR BARRIER
ROCK OR GRAVEL PLAIN WITH
NO LARGE SAND DUNES

D PLAYA

- 1 0 0 - ISOHYET-Averege ennuel reinfall,

in millimetres
- - - BOUNDARY OF SAND SEA

INTERMITTENT DRAINAG~howing
direction of flow

MAP lOCAT1CJ>l

bian Peninsula. Map based on landsat imagery. Isohyets estimated from vegetation signature and station records. Precipitation is
areas. (Fig. 238.)

of 2.6 km (1.6 mil from crest to crest. Their gentler elsewhere (chapter J, fig. 175 and tables 38, 39).
(stoss) slopes are covered by smaller crescentic Toward the east and south, in Al 'Uruq al
dunes, as shown in chapter J, fig. 174B. Thus, they Mu'tari~ah, these compound crescentic dune ridges
are compound barchanoid or crescentic forms simi- grade into complex ridges with star dunes on their
lar in size and pattern to dunes in the Peoples crests (McKee and Breed, 1976, fig. 56; chapter J, fig.
Republic of China, the western Sahara, and 174).

Digitized by Google
47"00 '

+
RED LINEAR DUNES EXTENDING 600 km (360 mi) across the westem Rub' al Khali to the foothills of Yemen. Skylab oblique
photograph Sl4-143-4643. Dunes in the Qa' amiyat region have an average crest-to-crest distance of 2.1 km (1.3 mi) and are com-
monly more than 100 km (60 mi) long. (Fig. 239.)

Dunes at the southeastern margin of the Rub' al ward-facing crescentic slipface segments of barcha-
Khali. along the border of Saudi Arabia with Oman noid ridge dunes in AI 'Uriiq al Mu'tari<!ah. The
(fig. 238) are mainly star dunes and isolated cres- southern barchanoid ridge dunes are white on
centic dunes (very wide barchans or very short false-color imagery in contrast with the yellow of
"transverse" ridges). The isolated crescentic dunes the dunes farther north. Their different size. orien-
in Oman at about lat 20 N.. long 55 E. have slip- tation. and color suggest that these dunes in the
faces that face north. opposite those of the south- southeastern part of the Rub' al Khali may be ac-

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 367

tively migrating northward under the influence of

strong southwest monsoon winds (fig. 238).
Star dunes along the eastern and southern
margins of the Rub' al Khali, mostly in Oman and
Yemen, are large isolated compound forms, shown
in detail on an aerial photograph (chapter J, fig.
179). The star dunes seem to "fan out" from the
mouths of intermittent streams which flow north-
ward from the southern highlands into the Rub' al
Khali. Star dunes are larger and redder in a north-
ward direction, where they merge with barchanoid
dune ridges, forming complex dunes that have
southeast-facing slipfaces (McKee and Breed, 1976,
fig. 56). This relationship suggests that some sand
in the star dunes along the southern margin of the
Rub' al Khali may be derived from the wadi
alluvium with which they seem, on the imagery, to
be associated.
West of AI 'Uruq al Mu'tari~ah region of mainly
crescentic dunes is a sand sea of dominantly linear
dunes. At the boundary between the two major
dune types (lat 20 N., long 53 E.) is a region of
0 0

complex dunes that have both linear and crescentic

elements (fig. 241). These include the "hooked"
NORTHEASTERN RUB' Al KHAlI, AI Jiwa - AI Batin area. land- dunes described by Holm (1960, p. 1371-1372; 1968,
sat imagery E1130- 06251 . Groups of small crescentic dune
ridges, in the area marked by smoke plumes from oil flares, seem
fig. 3). Very large compound linear dunes extend
to be migrating southward and overriding much larger com- west of the region of complex dunes for hundreds
pound
. ,
crescentic dunes of the northern AI 'Uruq al Mu'taridah. of kilometres, through the Qa'amiyat region to
(Fig. 240.) foothills in Yemen (fig. 239) . At the downwind end

NORTH-CENTRAL PART OF THE RUB' Al KHAlI, west of AI 'Uruq al Mu'tari~ah . landsat imagery E1166- 06254. Dune patterns
hange from large crescentic dunes in the east to linear (seif) dunes in the west. In the zone west of the compound crescentic dunes
are smaller crescentic ridges. Complex combinations of linear and crescentic dunes ("hooked dunes") occur in the center of the
area. (Fig. 241 .)

Digitized by Google
368 A STUDY OF GLOBAL SAND SEAS

of this sand sea. at the Jabal Tuwayq are the largest

linear dunes seen on Landsat imagery during the
present study (fig. 239; chapter J. fig. 169F and
tables 36. 37). They average 1.5 km (0.9 mil in
width and are spaced an average of 3.2 km (2 mil
apart. Sample area
Measurements in the sand sea of linear dunes
were made on eight 2.500-km2 (1.500-mi2) sample
areas of images. At the upwind end of the field.
there were 25 ridges per 50-km (30-mi) distance. rroo'
"-
The widths of the dunes and interdunes corridors III

are nearly equal. averaging 2 and 2.1 km (1 .3 mil. J

respectively. At the downwind end of the field.
dunes are fewer (18 per 50-km (30-mi) distance)
and narrower (0.7 km (0.4 mi. as opposed to wider
interdune areas (2.8 km (1 .8 mi. Although the up-
wind and downwind distributions of dunes thus
differ. a plot of the measurements of all the samples
against their distance downwind (expressed as a
percentage of the total length of the field) shows no
consistent trend.

The Wahiba Sands

Review of Previous Work
The Wahiba Sands is a sand sea of approximately
16.000 km2 (10.000 mi2). east of the Rub' al Khali. in
Oman (I. G. Wilson. 1973. table 1). The large linear 40 KM
dunes of the Wahiba Sands are Pleistocene features t--""----,-"'""'"----......'I---JI
and do not reflect the trends of present-day winds. 20 MI

according to Glennie (1970. p. 90-95). The largest

dune ridges are joined by numerous subsidary SOUTHERN PART OF THE WAHIBA SANDS, Oman. landsat
ridges oblique or parallel to the main trend. as imagery E1126 - 06024. Outlined area is a sample of compound
shown on aerial photographs (Glennie. 1970. fig. linear dunes on which measurements were made. (Fig. 242.)
74) . Thus. the ridges are compound linear dunes
(chapter J. table 25) . effective winds. as suggested by the Al MasTrah
sand rose (fig. 238) . A good correspondence of
Interpretation of Landsat Imagery dimensionless ratios (Strahler. 1958. p. 292-293) of
measured mean widths and mean lengths occurs in
Compound linear dunes in the Wahiba Sands the Wahiba Sands and the Namib Desert. Mean
(fig. 242) have a mean width of 1.1 km (0.7 mil and wavelengths (spacing) of linear dunes in both
are spaced an average of 1.6 km (1 mil apart. In- deserts is dissimilar (table 44). Spacing is much
dividual ridges are as much as 100 km (60 mil long greater between the linear dunes of the Namib than
and have an average length of 31.3 km (20 mil. As between the linear dunes of the Wahiba Sands (ta-
in the Namib sand sea (fig. 223). crescentic dunes ble 44).
occur along the coast. and linear dunes are farther
inland. At the north end of both the Namib sand sea Northern Sand Seas
and the Wahiba Sands. linear dunes are truncated Review of Previous Work
along an intermittent stream (Glennie. 1970. fig.
74). Linear dunes in the Wahiba Sands may be Sand seas in the northern part of the Arabian
alined subparallel to the resultant direction of desert include An Nafud. Ad Dahna'. Nafud ath

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 369

TABLE 44 - Planimetric similarity of linear dunes of the Wahiba Sands. Oman. and linear dunes of the Namib
Desert. South -West Africa. shown by comparison of scale ratios derived from their measurements
[Subacript to sand aea indicatea firat or aecond of pair. Subacript to L, i, and X indicatea firat or aecond of aand sea pair)

Mean Mean Mean Degree of

Sand aea dune L1 /L2 dune W/W2 dune );1/X 2 L1/iil L2/ii2 L/X1 L/X 2 ii1/X1 ii/X 2 correapon-
1enjrth width wavelength dence i
L i X

wahiba sanda1-- 31.25 1.06 1.62

1.16 1.20 0.74 29.5 30.7 19.3 12.27 0.65 0.40 Good
Naaib oeaert 2 27.0 .88 2.20

i oegree of correspondence ia baaed on the maxtmu. difference in aca1e ratioa. 0.01-0.20 - Excellent, 0.21-0.39 - Very Good,
0.40-0.59 - Good, 0.60-0.79 - Fair, 0.80-0.99 - Poor.

Thuwayrit. Nafiid as Sirr. and others southwest of 1373). An Nafud may furnish some of the sand of
Ad Dahna' (fig. 243). The largest and least studied Ad Dahna' and the other northern sand seas (Holm.
is An Nafud. which covers about 72.000 km2 1960. p. 1370).
(45.000 mi2) (I. G. Wilson. 1973. table 1: this Ad Dahna' is an arc-shaped sand sea that extends
chapter. fig. 243). Dunes of this region reportedly 1.200 km (750 km) southeastward from An Nafud
reach heights of 100 m (330 ft) (Holm. 1960. p. to the northern Rub' al Khali (Holm. 1960. p. 1370.

50' !i2"

IRAQ

SAUDI ARABIA

93 Md'IIIIt
-~
365 ~I
~-----~ _ _~___~I________~______~_____~~~______-L~--~~~~~~
o 91 100 191 200 250 KlOMElRES
I I
I
I I
I
I I

o 91 100 MLES

OUTLINE MAP OF NORTHERN ARABIAN PENINSULA. Resultant drift directions are parallel to elongation of An Nafud, Ad Dahna',
and other northern sand seas, shown by stippled area (sand sea patterns modified from Holm, 1960, fig. 1). Arrows on sand roses
indicate resultant drift directions; numbers near sand roses are drift potential, OP (on bottom), resultant drift potential, ROP (on top),
and ROPIDP in parentheses (computation of sand roses is described and terms defined in chapter F and in fig. 189). Sand rose for
Ha'j( is based on computations by Bagnold (1953) and is not to scale (OP and ROP are not available for Ha'il). (Fig. 243.)
$

Digitized by Google
370 A STUDY OF GLOBAL SAND SEAS

Linear dunes
to'

\ .-
Dome dunes
'.

. I
20 MI

PART OF AD DAHNA' OF SAUDI ARABIA. landsat false-color imagery E1209- 07041 . Complex dunes composed of linear ridges
and giant dome-shaped dunes fill the valleys, illustrating topographic control of sand sea boundaries. Dune sand appears yellow in
false-color imagery. Blue-green areas are mostly sandstone and limestone bedrock, with a thin sand cover (yellow). Red areas along
stream valleys are vegetated. (Fig. 244.)

fig. 1). Star dunes in Ad Dahna' are reported to be monly vegetated and crossed by low sand ridges
150-170 m (500-560 ft) high (Holm, 1960, p. 1370). (Holm, 1953, p. 108-110).
Southwest of Ad Dahna:' are numerous elongate Dome-shaped dunes in the NafUd ath Thuwayrat
arcs of eolian sand. Unlike Ad Dahna', sands of the (fig. 243) are commonly 1-1.5 km (0.5-1 mil in
inner arcs do not extend to the Rub' al Khali. The diameter and 100-150 m (330-500 ft) high; com-
northern sand seas are dominated by giant dome- pound varieties are as much as 5 km (3 mil long
shaped dunes (chapter J. fig. 184), each separated (Holm, 1953, p. 108-109). The upper surfaces of
from the next by deflation hollows that are com- some dome-shaped dunes have crescentic dune

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 371

ridges. so they are here classed as complex dunes the Arabian Peninsula. This situation results in
(chapter J. fig. 184). northwesterly ("shamal") winds over much of the
peninsula. On the southeast coast. surface wind
Interpretation of Landsat Imagery flow into the same region of low pressure results in
southwesterly winds. From August through
Dunes in the northern and southern parts of An
November the thermal low over Asia weakens. and
Nafiid are complex types. formed of linear ridges.
the center of high pressure near the eastern part of
some with star dunes. Compound crescentic dune
northern Africa shifts from the Mediterranean Sea
ridges occur along the west (upwind) margin and in
southeastward into Libya. This is a season of weak.
the center of the sand sea. These ridges are spaced
directionally variable winds over much of the Ara-
so closely that they obscure the interdune areas.
bian Peninsula.
The crescentic-shaped hollows observed in An Average annual rainfall is less than 100 mm (3.9
Nafud by Bagnold (1951. p. 85) and by Holm (1960. in.) at most stations near the An Nafud. Ad Dahna',
p. 1370. fig. 3) may be the slipfaces of these nearly and the Rub' al Khali (figs. 238. 243). Only Ha'il
overlapping dunes. (102 mm or 4 in.) and ~lalah (106 mm, or 4.1' in.)
Broad sheets of sand and linear dunes are charac- within the regions shown in figures 238 and 243
teristic of Ad Dahna' and the northern sand seas. have recorded more than 100 mm (3.9 in.) average
north of the valleys of Wadi Rima' and Wadi al annual rainfall. based on available data. Vegetation
Bitin (fig. 244). At the junction of the sand-filled (red on Landsat false-color imagery) is visible only
valleys of these intermittent streams. linear dunes in the sand regions near oases. From north to south
grade into dome-shaped dunes of the Nafud ath across the peninsula, average annual rainfall is 41
Thuwayrat. NafUd as Sirr. Nafud Shuqayyiqah. and
mm (1.6 in.) at Badanah. 35 mm (1.3 in.) at Raflia'.
the southern part of Ad Dahna'. Dome-shaped 98 mm (3.8 in.) at Dhahran, 84 mm (3.3 in.) at Ar
dunes in the Nafiid ath Thuwayrat. measured on
Riyan, and 58 mm (2.3 in.) at Al MasTrah , (U.S.
sample areas of Landsat imagery have a mean Naval Weather Service, 1974b).
diameter of 1.2 km (0.8 mil. They are the largest
dunes of this type observed during the study
(chapter J. table 42). The sand seas in which they Direction of Sand Drift
occur fill valleys that are bounded on both east and Annual resultant drift directions in the An Nafiid
west by sedimentary rock. Dome-shaped dunes in and Ad Dahna' are toward the east near Turayf and
these sand seas increase in size from west to east in Badanah and toward the south-southeast near
a downwind direction across each valley. Dhahran and Ar Riyan (fig. 243). The curvature of
annual resultant drift direction in the northern
Surface Wind Flow and Precipitation regions closely parallels trends of the An Nafiid, Ad
Surface wind circulation on the Arabian Penin- Dahna', and other sand seas. The curvature resem-
bles in scale and orientation the clockwise trend of
sula is mainly controlled during winter by the
annual resultant drift directions in Algeria (fig.
Sahara High of the African continent. and during
206).
summer by a large low-pressure cell on the Indian
subcontinent. In winter. clockwise flow of air from Monthly resultant drift directions in the An
the Sahara High and its extension over Arabia pro- Nafud are usually eastward during the winter
duces westward surface wind flow in northern months but shift toward the south in the summer.
Arabia and south-southwestward flow in the Rub' For example, monthly resultant drift directions at
al Khali (fig. 245A). During the spring and summer. Badanah (fig. 246) are toward the east from Decem-
the center of high pressure over eastern North ber through April but are toward the southeast
Africa shifts northwestward to approximately lat from June through October. At Dhahran, on the
35 N . long 15 E. (over the Mediterranean Sea; fig. Persian Gulf, the resultant drift direction is toward
245B). Also during this time. low pressure deepens the southeast most of the year but shifts to the
over Asia and. by summer. extends as a trough over south during March -May (fig. 247).

Digitized by Google
372 A STUDY OF GLOBAL SAND SEAS

SEA-LEVEl MEAN-PRESSURE CONTOURS (in millibars) on the Arabian Peninsula and adjacent
regions for (A) January and (B) June. Average trends in surface wind flow indicated by arrows. L, area
of low air pressure; H, area of high pressure. Trough of low pressure over Arabian Peninsula in June
indicated by dashed line. Modified from Crutcher and Meserve (1970). (Fig. 245.).

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 373

{)P=39 {)P=l00
N FEB. APR.

t DP=19
JAN.
{)P=28
FEB.
DP=33
MAR.
DP=37
APR.
N

t
{)P=32
AUG.

{)P=53 {)P=90 {)P=48

MAY JUNE JULY

{)P=46 DP=22
OCT. DEC.

SAND ROSES FOR 6 MONTHS FOR BADANAH, Saudi Arabia.

Resultant drift direction (arrow) is generally toward the east
from December through April, and toward the southeast from
June through October. Drift potential, in vector units, is given
DP=21 {)P=11 DP=12 DP=19
for each month. (Fig. 246.) SEPT. OCT. NOV. DEC.

Wind distribution in the An Nafud -northern Ad

SAND ROSES FOR 12 MONTHS FOR DHAHRAN, Saudi Arabia.
Dahna' region (fig. 243) range from unimodal
Compare with sand roses for Badanah (fig. 246). The wind
(Dhahran) to bimodal (Badanah) to complex regime at Dhahran is more directionally steady from month to
(RafIJii'). This range of wind regimes accounts for month than at Badanah; wind distribution is usually unimodal
the wide variety of dune forms in the region (fig. at Dhahran but may be bimodal or complex at Badanah. Resul-
244). tant drift directions (arrows) are more toward the south at
Annual resultant drift directions in the Rub' al Dhahran than at Badanah. Drift potential, in vector units, is
given for each month. (Fig. 247.)
Khali (fig. 238) are toward the south-southwest
near Al 'Ubaylah and toward the east along the
coast of the Trucial States near Ash Shariqah. vary little from month to month at Al 'Ubaylah, in
Along the southeastern part of this coast, the resul- the northern Rub' al Khali (fig. 249). This suggests
tant drift direction seems to be generally toward the that effective winds farther downwind from Al
north, based on data from the island station of Al 'Ubaylah in the western Rub' al Khali, in the region
MasTrah. of linear dunes, may also be relatively steady in
At Ash Shariqah, on the coast of the Trucial direction. The eastern and central parts of the Rub'
States in the northermost Rub' al Khali, resultant al Khali are characterized by crescentic and linear
drift directions are toward the east or southeast dunes (fig. 240) , which are often associated with
most of the year (fig. 248). Greatest directional narrow wind distributions (chapter F, fig. 105). Star
variability of effective winds occurs during the late dunes occur along the southern margin of the Rub'
summer through early fall (August and September, al Khali, a region which may be subject to wind
fig. 248) at Ash Shariqah. Resultant drift directions flow from the southwest during the summer.

Digitized by Google
374 A STUDY OF GLOBAL SAND SEAS

DP-30 DP-48
FEB. APR.
N

t DP-23
JAN.
DP-28
FEB.
DP-53
MAR.

t
DP-33
MAY
DP-17
AUG.

/ DP-32
APR.
DP-30
JUNE

~
..

DP-13 DP-17
SEPT. DEC.

SAND ROSES FOR 6 MONTHS FOR ASH SHARIQAH, Arabia.

Resultant drift direction (arrow) is generally toward the east
from December through May; toward the southeast from
August through September. late summer through early fall is the DP-<1 DP-7 DP-7
season of greatest directional variability of effective winds OCT. NOV. DEC.
(August sand rose) and is also the season of lowest drift poten-
tials. Drift potential, in vector units, is given for each month.
(Fig. 248.) SAND ROSES FOR 9 MONTHS FOR Al 'UBAYlAH, Arabia, in
the Rub' al Khali sand sea. Sand roses are not available for July,
Amount of Sand Drift August, and September. Monthly resultant drift directions (ar-
rows) are relatively steady toward the south or southeast for
The average drift potential of the northern sand most of the year. Resultant drift directions at AI 'Ubaylah con-
seas (wind stations shown in figure 243) is 489VU, a tinue the trend of a broad tradewind arc suggested by data from
high value compared to that of most other desert stations in the An NafUd farther north (fig. 243). Surface wind
data for sand roses published by permission of the Arabian
regions around the world (chapter F, table 16).
American Oil Co. Drift potential, in vector units, is given for
Highest drift potentials at stations in An Nafnd, Ad each month. (Fig. 249.)
Dahna', and the Rub' al Khali may occur in the
spring or early summer, depending upon location. less skies (ARAMCO Survey, 1974, p. 2). In the
At many stations along the Persian Gulf, the season northern sand seas (including the An Nafiid), high-
of high drift potentials is associated with strong est drift potentials occur in April at Badanah, in
north or northwesterly winds that are locally May at AI Qaniimah to the southeast, and in June
known as the shamal winds. During that season the at Dhahran still farther to the southeast (fig. 250).
wind may blow for 2 or 3 days at 22 -26 knots Much of the potential sand movement in the north-
(41-48 km/hr) , then diminish to 13-17 knots ern sand seas occurs during sharply defined
(24-31 km/hr) for several more days under cloud- seasons, rather than distributed evenly throughout

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 375

the year. In the Rub' al Khali, highest drift poten-

type of eolian landform in the Thar Desert. Clusters
tials at Ash Shiriqah and AI 'Ubaylah occur in of these dunes average 2.6 km (1.6 mil long and 2.4
March (fig. 251). At AI Ma,irah, the highest drift
km (1.5 mil wide. Parabolic dunes extend from the
Rann of Kutch northward and eastward to Bikaner
potentials occur in July (299 VU) at the peak of the
southwest monsoon winds. The drift potential at AI
and Jodhpur and occupy an area of at least 100,000
'Ubaylah, approximately 200 VV, suggests that the
km z (60,000 mi2). Smaller, apparently simple cres-
Rub' al Khali may not be as windy as An NafUd. centic dunes and short linear dunes occur in small
fields surrounded by the parabolic dunes.
The Thar Desert of India and wavelength Compound crescentic dunes with an average
of 1.3 km (0.8 mil occur in a broad band
Pakistan along the northwest margin of the Thar Desert,
Summary of Conclusions from approximately lat 2640' N., to lat 2940' N.
Eroded, heavily vegetated crescentic dunes occur
Vegetated compound parabolic dunes of a dis- east of Jodhpur and on the Gujariit Plain in fields
tinctive rakelike shape are the most widespread oriented parallel to the eastern shore of the Rann of
Kutch. Linear dunes, some of which may be com-
pound forms, are oriented northeast-southwest
100
I ,
, along both the east and the west margins of the
,
I '-"Blldaflh
,, I
desert. Sand streaks and sheets, similar in ap-
pearance to those observed on Landsat imagery of
,, /\
80
,, \
, I\
,
, other deserts, are oriented approximately north-
east-southwest along the east margin of the sand
sea near Jodhpur and Barmer.
,, I \ Average annual rainfall in the Thar Desert
,, I yAI QayfOmah ranges from less than 100 mm (3.9 in.) along the
~
z
,I \
;:) \
\
~~ ~I
60 Dhahfln

~

1\
I \

~ I
I \

~
I \ ..... AI 'Ubayllih
I
I \

~ ~40 l \

Ii;
~
~
\
\
,
15
~
~20
Ii;
15
,
,,
~--

oJ D
D
MONTHS OF THE YEAR
MONTHS OF THE YEAR
DRIFT POTENTIAL DURING THE YEAR at Ash Shariqah and AI
DRIFT POTENTIAL DURING THE YEAR at 8adanah, AI 'Ubaylah, Arabia. Location of stations shown in figure 238. 80th
Qaysumah, and Dhahran, Arabia. Location of stations shown in stations probably experience highest drift potentials in March.
figure 243. Season of highest drift potentials occurs early in the Monthly drift potentials are not available for AI 'Ubaylah for July
year in the northwestern An Naflid (8adanah) but later at through September. Data for AI 'Ubaylah published by permis-
locatities closer to the Persian Gulf (Dhahran). (Fig. 250.) sion of the Arabian American Oil Co. (Fig. 251.)

Digitized by Google
376 A STUDY OF GLOBAL SAND SEAS

northwest margin of the desert to more than 400 Pakistani studies, mainly for agricultural and sand-
mm (15.8 in.) along the southeast margin. Present- control purposes, are numerous. Recent geological
day effective winds in the Thar Desert are mainly fieldwork by Western scientists has been limited to
from the west-southwest or southwest (southwest the east margin of the desert, along the Aravalli
monsoon). These winds result from air flow into a Range (Goudie and others, 1973), and in the Rann
thermal low-pressure cell that develops over the In- of Kutch (Glennie and Evans, 1976) , and to the
dian subcontinent during the summer. Average an- Porali Plain west of the main dune areas, near
nual drift potentials are highest in the south and Karachi {Verstappen, 1966, 1968a; Snead and Frish-
decrease northward; however, relatively little wind man, 1968}. Modern geomorphological studies of
energy occurs at present in the Thar Desert com- the main dune area (Verstappen, 1968a, 1970) are
pared to other desert regions. Highest drift poten- based mostly on interpretations of aerial photo-
tials at most stations in the Thar Desert occur in graphs. Useful maps are the U.S. Department of
June, usually preceding the time of maximum rain- Commerce, U.S. Air Force Operational Navigation
fall by 1 or 2 months. Charts H -8 and J -8; and U.S. Army Map Service,
Present-day low-energy unimodal wind regimes Series 1301, Sheets NF 42. NG 42, NH 43, NG 43,
are roughly compatible with the parabolic, crescen- and NH 42; all maps are at a scale of 1:1,000,000.
tic, and linear dune morphologies in the Thar Landsat imagery has been used to map regional
Desert. Parabolic dunes and sand streaks aline with distribution and morphology of eolian sand in the
resultant drift directions of the present-day winds. Thar Desert. The map (fig. 252) shows eolian sand
The compound parabolic dunes, which are the most in relation to precipitation isohyets (in millimetres
widespread and distinctive feature of the sand sea, per year) and potentially effective sand-moving
may have formed several thousand years ago and winds.
not in the relatively low-energy wind regime of to-
day. The parabolic dunes occur in a region of ap- Review of Previous Work
proximately 200 mm (7.9 in.) average annual rain- Thirty-four samples of sand from fixed and ac-
fall and are mostly vegetated. Ground studies by tive dunes in the eastern part of Rajasthan were
other workers suggest that surfaces of these dunes collected and analyzed by Goudie, Allchin, and
are weakly cemented. However, less vegetated Hedge (1973, p. 247 -248). The samples had an
crescentic dunes in numerous places within the average median grain size of 0.148 mm, compared
sand sea, surrounded by vegetated linear and with the median diameters of 0.1-0.15 mm for sand
parabolic dunes, may be active in the present wind in the central Gujarit Plain, 0.125 mm in Punjab,
regime. and approximately 0.15 mm for sand in western Ra-
jasthan. A mean diameter of 0.14-0.15 mm was
Introduction determined for sand from the Porili Plain in
The Thar Desert extends westward in north- southern Pakistan (Goudie and others, 1973, p. 248).
western India from the Aravalli Range to the vege- The sand in general is well rounded (Wadia, 1939,
tated flood plain of the Indus River and from the p.291}.
Rann of Kutch northeastward into the State of Pun- Sand of the Thar Desert is composed mostly of
jab, India (fig. 252). The southern and western parts quartz, with some hornblende and feldspar. It in-
of the desert lie mostly in Pakistan. Physio- cludes calcareous grains, mainly foraminifer tests
graphically, the region consists of low hills and (Wadia, 1939, p. 291), and salt particles (Qadri,
sand dunes lying on sandy alluvium that is under- 1957, p. 169). Sand of the Porali Plain (Verstappen,
lain by caliche-cemented concretionary beds. Few 1968a, p. 208) also is mainly quartz but has a high
topographic barriers interrupt the sea of sand, ex- CaC03 content, due in part to the presence of
cept for scattered outcrops of limestone and foraminifer and ostracode remains. Fixed dunes of
sandstone of Jurassic age near Jaisalmer (fig. 252). the region are said to commonly have a caliche
Early work in the Thar Desert was by British crust.
geographers (Frere. 1870; Blanford. 1877). The The sands are derived from numerous sources in-
geology of India is described by Wadia {1939}. cluding deflating flood plain deposits, especially
Auden {1950}. and Krishnan (1960). Indian and along the Indus River Valley. weathered materials

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 377

from the Aravalli Range, and sand of former of 15-20 parabolic dunes that range in length from
shorelines (Seth, 1963, p. 449). Depth of sand in the 1 to 3 km (0.6-1.9 mil and attain heights of 30-70
Thar Desert is several metres (Krishnan, 1960, p. m (100-230 ft) have been described by Singh,
514). Horizontally bedded fine-grained slightly in- Ghose, and Vats (1972, p. 52). These parabolic
durated sandstone which "exactly resembles the dunes extend for great distances (as much as 6 km
windblown sand of the dunes" is reported to occur or 4 mi). Their arms are nearly parallel and more
at depths greater than 100 m (330 ft) in wells along uniformly spaced than in those dunes to the south.
a road connecting Karachi, Hyderabad, Barmer, and At about lat 26 N., remnants of parabolic dunes are
0

Jodhpur (Seth, 1963, p. 449). only faintly recognizable on aerial photographs,

Old, dissected dunes of the Gujarit Plain, in the and fixed longitudinal dunes predominate
southeastern part of the desert, are dark brownish (Verstappen, 1968&, p. 210-211).
red (7.5YR 5/8) (Goddard and others, 1948) and The fixed longitudinal dunes of the second
they protrude through white, fixed dunes (Goudie CAZRI category in the Thar Desert were observed
and others, 1973, p. 247-248). Sands in all types of by Verstappen (1968&, p. 202-203) to have the same
dunes on the PoraU Plain in Pakistan are light N. 55 E. orientation as dunes of similar type north
0

yellowish brown (10YR 6/4) (Goddard and others, of the MIanI Har Lagoon on the Poriili Plain (fig.
1948; Verstappen, 1968a, p. 204). Old fixed dunes in 206), which are 20-50 m (65-165 ft) high, 90-350
the Thar Desert are vegetated with perennial and m (300-1,150 ft) wide, and as much as 5 km (3 mil
annual grasses, shrubs, mainly Calligonum long. The fixed longitudinal dunes of the Thar, like
polygonoides and Haloxylon salicornicum, and low those of the Por81i Plain, are believed by Verstap-
scattered trees, especially Acacia sp. and Prosopis pen (l968a, p. 210, 218), to have been derived by
sp. (Roy, 1969, p. 16-17). blowing out of the noses of parabolic dunes. Both
Scientists at the Central Arid Zone Research In- parabolic and longitudinal dunes of the Poriili Plain
stitute (CAZRI) at Jodhpur, India have established (and by analogy, dunes of the second CAZRI
three categories of eolian sand bodies in the Thar category in the Thar) are tentatively assigned by
Desert, based on field determinations of relative age Verstappen (1966, p. 7; 1968&, p. 203-204) to the
and stabilization. These categories are (1) old dis- Atlanticum Period, circa 5,000 years before the
sected dunes of indeterminate type and sand present.
"shields," (2) stabilized parabolic, longitudinal, and Areas of active eolian deposition of the third
transverse dunes, and (3) active small-scale CAZRI category include the desert between
barchan, shrub-coppice, and low-longitudinal ridge Phalodi and Jodhpur, between Barmer and
dunes (Singh and others, 1972). Jaisalmer, between Loharu and Bhiwini, and along
Old well-vegetated dissected dunes and sand the courses of the Sutlej and old Ghaggar Rivers
shields (sheets?) of the first category occur in the (Seth, 1963, p. 449-450). The present northeast-
eastern part of the Thar Desert in the Jodhpur area, ward movement of sand and the concurrent
along the western slopes of the Aravalli Range, and development of active small-scale barchans, low
on the Gujarat Plain of India (Singh and others, longitudinal ridge dunes, and shrub-coppice dunes
1972, p. 51-52). Middle Stone Age artifacts are are attributed by most workers to the erosion of
associated with these dunes which are thought to sand from fixed dunes which have been overgrazed
have formed under conditions of aridity more ex- or subjected to farming (Khalil, 1957, p. 241; Roy,
tensive than those of the present (Goudie and 1969, p. 16-17; Chaudri, 1957, p. 144; Christian,
others, 1973, p. 243, 249, 252-255). 1959, p. 7 -8). On the Porali Plain, and in the Thar
Fixed parabolic dunes of the second category Desert, the only dunes presently forming, by reac-
northwest of Mandar, India, at about lat 25 N. (fig. tivation of the parabolic dunes, are barchans
0

252), are characteristically 35-70 m (115-230 ft) (Verstappen, 1968, p. 7; 1968a, p. 203, 210-211).
high but may be as much as 150 m ( 490 ft) high.
They average 2.5 km (1.5 mil long, are variably Interpretation of Landsat Imagery
spaced about 250 m (820 ft) apart, and are oriented Some of the dunes referred to in the literature
N. 50 -55 E. (Verstappen, 1968a, p.210). Farther have been observed on Landsat imagery of the Thar
0 0

inland near Jodhpur, Bikaner, and Jaisalmer, groups Desert (fig. 252). Eroded and heavily vegetated

Digitized by Google
378 A STUDY OF GLOBAL SAND SEAS

Ganganagar ,
3 -
~

a 50 100
I
I I I I I
a 25 50 75 MILES MAP LOCATION

DISTRIBUTION AND MORPHOLOGY OF EOLIAN SAND in the Thar Desert of India and Pakistan (above and facing page). Map
based on Landsat (ERTS) imagery. Isohyets are modified from Rao (1958) and from Pramanik (1952). (Fig. 252.)

crescentic dunes visible on images of the Gujarat Verstappen (1968a) are clearly visible in figure 253.
Plain may be some of the old, dissected dunes of the Dunes on the Porali Plain in Pakistan, however. are
first CAZRI category (also described by Goudie and not visible on Landsat imagery.
others, 1973). They are in a region that now Compound parabolic dunes with many arms in
receives 400-500 mm (16-20 in.) rainfall annually. rake-shaped clusters have been observed during
The old dunes are partly covered by vegetated sand this study only in the Thar Desert. Parabolic dunes
streaks. Clusters of compound parabolic dunes observed in the other sand seas studied are
described by Singh, Ghose, and Vats (1972), and by U-shaped, with long trailing arms (chapter J, fig.

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 379

EXPLANATION
LINEAR DUNES

D Simple, short

Cd Mostly simple, some compound

.....
CRESCENTIC DUNES
~" O
,"
Barchanoid ridges
\: ;
Compound barchanoid ridges
..., ''I. '
. ~: ..': COMPOUND PARABOLIC DUNES

SHEETS AND STREAKS

Simple
Compound, with linear elements

BEDROCK OUTCROP OR BARRIER

DESERT PLAIN WITH NO LARGE DUNES

PLAYA
VEGETATED PLAIN

WIND ROSE-Length of arm is proportional to

~;f
percentage of wind observations from a given
direction. Percentage of observations given where
too long for graphic portrayal

- .100 - ISOHYET-Average annual rainfall,

in millimetres
- - - BOUNDARY OF SAND SEA
. INTERMITTENT DRAINAGE-Showing direction
of flow
(Fig. 252. -Continued.)

183) . The parabolic dunes can be discerned on im-

agery from the north edge of the Rann of Kutch as
far north as Bikaner (fig. 252) and are the dominant
dune type of the Thar Desert. In the southernmost
Thar. along the Rann of Kutch. dune clusters are
massed (fig. 253). but farther inland, the clusters
are an average of 2.6 km (1.6 mil long and 2.4 km PARABOLIC DUNE ClUSTERS along the southern margin of the
(1.5 mil wide, have an average of seven arms, and. Thar Desert. landsat imagery El137 - 05214. (Fig. 253.)
although variably spaced, are an average of 2.3 km
(1.4 mil apart. The rakelike arrangement of the fields of crescentic dunes surrounded by the linear
crests of parabolic dune arms is apparent in a and parabolic dunes (fig. 255) showed no evidence
sketch by A. S. Goudie (written commun. to of vegetation. The zone of crescentic dunes shown
McKee, 1974) (fig. 254). in figure 255 includes an area sketched from aerial
Comparison of imagery made during the dry photographs by Verstappen (l968a, fig. 3. right) and
season with that made after the onset of the mon- described by as him containing active barchans. It
soon rains confirms that the linear dunes. and the probably represents a major area of modern eolian
parabolic dunes with which they are associated15 activity. The crescentic dunes. whose curved seg-
are seasonally covered by vegetation. However, ments are too small to measure on the imagery,
have wavelengths of as much as 0.5 km (0.3 mi).
The vegetated linear dunes shown in figure 255
IS NOTE. - The problem of distinguishing linear dunes from detached arms of
parabolic dunes on the basis of morphology alone is great. Parabolic arms have may be some of the "stabilized longitudinal dunes"
linear form and develop as parallel sand ridges but are structurally very different of the second CAZRI category. believed by Verstap-
from linear dunes (Chapter A. table 1: chapter E. table 9). Subsequent correct
genetic interpretation of the dune field depends on type of dune identified. pen (1968a, fig. 3. right) to have formed from
E. D. McKee. editor. associated parabolic dunes.

Digitized by Google
380 A STUDY OF GLOBAL SAND SEAS

70"00'

RAKElIKE ARRANGEMENT OF CRESTS of parabolic dune arms

comprising compound forms near Jodhpur, India (Iat 2630' N.,
long 725' E.). Sketch traced from a survey map of India based
on aerial photography (from A. S. Goudie, written commun. to
McKee, 1974). (Fig. 254.)

In the western part of the Thar Desert adjacent to

the Indus River flood plain, in a region that
generally receives less than 100 mm (3.9 in.) annual VEGETATED PARABOLIC AND LINEAR DUNES surrounding
rainfall, are fields of large dunes here interpreted as fields of barren crescentic dunes in the southwestern part of the
linear and crescentic types (figs. 252,256). The lat- Thar Desert. landsat imagery E1137-o5212. (Fig. 255.)
ter resemble the compound crescentic ridges com-
mon to many other sand seas described in this In the eastern part of the desert near Jodhpur and
chapter and compared in chapter J (tables 38, 39). Barmer are sheets and streaks of sand which appear
Parabolic dunes grade westward into these large as yellow and brown on false-color imagery in-
dunes, but generally dunes of the parabolic type are dicating that the sand is actually red and partly
not observed on imagery of the more arid, western vegetated.
part of the desert (fig. 252). Segments of compound
crescentic dunes, measured on imagery (fig. 256), Surface Wind Flow and Precipitation
average 1.5 km (1 mil from horn to hom, have an
average wavelength of 1.3 km (0.8 mil, and their Surface wind flow in the Thar Desert is from the
avalanche slopes, which face N. 5-10 W., have a southwest during the summer and from the north-
more westerly trend than do the northeast-facing east during the winter. The summer southwesterly
active crescentic dunes on imagery of the desert flow occurs when a thermal low-pressure cell
near Umarkot (fig. 255). The large linear dunes on develops over the Indian subcontinent due to solar
the west margin of the sand sea average 2 km (1.2 heating (fig. 245B). The winter northeasterly flow
mil wide and are commonly 20 km (12.5 mil long; is a result of subsidence from a high-pressure cell
they are interpreted here to be compound linear which develops over central Asia during this
dunes (fig. 256). season (fig. 245A).

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 381

Most of the desert receives more than 100 mm (4 These dunes grade into probable reversing dunes
in.) of rain per year. (See isohyets, fig. 252.) This and star dunes near mountain ranges along the
rainfall occurs mainly during the summer south- eastern and southern margins of the sand sea. To
west monsoon season, figure 257. the south and east, in the Ya-pu-Iai Sha-mo,
Annual resultant drift directions are toward the Yamalik Dunes, and Wu-lan-pu-ho Sha-mo, dome-
northeast at most stations in the region, as shown shaped dunes and star dunes are in elongate fields
on figure 252. Resultant drift directions trend more that parallel the trends of bounding valley sides.
northward from Bhuj (N. 73 E.) to Ganganagar (N.
0
Unusually broad crescentic ridges, not seen
3r E.). Resultant drift directions at all stations elsewhere during this study, occur in the T'eng-ko-
aline within 10 with the trends of large parabolic
0
li Sha-mo at the southeastern margin of the Ala
dunes visible on Landsat imagery, except near Shan Desert.
Jodhpur. Average annual rainfall in the Takla Makan
Monthly resultant drift directions show a south- Desert commonly is less than 100 mm (4 in.).
ward shift at most stations during the winter ( fig. Average annual rainfall in the Ala Shan Desert in-
258). creases from less than 100 mm (4 in.) in the north-
Maximum drift potentials at all Indian stations west to more than 200 mm (8 in.) in the southeast.
studied occur during June. Hyderabad and Effective winds in the Takla Makan Desert are
Jacobabad, in Pakistan, however, record maximums mainly from the north and northeast in the eastern
in July and May, respectively. The summer max- part of the Tarim Basin, and from the west and
imum of drift potential in the Thar Desert is illus- northwest in the western part. Winds in each area
trated in figure 259. Drift potentials are greater dur- tend to be directionally steady throughout the year.
ing all 12 months at southern stations and are Crescentic and linear dunes, or combinations of
progressively less toward the north (figs. 252, 258). these types, commonly occur in areas of the Takla
Makan Desert that have both unimodal and bimo-
dal present-day wind regimes. Star dunes and
Takla Makan and Ala Shan reversing dunes, common in many parts of the Ala
Deserts of the Peoples Shan Desert, may be attributable to the great direc-
tional variability of effective winds there. Little
Republic of China wind energy is available for sand movement in
either the Takla Makan or Ala Shan Deserts as
Summary of Conclusions compared with other desert regions. Wind energy
Major sand seas in northwestern China are in a in both deserts is greatest in the spring.
structural and topographic basin (Tarim Basin) in
the Takla Makan Desert and in low areas on stony Introduction
plains in the Ala Shan Desert farther east. The Landsat imagery has been used to map and
Takla Makan sand sea, west of the playa Lop Nor, describe the distribution and morphology of eolian
may be divided into four regions characterized by sand in the Taklan Makan and Ala Shan Deserts of
distinctive dune assemblages. Large compound the Peoples Republic of China. Sand seas in both
crescentic ridge dunes in the Cherchen Desert, deserts are centered at about lat 40 N. and are the
0

south and west of the Tarim River, grade south- highest latitude sand seas included in this study.
ward and westward into crossing sets of crescentic Meteorological data covering surface winds and
dune ridges with distinctive chevron patterns. precipitation at climatic stations in northern China
Dome-shaped and linear dunes occur in the north- were obtained mainly from the National Climatic
central part of the sand sea, and complex dunes oc- Center, Asheville, North Carolina. Additional data
cupy the south-central part of the sand sea. Dunes were obtained from reports by U.S. Army Air
that seem to be simple barchanoid, dome-shaped, Forces Weather Division (1945), Dalrymple,
and linear characterize the western Takla Makan Everett, and Wollaston (1970), and others. The
Desert, west of the Ho-t'ien Ho. sand-moving potential of surface winds is repre-
Compound crescentic dunes are the main dune sented by sand roses, and average annual rainfall is
type in approximately 33,000 kmz (20,500 miZ) of shown by precipitation isohyets on the maps (fig.
the Pa-tan-chi-lin Sha-mo of the Ala Shan Desert. 260,283).

Digitized by Google
382 A STUDY OF GLOBAL SAND SEAS

WESTERN THAR DESERT. landsat imagery E1101- 05201;

E1120- 05263. A, Compound linear dunes south of the junc-
tion of the Indus and Sutlej Rivers. Vegetation and salt flats oc-
cur in interdune areas subject to periodic flooding. 8, Com-
pound crescentic dunes south of Sukkur, Pakistan. The extreme
aridity of the western part of the desert is shown by the sharp
demarcation of irrigated, vegetated flood plain from the dune
country. (Fig. 256.)

Sand seas of the Takla Makan Desert in the

Tarim Basin of northwestern China (fig. 260) were
first described by 19th century Russian and
Swedish explorers (Bohdanowicz. 1892; Pvetsov
(1895). in Hedin. 1905a; Obrutschev. 1895; and
Hedin. 1905a. b) . Regional geology of the Tarim
Basin has been described by Norin (1941). Recent
Soviet and Chinese investigations include sampling
of eolian sand around the periphery of the Tarim
Basin but do not include studies of the main sand
seas in the interior of the desert. Useful maps of the
Takla Makan and Ala Shan Deserts are the U.S.
Army Map Service 1301 Series. sheets NJ 43. NJ 44.
NJ 46. NK 46. NK 47. NK 48. NI 49; and the U.S.
Department of Commerce. U.S. Air Force Opera-

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 383
7rf

Ganganagar/

1- (%l
",acoblbld

1- (f) THAR Bikaner(

~ ~ (%l
DESERT

JodhPU,;/
~I (..!)
Barmer( 27 3
"~derabld
~ H)'
~ (1)
J F M A M J J A
MON1ltS OF THE YEAR

RELATION OF RAINFAll PATTERNS to effective winds at

Jaisalmer and at Barmer, Thar Desert. Maximum rainfall precedes
time of maximum winds by about 1 month at each station.
Modified from McKee, Breed, and others (1974, fig. 27). (Fig.
257.)
OUTLINE MAP OF THE THAR DESERT, India and Pakistan, with
resultant drift directions for June (solid arrows) and January
tional Navigation Charts F-6, F-7, F-8, F-9, G-7, (dashed arrows). Also shown, drift potential (lower number)
G-8, and G-9. and resultant drift potential (upper number) for June and Janu-
Recent publications about the Takla Makan ary. January values are in parentheses. (Fig. 258.)
Desert are mostly in Russian or Chinese (listed in
Murzayev, 1967, p. 580-615), but some have been
translated into French (Petrov, 1962) and English
(Petrov, 1966, 1967; Murzayev, 1967). A glossary of
geographic names used in central Asia is included
by Murzayev in his publication (1967, p. 533-579).
A book on the physical geography of China
(Zaychikov and others, 1965, p. 533 -579) includes
information about the Takla Makan sand sea
gathered by the Sinkiang Multi-Field Expedition of
1957-59.
Early work in the Ala Shan Desert was by Scan-
dinavian geologists (H6mer, 1936; Norin, 1941). In
1957, Chinese and Soviet members of the Central ~~~~~~A~~~~~J~~A--~S---~-~N=-~D
Huang Ho Complex Expedition studied eolian pro-
MON1ltS OF THE YEAR
cesses in the Ala Shan and O-erh-to-ssu sand seas,
followed in 1958 by the Tsinghai-Hansui Expedi-
GRAPH OF DRIFT POTENTIAL at two representative stations in
tion of the Chinese Academy of Sciences, which the Thar Desert. Both Bhiij and Barmer have highest drift poten-
explored the eastern and southern parts of the Ala tials in June. Lower drift potentials at Barmer illustrate north-
Shan, including the T'eng-ko-li Sha-mo. In 1959, 17 ward decrease of drift potential within the desert. (Fig. 259.)

Digitized by Google
384 A STUDY OF GLOBAL SAND SEAS

DISTRIBUTION AND MORPHOLOGY OF EOLIAN SAND in the Takla Makan Desert, Peoples Republic of China. Map based on
landsat imagery. Isohyets are not recorded. (Fig. 260.)

field teams of Chinese and Soviet investigators P'en-ti. southern Dzungaria. and the Takla Makan
began a 3-year Complex Sand Expedition into the Desert. Field stations were established for studies
Ala Shan and O-erh-to-ssu Deserts, the Ch'ai-ta-mu of sand stabilization and reforestation. vegetation

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 385

91
EXPLANATION
LINEAR DUNES
Ka-shun-sha-mo
Simple. short

1_ _-;!f--~41 Compound. two sets superimpol8d

Complex. with scattered crescentic ridges

CRESCENTIC DUNES
Simple barchanoid ridges
Compound
Barchanoid ridges

Discontinuous barchanoid ridges

Barchanoid ridges with long straight segments

Complex crescentic ridges
With linear elements superimpoeed
Ridges chevron-shaped with linear and
crescentic elements superimposed
With undiscemable pattems

SIMPLE. AND SOME COMPOUND. STAR DUNES

DOME DUNES
~ Simple and compound
~
Compound
In linear chains

Elongated

311" SAND SHEETS AND STREAKS

Compound with linear elements

BEDROCK OUTCROP OR BARRIER

ROCK AND GRAVEL PLAIN WITH
NO LARGE DUNES
PLAYA
VEGETATED PLAIN

- - - BOUNDARY OF SAND SEA

37" ---'> INTERMmENT DRAINAGE-Showing direction
of flow

MAP LOCATION

50 100 150 KILOMETRES

I I I
I I I
o 25 50 75 MLES

(Fig_ 260. - Continued.)

was surveyed, and about 300 eolian sand samples center of the Tarim Basin (I. G. Wilson, 1973, table
were collected. Results of those surveys were 1), which is one of the largest basins with no ex-
reported by Petrov (1961, 1966, 1967), but locations
terior drainage (530,000 km2, or 330,000 mi2) on
of dune areas described in the Soviet literature are
Earth (Norin, 1941, p. 1). This basin is of late Terti-
difficult to relate to dunes of the Ala Shan Desert as
ary tectonic origin and is nearly ringed by high
observed on Landsat imagery. glaciated mountains composed of Paleozoic,
Mesozoic, and Tertiary strata (Norin, 1941, p. 4;
The Takla Makan Desert Suslov, 1961, p. 533-534). The surface of the Tarim
Review of Previous Work Basin slopes gently eastward, from an elevation of
1,400 m (4,600 ft) at So-ch'e to 780 m (2,560 ft) at
Sand seas of the Takla Makan Desert (fig. 260) Lop Nor, a 2,000 km2 (1,240 mi2) playa into which
occupy about 261,000 km2 (162,000 mi2) in the all drainage of the basin seeps (fig. 260). The

Digitized by Google
386 A STUDY OF GLOBAL SAND SEAS

Yarkand-Tarim-Ho'tien Ho drainage system carries 458-459), in which detrital material eroded from
water across the desert to Lop Nor during flood mountains around the Tarim Basin is transported
periods in July and August, when snow melts in the by streams eastward toward the playa, Lop Nor,
surrounding Tien Shan and Kunlun Mountains. and then the sand and silt are transported by winds
Other streams, such as the Ni-ya Ho and K'o-li-ya westward into the sand sea.
Ho, flow only a short distance across the piedmont Large dunes in the Cherchen Desert in the
and into the sands. eastern Takla Makan region (fig. 260) are described
The sand sea in the middle of the basin is ringed (Hedin, 19058, p. 270) as "regularly formed and
by a border of alluvial fans, braided river channels, linked together, crescentic in shape, with a steep
and swampy salt marshes (fig. 260). In historic leeward flank on the west and a long gentle slope
time, cultivated areas on the periphery of the sand on the east." Rows of subsidiary crescentic ridges
sea have been inundated by eolian sands, as the are on the long slopes of these crescentic ridges, as
flow of the Ni-ya Ho and other rivers fed by shown on Hedin's profile sketches (1905a, fIgS.
glaciers in the Kunlun Mountains have diminished 238-257, p. 271-275). Hedin apparently was the
(Norin, 1941, p. 30). first student of desert dunes to describe the form
Lowered water tables around the margins of the classified in chapter J as compound crescentic.
sand sea have contributed to the restriction of pop- These dunes were called "complex barkhan ridges"
lar forests which were extensive until about 1,500 by Petrov (1967, p. 295-298), who recognized that
years ago (V. M. Sinitsyn, in Zaychikov and others, they were of the same form as dunes in the Peski
1965, p. 588; Suslov, 1961, p. 557). Poplars, tamarisk, Karakumy, U.S.S.R. The dunes were described by
Russian thistle, dogbane, and reeds are now Hedin (1905b, p. 400) as "remarkably uniform" in
restricted to old stream valleys and interdune height, ranging only within 10-15 m (33-49 ft) of
marshes. Except for thin covers of grass (Aristida a measured 89.5 m (293.6 ft) and reaching a max-
pinnata, Agriophyllum arenarium, Corispermum imum of 100 m (330 ft).
sp. (Petrov, 1962, p. 147-150) and uncommon Between the dunes of the Cherchen Desert are
saltbushes (Haloxylon sp.), the sand seas are flat-floored interdune basins (the bahirs of Hedin,
"almost completely devoid of vegetation" (Petrov, 1905a, p. 234-243, 269-276, 311-328, 360-369;
1962, p. 147, 150; 1967, p. 306-328; Suslov, 1961, p. 1905b, p. 408-409). Interdune areas in the interior
558). of the Cherchen Desert are commonly covered to a
The Soviet-Chinese Complex Sand Expedition depth of 1-2 m (3-7 ft) by silt or horizontal layers
samples sands from peripheral areas of the Takla of extremely fine yellow clay that is eroded into
Makan Desert, including the K'a-t'e Sha-mo and yardangs (Hedin, 1905a, p. 317 -318, 321-324).
dune fields near Guma (Hotien) and A-k'o-su (fig.
260). All the sands were from relatively small dune Interpretation of Landsat Imagery
fields, many of which formed during histroic time Takla Makan sand seas west of Lop Nor can be
and are of "insignificant thickness" (Petrov, 1967, divided, on the basis of dune types observed on
p. 278-283, 295, 300-301, 308-335, table 33). Landsat imagery, into four sections (fig. 260): (1)
Peripheral eolian sands of the Tarim Basin are the Cherchen Desert in the eastern part of the- Takla
generally fine and include large amounts of silt. Makan Desert, (2) the north-central part of the
They are typically grayish yellow and are believed Takla Makan south of Kucha, (3) the south-central
to have formed in place from the winnowing of part of the Takla Makan, near K'o-li-ya Ho and Ni-
nearby alluvial sands (Petrov, 1967, p. 283, 301, 307, ya Ho, and (4) western sand seas between the Ho-
331, table 33). t'ien Ho and the Yarkand River.
The Soviet and Chinese scientists recorded no The main part of the eastern Takla Makan
sand samples from the large dunes in the main sand Desert, west of Lop Nor, is known as the Cherchen
seas of the Takla Makan Desert, but Petrov (1967, p. Desert. Very large compound crescentic dune
327) stated that the sands there are similar to the ridges in the northern part of the Cherchen Desert
muddy-yellow peripheral sands in texture and are separated by interdune basins commonly 1.7 km
composition. Dunes in the main sand sea may have (1 mil long and 1.2 km (0.8 mil wide. Interdune
formed by processes described by Hedin (1905&, p. areas near the Tarim River are subjected to periodic

Digitized by Google
 REGIONAL STUDIFS OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 387

COMPOUND CRESCENTIC DUNES and closed elongate inter- km2 (453- mj2) field of smaller crescentic (barchanoid) dune
dune areas in the Cherchen Desert, northern Takla Makan ridges encroaches westward over the larger compound
Desert south of the Tarim River, Peoples Republic of China. barchanoid ridges. Patterns of both sets of dunes are enhanced
landsat imagery E1128-04253. Dunes and interdune basins by snow. To the south, the pattern is a complex chevronlike
were measured in the outlined sample area. At the east (up- combination of linear and crescentic elements. (Fig. 261 .)
wind) edge of the sand sea (adjacent to the sample area) a 700-

flooding. Several that are adjacent to the river composed of segments with a mean width of 0.9 km
valley contain lakes, whereas others contain salt (0.6 mil and a mean length of 0.8 km (0.5 mil and
flats (fig. 261). Interdune areas observed on Landsat spaced an average of 0.9 km (0.6 mil apart. The
imagery of the Cherchen Desert closely resemble smaller ridges are oriented toward the southwest,
interdune areas in other sand seas, where they are as are the big dunes, in agreement with the resul-
classed by Glennie (1970, p. 60-61, 98) as "inter- tant drift direction of present-day winds (fig. 260).
dune inland sabkhas." Difference in size of the two sets of dunes might be
Measurements of the large barchanoid dunes in interpreted as evidence of two different episodes of
the Cherchen Desert (fig. 261) show that the ridges dune-building. A similar development of two sets
are composed of crescentic segments with a mean of crescentic dunes is described earlier in the sec-
horn-to-horn width of 3.2 km (2 mil and a mean tion on the eastern Rub' al Khali, Saudi Arabia (fig.
length of 2.2 km (1.4 mil and are spaced an average 240).
of 3 km (1.9 mil apart. They are the largest dunes of The compound crescentic dune ridges in the
compound crescentic form to be noted on Landsat Cherchen Desert grade westward through a zone of
imagery (chapter J, table 38). The large dunes are complex ridges into dome-shaped and linear dunes
overridden south of T'ieh-kan-li-k'o by a field of directly south of the Tarim River. Measurement of
much smaller barchanoid dune ridges (fig. 261) the dome-shaped dunes in this area shows that they

Digitized by Google
388 A STUDY OF GLOBAL SAND SEAS

Dunes in the south-central Takla Makan Desert

between Cherchen and Nochiang are mainly com-
plex ridges in which two trends (south and west).
can be observed on Landsat imagery (figs. 260. 261).
Linear elements predominate in some areas. and
crescentic elements predominate in others in the
southern part of the desert. The two trends result in
a chevronlike dune pattern first described by Hedin
(1905a. p. 364). The linear elements may represent
full development of the dunes that were called
"thresholds" by Hedin. for they connect and lead
up to adjacent crescentic dune ridges in at least one
of the regions he investigated during four crossings
of the desert (Hedin. 1905a. p. 315-327. 360-362.
figs. 263-264; 1905b. p. 409).
Sand seas west of the Ho-t'ien Ho are called the
Bel'kum and Tagkum Sands. divided at lat 3830' N.
by a large escarpment (Norin. 1941. p. 3). The
escarpment is the only major rock outcrop visible
on imagery of the interior of the main Takla Makan
sand sea and is composed of folded Tertiary strata
(Norin. 1941. p. 3). It seems to be an important
topographic barrier to sand movement. Compound
crescentic dune ridges are piled up along the north
(upwind) side. but such massed features are absent
on the south side. where only scattered linear
ridges. dome-shaped dunes. and simple barchanoid
ridges occur (fig. 262). N. A. Belyayevskiy (in
Murzayev. 1967. p. 56) reported that dunes near the
escarpment rise to 300 m (1.000 ft) above the sur-
rounding plain. The general size of dunes in the
o 40 KM western part of the Takla Makan Desert is given as
~~-""""--r-'"""--'"-I'lir--MI~I 60 m (200 ft) (Norin. 1941. p. 3). The western boun-
dary of the Takla Makan sand seas is at the
Yarkand River.
PROBABLE TOPOGRAPHIC CONTROL of position and height
of dunes by an escarpment along the Ho-t'ien Ho, an intermit- The Ala Shan Desert
tent stream that crosses the western Takla Makan Desert. Land-
sat imagery El078-Q4481. (Fig. 262.) Review of Previous Work
The Ala Shan Desert is a hilly plain about 700 km
have a mean diameter of 1.1 km (0.7 mil. A sample (435 mil east of the Takla Makan Desert. at the east
of these dunes is compared with dome-shaped end of the Ho-hsi-tsou-Iang (fig, 263). Sand seas oc-
dunes in other deserts (chapter J. tables 42. 43). cupy about 65 percent of the Ala Shan Desert, be-
Some of the dome-shaped dunes in the Cherchen tween the intermittent Jo Shui on the west, and the
Desert have linear "tails" that extend downwind perennial Huang Ho (Yellow River) on the east. To
(southward). subparallel to the resultant drift direc- the south, eolian sands are bounded by foothills of
tion at Kucha (fig. 260). The dome-shaped dunes the Nan Shan Range, and to the east, by the Ho-lan-
grade southward and westward into fields of short shan Range and the valley of the Huang Ho. To the
linear dunes which occupy the central portion of north. scattered dune fields occur in numerous
the Takla Makan sand sea northeast of Ho-t'ien Ho depressions on the stony plains of northern China
(fig. 260).

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 389

and Mongolia. Main sand seas in the Ala Shan p. 243) and extensive Quaternary and modem
Desert are known as the Pa-tan-chi-lin Sha-mo, alluvial-lacustrine sediments of the 10 Shui
Yamalik Dunes, Ya-pu-lai Sha-mo, T'eng-ko-li Sha- drainage system (Norin, 1941). The sand-carrying
mo, and Wu-lan-pu-ho Sha-mo (fig. 263). capacity of the 10 Shui is not known, but the
The Pa-tan-chi-lin Sha-mo supports shrubs amount of mud carried in suspension at normal
(Hedysarum mongolicum, Calligonum sp., Halox- high water was measured at Hei-ch'eng (Homer,
ylon sp., Cargana microphylla), sages (Artemisia 1936, p. 729-730, fig. 7) at more than 40,000 tons per
sph aeroceph ala) , gramen (Psammochlea villosa) , day, and most of the sediment is deposited
annual grasses (Agriophyllum gobicum and upstream (south) from the Ka-shun-no-erh and So-
Pugionium cornutum), and clumps of Ephedra sp. kuo-no-erh. Dominant west and northwest winds
In some interdune areas and especially around in- are reported to carry, in single day, several tons of
terdune lakes are reeds (Phragmites communis), sand eastward from the lower 10 Shui across the
dogbane (Apocynum hendersonii), and tamarisk gravel plains toward the sand seas of the Ala Shan
(Tamarix ramosissima) (Zaychikov and others, Desert (Homer, 1936, p. 735).
1965, p. 345, 547; Petrov, 1966, p. 252-253; 1967, p. Thickest sand deposits in the Ala Shan Desert oc-
26). Vegetation is reported to be sparse in the cur in its western, driest part (Petrov, 1962, p. 141).
Yamalik Dunes and is generally restricted to inter- Sand in the Pa-tan-chi-lin Sha-mo, in the center of
dune areas in the T'eng-ko-li Sha-mo. In the Ya-pu- the Ala Shan Desert, fills depressions and covers
lai Sha-mo, vegetation is mostly the shrub rock outcrops to an estimated depth of 150 m (490
(Hedysarum scoparium) , sagebrush (Artemisia ft) or more (Petrov, 1966, p. 243). Main sand ridges
sph aeroceph al a) , and grass (Agrioph yll urn in the Pa-tan-chi-lin Sha-mo move very slowly
arenarium). In the Wu-Ian-pu-ho Sha-mo, adjacent southeastward under the influence of dominant
to the Huang Ho (fig. 260), Ordos sage (Artemisia northwest winds (D. Fedorovich, 1961, in
ordosica), gramen, and Hedysarum scoparium Zaychikov and others, 1965, p. 53), but small
grow on the dunes, and Nitraria schoberi and even barchans and barchanoid ridges in the interdune
reeds (Phragmites communis) grow in interdune areas move northeastward as much as 14 m (46 ft)
areas (Petrov, 1966, p. 252). per year because orientation of the main dune
The Pa-tan-chi-lin Sha-mo in the central part of ridges affects wind directions in the interdune
the Ala Shan Desert (fig. 263) are fine grained and areas (petrov, 1966, p. 244).
red, (Petrov, 1966, p. 243). Sand of the T'eng-ko-li Height of dunes in the Ala Shan Desert are
Sha-mo, in the southern part of the desert, is "mud- reported (Petrov, 1962, p. 141; 1966, p. 243-244,
dy yellow, fine grained, and well winnowed"; very 252), as follows: Large crescentic dunes in the main
fine grains (0.25-0.1 mm) account for 98.3 percent body of the Pa-tan-chi-lin Sha-mo average 300 m
of the sand in some areas, whereas lag gravel and (990 ft), with a maximum height of about 400 m
coarse sand occur in the interdune areas (Petrov, (1,310 ft); isolated star and reversing dunes in the
1966, p. 247). The Yamalik Dunes, in the north- northern part of the field are as much as 90 m (295
eastern part of the desert, are composed of red sand ft) high. The Yamalik Dunes are as much as 20-25
deflated from the stony plains and gray sand win- m (65-80 ft) high, and dunes in the Wu-Ian-pu-ho
nowed from alluvium (Petrov, 1968, p. 251). The Sha-mo are as much as 25-30 m (80-100 ft) high
Ya-pu-lai Sha-mo, in the south-central part of the (Petrov, 1966, p. 251, 252).
Ala Shan Desert, is made up of fine sand (Petrov,
1966, p. 253). The Wu-Ian-pu-ho Sha-mo, at the east Interpretation of Landsat Imagery
margin of the desert, is made up of fine sand with Dunes of the Pa-tan-chi-lin Sha-mo cover about
some silt, according to Petrov, who stated that it is 33,000 km2 (20,500 mi2) in the center of the Ala
derived partly from local alluvium but also comes Shan Desert (fig. 264). Northwest (upwind) gentle
from the interior of the Ala Shan Desert. Sand and slopes of large crescentic dune ridges are covered
silt of the Wu-Ian-pu-ho Sha-mo commonly migrate with subsidiary crescentic dunes (fig. 264), so the
into the Huang Ho (Petrov, 1966, p. 252). dunes are compound crescentic (large barchanoid)
Suggested sources of sand in the Ala Shan Desert ridges in the classification system used for Landsat
are outcrops of Cretaceous sandstone (Petrov, 1966, imagery. They are the "complex barchan hill-

Digitized by Google
390 A STUDY OF GLOBAL SAND SEAS

a 50 100 150 KILOMETRES

rl ---'I~I--TI-JI-'I--~I
MAP LOCAnON a 25 50 75 MUS

DISTRIBUTION AND MORPHOLOGY OF EOLIAN SAND in the Ala Shan Desert, Peoples Republic of China (above and facing
page). Map based on Landsat imagery. Isohyets from Petrov (1967) and World Meteorological Organization station data. (Fig. 263.)

shaped ridges" of Petrov (1966, p. 243). Measure- Tarim Basin (fig. 260). Differences in dune patterns
ments of the crescentic segments on Landsat imag- (as opposed to differences in dune types) in the Pa-
ery (fig. 264) show that they have a mean width tan-chi-lin Sha-mo and Cherchen Desert seem to be
(horn to horn) of 2.9 km (1.8 mil and a mean length due mainly to differences in relative size and spac-
of 2.2 km (1.3 mil. The ridges are spaced an average ing of the dune and interdune elements (chapter 1).
of 2.9 km (1.8 mil apart. Numerous lakes occur in Compound crescentic dune ridges of the Pa-tan-
interdune areas. The crescentic dunes of the Pa- chi-lin Sha-mo are compared in chapter J (tables 38,
tan-chi-lin Sha-mo differ in pattern from those in 39) with similar dunes in other deserts.
the Cherchen Desert described earlier (fig. 261) but Toward the southeast margin of the Pa-tan-chi-
are closely similar to those of the Ak-bel Kum on lin Sha-mo (fig. 264B) dune ridges are shorter, dis-
the south and east shores of Po-ssu-t'eng Hu in the tinctly S-shaped, and sharp crested. They are in-

Digitized by Google
 REGIONAL STUmps OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 391

EX PLANAnON the central Asiatic high-pressure cell and a thermal

o Mosn Y SIMPLE. SOME COMPOUND. LINEAR DUNES
CRESCENTIC DUNES
low-pressure cell which develops over the Indian
subcontinent in the summer. During the winter,
D.' Simple barchanoid ridges subsidence from the central Asiatic high tends to
b Compound barchanoid ridges

Complex with
produce northerly to northeasterly winds across the
Takla Makan and Ala Shan Deserts (fig. 266A).
Bulbous ridges During the spring the central Asiatic high weakens,
Feathered ridges its center shifts 10 westward, and a thermal low
-".

Feathered ridges with linear elements

begins to develop over the Indian subcontinent (fig.
Reversing ridges
2668). At this time. both the Takla Makan and Ala
Shan Deserts lie in a zone between the high- and
Ridges with linear elements superimposed
low-pressure cells. and also during this time the
SIMPLE. AND SOME COMPOUND. STAR DUNES
highest drift potentials occur, as discussed later. By
SIMPLE AND COMPOUND DOME-SHAPED DUNES summer time (fig. 266C) low pressure dominates
SHEETS AND STREAKS much of Asia, especially India, and wind regimes in
".~
Io-,-,----d Compound. with linear elements the deserts are weaker and more directionally
BEDROCK OUTCROP OR BARRIER unstable than at other times of the year.
1----1
ROCK OR GRAVEL PLAIN WITH NO LARGE DUNES
Mountains that ring the Takla Makan Desert and
f-------l those in and around the Ala Shan Desert may
PLAYA OR SALT MARSH
deflect wind flow, resulting in variations from the
..........:1' VEGETATED PLAIN OR MARSH
trends of wind direction and strength as just dis-
- 2 0 0 - ISOHYET-Average annual rainfall. in millimetres cussed and as shown in figure 266. In addition, cold
- - - BOUNDARY OF SAND SEA frontal passages. associated with midlatitude
- INTERMITIENT DRAINAGE-Showing direction of flow cyclones, may produce the westerly winds ob-
(Fig. 263. -Continued.) served at some stations (figs. 267.268), particularly
terpreted here as reversing dunes. Isolated dunes at during the spring.
the extreme eastern and southern margins of this The Takla Makan Desert receives less than 100
sand sea are interpreted here as star dunes. mm (4 in.) precipitation per year (fig. 260). Average
Small crescentic dune ridges, dome-shaped annual precipitation is only 30 mm (1.2 in.) at
dunes, and small star dunes of the Ya-pu-Iai Sha- Charkhlik and Cherchen, and in the interior of the
mo, Yamalik Dunes, and Wa-Ian-pu-ho Sha-mo fill Tarim Basin is commonly zero (Zaychikov and
abandoned stream valleys and other low areas on others, 1965, p. 532, 555). In the Lop Nor area,
the stony plains of the eastern Ala Shan Desert. average annual precipitation is about 10 mm/yr (0.4
Dome-shaped dunes and star dunes in the Yamalik in./yr) (Petrov, 1967, p. 306).
Dunes tend to be in linear patterns that parallel the The climate of the Ala Shan Desert is transitional
trends of bounding valley sides (fig. 265A). from the desert climate of the Tarim Basin to the
Dunes in the T'eng-ko-li Sha-mo at the down- steppe climate of Inner Mongolia (Petrov, 1967, p.
wind (southeastern) margin of the Ala Shan Desert 22). Average annual rainfall increases from north-
are broad ridges of a complex variety not seen west to southeast in the Ala Shan Desert (fig. 263).
elsewhere. The broad crescentic ridges seem to lack The southern and eastern margins of the desert
main slipfaces, but the tops of the ridges have may receive approximately 200 mm (8 in.) of rain-
clusters of star dunes at irregular intervals (fig. fall per year (fig. 263).
2658). Southwest of the dunes are elongate salt flats
here interpreted as deflation hollows. Dunes in the Direction of Sand Drift
T'eng-ko-li Sha-mo are reported (Petrov, 1966, p. Annual resultant drift directions differ widely
247) to be more than 100 m (328 ft) high. from place to place within the Takla Makan Desert,
Surface Wind Flow and Precipitation but resultant drift directions at each station are
mostly steady throughout the year (fig. 260). At
Surface wind circulation over the Takla Makan Kucha, in the northern part of the Tarim Basin. an-
and Ala Shan Deserts is controlled principally by nual resultant drift directions are south or

Digitized by Google
392 A STUDY OF GLOBAL SAND SEAS

PA-TAN-CHI-LIN SHA-MO, Ala Shan Desert, Peoples Republic of the )0 Shui intermittent drainage system. B, East edge of the
of China. Landsat imagery E1424 - 03262; E1117 - 03222. A, The sands, where they are bounded by the Ya-pu-Iai Shan; sharply
main mass of sands, showing compound crescentic dune recurved S-shaped dunes are interpreted as reversing
ridges, many with interdune lakes. To the northwest are playas barchanoid ridges. Star dunes are also evident. (Fig. 264.)

southeastward (fig. 269), subparallel to the trend of ing wind patterns as suggested by Hedin (1905a, p.
the dome-shaped and linear dunes of the northern 152, 360-362); however, available wind data for
and central parts of the sand sea. In the western Nochiang, about 60 km (37 mil from these dunes
Tarim Basin -including the Bel'kum and Tagkum (fig. 260 and chapter F) do not confirm Hedin's in-
Sands -resultant drift directions may be toward terpretation.
the southwest as at Pa-ch'u or toward the southeast Annual resultant drift directions in the Ala Shan
as at So-ch'e or east as at Khotan (fig. 260). Resul- Desert (fig. 263) are toward the southeast except at
tant drift directions in the Cherchen Desert near Shan-tan, which is in an intermontane valley (fig.
T'ieh-kan-li-k'o show that the large transverse 263). Resultant drift directions are relatively steady
(crescentic) dunes normally are associated with throughout the year, but some wind regimes are
unimodal or acute bimodal wind distributions, such complex from month to month. For example,
as that at T'ieh-kan-li-k'o. Dunes in the southern monthly wind regimes tend to be complex at Chi-
part of the Cherchen Desert near Nochiang, lan-rai (fig. 268). The average RDPIDP (fig. 189) is
however, are complex combinations of linear and 0.40 for stations in the Ala Shan Desert versus an
crescentic ridges in a chevronlike pattern (figs. 260, average RDPIDP of 0.60 for stations near the Takla
261). This complexity may be a result of intersect- Makan Desert. This difference reflects the greater

Digitized by Google
 REGIONAL STUDI&) OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 393

directional variability of wind regimes in the Ala high-pressure cell, where winds tend to be weak,
Shan Desert and may account for the common oc- and because they are to some degree sheltered by
currence there of reversing dunes and star dunes mountains.
(fig. 264). Star dunes are associated with complex Drift potentials in the Takla Makan Desert are
annual wind distributions. highest in the northeast, near Charkhlik, and west
at Ai-ssu-la-k'o-chan, off the map at lat 3947' N.,
Amount of Sand Drift long 7547' E. Minimum values occur at stations
The average drift potential of stations in the Ala along the southern desert margin (fig. 260).
Shan Desert (129 VV) is slightly greater than the Springtime is the season of highest drift potentials
average drift potential of stations in the Talda in both the Takla Makan and Ala Shan Deserts
Makan Desert (82 VV), but both deserts have low (figs. 270, 271), because they are in the zone be-
drift potentials compared with other deserts around tween the high-pressure cell in China and the low-
the world (chapter F, table 16). These low values pressure cell in India (fig. 266B). Some stations also
may occur because the deserts are positioned have a winter season of relatively high drift poten-
relatively near the center of the central Asiatic tials (fig. 270 B) .

Digitized by Google
394 A STUDY OF GLOBAL SAND SEAS

~'-

T'.~o-Ii Sha-mo
A

DUNE FiElDS IN THE EASTERN ALA SHAN DESERT (above and

facing page). Landsat imagery E1224 - 03165; E1133 - 03112. A,
Yamalik Dunes fill valleys on the stony plains. B, Complex dune
ridges in the T'eng-ko-li Sha-mo, continuous to the southeast
with the area shown in A. To the east is the partly snow-covered
Ho-Ian-shan Range. East of Ho-Ian-shan is part of the Huang Ho
valley. (Fig. 265.)

Digitized by Google
 REGIONAL STUDIES OF SAND SEAS, USING LANDSAT (ERTS) IMAGERY 395

(Fig. 265. -Continued.)

A B c

MEAN SEA-LEVEL PRESSURE (in millibars of mercury) distribution over the Takla Makan and Ala Shan Deserts of China for 3 months
of the year. H, high-pressure cell; l, low-pressure cell. Trends of surface wind flow are shown by solid arrows; dashed arrows indi-
cate winds which may result from frontal passages or cyclones, or from deflection of prevailing winds by mountains. A, January; B,
April; C, June. (Fig. 266.)

Digitized by Google
396 A STUDY OF GLOBAL SAND SEAS

1
DP=1 DP=7 DP=7
FEB. APR. JUNE

~
DP=3 DP=1 DP=1
AUG. OCT. DEC.

SAND ROSES FOR 6 MONTHS FOR KHaT AN, Peoples

Republic of China, illustrating the eastward resultant drift direc-
tion along the southwestern margin of the Takla Makan Desert,
west of the region of complex dunes (fig. 260). As at So-ch'e
(fig. 260) effective winds are directionally steady throughout the
year. Drift potential, in vector units, is given for each month.
(Fig. 267.)

DP=4 DP=2
OCT. DEC.

SAND ROSES FOR 6 MONTHS FOR KUCHA, Peoples Republic

of China, illustrating southward resultant drift directions in the
north-central Tarim Basin. Monthly resultant drift directions
vary little because of directional steadiness of effective winds.
This station also exhibits low drift potentials characteristic of
stations near the Takla Makan sand seas. Drift potential, in vec-
tor units, is given for each month. (Fig. 269.)

DP=29
JUNE

SAND ROSES FOR 6 MONTHS FOR CHI-LAN-T'AI, Peoples

Republic of China, illustrating the complex wind regimes which
may occur from month to month in the mountainous, eastern
Ala Shan Desert. Resultant drift directions are relatively steady
(southeasterly, with the exception of August) during the year.
Drift potential, in vector units, is given for each month. (Fig.
268.)
Digitized by Google
 REGIONAL STUDIES OF SAND SEAS. USING LANDSAT (ERTS) IMAGERY 397

I I I I I I I
F M A M SON
M

VARIATION IN DRIFT POTENTIAL during the year at Chi-Ian-

t'ai and Mao-mu on the east and west margins of the Ala Shan
Desert, Peoples Republic of China. These stations have higher
average monthly and annual drift potentials than most stations
. th T kl Makan Desert b ' on with them,
est drift potent g. (Fig. 271.)

--"
?----
F M A M J J A S o
MONTHS OF THE YEAR

V RIFT POTENTIA ar at represen-

tat ithin the Takla A, Charkhlik,
C eriya. 8, Kucha, stations have
hi ntials in the spri n stations also
ha h drift potential r. Part 8 also il-
lu ase in average d p m the east end
of the Takla Makan Desert along the south edge toward Keriya.
(Fig. 270.)
Digitized by Google
 A STUDY OF GLOBAL SAND SEAS

GLOSSARY
Prepared by MARGUERITE GLENN

1Ir HIS GLOSSARY defines many of the terms high. The crest is asymmetrical and migrates upwind.
~ used in this Professional Paper; it is for the The stoss side is steeper than the lee side.
convenience of the reader who may not be a Adhesion wart. A small irregular wartlike sand accumulation
specialist on deserts or eolian deposits. Many of the made by wind that changes direction rapidly while
blowing on a moist sand surface.
common physiographic, geologic, meteorological,
and mathematical terms used are not defined here Akle (North Africa). See Fishscale dune pattern.
because their meanings are well established. Anchored dune. See Fixed dune.
Entries and their synonyms are printed in Angle of repose. Maximum angle at which loose detritus is
boldface. The most commonly used term for any stable (commonly 32 -34 for dry dune sand).
one meaning is the one that is defined, and the syn- Antiripplet. See Adhesion ripple.
onymous entries are referred to it. When more than Arm (of dune). The trailing outer extension of a parabolic or
one definition is accepted for an entry, the defini- star dune.
tions will be numbered to show the currently more Arroyo (Southwestern United States). Dry wash. or Wadi
common or preferred usage. Geographic areas of (North Africa). Channel of an ephemeral or intermittent
usage are shown in parentheses following some stream. It commonly has vertical walls of unconsoli-
terms. dated detritus 60 em (2 ft) or more high.
Many sources were used in preparation of this Avalanche bedding. Steeply inclined bedding in dunes pro-
glossary. Especially helpful were "A Desert Gloss- duced by avalanche of sand down the slipface of the
dune.
ary," by R. O. Stone (1967), and "Glossary of
Geology," published by American Geological In- Avalanche slope. See Slipface.
stitute (1972). Backshore. The part of a beach that is landward of a berm or
Some entries are terms used for the first time in beach crest. In opposition to foreshore.
this paper, and they are so identified; for example: Bahr (Lake Chad). Seasonally flooded interdune area.
(McKee, this paper). Bajada (Southwestern United States). A broad alluvial slope at
Acute Bimodal wind distribution. Wind distribution with two the base of a mountain range. Fan forms mayor may not
modes whose peak directions itnersect at an angle of 90 be present.
or less (Fryberger. this paper). Bajir (China). See Interdune corridor.
Adhesion ripple or Antiripplet. Irregular sand ridge Barchan, Barcan, Barchane, Barkan. or Barkhan. Crescent-
transverse to wind direction. formed when dry sand is shaped sand dune. Windward slope is gentle and lee
blown across a smooth moist surface. It may be 30-40 slope or slipface is at the angle of repose of dry sand.
em (12 -16 in.) long and a few centimetres (about 1 in.) Horns of the crescent extend downwind.

Digitized by Google
400 A SruDY OF GLOBAL SAND SEAS
Barchanoid dune. Any dune of the barchan or barchanoid- Compound linear dune. A major longitudinal dune ridge with
ridge type. small linear ridges formed on it. The minor ridges may
Barchanoid ridge. Asymmetrical wavy dune ridge. oriented be parallel to the main ridge or diverge from it. as in
transverse to wind direction with gently dipping wind- feathered seif dunes.
ward slopes and steeply dipping (32 -34) lee slopes. Contorted bedding. All disturbed strata. regardless of cause.
commonly separated from downwind wavy dune ridges where deformation occurred after deposition but before
by open interdune corridors. It consists of coalesced lithification. Structures are restricted mostly to simple
barchans in rows. sets of strata or cross-strata with unaffected sets of
Basin. A depressed area without a surface outlet. into which strata above and below.
runoff from the surrounding terrain drains. Couloir (North Africa). Open interdune corridor.
Basketweave dune pattern. Type of parallel wavy dune com- Crescentic dune. (1) See Barchan; (2) in chapters Jand K used
plex in which the dune ridges are offset. forming a in classification of Landsat (ERTS) imagery patterns for
chevron or herringbone pattern of ridges and interdune all barchaos. barchanoid ridge. and transverse type
hollows. dunes.
Berg wind (Namib Desert. South-West Africa). A local name Crescentic dune ridge. (1) See Barchanoid ridge; (2) used in
for an east wind. chapters J and K as a sand ridge or elongate dune with
Blowout or Reactivated dune. Saucer-. cup-. or trough-shaped arclike or barchanoid segments.
hollow formed by wind erosion on a preexisting sand Crescent ripple. A small ripple in the shape of a moon in its
deposit. The adjoining accumulation of sand derived first or last quarter. The horns or arms of the ripple ex-
from the depression. where readily recognizable. is tend downcurrent.
commonly considered part of the blowout. Crest. A dune summit. It is the highest natural projection of a
Brink. The top of the slipface of a dune. It may correspond to dune.
the crest. Cumulative frequency distribution curve or Cumulative
curve. A curve drawn to represent the percentage of oc-
Caliche. An essentially calcium-carbonate deposit remaining currences of a number of observations of a variable less
in or on soil following loss of water through evaporation than or greater than any given value for an entire sam-
or other processes. ple.
Calcrete. Hard. dense layer of caliche in soil.
Deflation. Removal or "blowing out" of fine-grained sediment
Captation. The trapping of irregular eolian sand masses by the wind.
against cliffs or mountains. thus forming climbing or
hanging dunes. Deflation basin. Hollow surface formed by removal of sand
and dust by eolian action. It commonly has a rim of
Chevron dune. A V-shaped variety of a parabolic dune. resistant material surrounding the depression.
Clay dune. Eolian accumulation of sand-sized clay aggregates. Deflation ripple. See Granule ripple.
Claypan. Playa or dry lake formed in a shallow. undrained Desert. An arid or semiarid region characterized by an excess
depression that has a hard. sunbaked surface of clay. of evaporation over precipitation.
Climbing dune. Sand deposited by wind against a cliff or Desert pavement. A sheetlike residual concentration of wind-
mountain slope. polished closely packed gravels. or rock fragments.
Coastal dune. Sand dunes on low-lying land abandoned or mantling a desert surface where wind has removed fine
built up by the sea. material. Gravels or fragments commonly are cemented
Cold desert. Arid region in which the mean annual tem- by mineralized solutions (AGI).
perature is less than 18C (64.4F). such as the Gobi Desert polish. Smooth and polished surfaces on rocks caused
Desert or some Arctic and Antarctic regions. where by the action of windblown sand and dust.
plant and animal life is restricted by low temperature. Desert varnish. Desert lacquer. or Desert patina. Surface
Colluvium. The sediment on a slope or at the base of a slope stain or crust of manganese or iron oxide which charac-
formed by mass-wasting or un concentrated sheet terizes many exposed rock surfaces in deserts. It coats
runoff. outcropping rocks. boulders. and pebbles. imparting a
Complex dune. A combination of two or more different dune brown or black color and a bright luster to their surfaces.
types in a single dune. in contrast to compound dune. Desiccation crack. Mud crack. Shrinkage crack. or Sun
which is a combination of two or more dunes of the crack. Small fissure or crack formed by shrinkage of
same type. clay or clayey beds in drying by the sun.
Complex wind distribution. Wind distribution with more Dikaka. Accumulation of dune sand covered by scrub or grass
than two modes or with poorly defined modes (Fry- vegetation; extended to include plant-root cavities in
berger. this paper). dune sediments (calcified root tubules).

Digitized by Google
 GLOSSARY 401
Discriminant analysis. A statistical procedure for classifying Erg (North Africa). See Sand sea.
subsequent samples into categories preViously defined Established dune. See Fixed dune.
and differentiated on the basis of samples from known-
populations. Extra dune. See Satellite dune.
Dissipation structure. Wavy. irregular layers in sediments. Extradune. Area marginal to dune field which contains sedi-
resulting from the modification of original stratification ments of the same age and source as the dunes (Lupe
by the concentration of colloidal materials (clay and Ahlbrandt. this paper).
minerals. hydrous iron oxides. and humic compounds) Falling dune. Sand accumulation (sloping at the angle of
O J. Bigarella. written commun . 1975). repose for dry sand) to the lee of a cliff or mountain side
Dome dune. Low circular or oval mound formed where dune as sand is blown off the top.
height is inhibited by unobstructed strong winds. It Fan (alluvial). A sedimentary deposit having a distinctive con-
generally lacks a slipface. vex upward surface that radiates downslope from a
point where a stream emerges from a canyon floor onto
Draa (North Africa). Large-scale accumulation of eolian sand.
a plain with more gentle gradient.
Used by I. G. Wilson (1973).
Drift potential (D.P). A measure. in vector units. of the sand- Fanglomerate. Indurated alluvial-fan gravel.
moving power of the wind. It is derived from reduction Feathered dune. A variety of compound linear dune in which
of surface wind data through a weighting equation and subsidiary ridges diverge from a main ridge.
usually represents 1 year (Fryberger. this paper). Festoon structure or Festoon-type cross bedding. A cross-
Dund (India). Interdune area filled by a lake. lamination pattern resulting from repetition of a two-
step process: (1) Erosion of plunging troughs and (2) fill-
Dune. Accumulation into a mound or ridge of windblown sedi-
ing of the troughs by thin concave upward layers
ment. commonly having a gentle upwind slope and one
generally conforming to the shape of the trough floors
or more steep lee slopes (slipfaces).
(fig. 157). Festoon pattern is most conspicuous on verti-
Dune-extradune depositional system. A system containing cal faces cut at right angles to longitudinal axes in a
all the sediments of a dune field and marginal area. It in- series of superimposed troughs plunging in the same
cludes dune and interdune sediments and associated ex- direction (Knight. 1929. p. 58). The troughs may be filled
tradune sediments. such as alluvial-fan. wadi. serir. symmetrically or asymmetrically.
stream. sabkha. playa. lake. beach. and tidal-flat
Fishhook dune. See Hooked dune.
deposits (Lupe and Ahlbrandt. this paper).
Fishscale dune pattern, Akle. or Peak and Fulji topography.
Dune-interdune depositional system. A system containing
Type of parallel wavy dune pattern in which the inter-
only dune and interdune sediments (Lupe and
dune areas are enclosed by crescentic elements of the
Ahlbrandt. this paper).
dune ridges.
Dune massif. A compound star dune.
Fixed dune. Nonmigratory dune fixed by vegetation or by ce-
Dune spacing index. A measure of the average number of mentation.
dune ridges crossing a 5O-km-long (31.1-mi-Iong) line
Fluid threshold shear velocity. That shear velocity at which
normal to the trend of dune ridges in a given area.
sand movement starts as a result of the shearing stresses
Dune valley. See Interdune. imparted to the sand by the wind. (See Bagnold. 1941.
p.88.)
Elb (Algeria). A transverse dune (Capot-Rey. 1945. p. 392). Frosted sand grains. Typical sand grain of eolian desert sedi-
Elephant-head dune (Colorado Desert. Calif. U.S.A.). Small ments. whose etched and pitted surface is created by im-
sand shadow whose windward face is covered with pact with other wind-transported grains and causes a
vegetation and with a long tapering snout of sand ex- scattering of light. It is the opposite of a glassy or
tending to the lee. polished sand grain.
Elongate blowout dune. Wind-excavated basin with a cres- Fulji or Fulje. (I) A depression (interdune) between barchans
cent-shaped rim of sand about the lee side. It differs or barchanoid ridges. especially where dunes are press-
from blowout dune in that it has an elongate form. and ing closely on one another (Arabia); (2) blowout or
there is slight migration of the basin and rim in the small parabolic dune (Australia).
direction of the prevailing wind. Giant crescent dune pattern. Type of parallel wavy dune pat-
Eolianite. All consolidated sedimentary rock formed from tern in which ridges are composed of coalesced
wind-deposited sand. commonly having large-scale megabarchans and are separated by open or closed inter-
eolian-type cross-strata. dune hollows.
Eolian sand. Wind blown sand. Gibber (Australia). Gravel plain.
Eolian sandstone. Sedimentary rock formed of consolidated Grain flowage. A type of sand avalanching in which in-
windblown sand. See Eollanite. dividual grains move independently of each other, in

Digitized by Google
402 A STUDY OF GLOBAL SAND SEAS
opposition to avalanching by mass movement of slump- Intermediate-energy wind environments. Wind environ-
ing. ments with an annual drift potential of 200 -399 vector
Granule ripple. Unusually large wind ripples composed in part units.
of detritus approaching or attaining granule size. 2-4 Irq. (1) (Arabia) Elongate dune or belt of elongate dunes
mm (0.08 -0.16 in.). They are commonly associated with (Stone. 1967. p. 255); (2) (North Africa) narrow
local concentrations of lag material. longitudinal dune or a seif (Stone. 1967. p. 263).
Granulometric control. The ability of various grain sizes to Isohyet. Line on a map connecting points which receive equal
control the shape or dimension of eolian bedforms. such amounts of average rainfall.
as ripples or dunes (I. G. Wilson. 1973). Khurd (Algeria). Guern. A high (80 -100 m (262 -328 ft
Graphical measures. Graphical method of obtaining statistical pyramid-shaped sand dune with curved slopes. formed
values of grain size in which selected values from points by the intersection of seifs (Capot-Rey. 1945. p. 393).
along a cumulative frequency curve are used in for- Kurtosis. The quality. state. or condition of peakedness or flat-
mulas to calculate the desired statistical data. ness of the graphic representation of a statistical dis-
Guern. See Khurd. tribution.
Hairpin dune. A greatly elongated parabolic dune that has Lag gravel. Residual accumulation of coarse detrital particles
migrated downwind. its horns drawn out parallel to (larger than sand size) from which the fine material has
each other. formed where wind is in conflict with been deflated.
vegetation and where winds are strong and of constant
Laguna (Southwestern United States). Shallow ephemeral lake
directions. in the lower part of a bolson. fed by intermittent
Hammada or Hamada. A bare rock surface in a desert. streams. See also Playa.
Hanging dune. See Falling dune. Lee dune. Dune formed to the lee of an obstruction.
Hardpan. Cemented or indurated layer of soil usually com- Lee slope, lee side, leeward side. The downwind side of a
posed of calcium carbonate. silica. iron oxide. or clay. at dune or other protuberance in the path of a wind.
or some distance beneath the ground surface. Leptokurtic (1) A frequency distribution that has a concentra-
High-energy wind environments. A wind environment with tion of values about its mean. greater than for the cor-
an annual drift potential of 400 vector units or more. responding normal distribution; (2) a narrow frequency-
Histogram. Multiple-bar diagram showing relative abun- distribution curve that is more peaked than the corres-
dances of specimens. materials. or other quantitative ponding normal distribution curve.
determinations divided into a number of regularly ar- Lettau equation for rate of sand drift. The equation suggested
ranged classes. by Heintz Lettau (University of Wisconsin. Madison.
Hooked dune or Fishhook dune. Dune consisting of a long written commun .. 1975) for the rate of sand drift. This
sinuous ridge forming the shaft and a well-defined cres- rate is a function of grain size and drag velocity (shear
cent forming the hook. velocity) of the wind. See chapter F. fig. 92.
Horns or Arms. The pointed ends of a dune. especially the for- Linear dune. Longitudinal dune. or Seif. Parallel. straight
ward extensions of a barchan dune. dune with slipfaces on both sides and with its length
many times greater than its width.
Hot desert. Arid region in which the mean annual temperature
is greater than 18C (64.6F) (Stone. 1967. p. 230). Loess. Soft. porous. yellow- or buff-colored accumulation of
wind-laid particles that are of silt size or smaller.
Impact threshold shear velocity. The minimum shear Longitudinal dune. See Linear dune.
velocity required to maintain sand in saltation. (See
Bagnold. 1941. p. 88.) Low-energy wind environment. A wind environment with an
annual drift potential of less than 200 vector units.
Impact threshold wind velocity. The minimum wind velocity
required to maintain sand in saltation. (See Bagnold. Low latitude desert. Hot and dry desert located between lats
1941. p. 32.) 15 and 35 N. or lats 15 and 35 S. in the region of
subtropical high-pressure air masses and the trade
Inselberg. Prominent residual steep-sided rocky hills rising winds.
abruptly from a plain or desert.
Interdune. Relatively flat area between dunes. Mean grain size. The diameter equivalent of the arithmetic
Interdune corridor. Interdune hollow. or Interdune area. mean or average grain size.
Continuous flat or gently sloping surface between dune Medano. Sand dune (Spanish term).
ridges or rows of dunes. Megabarchan. Giant barchan. 100 m (328 ft) or more in height.
Interior drainage. Centripetal drainage. or Internal drainage. These dunes commonly coalesced into ridges to form
Drainage toward the lowermost part of an enclosed parallel wavy dune patterns. Each crescent in the ridge
basin. normally has one major slipface.

Digitized by Google
 GLOSSARY 403
Me80kurtic. Closely resembling a nonnal frequency distribu- Onshore wind. Winds that blow from the sea onto a land sur-
tion - A distribution curve that is neither leptokurtic face.
(very peaked) nor platykurtic (flat across the top). Orographic desert. See Rain shadow desert.
Middle latitude desert. An arid region between lats 35 and SOO Oscillation ripples. Symmetrical ripple marks consisting of
N. or lats 35 and soo S. that owes its existence either to sharp. narrow. relatively straight crests between broadly
its location in the deep interior of a continent or to high rounded troughs.
mountains across the path of the moisture-bearing pre-
vailing wind. Oued. See Arroyo.

Migration (dune). Movement of a dune due to transfer of sand Packing. The manner of arrangement or spacing of the solid
from windward to the leeward side. particles in a sediment or sedimentary rock.
Miniature terrace-and-cliff structure. A succession of abrupt Pan. (1) Natural basin or depression that intermittently con-
rises or steps of about 2.54 mm (0.1 in.), each ascending tains water that may leave a saline deposit; (2) a dry
an inclined bedding plane. lake or playa.
Moderate-energy wind environment. A wind environment Parallel ripples. Ripple marks with relatively straight crests.
with an annual drift potential of 200 --399 vector units. oriented transverse to wave or current direction and
having a symmetrical or asymmetrical profile.
Modified lee slope or surface. A dune slipface or slipfaces
altered by slumping, grain flowage, or deposition of sus- Parabolic dune. A U-shaped or V-shaped dune representing a
pended load. type of blowout in which the middle part has moved for-
ward with respect to the sides or anns. Open end of U or
Moment measures. The expected values of each of the powers V faces upwind and the anns commonly are anchored
of a random variable that has a given distribution; a by vegetation. One to three slipfaces occur on the con-
weighted measure of central tendency; in sedimen- vex outer part of the nose and arms of the dune.
tology they are related to the center of gravity of the par-
ticle-size-distribution curve and are defined about the Parallel sinuous ripples. Ripple marks in which the horizontal
mean value of the variable. distance. nonnal to flow. is many times the wavelength.
and the crests swing from side to side as viewed in plan.
Monsoon. Type of wind system in which there is a complete or
almost complete reversal of prevailing direction from Parallel straight dune. Seif. or Longitudinal dune. A linear
season to season. It is especially prominent within trop- ridge of windblown sand whose length is much greater
ics on eastern sides of great landmasses. than its width. and which commonly has slipfaces along
Morphology. The shape or physiography of land-surface both sides. It moves in a direction parallel to its length.
features. Parallel wavy dune complex. A group of crescentic dune seg-
ments that have coalesced into wavy ridges. Each cres-
Mud polygon, Desiccation polygon, or Mud crack polygon. centic segment of the ridge commonly is as wide as it is
Block of fine-grained sediment, bounded by mud cracks long. and the segments are asymmetrical. All segments
or shrinkage cracks, that commonly has three to five of a ridge have their major slipfaces oriented nonnal to
sides. although in some examples it may have as many wind directions.
as eight sides.
Patina. A thin film of Desert varnish.
Narrow unimodal wind distribution. A wind regime that pro- Peak and Fulji topography. See Fishscale dune pattern.
duces a distribution of drift potential on a sand rose in Pediment. A broad surface or plain fonned in a desert terrain
which 90 percent or more of the drift potential lies with- by erosion at a mountain base. It may be bare bedrock or
in a 45 arc of the compass. veneered by a thin layer of alluvium.
Nefud. A high sand dune. See also Sand sea. Piedmont. A surface lying or fonned at the base of a mountain.
In desert regions it is a plain or slope and may include a
Observer bias. Tendency for observers to record wind direc- pediment. bajada. or both.
tions from prime directions instead of from intennediate
directions of the compass. Platykurtic. (1) A frequency distribution that has a concentra-
tion of values about its mean less than for the corres-
Obtuse bimodal wind distribution. A wind regime that pro- ponding nonnal distribution; (2) a broad. flat-topped
duces a distribution of drift potential on a sand rose that frequency distribution curve that is less peaked than the
has two modes whose peaks intersect at an angle greater corresponding nonnal distribution curve.
than 90.
Playa (Southwestern United States). Clay pan, Dry lake. Salt
Oghurd (North Africa). A massive summit above the general ftat. or Salt pan. An extremely flat vegetation-free area
sand dune level, or a massive mountainous dune resting of clay. silt, and (or) evaporite minerals in the lower-
on an underlying rocky topographic feature (Stone. most part of a basin having interior drainage in an arid
1967. p. 263). or semiarid region. It is fonned by a temporary lake
Omiribi (Kalahari Desert, Africa). See Arroyo. which has evaporated. leaving behind fine sediment.

Digitized by Google
404 A STUDY OF GLOBAL SAND SEAS
The tenn is also applied to the temporary lake. (Spanish Retention ridge or Precipitation ridge. A ridge fonned
tenn meaning beach. shore. orstrand.) parallel to shore by deposition of sand on the landward
Playa lake. Temporary or ephemeral lake in the desert. margin of the backshore. It accumulates against a
vegetation barrier and is a nearly stabilized dune.
Plinth. Lower and outer parts of seif. beyond the slipface boun-
daries. This part of the dune has never been subjected to Reversing dune. A dune that tends to grow upwards but mi-
sand avalanches (Bagnold. 1941. p. 229). grates only a limited distance because seasonal shifts in
direction of dominant wind cause it to move alternately
Polypyramid. A star dune. in nearly opposite directions.
Precipitation ridge (Cooper. 1958. p. 55). See Retention ridge. Rhourd. A pyramid-shaped sand dune. fonned by the intersec-
Prevailing or predominant wind. A wind that blows most of tion of other dunes (Aufrere. 1934. p. 139). See also
the time from one particular direction. Khurd.
Procedure bias. Error introduced into a wind summary when Ripple. Any topographic deviation on the bedding surface of a
observations originally recorded from 36 points of the sedimentary deposit that resembles a ripple of water.
compass are summarized into 16 directions. It results in Ripple height. The difference in elevation between the lowest
disproportionate representation of observations from point of a trough and the highest point of an adjoining
the prime directions of the compass at the expense of crest in a rippled sand surface.
the intennediate directions.
Ripple index. Wavelength of a ripple divided by height of a
Produne. The part of the dune lying beyond the dune front and ripple (or ratio of wavelength to height).
gently sloping down to the desert floor across which the
dune is advancing. Ripple symmetry index. The length of the horizontal projec-
tion of the stoss side divided by length of the horizontal
Pyramidal dune or Polypyramid. See Star dune. projection of the lee side of a ripple (or ratio of stoss side
to the lee side of a ripple).
Raindrop sand crater. A thin detached layer or shell of sand. Ripple wavelength. The horizontal distance between the
roughly circular with thin margin. fonned by raindrops deepest points of two troughs or the horizontal distance
striking dry sand. Center of the shell is depressed and between two crests in a rippled sand surface.
margin is raised slightly above the general sand level.
After drying. the shell remains firm and retains the Rising dune. See Climbing dune.
same shape; on a tilted sand surface it has a raised lower Roughness factor during sand driving (K'). The focus during
rim. sand driving. or that height. above the surface. at which
Rain shadow desert or Orographic desert. Arid land on the lines depicting logarithmic height versus wind velocity
lee side of a mountain or a mountain range which inter- for several wind velocities converge. See also Bagnold.
cepts moisture-laden air. 1941. p. 59.
Recurved dune. A dune in which a forward extension is Sabkha. See Sebkha.
curved back toward main part of dune as by deflection
of wind or by opposing action of two or more wind Salina. A place where crystalline salt deposits are fonned.
durections. See also Reversing dune and Sigmoidal such as a salt flat. a salt pan. a salada. especially a salt-
dune. incrusted playa or a wet playa.
Reg (Algeria). A large gravel deposit on a desert surface. It Saltation. Bounding movement of rock particles. usually sand
differs from serir in that most of the fine detritus has not size. when carried by wind.
been removed (Gautier. 1935. p. 62). Sand avalanche. Movement of large masses of sand down a
Resultant drift direction (RDD). The direction of the resul- dune face when the angle of repose is exceeded. or
tant drift potential. when the dune is disturbed. It may consist of mass
movement (slumping) or grain flowage or both.
Resultant drift potential (RDP). A measure in vector units. of
the net sand-moving power of the wind at a station. Sand sea. Erg (North Africa). Vast region covered with sand
Vector unit totals from various directions on a sand rose and occupied by dunes.
are treated as vector quantities and resolved tri- Sand drift. (1) A deposit of eolian sand fonned as the result of
gonometrically to a resultant. the magnitude of which is a decrease in wind velocity. It is commonly developed
the resultant drift potential. RDP. the direction of which on the leeward side of windbreaks. such as vegetation.
is the resultant drift direction. RDD. boulders. rock outcrops. or other irregular surface relief.
Resultant drift potential/drift potential (RDP/DP). A dimen- (2) Used in chapter F (this paper) as a general tenn for
sionless parameter which is a measure of the variability movement of sand by wind occurring in a desert or
of a wind regime without regard to specific directional along the shore.
components or total energy of environment. Sand dune. Ridge or pile of windblown sand. See Dune.

Digitized by Google
 GLOSSARY 405
Shear sorting. Sorting of sediments. as in a sand dune.
Sand fall. Sand swept over a cliff or escarpment. See also Fall-
ingdune. whereby smaller grains tend to move toward the zone of
greatest shear strain. the shear plane. and the larger
Sand glacier. Sand accumulation shaped as a broad fanlike
grains toward the zone of least shear. In some dunes it is
plain that develops where sand is blown up the sides of
characterized by a lamination of fine-grained minerals
a mountain. through passes and saddles. to spread out
and occurs in a zone between moving and residual sand
on the opposite side.
(Stone. 1967. p. 246).
Sand levee. See Whaleback. Shrinkage crack. See Desiccation crack.
Sand ridge. (1) A parallel straight or linear dune: (2) a Shrub-c:oppice dune. Small streamlined dune that forms to the
relatively coarse-grained high-crested widely spaced lee of vegetation on a smooth very shallow sand surface.
ripple of sand.
Sigmoidal dune. S-shaped steep-sided sharp-crested dune
Sand ripple. A ripple mark formed by wind in dry sand: the formed under the influence of alternating and opposing
wavelength depends on wind strength and remains con- winds of roughly equal velocities (Holm. 1957). See also
stant with time. See Ripple. Reversing dune and Recurved dune.
Sand rose. A circular histogram depicting the amount of sand Silcrete. A secondary deposit of siliceous material in soils of
potentially moved by winds from various compass hot. arid. and semiarid regions.
directions at a given geographic locality.
Silk (North Africa). See Seif.
Sand shadow. (1) Small mound or dune that forms around
clumps of vegetation. behind rocks. or around other Singing sand. or Acoustical. Barking. Booming. Musical.
obstructions: (2) eolian deposit formed in the lee of a Roaring. Sonorous. Sounding. Whispering. or Whis-
fixed obstruction. See also Sand drift. tling sand. Terms used for sand that. when sliding
down a slipface. emits a musical tone or a humming or
Sand sheet. Accumulation of sand in essentially flat laminae booming sound.
forming a sheetlike or blanketlike deposit. It has no slip-
faces. but has distinct geographic boundaries. Skewness. The state of a frequency distribution which
bunches on one side of the average and tails out on the
Sand streaks. See Stringers.
other side. Skewness results from lack of coincidence of
Sand strip. Long. narrow strip of sand that extends for a con- the mode. median. and arithmetic mean of the distribu-
siderable distance downwind from each hom of a dune. tion. It is measured by the quotient of the difference be-
Satellite dune or Extra dune. Auxiliary or smaller dune tween the arithmetic mean and mode divided by the
associated with main dune. standard deviation.

Sebkha. Sabkha. Sebkra. and many others (especially North Slipface. Steep face on the lee side of a dune. usually at the
Africa and Arabia). A flat area of clay. silt. or sand com- angle of repose of dry sand (commonly 30 --34).
monly with saline incrustations: a surface of deflation Slumping. A mass movement in which grains normally remain
in which dry. loose particles are removed down to the relatively cohesive within sand sheets. even though in-
level of the groundwater or to the zone of capillary con- dividual groups of laminae may be warped. stretched. or
centration (Friedman and Sanders. 1978): includes: (1) otherwise contorted. This movement is in contrast to
Coastal sebkha - Coastal flat that occurs just above sand-flow avalanching during which laminae are being
level of normal high tide in hot. arid. desert climate. (2) progressively destroyed by the separation and disper-
Inland sebkha - Flat inland areas of clay. silt. or sand. sion of the component grains.
often with saline incrustations. See also Playa. Clay
pan. (3) Interdune sebkha. Slump mark. A surface marking formed on a foreset deposit by
slumping or mass avalanching. It may consist of a crack
Seif. Sif. Sief. Er (North Africa and Arabia). Linear dune. or or step fault resulting from tensional stresses near the
Longitudinal dune. Linear dune which. in groups. com~ avalanche summit or of wrinkles and other compres-
monly makes a parallel straight pattern oriented in the sional structures in the zone of compression below.
direction of prevailing wind or winds. It charac- Dendritic and other varieties of slump marks also may
teristically has two slipfaces. form.

Semiarid. Partially arid. It refers to region in which the mean Southeast trades. Winds persistently from the southeast in the
annual rainfall is 30 --40 cm (12 -16 in.) according to Southern Hemisphere. caused by counterclockwise cir-
some geographers. or between 25 --50 cm (10 and 20 in.) culation about permanent oceanic high-pressure cells.
according to others. centered about lat 20 S.

Serir (Egypt and Libya). A rough desert surface of fairly large Stabilized dune. See Fixed dune.
rounded pebbles with no fine material remaining be- Stabilized free dune or active form. A dune stabilized by
tween pebbles (Gautier. 1935. p. 62). vegetation but still retaining its initial form.

Digitized by Google
406 A STUDY OF GLOBAL SAND SEAS
Star dune. A pyramidal dune. roughly star-shaped or resem- Trough cross strata. Set of cross-strata whose lower bounding
bling a pinwheel. with three or more radial buttresses or surface is curved surface of erosion. It results from
arms extending in various directions from a high central channeling and subsequent deposition. Overlapping sets
cone. Slipfaces dip in at least three directions. may show festoon pattern.
Stellate rose. A star dune. Two-storied barchan. A compound barchan.
Stoss side. Side of dune facing dominant wind direction and
direction from which dune has migrated. It is opposite Umbracer dune. Type of longitudinal dune that tapers to a
the lee side. point downwind. formed under constant wind direction
in the lee of an obstacle. It may form behind prominent
Straat (South Africa). See Interdune corridor. bedrock obstacles. or more commonly behind clumps of
Street. See Interdune corridor. bushes (Shrub-coppice dune).
Stringers. Elongate sand bodies that have no slipfaces. They Unidirectional dune field. Dune field in which dunes migrate
are varieties of sand sheets that have stringlike shape in only one principal direction.
and that commonly. though not exclusively. develop Upsiloidal dune. See Parabolic dune.
along the margins or at the downwind extremity of a
Uruq. (1) (Arabia) Elongate or seif dune (Stone, 1967, p. 256);
sheet. The term "streak" is used instead of "stringer" in
(2) (North Africa) chain of elongate dunes (Stone, 1967.
chapters J and K. which are based on observations of
p.264).
Landsat images.
Surface creep. The slow movement of large detrital grains as a U-shaped dune. See Parabolic dune.
result of being pushed by the impact of smaller saltating
grains. Vector unit (VU). Number arbitrarily used in computation and
derived from application of a weighting equation for
Surface wind. Wind near ground-surface level. It is usually sand movement by wind to surface wind data. The
measured at a height of 10 m (32.8 ft) or less. meaning of a vector unit in terms of sand transport
Suspended load. The part of the total wind load that is carried kg/m2/yr (lb/ftz/yr) will differ as different weighting
for a considerable period of time in suspension. free equations or other variables are used, but its value is
from contact with the desert floor. It consists mainly of directly proportional to the effective wind energy of the
silt and finer detritus. environment.
Suspension. A mode of sediment transport in which the up- Ventifact. Wind-polished or abraded pebble.
ward currents in eddies of turbulent flow are capable of
supporting weight of particles and keeping them in- Wadi. See Arroyo.
definitely held in the air. Wash. (1) See Arroyo; (2) alluvium. especially coarse
Swash zone. Area of a beach where water rushes up onto it alluvium.
following the breaking of a wave. Wedge-planar cross-strata. Set of cross-strata bounded by
Tabular-planar cross-strata. Set of cross-strata bounded by oblique erosion surfaces that separate it from other sets
planar. essentially parallel surfaces. of cross-strata.

Thematic map. Interpretive drawing. based on ERTS -1 color Weighting equation. A function (used in this report) to weight
mosaic. which expresses with patterns the relationships surface winds of different velocities according to their
of features observed on the imagery. relative ability to move sand. See chapter F.

Thermal low. An area of low atmospheric pressure. caused by Weighting factor for surface winds. A numerical expression
heating of the land surface by solar radiation. resulting of the rate at which wind of a certain velocity moves
in expansion of adjacent air. sand compared to winds of other velocities.

Traction. A mode of sediment transport in which particles are Whaleback or Sand levee. Coarse residue or platform built up
swept along. or are immediately above and parallel to. and left behind by the passage of a long continued suc-
the ground surface. Movement may be by rolling. slid- cession of seif dunes along the same path (Bagnold.
ing. dragging, pushing, or saltation. 1941. p. 242).

Trade wind desert. See Low latitude desert. Wide unimodal wind distribution. A wind regime that pro-
Transverse dune. Asymmetric sand ridge (leeward slope. duces a distribution of drift potential on a sand rose that
steep; windward, gentle) in which long dimension is has a single peak or mode on the compass and an arc of
normal to dominant wind direction. It has one slipface. more than 45.
It shows largely horizontal strata (apparent dip) in cross Winddrift dune or Windrift dune. Linear dune deposited on
sections parallel to crest, in contrast to curving strata of the lee slide of blowout areas.
barchanoid ridges viewed in similar sections. Wind regime. The pattern of winds characteristic of a particu-
Tropical desert. See Low Latitude desert. lar region.

Digitized by Google
 GLOSSARY 407
Wind rose. A circular histogram depicting the percentage oc- Windward slope. The upwind side of a dune.
currence of wind speed groups from a given set of com-
pass directions. at a certain geographic locality. It has Xerophyte. Plant adapted to an arid environment.
arms proportional in length to the number (or percen-
tage) of observations of wind from a given direction. Yardang or Yarding. (1) Sharp salient ridge of incompetent
Wind shadow. The area adjacent to a slope. cliff. or escarp- sedimentary material oriented parallel to the direction
ment which is protected from the direct action of the of the prevailing effective wind and presumably pro-
wind blowing over it. duced by the abrasion of wind-driven sand; (2) irregular
ridge bounded by rounded troughs formed by eolian ero-
Wind shadow dune. See Umbracer dune and Shrub-coppice sion; (3) landscape form produced in limestone or
dune. sandstone regions by infrequent rains combined with
Wind sweep. Windward slope of an advancing dune. or that wind action resulting in a surface covered with a fine
portion of the windward slope of a dune up which the and compact lacework of sharp ridges pitted by corro-
main wind currents pass. sion.

Digitized by Google
Digitized by Google
 409

A STUDY OF GLOBAL SAND SEAS

SELECTED REFERENCES
Ahlbrandt. T. S.. 1973. Sand dunes. geomorphology and ARAMCO Survey. Projects Construction Services (Shaik
geology. Killpecker Creek area. northem Sweetwater Mahboob. compiler). 1974. AER -716. Southem area cli-
County. Wyoming: Wyoming Univ. Ph. D. thesis. 174 p.; mate. estimated climatic information for Shaybah BPI
available from Xerox University Microfilms. Ann Arbor. 151-192 and BPI 169-29: Arabian American Oil Co.
Mich. [U.S.A.]. no. 73-25.539. 212 p. Dhahran. Saudi Arabia. 35 p. and appendix [34] p.
_ _ 1974a. Dune stratigraphy. archaeology. and the Auden. J. B. 1950. Introductory report on the ground-water
chronology of the Killpecker dune field. in Michael resources of westem Rajasthan: New Delhi. India Geol.
Wilson. ed.. Applied geology and archaeology - The Survey Bull .. ser. B. no. 1. 59 p.
Holocene history of Wyoming: Wyoming Geol. Survey
Aufrere. L60n. 1928. L'orientation des dunes et la direction des
Rept. Inv. 10. p. 51-60.
vents: [Paris] Acad. Sci. Comptes Rendus. v. 187. p.
_ _ 1974b. The source of sand for the Killpecker sand-dune
833-635.
field. southwestem Wyoming: Sed. Geology. v.lt. no. 1. p.
_ _ 1930. L'orientation des dunes continentales: Internat.
39--57.
Geog. Cong .. 12th. Cambridge. England. 1928. Proc .. p.
_ _ 1975. Comparison of textures and structures to dis-
220-231.
tinguish eolian environments. Killpecker dune field.
_ _ 1932. Morphologie dunaire et metflorologie saharienne:
Wyoming: Mtn. Geologist. v. 12. no. 2. p. 61 -63.
Assoc. Geographes Fran~ais Bull. 56. p. 34 --47.
Airy. G. B.. 1845. Tides and waves: [London] Encyclopedia
_ _ 1934. Les dunes du Sahara algerien: Assoc.
Metropolitans. v. 5. p. 241-396.
Geographes Franc;ais Bull. 83. p. 130-142.
Alimen. Henriette. 1965. The Quatemary era in the northwest
Sahara. in H. E. Wright. Jr . and D. G. Frey. eds . Interna- Averitt. Paul. Detterman. J. S .. Harshbarger. J. W.. Repenning.
tional studies on the Quatemary: Geol. Soc. America Spec. C. A .. and Wilson. R. F. 1955. Revisions in correlation
Paper 84. p. 273 -291. and nomenclature of Triassic and Jurassic formations in
Alimen. Marie-Henriette. Doudoux-Fenet. Daniele. Ferrere. southwestern Utah and northem Arizona: Am. Assoc.
Jacqueline. and Palau-Caddoux. Marguerite. 1957. Sables Petroleum Geologists Bull . v. 39. no. 12. p. 2515-2524.
quatemaires du Sahara nord-occidental (Saoura-Ougarta): Baars. D. L. 1961. Permian blanket sandstones of the Colorado
Algeria Service de la Carte GeOlogique Bull. 15. 207 p. Plateau. in J. A. Peterson and J. C. Osmond. eds .. Geome-
[1958]. try of sandstone bodies - A symposium. 45th Ann. Mtg .
Allen. J. E. and Balk. Robert. 1954. Mineral resources of Fort Atlantic City. N.J.. 1960: Tulsa. Okla.. Am. Assoc.
Defiance and Tohatchi quadrangles. Arizona and New Petroleum Geologists. p. 179-207.
Mexico: New Mexico Bur. Mines and Mineral Resources _ _ 1962. Permian System of Colorado Plateau: Am. Assoc.
Bull. 36. 192 p. Petroleum Geologists Bull. v. 46. no. 2. p. 149-218.
Allen. J. R. L. 1968. Current ripples - Their relations to pat- _ _ 1972. Southwestem area. in Permian System. in
tems of water and sediment motIon: Amsterdam. Nether- Geologic Atlas of the Rocky Mountain region. United
lands. North-Holland Publishing Co . 433 p. States of America: Rocky Mtn. Assoc. Geologists. p.
Almeida. F. F. M. de. 1953. 8otucat6. a Triassic desert of South 146-165.
America: Intemat. Ceol. Cong.. 19th. Algiers 1952. Bagnold. R. A .. 1933. A further joumey through the Libyan
Comptes Rendus. sec. 7. pt. 7. p. 9-24. desert: Geog. Jour . v. 82. p. 103 -129.
_ _ 1954. 8otucat6. a Traissic desert of South America: _ _ 1935. The movement of desert sand: Geog. Jour.. v. 85.
[Brazil] Div. Geologia e Mineralogia Notas Prelim. e p.342-369.
Estudos 86. 21 p. _ _ 1937. The transport of sand by wind: Geog. Jour . v. 89.
Almeida. F. F. N. de. and Barbosa. Octavio. 1953. Geologia das p. 409 --438.
quadri"culas de Piracicaba e Rio Claro estado de 88'0 Paulo _ _ 1941. The physics of blown sand and desert dunes: lon-
[Geology of the Piracicaba and Rio Claro quadrangles. Sib don. Methuen a: Co . Ltd. 265 p.
Paulo]: [Brazil] Div. Geologia e mineralogia 80letrm _ _ 1951. Sand formations in southem Arabia: Geog. Jour ..
143.96p. v. 117. pt. 1. p. 78-86.
American Geological Institute. (Margaret Gary. Robert _ _ 1953. The surface movement of blown sand in relation
McAfee. Jr. and C. L. Wolf. eds.). 1972. Glossary of to meterology. in Desert Research - Internat. Sym-
geology. with a foreword by Ian Campbell: Washington. posium. Jerusalem. 1952. Proc.: Israel Research Council
857 p. Spec. Pub. 2. p. 89--96.

Digitized by Google
410 A STUDY OF GLOBAL SAND SEAS
Baker, A. A., 1936, Geology of the Monument Valley -Navajo July 15 -31,1975: Boletfm Paranaense Geoci~ncias, no. 33,
Mountain region, San Juan County, Utah: U.S. Geol. p.133-167.
Survey Bull. 865, 106 p. Bigarella, J. J., Alessi, A. H., Becker, R. D., and Duarte, G. M.,
Baker, A. A., Dane, C. H., and Reeside, J. B., 1936, Correlation 1969, Textural characteristics of the coastal dune, sand
of the Jurassic formations of parts of Utah, Arizona, New ridge, and beach sediments: Boletfm Paranaense Geoci@n-
Mexico, and Colorado: U.S. Geol. Survey Prof. Paper 183, cias, no. 27, p. 15-00.
66 p. Bigarella, J. J., Becker, R. D., and Duarte, G. M., 1969, Coastal
Bamard, W. S., 1973, Duinformasies in die Sentrale Namib: dune structures from Paran~ [Brazil): Marine Geology, v.
Tegnikon [South Africa, Stallenbosch Univ.), December, 7, p. 5-55.
p.2-13. Bigarella, J. J., and Oliveira, M. A. M., 1966, Preliminary note
Bartenstein, Helmut, 1968, Present status of the Paleozoic about the directions of sand grain transportation at Fumos
paleogeography of northem Germany and adjacent parts and Botucatu in the seventieth part of Paranfl Basin
of northwest Europe, in D. T. Donovan, ed., Geology of [Brazil): Boletfm Paranaense Geografia, nos. 18 -20, p.
shelf seas - Inter-Univ. Geol. Cong., 14th, Proc.: Edin- 247 -256. [In Portuguese.)
burgh, London, Oliver and Boyd, p. 31 -54. Bigarella, J. J., and Popp, J. H., 1966, Contributions to the study
Beard, D. C., and Weyl, P. K., 1973, Influence of texture on of recent shoreline sediments; IV - Shore and dunes of
porosity and permeability of unconsolidated sand: Am. Barra do Sui (SC) [Santa Catarina, Brazil): Boletfm Para-
Assoc. Petroleum Geologists Bull., v. 57, no. 2, p. 349-369. naense Geografia, nos. 18 -20, p. 133 -149. [In Portuguese;
Beheiry, S. A., 1967, Sand forms in the Coachella Valley, EIlBlish abs.)
southem Califomia: Assoc. Am. Geographers Annals, v. Bigarella, J. J., and Salamuni, Riad, 1961, Early Mesozoic wind
57, no. 1, p. 25 -48. pattems as suggested by dune bedding in the Botucatu
Belly, Pierre-Yves, .1964, Sand movement by wind, with Ad- sandstone of Brazil and Uruguay: Geol. Soc. America Bull.,
dendum 2, by Abdel-Latif Kadib: U.S. Army Corps of v. 72, no. 7, p. 1089-1106.
Engineers, Coastal Eng. Research Center Tech. Memo. 1,
------1967, Some paleogeographic and paleotectonic features
[85 pl.
of the Parana Basin, in J. J. Bigarella, R. D. Becker, and I. D.
Bemer, R. A., 1969, Goethite stability and the origin of red
Pinto, eds., Problems in Brazilian Gondwana geology: In-
beds: Geochim. et Cosmochim. Acta, v. 33, no. 2, p.
temat. Symposium on the Gondwana Stratigraphy and
267-273.
Paleontology, 1st, Mar del Plata, 1967; p. 235 -301.
Besler, Helga, 1972, Geomorphologie der Dunen: Namib und
Blanford, W. T., 1877, Geological notes on the Great Indian
Meer, [South-West Africa, Swakopmund), v. 3, p. 25-35.
Desert between Sind and R8jput'na: India Geol. Survey
Bettenay, Eric, 1962, The salt lake systems and their associated
Records, v. 10, p. 10-21.
aeolian features in the semi-arid regions of Westem
Bohdanowicz, Karol, 1892, Geologicheskiia izsliedovaniia v
Australia: Jour. Soil Sci., v. 13, p. 10 -17.
Vostochnam Turkestanie-Geograficheskoe obshchestvo
Beukes, N. J., 1970, Stratigraphy and sedimentology of the Cave
SSSR Tibetskaia ekspeditsii 1889 -1890, Trudy no. 2
Sandstone Stage, Karroo System, in Internat. Union Geol.
[Geological research in eastem Turkestan - Works of
Sci. Symposium, 2d, South Africa, 1970, Proc. and
1889 -1890 Tibet expedition): St. Peterburg, Tipografiia M.
Papers - Gondwana stratigraphy and paleontology:
M. Stasiulevicha, 167 p. [In Russian.)
Marshalltown, Transvaal, South Africa, Geol. Soc. South
Africa, p. 321-341. Boocock, C., and van Straaten, O. J., 1962, Notes on the geology
Beydoun, Z. R., 1966, Geology of the Arabian Peninsula- and hydrogeology of the central Kalahari region,
Eastern Aden Protectorate and part of Dhufar: U.S. Geol. Bechuanaland Protectorate: Geol. Soc. South Africa
Survey Prof. Paper 560 -H, 49 p. Trans., v. 65, pt. 1, p. 125 -176.
Bigarella, J. J., 1965, Sand-ridge structures from Parana coastal Boulet, R., Guichard, E., and Vieillefon, J., 1971, Observations
plain [Brazil): Marine Geology, v. 3, p. 269-278. pedologiques et leurs relations avec les faits geomorphol-
_ _ 1972, Eolian environments - their characteristics, ogiques dans Ie delta su Sl!n~al - Comparison avec les
recognition, and importance, in J. K. Rigby and W. K. observations effectuees au Niger [Pedology and its rela-
Hamblin, eds., Recognition of ancient sedimentary en- tion to geomorphology in the Senegal delta - Comparison
vironments: Soc. Econ. Paleontologists and Mineralogists with observations made on the Niger): Assoc. Silnegal.
Spec. Pub. 16, p. 12 -62. Etude Quaternaire Ouest African, Bull. 29, p. 103-114.
_ _ 1973a, Paleocurrents and the problem of continental Brady, L. F., 1935, Preliminary note on the occurrence of a
drift: Geol. Rundschau, v. 62, no. 2, p. 447 -477. primitive theropod in the Navajo: Am. Jour. Sci., ser. 5, v.
_ _ 1973b, Textural characteristics of the Botucatu 30, no. 177, p. 210-215.
Sandstone: Boletfm Paranaense Geoci@ncias, no. 31, p. _ _ 1939, Tracks in the Coconino sandstone compared with
85-94. those of small living arthropods: Plateau, v. 12, no. 2, p.
_ _ 1975a, Structures developed by dissipation of dune and 32~4.
beach ridge deposits: Catena. v. 2, no. 1-2, p. 107 -152. Bramkamp, R. A., Gierhart, R. D., Brown, G. F., and Jackson,
_ _ 1975b, Lagoa dune field, State of Santa Catarina, R. 0., 1956, Geology of the southem Tuwayq quadrangle,
Brazil- A model of eolian and pluvial activity, in Inter- Kingdom of Saudi Arabia: U.S. Geol. Survey Misc. Geol.
national symposium on the Quaternary, southern Brazil, Inv. Map I -212A, scale 1:500,000.

Digitized by Google
 SELECTED REFERENCES 411
Bramkamp, R. A., Gierhart, R. D., Owens, L. D., and.Ramirez, Campbell, R. L. Jr., 1968. StratigraphiC applications of dip-
L. F., 1963, Geologic map of the western Rub' al Khali meter data in mid-continent: Am. Assoc. Petroleum
quadrangle, Kingdom of Saudi Arabia: U.S. Geol. Survey Geologists Bull.. v. 52. no. 9, p. 1700-1719.
Misc. Geol. Inv. Map 1-218A, scale 1:500,000. Campos, L. F. Gonzaga de, 1889, ~i10 geologica: [Brazil] ~o
Bramkamp, R. A., and Ramirez, L. F., 1958, Geology of the Paulo Commissro Geog. Geol. da Provincia Relatorio,
northern Tuwayq quadrangle, Kingdom of Saudi Arabia: 1889, Annexo, p. 21 --34.
U.S. Geol. Survey Misc. Geol. Inv. Map 1-20'7A, scale, Capot-Rey, Robert, 1943, La morphologie de L'Erg Occidental:
1:500,000. Algiers Univ. Inst. Recherches Sahariennes Travaux, v. 2,
_ _ 19598, Geology of the northwestern Rub' al Khali quad- p.69-103.
rangle, Kingdom of Saudi Arabia: U.S. Geol. Survey Misc. _ _ 1945, Dry and humid morphology in the western Erg:
Geol. Inv. Map I -213A, scale 1:500,000. Geog. Rev., v. 35, no. 3, p. 391--407.
_ _ 1959b, Geology of the Wadi al Batin quadrangle, _ _ 1947a, L'Edeyen de Mourzouk: Algiers Univ. Inst.
Kingdom of Saudi Arabia: U.S. Geol. Survey Misc. Geol. Recherches Sahariennes Travaux, v. 4, p. trl-l09.
Inv. Map 1-203A, scale 1:500,000 [1960]. _ _ 1947b, Chronique - Exploration de L'Erg Oriental:
_ _ 1960, Geographic map of the Jawf-Sakakah quadrangle, Algiers Univ. Inst. Recherches Sahariennes Travaux, v. 4,
Kingdom of Saudi Arabia: U.S. Geol. Survey Misc. Geol. p.181-187.
Inv. Map I -201B, scale 1:500,000. [1961]. _ _ 1953, Le Sahara franCj8is: Paris [France] Presses Univ.,
_ _ 1963, Geologic map of the Darb Zubaydah quadrangle, 584 p.
Kingdom of Saudi Arabia: U.S. Geol. Survey Misc. Geol. Capot-Rey, Robert, and Capot-Rey, Francoise, 1948, Ie deplace-
Inv. Map I -202A, scale 1:500,000. ment des sables ooliens et la formation des dunes df!ser-
Bramkamp, R. A., Ramirez, L. F., Brown, G. F., and Pocock, tiques, d'apres R. A. Bagnold: Algiers Univ. Inst. Recher-
A. E., 1963, Geologic map of the Wadi ar Rimah quad- ches Sahariennes Travaux, v. 5, p. 47--80.
rangle, Kingdom of Saudi Arabia: U.S. Geol. Survey Misc. Carroll, Dorothy, 1958, Role of clay minerals in the transporta-
Geol. Inv. Map 1-206A, scale 1:500,000. tion of iron: Geochim. et Cosmochim. Acta, v. 14, nos.
Breed, C. S., and McKee, E. D., 1973, Sand dunes in desert 1-2, p. 1-27.
areas: Tucson, Ariz., Remote sensing in arid lands, 4th Chaudhri, I. I., 1957, Succession of vegetation in the arid
Ann. Conf., Arizona Univ., Proc., p. 160-171. regions of West Pakistan Plains, in Food and Agricultural
Broin, T. L., 1957. Stratigraphy of the Lykins Formation of Council of Pakistan and UNESCO, Symposium on soil ero-
eastern Colorado: Colorado Univ. Ph. D. thesis. 258 p.; sion and its control in the arid and semiarid zoneS,
available through Univ. Microfilms. Ann Arbor, Mich. Karachi, Nov. 1957: p. 141-158.
Brookfield. Muriel. 1970. Dune trends and wind regime in Chavaillon, Jean, 1964, ttude stratigraphique des formations
central Australia, in Piedmont plains and sand formations Quaternaires du Sahara nord-occidental (Colomb-~har Ii
in arid and humid tropic and subtropic regions: Zeitschr. Reggane): [France] Centre Recherches sur les Zones
Geomorphologie. Supp. 10. p. 121-153. Arides, stlrie ~logie no. 5, 393 p.
Brown. G. F .. 1960. Geomorphology of western and central Chepil, W. S .. 1945, The transport capacity of the wind, Pt. 3 of
Saudi Arabia: Internat. Geol. Cong.. 21st. Copenhagen. Dynamics of wind erosion: Soil Sci., v. 50, no. 6, p.
Rept. 21. pt. 21. p. 150-159. 475--480.
Brown. G. F . Layne. Newton. Goudarzi, G. H . and MacLean. Chepil, W. S., and Woodruff, N. P., 1963, The physics of wind
W. H . 1963, GeologiC map of the northeastern Hijaz quad- erosion and its control, in A. G. Norman, ed., Advances in
rangle. Kingdom of Saudi Arabia: U.S. Geol. Survey Misc. agronomy, v. 15: New York, London, Academic Press, p.
Geol. Inv. Map I -205A. scale 1:500.000. 211-302.
Bryson. R. A . 1974. A perspective on climatic change: Science. Christian, C. S., 1959, An arid zone research institute for India:
v. 184, no. 4138. p. 753 -760. Arid Zone [UNESCO Newsletter], no. 5, p. 6--9.
Bucher. W. H .. 19198. On ripples and related sedimentary sur- Chu, Chen-jen, Lieu, Hua-hsun, Ch'en, En-chiu, Wu. Kung-
face forms and their paleogeographic interpretation: Am. ch'eng, Mi, Kuo-yuan, Hsiao, Yu-ch'uan, Meng, Te-cheng,
Jour. Sci., 4th ser . v. 47. no. 279, p. 149-210. and Hsu. Chen-fu. 1962. The natural characteristics and
_ _ 1919b. Fossil ripples and their paleogeographic in- the melioration of the southwestern section of the
terpretation. pt. 2 of On ripples and related sedimentary Taklamakan Desert of Sinkiang. in Academia Sinica Sand
surface forms and their paleogeographic interpretation: Control GrouP. eds., Research on sand control: U.S. Joint
Am. Jour. Sci., 4th ser., v. 47, no. 280, p. 241-269. Publications Research Service JPRS 19993, Tech. Trans.
Bunker. D. G., 1953. The south-west borderlands of the Rub' al 63 -31183, p. 1 --46. [Figs. omitted.]
Khali: Geog. Jour., v. 119. pt. 4. p. 420--430. Conant, L. C., and Goudarzi, G. H., compilers, 1964. Geologic
Camp. C. L., 1936, A new type of small bipedal dinosaur from map of the Kingdom of Libya: U.S. Geol. Survey Misc.
the Navajo sandstone of Arizona: California Univ. Pubs. Geol. Inv. Map I -350A. scale 1:2.000,000.
Geol. Sci. Bull . v. 24. no. 2, p. 39~5. _ _ 1967, Stratigraphic and tectonic framework of Libya:
Campbell. C. V.. 1971. Depositional model- Upper Cre- Am. Assoc. Petroleum Geologists Bull., v. 51. no. 5, p.
taceous Gallup beach shoreline. Ship Rock area, north- 719-730.
western New Mexico: Jour. Sed. Petrology. v. 41, no. 2. p. Cooke, R. U.. and Warren, Andrew, 1973, Geomorphology in
395--409. deserts: Los Angeles. California Univ. Press, 374 p.

Digitized by Google
412 A STUDY OF GLOBAL SAND SEAS
Cooper. W. S.. 1958. Coastal sand dunes of Oregon and Eckel. E. B. 1959. Geology and mineral resources of
Washington: Geol. Soc. America Mem. 72. 169 p. Paraguay - A reconnaissance. with sections on Igneous
_ _ 1967. Coastal dunes of California: Geol. Soc. America and metamorphic rocks. by Charles Milton and E. B. Eckel.
Mem. 104. 131 p. and Soils. by P. T. Sulsona: U.S. Geol. Survey Prof. Paper
Cordani. U. G . and Vandoros. Paul. 1967. Basaltic rocks of the 327.110 p.
Paran6 Basin. in J. J. Bigarella. R. D. Becker. and 1. D. Pinto. Emmett. W. R.. Beaver. K. W . and McCaleb. J. A . 1972. Penn-
eds.. Problems in Brazilian Gondwana geology: Internat. sylvanian Tensleep reservoir. Little Buffalo Basin oil field.
Symposium on the Gondwana Stratigraphy and Paleon- Big Hom Basin. Wyoming: Mtn. Geologist. v. 9. no. 1. p.
tology. 1st. Mar del Plata. 1967. p. 207 -231. 21-31.
Cornish. Vaughan. 1897. On the formation of sand dunes: Engel. C. G . and Sharp. R. P . 1958. Chemical data on desert
Geog. Jour . v. 9. p. 278~. varnish: Geol. Soc. America Bull . v. 69. no. 5. p. 487 -518.
Cressey. G. B. 1928. The Indiana sand dunes and shore lines of Falke. Horst. ed.. 1972. Rotliegend - Essays on European
the Lake Michigan Basin: Geog. Soc. Chicago. Bull. 8 Lower Permian: Leiden. Netherlands. E. J. Brill &: Co . In-
[Chicago Univ. Press). 80 p. ternat. Sed. Petrographical Ser. v. 15. 299 p.
Crone. A. J. 1974. Experimental studies of mechanically in- Finkel. H. J. 1959. The barchans of southern Peru: Jour.
filtrated clay matrix in sand: Geol. Soc. America Abs. with Geology. v. 67. no. 6. p. 614-647.
Programs. v. 6. no. 7. p. 701. _ _ 1961. The movement of barchan dunes measured by
_ _ 1975. Laboratory and field studies of mechanically in- aerial photogrammetry: Photogrammetric Eng. v. 27. no.
filtrated matrix clay in arid fluvial sediments: Boulder. 3. p. 439-444.
Colo . Colorado Univ. Ph. D. thesis. 162 p.: available from Fisher. R. A . 1936. The use of multiple measurements in tax-
Xerox Univ. Microfilms. Ann Arbor. Mich .. no. 76-3897. onomic problems: Annals of Eugenics. v. 7. p. 179-188.
174 p. Flint. R. F . and Bond. Geoffrey. 1968. Pleistocene sand ridges
Crutcher. H. L. 1957. On the standard vector deviation wind and pans in western Rhodesia: Geol. Soc. America Bull . v.
rose: Jour. Meteorology. v. 14. p. 28-33. 79. no. 3. p. 299-314.
_ _ 1966. Components of 1.000-mb winds (or surface Folk. R. L. 1962. Of skewnesses and sands: Jour. Sed.
winds) of Northern Hemisphere: Washington. U.S. Govt. Petrology. v. 32. no. 1. p. 145-146.
Printing Office. [NAVAIR 50 -IC -51). 74 p.
Crutcher. H. L. and Baer. Ledolph. 1962. Computations from _ _ I968a. Bimodal supermature sandstones - Product of
elliptical wind distribution statistics: Jour. Applied the desert floor: Internat. Geol. Cong . 23d. Prague. 1968.
Meteorology. v.I. p. 522 -530. Proc . Genesis and classification of sedimentary rocks. sec.
Crutcher. H. L. and Meserve. J. M. 1970. Selected level 8. p. 9-32.
heights. temperatures and dew points for Northern Hemis- _ _ 1968b. Petrology of sedimentary rocks: Austin. Tex.
phere: Washington. U.S. Govt. Printing Office. [NAVAIR Hemphill. 170 p.
___1969. Grain shape and diagenesis in the Simpson Desert.
5O-1C-52. Sept. 1968). 410 p.
Cuvillier. J. 1930. R~vision du nummulitique ~yptien: Institut Northern Territory. Australia: Geol. Soc. America Abs.
d'Egypte M~m . v. 16. 371 p. [In French.) with Programs for 1969. pt. 7. p. 88-69.
Dake. C. L. 1921. The problem of St. Peter sandstone: Missouri _ _ 1971a. Longitudinal dunes of the northwestern edge of
Univ. School Mines and Metallurgy Bull . Tech. Ser. v. 6. the Simpson Desert. Northern Territory. Australia. (pt.) 1.
no. 1. 225 p. Geomorphology and grain size relationships: Sedimen-
Dalrymple. P. C . Everett. K. R . and Wollaston. Sarah. 1970. tology. v. 16. p. 5-54.
Environments of Central Asian Highlands: U.S. Army _ _ 1971b. Genesis of longitudinal and oghurd dunes eluci-
Natick Labs. Tech. Rept. 71-19-ES. Earth Sci. Lab. Ser. dated by rolling upon grease: Geol. Soc. America Bull . v.
ES-62. 58 p. 82. no. 12. p. 3461-3468.
Daveau. Suzanne. 1965. Dunes ravinees et dep)ts du Quater- _ _ 1976. Reddening of desert sands. Simpson Desert.
naire Recent dans Ie Sahel Mauritanien: Rev. ~g. Af- Northern Territory. Australia: Jour. Sed. Petrology. v. 46.
rique Occidentale. v.I. no. 2. p. 7 -47. no. 3. p. 604 -615.
Denny. C. S . Warren. C. R. Dow. D. H . and Dale. W. J. 1968. A Folk. R. L. and Ward. W. C . 1957. Brazos River bar - A study
descriptive catalog of selected aerial photographs of in the significance of grain-size parameters: Jour. Sed.
geologic features in the United States: U.S. Geol. Survey Petrology. v. 27. no. 1. p. 3-27.
Prof. Paper 590. 79 p. Fourtau. R..l002.Sur Ie gres nubien: [paris) Acad. Sci. Comptes
Dingler. J. R. 1975. Wave-formed ripples in nearshore sands: Rendus. v. 135. p. 803-604.
San Diego. Calif. California Univ. Ph. D. thesis. 136 p.; Frazier. D. E . and Osanik. A . 1961. Point-bar deposits. Old
available from Xerox Univ. Microfilms. Ann Arbor. Mich . River Locksite. Louisiana: Gulf Coast Assoc. Geol. Socs.
no. 75-11.227. 152 p. Trans . v. 11. p. 121-137.
Dresch. Jean. and Rougerie. G . 1960. Observations morpholo- Frere. H. B. E . 1870. Notes on the Runn [Rann) of Cutch and
giques dans Ie Sahel du Niger: Rev. Geomorphologie Dy- neighboring region: Royal Geog. Soc. [London) Jour. v. 40.
namique. v.11. no. 4-6. p. 49-58. pn French; English abs.) p.181-207.
Dubief. Jean. 1952. Le vent et Ie ~placement du sable au Friedman. G. M. 1961. Distinction between dune. beach. and
Sahara: Algiers Univ. Inst. Recherches Sahariennes Tra- river sands from their textural characteristics: Jour. Sed.
vaux. v. 8. p. 123 -164. Petrology. v. 31. no. 4. p. 514-529.

Digitized by Google
 SELECTED REFERENCES 413
_ _ 1967. Dynamic processes and statistical parameters _ _ 1970. Notes on some major dune types in southern
compared for size frequency distribution of beach and Africa: South African Geog. Jour. v. 52. p. 93 -101.
river sands: Jour. Sed. Petrology. v. 37. no. 2. p. 327 -354. _ _ 1972. Climate. weathering. crust formation. dunes and
_ _ 1966. Textural parameters of beach and dune sands. in fluvial features of the Central Namib Desert. near
Abstracts for 1965: Geol. Soc. America Spec. Papers 87. Gobabeb. South-West Africa: Madoqua. Series II. v.I. no.
p.60. 54 -64. p. 15 -31.
_ _ 1973. Textural parameters of sands - Useful or use- Goudie. A. S . Allchin. Bridget. and Hedge. K. T. M . 1973. The
less?: Geol. Soc. America Abs. with Programs. v. 5. no. 7. former extensions of the Great India Sand Desert: Geog.
p.626-627. Jour . v. 139. pt. 2. p. 243-257.
Friedman. G. M . and Sanders. J. E . 1978. Principles of sedi- Gradzinski. Ryszard. and Jerzykiewicz. Tomasz. 1972. Addi-
mentology: New York. John Wiley &: Sons. tional geographical and geological data from the Polish-
Furst. Manfred. 1970. Beobachtungen an quartaren Buntsedi- Mongolian palaeontological expeditions. in Results of the
men ten der zentralen Sahara [Observations on variegated Polish-Mongolian palaeontological expeditions. Pt. 4:
sediments in the central Sahara]: Abhandlungen des Palaeontologia Polonica. no. 27. p. 17-30.
Heissischen Landesamtes fUr Bodenforschung. no. 56. p. _ _ 19748. Dinosaur- and mammal-bearing aeolian and
129-150. associated deposits of the upper Cretaceous in the Gobi
Garlick. W. G .. 1969. Special features and sedimentary facies of Desert (Mongolia): Sed. Geology. v. 12. p. 249-278.
stratiform sulphide deposits in arenites [with discussion]. _ _ 1974b. Sedimentation of the Barun Goyot Formation. in
in C. H. James. ed. Sedimentary ores. ancient and modem Results of the Polish-Mongolian palaeontological expedi-
[revised]: Inter-Univ. Geol. Cong . 15th. Univ. Leicester tions. Pt. 5: Palaeontologia Polonica. no. 30. p. 111-146.
[England]. Dec. 1967. Proc . Leicester Univ. Dept. Geology
Spec. Pub. 1. p. 107 -169. Greeley. Ronald. Iversen. J. D . Pollack. J. B. Udovich. Nancy.
Garrels. R. M . and Christ. C. L. 1965. Solutions. minerals. and and White. Bruce. 1974. Wind tunnel simulations of light
equilibria: New York. Harper and Row. 450 p. and dark streaks on Mars: Science. v. 183. no. 4127. p.
Gautier. E. F . 1935. Sahara - The Great Desert: New York. 847-849.
Columbia Univ. Press. 280 p. [A translation by D. F. Greenwood. Brian. 1969. Sediment parameters and environ-
Mayhew.] ment discrimination - An application of multivariate
Gees. R. A . 1965. Moment measures in relation to the deposi- statistics: Canadian Jour. Earth Sci.. v. 6. p. 1347 -1358.
tional environments of sands: Eclogae Geol. Helvetiae. v. Gregory. H. E.. 1917. Geology of the Navajo country-A
58. no. 1. p. 209-213. reconnaissance of parts of Arizona. New Mexico. and
Gevers. L. W . 1936. The morphology of western Damaraland Utah: U.S. Geol. Survey Prof. Paper 93.161 p.
and the adjoining Namib Desert of South-West Africa: Griffiths. J. C . 1967. Scientific method in analysis of sedi-
South African Geog. Jour . v. 19. p. 61-79. ments: New York. McGraw-Hill Book Co .. 508 p.
Gill. W. D. 1967. The North Sea basin: World Petroleum Cong. Griffiths. J. F . and Soliman. K. H . 1972. The northern desert
7th. Mexico. 1967. Proc . Origin of oil. geology. and (Sahara). Chap. 3. in H. E. Lansberg. ed .. World survey of
geophysics. v. 2. p. 211-219. climatology. climates of Africa. v. 10: Amsterdam. New
Gilmore. C. W .. 1926. Fossil footprints from the Grand Canyon: York. London. Elsevier Publishing Co .. p. 75-132.
Smithsonian Misc. Colin . v. 77. no. 9. 41 p. Grove. A. T . 1958. The ancient erg of Hausaland. and similar
_ _ 1927. Fossil footprints from the Grand Canyon: second formations on the south side of the Sahara: Geog. Jour . v.
contribution: Smithsonian Misc. Colin . v. 80. no. 3. 78 p. 124. pt. 4. p. 526 -533.
_ _ 1928. Fossil footprints from the Grand Canyon: third _ _ 196Oa. Geomorphology of the Tibesti region. with
contribution: Smithsonian Misc. Colin . v. 80. no. 8. 16 p. special reference to western Tibesti: Geog. Jour .. v. 126. pt.
Gilreath. J. A . and Maricelli. J. J. 1964. Detailed stratigraphic 1. p. 18-31.
control through dip computations: Am. Assoc. Petroleum _ _ 196Ob. [A note] in J. R. V. Prescott and H. P. White.
Geologists Bull . v. 48. no. 12. p. 1902 -1910. Sand formations in the Niger Valley between Niamey and
Glennie. K. W . 1970. Desert sedimentary environments. in Bourem: Ceog. Jour . v. 126. pt. 2. p. 202 -203.
Developments in sedimentology [v.] 14: Amsterdam. lon- _ _ 1969. Landforms and climatic change in the Kalahari
don. New York. Elsevier Publishing Co . 222 p. and Ngamiland: Geog. Jour . v. 135. pt. 2. p. 191 -212.
_ _ 1972. Permian Rotliegendes of northwest Europe in- Grove. A. T .. and Warren. Andrew. 1968. Quaternary land-
terpreted in light of modem desert sedimentation studies: forms and climate on the south side of the Sahara: Geog.
Am. Assoc. Petroleum Geologists Bull.. v. 56. no. 6. p. Jour . v. 134. pt. 2. p. 194 -208.
1048-1071.
Glennie. K. W . and Evans. G .. 1976. A reconnaissance of the Hack. J. T .. 1941. Dunes of the western Navajo country: Geog.
recent sediments of the Ranns of Kutch. India: Sedimen- Rev . v. 31. no. 2. p. 240-263.
tology. v. 23. p. 625-647. Hamblin. W. K . 1961. Micro-cross-Lamination in upper
Goddard. E. N . Chm . and others. 1948. Rock-color chart: Natl. Keweenawan sediments of northern Michigan: Jour. Sed.
Research Council (reprinted by Ceol. Soc. America. 1951. Petrology. v. 31. no. 3. p. 390-401.
1970).6 p. Hand. B. M .. 1967. Differentiation of beach and dune sand.
Goudie. A. S.. 1969. Statistical laws and dune ridges in using settling velocities of light and heavy minerals: Jour.
southern Africa: Geog. Jour . v. 135. pt. 3. p. 404-406. Sed. Petrology. v. 37. no. 2. p. 514-520.

Digitized by Google
414 A STUDY OF GLOBAL SAND SEAS
Hanley. J. H . and Steidtmann. J. R . 1973. Petrology of Homer. N. G . 1936. Geomorphic processes in continental
limestone lenses in the Casper Formation. southemmost basins of central Asia: Intemat. Geol. Cong.. 16th.
Laramie Basin. Wyoming and Colorado: Jour. Sed. Washington. D.C .. 1933. Rept.. v. 2. p. 721-735.
Petrology. v. 43. no. 2. p. 428 -134. Hubert. J. F . 1960. Petrology of the Fountain and Lyons Forma-
Hanley. J. H . Steidtmann. J. R . and Toots. Heinrich. 1971. tions. Front Range. Colorado: Colorado School Mines
Trace fossils from the Casper Sandstone (permian). Quart . v. 55. no. 1. p. 1-242.
southem Laramie Basin. Wyoming and Colorado: Jour. Huntington. Ellsworth. 190'7. Some characteristics of the
Sed. Petrology. v. 41. no. 4. p. 1065 -1068. glacial period in nonglaciated regions: Geol. Soc. America
Harms. J. C . and Fahnestock. R. K. 1965. Stratification. bed Bull .. v. 18. p. 351-388.
forms. and flow phenomena (with an example from the Jones. J. R. 1971. Ground-water provinces of Libyan Arab
Rio Grande). in Primary sedimentary structures and their Republic. in G. Carlyle. ed . Symposium on the geology of
hydrodynamic interpretation - A symposium: Soc. Econ. Libya. Tripoli. 1969: Univ. Libya Fac. Sci. Press. p.
Geologists and Mineralogists Spec. Pub. 12. p. 84 -115. 449-157.
Harms. J. C . MacKenzie. D. B. and McCubbin. D. G . 1963. Jordan. G. F . 1962. Large submarine sand waves: Science. v.
Stratification in modem sands of the Red River. Louisiana: 136. no. 3519. p. 839-648.
Jour. Geology. v. 71. no. 5. p. 566--580. Kidir. Uz16. 1966. Az eolikus felszlni form'k termilszetes
Harshbarger. J. W . Repenning. C. A . and Irwin. J. H . 1957. rendszere [Natural systems of eolian landforms]: Foldrajzi
Stratigraphy of the uppermost Triassic and the Jurassic ~rteslto. v. 15. no. 4. p. 413-146.
rocks of the Navajo country [Colorado Plateau]: U.S. Geol. Kawamura. R .. 1951. Study on sand movement by wind: Tokyo
Survey Prof. Paper 291. 74 p. Uni\<. Inst. Sci. and Technology Rept . v. 5. no. 314.
Hatchell. W.O . 1967. A stratigraphic study of the Navajo Kent. P. E . and Walmsley. P. J. 1970. North Sea progress: Am.
Sandstone. (Upper Triassic(?) -Jurassic). Navajo Moun- Assoc. Petroleum Geologists Bull . v. 54. no. 1. p. 188-181.
tain. Utah and Arizona: Albuquerque. N. Mex . New Mex- Khalil. M. S . 1957. The role of vegetation in checking erosion
ico Univ .. unpub. M.S. thesis. 121 p. in arid and semiarid tracts. in Food and Agricultural
Haughton. S. H . 1969. Geological history of southem Africa: Council of Pakistan and UNESCO. Symposium on soil ero-
Cape Town. Geol. Soc. South Africa. 535 p. sion and its control in the arid and semiarid zones.
Haun. J. D. and Kent. H. C . 1965. Geologic history of Rocky Karachi. Nov. 1957: p. 237 -280.
Mountain region: Am. Assoc. Petroleum Geologists Bull .. Kiersch. G. A . 1950. Small-scale structures and other features
v. 49. no. 11. p. 1781-1800. of Navajo sandstone. northem part of San Rafael Swell.
Hedin. Sven. 1896. A joumey through the Takla-Makan Desert. Utah: Am. Assoc. Petroleum Geologists Bull . v. 34. no. 5.
Chinese Turkistan: Geog. Jour . v. 8. July to December. p. p.923-942.
284-278. King. D. 1960. The sand ridge deserts of South Australia and
_ _ 1899. Through Asia: New York. London. Harper &: related aeolian landforms of Quatemary arid cycles: Royal
Bros . v.I. p. 1-649; v. 2. p. 653-1255. Soc. South Australia Trans .. v. 83. p- 99-106.
_ _ 1905a. The Tarim River. in Scientific results of a Knight. S. H . 1929. The Fountain and Casper formations of the
joumey in central Asia. 1889-1902. v. 1: Stockholm. Laramie Basin - A study on genesis of sediments: Wyom-
Swedish Army Lithographic Institute. 523 p. ing Univ. Sci. Pub .. Geology. v.I. no. 1. p. l-a2.
_ _ 1905b. The central Asian deserts. sand-dunes. and Koch. G. S. Jr . and Link. R. F . 1970. Statistical analysis of
sands. Pt. 2. in Scientific results of a joumey in central geological data: New York. John Wiley &: Sons. 375 p.
Asia. 1899-1902. v. 2: Stockholm. Swedish Army Krishnan. M. S . 1960. Geology of India and Burma [4th ed.]:
Lithographic Institute. p. 379 -718. Madras. India. Higginbothams. Ltd .. 604 p.
Henderson. Junius. 1924. Footprints in Pennsylvanian sand- Krumbein. W. C . and Graybill. F. A .. 1965. An introduction to
stones of Colorado: Jour. Geology. v. 32. no. 3. p. 226-229. statistical models in geology: New York. McGraw-Hill
Hesselberg. T . and Bjorkdal, E . 1929. tiber das Verteilungs- Book Co . 475 p.
gesetz der Windunruhe: Beitraege zur Physik der Frein At- Kuenen. P. H . 1960. Experimental abrasion; (pt.] 4. Eolian ac-
mosphare. v. 15. p. 121-133. pn German.] tion: Jour. Geology. v. 68. no. 4. p. 427 -149.
Hills. E. S. ed. 1966. Arid Lands - A geographical appraisal: Kuenen. P. H . and Perdok. W. G . 1962. Experimental abrasion;
London. Methuen &: Co .. Ltd. and Paris. UNESCO. 461 p. [pt.] 5. Frosting and defrosting of quartz grains: Jour.
Hogue. Mominul. 1975. An analysis of cross-stratification of Geology. v. 70. no. 6. p. 648~.
Gargaresh calcarenite (Tripoli. Libya) and Pleistocene Kukal. Zdenek. 1971. Geology of recent sediments: London.
Paleowinds: Geol. Mag . v. 112. no. 4. p. 393 -101. New York. Academic Press. 490 p.
Holm. D. A . 1953. Dome-shaped dunes of the central Nejd. Laming. D. J. C .. 1966. Imbrication. paleocurrents. and other
Sauda Arabia: Internal. Geol. Cong .. 19th. Algiers. 1952. sedimentary features in the lower New Red Sandstone.
Comptes Rendus. sec. 7. pt. 7. p. 107 -112. Devonshire. England: Jour. Sed. Petrology. v. 36. no. 4. p.
_ _ 1957. Sigmoidal dunes - A transitional form [abs]: 940-959.
Geol. Soc. America Bull .. v. 88. no. 12. pt. 2. p. 1746. Langmuir. Donald. 1971. Particle-size effect on the reaction
_ _ 1960. Desert geomorphology in the Arabian Peninsula: goethite = hematite + water: Am. Jour. Sci.. v. 271. no. 2.
Science. v. 132. no. 3437. p. 1369-1379. p.147-156.
_ _ 1968. Sand dunes. in R. W. Fairbridge. ed . The en- Le Ribault. Lorc. 1974. Leroscopie. m~thode de determination
cyclopedia of geomorphology: New York. Reinhold Book de I'histoire g~logique des quartz d~tritiques [Exoscopy.
Corp .. p. 973 -979. method to determine the geologic history of detrital

Digitized by Google
 SELECTED REFERENCES 415
quartz]: Rev. Geographie Phys. et G&>logie Dynam . v. 16. _ _ 1952. Uppermost Paleozoic strata of northwestern
no. 1. p. 119-130. [In French.] Arizona and southwestern Utah. in Utah Geol. Soc.
Lewis. A. D. 1936. Sand dunes of the Kalahari within the bor- Guidebook to Geology of Utah: no. 7. p. 52 -55.
ders of the Union: South African Geog. Jour . v. 19. p. _ _ 1954. Stratigraphy and history of the Moenkopi Forma-
22-32. tion of Triassic age: Geol.. Soc. America Mem. 61. 133 p.
Leprun. J. C . 1971. Nouvelles observations sur les formations _ _ 1957. Primary structures in some Recent sediments:
dunaires sableuses fix~es du Ferlo nord occidental Am. Assoc. Petroleum Geologists Bull . v. 41. no. 8. p.
(Senegal): Assoc. Senegal. Etude Quarternaire Quest Afri- 1704-1747.
cain. Bull. 31 -32. p. 69 -78. _ _ 1966. Structures of dunes at White Sands National
Lindsay. J. F . 1973. Reversing barchan dunes in Lower Victoria Monument. New Mexico (and a comparison with struc-
Valley. Antarctica: Geol. Soc. America Bull . v. 84. no. 5. p. tures of dunes from other selected areas): Sedimentology.
1799-1806. v. 7. no. 1. Spec. Issue. p. 3-89.
Logan. R. F . 1960. The central Namib Desert. South-West McKee. E. D. and Bigarella. J. J. 1972. Deformational struc-
Africa: Natl. Acad. Sci. - Natl. Research Council Pub. 758. tures in Brazilian coastal dunes: Jour. Sed. Petrology. v. 42.
162 p. no. 3. p. 670-881.
Maack. Reinhard. 1941. Algumas observac~es j} respeito da ex- McKee. E. D. and Breed. C. S . 19748. An investigation of major
istancia e da extensao do arenito superior ~o Bento ou sand seas in desert areas throughout the world. in Third
Caiu' no estado do Parana [Some observations regarding Earth Resources Technology Satellite -1 Symposium.
the existence and enlargement of the Superior sao Bento Washington. D. C . Dec. 10 -14. 1973: Natl. Aeronautics
or Caiu' Desert. Paran']: Curitiba. Paran6. Brazil Mus. Ar- and Space Admin. Spec. Pub. NASA SP -351. p. 665 -879.
quivos Paranaense. v.I. p. 107-129. _ _ 1974b. Preliminary report on dunes. in Skylab 4 visual
Mabbutt. J. A . 1955. Erosion surfaces in Namaqualand and the observations project report. JSC -09053. June 1974. NASA
ages of surface deposits in the south-western Kalahari: TM X-58142. p. 9-1 to 9-9.
Geol. Soc. South Africa Trans. and Proc . v. 58. p. 13-30. _ _ 1976. Sand seas of the world. in R. S. Williams. Jr. and
_ _ 1967. Denudation chronology in central Australia; W. D. Carter. eds. ERTS -1. A new window on our planet:
structure. climate. and landform inheritance in the Alice U.S. Geol. Survey Prof. Paper 929. p. 81-88.
Springs area. in J. N. Jennings and J. A. Mabbutt. eds . _ _ 1977. Desert sand seas. chap. 2. in G. P. Carr and others.
Landform studies from Australia and New Guinea. with a eds.. Skylab explores the Earth: Natl. Aeronautics and
foreword by Prof. E. S. Hills: Canberra. Australian Natl. Space Admin. Spec. Paper 380. [NASA SP - 380]. p. 5 - 48.
Univ. Press. p. 144 -181. McKee. E. D. Breed. C. S . Fryberger. S. G . Gebel. Dana. and
_ _ 1968. Aeolian landforms in central Australia: Aus- McCauley. C. K. 1974. A synthesis of sand seas
tralian Geog. Studies. v. 6. no. 2. p. 139-150. throughout the world - Type III. Final Report for 1 July
Mabesoone. J. M. 1963. Coastal sediments and coastal develop- 1972 -31 March 1974: NASA Earth Resources Survey
ment near cadiz (Spain): Geologie en Mijnbouw. no. 2. p. Program. NASA -CR -139268.95 p. [Goddard Space Flight
29-43. Center. Greenbelt. Md.]
McConnell. R. B.. 1959. Notes on the geology and geo- McKee. E. D. and Douglass. J. R. 1971. Growth and movement
morphology of the Bechuanaland protectorate: Internat. of dunes at White Sands National Monument. New Mex-
Geol. Cong.. 20th. Mexico. 1956. Assoc. de Servicios ico. in Geological Survey research 1971: U.S. Geol. Survey
GtIologicos Africanos. Actas y Trabajos de las reuniones Prof. Paper 750-0. p. DI08-0114.
celebradas en Mexico en 1956. p. 175 -186. McKee. E. D. Douglass. J. R . and Rittenhouse. Suzanne. 1971.
MacGregor. P. A . 1955. Development of the theories of the cir- Deformation of lee-side laminae in eolian dunes: Geol.
culation of the atmosphere over South Africa: South Soc. America Bull . v. 82. no. 2. p. 359-378.
African Geog. Jour . v. 49. p. 41-52. McKee. E. D. and Moiola. R. J. 1975. Geometry and growth of
McKee. E. D.. 19348. An investigation of the light-colored the White Sands dune field. New Mexico: U.S. Geol.
cross-bedded sandstones of Canyon de Chelly. Arizona: Survey Jour. Research. v. 3. no. 1. p. 59-88.
Am. Jour. Sci.. 5th ser. v. 28. no. 165. p. 219-233. McKee. E. D. Oriel. S. S . Berryhill. H. L. Jr . Cheney. T. M .
_ _ 1934b. The Coconino sandstone - Its history and Cressman. E. R. Crosby. E. J. Dixon. G. H . Fix. C. E.
origin: Carnegie Inst. Washington Pub. 440. p. 77 -115. Hallgarth. W. E.. Ketner. K. B.. MacLachlan. M. E .
_ _ 1938. Structures in modern sediments aid in interpret- Maughan. E. K. McKelvey. V. E. Mudge. M. R. Myers. D.
ing ancient rocks: Carnegie Inst. Washington Pub. 501. p. A . and Sheldon. R. P . 1967. Paleotectonic maps of the
683-694. Permian SyStem: U.S. Geol. Survey Misc. Geol. Inv. Map
_ _ 1940. Three types of cross-lamination in Paleozoic 1-450. 184 p.
rocks of northern Arizona: Am. Jour. Sci.. v. 238. no. 11. p. McKee. E. D.. Oriel. S. S . Swanson. V. E. MacLachlan. M. E.
811-824. MacLachlan. J. C . Ketner. K. B. Goldsmith. J. W . Bell. R.
_ _ 1944. Tracks that go uphill [in Coconino sandstone. Y. and Jameson. D. J. 1956. Paleotectonic maps of the
Grand Canyon. Arizona]: Plateau. v. 16. no. 4. p. 61-72. Jurassic system. with a separate section on paleogeogra-
_ _ 1945. Small-scale structures in Coconino sandstone of phy. by R. W. Imlay: U.S. Geol. Survey Misc. Geol. Inv.
northern Arizona: Jour. Geology. v. 53. no. 5. p. 313-325. Map 1-175.
_ _ 1947. Experiments on the development of tracks in fine McKee. E. D. Reynolds. M. A . and Baker. C. H . Jr . 19628.
cross-bedded sand: Jour. Sed. Petrology. v. 17. no. 1. p. Laboratory studies on deformation in unconsolidated sedi-
23-28. ment. in Short papers in geology. hydrology. and topogra-

Digitized by Google
416 A STUDY OF GLOBAL SAND SEAS
phy: U.S. Geol. Survey Prof. Paper 450 - 0, p. 0151- 0155. (pennsylvanian) of Wyoming and its cephalopod fauna:
_ _ 1962b, Experiments on intraformational recumbent Jour. Paleontology, v. 10, no. 8, p. 715-738.
folds in crossbedded sand, in Short papers in geology, hy- Miller, R. L., and Kahn, J. S., 1962, Statistical analysis in the
drology, and topography: U.S. Geol. Survey Prof. Paper geological sciences: New York, John Wiley &: Sons, 483 p.
450 -D, p. 0155 -D180. Moiola, R. J., Spencer, A. B., and McKee, E. 0.,1973, Linear dis-
McKee, E. D., and Tibbitts, G. C., Jr., 1964, Primary structures criminant analysis - A technique for analyzing sand
of a seif dune and associated deposits in Libya: Jour. Sed. bodies: Geol. Soc. America Abs. with Program., 1973 Ann.
Petrology, v. 34, no. I, p. 5 -17. Mtgs., v. 5, no. 7, p. 741-742.
Madigan, C. T., 1936, The Australian sand-ridge deserts: Geog. Moiola, R. J., Spencer, A. B., and Weiser, Daniel, 1974,
Rev., v. 26, no. 2, p. 205-227. Differentiation of modern sand bodies by linear discrimi-
_ _ 1938, The Simpson Desert and its borders: Royal Soc. nant analysis: Gulf Coast Assoc. Geol. Socs. Trans., v. 24,
New South Wales Jour. and Proc. , v. 71, pt. 2, p. 503 -535. p.321-326.
_ _ 1948, The Simpson Desert Expedition, 1939, Scientific Moiola, R. J., and Weiser, Daniel, 1968, Textural parameters-
reports; No.6, Geology - The sand formations: Royal Soc. An evaluation: Jour. Sed. Petrology, v. 38, no. 1, p. 45-53.
South Australia Trans., v. 70, p. 45 -63, 205 -227. Monod, Th~ore, 1958, MajAbat al-KoubrA; Contribution il
Maher, J. C., 1954, Lithofacies and suggested depositional en- l'filde de l"'empty quarter" ouest-saharien: FranCj8is
vironment of Lyons sandstone and Lykins formation in d'Afrique Noire Inst. Mfim. 52, 408 p.
southeastern Colorado: Am. Assoc. Petroleum Geologists _ _ 1961, Maj&bat al-KoubrA (suppI6ment): Fran~ais
Bull., v. 38, no. 10, p. 2233-2239. d'Afrique Noire Inst. Bull. 23, no. 3, p. 591-637.
Manton, W. I., 1968, The origin of associated basic and acid Munsell Color Co., 1954, Munsell soil color charts: Baltimore,
rocks in the Lebombo-Nuanetsi igneous province, Md.
Southern Africa, as implied by strontium isotopes: Jour. Murzayev, E. M., 1967, Nature of Sinkiang and formation of the
Petrology, v. 9, pt. I, p. 23 -39. deserts of central Asia: U.S. Joint Publications Research
Margolis, S. V., and Krinsley, D. H., 1971, Submicroscopic Service JPRS 40,229, IT 67 -30944, 617 p. [Technical
frosting on eolian and subaqueous quartz sand grains: translation from Russian - Priroda sin'tszyana I for-
Geol. Soc. America Bull., v. 82, no. 12, p. 3395-3406. mirovaniye pustyn' tsentral'nay Azii: Akad. Nauk SSSR
Martins, L. R., 1967, Aspectos texturais e deposicionais dos Inst. Geograffi, 1966.]
sedimentos praiais e (,olicas da plan(cie costeira do Rio Nairn, A. E. M., ed., 1961, Descriptive palaeoclimatology: New
Grande do Sui [Brazil]: Escola de Geologia, Rio Grande do York, Interscience Publishers, 380 p.
Sul. Fed. Univ., Spec. Pub. 13, 102 p. _ _ 1964, Problems in palaeoclimatology, in NATO
Marzolf, J. E., 1969, Regional stratigraphic variations in prim- palaeoclimates conf., Univ. Newcastle-upon-Tyne and
ary features of the Navajo Sandstone, Utah: Geol. Soc. Durham, England, 1963, Proc.: New York, Interscience
America Abs. with Programs for 1969, pt. 5, p. 50-51. Publishers, 705 p.
Mason, C. C., and Folk, R. L., 1958, Differentation of beach, Newell, N. D., and Rigby, J. K., 1957, Geological studies on the
dune, and aeolian flat environments by size analysis, Great Bahama Bank, in Regional aspects of carbonate
Mustang Island, Texas: Jour. Sed. Petrology, v. 28, no. 2, p. deposition: Soc. Beon. Paleontologists and Mineralogists
211-226. Spec. Pub. 5, p. 15 -72.
Maughan, E. K., and Wilson, R. F., 1960, Pennsylvanian and Newell, N. D., Rigby, J. K., Whiteman, A. J., and Bradley, J. S.,
Permian strata in southern Wyoming and northern Col- 1951, Shoal-water geology and environments, Eastern
orado, in Guide to the geology of Colorado: Geol. Soc. Andros Island, Bahamas: Am. Mus. Nat. History Bull., v.
America, Rocky Mtn. Assoc. Geologists, and Colorado Sci. 97, art. I, p. 1-29.
Soc., p. 34-12. Norin, Eric, 1941, The Tarim Basin and its border regions:
Meigs, Peveril. 1953, World distribution of arid and semi-arid Leipzig Akad. Verlagsgesellschaft, Regionale Geologie der
homoclimates, in Reviews of research on arid zone hy- Erde, v. 2, pt. 4b, 40 p.
drology: Paris, Arid Zone Programme -I, UNESCO, p. Norris, R. M., 1968, Barchan dunes of Imperial Valley, Califor-
203-209. nia: Jour. Geology, v. 74, no. 3, p. 292-306.
Meinster, B., and Tickell, S. J., 1976, Precambrian aeolian _ _ 1969, Dune reddening and time: Jour. Sed. Petrology, v.
deposits in the Waterberg Supergroup: South Africa Geol. 39, no. I, p. 7 -11.
Soc. Trans., p. 191-199. Norris, R. M., and Norris, K. S., 1961, Algodones dunes of
Melton, F. A., 1940, A tentative classification of sand dunes- southeastern California: Geol. Soc. America Bull., v. 72,
Its application to dune history in the southern High no. 4, p. 805 -620.
Plains: Jour. Geology, v. 48, no. 2, p. 113 -174. Obrutschev, V. A., 1895, [Uber] Ueber die verwitterungs und
Merk, G. P., 1960, Great sand dunes of Colorado, in Guide to deflationsprocesse in Central-Asien: St. Petersburg,
the geology of Colorado: Geol. Soc. America, Rocky Mtn. Russisch-Kaiserlich Mineralogische Gesell. Verb., ser. 2,
Assoc. Geologists, and Colorado Sci. Soc., p. 127 -129. no. 33, p. 229 -272.
Merriam, Richard, 1969, Source of sand dunes of southeastern Opdyke, N. D., 1961, The palaeoclimatological significance of
California and northwestern Sonora, Mexico: Geol. Soc. desert sandstone, chap. 3, in A. E. M. Nairn, ed., Descrip-
America Bull., v. 80, no. 3, p. 531-533. tive Palaeoclimatology: New York, Interscience Pub-
Miller, A. K., and Thomas, H. D., 1938, The Casper Formation lishers, p. 45-60.

Digitized by Google
 SELECTED REFERENCES 417
Opdyke. N. D. and Runcorn. S. K. 1960. Wind direction in the Durham. England. 1983. Proc.: New York. Interscience
Western United States in the Late Paleozoic: Geol. Soc. Publishers. p. 394-405 [l965J.
America Bull . v. 71. no. 7. p. 959-9'71. Powers. R. W . Ramirez. L. F.. Redmond. C. D. and Elberg. E.
Pacheco. J. A. do Amaral. 1913. Notas sobre a geologia do valle L.. Jr . 1988. Geology of the Arabian Peninsula - Sedi-
do Rio Grande a partir da f6z do Rio Pardo at6 a sua con- mentary geology of Saudi Arabia: U.S. Geol. Survey Prof.
fluencia com 0 Rio Paranahyba: explorasi'o do Rio Grande Paper 560 -0. 147 p.
e de seus affluentes: [BrazilJ Sil'o Paulo Commissao Geog. Pramanik. S. K.. 1952. Hydrology of the Rajasthan Desert-
e Geol.. p. 33 ~8. Rainfall. humidity. and evaporation. in Symposium on
Page. W. D. 1972. The geological setting of the archaeological the Rajputana Desert. India. 1952. Proc.: India Nan. Inst.
site at Qued el Akarit and the paleoclimatic significance of Sci. Bull. 1. pt. 4. Meteorology and Hydrology. p. 183 -197.
gypsum soils. southern Tunisia: Boulder. Colo . Colorado Prescott. J. R. V .. and White. H. P.. 1960. Sand formations in the
Univ. unpub. Ph. D. thesis. 111 p.: available from Xerox Niger Valley between Niamey and Bourem: Geog. Jour . V.
Univ. Microfilms. Ann Arbor. Mich . no. 73 -1812. 121 p. 126. pt. 2. p. 200 -203.
Peabody. F. E . 1940. Trackways of Pliocene and Recent sala- Price. W. A . 1950. Saharan sand dunes and the origin of the
mandroids of the Pacific Coast of North America: longitudinal dunes - A review: Geog. Rev . V. 40. no. 3. p.
Berkeley. Calif. California Univ . unpub. M.A. thesis. 462--465.
Pederson. S. L. 1953. The stratigraphy of the (pennsylvanian Pryor. W. A . 1971a. Petrology of the Permian Yellow Sands of
and Permian) Casper and Fountain formations of northeastern England and their North Sea basin
southeastern Wyoming and north-central Colorado: equivalents: Sed. Geology. V. 6. p. 221 -254.
Laramie. Wyo . Wyoming Univ . unpub. M.A. thesis. 87 p. _ _ 1971b. Petrology of the Weissliegendes sandstones in
Peirce. H. W . 1983. Stratigraphy of the De Chelly Sandstone of the Harz and Werra-Fulda areas. Germany: Geol.
Arizona and Utah: Tucson. Ariz. Arizona Univ . Ph. D. Rundschau. V. 60. no. 2. p. 524-551.
thesis. 206 p.: available from Univ. Microfilms. Ann Ar- _ _ 1973. Permeability-porosity patterns and variations in
bor. Mich . no. 63 -3650. 291 p. some Holocene sand bodies: Am. Assoc. Petroleum
Pesce. Angelo. 1988. Gemini space photographs of Libya and Geologists Bull . V. 57. no. 1. p. 162 -189.
Tibesti - A geological and geographical analysis: Tripoli. Qadri. S. M. A . 1957. Wind erosion and its control in Thar. in
Petroleum Exploration Soc. Libya. 81 p. Food and Agricultural Council of Pakistan and UNESCO.
Peterson. J. A . 1972. Jurassic System. in Geologic atlas of the Symposium on soil erosion and its control in the arid and
Rocky Mountain region. United States of America: Rocky semiarid zones. Karachi. Nov. 1957: p. 169 -184.
Mtn. Assoc. Geologists. p. 177 -189. Raisz. Erwin. 1952. Landform map of North Africa: Boston
Peterson. J. A . and Osmund. J. C . eds . 1961. Geometry of [Mass.J. 02114.130 Charles St.. scale 1:3.801.800.
sandstone bodies - A symposium. 45th Ann. Mtg. Atlan- Ramirez. L. F .. Elberg. E. L. Jr.. and Helley. H. H .. 1983a.
tic City. N. J. 1960: Tulsa. Okla. Am. Assoc. Petroleum Geologic map of the south central Rub' al Khali quad-
Geologists. 240 p. rangle. Kingdom of Saudi Arabia: U.S. Geol. Survey Misc.
Petrov. M. P . 1961. Geographic explorations in the deserts of Geol. Inv. Map 1- 219A. scale 1:500.000.
Central Asia: U.S. Joint Publications Research Service _ _ 1963b. GeologiC map of the southeastern Rub' al Khali
JPRS 9221.31 p. [Technical translation from Russian.J quadrangle. Kingdom of Saudi Arabia: U.S. Geol. Survey
_ _ 1962. Types de deserts de l' Asie centrale: Annales de Misc. Geol. Inv. Map I-220A. scale 1:500.000.
Geographie. no. 384. v. 71. p. 131-155. Range. Paul. 1927. Geologie der Kiistenwiiste Siidwestafrikas
_ _ 1988. The Ordos. Alashan and Peishan. v. 1 of The zwichen dem Kuiseb und der Lilderitzbucht - Eisenbahn:
deserts of Central Asia: U.S. Joint Publications Research Internat. Geol. Cong . 14th. Madrid. 1926. Comptes Ren-
Service JPRS 39.145. p. 241-260. [Translated from Rus- dus. V. 3. p. 901 -005.
sian.J Ratner. Benjamin. 1950. A method for eliminating directional
_ _ 1967. The Hoshi Corridor. Tsaidam and Tarim Basin. v. bias in wind roses: Monthly Weather Review. v. 78, no. 10,
2 of The deserts of Central Asia: U.S. Joint Publications p.185-188.
Research Service JPRS 42.772. 335 p. [Translated from Rao, K. N., 1958. Some studies on rainfall of Rajasthan with
Russian.J particular reference to trends: Indian Jour. Meteorology
Pettijohn. F. J. Potter. P. E. and Siever. Raymond. 1972. Sand and Geophysics, v. 9, no. 2, p. 97 -116.
and sandstone: Berlin. New York. Springer-Verlag. 618 p. Rawson, R. R., and Turner, C. E., 1974, The Toroweap Forma-
Poldervaart. Arie. 1955. Kalahari sands: Pan-African Congo on tion - A new look, in T. N. V. Karlstrom. G. A. Swann,
Prehistory. 3rd. Livingstone. Northern Rhodesia. 1955. p. and R. L. Eastwood. eds., Geology of northern Arizona,
106-114. with notes on archaeology and paleoclimate. pt. 1 of
Poole. F. G . 1962. Wind directions in late Paleozoic to middle Regional studies: Geol. Soc. America Rocky Mtn. Sec.
Mesozoic time on the Colorado Plateau. in Short papers in [Guidebook 27, pt. IJ, p. 155-190.
geology. hydrology. and topography: U.S. Geol. Survey Read, C. B., 1951, Stratigraphy of the outcropping Permian
Prof. Paper 450-0. p. D147 -0151. rocks around the San Juan Basin, in New Mexico Geol.
_ _ 1964. Palaeowinds in the Western United States. in A. Soc. Guidebook 2d Field Conf., The south and west sides
E. M. Nairn. ed . Problems in palaeoclimatology: NATO of the San Juan Basin, New Mexico and Arizona, 1951: p.
Palaeoclimates Conf. Univ. Newcastle-upon-Tyne and 60-84.

Digitized by Google
418 A STUDY OF GLOBAL SAND SEAS

Read, C. B., and Wanek, A. A., 1961, Stratigraphy of outcrop- features on Mars, 2, Mariner 9 global results: Jour. Geo-
ping Permian rocks in parts of northeastem Arizona and phys. Research, v. 78, no. 20, p. 4163-4196.
adjacent areas: U.S. Geol. SUlVey Prof. Paper 374-H, 10 p. Sahu, B. K., 1964, Depositional mechanisms from the size
Reiche, Parry, 1938, An analysis of cross-lamination - The analysis of clastic sediments: Jour. Sed. Petrology, v. 34,
Coconino sandstone: Jour. Geology, v. 46, no. 7, p. no. 1, p. 73 - 83.
905-932. Salamuni, Riad, and Bigarella, J. J., 1967, The BotucatU Forma-
Reineck, H. E., and Singh, I. B., 1973, Depositional sedimentary tion, in J. J. Bigarella, R. D., Becker, and I. D., Pinto, eds .
environments with reference to terrigenous clastics: New Problems in Brazilian Gondwana geology: Intemat. Sym-
York, Heidelberg, Berlin, Springer-Verlag, 439 p. posium on the Gondwana Stratigraphy and Paleontology.
Reynolds, M. W., Ahlbrandt, T. S., Fox, J. E., and Lambert, P. 1st. Mar del Plata, 1961, p. 197 -206.
W., 1975, Description of selected drill cores from Paleozoic Schiffers, Heinrich. 1971. Die Sahara und ihre Randgebiete:
rocks, Lost Soldier oil field, south-central Wyoming: pt. I, Darstellung eines Naturgrossraumes Gesamtbearbeitung,
Wells 114A: tract 13, C -128: tract 4, C -14: tract 10, T -1: Bd. 1, Physiogeographie: Munich, Weltfonun Verlag,
tract 10, C -2: U.S. Geol. Survey Open-File Rept. 75 -662, 674p.
73 p. Schlee, John. Uchupi. Elazar. and Trumbull. J. V. A., 1964.
_ _ 1976, Description of selected drill cores from Paleozoic Statistical parameters of Cape Code beach and eolian
rocks, Wertz oil field, south-central Wyoming: pt. I, Wells sands. in Geological Survey research 1964: U.S. Geol.
27 ABC: 46 ABC8t:E: 42 B: West Wertz 2: U.S. Geol. Survey Survey Prof. Paper 501~. p. 0118 ~122.
Open-File Rept. 76 -377,56 p. Schuchert. Charles. 1918. On the Carboniferous of the Grand
_ _ 1977, Preliminary description of cores from selected Canyon of Arizona: Am. Jour. Sci . 4th ser. v. 45. p.
wells, Wertz oil field, south-central Wyoming: pt. 2: U.S. 347-361.
Geol. Survey Open-File Rept. Schulze. R. R . 1965. General SUlVey. pt. 8 of Climate of South
Reynolds, M. W., and Fox, J. E., 1973, Description of selected Africa: South Africa Weather Bur. Pub. WB 28. 322 p.
drill cores from Paleozoic and Precambrian rocks, Lost _ _ 1972. South Africa. chap. 15. in H. E. Lansburg. ed.
Soldier oil field. south-central Wyoming: pt. II. wells tract World survey of climatology. climates of Africa. v. 10:
11, T -1: l09A: tract 13, C -115; 103; tract 13, C -127: tract Amsterdam. New York. London. Elsevier Publishing Co.
14. T -11; tract 9. M -1: U.S. Geol. Survey Open-File Rept. p.501-586.
76-64,50 p. Seely. M. K., and Sandelowsky. B. H . 1974. Dating the regres-
Rice, C. M . 1951. Dictionary of geological terms (exclusive of sion of a river's end point: South Africa Archeological Soc.
stratigraphic formations and paleontologic genera and Bull . Goodwin Series no. 2. p. 61 -64.
species): Ann Arbor. Mich., Edwards Bros . 465 p. Selley. R. C . 1968. A classification of paleocurrent models:
Ries, Heinrich, and Conant, G. 0 . 1931. The character of sand Jour. Geology. v. 76. no. 1. p. 99-110.
grains: Am. Foundrymen's Assoc. Trans., v. 2. no. 10. p. _ _ 1970, Ancient sedimentary environments - A brief
353-392. survey: Ithaca. N.Y., Comell Univ. Press. 237 p.
Rigby. J. K. and Hamblin. W. K., eds., 1972. Recognition of an- Seth. S. K., 1963. A review of evidence conceming changes of
cient sedimentary environments: Soc. Econ. Paleon- climate in India during the protohistorical and historical
tologists and Mineralogists Spec. Pub. 16, 340 p. periods. in Changes of climate: UNESCO and World
Rittenhouse. Gordon, 1971. Pore-space reduction by solution Meteorol. Organization Symposium. Rome. 1961. Proc.
and cementation: Am. Assoc. Petroleum Geologists Bull . Arid Zone Research [v.] 20, p. 443-454.
v. 55, no. 1, p. 80-91. Sevon. W. D . 1966. Distinctions of New Zealand beach. dune.
Rogers, A. W., 1934. The build of the Kalahari: South African and river sands by their grain-size distribution charac-
Geog. Jour .. v. 17, p. 3 -12. teristics: New Zealand Jour. Geology and Geophysics. v. 9.
_ _ 1936. The surface geology of the Kalahari: Royal Soc. p.212-223.
South Africa Trans .. v. 24. p. 57-aQ. Sharp. R. P . 1963. Wind ripples: Jour. Geology. v. 71. no. 5. p.
Rogers, John. and Tankard, A. J. 1974, Surface textures of some 617-636.
quartz grains from the west coast of southem Africa: _ _ 1966. Kelso dunes. Mohave Desert. Califomia: Geol.
Southem Africa Electron Microscopy Soc. Proc., v. 4. p. Soc. America Bull . v. 77. no. 10. p. 1045-1073.
55-56. Shea. J. H . 1974. Deficiencies of clastic particles of certain
Roy. B. B., 1969. Technical report - Progress of research, in sizes: Jour. Sed. Petrology. v. 44. no. 4. p. 985-1003.
Central Arid Zone Research Inst.. Jodhpur, Ann. Rept.: p. Shepard. F. P . and Young. Ruth. 1961. Distinguishing between
12-20. beach and dune sands: Jour. Sed. Petrology. v. 31. no. 2. p.
Royal Navy and South African Air Force. 1944. Introduction: 196-214.
The west coast of Africa from River Congo to Olifants Shergold. J. H . Druce. E. C., Radke. B. M . and Draper. J. J.
River. with an appendix on the climates of St. Helena and 1976. Cambrian and Ordovician stratigraphy of the eastem
Tristan da Cunha. pt. 1. in Weather on the coasts of portion of the Georgina Basin. Queensland and eastem
southem Africa. v. 2: p. iii -viii; pt. 1. p. 1-61. Northem Territory: Intemat. Geol. Cong. 25th. Australia.
Rumney. G. R .. 1968. Climatology and the world's climates: 1976. Field Excursion 4C Guidebook. 54 p.
New York, Macmillan Co .. 656 p. Sherzer, W. H . and Grabau. A. W., 1909. The Sylvania
Sagan. C . Veverka. J.. Fox. P .. and others. 1973. Variable Sandstone; its distribution. nature and origin. chap. III. in

Digitized by Google
 SELECTED REFERENCES 419
A. W. Grabau and W. H. Sherzer. The Monroe formation stratification in sandstone of the Casper Formation.
of southern Michigan and adjoining regions: Michigan southernmost Laramie Basin. Wyoming: Geol. Soc.
Geol. and BioI. Survey Pub. 2. p. 61 - 66. America Bull.. v. 85. no. 12. p. 1835-1842.
Shotton. F. W .. 1937. The Lower Bunter sandstones of north Stokes. W. L. 1961. Fluvial and eolian sandstone bodies in Col-
Worcestershire and east Shropshire [England]: Geol. Mag. orado Plateau. in J. A. Peterson and J. C. Osmond. eds .
no. 662. v. 74. no. 12. p. 534-553. Geometry of sandstone bodies - A symposium. 45th Ann.
_ _ 1956. Some aspects of the New Red desert in Britain: Mtg. Atlantic City. N.J . April 25-26. 1960: Tulsa. Okla .
Liverpool and Manchester Geol. Jour . v.I. pt. 5. p. Am. Assoc. Petroleum Geologists. p. 151 -178.
450-465. Stone. R. 0 .. 1967. A desert glossary: Earth-Sci. Rev . v. 3. no. 4.
Simons. F. S . 1956. A note on Pur-Pur dune. Vim Valley. Peru: p.211-266.
Jour. Geology. v. 64. no. 5. p. 517 -521. Strahler. A. N . 1956. Dimensional analysis applied to fluvially
Singh. S . Ghose. B. and Vats. P. C . 1972. Genesis. orientation. eroded landforms: Geol. Soc. America Bull . v. 69. no. 3. p.
and distribution of sand dunes in arid and semi-arid 279-299.
regions of India: Central Arid Zone Research Inst.. Stride. A. H . 1963. Current-swept sea floors near the southern
Jodhpur. Ann. Rept.. p. 50 -53. half of Great Britain: Geol. Soc. London Quart. Jour . v.
Sinitsyn. V. M .. 1959. Tsentralnaya Aziya: Moscow. Gosudar. 119. pt. 2. no. 474. p. 175 -199.
Izd. Geog. Lit.. 456 p. Suslov. S. P . 1961. Physical geography of Asiatic Russia
Smith. D. B. 1972. The Lower Permian in the British Isles. in [English translation by N. D. Gershevsky]: San Francisco.
Horst Falke. ed . Rotliegend - Essays on European Lower London. W. H. Freeman a: Co . 594 p. [First published in
Permian: Leiden. Netherlands. E. J. Brill a: Co . p. 1-33. Russian in Leningrad and Moscow. 1947.]
Smith. D. B. and Francis. E. A . 1967. Geology of the country Tanner. W. F . 1964. Eolian ripple marks in sandstone: Jour.
between Durham and West Hartlepool [England]: Great Sed. Petrology. v. 34. no. 2. p. 432-433.
Britain Geol. Survey Mem . 354 p. _ _ 1965. Upper Jurassic paleogeography of the Four Cor-
Smith. H. T. U . [1943]. Aerial photographs and their applica- ners region: Jour. Sed. Petrology. v. 35. no. 3. p. 564-574.
tions: New York. D. Appleton-Century Co. 372 p. _ _ 1968. Numerous eolian ripple marks from Entrada For-
_ _ 1946. Sand dunes: New York Acad. Sci. Trans . ser. 2. v. mation: Mtn. Geologist. v. 3. no. 3. p. 133 -134.
8. no. 6. p. 197 -199. _ _ 1967. Ripple mark indicies and their uses: Sedimen-
_ _ 1953. Classification of sand dunes [abs.]. in ~rts ac- tology. v. 9. no. 2. p. 89 -104.
tuels et anciens: Internat. Geol. Cong . 19th. Algiers. 1952. Terwindt. J. H. J. 1971. Sand waves in the southern bight of the
Comptes Rendus. sec. 7. p. 105. pn English.] North Sea: Marine Geology. v. 10. no. 1. p. 51 ~7.
_ _ 1956. Giant composite barchans of the northern Peru- Thesiger. W . 1949. A further journey across the Empty
vian deserts [abs.]: Geol. Soc. America Bull . v. 67. no. 12. Quarter: Geog. Jour . v. 113, p. 21-46.
pt. 2. p. 1735. Thompson. D. B.. 1969. Dome-shaped aeolian dunes in the
_ _ 1963. Eolian geomorphology. wind direction. and Frodsham Member of the so-called "Keuper" Sandstone
climatic change in North Africa - final report: U.S. Air Formation (Scythian-7 Anisian; Triassic) at Frodsham.
Force Cambridge Research Lab.. Geophysics Research Cheshire (England): Sed. Geology. v. 3. no. 4. p. 263 -289.
Directorate [Contract AF 19(628) -298]. AFCRL~-443. _ _ 19708. Sedimentation of the Triassic (Scythian) red
49p. pebbly sandstones in the Cheshire Basin and its margins:
_ _ 1965. Dune morphology and chronology in central and Geol. Jour., v. 7. pt. 1. p. 183-261.
western Nebraska: Jour. Geology. v. 73. no. 4. p. 557 -578. _ _ 1970b. The stratigraphy of the so-called "Keuper"
_ _ 1968. Nebraska dunes compared with those of North Sandstone Formation (Scythian-7 Anisian) in the Permo-
Africa and other regions. in C. B. Schultz and J. C. Frye. Triassic Cheshire Basin: Geol. Soc. London Quart. Jour . v.
eds . Loess and related eolian deposits of the world: Inter- 126. p. 151-181.
nat. Assoc. Quaternary Resarch Cong.. 7th. Colorado. Thompson. M. L. and Thomas. H. D. 1953. Systematic paleon-
U.S.A . 1965. Proc . v. 12. p. 29-47. tology of fusulinids from the Casper Formation: Wyoming
_ _ 1969. Photo-interpretation studies of desert basins in Geol. Survey Bull. 46. pt. 2. p. 15-56.
northern Africa - Final report (pt. 1). March 1963 -July Thompson. W.O . 1949, Lyons sandstone of Colorado Front
1968: U.S. Air Force Cambridge Research Lab .. Range: Am. Assoc. Petroleum Geologists Bull . v. 33. no. 1.
AFCRL ~ ~590. 60 p. p.52-72.
Snead. R. E . and Frishman. S. A . 1968. Origin of sands on the _ _ 1954. Lyons sandstone in Denver basin [Colo.] [abs.]:
east side of the Las Bela valley. West Pakistan: Geol. Soc. Am. Assoc. Petroleum Geologists Bull .. v. 38. no. 6, p.
America Bull . v. 79. no. 11. p. 1671-1676. 1311.
South African Weather Bureau. 1960. Climate of South Africa; Tieje. A. J.. 1923. The redbeds of the Front Range in Col-
pt. 6. Surface winds: Pretoria. 202 p. orado - A study in Sedimentation: Jour. Geology. v. 31.
Stanley. K. 0 .. Jordan. W. M. and Dott. R. H .. Jr. 1971. New hy- no. 3. p. 192-207.
pothesis of Early Jurassic paleogeography and sediment Tillman. R. W .. 1971. Multiple group discriminant analysis of
dispersal for Western United States: Am. Assoc. grain size data as an aid in recognizing environments of
Petroleum Geologists Bull . v. 55. no. 1. p. 10-19. deposition: Internal. Sedimentological Cong.. 8th.
Steidtmann. J. R . 1974. Evidence for eolian origin of cross- Heidelberg. Program with Abs . p. 102.

Digitized by Google .
420 A STIJDY OF GLOBAL SAND SEAS

Trask. P. D. 1931. Compaction of sediments: Am. Assoc. _ _ I968a. On the onglO of longitudinal (seif) dunes:
Petroleum Geologists Bull.. v. 15. no. 3. p. 271-276. Zeitschr. Geomorphologie. v. 12. no. 2. p. 200-220.
Tricart. Jean. 1959. ~morphologie dynamique de la moyenne _ _ 1968b. Geomorphology at the Central Arid Zone
vallee du Niger (Soudan): Annales G~graphie. no. 368. p. Research Institute. India: Nature and Resources. v. 4. no. 3.
333-343. p.6-9.
_ _ 1965. Rapport de la mission de reconnaissance _ _ 1970. Aeolian geomorphology of the Thar Desert and
8eomorphologique de la vall6e moyenne du Niger: Institut paleo-climates. in Piedmont plains and sand-fonnations in
Franyais d'Afrique Noire Mtim. 72. 196 p. arid and humid tropic and subtropic regions: Zeitschrift
Tricart. Jean. and Brochu. Michel. 1955. Le grand erg ancien du fur Geomorphologie. supp. v. 10. p. 104-120.
Trarza et du Cayor (Sud-Ouest de la Mauritanie et Nord Verstappen. H. T.. and van Zuidam. R. A.. 1970. Orbital
du senflgal): Rev. Geomorphologie Dynam . v. 4. p. photography and the geosciences - A geomorphological
145-176. example from the Central Sahara: Geoforum. v. 2. p.
Twenhofel. W. H . 1926. Treatise on sedimentation: Baltimore. 33-47.
Md . The Williams &: Wilkins Co . 661 p. Visher. G. S .. 1971. Depositional processes and the Navajo
_ _ 1945. The rounding of sand grains: Jour. Sed. Petrology. Sandstone: Geol. Soc. America Bull.. v. 82. no. 5. p.
v. 15. no. 2. p. 59-71. 1421-1423.
Twidale. C. R. 1972. Evolution of sand dunes in the Simpson Wadia. D. N . 1939. Geology of India: London. Macmillan and
Desert. central Australia: Inst. British Geographers Pub. Co . 460 p.
56. [reprinted from Trans .. July 1972]. p. 77-100. _ _ 1953. Geology of India [3d ed.]: London. Macmillan and
Udden. J. A . 1898. The mechanical compositions of wind Co . 531 p.
deposits: Augustana Library Pub. no. 1.69 p. Walker. T. R . 1967. Fonnation of red beds in modem and an-
_ _ 1914. Mechanical composition of clastic ...sediments: cient deserts: Geol. Soc. America Bull . v. 78. no. 3. p.
Geol. Soc. America Bull . v. 25. p. 655 -744. 353-368.
U.S. Air Force [MATS]. 1956. Aeronautical chart 541. Sebha _ _ 1976. Diagenetic origin of continental red beds. in Horst
Oasis. Algeria-Libya. Falke. ed .. The continental Pennian in west. central and
U.S. Anny Air Forces Weather Division. 1945. Weather and cli- south Europe: NATO Advanced Study Inst.. Mainz. Ger-
mate of China. Pts. A and B: Rept. 890; U.S. Air Weather many. 1975. Proc . p. 240 -282.
Service Tech. Rept. 105--34. Walker. T. R. and Crone. A. J. 1974. Mechanically-infiltrated
U.S. Geological Survey. 1963. Map of the Arabian Peninsula: clay matrix in desert alluvium: Geol. Soc. America Abs.
U.S. Geol. Survey Misc. Geol. Inv Map 1-270 B-2. scale with Programs. v. 6. no. 7. p. 998.
1:200.000. Walker. T. R . and Hanns. J. C . 1972. Eolian origin of flagstone
U.S. Naval Weather Service. 1967. World-wide airfield summ- beds. Lyons Sandstone (Pennian). type area. Boulder
aries - Middle East: Asheville. N.C . U.S. Nat!. Weather County. Colorado. in Environments of sandstone. carbon-
Records Center, v. 2. pt. 1. p. 1-334; v. 2. pt. 2, p. 335--698. ate. and evaporite deposition: Mtn. Geologist. v. 9. no. 2-3.
_ _ 1968. World-wide airfield summaries - Africa (north- p.279-288.
ern half): Asheville, N.C . U.S. Nat!. Weather Records Walker. T. R .. and Honea. R. M. 1969. Iron content of modem
Center. v. 9. pt. 1. 559 p. deposits in the Sonoran desert - A contribution to the
_ _ 1974a. World-wide airfield summaries - China. North origin of red beds: Geol. Soc. America Bull.. v. 80. no. 3. p.
Korea. Mongolia: Asheville, N.C., U.S. Nat!. Weather 535-543.
Records Center, v. 12. pt. 1. 371 p.; v. 12. pt. 2.322 p. Wallington. C. E .. 1968. A method of reducing observing and
_ _ 1974b. World-wide airfield summaries - Middle East: procedure bias in wind-direction frequencies: Meteor-
Asheville. N.C .. U.S. Nat!. Weather Records Center. v. 2. ological Mag . v. 97. no. 1155. p. 293 -302.
pt. 2. 338 p. Walther. Johannes. 1888. [ijber] Ueber Ergebnisse einer
U.S. Weather Bureau and U.S. Anny Air Forces, 1944, North- Forschungsreise auf der Sinaihalbinsel und in der
ern Africa: U.S. Weather Bureau Prelim. Rept. 41. p. 1-23. Arabischen WUite: Berlin Gesellschaft fiir Erdkunde
Urvoy. Y., 1935. Terasses et changements de climat quater- Verh . 15. p. 244 -255.
naires a I'est du Niger: Annales de G60graphie. v. 44. no. Warren. Andrew. 1969. A bibliography of desert dunes and
249. p. 254 -263. associated phenomena. in W. G. McGinnies and B. J. Gold-
_ _ 1942. Les bassins du Niger; etude de g'OOgraphie phys- man. eds . Arid lands in perspective: Am. Assoc. Adv. Sci.
ique et de pal~ogeographie: Institute Fransais de L'Afrique and Arizona Univ. Press. p. 75-99.
Noire. Mem. no. 4. 141 p. _ _ 1970. Dune trends and their implications in the central
Vail. C. E.. 1917. Lithologic evidence of climatic pulsations: Sudan. in Piedmont plains and sand-fonnations in arid
Science. New Ser.. v. 46, p. 90-93. and humid and subtropic regions: Zeitschrift fUr Geo-
Van de Kamp. P. C., 1973, Holocene continental sedimentation morphologie. supp. v. 10. p. 154 -180.
in the Salton basin, California - A reconnaissance: Geol. _ _ 1972. Observations on dunes and bi-modal sands in the
Soc. America Bull .. v. 84, no. 3,827--&48. Tenere Desert: Sedimentology. v. 19. no. 1 - 2. p. 37 - 44.
Verstappen, H. T .. 1966. Landfonns. water, and land use west Washburne. C. W .. 1930. Petroleum geology of the State of S8'0
of the Indus Plain: Nature and Resources. v. 2, no. 3. p. Paulo. [Brazil]: 88'0 Paulo Commisa'o Geog. e Geol.
6-8. Boletfm 22. 272 p.

Digitized by Google
 SELECTED REFERENCES 421
Weimer. R. J. and Land. C. B. Jr . 1972. Lyons Formation (Per- _ _ 1973. Ergs: Sed. Geology. v. 10. no. 2. p. 77 -106.
mian). Jefferson County. Colorado - A fluvial deposit. in Wilson. R. F . 1958. The stratigraphy and sedimentology of the
Environments of sandstone. carbonate. and evaporite Ourassic) Kayenta and (?Triassic) Moenave formations.
deposition: Mtn. Geologist. v. 9. no. 2-3. p. 289-297. Vermilion Cliffs region. Utah and Arizona: Stanford Univ.
Wellington. J. H.. 1955. Southern Africa - A geographical unpub. Ph. D. thesis. 167 p.
study: v.I. Physical geography: London. Cambridge Univ. Wopfner. H . and Twidale. C. R.. 1967. Geomorphological
Press. 528 p. history of the Lake Eyre basin. in J. N. Jennings and J. A.
Wentworth. C. K.. 1932. The mechanical composition of sedi- Mabbutt. eds . Landform studies from Australia and New
ments in graphic form: Iowa Univ. Studies in Natural Guinea: Cambridge [England] Univ. Press. p. 119-143.
History. v. 14. no. 3. 127 p. Yaalon. D. H . and Ganor. E . 1973. The influence of dust on
White. L. P . 1971. The ancient erg of Hausaland in south- soils during the Quaternary: Soil Sci . v. 116. p. 146-155.
western Niger: Geog. Jour . v. 137. pt. 1. p. 69-73. Yakubov. T. F . and Bespalova. R. Y. 1961. Soil-formation
Wilks. S. S. 1962. Mathematical statistics: New York. John processes during the invasion of sands by plants in the
Wiley 8t: Sons. 644 p. northern deserts of the Caspian region: Soviet Soil Sci . no.
Williams. G. E . 1968. Formation of large-scale trough cross- 6.p.651~.
stratification in a fluvial environment: Jour. Sed. Yu. Shou-chung. Li. Po. Tsai. Wei-ch'i and T'an. Chien-an.
Petrology. v. 38. p. 138-140. 1962. A survey of the Gobi of western Inner Mongolia and
Wills. L. J. 1948. The paleogeography of the Midlands: London. Pa Tan Chi Lin desert. in Academia Sinica Sand Control
Hodder and Stoughton. Ltd . 144 p. Group, eds., Research on sand control: U.S. Joint Publica-
_ _ 1956. Concealed coalfields - A paleogeographical tions Research Service JPRS 19993, Tech. Trans.
study of the stratigraphy and tectonics of mid-England in 63 -31183. p. 244 -276.
relation to coal reserves: London. Glasgow. Blackie and Zaychikov. V. T., [ed.], and others, 1965. The physical geogra-
Son. 208 p. phy of China: U.S. Joint Publications Research Service
Wilson. I. G . 1971a. Desert sandflow basins and a model for the JPRS 32,119. TT 65-32612, 634 p. [Technical translation
development of ergs: Geog. Jour . v. 137. pt. 2. p. 180-199. from Russian.]
_ _ 1971b. Journey across the Grand Erg Oriental: Geog. Zingg, A. W., 1953, Wind tunnel studies of the movement of
Mag . v. 43. p. 264-270. sedimentary material, in J. S. McNown and M. C. Boyer,
_ _ 1972. Aeolian bedforms - Their development and eds., Hydraulics Conf., 5th, Iowa [U.S.A.], 1952, Proc.:
origins: Sedimentology. v. 19. no. 3--4. p. 173-210. Iowa Univ .. p. 111-135.

Digitized by Google
Digitized by Google
 A STUDY OF GLOBAL SAND SEAS

INDEX
[Italic page numbers indicate major references)

A Page Page Page

Apocynum hendersonii .. .. . . .. .. .. .. . 389 Beni Abbes, Algeria, star dune in ... 142, 167
Ablekwasputs, Africa .. ".,', ... ,',. 355 Aprons.. .... ...... ... ........ ...... 104 Berg Winds ....................... 352
Acacia sp. . . . . . . . . . . . . . . . . . . . . . . . . 358 Arabian sand S88S 382 Bilma, Niger, linear dune in ...... 164
Ad Dahns'. . . . . . . . . . . . . . . . . . . . . . . .. 362, 366 barchan dune In .. . .. . .. . .. . .. .. 363 Blowout. See Dune, blowout.
Africa. See Abiekwasputs, Cave Sand- crescantic dune in .. .. .. . .. .. .. . 362 Bonito Canyon .................... 206
stone, Dwyka Group, Kal- dome dune in .. . .. . . .. . .. .. .. . .. 362 Bordj Omar Drlss, Algeria, slar dune
ahari Desert, Karroo Sy. hooked dune in .. .. . .. . .. . .. .. .. . 367 in ..................... 169
tem, Mariental, Nambi linear dune in .. .. .. .. . . .. .. . .. .. 362 Botucat~ Sandstone, Brazil ..... 190,197,233
Desert, Pelican Point, Up. slar dune in .. .. .. .. .. .. .. .. .. .. 362 stratification in ... ..... ...... 235
ill8ton. Arenites. feldspethic, of Zambia. . . . . . 190 Brady field ......................... 243
See also individual countries. Aristida sp. . .. . .. . .. .. .. .. .. . ... 336, 340. 386 Brazil, coastal dune in . . . . . . . . . .. . ... 6, 24, 30
AgriophyHum sp. . . . . . . . . . . . . . . . . . .. 366, 369 Arkose, in Casper Formation. . . . . . . . 221 interdune structure in . . . . . . . . . . 109
Akjoujt, Mauritania, linear dune in. . . . 261 in Lyons Sandstone. . . . . . . . . . . . . . 217 studies in . . . . . . . . . . . .. . . . . . . . . . . 89
Al HaMil al Asway, Libya. . . . . . . . . . . . 79 Arroyos............................ 196 See o/so Cabo Frio, Florlnopolis,
AI'Ur\lq al Mu'taridah, crescentic dune Artemisia sp. ....................... 389 Guairamar, 'ardlm Sio
in .................... 362 Atar, Mauritania, linear dune in . . . . 261 Pedro, ,08quina Beach,
AI'uwaynit, Libya. . . . . . . . . . . . . . . . . . 66 Atlanticum Period. . . . . . . . . . .. . . .. . . 377 'oaquina Hill, Lagoa dune
Ala Shan Desert, China. . . . . . . . . . . . 381 Augite ........................ " 65, 69, 79 field, Parana Besin, PSrto
barchanoid dune in .... , ..... , .. 152, 155 Australia. See Giles. Great Sandy De. Novo, Praia do Laste,
crescentic dune in . . . . . . . . . . .. . 267 ert, Great Victoria Des- Sambaiba Sandstone,
star dune in ......... ,......... 275 ert, Oodnadatta, Simp. Santa Caterina.
Algeria. See Beni Abbes: Bordj Omar son Desert. Break-aperts ....................... 112
Driss; Erg Irarrarene: Australian dunes, clay coat11J8s on sand Breccias ........................ 112, 121
Grand Erg Occidental: of ..................... 72 Buckhorn Wash, Utah............ .. 212
Grand Erg Oriental: Avalanches, wet.................... 133 Bushveld Sandstone, Transvaal. See
Hassi Messaoud: Sahara, Avalanching .............. 103, 122, 133, 192. Cave Sandstone.
northern. 200, 218, 232, 258
Algodones Dunes, California .. . . . . . . 267 Azlac Sandstone, Nevada and Califor- C
Alluvial fans .............. '" 27,122,221,386 nia .................... 190,209
sand ....... ,................. 27 Cabo Frio, Brazil, parabolic dune in.. . 95
sediments, unconsolidated....... 29 B Caiu; Sandstone .... ........ ....... . 234
Alluvium ........................... 376 Calcite ........................ 191, 211
modern ........... , ... ....... 72 Bahama Islands. carbonate sands in .. 24 Caliche............................. 376
Alteration, mineral, as cause of red pig- Boirdeslherla borbosoi ............. 235 carbonate. . . . . . . . . . . . . . . . . . .. . . 67
ment ... ,., .... , .... 65 Bejirs .......................... 282 gypsum ... ........... ... ....... 67
postdepositional chemical. . . . . . 191 Barra do SuI, Brazil, dome dune in ... 99 zones .......................... 66
Aluminum silicates. . . . . . . . . . . . . . . . . . 69 Berun Goyot Formation, Mongolia ... 190,230 Colligonum sp. ...................... 340, 369
An Nafiid. Saudi Arabia ............ 363, 366 contorted beddill8 in ........... . 231 Candonopsis pyriformis candona . . . . . . 235
crescentic dune in . . . . . . . . . . . . . . 275 feldspar in .................... 231 Canyon De Chelly .. . .. .. . .. .. .. .. .. . 205
Analysis. discriminant............... 31,55 interdune deposit in ............ . 230,232 Carbonate ............... . . . .. . .. . . . 76
energy-dispersive X-ray . . . . . . . . . 66 quartz in ..................... .. 231 caliche ..... ............... .... 67
linear.......................... 57 stratification in ............... .. 231 clast........................... 67
linear regression.. .. . . .. .. .. .. .. 143 transverse ridge dune in ........ . 223 sands, in Behama Islands. . . . . . . 24
regression . .. . .. .. .. .. .. .. . .. .. . 262 Basin, interior desert .............. .. 215 Cargona microphyllo ................ 369
of ripple marks, in Coconino Sand- slnki118 ........................ 191.196 Carmel Sea.... ...... ... ... ... ..... . 215
stone .................. 200 Beach sands ....................... . 55 Casper Formation. Wyomill8 ...... 190,195,221
scannill8 electron microscope . . . . 66 Beddill8, contorted ................. . 112, 189 contorted beddlll8 in . . . . . . . . . . . 223
whole phi ...................... 56 in Barun Goyot Formation ....... 231 feldspar in .. . . . . . . . . . . . . . . . . . . 221
X-ray diffraction. . . . . . . . . . . . . . . . 66 in Casper Formation ............ 223 festoon structures in . . . . . . . . . .. 221, 222
Anhydrite .......................... 196,225 in Coconino Sandstone .......... 201 interdune deposits in . . . . . . . . . . . 224
Antarctica. See Victoria Valley. in Navajo Sandstone ........... . 213 limestone in .................... 221
Aouk~r dunes. Sahara desert, tran&- penecontemporaneous contorted. orthoquartzite in . . . . . . . . . . . . . . 221
A versedunein........... 271 in Navajo Sandstone ... 216 quartz In . . . . . . . . . . . . . . . . . . . . . . . 221
Aoukar sand seas, crescentic dune in Bedforms, categorized .............. . 173 ripple marks in . . . . . . . . . . . . . . . 223
................... 310,313. 317 Bedrock. in Libya .................. .. 65. 72 stratification in .. .. .. .. . .. . .. . . . 222
423

Digitized by Google
 Page
on. Wyoming-Con CrosHtrata-Co
Is In ............. . trough .... 192, 221 ntinued
e. Africa ......... . tubular-plan 97,99 Desert ......... .
barchan dune in................ 237 Cyperaceae ........................ . 122 In Sahara of Mali .. . . . .. . . .. 269
feldspar In . . . . . . . . . . .. . . . . . . . . . 237 in Vallee de L' Azaouk . . . . . . . 317
parabolic dune in . . . . . . . . . . . . . . . 237 o barchanoid . . .. .. . . .. . . .. . .. .. . . 151
quartz in . .. . . . . ... . .. . . ... .. . .. 237 in Ala Shen Desert. . . . . . . . .. 152, 155
ripple marks In . . . .. .. . . . .. .. . . . 238 Data. quarter phi grain-size ......... . 24,55 in Cherchen Desert . . . . . . . . . 154
stratification in . . . . . .. . .. . .. . . . . 237 surface wind .................. . 141 in Namib Desert ........ '" . 153
tracks of vertebrate animals in . . . 237 whole phi ..................... . 56 in Navajo Sandstone........ 215
dune in ........ .. De CheUy Sandst 190,204 Makan Desert.. 14
feldspar in 205
interdune de 197,206 Desert .......... ..
orthoquartzi 205 ba Sands........ ..
atlon. postdepositio ripple mark 207 ironments of .... .
ert. China, barcha stratificatio 206 Dune. transverse.
une in ........... . trails of inve 207 ridge. In Killpeck
crescentic dunes in. . .. . . ... .. . .. 381 transverse dune in ............. . 197 dune field.. .. .. .. .. . .. . 91
Chert .............................. 191 Defiance uplift ..................... . 208 In Mao-Mu (Ting-hein) . . . . . . . 155
Cheshire Basin, England............. 197 Deposits, accretion ................. . 103 in White Sands National Mon-
Chevron. . . . . . . . . . . . . . . . . . . . . . . . . . . . 24, 381 aUuvial ....................... . 226 ument ................ 90,97, 152
China. See Ala Shen Desert, Cherchen fan ....................... . 196 blowout .................. 90,94,118,192
Desert, Mao-Mu. Pa-taD- avalanche ..................... . 112 defined .................... 11
chi-lin Sha-mo. Takla carbonate ..................... . 66 in LagOB dune field.. .. .. .. 114
akan Desert. See also
dune ............ . flood plain. 196 dune field ........
te .............. .. fluvial .... 23
I ............... .. fluviatile. . . 27, 191, 196 . .. .... 24, 30, 36, 7
glacial .... 236
........... 23,67 glaciofluviat 27 Sendstone ...... .
authlgemc ..................... . impermeabl .... .. .. . . . . . .. .. . . 23 coor ......................... ..
coatings, on sand. . . . . . . .. .. .. . . . 69, 78 inland dune. . .. . . .. . . . . .. .. .. . . . 55 as result of eirborne dust ... . 63
of Australian dunes. . . . . . . . . 72 interdune ................. 23,87,107.191 as result of climete ......... . 64,67
Climate, of Arabian Sand Seas........ 371 in Barun Goyot Formation. .. 230, 232 as result of grein size ....... . 80
effect on dune color.. . . . . . . . . .. . 64, 67 in Casper Formation. . . . . . . . 224 as result of mineral alteration 65
effect on dune formation .. 78.106,122.169 in De CheUy Sandstone ...... 197,206 as result of moisture ....... . 65, 79
of Kalahari Desert. . . . . . . . . . . . . . 380 In Fezzan .. . . .. . .. . . .. .. . . . 55. 109 explanation of ............. . 63
Desert ........... . in Lyon 216
Sahara ........ .. in Nava 09,212,214
in Rolli 225 Desert ......... .
Sahara ......... . in Whit on- Makan Desert ... .
akan Desert ..... . um 109,206
sert ............ .. See also
Sahera ......... . lacustrine. 191, 196 .. ......... 142,1
Coconino County, Arizona. herchan lag ......... ,. 191,194,218,221,225,389 in Al Uruq al Mu'taridah . . . . . 362
dune in................ 90 marine. , ........... , ....... 191, 196, 209 in Ala Shan Desert. . . . . . . . . . 267
Coconino Sandstone, Arizone ....... 3, 190, 197 permeable .............. , . . . . . . . 23 in An Nafud, Saudi Arabia. . . 275
contorted badding in . . . . . . . . . . . . 201 river channel.. .. . . .. . . . . .. . . . . . 196 InAouk;rsandseas ..... 310,313,317
rain pits in . . .. . . .. . . . . . .. . . . . . . 200 sedimentary .. . . .. .. . . .. . .. .. .. . 219 In Arabian sea sands. . . . . . . . 362
ripple marks in ................. 199 serir .. ... . . .. . . .. .. . . .. . .. . . .. . 24, 218 in Cherchen Desert, China. . . 361
slump marks in . . .. .. . . . . . . . .. .. 200 stream . . . . . . . . . . . . . . . . . . . . . . . .. 209, 219 defined .................... 256. 309
on in ............ . subaqueous .3,233,243 Desierto, Mexico ..
vertebrate animals Detritus. . . . . . . . 09. 193, 196 Sand Dunes Nation
rial ............ .. arkosic .... 219 ument ......... ..
Dinosaur remains 214,230.232 ska Sand Hills ... .
au .............. . Dolomite ..... . 225 Desert ........ 26
delta. fluvial sand . Dome-shaped du e. rn Sahera ...... .
assemblage, of Sa Doney Crater, A 201 -chi-lin She-mo... .
facies, Botuca~ ....... . 235 Drift potential .......... 146, 150,316,325.345, in Persia Gulf COBSt ........ 269
Corispermum sp.................... . 386 353,362,374,376,391 in Pew Karakumy ......... . 267
Comulaca sp....................... . 340 associated with barchanoid dunes 152 in Rub'al Kahali .......... .. 362
Couloirs ........................... . 282 associated with linear dunes. . . . . 155 in Sahara Desert. .......... . 310
Cow Springs Sandstone. . . . . . . . . . . . . . 190 associated with star dunes. ... . . . 165 in southern Sahara. . .. . .... 317
Cretaceous .................. 65, 233, 338, 389 Dune ................................. 4, 8, 69 In Takla Makan Desert ...... 367.369
Early.......................... 225 barchan ..... 9,24,28,88.89,141,151,258 in Thar Desart. ............. 271, 375
InA~b ~3 de LAzaouk .... .
ry rocks ....... .. in Cave 237 rn Sahara ...... ..
in Coco ne . 90 Dune, star.
............. 5,87, in Libya 269 of ............. ..
212,218 inNam 153 , in LagOB dune fiel
in Rub'a 269 ... 24. 26, 66, 98. 1

:"VU I.J] _ "-'

 INDEX 425
Page Page E Page
Dune-Continued Dune-Continued
produne-Contlnued dome-Continued Earth Resources Technology Satellite.
in Arabian sea sands ....... . 362 pyramidal. See Dune. star. See Images. ERTS-1; Landsat-!.
in Barra do Sui ........... .. 99 raversing ........ 29.68.103.114.123.195 Egypt. See Great Western Desert.
in Frodsham Member ...... . 197.226 deFined .. .................. 12 England. See Cheshire Basin. Frodsham
in Grand Erg Occidental .... . 280 in Great Sand Dunes National Member of so-called Keu-
in Killpecker dune field ..... . 98.280 Monument ............. 104.195 per Sandstone Formation.
il' Nafiid as Sirr .......... .. 362 in Kelso dune Field. Mojave Keuper Sandstone. Lower
In Nafu~ ath Thuwayriit .... . 362 Desert................. 105 Mottled Sandstone.
in Nafiid Shuqayyiqah ...... . 362 in Lagoa dune field. . . . . . . . .. 105. 123 Yellow Sands of Durham.
in Pararui ................. . 99 in Namib Desert. . . . . . . . . .. 346. 349 Entrada Sandstone.. .. .. . .. . . .. .. . .. 190
in Praia do Leste ........... . 99 in Takla Makan Desert. . . . . . 389 Environments. depositional. . . . . . . . . . . 27
In Santa Caterina. . .. .. .. .. . 99 in Victoria Velley . . . . . . . . . . . 104 Eolianites. . . . . . . . . . . . . . . . . . . . . . . . . . . 24
in Takla Makan Desert .. 280.381.391 in White Sands National Mon- Ephedra sp. .. . .. .. . .. .. .. . .. . . .. .. . . 389
in White Sands National Mon- ument . ... ... ... ... .... 106 Epidote ...................... 65.69.231.357
ument .............. 98. 152. 279 self ............................ 24 Erogrostls sp. .... . .. .. .. . . .. . ... .. .. 357
double berchan . . . . . . . . . . . . . . . . . 5 in Sabha .. .. .. .. .. .. .. .. .. . 103 Erg Bilma Region. Niger. linear dune
forms.......................... 280 See also Dune. linear; Ridge. in .... .... ............. 164
hanging ........................ 24.127 linear. 'Erg Chech. Algeria ............... 326. 329. 335
in Lagoa dune Field. . . . . . . . . . 127 simple. deFined ................. 260 Erg er Raoul.. .. .. .. . . .. . .. . .. .. .. .. 326. 335
hooked. in Arabian sea sands. . . . 367 linear. in Great Sandy Erg Hammami Region. Mauritania. lin-
immobile. in western Sahara..... 312 Desert ............. 141. 157. 163 ear dune in .. . .. . .. . .. . 164
inland... ... .... . .. ... .... .... . . 27. 44 linear. in Great Victoria Erg-Iguidi .. .. .. .. . .. .. .. .. .. .. .. .... 326. 335
linear .............. 102.107.109.141.155. Desert ............. 141. 157. 163 Erg Irarrarene. Algeria. ster dune in . . 169
193.280.261.313.317 linear. in Kalahari Erg Makteir. Mauritania. linear dune In 142
in Arabian sea sands. . . . . . . . 362 Desert ............. 141. 156. 163 Ergs................................ 310
in Akjoujt .. .. .. . .. .. .. .. .. . 261 star ...... 24.29.88.106. 141.165.261.275 ERTS-t. See Images. ERTS-1; Landsat-I.
inAtar..................... 261 in Ala Shan Desert.. . .. . .. .. 275 EslherieUo sp. . .. . .. .. .. . .. . . .. . .. .. . 235
inBUma.................... 164 in Arabian sea sands. . . . . . . . 362 Eljo Sandstone. South-West Africa.
defined .................... 11. 258 in Beni Abbes .. .. .. .. .. .... 142. 167 See Cave Sandstone.
in Erg Bilma Region......... 164 in 80rdj Omar Driss . . . . . . . . . 169 Eueslherio mendesi .. . .. . . . .. . .. .. . .. 235
in Erg Hammami Region..... 164 deFined .................... 9. 259 Evaporites.. ... .... ..... .... ... . . .. . 217
in Erg Makteir. . . . . . . . . . . . . . 142 in Erg Irarrarene........... 169 Experiments. wave-tank ............. 177
in Fort Gouraud. . . . . . . . . . . .. 156. 164 in Ghudiimis ................ 142. 165 Extradune. defined .. . . . .. . .. . .. .. . .. 243
in Giles. Australia. . . . . . . . . . 163 in Gran Desierto ............ 107. 275
interdune surfaces in . . . . . . . 109 in Grand Erg Occidental ..... 165. 326 F
in Jabel Tuwayq . . . . . . . . . . . . 368 in Grand Erg Oriental ..... " . 107. 142.
junctures in ................ 156 165.275.326 Fans. alluvial ................ 27.122.221.386
in Kalahari Desert. . . . . . . . . . 360 in Hassi Messaoud. . . . . . . . . . 167 Faulls .............................. 112. 121
in Mariental.. . .. . . .. .. .. . .. 163 in high-energy wind regime. . 142 Fauna. See Vegetation.
in Namib Desert. . . . . . . . . . .. 346. 349 interdune deposits. . . . . . . . . . 111 Feidjs ............................. . 282
in Navajo Indian Reservation 261 in Irq as Subey . .. .. . . . .. . .. 277 Feldspar .......................... . 124. 191
in Peski Karakumy ......... 141. 161 in low-energy wind ragime . . . 142 authigenic ..................... . 228
in Qu'amiyat .. ...... .... .. 367 in Namib Desert. . . . . . .. 275. 346. 349 in Barun Goyot Formation ....... . 231
in Rotliengendes . . . . . . . . . . . . 226 in northern Sahara ...... 326.338.341 in Casper Formation ............ . 221
in Rub'al Khali .............. 260. 362 in Ramlet es Sahama . . . . . . . . 363 in Cave Sandstone ............. . 237
in Sabha . .. .. .. .. . . .. .. .... 156. 161 inRubaIKhali~ ........ 107.275.362 in De Chelly Sandstone ......... . 205
in Sahara Desert ............ 261. 310 in Sierra del Rosario. . . . . . . . 275 in Frodsham Member ........... . 228
inSimpsonDesert ........... 141.157 in Takla Makan Desert. . . . . . 389 in Kalahari Desert ............. . 356
in southern Sahara......... 317 in Zalim. Saudi Arabia...... 106 in Lower Mottled Sandstone ..... . 227
in Takle Makan Desert...... 381 stringer. in Kalahari Desert. . . . . . 359 in Lyons Sandstone. . . . .. . . . . . . . . 218
in Upington . . . . . . . . . . . . . . . . . 156 transverse 24.28.88.93.141.151.191.258 in Navajo Sandstone............ 211
in western Sahara .......... 310. 313 in Aoukir dunes. .. . . .. . . .. . 271 in Thar Desert.. .. . . .. .. .. . . .. . . 376
longitudinal. in Namib Desert. . . . 349 in Cave Sandstone. . . . . . . . . . 237 Feldspathic sand.. . . .. . .. . . .. .. .. . . . 24
See also Dune. linear: Ridge. in De Chelly Sandstone...... 197 Killpecker dune Field. . . . . . . . . . . . 29
linear. deFined ..................... 9 Lyons Sandstone. .. .. . . . . .. . . .. . 220
massifs. See Dune. star. in Frodsham Member. . . . . . . 229 Ferric hydrate ...................... 63.68.78
migration .............. 108. 227. 233. 367 in Killpecker dune Field. . . . . . 94 See also Limonite.
oghurd. See Dune. star. in Namib Desert. .. . .. . . .. . . 349 Ferromagnesian silicate minerals. . . . . 65
parabolic .............. 24. 28. 88. 94. 114. in Navajo Sandstone.. .... . . 215 alteration of .. . .. . .. . .. .. . . .. . .. 63
122.193.220.261.277 in Rotliegendes. .. .. . . .. . .. . 226 See also Augite. Epidote. Horn-
in Cabo Frio.. . . .. . .. .. . . .. 95 in Takla Makan Desert. . . . . . 392 blende.
in Cave Sandstone. . . . . . . . . . 237 in White Sands National Festoon cros&-Iamination ............ . 221
defined .................... 11. 259 Monument ............ 93. 97. 193 Festoon structures. Casper Formation. 221.222
in Indiana dunes .. .. .. .. .. .. 95 transverse ridge. in Barun Goyot Fezzan. Libya. inland dune deposit in .. 55
in Lagoa dune field . . . . . .. 96. 114. 122 Formation ............. 233 interdune deposit in ............ . 55. 109
in Lyons Sandstone . . . . . . . . . 220 in Namib Desert.. . .. . . .. . . . 153 Floods. flash. errect on dunes ........ . 124. 130
in Thar Desert ........... 95. 277. 375 varieties ....................... 260 Florinopolis. Brazil ................. . 196
in While Sands Nationel Mon- Dust. airborne...................... 73 Fluvial sand ....................... . 30
ument .............. 95.152.277 as cause of dune color. . . .. . .. . .. 63 in Colorado River delta ......... . 30
produne... ..................... 24 Dwyka Group. Africa.. . .. .. .. . . .. . .. 236 Fluviatile deposits .................. . 27.228

Digitized by Google
426 INDEX
Paga Page Page
Folding. penecontemporaneous. . . . . . . 5 Gualramar. Parana. Brazil. retention Kaolin......... .... ......... ........ 205
Folds .................. 112.121.213.223.230 ridge In................ 97 Kaolinite ........................... 228
gentle.......................... 112 Gujarit Plain. India . . . . . . . . . . . . . . . . . 375 in Libya. . . . . . . . . . . . . . . . . . . .. .. . 69. 73
stairstep ....................... 105 Gulf of Sidra.. .. .. .. .. .. . .. .. .. . .. .. 67 Karroo System. Africa. . . .. . . . . . . . .. . 236
Footprints. See Tracks. ofvertebrate Gypsum ......................... 78. 191. 196 Kayenta Formation.. .. .. .. .. .. .. .. .. 211
animals. caliche... ....... ... .... ........ 67 Kelso dune field. Molave Desert. Cali-
Foraaets ............ 89.105.121.192.216.218 sand. in Wbite Sands National fornia. elongate ridge in. 105
Forest Sandstone. Rhodesia. See Cave Monument ............. 24. 28 raverslng dune in . .. . . . . . .. . . . . . 105
Sandstone. Keuper Sandstone. England. . . . . .. 190. 197. 228
Fort Flatters. Algeria. See Dordl Omar H Killpecker dune field. Wyoming. . . . . . . 6. 89
Orin. Algeria. barchanoid ridge dune in . . . . . . . . 91
Fort-Gouraud. Mauritania. linear dune Halite.............................. 226 dome dune in .. .. .. .. .. . .. .. .. .. 96. 280
in ..................... 156.164 Ha/oxylon sp. ....................... 386. 389 feldspe!hic sand . . . . . . . . . . . . . . . . 29
F088ils. marine. . . . . . . . . . . . . . . . . . . . . . 224 Hassi MB8BBoud. Algeria. star dune In. 167 interdune structure in . . . . . . . . . . . 109
F088ils. See Dinosaur ramains. Trace Hedysorum sp. . . . . . . . . . . . . . . . . . . . . . . 389 quartzosa sand ................ . 29
fossils. Hematite. in Libya ................ " 63. 69. 78 transverse dune in . . . . . . .. . .. . . . 94
Fountain Formation ................ . 221 in RotliegendB8 .. .. . . .. .. . .. .. .. 225 Kinlichee .. .. .. .. .. .. .. . .. .. .. . .. .. . 207
Four Corners araa .................. . 205 Hermit Shale....................... 197 Kyanite. detrital.................... 357
Frodsham Member of ~lled Keuper Holocene ........................... 67
Sandstone Formation. Hornblende .................... 65. 69. 79. 376 L
England .............. . 228 Humic compounds ................... 122.127
dome dune In .................. . 197. 228 Lacustrine deposits. . . . . . . . . . . . . . .. 27. 29. 228
feldspar in ................... .. 228 Lagoa dune field. Brazil.... .......... 114
illite in ........................ . 228 blowout dune in.... ............. 114
quartz in ...................... . 228 illite. in Frodsham Member. . . . . . . . . . . 228 captation dune in .. . .. . .. . .. .. .. 127
transverse dune in ............. . 229 in Libya. . . . . . . . . . . .. . . . .. . . . . . . 69. 73 dlssipetion of dune in .... , . . . . . . . 124
Front Range highlands. of Lyons Sand- 1llite-montmorillonite. in Libya. . . . . . . . 69. 73 granite in ...................... 114.124
stone.................. 218 Ilmenite ............................ 69. 356 hanging dunes in.. ........ ...... 127
Frosting ........................ 189. 191. 211 Images. ERTS- 1 ........... 6, 95. 102. 141. 150. interdune araas in .............. 122. 130
process ........................ 24 169. 173. 257. 309 parabolic dune in ............. 96. 114. 122
FuljB8 .............................. 282 India. See Gulariit Plain. Thar Desert. ponds in . . . . . . . . . . . .. . . . . . . . . . . . 130
Indiana dunes. Lalce Michigan shora. raversing dune in ............... 105.123
G perabolic dune in . . . . . . . 95 soil horizons in .. .. .. .. .. .. . .. .. . 130
Indurated roclcs . . . . . . . . . . . . . . . . . . . . . 24 stringers in .................... . 127
Ganado.... ........... ....... ....... 206 Interdune araa. 107. 122. 189.220.282.336.389 Lake. duert ....................... . 225
Garnet ............................. 356 in Lagoa dune field ............. 122.130 Lamb Point Tongue ................. . 211.214
Ghudamis. Libya. sand rose for. . . . . . . 165 in Rub 'al Khali .. .. . .. . .. . .. .. .. . 282 Laminae. fadaout. .................. . 112
star dune in . . . . . . . . . . . . . . . . . . .. 142. 165 in Sahara DBBert . . . . . . . . .. . . . . . . 282 Landsat-l ......................... . 6
GilB8. Australia. lineer dune in. . . . . . . 163 in White Sands National Monu- See also lmagu. ERTS-l.
GlacioDuviatile deposits. . . . . . . . . . . . . 27 ment .................. 282 Landsat ilnagery. See Images.
Gobi DeBBrt. Mongolia. . . . . . . . . . . . . . . 230 Interdune deposits. See Deposits. intel'- ERTS-1.
Goethite. . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 dune. Laramie Besin. Wyoming ........... . 221
Gondwana beds. Africa. . . . . . . . . . . . . . 236 Interdune sabkhas.................. 193 Lateritic soils ...................... . 63
Grain. frosted. . . . . . . . . . . . . . . . . . . . . . . 31 Iron. source of . .. . .. .. . .. .. .. . . .. .. . 69 Lava .............................. . 236
size. effect on pigment. . . . . . . . . . . 60 Iron-bearing minerals ............... 63. 69. 79 Lettau equation for sand drift ........ 141.145
GrominBOe ......................... 122 Iron oxide ...... 7.63.89.78.191.211.227.338 Leupp. Arizona ..................... 5
Gran Desierto. Mexico. crB8centic dune Irq as Subay. Saudi Arabia. star dune Lias. See Jurassic. Lower.
In ..................... 269 in ..................... 277 Libya. bedrock in.. .. . .. .. .. .. .. .. .. . 216
star dune in .................... 107.275 hematite in ..................... 63.69. 78
Grand Canyon.. .. .. .. .. .. .. .. .. .. .. 3. 197 illite in . . . . . . . . . . . . . . . . . . . . . . . . . 69. 73
Grand Erg Occidental. Algeria. . . . . . . . 329 illite-montmorillonite in . . . . . . . . . . 69. 73
dome dune in .. .. .. . .. .. . .. .. .. . 280 Jabal as Sawdii: Libya. . . . . . . . . . . . . . . 79 lceollnite in .. . .. .. .. .. .. .. . .. .. . 69. 73
quartz in . . . . . . . . . . . . . . . . . . . . . . . 334 Jabal Tuwayq. linear dunB8 in . . . . . . . . 368 studies in . . . . . . . . . . . . . . . . . . . . . . . 89
star dune in .................... 165.326 Jardim sao Pedro. Para~. Brazil. re- wutern. aolian deposits. . . . . . . . . 63
Grand Erg Oriental. Algeria and Tun- tention ridge in . .. .. . . .. 97 See also AI Harul al Asway; FIIZZIIJI;
isia.................... 335 Joaquina Beach. Brazil. . . . . . . . . . . . . . . 114 Ghudiimis; Jabal as Saw-
star dune in ........ 107.142.165.275.328 Joaquina Hill. Brazil. . . . . . . . . . . . . . . . . 127 dii; Ramiat Zallaf; Ramil
See also GbudSmis. Libya. JuncturB8. of linear dunB8 . . . . . . . . . . . . 156 al Kabi'r: Sabha; Sahara.
Granite. at Lagoa dune field. . . . . . . . .. 114. 124 Jurassic. . . . . . . . . . . . . .. 189. 190. 209. 233. 376 northern; 1a~i' Awb8rl;
Graat Dune at Kerzaz. ... . .. ...... ... 335 JuraSSiC. Late. . . . . . . . . . . . . . . . . . . . . . . 235 :lall.rii' Marzuq; 1aiiii'
Graat Escarpment ................... 346.349 JuraSSiC. Lower . . . . . . . . . . . . . . . . . . . . . 209 labyanah.
Great Sand Dunes National Monument. Libyan DB8ert. barchan dune in ...... . 269
Colorado .............. 29. 89 K LlmB8tone ......................... . 376
crascentic dune in .. . .. . .. . .. .. . 269 in Casper Formation ............ . 221
raversing dune in ............... 104. 195 Kalahari DBBert. Africa. . . . . . . . . . . . . . 355 lenticular ..................... . 225
Great Sandy Desert. Australia. simple feldspar in .. .. . .. . .. .. . .. .. .. .. 356 marine ........................ . 65
linear dune In ...... 141.157. 163 linear dune in .. .. .. . .. .. .. . .. .. . 380 aphanitic ................. . 221
Greet Victoria Desert. Australia. sim- quartz in . . . . . . . . . . . . . . . . . . . . . . . 356 In Navajo Sandstone ........... . 211
ple linear dune in ... 141.157.163 simple linear dune in ........ 141.156.163 Limonite ........................... . 63.68
Great Western DBBert. Egypt. . . . . . . . . 329 striliger dune in .. . .. . . .. .. .. . .. . 359 Linear discriminant anelysis ........ . 57
Green River Desert.................. 212 Kanab area. . . . . . . . . . . . . . . . . . . . . . . . . 212 Linear ragression analysis .......... . 143

Digitized by Google
 INDEX 427
Page Page Page
Littoral sand ....................... . 27.30 MosaiCS. ERTS ..................... . 6 Paleozoic. Upper............ .. .. .. .. 243
Loet Soldier Field. Wyoming ....... 243 Mudstone ......................... . 229.237 Palygorskite . , ..................... 67,73.78
Lower Bunter Sandstone. See Lower Para~y. See Mislones Formation.
Mottled Sandstone. N Para~ BaBin ................ , . .. . . 233
Lower Mottled Sandstone. England 190.197.227 ParanS. Brazil. dome dune In . , . . 99
feldspar In . . . . . . . .. .. . . . . . 227 N summaries.. . . . . .. . .. . . . . . . . . .. . . 143 Parnaiba BaBin ............. . . .. . . .. 233
quartz In .. .. .. .. .. .. .. .. .. .. .. 227 Nafiid as Sirr .. .. .. . .. .. .. .. .. .. .. .. 369 Pa-tan-chi-lin Sha-mo. China. crescantic
stratification In . . . . . . . . . . . . . . 227 doma dunes in . .. . . . . . . . . .. . 362 dune In ................ 381
Lyons Sandstone. Colorado .... 190.197.217 Nafiid ath Thuwayrat . . . . . . . . 366 Peel. colloidal. .. . .. .. .. .. .. .. .. .. .. 127
arkosa In .. .. .. .. . .. .. .. .. . .. .. . 217 dome dune In . . . . . . . .. . . . . . . . 362 Pelican Point, South-West Africa..... 153
coastal dunes In . . . . . . . . . . . . . . .. . 220 Nariid Shuqayyiqah. dome dune in . . 362 Pennsylvanian sandstone ..... 35.190.221
feldspar In .. .. .. .. .. .. .. .. .. .. . 218 Namib Desart. South-West Africa. . . 346 Permeabllity . . . . . . . . . . .. . . . . . .. . . 33
Front Range highlands. .. . . . . 218 barchan dune in .. .. .... . .. .. .. . 153 Permian Rotliegendes. See Rotlie-
Interdune deposits In . . . . . . . . . 218 barchanoid dune in....... ....... 153 gendes.
parabolic dunes In .... .. .. . .. .. . 220 complex dune In .. .. .. .. . .. .. ... 346. 349 Permian sandstone ......... 35.190.197.204.
quartz In . . . .. . . . . . . . . . . . .. . . . 218 crascentic dune In .......... 269. 346. 349 217.221.225.227
quartzosa sand in .. .. .. .. .. .. .. . 220 linear dune in .. .. .. .. .. .. .. .. ... 346. 349 Permian Satanka Formation. See Satan-
rain pits In . . . .. . . . . . . . ... . . . 218 longitudinal dune In .. .. .. .. .. .. 349 ka Formation.
ripple marks In .. .. .. .. .. .. .. .. . 218 quartz in .. .. .. .. . .. .. .. .. .. .. .. 349 Persian Gulf Coast. crescentic dune In . 289
slump marks In . . . . . . . . .. . . . . 218 reversing dune in .. .. .. .. .. .. ... 346. 349 Peru. See Pur-pur dune.
stratification In . .. .. .. .. .. .. . .. 218 star dune In ................. 275. 346. 349 Peskl Karakumy. U.S.S.R.. crescantic
trails of invertebrate animals in.. 219 transversa ridge dune In ........ 153, 349 dune In ................ 267
Navajo Indian Reservation. Arizona. simple linear dune In .... , .... 141. 181
M linear dune In . . . . . . 261 . Petroleum .......................... 190.243
sand streaks in ................. 281 Phi grade scale. . . . . . . . . .. . . . . .. . . . 23
Magnetite .......................... 89 Navajo Sandstone ............ 189. 190.209 Photography, aerial ............ 173,257. 309
Makgabeng Formation. Transvaal.. . . 191 barchanoid dune in...... ........ 215 Skylab ......................... 257. 309
Mali. See Sahara. southarn; Sahara. contorted bedding In ............ 213 still. in laboratory. . . . . . . . .. . . . . 173
western. faldspar in ..................... 211 Phrogmites communls . . . . . . . . . . . . . . 389
Mantle. granitic. . . . . . . . . . . . .. . . . . . . 127 interdune deposit in ....... 209.212.214 Physics. sand .. .. .. .. .. .. .. .. .. .. .. 4
Ma~Mu (Tlng-hsin). China. barchanoid limestone In .................... 211 Piramboia Facies. . . . . . . . . . . . . . . . . . . 235
ridge dune in . . . . . . . . . . . 155 orthoquartzite in . . . . . . . . . . . . . . . 211 Pitting. . . . . .. . . . . . .. . . . . . . . . . . . .. . . 189
Maps. thematic ....... . . . . . . . . . . . . . 6 penecontemporaneous contorted Plateau du Mangueni . . . . . . . . . . . . . . . . 340
Marble Canyon. . . . . . . . . . . . . . . . . . . . 197 bedding in .. .......... . 216 Playa deposits .................. 27. 29. 232
Mariental. South-West Africa. linear quartz in .. ..... ...... ... .... 211 Pleistocene .................. 357.383.366
dunes in ............. 163 ripple marks in .. .. . .. .. .. .. .. . . 212 early........................... 87
sand rose for. . . . . . . . . . . . . . . . . . 163 stratification In .. .. . . .. .. .. .. . .. 212 late............................ 349
Marine aphanitic limestone. . . . . . . . . . 221 transverse dune in.............. 215 sandstones ..................... 334
Marl. ... ... ... ....... .......... .... 109 Nazlini Canyon. . . . . . . . . . . . . . . . . . . . 207 Pond. desert.. .. .. .. .. .. .. .. .. .. .. .. 196
Material. detrital. . . . . . . . . . . . . . . 127 Nebraska Sand Hills. crescentic dune ephemeral...................... 214
Mauritania. See .\kjoujt; Atar; Erg In ........... , ......... 267 interdunel . . . . . . . . . . . .. . . . . . . 109
Hammami Region; Erg Negegt Basin. Mongolia. . . . . . . . . . . . . 230 In Lagoa dune field. Brazil . . . . . . . 130
Makteir; Fort-Gourard; Niger. See Bilma; Erg Bilma Region; Porali Plain. Pakistan. . . . . . . . . . . . . . . 376
Sahara. southern; Saha- Sahara, southern. Porto Novo. BrazU. retention ridge in . . 97
ra. western. Nitraria schoberi . .. . .. . .. .. . .. .. .. .. 369 Praia do Laste. Brazil. dome dune in . . 99
Measurement. systems of ............ 17 Northern Sahara. crescentic dune In . 349 Precambrian. . . . . . . . . . . . . . . . . . . . . . . 190
Mesozoic ..... 15.65.81.190.197.243.383.385 star dune in ................. 326.338,341 PrB88rvation. of markings.. . .. ....... 5
Mexico. See Gran Desierto. Nortbern Sand Seas.. .. . .. . . .. .. .. .. 368 PsonunochJeo villoso . . . . . . . . . . . . . . . . 389
Mica ........................... 205.228.231 Nugget Sandstone .................. 190,209 Pugionium comutum . . . . . .. . . . . . . . . 389
Minerals. ferromagneslan silicate. altar- Nyamandlovu Sandstone, Rbodesla. Pur-pur dune. Peru. . .. . .. . . .. .. .. .. . 267
ation of .............. 63 See Cave Sandstone.
Miocene ........................... . 347 Q
sandstones .................... . 334 o
Misiones Formation. Paraguay... . . . . 235 Qa'amiyat. compound linear dune In.. 387
Model. dune-extradune . . . . . . . . . . . . . . 243 Oliveni. Hawaii ...... .. .. .. .. .. .. .. . 24 Quartz ....................... 24.85,124.191
Moenkopi Formation. Arizona. . . . . . . 190 Oodnadatta. Australia. wind regine at. 157 in Barun Goyot Formation. . . . . . . . 231
Mogollon Rim. Grand Canyon......... 197 Orthoquartzite ................... 65,191.205 in Casper Formation. .. .. .. . .. .. . 221
Moisture. effect on dune color. . . . . . . . 65. 79 in Casper Formation. . . . . . . . . . . 221 in Cave Sandstone.. .. . .. .. .. . .. 237
effect on dune formation ........ 78. 94. 97. in De Chelly Sandstone. . . . . . . . . 205 In Frodsham Member. . . . . . . . . . 228
113.194 in Navajo Sandstone. . . . . . . . . . . . 211 in Grand Erg Occidental. . . . . . . . . 334
role in structure type ........... . 112 Oved Saoura . . . . . . . . . . . . . . . . . . . . . . . . 334 in Kalahari Desert. . . . . . . . . . . . . . 356
Mongolia. See Gobi Desert. Overthrusts ........................ 112 in Lower Mottled Sandstone. . . . . . 227
Monument Valley .................. . 206 in Lyons Sandstone.............. 218
"Morphology. Provenance. and Move- P In Namib Desart .. .. .. .. .. .. .. .. 349
ment of Desart Sea Sands. In Navajo Sandstone.. .. .. .. . .. . 211
The" .................. 6 Puchecoio ocuminoto . .. . .. . .. .. .. . .. 235 In Rotliegendes . .. . .. .. .. .. .. .. . 225
Morphometry. of crescentic dunes 265. 269. 297 Puchecoio rodriguesi ... . . . . . . . . . . . . . 235 In Sabri'Marziiq . . . . . . . . . . . . . . . 339
of dome dunes .................. 281. 302 Pakistan. See Poriili Plain. Thar Desart. san'3sfone ...................... 197.205
of linear dunes .................. 261.296 Po1oeolimnodio petrii .. .. .. .. .. .. .. .. 235 in Thar Desert. . . . . . . . . . . . . . . . . 376
of parabolic dunes.. . .. . .. . .. . .. 277 Paleosol II. horizon) . . . . . . . . . . . . . . . . . 124 See oJso Cbevron.
of star dunes ................... 275. 299 Paleozoic ... 15,65,190.197.217.339,363.385 Quartzite. feldspathic ............... 217

Digitized by Google
428 INDEX
Page Page Page
Quartzite-ConUnued Sahera Desert-Cotinued Shamal winds ....................... 371. 374
lenticular dolomitic............. 217 linear dune in .................. . 261.310 Shatt al larid .. .. .. .. .. .. .. .. .. .. .. . 336
Quartzoaa 88nd . . . . . . . . . . . . . . . . . . . . . 24 of Mali. berchan dune in ........ . 269 Shurtz Sandstone ................... 211
Killpecker dune field . . . . . . . . . . . . 29 of Mauritania. 88nd streaks in ... . 281 Sidra. Gulf of .. .. .. .. .. .. .. .. .. .. .. . 67
In Lyons Sandstone. . . . . . . . . . . . . . 220 northern. of Algeria ............ . 328 Sierre del R088rlo. star dune in . . . . . . . 275
Quaternary alluvium. . . . . . . . . . . .. 72. 334. 389 of Libya ................... . 328 SWca ........................ 69.73. 191. 211
Quaternary dune 88nd . . . . . . . . . .... . . 72 of Tunlsla ................. . 328 SWcate. aluminum .................. 69
southern. of Mall. ............. .. 317 ferromagneslan .. .. .. .. .. .. .. .. . 79
R of Mauritania ............ .. 317 Silt content. . . .. . . . . . . . . . . . . .. . . . . . . 23
of Niger ................... . 317 Siltstone ........................ 211.217.229
Rainplts ........................... 5.87.191 western. of Mali ............... . 310 calcareous..................... 214
In Coconino Sandstone. . . . . . . . . . . 200 of Mauritania ............ .. 310 micaceous ...................... 208.217
in Lyona Sandstone. . . . .. . . .. . . . . 218 of Sanegal. ................ . 310 Simpson Desert. Australia . . . . . . . . . . . 63. 78
Remlat as Sallama. star dune In . . . . . . . 383 Sahel...... ... ....... .............. 317 simple linear dune in ............ 141. 157
Ramiat ZelliT. Libya. . . . . . . . . . . .. . . .. 327. 338 jajrii'AwbBri. Libya ....... 83.66.79.326.338 Size. greln . . . .. . . . . . . . . . . . . . . .. . . . . . 23
Ramll al KabTr. Libya................ 329 ~irii'Marziiq. Libya ....... 63. 66. 79. 328. 339 Skeleton. vertebrete ................. 231. 232
"Red Desert SoUs" .................. 66 . quartz in . . . . . . . . .. . . . . . . .. . . . . . 339 Skylab4.. .......................... 6
Regression analysis .. .. .. .. .. .. .. .. .
Remote sensing. See Images. ERTS-l.
Reptile remains. In Cave Sandstone. . .
262

237
tr.~~~~~~: .~i~:.~ ::::::::::: ::: 79.~:
Sambelbe Sandstone. Brazil .......... 190.233
Slipfaces. study of.. . . .. . .. . .. . . .. . .. 87. 200
Slump marks ................... 5.87.189.191
In Coconino Sandstone. . . . . . . . . . 200
Rhodesia. See Cave Sandstone. Sampling ........................... 23 In Lyons Sandstone.. .. .. .. . .. .. . 218
Ridge .............................. 24 San Rafael Swell .. .. .. . .. .. .. .. .. .. . 211 Slumping ................... 104.113.121.231
beach dune . . . . .. . . . . . .. . .. . . .. . 97 Sand. alluvial. . . . . . .. . . . . . .. . . . . . . . . 23 Soil horizons . . . . . . . . . . . . . . .. . . . . .. . . 130
defined ........................ 11 drift. See Sand migration. In LagOB dune field..... ... ...... 130
elongate in Kelso dune field...... 105 feldspethlc. in Killpecker dune Sonoran Desert .......... .. .. .. .. .. . 30
linear .......................... 88 field ................... 29 Soesus Vlei .......... .. .. .. .. .. .. .. . 352
precipitation. United States west littorel ........ . . .. . . . . . . . . . . . . . 23 Southern Sahara. crescentic dunes In. 317
COBst .................. 96 migration ...................... 141 linaarduneln............... .... 317
retention . . . . . . .. . .. . . . . . . . . . . . . 96 quartzose. in Killpecker dune field. Spanish Sahara .. .. .. .. .. .. .. .. .. .. . 310
In Gualramar .. .. .. .. .. .. .. . 97 Wyoming .............. 29 Staurolite .......................... 357
in lardlm sao Pedro. Brazil. . 97 ridges. See Dune. linear. Stereonets . .. . . . . . . . . . . . . . . . . . . . . . . . 97. 194
in Porto Novo. Brezil . . . . . . . . 97 rose ............... 141. 147. 163. 165. 310 Stipogrostis sp.. .. . .. .. .. .. .. .. .. .... 349. 357
88nd. See Dune. linear. serir ........................... 23 Straate............................. 282
Ripple index (Rl) .... 174.192.200.224 sheets .. 11.87.156.167.281.281.313.375 Strata. brecciated................... 112
Ripple marks ................... 5.87.189.191 In Upington................. 156 marine.. ... ....... ... .......... 190
in Casper Formation. . . . . . . . . . . . . 223 InValleedeL' Azaouk....... 320 Stratification. of Barun Goyot For-
In Cave Sandstone . . . . . . . . . . . . . . 238 See olso Stringers. mation................. 231
in Coconino Sandstone. . . . . . . . . . . 199 shields .. . . . . . . . . . . . . . . . . . . . . . . . 377 of Butucatu Sandstone. . . . . . . . . . . 235
in De Chelly Sandstone. . . . . . . . . 207 streaks ..................... 281.281. 375 of Casper Formation. . . . . . . . . . .. . 222
In Lyons Sandstone.............. 218 in Navajo Indian Reservation 281 of Cave Sandstone.. .. .. .. .. .. .. 237
in Navajo Sandstone. . . . . . . . . . . . 212 In Sahara of Mauritania ..... 281 of Coconino Sandstone. . . . . . . . . . . 199
Ripple symmetry index (RSI).......... 174 See olso Stringers. of De Chelly Sandstone. . . . . . . . . . 206
Ripples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24. 173 Sandstones. ancient................. 15.189 of Frodsham Member. . . . . . . . . . . . 229
adhesion ................... 192.196.226 origin.......................... 189 of Lower Mottled Sandstone. . . . . 227
River sands. . . . . . . . . . . . . . . . . . . . . . . . . 55 Santa Caterina. Brezil. dome dune in . . 99 of Lyons Sandstone. . . . . . . . . . . . . . 218
Rocky Mountains. Colorado and Wyo- Santana facies. . . . . . . . . . . . . . . . . . . . . . 235 of Navajo Sandstone.. . ..... . ... 212
ming .................. 243 Sarir Kalanskiyu . . . . . . . . . . . . . . . . . . . . 329 of Rotliegendes .. .. .. .. .. .. .. .. . 228
Rotliegendes. northwestern Europe ... 190.225 Sarir Tibasti . . . . . . . . . . . . . . . . . . . . . . . . 79 Straak. defined. . . . . . . . . .. . . . . . . . .. . . 11
hematite in .. .. .. .. .. .. .. .. .. .. . 225 Satanka Formation.................. 217 Stringers .................... 87.127.261.281
Interdune deposits in . . . . . . . . . . . . 225 Satellite sensing. See Images. ERTS-1. defined .............. .......... 11
linear dune in . . . . . .. . . . . . . . . . . . . 226 Saudi Arabia. studies in. . . . . . .. . . . . . . 89 In Lagoa dune field. Brazil...... . 127
quartz In .. .. .. . .. .. .. . .. .. .. . .. 225 See olso An Nal'iid. Irq as Subay. Structures. deformational ............ 111. 121
stratification in .. .. .. .. .. .. . .. .. 229 Rub 'al Kheli. Wahiba diuipetion ..................... 127
transverse dunes in .. . .. .. .. . .. . 226 Sands. Zalim. interdune ...................... 107
Rub'al Kheli. Saudi Arabia. . . . . . . . . . . 363 Scanning electron microscope analysis 66 in Brazil................... 109
barchan dune in .. .. .. .. .. .. .. .. 269 Sebkha. interdune. . . . . . . . . . . . . . . . . . . 225 in Killpecker dune field. . . . . . 109
crescentic dune in .. .. .. .. .. .. .. 362 inland........... .... ....... .... 196 penecontemporaneous .. 111.121.229.231
interdune areas in . . . . . . . . . . . . . . 282 Sediment arkosic. . . . . . . . . . . . . . . . . . . . 224 rotated..... ... . ... ...... ... .... 112
linear dune in .. .. . .. . .. .. .. .. ... 260. 382 Sedimentary rocks. Cretaceous. . . . . . . 85 Studies. 88nd-texture .... .. .. .. .. .. .. 4
star dune in ................. 107.275.362 Tertiary. . . . . . . . . . . . . . . . . . . . . . . . 85 statistical ...................... 5
Sedimentary structures. of dunes. . . . . 89 textural . . . . . . . . . . . . . . . . . . . . .. . . 24
S minor .......................... 5 See olso Texture.
primary........................ 5 Subarkose beds..................... 205
Sabhii. Libya ....................... . 8.66 Sediments. alluvial. unconsolidated. . . 29 calcareous ..................... 221
linear dune in .................. . 156.161 fluvial.... .... ....... ...... .... 347
88nd rose for .................. . 165 Salfs. See Dune. linear. T
sel dune in .................... . 103 Senegal. See Sahara. western.
Sahara Desert ..................... . 84 Sensing. remote. See Images. ERTS-1. TaCU8rembO Sandstone. Uruguay. . . . . 235
barchan dune in ............... . 269 Serra Geral ba881tic lava flows. . . . . . . 234 Takla Makan Desert. Chine . . . . . . . . .. 381. 385
crescentic dune in ............ .. 310 Settling. mineral . .. .. .. .. . .. .. .. . .. . 24 barchanoid dune in. 142. 152.381.387. 389
interdune area In .............. . 282 Shale ............................ 85.217.237 complex dune in ................ 381. 387

Digitized by Google
 INDEX 429
Page Page Page
Takla Makan Deaart. China-Continued Treils of invertebrate animals-Continued
cl'BllCllntic dUDB in . . . . . . . . . . . .. 387. 389 in Lyons SandstODB. . .. . . . . . 219 White Rim SandstODB................ 190
dome duna in .............. 280.381.391 Trensport distence. effect on pigment. 80 White Sands National Monument ...... 6. 33.89
liDBar dune in ....... " . . . . . . . 381 Transvaal. See Ceve Sendstone. barchenoid ridge dune In ..... 90. 97. 152
reversing dune in . . . . . . . . . . 389 Makgabeng Formation. dome duna in . . . . . . . . . . . . .. 98. 152. 279
star dune in ............... . 389 Waterberg Supergroup gypsum send in . . . . . . . . . . . . 24. 28
trensverse dune in . . . . . . . . . . . . . 392 Trenching ...................... 5. 89 interdune area in . . . . . . . . . . . . 282
Tamarilr ramosiBsima . . . . . . . . . . . . . . . 389 Tria88lc ..... 189.190.197.209.228.236.357 interdune deposit in . . . . . . . . . 109. 206
Tartm Besin .................... 381. 385 Tunisia. gypsum caliches In . . . . . . 67 parabolic dune in ......... 95. 152. 277
Tensleep Sandstonea. Wyaming .. 35.221.237 See olso Grand Erg Ortental; Sa reversing dune in............... 106
Terminology. dune.................. 284 hera. northern. transverse dune in .......... 93.97. 193
Tertiary ................... 65.347.357 plaDBr cl'088-8tratification . . . . . . . 206
late............................ 385 U.V
sedimentary racks . . . . . . . . . . . 65 Wind. effect on dune formation ... 73. 94. 99.
volcanic rocks . . . . . . . . . . . . . 65
102.107.152.227.243.373
U.S.S.R. See Peski Karakumy.
Textural parameters . . . . . . . . . . . 24. 55 effective ........... 141. 155. 262. 280.
Uplngton. South Africa. linear dune in. 156
send sheets in.................. 156
282.317.343.376
Texture. of Berun Goyot Formation. . 231 See olso Wind. send-moving.
of BotucatU Sandstone. . . . . . . . . . 235 Upper Nemegt Beds. See Berun Goyat
energy .................... 141. 150
of Caspar Formation. . . . . . . . . . . 221 Formation.
regime ............. 8. 141. 157.327.336.
of Cave SandstoDB . . . . . . . . . . . 237 Uruguay. SeeTacuarem~.
341.354.361.373
of Coconino Sandstone. . . . . . . . 199 aSSOCiated with berchenoid
of De Chelly SandstoDB . . . . . . 205 ValleedeLAzaouk. barchenduDBln.. 317
dunes................. 152
of Frodshem Member. . . . . . . . . 228 crescentic dune in .............. 321
associated with linear dunes. 155. 262
of Lower Mottled Sandstone. . . . 227 send sheets in ................ 320
aSSOCiated with star dunes. . . 165
of Lyons Sandstone. . . . . . . . . . . . . . 217 Vegetation .... 122. 235. 340. 349. 357. 386. 389
classification of .......... 141. 148
of Namib Deaart . . . . . . . . . . . . . . . 349 effect on dune formation ....... 94.97.235.
sendmoving ............... 87. 103. 141
of Navajo SendstoDB . . . . . . . . . . 210 336.358
of RoWegendes . . . . . . . . . . . . . 225 Victoria Valley. Antarctica. reversing Wingate Sandstone.................. 190
Ther Desert. India end Pekistan . . . . . 375 dune in ................ 104 Wu.lan-pu.ho Shemo . . . . . . . . . . . . . . . . 381
barchenoid dune in. . . . . . . . . . . . . . 154 Volcanic racks. Tertiary............. 65 Wyoming. western. parabolic dune in . 96
crescentic dune in ............ 271. 375
feldspar in .. . . . . . . . . . . . . . . . . . . 376 W X.Y.Z
parabolic dune in ............ 95. 277. 375
quartz in . . . . . . . . . . . . . . . . . . . . 376 Waddan. Libya. . . . . . . . . . . . . . . . . . . . . 66 X.ray analYSis. energy-dispersive . . . . . 66
Thrusts. stairstep . . . . . . . . . . . . . . . . 105 Wadis ............................ 196. 225 X.ray diffraction analysis. . . . . . . . . . . . 66
Toroweap Formation. Grand Canyon.. 197 Wahiba Sands. Arabia. . . . . . . . . . . . . .. 363. 368
Tourmaline ...... . . . . . . . . . . . . .. 334. 356 barchenold dune in. . . . . . . . . . . . 154 Yamalik Dunes ..................... 381. 389
Trace fD88ils. in Casper Formation. . . . 224 Warps. See Folds. gentle. Yapulai Shemo .................... 381
Traces. faunal. . . . . . . . . . . . . . . . . . .. 190. 202 Waterberg Supergroup. Transvaal. . . . 191 Yardangs. . . . . . . . . . . . . . . . . . . . . . . . . . 386
Tracings. laboratory . . . . . . . . . . . . . . . . 173 Weber Sandstones ................ 35.190.243 Yellow Sands of Durham. England .... 190.226
Tracks. of vertebrete animals .... 189. 193. 197 Wentworth grade scale. . . . . . . . . . . . . . 23
in Cave Sendstone . . . . . . . . . . . . . 237 Wertz field. Wyoming... . . . . . . . . . . . 243 Zallm. Saudi Arabia. star dune in . . . . . 106
Treils. of invertebrate animals. in C~ Western Sahara. crescentic dune in ... 310. 313 Zambia. feldspathic arenites in....... 190
nino Sandstone. . . . . . . . 202 immobile dune in................ 312 Zion Canyon. Utah ................... 212. 214
in De Chelly Sandstone. . . . . . . . 207 linear dune in .................. 310. 313 Zircon ............................. 334. 356

Digitized by Google
Digitized by Google .
Digitized by"GOOg Ie
 ,
~Q
.- ec

Scripps Institution of Oceanography Library

ersi Cal ia, Dieg
DATE DUE
--
M
- -
I
--' -
30 ,~ -- -- -
. M., nrMT
..... A
-v

~ - -- -- -
JU"" "'t .;r.tr."'

-- P 25 teL -- -- -

-- -- - -- -- -

--

-- -- - -- -- -

-- -- - -- -- -

-- -- - -- -- -

-- -- - -- -- -

--
- --
SI 23
--
-I -- -
UCSD LibT'.
 ' .
-..

Digitized by Google