See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/248792888

Elastic wave velocities within oceanic layer 2 from sonic full waveform logs in
Deep Sea Drilling Project Holes 395A, 418A, and 504B

Article  in  Journal of Geophysical Research Atmospheres · June 1990

DOI: 10.1029/JB095iB06p09189

CITATIONS READS
22 10

3 authors:

Daniel Moos Philippe Adrien Pezard

independent French National Centre for Scientific Research
144 PUBLICATIONS   3,409 CITATIONS    206 PUBLICATIONS   2,859 CITATIONS   

SEE PROFILE SEE PROFILE

Mike Lovell
University of Leicester
157 PUBLICATIONS   1,060 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Deep drilling of ocean basement 1: DSDP/ODP Hole 504B (equatorial Pacific Ocean) View project

Gas Monitoring in shallow depth View project

All content following this page was uploaded by Philippe Adrien Pezard on 04 February 2020.

The user has requested enhancement of the downloaded file.

 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 95, NO. B6, PAGES 9189-9207, JUNE 10, 1990

Elastic Wave Velocities Within Oceanic Layer 2 From Sonic Full Waveform
Logs in Deep Sea Drilling Project Holes 395A, 418A, and 504B
DANIEL MOOS

Department of Geophysics, Stanford University, Stanford, California

PHILIPPE PEZARD

Borehole Research Group, Lamont-Doherty Geological Observatory, Palisades, New York

MICHAEL LOVELL

Department of Geology, University of Nottingham, Nottingham, England

Multichannel full waveform acoustic logs were recorded in Deep Sea Drilling Project (DSDP) hole
418A during Ocean Drilling Program (ODP) leg 102, DSDP hole 395A during ODP leg 109, and DSDP
hole 504B during ODP leg 111, to provide nearly continuous measurementsof elastic wave velocities
as a function of depth within oceanic layer 2 for different spreading rates and crustal ages. The
velocities depend primarily on the morphologyof the basalt. Massive units have Vp above 5 km/s, Vs
above 2.8 km/s, and Vp/Vs below 1.9. Their velocities increase with depth as expected from crack
closure due to confiningstressand are lower where fractures are present. Vp and Vs within pillows are
quite variable but in general are lower than in massive units, and Vp/Vs is higher. Velocities within
breccias are similar to or lower than those of pillows. Within a given morphology the elastic properties
do not depend on spreadingrate. Within pillows, velocity variations depend primarily on the degree
of alteration infilling and the mineralogy of the infilling matedhal,and secondadhlyon conditions at
emplacement. Coherent shear arrivals are sometimesnot found within shallow pillows, and in extreme
casesthe compressionalarrival is also incoherent, because of scattering resulting from the similarity
between the sonic wavelength and the pillow size. This "sampling bias" in the sonic log is the most
likely cause of somewhat higher average sonic than seismic velocities measured within extrusive
basalts at shallow depths. Other causes include the fact that laterally finite massive units may be
oversampledby borehole measurementsand that voids between pillows and fractures, whose size is
within an order of magnitude of the sonic wavelength, have less effect on sonic than on seismic
velocities. The crustal section can be separatedonly roughly into seismic layers on the basis of the
sonic velocities. Changesin velocity with depth within the extrusive section at a single site giving rise
to previously defined seismiclayers 2A and 2B are due to a changewith depth in the properties of the
pillows, associated with differences in the amount and type of infilling matedhals,rather than to a
change with depth in the relative proportion of pillows and massive units. Because alteration occurs
early and its history is different at different depths,depth-dependentvariationsin pillow properties are
(1) establishedat young age and (2) persisteven in old crust. Within the intrusive section of DSDP hole
504B, velocities are similar to or higher than those measured in saturated minicores at atmospheric
pressure, and seismic and sonic velocities are approximately equal, suggestingthat little large-scale
porosity is present.

INTRODUCTION within ocean basalts. These velocities can be related directly

to the morphology and alteration history of the basalts and
With the routine acquisition during Ocean Drilling Pro- reveal details of the velocity structure of the crust at length
gram (ODP) cruises of high-quality geophysical log measure- scales of 10 cm to a few meters (the scale of macroscopic
ments in Deep Sea Drilling Project (DSDP) and ODP bore- fracturing, the size of pillows, and the thickness of individual
holes (this volume), detailed profiles of the physical flow units). Thus they provide insight into the vertical
properties of the oceanic crust as a function of depth in crust heterogeneity of the oceanic crust and into the causes of the
of varying age and origin are now being obtained. This paper observed seismic velocity structure [e.g., Houtz and Ewing,
presents the results of analyses of multichannel full wave- 1976; Spudich and Orcutt, 1980; Purdy, 1987; Vera, 1989;
form acoustic logs in DSDP holes 395A, 418A, and 504B. Harding et al., 1989]. Comparison of the velocities deter-
Compressional and shear velocities for sonic waves propa- mined by sonic logging to elastic wave velocities measured
gated vertically along the borehole wall are calculated over in cores at ultrasonic frequencies, and to seismic velocities
travel paths of several meters at frequencies of several near the boreholes, further characterizes the large-scale
kilohertz.
porosity and void structure.
The results represent the first direct, combined measure-
ments of compressional and shear elastic wave velocities
CRUSTAL STRUCTURE FROM SEISMIC VELOCITIES
Copyfight 1990 by the American Geophysical Union.
Paper number 90JB00261. Early seismic refraction experiments [e.g., Houtz and
0148-0227/90/90JB-00261 $05.00 Ewing, 1976] revealed a gross layered structure for the
9189
9190 Moos ET AL.: ELASTIC VELOCITIES IN OCEANIC LAYER 2

ophiolite Pacific Ocean Crust

model
magma chamber, but there is still no clear seismic differen-
layered Fanfare ESP tiation of the overlying layer into 2B (pillows and flows) and
model* 2 model
15Ma' _<180Ka"
2C (sheeted dikes). Harding et al. [1989] conclude that
seismic layers 2A (the thin surficial low-velocity layer), 2B
(the remainder of the basaltic sectionto the top of the magma
mostly chamber), and 3 (the solidified magma chamber itself) are
pillows
• iii??•i
2A well represented in young crust but that 2C is simply a
transition region between layers 2 and 3. Vera [1989] agrees
._>e and further suggests that this transition lies above the
mostly =m
2B
flows dike-gabbro contact and is perhaps a metamorphic, rather
than a lithologic, boundary. These new results are summa-
rized in Figure 1, panel c.
sheeted 2C The difficulty of relating seismic structure to drilling
dikes results is illustrated by the summary of results from drilling
in DSDP hole 504B by Becker et al. [1989]. Although the
preponderance of pillows in the uppermost -800 m of
gabbros
3 a
*Houtz and Ewing, 1976
b c
basement drilled at site 504 indicates that these materials are
primarily extrusive, the boundariesseparatingseismiclayers
'SpudichandOrcutt,
1980 :• • • • • -• 2A, 2B, and 2C are unclear. Similar problems are encoun-
**Hardinget al., 1989 Vp (km/s) tered in other DSDP holes. Because of this uncertainty we
will avoid discussingthe velocity structure at sites 395,418,
Fig. 1. Schematic illustrating a range of models of the oceanic and 504 in terms of the seismic layering, except to point out
crust derived from seismicdata, compared to a "reference" ophio-
the positions and thicknesses of layers 2A and 2B from
lite. The progressiverefinement of these models is a consequenceof
both higher-resolutiondata acquisition and improved analysistech- previous studies.
niques.
SITE DESCRIPTIONS

basaltic layer of the oceanic crust (Figure 1, panel a). Layer Multichannel full waveform acoustic logs have been re-
2A, with P wave velocities averaging 3.64 km/s; layer 2B, corded to date in oceanic basalts at three sites, the locations
with velocities averaging 5.19 km/s; and layer 2C, with of which are shown in Figure 2. Hole 395A was drilled into
velocities averaging6.09 km/s, are underlain by layer 3, with 7.3-Ma crust near the center of magnetic anomaly 4, approx-
P wave velocities averaging 6.87 km/s. By analogy with imately 110 km west of the mid-Atlantic Ridge. Hole 418A
ophiolites and direct comparison to the core recovered by was drilled into 110-Ma crust on the southern tip of the
Bermuda Rise in the western Atlantic Ocean. The crust at
DSDP drilling, layer 2A is assumed to be composed prima-
rily of pillow basalts, 2B of pillows and massiveunits, 2C of sites 395 and 418 is inferred to have been produced at the
sheeted dikes, and 3 of gabbros. The thickness of layer 2A same point (approximately 23øN) on the mid-Atlantic Ridge.
Hole 504B was drilled into 5.3-Ma crust about 200 km south
decreases with increasing age, in the Pacific from 700 m at
the ridge to 100 m in 30-Ma crust and in the Atlantic from of the Costa Rica Rift. As such, these three holes sample
1500 m at the ridge to 100 m in 60-Ma crust. Spudich and young and old crust generated at the slow spreading rate
Orcutt [1980] pointed out that the seismic layering was an mid-Atlantic Ridge and young crust generated at the inter-
artifact introduced by the refraction geometry and developed mediate spreading rate Costa Rica Rift.
an alternative model using inversion techniques and syn-
Hole 395A
thetic seismograms that included velocity gradients (e.g.,
Figure 1, panel b). However, seismicand borehole geophys- Hole 395A, drilled during DSDP leg 45, penetrates 93 m of
ical data are still interpreted using the terminology of Houtz sediments and 571 m of basement [Melson et al., 1978]. The
and Ewing [1976]. uppermost112 m were cased, and well bore instabilitiesat a
More recent seismic investigations [Purdy, 1987] reveal brecciated dolerite zone caused loss of the hole below that
extremely low velocities (<2.35 km/s) at the top of basement depth, leaving the basement section open between 112 and
in very young (mid-Atlantic Ridge) crust. Young (< 180 kyr) 609 meters below sea floor (mbsf). Core recovery was
crust on the East Pacific Rise (EPR) has a -100- to 200- extremely low, ranging from less than 6% in a sedimentary
m-thick surfacelayer with P wave velocities of 2.35-2.6 km/s breccia at the top of basement(now largely behind casing)to
and S wave velocities below 0.8 km/s [Harding et al., 1989; 32% in the interval 174-210 mbsf. The recovered basalts
Vera, 1989]. Results further off the axis of the mid-ocean were subdivided into phyric and aphyric units. Combined
ridges suggestthat this feature is ephemeral. For instance, a analysisof recovered cores and of geophysicallogs recorded
singlevelocity of 4.1 km/s provides an adequatefit to seismic during ODP leg 109 resulted in the lithostratigraphic column
resultsfor the uppermost250 m of 7.3-Ma Atlantic crust near shown in Figure 3. Although the recovered material includes
DSDP hole 395A [Purdy, 1987], and in 15-Ma Pacific crust most of the lithologies inferred from later logging, the log
the uppermost low-velocity layer is similarly not observed measurements indicate their true vertical extent. For exam-
[Spudich and Orcutt, 1980]. On the EPR, velocities increase ple, the cobbles (unit 3) are thicker than indicated in the
abruptly at the base of the shallow low-velocity layer and core, and the positions and thicknessesof breccia units 13,
continue to increase smoothly with depth to the inferred top 15 and 17 are somewhat different than a standard treatment
of the magma chamber at about 1.3 km. At 180 ka, layer 3 of the recovered cores might suggest.
velocities of more than 6.5 km/s replace the low-velocity All of these materials have been altered by seawater to
 Moos ET AL.: ELASTIC VELOCITIES IN OCEANIC LAYER 2 9191

45øNi.:::::..../.½!:...,,t:.
I.';'.L"•
• •:..: , .:::':?/'
)•:.":.'.,::'...•f:
'
...::...,..,.,/,•,•,
½...•,'i // .,,-,-
]•
' // I I / !
I i / !I/I/ / / I I
/ / i // d:::'-..
, ..'.'.••:'
I ':':'•:..":'
'"':'"" . .:?..:7."• 180 • 35 • • • 38 • I! ! ! ! • ¾:.'.'...'..'
-':......
• ..'½-"'.' ...:-:..'-•':",.•"/ /, / . , , u I_•38 • • / {i':.•:
I .'.:."•:Y-
'..i::.:-•:-:• / / // / 81
• • • 63
• 53 t ß 5363
ß 81 •?'i';.
i;:.i:!:.::.. .
I .,, _ ^zorn, -'.
'....
ß-.'
I :..:..','...."?:.,•• ., - / ,"• ?7øg½'o,7-,,"
';"-'- / • /-"-..'•'"'"':;'1
"."
•"
......
I ...::•.':.:.:j,• / / •errnuaa / / ,,.,•, /•/'& ,.',zer,•.•- ,', ! ../,.:.•:..': ' I
[:'...::..:.....'
.....'::i;'[ I / I " // ,/ .,• /,/'/ø 38 /•4 •:•<" / Madeira[.;':.'?
' J
30ø'•:-";"';••'":?'.-• , ,•/ • / / 4•-- z,/_/ / / zgø ?' / / 2:'::'
I• - '•i:.:•.l, '•4//114 / , , "%•ti• •'-•-/.z_ / / , _/ • M..,•':""' I
I •:.'::".?,• /,,•,•'• • / I / / i' '•" /' / '/ ø'ø• •' -.:•'""'"' I
II %i!.
• •
•.1:•
/ TM
©.418A ,
•on••. •.I/
P- --.. -•. /
// ./// ,/
/ / / /
Canaries
•F.":
/,.:%
,
I• ..•
'"••
z • 81 /
•'r'-,•,•Z.
.....
395A/
• I
z
/"-'::-':':"
•'•.•7.-'
œ"":'"'"
'"• - - '"'::"::'""•' ,, ,.... / 63 =•--"•""'•.
ß/ , ,.,•.
t-':"
'••'
-.":
•( -..-.• SanJuan • 53• t •o • •o {{(i:.;' • I
I':•ff
' (:• :•• ,• . / • • .., 9,,/9 ' 53•ø 8• Cape
Verde, ):•:; •00kr•

.... •,' ' 'X',

__%.... ½ ,• &• '¾.-:..'-

1' ¾.::.::L:
.... ß
! ½• ,,.•'..-:••' -':?,_L ATLANTIC
OCEAN FZ
0ø• ß,•U•,:' .
75øW 60 ø 45 ø 30 ø 15 ø

Fig. 2. Locationsof DSDP holes395A, 418A, and 504B, alongwith crustalagesin the Atlantic oceandeducedfrom
magneticanomalies[after Hyndman and Salisbury, 1984].

some degree. The predominant alteration occurred at low Hole 504B

temperatures,with carbonate veins, smectites(usually sa-
Hole 504B was drilled duringDSDP legs 69, 70, and 83 and
ponite), zeolites, goethite, opal, and gypsum present in
ODP leg 111; at the completion of ODP leg 111 the hole had
decreasing amounts [Lawrence et al., 1978]. Veins with
reached a depth of almost 1563 mbsf [Becker et al., 1989].
carbonate or other filling, altered basalt glass, and filled
The hole penetrates 274.5 m of sediments before entering
vesicles arecommon. Olivineis sometimes alteredin pl•c•e• basement.Figure 5 showsthe lithostratigraphydetermined
althoughassociatedplagioclaseand pyroxeneare generally
by combined analysis of borehole televiewer logs [Cann et
quite fresh [Juteauet al., 1978].Alteration infillingincreases
al., 1983; Anderson et al., 1985; Newmark et al., 1985;
toward the baseof the cored section,particularlywithin unit
18. Shipboard Scientific Party, 1988] and cores. Pillows and
minor flows are encountered in the uppermost 571.5 m of
basement (846 mbsf), the "extrusive" basalts. Below that
Hole 418A depth is a 209-m transition zone of intermixed pillows, flows,
massive basalts, and dikes. From 780.5 m subbasement
Hole 418A was drilled during DSDP legs 52-53 through
(1055 mbsf) to total depth, sheeted dikes and massive units
324 m of sedimentsand penetrated544 m into oceaniclayer
2 basaltsto a total subbottomdepth of 868 m [Donnelly et predominate. The approximate boundaries of seismiclayers
al., 1979]. Core recovery averaged 72%. The recovered 2A, 2B, and 2C are indicated in the figure [after Becker et al.,
1989].
basaltsare quiteuniformin compositionandare porphyritic,
with plagioclaseas the dominant phenocryst, olivine, and Alteration in the uppermost 1075 m of basement is sum-
minor amountsof spinel. Figure 4 showslithostratigraphy marized by Honnorez et al. [1983] and by Alt et al. [1985].
inferred from recovered cores and from geophysicallogs An uppermost zone (to 584.5 mbsf) is characterized by
recorded during ODP leg 102. As in DSDP hole 395A, log oxidation and hydration with the addition of potassiumand
analysis suggestedsome small shifts in the thicknessesand the loss of silica. Alteration halos are seen around fractures,
positionsof the lithologic units, and in this well, thin altered and pore spacesare pervasively filled with saponite. From
units between the massive units 2B, 2C, and 2D can be seen 584.5 to 835.5 mbsf, calcium is removed, and magnesiumis
in the logs [Broglia and Moos, 1988]. added by alteration. In this interval, groundmass alteration
Alterationis predominantlydue to low-temperatureinter- occurs around fractures, and disseminatedpyrite is found.
action with seawater and increasesthe potassiumand hy- The shallow zone is inferred to have been produced by first
droxyl content,addingsmectite,calcite,pyrite, and zeolites oxidative and then anoxic processes, whereas the lower
primarily as veins and vesicle fillings but also replacing zone is inferred to have formed either at low water/rock
olivine and interstitial groundmass [Holmes, 1988]. Alter- ratios or with seawater which had previously reacted with
ation fillingvoidsbetweenpillowsin the uppermost190m of basalt. Alteration in the interval from 836 to 898 mbsf, at
basement(units 1, 5, and 6A) reachesas much as 25% by temperatures above 100øC, first resulted in the formation of
volume; infilling decreasesabruptly below about 514 mbsf clay, with later formation of zeolites, anhydrite, and calcite,
(the base of breccia unit 6A), and thus the position of the as the fluid composition evolved. In the transition zone and
ancestrallayer 2A/2B boundaryis chosenat that point [e.g., the dikes (below 898 mbsf), three alteration stagesoccurred;
Broglia and Moos, 1988]. initially high (200ø-250øC)temperatures and partially reacted
9192 Moos ET AL.: ELASTIC VELOCITIES IN OCEANIC LAVER 2

Lithology Unit

not
logged 1

2 A2
400
150

• 4 P2
200

5 P3

250
'6
500
7
P4

300

10
11 P5
12

13
350

600
400

??

A3

450 16

500
• '17
700

LEGEND
LEGEND
550 - 18

?? Pillow
basalt • Pillow
basalt
Massive
basalt '• Altered
pillow
A4 Breccia •:[• Massive
pillow
19
600
-- not logged - Cobbles I Massive
basalt
• Breccla
Fig. 3. Geologic column based on core recovered from DSDP
hole 395A during DSDP leg 45. Small numbers in the right-hand
column indicate lithologic units, labeledand positionedaccordingto Fig. 4. Lithology as a function of depth in DSDP hole 418A.
standardDSDP procedures[Melson et al., 1978]. Chemical units are Numbers and divisions in the right-hand column indicate lithostruc-
preceded by the letter A (aphyric) or P (phyric). The positionsand tural units inferred from core recovery [Donnelly et al., 1979]. In the
thicknesses of specific units in the left-hand column have been left-hand column the depths and thicknesses of these units have
shifted slightlyon the basisof geophysicallog data collectedduring been adjusted on the basis of geophysicallog data acquired during
ODP leg 109. The position of the seismiclayer 2A/2B boundary is ODP leg 102, and the position of the ancestrallayer 2A/2B boundary
chosen at the base of breccia unit 17, on the basis of log response at the base of breccia unit 6 is dictated by the abrupt decreasein the
[Moos, 1990]. degree of alteration infilling [Broglia and Moos, 1988].
 Moos ET AL ß ELASTIC VELOCITIES IN OCEANIC LAYER 2 9193

Llthology Unit

2A

Llthology Unit
836

5•, Llthology Unit

53 1350

900
4
400 5
1400
9

16

17
7o
71

IOO0

5OO NO RECOVERY

87--
-89

• =4
25

1562.3
1100
LEG 111
6OO 28

30A

115
33

1200 118 j
.

700

37 o '

LEGEND

] Pillow
basalt
Massive
basalt
'• Breccla
45

47

836 1350

LEGS 69 AND 70 LEG 83

Fig. 5. Geologiccolumnbasedon corerecoveredfrom DSDP hole504BduringDSDP legs69, 70, and83 andODP

leg 111[Cannet al., 1983;Andersonet al., 1985;ShipboardScientificParty, 1988].Smallshiftsin the positionsof the
individual units based on borehole televiewer data [Newmark et al., 1985] are not resolvable on the scale of this figure.
The position of the layer 2A/2B boundaryis uncertain. The base of layer 2B and the top of layer 2C (the
extrusive/intrusiveboundary) are arbitrarily chosen where the first dikes and the last pillows were recovered,
respectively.The transitionzone is the interval betweenthesedepths.

seawater deposited vein minerals, with a second stage of Core Physical Properties
vein mineral depositionat temperaturesup to 380øC.Later,
zeolites formed at temperaturesbelow 200øC.The degree of Shipboard measurementsof the porosity and compres-
alteration is apparently controlled by permeability and is sional wave velocity within cores recovered from holes
more pervasive in the transition zone, the uppermost 100 m 395A, 418A, and 504B were supplementedby further post-
of dikes, and within highly fractured intervals [e.g., Ship- cruise measurements. Porosities range from a few percent to
board Scientific Party, 1988]. more than 10%. In general, compressionalvelocities in-
9194 Moos ET AL..' ELASTIC VELOCITIES IN OCEANIC LAYER 2

Time (ms) acousticenergyin the fluid and if the signalcontainsenergy

o i of sufficientlylow frequency. Following the shear wave is
the first normal mode or pseudo-Rayleighwave, which
propagatesat a velocity slightly below the shearvelocity of
the formation. The Stoneley wave, which is evanescentin
both the fluid and the formation, propagatesat frequencies
belowthat of the refractedwavesat a velocitythat is slightly
lessthan the acousticwave velocity in the fluid. Chengand
Toks6z [1981] and Paillet and White [1982] discussin further
detail the theory of acousticwave propagationin boreholes.
The full waveform acousticloggingtool usedin this study
consistsof a magnetostrictivesourceseparatedfrom a set of
12 piezoelectric receivers by a variable length acoustic
isolator sectionand is centered in the hole by two sets of
bowspring centralizers, one located above the source and
one below the receivers.The receiversare spacedat 0.15-m
intervalswithin an oil-filledhydraulichose,forminga 1.65-m
array. The sourceto first receiver separationcan be varied
shear by changingthe length of the acoustic isolator section. The
configurationof the tool and the processingappliedto the
' compressional
data after recording are presented for each well in Table 1.
Fig. 6. Typical waveforms recorded in a hole drilled through Further details of the recording and processingtechniques
granodioritc. are presented below.

crease with decreasing porosity, ranging from less than 5 Hole 395A
km/s in the mostalteredporousmaterialsto morethan 6.5
A full-waveform sonic log was run in DSDP hole 395A
km/s. No clear differencewas observedbetweenthe prop-
erties of cores taken from fresh massive basalts or fresh duringODP leg 109from just insidethe casingat 112mbsfto
pillows from within the extrusive sections(holes 395A and approximately 590 mbsf [Moos, 1990]. The lowermost sec-
418A and hole 504B above 846 mbsf). However, samples tion of available hole was not logged, as a 10-m-longtem-
recovered from below the transition zone within hole 504B perature/magnetometersonde was attached to the bottom of
had uniformly low porosities(below 3%) and high velocities the tool during the logging run. The raw data were contam-
[Christensen and Salisbury, 1985; Shipboard Scientific inated by a strongmode traveling along a steel cable within
Party, 1988]. the receiver sectionof the tool, in part becauseof this added
weight. However, strongarrivals refracted throughthe for-
DATA COLLECTION mation could be observed within the lowermost 100 m of the
hole and in isolatedsectionsat shallowerdepth.To eliminate
Sonicloggingtoolsare designedto optimizedetermination
the tool wave, the data were filtered using a zero-phase
of the compressionalwave velocity within the rock sur-
3dB/octaveButterworth filter, with 3-dB down points at 8
rounding a well bore from the measurement of the travel
and 30 kHz.
time of refracted compressionalwave energy along the
The filtered data are shown in Figure 7. Although the
borehole wall. Additional information, includingin many
casesthe shear wave velocity within the rock surrounding removal of the low-frequencytool wave also severelyatten-
the well bore, can be extracted from analysisof full wave- uated the shear and Stoneley wave modes, variations in the
form acousticlogs. However, accurateinterpretationof the first-arrivingP wave mode amplitudeare quite clear. Zones
full waveformsrequires some theoreticalknowledgeof the with strong, coherent arrivals occur at 180-185, 190-198,
properties of the generated wave field. 251-261, and 291-307 mbsf, and the interval from 320 to 420
Figure 6 shows a set of waveforms recorded in grano- mbsf is characterizedby steadily increasingwaveform co-
diorite, which show clearly the characteristicsof full wave- herence with depth. Below 420 mbsf the coherence de-
form sonic data. As shown in the figure, the first energy creasesabruptly and then increaseswith depth until at 520
arrivingat the receiversis a refractedcompressional wave, mbsf,strongringingarrivalsare present.Thesepersistto the
followed by a refracted shear wave, if the formation shear total loggeddepth, with the exception of a narrow zone at
wave velocity is higher than the propagationvelocity of 555 mbsf correspondingto a washed-out interval.

TABLE 1. Recording and ProcessingCharacteristicsfor the Multichannel Sonic Data Used in

This Study

Hole Spacing,m Tool Motion Depth Interval, m Gain Filter, kHz

395A 2.95 down 0.5 fixed 8-30
418A 1.95 up 0.3 fixed 5-20
504B 1.95 down 0.4-0.5 variable 4-20
 Moos ET AL.' ELASTIC VELOCITIESIN OCEANIC LAVER 2 9195

Time (ms) Velocity (km/s)

0.0 1.0 2.0 3.0 654 3 2 1.5

Ii• I ii , Ir

200

300

4OO

500

50 100 150 200

Slowness (l•s/ft)

Fig. 7. (Left) Sonicwaveformsrecordedat receiver1, 2.95m belowthe source,in DSDP hole395AduringODP

leg 109.The datawererecordedduringa singleloweringof thetool,anddepthsareinferredfromtheobserver' s notes.
The waveformswere filtered usinga band-passfilter with 3-dB down points at 8 and 30 kHz. (Right) Single-pass
semblance analysiscalculated
usingall 12receiverswith a timewindowof 300/xsis shown.The semblance is depicted
as solid above 0.2.
9196 Moos ET AL.' ELASTICVELOCITIESIN OCEANICLAYER 2

Time (ms) Velocity (km/s)

0.0 0.5 1.0 1.5 2.0 2.5 6 5 4 3 2 !.5
i

500 -

600 -

700 -

50 I00 150 200

Slowness (l•s/ft)

Fig. 8. (Left) Sonicwaveformsrecordedat receiver4, 2.55 m belowthe source,in DSDP hole418A duringODP
leg 102.The datawererecordedwith the tool movinguphole,anddepthswerereaddirectlyfrom a depthencoderon
the wire line. The waveformswerefilteredusinga band-pass filterwith 3-dBdownpointsat 5 and 20 kHz. (Right)
Single-pass semblanceanalysiscalculated
usingall 12receiverswitha timewindowof 300/xsis shown.The semblance
is depicted as solid above 0.2.
 Moos ET AL.' ELASTIC VELOCITIES IN OCEANIC LAYER 2 9197

Time (ms)
0.0 1.0
Velocity
(kin/s) Hole 418A
? ,• 12
Full waveform acoustic log data were recorded in DSDP
hole 418A from just above the bottom of the hole at 789 mbsf
to just above a bridge at 464 mbsf, with the sonde moving at
a constant rate up the hole and with a constant gain applied
to the data [Moos, 1988]. This is the optimum mode of
recording, but of the three boreholes, hole 418A is the only
400 one where data were recorded in this manner. The upper-
most 140 m of basement were not logged with the multichan-
nel logging tool. To remove "road noise" associated with
the motion of the tool, the data were filtered prior to analysis
using a zero-phase 3dB/octave Butterworth filter with 3-dB
500
down points at 5 and 20 kHz.
Figure 8 presents the filtered waveforms recorded at
receiver 4, 2.4 m below the source. The first arrival is the
refracted compressional wave. The low-frequency constant
600 arrival time phase at about 1.7 ms is the Stoneley wave. The
shear wave can be seen clearly in the lowermost 50 m of the
hole and within massive units (for instance, units 10 and
ß.,,_,j. Its arrival time, shortly after the compressional first
break, varies in the same sense but more than the compres-
700
sional arrival as the elastic properties of the formation
change. A breccia (cored unit 6A) can be seen by late
arrivals at 504-514 mbsf. Within the altered pillows above
this depth, amplitude and coherence are much lower than in
the remainder of the hole. A narrow breccia/fractured inter-
800
val at about 620 mbsf is associated with lower amplitudes
and later arrivals. Comparing these data with those obtained
at holes 395A and 504B reveals considerably higher ampli-
tudes within this section of older crust.

Hole 504B

A multichannel sonic log was recorded in DSDP hole 504B

during ODP leg 111 from the casing to a depth of 1526 mbsf.
The log was recorded in segmentswhile the tool was moving
downhole, and the data density depends on the rate of
descent. Two logging runs were made. The first was prior to
leg 111 drilling and recorded data in the open hole from the
bottom of the casing to a depth of 776 mbsf, at which point
I100
the log was stopped to avoid exceeding the temperature
limitations of the logging sonde. The second set of data was
recorded from 756 to 1526 mbsf, again with the tool moving
downhole. The first pass was started and stopped a number
of times, and so although over 2600 sets of data were
1200
recorded, the data density in the open hole was considerably
less than the 0.13-m interval that a simple average would
suggest. In the uppermost 150 m one complete suite was
recorded approximately every 0.52 m; from 415 to 591 mbsf
1300 this was reduced to every 0.4 m; and in the remainder of the
hole an entire suite was recorded every 0.5 m. The downhole
gains also varied during the recording. Gains were quite high
and uniform across the receiver array until a depth of 534
mbsf and were reduced by 50% at that depth, remaining
1400

Fig. 9. (Opposite) (Left) Sonic waveforms recorded at receiver

1, 1.95 m below the source, in DSDP hole 504B during ODP leg 111.
1500
The data were recorded during two lowerings of the tool, and depths
are inferred from the observer's notes. The waveforms were filtered
using a band-pass filter with 3-dB down points at 4 and 20 kHz.
50 I00 150 200
Slowness (•zs/ft)
(Right) Single-passsemblanceanalysiscalculated usingall 12 receiv-
ers with a time window of 300 /•s is shown. The semblance is
depicted as solid above 0.25.
9198 Moos ET AL.: ELASTIC VELOCITIES IN OCEANIC LAYER 2

Time (ms) coherenceand 0 is a perfectly random signal. The semblance

0 1 2 for a given arrival will be highest when the lag at each
receiver is the product of the true inverse velocity and the
offset. The typical window length is 200/as, or about 2 full
cycles of a 10-kHz signal. The start time of the window
varies, depending on which arrival is being studied. To
determine the propagation velocity of a discrete phase, the
window is fixed by selectinga start, or offset, time at the first
receiver, as shown in Figure 10. Alternatively, the start time
can be fixed at the source to study the entire wave train (P,
$, and the borehole modes) in a single pass.
Figure 11 shows a semblance panel, with semblance
magnitude plotted as a function of start time and slowness.
The peaks corresponding to the P, S, and Stoneley waves
are labeled, along with paths through the data for a start time
fixed at the source (path a) and at the first receiver for the P
wave velocity calculation (path b) and the S wave velocity
calculation (path c).
AT( = SLOWNESS)
AX The results of a single-pass semblance calculation (as
illustrated schematically by path a in Figure 11) using the
Fig. 10. Schematicshowingthe path throughfull waveform data entire set of 12 receivers in each well with a range of
of a semblance window optimized for the determination of shear
wave velocity, for data collected at 628 mbsf in DSDP hole 418A
slownessesbetween45 and 210/as/ft (velocitiesbetween6.8
[after Moos, 1988]. The parameters for the semblance calculation and 1.4 km/s) are shown along with the waveforms in Figures
are the offset time (in this case fixed at the first receiver) and the 7-9. The semblancepeak at about 50-70/as/ft (4.35-6.1 km/s)
slowness, which defines the move-out of the window across the correspondsto the compressionalwave, the peak at 80-120
receiver array. The window length of 200 /as is that used for the
determination of final velocities from the full waveform data in holes
/as/ft (2.54-3.81 km/s) correspondsto the shear wave, and
395A, 418A, and 504B. Stoneley and direct fluid arrivals are associated with the
semblance peak at about 200 /as/ft (1.52 km/s). In these
figuresvariations in semblanceare quite clear and generally
constant from that point to total depth. The data were correlate with variations in wave mode amplitudes. For
filtered using a zero-phase 3dB/octave band-pass Butter- example, in hole 418A the breccia unit 5 (above 500 mbsf)
worth filter with 3-dB down points at 4 and 20 kHz, to has very low semblances, and the massive units 9 and 10
remove low-frequency "road noise" due to tool motion and (676-686.5 mbsf) have very high semblancesand consistent
high frequencies generated by the downhole electronics. low slownesses (high velocities). Similarly, in holes 504B
The filtered waveforms are presented in Figure 9. The
waveforms are quite variable in character in the uppermost
570 m, where there is little evidence of a depth-dependent
trend. Within this depth range, prominent high-amplitude Velocity(km/s)
intervals correlate with massive units (for example, unit 2D). 654 3 2 1.5 //
The gain reduction at 534 mbsf does not significantly affect 1500 I I I I ._ I
St
the character of the data, although the reduction in sampling
density at 591 mbsf is clear. The transition zone (-•570 mbsf
to --•780 mbsf) has very low amplitude data, below which
amplitudes increase and arrival times decrease through the .E 1000
dike screen and sheeted dikes. Although the first-arriving P -• ½ S
wave is generally quite clear, the shear arrival is difficult to
discern. With the exception of short intervals a late-arriving
low-frequency mode (the Stoneley wave) is evident through-
5OO
out the logged depth.
40 90 140 190 240

Slowness (ias/ft)
ELASTIC VELOCITIES FROM FULL WAVEFORMS
Fig. 11. The magnitudeof the semblanceas a function of offset
The velocities of compressional and shear waves are time at the first receiver and slowness(inverse velocity), for the data
calculated using a modified semblancetechnique. This tech- shown in Figure 7. The P, S, and Stoneley velocities can be
nique has been employed by a variety of researchers to computed from the positions of the associated maxima. One-
dimensional trajectories through the data illustrate (path a) a 260-/xs
analyze sonic waveforms and was first discussedby Kimball offset time fixed at the source, for a source-to-first-receiver offset of
and Marzetta [1984]. Semblance is simply a measure of the 1.95 m, (path b) an offset time fixed at the first receiver to optimize
coherenceof a windowed portion of a signalrecorded by an the measurement of P wave velocity, and (path c) an offset time
array of receivers. The window is chosen to coincide with a fixed at the first receiver to optimize determination of the S wave
mode whose velocity one wishes to determine, as illustrated velocity. Trajectory a defines the single-passpath for which sem-
blance is displayed as a function of depth in Figures 8 and 9. Both
in Figure 10. Its position at each receiver is the product of the semblanceimage and the path for data from hole 395A (Figure 7)
the receiver offset and an assumed slowness (inverse veloc- are somewhat different, because of the longer source-to-
ity). Semblance varies between 0 and 1, where 1 is perfect first-receiver offset (see Table 1).
 Moos ET AL ' ELASTIC VELOCITIES IN OCEANIC LAYER 2 9199

and 395A, massive units are associated with high P wave

semblances and velocities. velocity
(km/s)velocity
(km/s)I (mbsf) VpNs
The final velocities were calculatedusing start times fixed 1
I
2 3 4

at the first receiver, as illustrated by path b for the P wave 234567 1 2 3

and path c for the shear wave in Figure 11. Only the six
closest receivers were used in these calculations, and veloc-
ity ranges were restricted to 6.5-3.2 km/s (compressional
window) and 4.3-1.7 km/s (shear window). The choice of
receivers was a compromisebetween optimizingthe discrim-
ination of short-wavelengthvelocity variations by keeping
the array length small and improving the stability of the
method by increasing the number of receivers. The veloci-
ties determined in this way were consideredmeaningful only
if the semblance was above 0.4. In hole 504B this criterion
was relaxed somewhat, as semblances in this well were low
becauseof "road noise" and the occasionalpresencewithin
the waveform suites of "dead" traces. In hole 418A, veloc-
ities in the depth interval above 490 mbsf were determined
from Schlumbergerlong-spacedsonic data, using the proce-
dures outlined by Barton et al. [1989].
The results are plotted in Figures 12, 13, and 14. P wave
velocities are shaded where they are above 5.0 km/s, and S

P-wave
] S-wave
] Depth
velocity(km/s) velocity(km/s) (m bsf)
1 2 3 4
Vp/Vs

i
2 3456 7 1 2 3

Fig. 13. Compressionaland shear velocities and their ratio as a

function of depth below sea floor in DSDP hole 418A. Also shown
are ultrasonicvelocities measuredin core samplesrecovered during
drilling [Donnelly et al., 1979], and two different seismic compres-
sional velocity profiles determined from oblique seismic data by
Swift et al. [1988]. The solid line shows interval velocities for
near-vertical travel paths between adjacent geophonepositions. The
dashed line shows a velocity gradient based on inversion of the
oblique data.

wave velocities where they are above 2.8 km/s. Also plotted
in these figures are seismic velocity profiles from experi-
ments near each hole and shipboard-determinedcompres-
sional velocities measured at ultrasonic frequencies within
core samples.

Hole 395A

The compressionalvelocity computed from the filtered

data acrossthe entire depth interval (Figure 12) rangesfrom
about 3.0 to more than 6.0 km/s with an average of 4.7 km/s.
Although compressional velocities could be determined
throughoutmost of the loggedsection, shear wave velocities
couldbe determinedin lessthan 30% of the depth range. The
Fig. 12. Compressionaland shear velocities and their ratio as a massive units (188-210 mbsf) yielded reliable shear and
function of depth below sea floor in DSDP hole 395A. Also shown compressionalvelocities, and shear wave velocities were
are ultrasonicvelocities measuredin core samplesrecoveredduring generally reliable in the interval 520-580 mbsf. Where both
drilling [Melson et al., 1978] and two different seismic compres- shear and compressionalvelocities were measured, Vp/Vs
sional velocity profiles at the site. The dashedline showsa constant
velocity of 4.47 km/s [from Jacobson et al., 1984]; the solid line
ranges from 1.7 to 2.1. The average Vp/Vs is 1.86. Labora-
shows a constant velocity in the uppermost 250 m of basementof tory velocities are generally higher than sonic velocities
4.12 km/s [from Purdy, 1987]. within the pillows at shallow depth, although within massive
9200 MoosETAL..'ELASTIC
VELOCITIES
IN OCEANIC
LAYER2

units and the deeperpillows, averageultrasonicand sonic

velocitiesare approximatelyequal.
velocdy
(km/s)velocity

234567
1
I
2

I
(km/s) (mb•i)
3

1
Vp/Vs
2
• 3
Jacobsonet al. [1984]tabulatedthe velocityresultsof four
seismicsurveys,includingtwo near site 395. The bestmodel
fit of theirseismicdatarequiredan averagevelocityof 4.47
km/sin thedepthintervalpenetrated by the hole.However,
Hussonget al. [1978]earlierfounda higherregionalaverage
velocityof 4.61 km/s, belowa 150-m-thicksedimentcap.
The resultsof a recent surveyare consistentwith a constant
velocityof 4.12 (+0.05) km/s in the uppermost250 m near
site395 [Purdy,1987].Theseare all somewhathigherthan
the average of the sonic compressionalwave velocities.
Giventhe fact that the lowestvelocitymaterialscouldnot be
measured by the log, the differences between seismic and
sonicvelocitiescouldeasilybe simplydueto "samplebias."

Hole 418A

Figure 13 showsresultsof the semblanceanalysisfor

DSDPhole418A. Vp variesbetween4 and6.4 kin/s,andVs
variessimilarlybetween2 and 3.5 km/s. Vp/Vs generally
rangesfrom 1.75to 2.1. P wavevelocitieswithinthepillows
abovebrecciasubunit6A (506-514mbsf)are quitevariable
and generally are below 4.5 km/s, whereas below subunit
6A, velocitieswithin the pillowsgenerallyexceed5 km/s.
Massiveunits,includingthoseat shallowdepth(units2B,
2C, and 2D) have very highvelocities(above6 km/s and 3.5
km/sfor P andS waves,respectively).Velocitiesare quite
low in the basaltbreccia.As at hole 395A, laboratory-
determinedultrasonicvelocitiesgenerallyboundthe log
velocitiesfrom above;withinmassiveunits,log andlabora-
tory results are approximatelythe same.
Seismic velocities in hole 418A were determined from an
obliqueseismicexperiment (OSE) [Swiftet al. 1988]using
differences
in verticaltraveltimesto geophones clampedat
41, 81,230, 330, and 430 m into basement. These velocities
are shownin the figure,alongwith a velocityprofiledeter-
minedfrominversionof theOSE data.The seismic velocity
within the interval41-81 m into basementis quite low,
possiblybecauseof the fact that the uppergeophonewas
clampedin a laterally finite massiveunit, and the lower
geophonein a pillow;unusuallyearly arrivalsfor propaga-
tion along the massiveunit would yield velocitiesfor the
intervalthat are too low. Even excludingthis interval,the
sonicvelocitiesare somewhathigheroverall than the inter-
val seismicvelocities.Interestingly,the inversionresultsare
moreconsistent withthelogdata.Swiftet al. [1988]suggest
that differencesbetweenlog and seismicresultsare due to
the fact that hole 418A was drilledthroughanomalously
competentrock, and thereforethe seismicdata betterrep-
resent the regional character of the crust.

Hole 504B

Figure 14 showscompressionaland shear velocitiesas a

functionof depthin DSDPhole504B.Compressional
veloc-

Fig. 14. (Opposite) Compressionaland shear velocities and

theirratio as a functionof depthbelowseafloorin DSDP hole504B.
Also shownare ultrasonic
velocitiesmeasured
in core samples
recoveredduringdrilling[Cannet al., 1983;Andersonet al., 1985;
ShipboardScientificParty, 1988]andaveragevelocitiesfrom verti-
cal travel time differencesto well bore clampedgeophonesat
selecteddepths[ShipboardScientificParty, 1988].
 MoosETAL.:ELASTIC
VELOCITIES
IN OCEANIC
LAYER
2 9201

TABLE2. Results
of Semblance
Analyses
ofWaveform
Suites
at Selected
Depths
WithinDSDPHoles395A,418A,and504B
LithologicUnit Morphology(From
Hole Depth,
mbsf (From
Core) Core) Velocity,
km/s Ratio Semblance
395A 208 4 (bottom) massivebasalt 5.7/3.2 1.78 0.8/0.7
418A 684 10 massivebasalt 5.8/3.2 1.81 0.7/0.6
504B 318 2D massive(aquifer
cap?) 6.1/3.0 2.01 0.9/0.8
504B 969 61 massivebasalt/fractured 5.5/3.1 1.77 0.6/0.5
504B 1407 163 massive(dikes) 6.5/3.4 1.91 0.5/0.7
418A 484 5 pillow
(shallow) 4.7/none ... 0.4/none
504B 275 1 pillow 3.2/none "- 0.6/none
504B 326 3A pillow
(acquifer?) 3.6/none '" 0.8/none
327.5 3.2/none '" 0.75/none
504B 825 48 pillow 5.5/none -" 0.4/none
395A 174 3 (bottom) cobbles 4.9/none ... 0.75/none
395A 472 pillow 4.5/none ... 0.8/none
504B 393 3C pillow
(highrecovery) 5.4/2.7 2.00 0.8/0.8
418A 584 6B pillow 6.0/2.8 2.14 0.5/0.4
418A 764 13C pillow 6.0/3.3 1.82 O.8/0.5
395A 543 pillow
(altered/infilled) 5.3/2.7 1.96 0.8/0.7
395A 585 pillow
(highly
altered) 5.6/2.9 1.93 0.9/0.8
418A 509 6A breccia 4.4/2.3 1.91 0.6/0.4
395A 350 breccia 3.6/none '" 0.8/none

ities could be determined throughout the logged depth, breccias,or massivebasalts.In thissectionthe propagation
although in largesections
of thehole,shearsemblanceswere characteristicsof each of these will be studiedby consider-
below0.4, andthusshearwavevelocitiesareconsidered less ingmeasurements
of VpandVsanda qualitative
estimate
of
reliable.Velocitiesgenerallyincreasewith depth,but in a coherencedeterminedfrom analysis of semblancepanels
stepwisefashion.Compressional velocitiesas low as 3.2 (seeFigure11).Thevelocities arethe semblance-weighted
km/swere measuredin the shallowestpillows,but below 350 averageof the slownesses withinthe semblance maximum
mbsf,Vp is above4 km/sevenin thepillowunits.Vsvaries for eachphase,computed withina windowchoseninterac-
between 2 and 3 km/s above 900 m. Within the sheeteddikes, tivelywhichspansa rangeof slownesses (+-10/as/ft)and
highvelocities (Vp above6 km/s;Vs above3.5 km/s)were start-times (200/as),asillustrated
onthefigure.Thevalueof
measured. Where both velocities were measured, Vp/Vs is thevelocitycalculated in thisway is relativelyinsensitive
to
generally between1.8and2.1, exceptwithinthetransition the size of the window and is generallyquite closeto the
zone,whereratiosareashighas2.5.Below1250mbsfVp/Vs velocityvaluecomputed from the (slowness, starttime)
is between 1.75 and 1.9. pointforwhichtheactualmaximum semblance isfound.The
Shipboardlaboratorymeasurements on recoveredcore qualitativeestimateof coherence is simplythe maximum
[Cannet al., 1983;Andersonet al., 1985;ShipboardScien- semblancewithin the window. A given phaseis assumedto
tificParty, 1988]generally
boundthe sonicvelocities
from be presentonlyif a semblance
maximumis foundin the
above,with laboratoryand sonicmeasurements
beingap- expectedregionof the semblance panel.
proximately
equalwithinmassiveunits.Laboratoryultra- Table 2 showsa summaryof the resultsfor a set of 19
sonicvelocitiesmeasuredduringleg 111are generallylower waveforms, sampling eachof the differentlithologieswithin
than sonic velocities below 1325 mbsf. They are also sys- eachborehole.Comparison of thesedatabothwithina given
tematicallylower than velocitiesmeasuredduringleg 83, well and acrossthe set of three wells reveals the range of
whichsuggests perhapsa systematic differenceassociated characteristicsof wavepropagation withinthesebasaltsand
with a changein the measurementtechnique. providesinsights intothevelocityresults whicharethemain
The compressional velocitiesarecloseto thosemeasured theme of this paper.
by vertical seismicprofiling(VSP) duringODP leg 111 Massivebasaltshave highvelocities,essentiallyequalto
[Shipboard Scientific Party, 1988],exceptin the shallowest those determined in the laboratory on cores, and high
section(above500 mbsf),wherelog velocitiesare higher semblancecoherences,even at shallowdepths.Fracturing
than seismic velocities, and within the transition zone,
within massiveunits, as, for example,at 969 mbsf in hole
where sonic velocities are lower than seismic velocities. In
504B,lowersthe velocitiesbut in this casedoesnot affect
contrast to results at holes 395A and 418A, relatively com-
the velocityratio. The semblance is a bit lower, however.
pletesoniclogdatain theuppermost fewhundred metersof Pillowshave quite variablecompressional wave veloci-
basementrevealsignificantly higherlog than seismicveloc-
ities.
ties, rangingfrom lessthan 3.5 to as muchas 6 km/s.
Althoughalmostall of thesepillowunitsare expectedto
propagate a refractedshearwave,sheararrivalsare often
SLOWNESS/ARRIVAL TIME PROCESSING notdetected,except(for example)in unitscharacterized by
OF FULL WAVEFORM SUITES
highrecovery(e.g., 393mbsfin hole504B)or significant
In order to studythe systematics of wave propagation amountsof infilling(e.g., the lowermostsectionof hole
within the differentmorphologicunits encounteredin oce- 395A)or a "massive"character (e.g., 764m in hole418A).
aniclayer2, we followthe systememployedto describethe This is likely due to scatteringwithin the pillowswhich
recoveredcoresand categorizetheseunitseitheras pillows, affects the shear wave more than the compressionalwave.
9202 MoosETAL.'ELASTIC
VELOCITIES
IN L)CEANIC
LAYER
2

a MASSIVE UNITS b PILLOWS

DSDP HOLE 504B 5.4 Ma
DSDP HOLE 504B 5.4 Ma
4.0 øI 4.0

. •

3.0-
•- 3.0-
E
I
,

._

2.0-
2.0-

1.0
I I I I I 1.0

2.0 3.0 4.0 5.0 6.0 7.0

2.0 3.0 4.0 5.0 6.0 7.0

Compressionalvelocity (km/s)
Compressional
velocity(km/s)
DSDP HOLE 395A 7.3 Ma DSDP HOLE 395A 7.3 Ma
4.0 4.0

3.0 •
3.0-

; i i ':'
2.0 m 2.0 -

1.0
I
1.0
i i

2.0 3.0 4.0 5.0 6.0 7.0 2.0 3.0 4.0 5.0 6.0 7.0

Compressionalvelocity (km/s) compressionalvelocity (km/s)

DSDP HOLE 418A 110 Ma DSDP HOLE 418A 110 Ma

4.0 4.0
t
, i

[ ,

3,0 i 3.0

2.0

1.0 1.0
i I i

2.0 3.0 4.0 5.0 6.0 7.0 2.0 3.0 4.0 5.0 6.0 7.0
Compressional
velocity
(km/s) Compressional
velocity
(km/s)
Fig. 15. Summary crossplot
of Vsversus
Vpin (a) massive
units,(b)pillows,
and(c)breccias,
Dataareshown
separately
foreachofthethreeboreholes
tohighlight
differences
between basalts
ofdifferent
ageandorigin.
Datafrom
breccias
inhole395Awhere
bothVpandVsweremeasured
weretoofewtobeincluded
inthissummary.
Linesof
constant
Vp/Vs(1.8and2.0forthetopandbottomlines,respectively)
areshown,along
with(dashed
line)a linearleast
squares
fittoultrasonic
velocities
measured
atinferred
insitudifferential
t}ressures,
averaged
bySalisbury
etal.[1985].
 Moos ET AL.' ELASTIC VELOCITIES IN OCEANIC LAYER 2 9203

c BRECCIA infilling, on the basis of both log response and core descrip-
tions [see Moos, 1990].
DSDP HOLE 504B 5.4 Ma
4.0
DISCUSSION

Elastic Wave Velocities and Lithologic Type

3.0 i To study more systematically the differences between

lithologic types within this set of three wells, Figure 15
presents cross plots of Vs versus Vp for massive basalts,
pillows, and breccias. The numbers plotted in this figure
correspond to the number of points with Vp and Vs values
2.0 m
within +-0.1 km/s of the center of each character. Superim-
posed on the plots are lines of constant Vp/Vs. The top line
is for Vp/Vs = 1.8. The bottom line is for Vp/Vs = 2.0. A
linear least squares fit to averaged P and S wave velocities
1.0 measured in cores from DSDP hole 504B at atmospheric
pore pressure and confining pressure equivalent to the
2.0 3.0 4.0 5.0 6.0 7.0
differential pressure at the sample depth, as reported by
Compressional velocity (km/s) Salisbury et al. [1985], is also shown on these plots.
Differences between the elastic velocities measured in
massive units intersected by these three holes are quite
DSDP HOLE 418A 110 Ma subtle. These units behave like classic microcracked solids
4.0
[e.g., O'Connell and Budianski, 1974; Cheng and ToksOz,
1981], with Vp/Vs decreasing with increasing Vp in holes
504B and 418A. Too few data points were recorded in hole
395A to comment on the trend there. Vp/Vs is below 2.0 with
3.0 - the exception of a very few points. The velocity ratio is
highest in hole 395A and lowest in hole 418A. This could be
ascribed to a slight change in the crack aspect ratio distri-
bution within 110-Ma crust compared to crust of much
younger age. The highest velocities were recorded in hole
2.0- 504B and represent the deeper sections of that hole, where
confining pressure has presumably closed more of the mi-
crocracks. In situ the shear velocity in massive units in-
creases with increasing compressional velocity somewhat
1.0
more rapidly than in cores from hole 504B measured at
2.0 3.0 4.0 5.0 6.0 7.0 equivalent differential pressures. At low compressional ve-
locities, Vs is lower than in the cores, and at high compres-
Compressional velocity (km/s)
sional velocities, Vs is higher than in the cores. This may be
Fig. 15. (continued) a consequence of incomplete closure of thin, stress relief
microcracks generated in the cores during removal from the
in situ stress state.
In pillows the results are quite different. Velocities are
The similarity between pillow size and sonic wavelength considerably higher in DSDP hole 418A than in the other
results in efficient scatteringfrom voids between pillows and holes. The lowest velocities were recorded in DSDP hole
from pillow rims. As discussed in more detail by Moos 504B. The filter applied to the data from hole 395A affected
[1988], there is a loss of high frequencies for both P and S the shear wave, and thus data from the lowest-velocity
wave arrivals within pillows, but because of the pillow size pillows in that well are incomplete. Consequently, hole 504B
and the shorter shear wavelength, the S wave is more pillows may be more representative of the range of values for
stronglyaffected. Variations in the properties of pillows may young crust. The 395A data (mostly from deeper in that hole)
be a function either of variations in pillow morphology at overlie the lower Vp/Vs ratio, higher velocity data from hole
emplacement or of variations in the amount and type of 504B. For pillows with similar P wave velocities the velocity
infilling materials, or a combination of both, as discussedin ratio in hole 418A is a bit lower than in hole 504B,,which
more detail below. suggeststhat fewer small aspect ratio cracks are present in
The two breccias are quite different. In hole 418A at 509 older crust.
mbsf, both compressionaland shear waves are transmitted. Surprisingly, the laboratory relationship between P and S
In hole 395A the breccia at 350 m does not transmit shear wave velocities fits the in situ pillow data quite well. This
energy. Again, these differences may be a consequence of suggeststhat pore aspectratio distr.ibutions affecting sonic
emplacement and/or alteration history. For example, the velocities are similar to those affecting ultrasonic velocities.
breccia in hole 418A is completely infilled with alteration In other words, macroscopic fractures and voids, which are
materials [Broglia and Moos, 1988; Holmes, 1988]. In con- likely to have a quite different aspect ratio distribution than
trast, breccias in hole 395A have considerably less alteration microcracks, do not greatly affect sonic velocities calculated
9204 Moos ET AL.: ELASTIC VELOCITIES IN OCEANIC LAYER 2

by semblance.Moos et al. [ 1986]foundthat velocityratiosin greatly improve the material's ability to transmit shear
pillows and flows were different in the shallowest 150 m of energy, whereas in hole 395A the alteration which increases
basement in hole 504B. Effective medium theories would the velocities also significantlyenhancesthe material's abil-
predictthis resultif the voidsbetweenpillowswere ordersof ity to transmit sonic waves. In hole 504B, alteration at
magnitudesmallerthan the sonicwavelength.However, the intermediatetemperaturesallows the propagationof shear
propertiesof the matrix basalt are quite similar for pillows energywithin the pillows, except within the shallowaquifer
and flows [e.g., Broglia and Moos, 1988], and the size of the (unit 3A).
pillows is similarto the sonicwavelength.Thus the effect of These changesin the characterof the pillows with depth
the voids between pillows on sonic velocities cannot be due to changesin alteration and infilling correspond in all
describedusingeffectivemediumtheories.One possibilityis three holes to the depth ranges within which boundaries
that either cooling cracks within the pillows or the pillow between the oceanic layers are placed. For example, the
contacts themselves contribute a componentof microcrack ancestrallayer 2A/2B boundary in hole 418A is placed at the
porosity, which affects the velocities in the same manner as base of the breccia unit 6A, which separatesmaterial filled
do the stressrelief microcracksin the laboratorysamples. with large amounts of clays from less infilled materials at
Data from breccias represent only one interval in hole greater depths. In hole 395A the steady increasein velocity
418A and two intervals in hole 504B; no shear velocities over the depth range 450-520 mbsf associated with infilling
were recorded in the breccias of hole 395A. Although and alteration of the pillows, leading to the presence of
brecciasare the least important structuraltype volumetri- coherent shear arrivals, led to the placement of the layer
cally, they may be quite important structurally, representing 2A/2B boundary within that depth interval [e.g., Hyndman
faults or "sedimentary" talus and providing high- and Salisbury, 1984]. Thus changesin velocitieswith depth
permeability conduits for geothermal fluids. Velocities are givingrise to the zonation of oceaniclayer 2 extrusives(e.g.,
generally quite low, although not as low as in the slowest Figure 1) may be due to changesin the type or amount of
pillows. In breccias from hole 418A, Vp/Vs decreaseswith alteration, as suggestedby Vera [1989] for the boundary
Vp; in breccias from hole 504B it increases. between layers 2 and 3.

Elastic Wave Velocities and Alteration Effects of Emplacement on Elastic Wave Velocities
Alteration infilling is the largest factor contributing to Although alteration infilling is the primary influenceon the
variations in the elastic velocities of pillows. However, the elastic properties of pillows, there is some evidence that
importance of alteration infilling in raising the velocity of depositionalphenomenaaccompanyingthe emplacementof
oceanicpillows dependson the nature of the infilling mate- these basalts is also important. Pezard [this issue] discusses
rial. For example, alteration infilling by weak clays in the the variation of electrical properties as a function of depth
uppermostintervals of DSDP hole 418A raisesthe compres- within individual units in hole 504B. In particular, he finds a
sional wave velocity significantly,but velocities within this systematicchangein the electrical propertieswithin individ-
interval still do not approachthoseat greater depths.Veloc- ual massiveunits of a particular extrusive sequence,related
ities within the lowermost portion of hole 395A are quite to changes in the magma source.
high, because of significantamounts of "harder" infilling Similarly, Moos [1990] delineates systematic changes in
materials. Comparison of the data from these two holes properties of the different chemical units within the cored
indicates that the effects of alteration on the velocities of sectionof DSDP hole 395A, on the basis of natural gamma,
pillows is not a simple function of age, as at equivalent resistivity, and sonic measurements. These systematics
depthsbelow a few hundred meters subbasementwithin hole were previously recognized in this hole by Mathews et al.
418A at 110 Ma and hole 395A at 7.3 Ma, velocitiesare quite [ 1984],but the poor data quality of the logs used in that study
similar. However, in the shallowercrust, systematicinfilling constrainedfurther analysis. The log variations reveal that
with clays may result in a more gradual changewith age in within each chemical unit, gamma intensity decreaseswith
the properties of the basalts. The absence of a shallow zone depth, and resistivity and sonic velocities increase with
in DSDP hole 417D (drilled near site 418) similar to that in depth, both within the pillows and within individual thin flow
hole 418A suggeststhat the degree of infilling in old crust units. Generally, each unit is initiated by a massiveunit at its
may be quite variable [Donnelly et al., 1979]. Thus the base, grading upward into less massive, more altered, more
degreeto which infilling occursin old crust is quite likely a rubbley pillows, with the highestdegreeof alteration and the
function of locally varying conditionssuchas the accessibil- least competent material at the top. This is then capped by
ity of the material to seawater or other fluids. another flow which initiates the next chemical unit. This
The degreeand type of infilling material also influencesthe characteristic sequence is particularly pronounced in the
scatteringproperties of oceanic pillow basalts. The fact that phyric units which make up most of the uppermost350 m of
shearwaves do not propagateefficiently at sonicfrequencies basement, although it is much less evident in the aphyric
within oceanicpillows at shallow depthsis a consequenceof basalts.
scatteringof the sonicwave from voids between the pillows. Evidence for a similar trend can be found in the compres-
This is more prevalent at young ages (holes 504B and 395A) sional velocities in hole 418A. Several such sequencescan
but also occurswithin older crust which has been infilled by be observed in unit 5, of which the intervals 410-435 mbsf
alteration products (as in the shallow sectionsof hole 418A). and 435-465 mbsf are most obvious. In each of these a
As a result, shear velocities cannot be determined for the high-velocity "massive" unit initiates a sequencein which
"slowest" pillows. At greater depths, both shear and com- velocities decrease upward, until the onset of another high-
pressional waves can be seen, even in young crust. For velocity unit which starts the next sequence.
example, in hole 418A, low-temperature alteration does not Pezard [this issue] suggestson the basis of the electrical
 MoosETAL.:ELASTIC
VELOCITIES
IN OCEANIC
LAYER
2 9205

logsa similar
sequence points valueto helpconstrain
forhole504Bandinparticular suchmodels.
Forexample,
thesonic
to massiveunit 2D cappinga shallowaquiferwithinlayer resultsdemonstratethat velocitiescharacteristicof layer 2A
2A, and massiveunit 27 whichprovidesa permeability arepresent in hole504Bonlyin theuppermost 100m andin
barrierandseparates different
alterationregimeswithinhole hole 395A only in the uppermost150 m. In hole 418A,
504B,asevidence of thissequence. As he suggests,sucha semblancevelocitiesof the Schlumbergerdata, for which
sequence is idealfor creatingnarrowconfined aquifersto lowervelocitieswererecordedabovethe subunit6a breccia,
controlhydrothermal withinthe extrusiveba- andthelargeamounts
circulation of alterationinfillingin thesepillows
salts.Sharpboundaries between different
alterationregimes suggestthatthe uppermost 185m of thisholemayat one
in theseholes(for example,in hole418Aat brecciaunit6A) timehavehadhigherporosities andhencevelocitieschar-
maybe similarlydueto structural controlof fluidflow. acteristic
of layer
2A [Brogliaand Moos,1988].However,in
all three holes, massiveunits with high P and S wave
velocities
arepresentevenwithintheshallowest
intervals.
Relationship
BetweenSonicand SeismicVelocities
Seismicand sonicvelocitiesdo not generallyagreewithin SUMMARY AND CONCLUSIONS
the shallowestextrusivebasalts,evenwithin the limits of the
measurements.Thisispredominantlydueto thelossof sonic Analysisof full waveformmultichannel
sonicdata in
velocityinformation
in the leastcompetent (slowest)pil- DSDP holes395A,418A, and504Bprovidescompressional
lows. The benefitof the semblanceanalysistechniqueis that and shear sonic velocitieswithin the extrusive sectionof
spurious velocitymeasurements canbe eliminated usinga bothyoungandoldcrustgenerated on the slow-spreading
reasonablecriterion(the semblancemagnitude),and thus mid-Atlantic Ridgeand for both extrusiveand intrusive
these results are less contaminated by error than those basalts withinyoungcrustproduced at thefaster-spreading
reported previously. Swiftet al. [1988]suggested thatthe Costa Rica Rift. Velocities within these oceanicbasaltsare
sonicresultsare not representative of the crustover the systematically higherfor massive unitsthanfor pillows.
largelateralareassampled seismically.Furthermore, the Breccias, whicharevolumetrically unimportant in determin-
averaging of the seismicvelocities overthe seismic wave- ingcrustalseismic velocities, havevelocities slightlylower
lengthmayalsoexplaindifferences in velocity[e.g.,Mutter thanthosewithinadjacentpillows.Thereis no measurable
and Newmark,1986],as laterallyfiniteunitswill be effec- difference
betweentheproperties
of morphologically
similar
tivelyoversampled by the soniccompared to the seismic basaltsthat is relatedto spreadingrate. Differencesin the
data. Thus sonicor core "stratigraphy"shouldbe treated grossvelocitystructure withinthe extrusivesectionof
with cautionin generatinggroundtruthvelocity-depthpro- youngcrustaredueto differences in therelativepercentages
files. However, there is almostcertainlya relationship of pillows and flows.
betweenlateralandverticalheterogeneity,
and measuresof Massivebasaltsare generallycharacterizedby compres-
the variabilityof velocityverticallymay be of value to sionalvelocitiesabove5 km/s and shearvelocitiesabove2.8
the sonicresultslaterallyawayfromthebore- km/s(Vp/Vsbelow1.9),equivalent
extrapolate to velocitiesmeasured
hole. oncoresamples at ultrasonic
frequencies.
Velocities in situ
A numberof authorshave suggested that horizontalve- increasewith depth,becauseof the effectof confining
pressure
locities are different than vertical velocitiesbecauseof the whichcloses smallaspectratiocracks,leadingto
presence of macroscopic fractures [e.g., Stephen,1981]. highersonicthanultrasonic
velocities
(measured at atmo-
Thiswill be a problemif sonicvelocities measured in the spheric
confining
pressure)
withinthesheeted dikesof hole
verticaldirectionare comparedto seismicrefractionveloc- 504B.Fracturingwithinmassivebasalts(asis observedin
ities or the resultsof velocity analysesfrom "diving" rays the transitionzonepenetrated
by hole504B)reducesboth
for compressional
[e.g.,Swiftet al., 1988].However,thisis nota problem andshearvelocities
withoutaffecting
their
the interval velocitiesmeasuredbetweenreceiversin a VSP ratio.

asin hole504B.Finally,theeffectof large- Velocitieswithinpillowsare quitevariable,rangingfrom

configuration,
scaleheterogeneitiessuchas fracturesandvoidsbetween Vp < 3.0km/sto velocities approaching thoseof massive
pillowsis quitedifferent
for seismic
andsonicwaves.The basalts. Shear velocities could not be determinedwithin
seismicwaves "see" a slower effective medium, whereas significant
portionsofthesections studied.However,where
the sonic waves "see" a faster more heterogeneous one. both velocitieswere measured,Vp/Vs rangesfrom 1.7 to
Seismicand sonic velocitiesgenerallyagree within the more than 2.1.
intrusivesof hole 504B. Within the extrusivesthe seismicis Differencesbetweenthe propertiesof pillowsof different
sometimes lower, andsometimes higher,thanthe averaged depthsandagesare primarilydue to differences in the
sonicvelocities.Consequently, anelasticdispersion is prob- amount andtypeof alterationinfilling.At shallowdepthsin
ably not importantin explaining the differences between young crust,openvoids between pillows sharplylowersonic
sonicand seismicvelocities.If it were, (1) the sonicveloc- velocitiesand severelyattenuatesonicenergy,makingac-
itiesshouldalwaysbe greaterthanseismic velocities;(2) the curatedeterminationof velocitiesextremelydifficult.This is
qualityfactorQ wouldhaveto bequitevariable, to explain apparent in hole395Aaboveabout500mbsfandat various
the variationof the velocitydifference;and (3) one would depths, including
a shallow"acquifer"
inhole504B.Pillows
expectdifferences in seismic andsonicvelocities to persist withinthedeeper extrusive
section
have significantly
higher
even into the intrusive section of hole 504B. velocitiesand propagateelasticwavesmore efficiently,
Althoughit is temptingto use the sonicvelocitiesto becauseof infillingwith "hard" alterationproducts(e.g.,
produce a layered modelofthebasaltic crust,similartothat calciteand feldspars),whichoccursat a fairly youngage.
interpreted fromseismic data,theabovediscussion suggestsTheproperties ofthisdeeperextrusive
section donotchange
that this shouldbe done with caution. The logs can be of muchwith ageor with spreading
rate for crustolderthan
9206 Moos ET AL.: ELASTIC VELOCITIES IN OCEANIC LAYER 2

about 5 Ma. The increase in velocity with crustal age in the vol. 69, U.S. Government Printing Office, Washington, D.C.,
1983.
shallower extrusive pillows is due to alteration infilling with
claysand weaker materialsand can raisetheir compressional Cheng, C. H., and M. N. T6ksoz, Elastic wave propagationin a
fluid-filledborehole and syntheticacousticlogs, Geophysics,46,
velocities by more than 1 km/s over the time interval from 1042-1053, 1981.
7.3 to 110 Ma, which is sufficientto explain the disappear- Christensen,N. I., and M. H. Salisbury, Seismic velocities, densi-
ance of seismic layer 2A. However, differencesbetween the ties, and porosities of layer 2B and layer 2C basalts from hole
properties of shallow pillows filled with clays and those of 504B, Initial Rep. Deep Sea Drill. Proj., 83, 367-370, 1985.
deeperpillows filled with "harder" alterationproductsper- Donnelly, T., et al., Initial Reports of the Deep Sea Drilling Project,
sist. vol. 51-53, U.S. Government Printing Office, Washington,D.C.,
1979.
In the extrusive pillows sonic velocities are generally Harding, A. J., J. A. Orcutt, M. E. Kappus, E. E. Vera, J. C.
lower than laboratory-determinedvelocities, becauseof the Mutter, P. Buhl, R. S. Detrick, and T. M. Brocher, Structure of
presence in situ of large-scale voids and fractures, and/or young oceanic crust at 13øN on the East Pacific Rise from
"softer" materials, not sampledby coring. However, some expandingspread profiles, J. Geophys. Res., 94, 12,163-12,196,
1989.
laboratory-determinedvelocitiesare equal to or even lower Holmes, M. A., Evidence for continuous and discontinuous alter-
than those in situ, suggestingthat in some cases lower- ation in DSDP hole 418A basalts and its significanceto natural
velocity materials have been successfullyrecovered. gamma-ray log readings, in Proceedings of ODP, Scientific Re-
Averaged sonic velocities within the entire extrusive sec- sults, vol. 102, edited by M. H. Salisbury et al., pp. 136-154,
tion are generally equal to or slightly higher than seismic Ocean Drilling Program, Houston, Tex., 1988.
Honnorez, J., C. Laverne, H.-W. Hubberton, R. Emmermann, and
velocities. This difference is also partly a function of sam-
K. Muehlenbachs,Alteration processesin layer 2 basalts from
pling bias, as (1) sonic velocities within the slowestpillows Deep Sea Drilling Project hole 504B, Costa Rica Rift, Initial Rep.
are difficult to determine, raisingthe "average" sonicveloc- Deep Sea Drill. Proj., 69, 509-546, 1983.
ity, and (2) laterally finite flows penetrated by the drill bit Houtz, R., and J. Ewing, Upper crustal structure as a function of
may not affect velocities in the larger rock volume sampled plate age, J. Geophys. Res., 81, 2490-2498, 1976.
by the seismicwaves. However, it is alsolikely that the scale Hussong, D. M., P. B. Fryer, J. D. Tuthill, and L. K. Wipperman,
The geologicaland geophysicalsettingnear DSDP site 395, North
of the sonicwave is suchthat it scattersfrom heterogeneities Atlantic Ocean, Initial Rep. Deep Sea Drill. Proj., 45, 23-37,
that behave like flaws in an effective medium at the scale of 1978.
the seismicmeasurement.Within the deeper sectionof hole Hyndman, R. D., and M. H. Salisbury, The physical nature of
418A, and within the intrusives in hole 504B, sonic and young upper oceanic crust on the mid-Atlantic Ridge, Deep Sea
seismic velocities are similar. Drilling Project hole 395A, Initial Rep. Deep Sea Drill. Proj., 78B,
839-848, 1984.
Jacobson, R. S., R. Adair, and J. Orcutt, Preliminary seismic
refraction results using a borehole seismometer in Deep Sea
Acknowledgments. The existenceand quality of the ODP ship- Drilling Project hole 395A, Initial Rep. Deep Sea Drill. Proj., 78B,
board geophysical logging program is due to R. N. Anderson of 783-794, 1984.
Lamont-Doherty Geological Observatory. We also gratefully ac- Juteau, T., F. Bingol, Y. Noack, H. Whitechurch, M. Hoffert, D.
knowledgethe help and cooperationof the OceanDrilling Program Wirrman, and C. Courtois, Preliminary results: Mineralogy and
staff and shipboard personnel, without which none of these data geochemistryof alteration products in leg 45 basement samples,
couldhave been collected.K. Becker was onboardfor all three legs Initial Rep. Deep Sea Drill. Proj., 45, 613-646, 1978.
duringwhich the data was collectedand was chief scientiston leg Kimball, C. V., and T. L. Marzetta, Semblance processing of
111. We also thank the other chief scientists, J. Scott and M. borehole acoustic array data, Geophysics,49, 274-281, 1984.
Salisburyon leg 102, W. Bryan and T. Juteau on leg 109, and H. Lawrence, J. R., J. J. Drever, and M. Kastner, Low temperature
Sakai on leg 111, for their help and support.D. Shahand K. Peters alteration of basaltspredominatesat site 395, Initial Rep. Deep
helped implement the data collection and analysis software. The Sea Drill. Proj., 45, 609-612, 1978.
authors of this paper greatly benefited from discussionswith D. Mathews, M., M. H. Salisbury,and R. Hyndman, Basementlogging
Goldbergand C. Broglia. Constructivereviews by P. Lysne and F. on the mid-Atlantic Ridge, Deep Sea Drilling Project hole 395A,
Paillet greatly improved the manuscript. The first author was Initial Rep. Deep Sea Drill. Proj., 78B, 717-730, 1984.
supported during this work by NSF contracts JOI-66-84 and OCE- Melson, W. G., et al., Initial Reports of the Deep Sea Drilling
8704609. Project, vol. 45, U.S. Government Printing Office, Washington,
D.C., 1978.
Moos, D., Elastic properties of 110-Ma oceanic crust from sonic full
REFERENCES waveformsin DSDP hole 418A, in Proceedingsof ODP, Scientific
Results, vol. 102, edited by M. H. Salisbury et al., pp. 49-62,
Alt, J. C., C. Laverne, and K. Muehlenbachs, Alteration of the Ocean Drilling Program, College Station, Tex., 1988.
upper oceanic crust: Mineralogy and processesin Deep Sea Moos, D., Petrophysical results from logging in DSDP hole 395A,
Drilling Project hole 504B, leg 83, Initial Rep. Deep Sea Drill. ODP leg 109, in Proceedings of ODP, Scientific Results, vol.
Proj., 83, 217-247, 1985. 106/109,edited by R. Detrick et al., pp. 237-253, Ocean Drilling
Anderson, R. N., et al., Initial Reports of the Deep Sea Drilling Program, College Station, Tex., 1990.
Project, 83, U.S. GovernmentPrintingOffice, Washington,D.C., Moos, D., D. S. Goldberg, M. A. Hobart, and R. N. Anderson,
1985. Elastic-wave velocities in layer 2A from full waveform sonic logs
Barton, C., D. Moos, and J.-P. Blangy, Analysis of full waveform at hole 504B, Initial Rep. Deep Sea Drill. Proj., 92,563-570, 1986.
acousticloggingdata at ODP site 642--Outer VOring Plateau, in Mutter, J. C., and R. L. Newmark, Seismic responseof detailed
Proceedingsof ODP, Scientific Results, vol. 104, edited by O. structureat DSDP hole 504B: Synthetic seismogramsin the tau-p
Eldholm, et al., pp. 953-964, Ocean Drilling Program, College domain, Initial Rep. Deep Sea Drill. Proj., 92,571-584, 1986.
Station, Tex., 1989. Newmark, R. L., R. N. Anderson, D. Moos, and M.D. Zoback,
Becker, K., et al., Drilling deep into the oceanic crust at hole 504B, Sonic and ultrasonicloggingof hole 504B and its implicationsfor
Costa Rica Rift, Rev. Geophys,27, 79-102, 1989. the structure,porosity, and stressregime of the upper 1 km of the
Broglia C., and D. Moos, In situ structureand propertiesof 110Ma oceanic crust, Initial Rep. Deep Sea Drill. Proj., 83, 479-510,
crustfrom geophysicallogsin DSDP hole 418A, in Proceedingsof 1985.
ODP, ScientificResults, vol. 102, editedby M. H. Salisburyet al., O'Connell, R. J., and B. Budianski, Seismic velocities in dry and
pp. 29-47, Ocean Drilling Program, College Station, Tex., 1988. saturatedcracked solids, J. Geophys. Res., 79, 5412-5426, 1974.
Cann, J. R., et al., Initial Reports of the Deep Sea Drilling Project, Paillet, F. L., and J. E. White, Acoustic modesof propagationin the
 MoosETAL..'ELASTIC
VELOCITIES
IN OCEANIC
LAYER
2 9207

boreholeand their relationshipto rock properties,Geophysics, oceaniccrustfroman ObliqueSeismicExperimentat hole418A,

47, 1215-1228, 1982. westernNorth Atlantic, in Proceedingsof ODP, ScientificRe-
Pezard,P. A., Electrical
propertiesof mid-ocean ridgebasaltand sults,vol.102,editedbyM. H. Salisbury etal., pp.97-126,Ocean
implicationsfor thestructureof theoceanic crustat DeepSea DrillingProgram,CollegeStation,Tex., 1988.
DrillingProjectsite504,J. Geophys. Res.,thisissue. Vera,E. E., Seismic structure
of 0- to 4.5-m.y.-old oceaniccrust
Purdy,G. M., Newobservations oftheshallowseismic
structure
of between9øand 13øNon the East PacificRise,Ph.D. Thesis,127
young oceanic crust,J. Geophys. Res.,92,9351-9362,
1987. pp., ColumbiaUniv., New York, 1989.
Salisbury,
M. H., N. I. Christensen,
K. Becker,andD. Moos,The
velocitystructureof layer2 at DSDPsite504fromlogging and M. Lovell,Department of Geology,Universityof Nottinghmn,
laboratoryexperiments, InitialRep.DeepSeaDrill. Proj.,83, NottinghamNG7 2RD, England.
529-539, 1985. D. Moos,Department
of Geophysics,
Stanford
University,
Stan-
Shipboard
Scientific
Party,Site504,inProceedings
ofODP,Initial ford, CA 94305.
Report,
PartA, vol.111,editedbyK. Becker
et al.,pp.35-252, P. Pezard,Borehole
Research
Group,Lamont-Doherty
Geologi-
OceanDrillingProgram,CollegeStation,Tex., 1988. cal Observatory,Palisades,NY 10964.
Spudich,
P., andJ. A. Orcutt,A newlookat theseismic
velocity
structureof the oceaniccrust,Rev. Geophys.,18,627-645, 1980.
Stephen,
R. A., Seismic
anisotropy
observed
inupper
oceanic
crust, (Received May 9, 1989;
revised January 22, 1990;
Geophys.
Res. Lett., 8, 865-868,1981.
Swift,S. A., R. A. Stephen,
andH. Hoskins,Structure
of upper acceptedJanuary22, 1990.)

View publication stats