Download the full version of the ebook now at ebookultra.

com

Climate and Oceans A Derivative of

Encyclopedia of Ocean Sciences Second Edition
John H. Steele

https://ebookultra.com/download/climate-and-
oceans-a-derivative-of-encyclopedia-of-ocean-
sciences-second-edition-john-h-steele/

Explore and download more ebook at https://ebookultra.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Encyclopedia of Life Sciences 1st Edition John Wiley &

Sons Ltd

https://ebookultra.com/download/encyclopedia-of-life-sciences-1st-
edition-john-wiley-sons-ltd/

ebookultra.com

Atmosphere Ocean and Climate Dynamics An Introductory Text

1st Edition John Marshall

https://ebookultra.com/download/atmosphere-ocean-and-climate-dynamics-
an-introductory-text-1st-edition-john-marshall/

ebookultra.com

Oceans A Scientific History of Oceans and Marine Life 1st

Edition Michael Allaby

https://ebookultra.com/download/oceans-a-scientific-history-of-oceans-
and-marine-life-1st-edition-michael-allaby/

ebookultra.com

Microbial Ecology of the Oceans Second Edition David L.

Kirchman

https://ebookultra.com/download/microbial-ecology-of-the-oceans-
second-edition-david-l-kirchman/

ebookultra.com
Encyclopedia of Hydrological Sciences 1st Edition Malcolm
G. Anderson

https://ebookultra.com/download/encyclopedia-of-hydrological-
sciences-1st-edition-malcolm-g-anderson/

ebookultra.com

Encyclopedia of Statistical Sciences 2nd ed Edition Samuel

Kotz

https://ebookultra.com/download/encyclopedia-of-statistical-
sciences-2nd-ed-edition-samuel-kotz/

ebookultra.com

Fishes of the Maldives Indian Ocean 4th Edition Rudie H.

Kuiter

https://ebookultra.com/download/fishes-of-the-maldives-indian-
ocean-4th-edition-rudie-h-kuiter/

ebookultra.com

Encyclopedia of global warming climate change 2nd edition

Edition Philander

https://ebookultra.com/download/encyclopedia-of-global-warming-
climate-change-2nd-edition-edition-philander/

ebookultra.com

Elgar Encyclopedia of Climate Policy 1st Edition Daniel J.

Fiorino

https://ebookultra.com/download/elgar-encyclopedia-of-climate-
policy-1st-edition-daniel-j-fiorino/

ebookultra.com
Climate and Oceans A Derivative of Encyclopedia of
Ocean Sciences Second Edition John H. Steele Digital
Instant Download
Author(s): John H. Steele, Steve A. Thorpe, Karl K. Turekian
ISBN(s): 9780123785558, 0123785553
Edition: 1
File Details: PDF, 28.44 MB
Year: 2010
Language: english
 Climate & Oceans

Editor-in-Chief
John H. Steele
Marine Policy Center, Woods Hole Oceanographic Institution, Woods Hole,
Massachusetts, USA

Editors
Steve A. Thorpe
National Oceanography Centre, University of Southampton,
Southampton, UK
School of Ocean Sciences, Bangor University, Menai Bridge, Anglesey, UK

Karl K. Turekian
Yale University, Department of Geology and Geophysics, New Haven,
Connecticut, USA

Subject Area Volumes from the Second Edition

Climate & Oceans edited by Karl K. Turekian

Elements of Physical Oceanography edited by Steve A. Thorpe
Marine Biology edited by John H. Steele
Marine Chemistry & Geochemistry edited by Karl K. Turekian
Marine Ecological Processes edited by John H. Steele
Marine Geology & Geophysics edited by Karl K. Turekian
Marine Policy & Economics guest edited by Porter Hoagland, Marine Policy Center,
Woods Hole Oceanographic Institution, Woods Hole, Massachusetts
Measurement Techniques, Sensors & Platforms edited by Steve A. Thorpe
Ocean Currents edited by Steve A. Thorpe
The Coastal Ocean edited by Karl K. Turekian
The Upper Ocean edited by Steve A. Thorpe
CLIMATE & OCEANS
A DERIVATIVE OF ENCYCLOPEDIA OF
OCEAN SCIENCES, 2ND EDITION

Editor

KARL K. TUREKIAN

Amsterdam • Boston • Heidelberg • London • New York • Oxford

Paris • San Diego • San Francisco • Singapore • Sydney • Tokyo
Academic Press is an imprint of Elsevier
Academic Press is an imprint of Elsevier
32 Jamestown Road, London NW1 7BY, UK
30 Corporate Drive, Suite 400, Burlington, MA 01803, USA
525 B Street, Suite 1900, San Diego, CA 92101-4495, USA

Copyright ^ 2010 Elsevier Ltd. All rights reserved

Material in the work originally appeared in Encyclopedia of Ocean Sciences (Elsevier Ltd., 2001) and Encyclopedia of Ocean Sciences,
2nd Edition (Elsevier Ltd., 2009), edited by John H. Steele, Steve A. Thorpe and Karl K. Turekian.

The following articles are US government works in the public domain and are not subject to copyright:
Satellite Oceanography, History and Introductory Concepts
Satellite Passive-Microwave Measurement of Sea Ice
Science of Ocean Climate Models

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic,
mechanical, photocopying, recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44)(0)
1865 843830; fax (+44)(0) 1865 853333; email: [email protected]. Alternatively you can submit your request online by visiting
the Elsevier website at (http://elsevier.com/locate/permissions), and selecting Obtaining permissions to use Elsevier material

Notice
No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein,
Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Control Number: 2009907115

ISBN: 978-0-08-096482-9

For information on all Elsevier publications

visit our website at www.elsevierdirect.com

PRINTED AND BOUND IN ITALY

10 11 12 13 14 10 9 8 7 6 5 4 3 2 1
 CONTENTS

Climate and Oceans: Introduction ix

OCEAN CIRCULATION & CLIMATE

Ocean Circulation N C Wells 3
Science of Ocean Climate Models S M Griffies 14
General Circulation Models G R Ierley 22
Ocean Circulation: Meridional Overturning Circulation J R Toggweiler 27
North Atlantic Oscillation (NAO) J W Hurrell 33
Abrupt Climate Change S Rahmstorf 41

TROPICAL CLIMATE CONTROLS

Pacific Ocean Equatorial Currents R Lukas 49
El Niño Southern Oscillation (ENSO) K E Trenberth 57
El Niño Southern Oscillation (ENSO) Models S G Philander 70
Monsoons, History of N Niitsuma, P D Naidu 76

HEAT & WATER BALANCE

Freshwater Transport and Climate S Wijffels 87
Heat and Momentum Fluxes at the Sea Surface P K Taylor 95
Heat Transport and Climate H L Bryden 103
Sea Surface Exchanges of Momentum, Heat, and Fresh Water Determined by Satellite L Yu 110
Evaporation and Humidity K Katsaros 120
Upper Ocean Heat and Freshwater Budgets P J Minnett 127

CRYOSPHERE
Antarctic Circumpolar Current S R Rintoul 141
Arctic Ocean Circulation B Rudels 154
Sub Ice-shelf Circulation and Processes K W Nicholls 169
Ice–Ocean Interaction J H Morison, M G McPhee 179
Coupled Sea Ice–Ocean Models A Beckmann, G Birnbaum 190
Ice Shelf Stability C S M Doake 201
Noble Gases and the Cryosphere M Hood 210
Sea Ice: Overview W F Weeks 214
Sea Ice Dynamics M Leppäranta 223

v
vi CONTENTS

Sea Ice P Wadhams 234

Sub-Sea Permafrost T E Osterkamp 252

SATELLITE MEASUREMENTS
Satellite Oceanography, History, and Introductory Concepts W S Wilson, E J Lindstrom,
J R Apel 265
Satellite Passive-Microwave Measurements of Sea Ice C L Parkinson 280
Satellite Remote Sensing of Sea Surface Temperatures P J Minnett 291

PALEOCLIMATE MARINE RECORD

Millennial-Scale Climate Variability J T Andrews 305
Oxygen Isotopes in the Ocean K K Turekian 314
Cenozoic Climate – Oxygen Isotope Evidence J D Wright 316
Determination of Past Sea Surface Temperatures M Kucera 328
Plio-Pleistocene Glacial Cycles and Milankovitch Variability K H Nisancioglu 344
Paleoceanography E Thomas 354
Paleoceanography, Climate Models in W W Hay 362
Paleoceanography: Orbitally Tuned Timescales T D Herbert 370
Paleoceanography: the Greenhouse World M Huber, E Thomas 378
Holocene Climate Variability M Maslin, C Stickley, V Ettwein 389
Plankton and Climate A J Richardson 397
Coccolithophores T Tyrrell, J R Young 407
Benthic Organisms Overview P F Kingston 416
Benthic Foraminifera A J Gooday 425
Past Climate from Corals A G Grottoli 437

CARBON SYSTEM
Carbon Cycle C A Carlson, N R Bates, D A Hansell, D K Steinberg 449
Carbon Dioxide (CO2) Cycle T Takahashi 459
Ocean Carbon System, Modeling of S C Doney, D M Glover 467
Radiocarbon R M Key 477
Stable Carbon Isotope Variations in the Ocean K K Turekian 493
Cenozoic Oceans – Carbon Cycle Models L Franc¸ois, Y Goddéris 494

EFFECTS & REMEDIATION

Sea Level Change J A Church, J M Gregory 507
Sea Level Variations over Geological Time M A Kominz 513
Glacial Crustal Rebound, Sea Levels, and Shorelines K Lambeck 522
 CONTENTS vii

Economics of Sea Level Rise R S J Tol 532

Land–Sea Global Transfers F T Mackenzie, L M Ver 536
Effects of Climate Change on Marine Mammals I Boyd, N Hanson 546
Fisheries and Climate K M Brander 550
Seabird Responses to Climate Change David G Ainley, G J Divoky 558
Carbon Sequestration via Direct Injection into the Ocean E E Adams, K Caldeira 566
Iron Fertilization K H Coale 573
Methane Hydrates and Climatic Effects B U Haq 585

APPENDICES
Appendix 1: SI Units and Some Equivalences 593
Appendix 2: Useful Values 596
Appendix 3: Periodic Table of the Elements 597
Appendix 4: The Geologic Time Scale 598
Appendix 5: Properties of Sea Water 599
Appendix 6: The Beaufort Wind Scale and Seastate 604

INDEX 607
 CLIMATE AND OCEANS: INTRODUCTION

The role of the oceans in climate locally has been known for a long time. The warming and cooling of land
masses close to the oceans is controlled in part by the ocean temperature since the high heat capacity of water
dampens the seasonal changes in heating by sunlight received by adjacent continental areas as a function of
the seasons. In lower latitudes the oceans dominate the air temperature of islands giving rise to the de-
scription of the climate on these islands as ‘‘maritime climates.’’
These effects globally are more extensive than these two elementary examples. Ocean currents and at-
mospheric circulation over the oceans cause a variety of interactions such as the ENSO (El Niño Southern
Oscillation) process and Gulf Stream transport of heat northward in the Atlantic Ocean–atmosphere inter-
actions are thus an important part of the local as well as the global climate regime as a function of seasons
and location.
The role of the cryosphere in influencing climate and in turn being influenced by climate is an important
part of the ocean-climate relationship, especially as we seek to understand the role of potentially increasing
global warming as a result of the enhancement, by human actions, of radiatively important gases, such as
carbon dioxide, and aerosols.
The oceans play an important role in deciphering the history of climate change over the past several
million years, during the ice ages, as well as over the past hundreds of million years. The source of our
information comes from deep sea cores. There the changing ocean proxies for climate changes are preserved
in the sedimentary components. The study of the distribution of the remains of marine tests provide not only
evidence of ecological changes with time including temperature, but marine tests also provide the material for
detailed studies of oxygen isotope variations, sensitive indicators of temperature and continental ice storage
typical of the ice ages. The chemistry of these tests and other recorders of past ocean environments, such as
coral reefs, provide many other proxies including element ratios and isotopic indicators of environment.
Properly dated cores allow the calibration of patterns of variation as a function of time and associated
climate indicators. The role of orbital forcing in controlling climate has been indicated by detailed studies of
hundreds of cores raised by standard piston coring techniques and more recently by ocean drilling.
The coupling of the knowledge of contemporary ocean-climate relationships to the ancient record allows
for a reconstruction of ancient locations of critical currents. In addition the influence of land mass and
elevation on controlling ancient ocean-climate relationships has been inferred from these records.

Karl K. Turekian
Editor

ix
OCEAN CIRCULATION & CLIMATE
 OCEAN CIRCULATION
N. C. Wells, Southampton Oceanography Centre, dynamical oceanography (see General Circulation
Southampton, UK Models). This integrated approach allows hypoth-
Copyright & 2001 Elsevier Ltd. eses to be made that can be tested by comparison
with observations. Furthermore, mathematical
models of the ocean circulation, based on the dy-
namical principles, can be constructed and tested
against observations (see El Niño Southern Oscil-
Introduction lation (ENSO) Models).
This article considers how ocean circulation is
This article discusses the following aspects of ocean measured, how the major processes at work are de-
circulation: what is meant by the term ocean circu- termined and the consequences of the ocean circu-
lation; how the ocean circulation is determined by lation on the climate system.
measurements and dynamical processes; the con-
sequences of this circulation on the Earth’s climate.

How is the Ocean Circulation

What is Ocean Circulation? Determined?
The ocean circulation in its simplest form is the The determination of ocean currents involves meas-
movement of sea water through the ocean, which urement of the displacement of an element of fluid
principally transfers temperature and salinity, from over a measured time interval. The position of the
one region to another. Temperature differences be- measurement is defined mathematically in a Cartesian
tween regions cause heat transfers. Similarly, differ- coordinate system (Figure 1) where x is positive
ences in salinity produce transfers of salt. On the eastward direction (lines of constant latitude), y is
time scale of the ocean circulation the inputs and positive northward direction (geographic North
exports of salt into and out of the ocean make a Pole), and z is positive upwards; z ¼ 0 corresponds to
negligible contribution to overall salinity and so mean sea level. Without ocean currents and tides the
variations in salinity occur by the addition and re- sea level would be an equipotential surface, i.e., one of
moval of fresh water into and out of the ocean. constant potential energy. The z coordinate is per-
Two major processes control the ocean circulation: pendicular to the equipotential surface. The origin is
the action of the wind and the action of small density the intersection of the Greenwich meridian (Universal
differences, produced by differences in temperature meridian) and the equator with mean sea level.
and salinity, within the ocean. The former process is The coordinates of a parcel of water can be de-
the wind driven circulation whereas the latter is the termined by the Global Positioning System (GPS).
thermohaline circulation. Although it is useful to This satellite-based system provides a horizontal
separate these two processes to better understand the position with an accuracy of better than 100 m,
ocean circulation, they are not independent from which is sufficient for large-scale flows in the ocean.
each other. Large-scale flows are at least 10 km in spatial scale
Ocean circulation is in reality a very complex sys- and have timescales of at least a day. A pressure
tem, as the flows are not steady in time or space. They device attached to the current meter normally de-
are turbulent flows that show variability on scales termines the vertical position.
from the largest scale of the ocean basins (see El Niño There are two mathematical methods for defining
Southern Oscillation (ENSO), North Atlantic Oscil- the displacement of the fluid. One is to measure the
lation (NAO)) to the smallest scales where the energy velocity of the fluid at a fixed point in the ocean, and
is finally dissipated as heat. This turbulent structure of the other is to follow the element of the fluid and to
the ocean means there are fundamental limitations on measure its velocity as it moves through the ocean.
the predictability of its behavior. The first method is known as a Eulerian description
Because of this inherent complexity ocean- and the second is a Lagrangian description of flow. In
ographers have approached ocean circulation by principle, the two methods are independent of each
using a combination of observational methods other. This means that a Eulerian measurement can
including ships, buoys Moorings and satellites, not provide Lagrangian currents, and Lagrangian
combined with the mathematical methods of measurements can not provide Eulerian currents.

3
4 OCEAN CIRCULATION

Eulerian Reflector

v (x, y, z, t ) Tripod tower

Navigation light
Measurement of velocity of the
fluid at fixed point (x, y, z ) Battery case Wind recorder

Instrument buoy
z v Rigid tripod

y Current
meters

0 x

For both diagrams _

Acoustic
x = eastward direction release for
y = northward direction Anchor weight recovery
(A) z = vertical upward

Lagrangian

v (a,t )

Measurement of the velocity of

Figure 2 A typical current meter mooring.
a fluid element. a is the position
vector from the origin to the fluid
element.

P′
a flow is difficult to measure directly, and will be dis-
z P cussed later in this article.
a′ First, the Eulerian method is considered. The
measurement of the flow at a fixed point in the ocean
a is only straightforward when a fixed position can be
y maintained, for instance with a current meter at-
tached to the bottom of the ocean or to a pier on the
coast. Most measurements have to be made well
away from land. This is achieved by attaching the
current meter to a mooring which is attached to
0 x weights and then deployed (Figure 2). The position of
Element of fluid moves a small a mooring can be determined by GPS. The current
distance a in a time t, from P to P′. meter may be a rotary device or an acoustic device.
The rotary current meter measures the number of
The velocity is the a revolutions over a fixed period, whereas the acoustic
t
In infinitesimal time a = v (a,t )
one measures the change in frequency of an emitted
(B) t sound pulse caused by the ocean current (i.e., it uses
the Doppler effect). Moorings may be deployed for
Figure 1 Eulerian and Lagrangian specifications of flow. periods of up to 2 years. In the analysis of the record
it is normal to remove the high frequency variability
Having defined the two mathematical methods of less than 1 day caused by tides by filtering the data.
how the currents are measured in practice is now A Lagrangian measurement of current can be de-
considered. Initially, these methods will address only termined by following an element of water with a
the measurement of the horizontal flow. The vertical float. The horizontal displacement of the water over
 OCEAN CIRCULATION 5

Each method gives different information on

the flow field. A mooring will give a time series of the
horizontal current, whereas a float will give the
3_ 6 m aluminum trajectory of the horizontal displacement of the par-
mast with flag, light,
and radar reflector cel. It is worth remarking that most of the infor-
mation on the surface ocean circulation has come
from mariners’ observations of the ships set, a
Styrofoam float method which has been used since the nineteenth
century. However, these measurements have their
limitations since they are neither eulerian nor
lagrangian measurements and additional influences
(e.g., wind effects on the ship) may cause errors.
3_ 6 m iron pipe This information can be analyzed in many differ-
ent ways to discern the major current systems. From
a set of moorings deployed for a few years across,
say, the Gulf Stream, the mean flow (i.e., the average
Chain ballast of all the current measurements) and its variability
can be determined. The mean flow could be calcu-
lated over a particular time period. This time period
Manila line is limited by the period of deployment, which is of
the order of 2 years. This is rather short for a cli-
matological mean, and a much longer period of 10
1.3 mm piano wire years is desirable. A few longer time series of currents
have been determined for the Gulf Stream in the
Florida Straits and for the Antarctic Circumpolar
Current in the Drake Passage (see Antarctic Cir-
cumpolar Current).
8.5 m parachute Recall ocean currents are turbulent and therefore
have variability on a whole range of timescales.
Hence the mean flow gives no information on the
variability of the flow. However, the statistics of the
Manila line flow can be calculated, based on the kinetic energy
(KE). The kinetic energy/unit volume is defined as:

1  
KE ¼ r u2 þ v2
Chain ballast 2
where r is the density of the sea water and u and v
are the eastward and northward components of the
horizontal flow, respectively.
Figure 3 A typical drifter with a parachute drogue of a few If the time mean current is defined as ū and u0 as
meters below the surface. It will follow the current at the depth of
the deviation from ū at any time, the mean kinetic
the parachute.
energy (KEM) and eddy kinetic energy (EKE) can be
defined by:
a small interval of time defines the Lagrangian cur- 1  
rent. Figure 3 shows typical float designs that are KEM ¼ r ū2 þ v̄2
2
used. The position of the float can be determined by
two methods. A float that has a surface satellite
1  
transmitter/receiver can have its position determined EKE ¼ r u02 þ v02
by GPS, whereas a subsurface float would use an 2
acoustic navigation system. Some floats can descend These two numbers give quantitative measures on
to a predetermined depth, maintain that depth for a the mean and variability of the flow respectively. The
few weeks and then return to the surface for a pos- ratio EKE/KEM gives a measure of the relative
ition fix. This technique allows the current to be variability of the flow. If the ratio is very much less
measured down to depths of 1 km below the surface. than 1 then the flow is steady, whereas if the ratio is
6 OCEAN CIRCULATION

approximately equal to 1 then the flow is very change from one season to another or it may show
variable. faster variation due to eddies.
Figure 4 shows the variability of the flow in the To address this variation time series analysis, such
Agulhas Current, which is an intense and highly as Fourier analysis, can be used to determine the KE
variable current off the coast of South Africa. of the flow for different time periods. Fourier an-
Although this ratio gives a measure of the vari- alysis produces a spectrum of the KE, either in fre-
ability of the current, it does not give any idea of the quency for a time series, or in wave number for a
exact time or space scales over which the current is spatial variation in flow. Figure 5 shows the analysis
varying. For example, the current may show a slow of a time series into its component frequencies. If the
current is varying on all timescales the spectrum
would be flat, but if there was only one dominant
_ 0.5 _ 0.5
period, it would peak at that one frequency. This
particular analysis shows that the current is varying
_ 1.0 _ 1.0 at the tidal frequency and the inertial frequency both
60
_ 1.5 _ 1.5 at the high frequency end of the spectrum. The in-
60

_ 2.0
ertial frequency is given by 2O sin f where O is the
_ 2.0
rotation rate of the earth and f is the latitude. At the
Depth (km)

_ 2.5 _ 2.5 lower frequency end, which corresponds to the ocean

_ 3.0 _ 3.0 circulation frequency, there is a broad band of high
kinetic energy. This band is due to eddies which
_ 3.5 _ 3.5

_ 4.0 _ 4.0

_ 4.5 _ 4.5
_2
_ 5.0 units cm 2 s _ 5.0 100 m
_ 42 _ 41 _ 40 _ 39 _ 38 _ 37 _ 36 _ 35 _ 34 _ 33
(A) Latitude

500 m

Semidiurnal tide
409
_ 0.5 300
354
6 _ 0.5
81 13
245
_ 1.0 191
81 _ 1.0 1000 m

Inertia
_ 1.5 _ 1.5
27

27

Kinetic energy

2000 m
_ 2.0 _ 2.0
Depth (km)

_ 2.5 _ 2.5
27

27

_ 3.0 _ 3.0

_ 3.5 _ 3.5

_ 4.0 _ 4.0

_ 4.5 _ 4.5
2 _2
_ 5.0 units cm s _ 5.0
100 m
_ 42 _ 41 _ 40 _ 39 _ 38 _ 37 _ 36 _ 35 _ 34 _ 33
(B) Latitude

Figure 4 The mean kinetic energy (KEM) (A) and the eddy
kinetic energy (EKE) (B) in a north–south slice through the
Agulhas Current system at 14.41E. The KEM maximum
year month week
corresponds to the mean position of the Agulhas Return 3 2 1
Current (Eastward flow) between 401 and 411S, and the 10 10 10
Agulhas Current (Westward flow) between 371S and 381S. The Period (h)
EKE distribution is much broader than KEM, which shows the
large horizontal extent of the flow variability. The ratio of EKE/ Figure 5 Frequency spectrum of kinetic energy from four
KEM is typically about a third, which indicates a very variable depths at site D (391N, 701W), north of the Gulf Stream. Note the
current system. (Reproduced from Wells NC, Ivchenko V and two high-frequency peaks, coinciding with the inertial period
Best SE (2000) Instabilities in the Agulhas Retroflection Current (19 h) and the semidiurnal tide (12.4 h). (Reproduced from
system: A comparative model study. Journal of Geophysical Rhines PB (1971) A note on long-period motions at site D. Deep
Research 105: 3233–3246.) Sea Research 18: 21–26.)
 OCEAN CIRCULATION 7

cause fluctuations of currents on timescales of weeks These tracers can be measured with high accuracy in
to months. a few laboratories around the world and from their
For these mean climatological currents, our distributions at different times, the three-dimensional
knowledge has been augmented by the application of circulation can be estimated. This method reveals the
the dynamic method. This method is based on the time history of the ocean circulation wherever the
observation that large-scale ocean currents are in tracer is measured. This is very different information
geostrophic balance, over large areas of the ocean. from that provided by the methods previously dis-
Geostrophic balance means that the Coriolis force cussed, but nonetheless it can reveal unique aspects
balances the horizontal pressure gradient force. The of the flow. For example, nuclear fallout deposited in
geostrophic flow is a good approximation to the flow the surface layer of the Nordic seas in the 1960s was
in the interior of an ocean outside the equatorial located in the deep western boundary current 10
region. The horizontal pressure gradient is dependent years later.
on the slope of the ocean surface and the horizontal
variation of the density distribution within the
ocean. In the future, the former may be determined An Ocean General Circulation Model
by satellite measurements of the sea surface height An ocean general circulation model is composed of a
and the geoid1 but at present we do not have an set of mathematical equations which describe the
accurate geoid at sufficiently high resolution to
time-dependent dynamical flows in an ocean basin.
measure the sea surface slope. The latter can be de-
The basin is discretized into a set of boxes of regular
termined from temperature, salinity and pressure horizontal dimensions but variable thickness in the
measurements that have been made over large ocean
vertical dimension. The horizontal flow (northward
areas during the last century. The dynamic method
and eastward components) is predicted by the mo-
allows the determination of the vertical shear of the
mentum equation (Figure 6A) at the corners of each
horizontal geostrophic current, and therefore to de-
box (Figure 7).
termine the absolute geostrophic current, additional
The forcing for the flow may come from the wind
measurements are required. For example if the cur-
stress (the frictional term in the momentum equa-
rent has been measured at a particular depth then the
tion) and from the surface buoyancy fluxes, arising
dynamic method can be referenced to that depth and from heat and freshwater (precipitation þ runoff–
the vertical profile of current can be obtained.
evaporation) exchange with the atmosphere and
The recent World Ocean Circulation Experiment
adjacent landmasses. These buoyancy fluxes change
(WOCE) hydrographic program has provided more
the temperature and salinity in the surface layer of
measurements of the ocean than all previous hydro-
the ocean. The surface water masses are then sub-
graphic programs and will give the most com-
ducted into the interior of the ocean by the vertical
prehensive assessment of climatological horizontal
and horizontal components of the flow, where they
ocean flow to date.
are mixed with other water masses.
Recall that the vertical circulation of the ocean The processes of transport and mixing are de-
cannot be measured directly because it is technically
scribed by the temperature and salinity equations
too difficult. Current indirect methods used to de-
(Figure 6B and C), at the center of each ocean box
termine the vertical circulation rely on the use of
(Figure 7). From these two equations the seawater
mathematical approaches, such as dynamical mod-
density, and thence the pressure can be obtained for
els, or the use of chemical tracers.
each box. The horizontal pressure gradient is then
Observations of temperature and salinity can be
determined for the momentum equation, and the
inserted into a mathematical ocean general circu-
vertical velocity is calculated from the horizontal di-
lation model (see below) which allows the three-di- vergence of the flow. This set of time-dependent
mensional, circulation to be determined, subject to
equations can then be used to describe all the dy-
limitations in the accuracy of the model.
namical components of the flow field, provided that
Chemical tracers have been inadvertently injected
suitable initial and boundary conditions are specified.
into the ocean from nuclear tests in the 1960s and
from industrial processes (e.g., chlorofluorocarbons).
Naturally occurring tracers such as 14C also exist. Wind-driven and Thermohaline
Circulation
1
The geoid is an equipotential surface, which would be repre-
The wind-driven circulation is considered first. The
sented by the sea level of a stationary ocean. Ocean currents cause surface layer of the ocean is directly driven by
deviations in sea level from the geoid. the surface wind stress and is also subject to the
8 OCEAN CIRCULATION

Momentum equation

Rate of change = Transport of momentum + Coriolis + Horizontal + Friction

of horizontal across faces of box acceleration pressure
momentum gradient
in box
(Rotation (Gradients in sea (Internal dissipation
of earth) level and in density) and at ocean
(A) boundaries)

Temperature equation
Rate of change = Transport of temperature + Mixing of temperature
of temperature across faces of box between box and
in box adjacent boxes
(B)

Salinity equation
Rate of change = Transport of salinity + Mixing of salinity
of salinity across faces of box between box and
in box adjacent boxes
(C)

Figure 6 The basic equations for an ocean general circulation model. (A) Momentum equation; (B) temperature equation; (C)
salinity equation. (Reproduced from Summerhayes and Thorpe, 1996.)

the Northern Hemisphere and 901 to the left in the

Earth′s Heat loss
Southern Hemisphere. This is due to the combined
rotation
action of the wind stress on the ocean surface and the
Coriolis force. These Ekman flows can converge and
produce a downwelling flow into the interior of the
ocean. Conversely a divergent Ekman transport will
produce an upwelling flow from the interior into the
Wind surface layer.
This type of flow is known as Ekman pumping,
and is directly related to the Curl of the Wind Stress
(see ). It is of fundamental importance for the driving
of the large-scale horizontal circulation, in the upper
layer of the ocean. For example, between 301 and
u u
501 latitude the climatological westerly wind, drives
T
u Su an Ekman flow equatorward, whereas between 151
and 301 latitude the trade winds drive an Ekman
flow polewards. At about 301 latitude the flows
converge and sink into the deeper ocean. Before
discussing the influence of Ekman pumping on the
interior ocean circulation the role of density is
considered.
Figure 7 A schematic of the model boxes in an ocean general The density of sea water increases with depth.
circulation model. The equations (Figure 6) for momentum are From hydrographic measurements of density, the
solved at the corners of the boxes (u), while the temperature (T),
and salinity (S) equations are solved at the centers of boxes. The
horizontal variation of the depth of a chosen density
model is forced by climatological wind stress, surface heat fluxes, surface can be mapped. These constant density sur-
and freshwater fluxes. (Reproduced from Summerhayes and faces are known as isopycnals. The flow tends to
Thorpe, 1996.) move along these surfaces and therefore the vari-
ations in the depth of these surfaces gives a picture of
the horizontal flow in the deep ocean, away from the
exchange of heat and fresh water between ocean and surface layer and benthic layer. The isopycnal sur-
atmosphere. This layer, which is typically less than faces dip down in the center of the subtropical
100 m in depth, is referred to as the Ekman layer. gyre at about 301. The formation of this lens of
That is a steady wind stress causes a transport of the light warm water is related to the climatological
surface water 901 to the right of the wind direction in distribution of surface winds, which produce a
 OCEAN CIRCULATION 9

convergence of Ekman transport towards the center climatological surface wind circulation and are found
of the gyre, and a downwelling of surface waters into in all the ocean basins.
the interior of the ocean. At the center of the lens, the The surface layer is also subject to heating and
sea surface domes upwards reaching a height of 1 m cooling, and the exchange of fresh water between
above the sea surface at the rim. Due to hydrostatic ocean and atmosphere, both of which will change the
forces the main thermocline is depressed downwards density of the layer. For example, heat is lost over the
to depths of the order of 500–1000 m (Figure 8). Gulf Stream on the rim of the light water lens of the
The surface horizontal circulation flows antic- subtropical gyre. Recall that flow tends to follow
yclonically around the lens with the strongest currents isopycnal layers and these layers will slope down-
on the western edge, where the slope of the density wards towards the center of the gyre. Cooling of the
surface reaches a maximum. These are geostrophic waters in the Gulf Stream leads to the sinking of
currents, where there is a balance between the Cor- surface waters to produce a water mass known as
iolis force and the horizontal pressure gradient force. 181C water. This water, which is removed from the
Generally, the circulation in the subtropical gyres is surface layer, will slowly move along the isopycnal
clockwise in the Northern Hemisphere and anti- layers into the thermocline. As it moves clockwise
clockwise in the Southern Hemisphere. These large- around the gyre it will be subducted in to the deeper
scale horizontal gyres are ultimately caused by the layers of the thermocline, in a spiral-like motion
(Figure 9). The deepest extent of the main thermo-
cline is located in the subtropical gyre to the west of
Bermuda on the eastern edge of the Gulf Stream ra-
ther than in the center of the ocean basin.
2 km
This asymmetry of the gyre is related to the beta
H
effect, i.e., the change of the Coriolis parameter with
6 km H latitude.
F
F

Cooling Cooling
Heating Heating
0 Upwelling in the
subpolar gyre
Thermocline

1 km A/W Warm, salty water

spirals downward
NADW
Thermo-haline
Circulation ABW
6 km
60˚N 30˚ Eq 30˚ 60˚S

Figure 8 A representation of the meridional average section

Heat and salt diffuse
through the atmosphere for December–February in the Northern across pycnocline
Hemisphere. The cells are the Hadley Cell (H) and Ferrel Cell (F).
Pyc
The strength of the cells is represented by the solid contours noc
line
which are in units of 40 megatonnes/second, whilst the dashed Subpolar gyre
contours are in units of 20 megatonnes/second. Note the
predominantly downward motion at B30 degree latitude,
associated with the subtropical anticyclones, and the strong Subtropical gyre
upward motion at equatorial latitudes which is associated with the Below the pycnocline the
Inter-Tropical Convergence Zone. A meridional transect through deep waters have more
uniform temperature and
the Atlantic Ocean, showing the position of the main thermocline. salinity characteristics
The small arrows represent the wind driven downwelling (Ekman
pumping) at B30 degree latitude, and the equatorial upwelling, Figure 9 Schematic representation of the wind-driven
which occurs within and above the main thermocline. The North circulation in the subpolar and subtropical gyre of an ocean
Atlantic Deep Water (NADW) is produced in the Labrador and basin. The wind circulation causes a convergence of Ekman
Nordic Seas and is the predominant deep water mass by volume. transport to the center of the subtropical gyre and downwelling
The Antarctic Intermediate Water (AIW) is produced at B501S, into the interior. Conversely in the subpolar gyre there is a
and by virtue of its salinity is lighter than the NADW. In contrast divergence of the Ekman transport and upwelling from the interior
Antarctic Bottom Water is the most dense water mass in the into the surface layer. This Ekman pumping is responsible for the
worlds ocean and is formed in the Weddell and Ross Seas. gyre circulations (see text for details). The western boundary
These deep flows form part of the thermo-haline circulation. The currents are depicted by the closeness of the streamlines. They
vertical scales are exaggerated in the lower troposphere and in are caused by the poleward change in the Coriolis force known as
the upper ocean. The horizontal scale is proportional to the area the BETA effect. (Reproduced from Bean MS (1997) PhD thesis,
of the Earth’s surface between latitude circles. University of Southampton.)
10 OCEAN CIRCULATION

The subtropical gyres are one of the most well- the ocean have their origin in the polar seas. These
studied regions of the ocean, and our understanding seas experience strong cooling of the surface, par-
is therefore most developed in these regions. These ticularly in the winter seasons. In the North Atlantic,
gyres occur in the surface and thermocline regions of there are connections through the Nordic seas to the
the ocean and are primarily controlled by the wind Arctic Ocean, from which sea ice flows. Heat energy
circulation, with modifications due to heating and melts the ice in the North Atlantic and the melt water
cooling of the surface. The question now arises of gives rise to further cooling. There are two effects on
why thermoclines are seen in the ocean. For example, the density of the water: cooling increases the density
why is the warm water not mixed over the whole whereas surface freshening, due to ice melt, decreases
depth of the ocean and why is the average ocean the density of the water. The former process usually
temperature about 31C. dominates and hence denser waters are produced.
To explain the observed behavior thermohaline These dense cold waters flow into the Atlantic
circulation, which is generated by small horizontal through the East Greenland and West Greenland
differences in density, due to temperature and salin- Currents and then into the Labrador Current. These
ity, between low and high latitude is considered. cold waters mix and sink beneath the warm North
How does it work? Consider an ocean of uniform Atlantic Current.
depth and bounded at the equator and at a polar In addition to surface polar currents there are also
latitude. We will assume it has initially a uniform deep ocean currents. The cold saline water entering
temperature and is motionless (for the moment the from the Nordic seas mixes as it sinks to the abyssal
effect of salinity on density are ignored). This layers of the ocean and moves southward as a deep
hypothetical ocean is then subject to surface heating current along the western boundary of the Atlantic.
at low latitudes and surface cooling at high latitudes. This water mass is known as NADW (North Atlantic
In the lower latitudes the warming will spread Deep Water) and it is the most prominent and volu-
downwards by diffusion, whereas in high latitudes minous of all the deep water masses in the global
the cooling will spread downwards by convection ocean. It flows into the Antarctic Circumpolar Current
which is a much faster process than diffusion. The from where it flows into the Indian and Pacific Ocean.
heavier colder water will induce a higher hydrostatic In addition to NADW, colder denser water, Antarctic
pressure at the ocean bottom than will occur at low Bottom Water (AABW) enters the Southern Ocean
latitudes. The horizontal pressure gradient at the from the Antarctic shelf seas. It is not as voluminous
ocean bottom is directed from the high latitudes to as NADW but it flows northwards in the deepest
the lower latitudes, and will induce an equatorward layers into the Atlantic, where it can be distinguished
abyssal flow of polar water. The flow can not move as far north as 301N. These deep flows upwell into the
through the equatorial boundary of our hypothetical thermocline and surface waters where they return to
ocean and therefore will upwell into the upper layer the North Atlantic. This global thermohaline circu-
of the tropical ocean, where it will warm by dif- lation has been termed the global conveyor circulation
fusion. The flow will then return polewards to the to signify its role in transporting heat and fresh water
high latitudes where it will downwell into deepest (Figures 10 and 11) around the planet.
layers of the ocean to complete the circuit. It is found How does this circulation explain the thermocline?
that the downwelling occurs in narrow regions of the The rate at which these cold deep abyssal waters are
high latitudes whereas upwelling occurs over a very produced can be estimated and it is known for a
large area of the tropical ocean. This hypothetical steady state in the ocean that production has to be
ocean demonstrates the key role of the deep hori- balanced by removal. A large-scale upwelling of the
zontal pressure gradient, caused by horizontal vari- abyssal waters into the thermocline produces this re-
ations in density, for driving the flow. moval. Our simple conceptual picture is of warm
To explain the observed thermohaline circulation, thermocline water mixing downwards, balanced by a
this hypothetical ocean has to be modified to take steady upwelling of the cold abyssal layers. Without
into account the Coriolis force, which causes the the upwelling, the warm waters would mix into the
deep abyssal currents to flow in narrow western deepest layers of the ocean.
boundary currents, the effect of salinity on the
density (the haline component of the flow), asym-
metries in the buoyancy fluxes between the Northern The Role of Fresh Water in Ocean
and Southern Hemispheres and the complex bathe-
Circulation
metry of the ocean basins.
There follows a descriptive account of the ther- The present discussion has shown that the wind-
mohaline circulation. The deepest water masses in driven circulation and the thermohaline circulation
 OCEAN CIRCULATION 11

6
60˚N 0.95
0.77
4
0.88
TA+TO (PW), annual

30˚N 0.43
2 FP
FP + 0.02 0.69
0˚ 0.06
0 FP _ 0.51 FP + 0.30 0.13
30˚S
_2 FP + 0.24
60˚S FA + 0.17 FA + FP _ 0.13 FA
_4
60˚E 120˚E 180˚ 120˚W 60˚W 0˚
(A) _ 6
4 Figure 11 An estimate of the transfer of fresh water (  109 kg
s1) in the world ocean. In polar and equatorial regions
TA (PW), annual

2 precipitation and river run-off exceed evaporation, and hence

there is an excess of fresh water, whereas in subtropical regions
there is a water deficit. A horizontal transfer of fresh water is
0
therefore required between regions of surplus to regions of
_2 deficit. FP and FA refer to the freshwater fluxes of the Pacific–
Indian throughflow and of the Antarctic Circumpolar Current in the
Drake Passage, respectively. (Reproduced from Wijffels SE,
(B) _ 4 Schmitt R, Bryden H and Stigebrandt A (1992) Transport of
4 freshwater by the oceans. Journal of Physical Oceanography 22:
155–162.)
TO (PW), annual

2
and cooling in winter, sink to the deepest layers of
0 the basins. At the Straits of Gibraltar, this dense sa-
_2
line layer flows out beneath the incoming fresher and
cooler Atlantic water. This Mediterranean water
_4 forms a distinct layer of high salinity water in the
(C) 80˚S 60˚ 40˚ 20˚ 0˚ 20˚ 40˚ 60˚ 80˚N eastern Atlantic Ocean. Similar behavior occurs at
Bab el Mandeb adjacent to the Gulf of Aden.
Figure 10 Poleward transfer of heat by (A) ocean and The influence of fresh water is more substantial in
atmosphere together (TA þ TO), (B) atmosphere alone (TA), and
(C) ocean alone (TO). The total heat transfer (A) is derived from
the polar oceans. A given amount of fresh water will
satellite measurements at the top of the atmosphere, that of the have a greater effect on density at low temperatures
atmosphere alone (B) is obtained from measurements of the than at high temperatures, because the thermal ex-
atmosphere, and (C) is calculated as the difference between (A) pansion of sea water decreases with decreasing
and (B). (1 Petawatt (PW) ¼ 1015 W.) Data compiled from three temperature. At higher latitudes there is a net add-
sources. (Reproduced from Carrissimo BC, Oort AH and Von der
Harr TH (1985) Estimating the meridional energy transports in the
ition of fresh water into the oceans, which arises
atmosphere and ocean. Journal of Physical Oceanography 15: from the excess of precipitation over evaporation
52–91.) and the melting of sea ice moving towards the
equator from the polar regions.
are major components of ocean circulation, which The addition of fresh water adds buoyancy to the
are ultimately driven by the surface wind stress and surface layer while cooling removes buoyancy,
buoyancy fluxes. Buoyancy fluxes are the net effect therefore the fresh water will tend to reduce the ef-
of heat exchange and the freshwater exchange with fect of the cooling. In the Arctic Ocean the surface
the overlying atmosphere. It has been shown that layer is colder but less dense than the warmer layer at
heat exchange is a major process explaining exist- B100 m and therefore is in equilibrium. . This stable
ence of both the thermocline and the deep abyssal halocline in the Arctic Ocean reduces the vertical
water but what is the role of the fresh water in ocean heat flux in to the deep ocean.
circulation? In the subpolar oceans, the addition of fresh water
In the subtropics there is net removal of fresh reduces the density of the surface layer and can re-
water by evaporation. This increases the salinity of duce the prevalence of deep convection. This hap-
the water which, in turn, increases the density of the pened in the late 1960s when fresh water, probably
thermocline waters. Normally this effect is opposed from excessive ice in the Arctic Basin, melted in the
by heating, which lightens the water. However, in the subpolar gyre. The effect on the thermohaline cir-
Mediterranean and the Red Sea evaporation pro- culation is unknown, but it is believed from model-
duces salient waters, which by virtue of their salinity ing studies that the decrease in the production of
12 OCEAN CIRCULATION

deep waters reduced the thermohaline circulation of oceans respond to the surface wind stress on seasonal
the ocean. timescales, which allows a strong coupling between
the ocean and atmosphere to take place. This gives
rise to phenomena such as the El Niño Southern
What are the Consequences of this Oscillation (see El Niño Southern Oscillation
Circulation on the Climate System? (ENSO)). The subtropical gyres respond to changes
in the wind circulation on decadal timescales,
The effects of the ocean circulation on the climate whereas the deep thermohaline circulation respond
can be understood in terms of the heat capacity of on millenial timescales. There is some evidence for
the ocean. The heat capacity of a column of sea rapid changes of local parts of the thermohaline
water only 2.6 m deep is equivalent to that of a circulation on timescales 50 years.
column of whole atmosphere and therefore the ocean Observations of the deep ocean are far fewer in
heats and cools on a long timescale compared with number than at the ocean and land surface. The
the atmosphere. longest continuous data set is a deep station at Ber-
It is known that there is a poleward gradient of muda that commenced operations in 1954. Obser-
temperature, which is driven by the thermal radi- vations from cruises in the earlier part of the century
ation imbalance between the low and high latitudes. are of unknown quality and therefore it is difficult to
In response to this temperature gradient there is a know whether differences are due to the use of dif-
flow of heat from the warmer to cooler latitudes. ferent instruments or to real changes in the ocean. It
Both the atmosphere and ocean circulations transfer is only since the 1950s that such changes have been
this heat from low to high latitudes by a variety of accurately measured. Figure 12 shows changes in the
mechanisms. In the low latitudes of the atmosphere temperature for that period of time across the At-
there is the Hadley cell which transfers low tem- lantic. These changes are of the order of a few tenths
perature air in the lowest levels via the trade winds
towards the equator and transfers warmer air pole-
0
ward in the upper troposphere (Figure 8). At higher
latitudes anticyclones and cyclones and their ac-
companying upper air jet streams transfer heat
polewards. In the ocean, the wind-driven Ekman 1000
currents transfer heat as surface waters move across
latitude circles. This water is returned deeper in the
ocean at a different temperature from that of the
surface water. The ocean gyres carry heat towards 2000
higher latitudes since the poleward flows of the
Pressure (dbar)

western part of the gyres are warmer than the

equatorward flows in the eastern parts of the gyre. 3000
Finally, and not least, is the contribution of the
thermohaline circulation, which transports warm
surface and thermocline waters to the highest lati-
tudes and returns cold water to lower latitudes. 4000
Figures 10 and 11 show the heat transport and fresh
water transport in the ocean.
A major difference between the atmosphere and 1992_1957
5000
ocean is the relative speed of their circulation. The 1992_1981
atmosphere circulation is a fast system, responding 1981_1957

on timescales of days to weeks. For example, wea-

ther systems in temperate latitudes grow and decay 6000
on timescales of a few days. By contrast, the ocean _ 0.4 _ 0.3 _ 0.2 _ 0.1 0.0 0.1 0.2 0.3 0.4
tends to be slower in its response. The fastest part of Zonally averaged temperature difference (˚C)
the system are the surface Ekman layers which re-
spond to changes in the surface wind circulation on a Figure 12 Temperature changes (1C) in the subtropical North
Atlantic (241N), 1957–1992. The measurements have been
timescale of one or two days. Changes in wind cir-
averaged across 241N between North Africa and Florida.
culation can cause planetary waves which will (Reproduced from Parrilla G, Lavin A, Bryden H, Garcia M and
change sea level and surface temperature on monthly Millard R (1994) Rising temperatures in the subtropical North
to seasonal timescales. In particular, the equatorial Atlantic. Nature 369: 48–51.)
 OCEAN CIRCULATION 13

of a degree over periods of 15 years. As the heat that the thermohaline circulation may be reduced or
capacity of the oceans is very much larger than that turned off completely when significant excess fresh
of the atmosphere, these changes in temperature in- water is added to the subpolar ocean. In the event of
volve very significant changes in the heat content of thermohaline circulation being significantly reduced
the ocean. The World Ocean Circulation Experiment or stopped, it may take many centuries before it re-
from 1990 to 1997 has provided measurements of turns to its present value. (see Abrupt Climate
ocean properties such as temperature, salinity, and Change, North Atlantic Oscillation (NAO))
chemical tracers as well as current measurements on In view of the current levels of uncertainty, it is
a global scale. This set of high quality measurements necessary to continue to monitor the ocean circu-
will provide the baseline from which future changes lation, as this will provide the key to the under-
in ocean circulation can be determined. standing of the present circulation and enhance our
Despite the brief record of deep ocean obser- ability to predict future changes in circulation.
vations, sea surface temperature measurements of
reasonable quality go back to the late nineteenth
century. These measurements can be used to assess
changes in the surface layers. Salinity measurements See also
are fewer and not as reliable but, nevertheless, Abrupt Climate Change. Antarctic Circumpolar
changes can be still detected. Current. El Niño Southern Oscillation (ENSO).
Salinity measurements in the Mediterranean over Freshwater Transport and Climate. General
the last century have shown a warming of the Circulation Models. Heat Transport and Climate.
Western Mediterranean Deep Water of 0.11C and North Atlantic Oscillation (NAO).
increase of 0.05 in salinity. The reasons for this
change are not known, but it has been speculated
that the change in salinity may be attributed to a
reduction in the freshwater flow due to the damming Further Reading
of the Nile and of rivers flowing into the Black Sea.
Gill AE (1982) Atmosphere–Ocean Dynamics. London:
An important recently identified question is the
Academic Press.
stability of the thermohaline circulation. The ther- Siedler G, Church J, and Gould (2001) Ocean Circulation
mohaline circulation is driven by small density dif- and Climate. London: Academic Press.
ferences and therefore changes in the temperature Summerhayes CP and Thorpe SA (1996) Oceanography –
and salinity arising from global warming may alter An Illustrated Guide. London: Manson Publishing.
the thermohaline circulation. In particular, theore- Wells NC (1997) The Atmosphere and Ocean: A Physical
tical modeling of the ocean circulation has shown Introduction, 2nd edn. Chichester: John Wiley.
 SCIENCE OF OCEAN CLIMATE MODELS
S. M. Griffies, NOAA/GFDL, Princeton, NJ, USA motion involved. We start at the large scale, where a
Published by Elsevier Ltd.
typical ocean basin is on the order of 103–104 km in
horizontal extent, with depths reaching on average
to c. 5 km. The ocean’s massive horizontal gyre and
Introduction: Models as Essential overturning circulations occupy nearly the full extent
Tools for Ocean Science of these basins, with typical recirculation times for
The rich textures and features of the global ocean the horizontal gyres decadal, and overturning time-
circulation and its tracer distributions cannot be fully scales millennial.
studied in a controlled laboratory setting using a At the opposite end of the spectrum, the ocean
direct physical analog. Consequently, ocean scientists microscale is on the order of 10  3 m, which is the
have increasingly turned to numerical models as a scale where molecular viscosity can act on velocity
rational experimental tool, along with theoretical gradients to dissipate mechanical energy into heat.
methods and a growing array of observational This length scale is known as the Kolmogorov
measurements, for rendering a mechanistic under- length, and is given by (v3/e)1/4, where v ¼ 10  6
standing of the ocean. Indeed, during the 1990s and m2 s  1 is the molecular kinematic viscosity for
early 2000s, global and regional ocean models have water, and e is the energy dissipation rate. In turn,
become the experimental tool of choice for many molecular viscosity and the Kolmogorov length
oceanographers and climate scientists. imply a timescale T ¼ L2/vE1 s.
The scientific integrity of computer simulations of When formulating a computational physics prob-
the ocean has steadily improved during the past lem, it is useful to estimate the number of discrete
decades due to deepened understanding of the ocean degrees of freedom required to represent the physical
and ocean models, and from enhanced computer system. Consider a brute force approach, where all
power facilitating more realistic representations of the space and timescales described above are
the huge range of scales relevant for ocean fluid dy- explicitly resolved by the ocean simulation. One-
namics. Additionally, ocean models are a key com- second temporal resolution over a millennial time-
ponent in earth system models (ESMs), which are scale climate problem requires more than 3  1010
used to study interactions between physical, chem- time steps of the model equations. Resolving space
ical, and biological aspects of the Earth’s climate into regions of dimension 10  3 m for an ocean with
system. ESMs are also used to help anticipate volume roughly 1.3  1018 m3 requires 1.3  1027
modifications to the Earth’s climate arising from discrete grid cells (roughly 104 times larger than
humanity’s uncontrolled greenhouse experiment. Avogadro’s number!). These numbers far exceed the
Quite simply, without computer models, our ability capacity of any computer in use today, or for the
to develop a robust and testable scientific basis for forseeable future. Consequently, a truncated de-
ocean and climate dynamics of the past, present, and scription of the ocean state is required.
future would be absent.
There is a tremendous amount of science forming
Truncation Methods
the foundations of ocean models. This science spans
a broad interdisciplinary spectrum, including physics, Three general approaches to truncation are em-
mathematics, chemistry, biology, computer science, ployed in fluid dynamics. One approach is to coarsen
climate science, and all aspects of oceanography. This the space time resolution. Doing so introduces a loss
article describes the fundamental principles forming of information due to the unresolved small scales.
the basis for physical ocean models. We are particu- Determining how the resolved scales are affected by
larly interested here in global ocean climate models the unresolved scales is fundamental to computa-
used to study large-scale and long-term phenomena tional fluid dynamics, as well as to a statistical de-
of direct importance to climate. scription of fluid turbulence. This is a nontrivial
problem in subgrid-scale (SGS) parametrization,
Scales of Motion and the a problem intimately related to the turbulence clos-
ure problem of fluid dynamics.
Subgrid-scale Problem The second truncation method filters the con-
To appreciate the magnitude of the task required to tinuum equations by truncating the fundamental
simulate the ocean circulation, consider the scales of modes of motion admitted by the equations.

14
 SCIENCE OF OCEAN CLIMATE MODELS 15

The result is an approximation to the original phys- kinetic energy through the process of baroclinic in-
ical system. The advantage is that reducing the ad- stability. Such eddy features are the norm rather than
mitted motions also reduces the space timescales the exception in most of the ocean. They provide a
required to simulate the system, and/or it simplifies chaotic or turbulent element to the ocean general
the governing equations thus facilitating compu- circulation. Hence, quite generally the ocean flow is
tationally cheaper simulations. The hydrostatic ap- not smooth and laminar, unless averaging over many
proximation is a prime example of filtering used in all years. Rather, it is turbulent and full of chaotic fine-
present-day global ocean climate models. Here, ver- scale motions which make for an extremely chal-
tical motions are assumed to possess far less energy lenging simulation task.
than horizontal, thus rendering a simplified vertical Mesoscale eddies have scales on the order of
momentum balance where the weight of fluid above a 100 km in the middle latitudes, 10 km in the higher
point in the ocean determines the pressure at that latitudes, and their timescale for recirculation is on
point. The Boussinesq approximation provides an- the order of months. The length scale is related to the
other filtering of the fundamental equations. In this first baroclinic Rossby radius, which is a scale that
case, the near incompressibility of seawater is ex- arises in the baroclinic instability process. Further-
ploited to eliminate all acoustic motions by assuming more, mesoscale eddies are the ocean’s analog of
that fluid parcels conserve volume rather than mass. atmospheric weather patterns, with atmospheric
The Boussinesq approximation is commonly used in weather occurring on scales roughly 10 times larger
ocean climate modeling, but it is becoming less so due than the ocean eddies. Differences in vertical density
to its inability to provide a prognostic budget for sea stratification account for differences in length scales.
level that includes steric effects. Steric effects are as- Ocean tracer properties are strongly affected by
sociated with water expansion or contraction arising the mixing and stirring of mesoscale eddies. For ex-
from density changes, and such changes are a key ample, biological activity is strongly influenced by
aspect of the ocean’s response to anthropogenic cli- eddies, especially with the strong upwelling in the
mate change, with a warmer ocean occupying a larger cyclonic eddies bringing high nutrient water to the
volume which raises the sea level. surface. Additionally, eddies are a leading order
A final truncation method considers a much feature transporting properties such as heat and
smaller space and time domain, yet maintains the freshwater poleward across the Antarctic Circum-
very fine space and time resolution set by either polar Current. They also mix properties across the
molecular viscosity (direct numerical simulation time-mean position of the Gulf Stream and Kuroshio
(DNS)), or somewhat larger viscosity (large eddy Currents, whose jet-like currents continually fluctu-
simulation (LES)). Both DNS and LES are important ate due to baroclinic instability.
for process studies aimed at understanding the Given the smaller scales of mesoscale eddies in
mechanisms active in fine-scale features of the ocean. the ocean than in the atmosphere, the problem of
Insights gained via DNS and LES have direct appli- simulating these ocean features is 10  10  10 more
cation to the development of rational SGS para- costly than in the atmosphere. Here, two factors of
metrizations of use for ocean climate models. 10 arise from the horizontal scales, and another from
the associated refinement in temporal resolution. It is
not rigorously known what grid resolution is re-
Ocean Mesoscale Eddies
quired to resolve the ocean’s mesoscale eddy spec-
The Reynolds number UL/v provides a dimensionless trum. However, preliminary indications point to
measure of the importance of advective effects relative 10 km or finer globally, with roughly 50–100 vertical
to viscous or frictional effects. Consider a large-scale degrees of freedom needed to capture the verti-
ocean current, such as the Gulf Stream, with a velocity cal structure of the eddies, as well as the tight vertical
scale U ¼ 1 m s  1, length scale L ¼ 100 km, and gradients in tracer properties. This resolution is far
molecular viscosity v ¼ 10  6 m2 s  1. In this case, the coarser than that required to resolve the Kolmogorov
11
Reynolds number is ReE10 , which is very large and scale (as required for DNS). Nonetheless, globally
so means that the ocean fluid contains extremely resolving the ocean mesoscale eddy spectrum over
turbulent regimes. At these space and timescales, the climate timescales remains beyond our means, with
effects of rotation and stratification are both critical. the most powerful computers only just beginning to
The turbulence relevant at this scale is termed geo- be applied to such massive simulations.
strophic turbulence. A geostrophically turbulent fluid Ocean climate modelers thus continually seek en-
contains numerous mesoscale eddy features, which hanced computer power in an aim to reduce the level
result from a conversion of potential energy (imparted of space time resolution coarsening. The belief,
to the ocean fluid by the atmospheric forcing) to largely reflected in experience with simulations of
16 SCIENCE OF OCEAN CLIMATE MODELS

refined resolution, is that reducing dependence on Posing the Problem of Ocean

SGS parametrizations, such as those parametrizing Modeling
mesoscale eddies, improves the simulation integrity.
One reason for this improvement is that many SGS The ocean is a forced and dissipative system. Forcing
parametrizations are far from relevant for all of the occurs at the upper boundary from interactions with
multiple flow regimes of the ocean. Additionally, no the atmosphere, rivers, and sea ice, and at its lower
SGS parametrization commonly used in ocean cli- boundary from the solid Earth. Forcing also occurs
mate models includes the stochastic effects of tur- from astronomical effects of the Sun and Moon
bulent eddies, and resolving these effects may be which produce tidal motions.
important for climate variability and predictability. Important atmospheric forcing occurs over basin
Finally, refined resolution allows for a better repre- scales, with timescales set by the diurnal cycle, syn-
sentation of the very complex land–sea boundaries optic weather variability (days), the seasonal cycle,
that strongly influence ocean currents and water and interannual fluctuations such as the North
mass properties. Hence, deducing the impact on Atlantic Oscillation and El Niño. Atmospheric mo-
ocean climate of explicitly resolved mesoscale eddies, mentum and buoyancy fluxes are predominantly re-
and other fine-scale currents such as occur near sponsible for driving the ocean’s large scale
boundaries, is the grand challenge of ocean climate horizontal and overturning circulations. Additional
modeling in the early twenty-first century. influences include forcing at continental boundaries
As grid resolution is refined so that mesoscale eddy from river inflow and calving glaciers, as well as in
features are admitted, the resulting eddy permitting polar regions where sea ice dynamics greatly affect
simulations become strongly dependent on the in- the surface buoyancy fluxes. Dissipation in the ocean
tegrity of the numerical methods used for the trans- generally occurs at the microscale, with turbulent
port of fluid properties. Although the representation processes converting mechanical energy to heat. En-
of transport is important at coarse resolution, the ergy moves from the large to the small scales through
fluid is also very viscous due to the huge viscosity nonlinear advective transport. In short, the ocean
required at the noneddying resolutions. Hence, a circulation emerges from an enormous number of
more energetic simulation, with enhanced energy in processes spanning a huge space and time range.
the resolved scale and much stronger fronts between Furthermore, modes of ocean dynamical motions
ocean properties, generally stresses the ability of span space and timescales even greater than the
transport algorithms to maintain physically appro- forcing scales, with ocean circulation patterns, es-
priate behavior. The integrity of the discrete trans- pecially those in the abyssal regions, maintaining
port operators is thus arguably the most critical coherence over global scales for millennia.
feature of a numerical algorithm that sets the fidelity From a mathematical physics perspective, the
of eddying simulations. problem of modeling the ocean circulation is a
Although mesoscale eddy permitting simulations problem of establishing and maintaining the kin-
are becoming more feasible as computer power in- ematic, dynamic, and material properties of the
creases, there remain important problems where SGS ocean fluid: a fluid which is driven by a multitude of
parametrizations must be used to partially capture boundary interactions and possesses numerous in-
features of the unresolved mesoscale spectrum, as ternal transport and mixing processes. Kinematic
well as smaller scales. The longer the timescale of the balances are set by the geometry of the ocean do-
phenomena studied, the greater the need for SGS main, and by assuming that the fundamental fluid
parametrization. In particular, no mesoscale eddy parcel constituents conserve mass. Dynamical bal-
simulation has yet been run for more than a few ances result from applying Newton’s law of motion
decades. Climate modelers must therefore carefully to the continuum fluid so that the acceleration of a
employ truncation methods to make progress. De- fluid parcel is set by forces acting on the parcel, with
pending on particulars of the unresolved motions, the dominant forces being pressure, Coriolis, gravity,
the resulting SGS parametrizations can, and do, af- and friction. And material balances of tracers, such
fect in nontrivial manners the simulation’s physical as salt, heat, and biogeochemical species, are affected
integrity. Consequently, a great deal of intellectual by circulation, mixing from turbulence, boundary
energy in ocean model algorithm development fluxes, and internal sources and sinks. The ocean
relates to establishing numerically robust and modeling problem also involves, at a fundamental
physically rational SGS parametrizations. This is an level, a rational parametrization of SGS processes,
extremely difficult problem. Nonetheless, some pro- such as fine-scale convective mixing in the upper
gress has been made during the past decades, at least ocean mixed layer due to strong mechanical and
enough to know what not to do in certain cases. buoyant interactions with the atmosphere, breaking
 SCIENCE OF OCEAN CLIMATE MODELS 17

internal waves induced from tides rubbing against where @ t measures time changes at a fixed space
the ocean bottom, and mesoscale eddies spawned by point, and v is the parcel’s velocity. The transport or
baroclinically unstable currents. From these balances advective operator v  r reveals the fundamentally
emerges the wonderful richness of the ocean circu- nonlinear character of fluid dynamics arising from
lation and its variety of tracer distributions. Faith- motions of fluid parcels with velocity v. In the
fully emulating this system using computer Eulerian perspective, mass conservation takes the
simulations requires a massive effort in scientific and form
engineering ingenuity and collaboration.
@t r þ r  ðrvÞ ¼ 0 ½3

Fundamental Budgets and Methods Tracer Budgets

We now aim to place the previous discussions onto a
In addition to freshwater, a seawater parcel contains
more rigorous mathematical basis. Doing so is a
numerous constituents known as tracers. Heat (a
necessary first step toward developing a suite of
thermodynamical tracer) and salt (a material tracer)
equations amenable to numerical methods for use on
are seawater’s two active tracers. Heat and salt,
computers. Due to limitations in space, the material
along with pressure, determine the mass density
here will be quite terse and selective, allowing many
through a complex empirical relation known as
results to be presented without full derivation and
the equation of state. Density in turn impacts the
many topics to be omitted.
hydrostatic pressure which then affects currents.
The basic equations of ocean fluid dynamics can
Other ocean tracers include radioactive species, such
be readily formulated by focusing on the dynamics of
as those introduced in the 1950s and 1960s from
a mass conserving parcel of seawater. A fluid parcel is
nuclear bomb tests; biological tracers such as
a region that is macroscopically small yet micro-
phytoplankton and zooplankton fundamental to
scopically large. That is, from a macroscopic per-
ocean ecosystems; and chemical elements such as
spective, the parcel’s thermodynamic properties may
carbon, iron, and nitrogen prominent in ocean bio-
be assumed uniform, and the methods of continuum
geochemical cycles.
mechanics are applicable to describing the mechanics
In describing the evolution of tracer within a sea-
of an infinite number of these effectively infinitesimal
water parcel, it is convenient to consider the tracer
parcels. However, from a microscopic perspective,
concentration C, which represents a mass of tracer
these fluid parcels contain many molecules (on the
per mass of seawater for material tracers. In addition
order of Avogadro’s number), and so it is safe to
to material tracers, we are concerned with a ther-
ignore the details of molecular interactions. Regions
modynamical tracer that measures the heat within a
of a fluid with length scales on the order of 10  5 m
fluid parcel. In this case, C is typically taken to be the
satisfy these properties of a fluid parcel.
potential temperature or potential enthalpy.
The evolution of tracer mass within a Lagrangian
Mass Conservation parcel of mass conserving fluid is given by
The mass of a parcel is written as dM ¼ rdV, where r
dC
is the mass per volume (i.e., the density), and dV is r ¼ r  J ½4
the parcel’s volume. Assuming the parcel mass to be dt
conserved as the parcel moves through the fluid leads where for simplicity we ignore tracer sources. The
to the differential statement of mass conservation tracer flux J arises from SGS transport of tracer in the
d absence of mass transport. Such transport consists of
lnðdMÞ ¼ 0 ½1 diffusion and/or unresolved advection. As this flux is
dt
not associated with mass transport, it vanishes when
where we assume there to be no internal sources of the tracer concentration is uniform, in which case the
mass. The time derivative d/dt measures changes tracer budget reduces to the mass budget of eqn [1].
occuring with respect to the moving parcel. This Use of the material time derivative relation [2] and
Lagrangian or material perspective complements the mass conservation [3] renders the Eulerian conser-
Eulerian perspective rendered by viewing the fluid vation law for tracer
from a fixed point in space. Transforming to the
Eulerian perspective leads to the relation @t ðrCÞ þ r  ðrCvÞ ¼ r  J ½5

d The convergence  r  (rCv) is the flux form of ad-

¼ @t þ v  r ½2
dt vective tracer transport. As stated in the section titled
18 SCIENCE OF OCEAN CLIMATE MODELS

‘Ocean mesoscale eddies’, advective transport is for ocean climate modeling. The Coriolis force
critical, especially in eddying simulations where both per mass is written f ẑ4v, with the Coriolis para-
the tracer and velocity distributions possess non- meter f ¼ 2O sin f, where O ¼ 7.292  10  5 s  1 is
trivial structure. The flux form of advective tracer the Earth’s angular rotation rate, and f is the lati-
transport is employed in ocean climate models rather tude. Gradients in pressure p impart an acceleration
than the alternative advective form rv  rC. The on the fluid parcel, with the parcel accelerated from
reason is that the flux form is amenable to con- regions of high pressure to low.
servative numerical schemes, as well as to a finite The friction vector F arises from the divergence of
volume interpretation described in the section titled internal friction stresses. In an ocean model, these
‘Basics of the finite volume method’. stresses must be sufficient to maintain a near-
The specification of SGS parametrizations ap- unit-grid Reynolds number. Otherwise, the simu-
pearing in the flux J is critical for the tracer equation, lation may go unstable, or at best produce unphysical
especially in models not admitting mesoscale eddies. noise-like features. This constraint on the numerical
The original approach, whereby the most common simulation is unfortunate, since a unit grid Reynolds
class of ocean climate models employed a diffusion number requires an effective viscosity many orders
operator oriented according to geopotential surfaces, of magnitude larger than molecular, since the grid
greatly compromised the simulation’s physical in- sizes (order 104–105 m) are much larger than the
tegrity for climate purposes. The problem with the Kolmogorov scale (10  3 m). This level of numerical
horizontally oriented operators is that they introduce friction is not based on physics, but arises due to the
unphysically large mixing between simulated water discrete lattice used for the numerical simulations.
masses. Given the highly ideal fluid dynamics of the Various methods have been engineered to employ the
ocean interior, most of the tracer transport in the minimal level of friction required to meet this, and
interior arising from eddy effects occurs along a lo- other, numerical constraints, so as to reduce the ef-
cally referenced potential density direction, other- fects of friction on the simulation. Such methods are
wise known as a neutral direction. This mixing ad hoc at best, and lead to some of the most un-
preserves water mass properties over basin scales for satisfying elements in ocean model practice since the
decades. Altering the orientation of the model’s dif- details of the methods can strongly influence the
fusion operator from horizontal–vertical to neutral– simulation.
vertical brought the models more in line with the real The hydrostatic approximation mentioned in the
ocean. In addition, mesoscale eddies stir tracers in a section titled ‘Scales of motion and the subgrid-scale
reversible manner. This stirring corresponds to an problem’ exploits the large disparity between hori-
antisymmetric component to the SGS tracer trans- zontal motions, occurring over scales of many tens to
port tensor. The combination of neutral diffusion hundreds of kilometers, and vertical motions, oc-
and skew diffusion have become ubiquitous in all curring over scales of tens to hundreds of meters. In
ocean climate models that do not admit mesoscale this case, it is quite accurate to assume that the
eddies, even those models not based on the geopo- moving fluid maintains the hydrostatic balance,
tential vertical coordinate. whereby the vertical momentum equation takes the
form
Linear Momentum Budget
@z p ¼ rg ½7
The linear momentum of a fluid parcel is given by
vrdV. Through Newton’s law of motion, momentum This approximation is a fundamental feature of the
changes in time due to the influence of forces acting ocean’s primitive equations, which are the equations
on the parcel. The forces acting on the ocean fluid solved by all global ocean climate models in use
parcel include internal stresses in the fluid arising today. By truncating, or filtering, the vertical mo-
from friction and pressure; the Coriolis force due to mentum budget to the inviscid hydrostatic balance,
our choice of describing the ocean fluid from a ro- we are obliged to parametrize strong vertical mo-
tating frame of reference; and gravity acting in the tions occurring in convective regions, since the
local vertical direction. The resulting equation for primitive equations cannot explicitly represent these
linear momentum takes the form motions. This has led to various convective para-
dv metrizations in use by ocean climate models. These
r ¼ rg ẑ  f ẑ4rv  rp þ rF ½6 parametrizations are essential for the models to
dt
accurately simulate various deep-water formation
In this equation, gE9.8 m s  2 is the acceleration processes, especially those occurring in the open
due to gravity, which is generally assumed constant ocean due to strong buoyancy fluxes.
 SCIENCE OF OCEAN CLIMATE MODELS 19

General Strategy for Time-stepping Momentum written over an arbitrary finite region as
Acoustic waves are three-dimensional fluctuations in Z Z Z  Z Z
the pressure field. They travel at roughly 1500 m s  1. @t rC dV ¼ dAðn̂Þ n̂  ðvrel C þ JÞ
There is no evidence that resolving acoustic waves is
essential for the physical integrity of ocean climate ½8
models. The hydrostatic approximation filters all
acoustic modes except the Lamb wave (an acoustic The volume integral is taken over an arbitrary fluid
mode that propogates only in the horizontal dir- region, and the area integral is taken over the
ection). As the Lamb wave is close in speed to the bounding surface to that volume, with outward
external gravity waves, it can generally be subsumed normal n̂ and area element dAðn̂Þ . The velocity vrel is
into an algorithm for gravity waves, and so is of little determined by the relative velocity of the fluid parcel,
consequence to ocean model algorithms. v, and the moving boundary. As the advective and
External or barotropic gravity waves are roughly diffusive fluxes penetrate the boundary, they alter
100 times the speed of the next fastest internal wave the tracer mass in the region. The budget for the
or advective vector linear momentum can also be written in this
pffiffiffiffiffiffiffi signal. In their linear form, they travel at
speed gH , with H the ocean depth and g the ac- form, with the addition of body forces from gravity
celeration of gravity. In the deep ocean, they are and Coriolis which act over the extent of the
about 5–10 times slower than acoustic waves. volume. Once formulated in this manner, the dis-
External gravity waves are nearly two-dimensional cretization problem shifts from fundamentals to de-
in structure, so they are largely represented by dy- tails, with details differing on how one represents the
namics of the vertically integrated fluid column. This SGS behavior of the continuum fields. This then
property motivates the formulation of primitive leads to the multitude of discretization methods
equation model algorithms which split the relatively available for such processes as transport, time-
fast vertically integrated dynamics from the slow and stepping, etc.
more complicated vertically dependent dynamics.
Doing so allows for a more efficient time-stepping
method to update the ocean’s momentum field. De-
Elements of Vertical Coordinates
tails of this split are often quite complex, and require The choice of how to discretize the vertical direction
care to ensure that the overlap between the fast and is the most important choice in the design of a nu-
slow modes is trivial, or else suffer consequences of merical ocean model. The reason is that much of the
an unstable simulation. Nonetheless, these algo- model’s algorithms and SGS parametrizations are
rithms form a fundamental feature in all ocean cli- fundamentally influenced by this choice. We briefly
mate models. outline here some physical considerations which
may prejudice a choice for the vertical coordinate.
For this purpose, we identify three regimes of the
Basics of the Finite Volume Method ocean germane to the considerations of a vertical
coordinate.
The previously derived budgets for infinitesimal fluid
parcels represent a starting point for ocean climate • Upper ocean mixed layer. This is a generally tur-
model algorithm designs. The next stage is to pose bulent region dominated by transfers of mo-
these budgets on a discrete lattice, whereby the mentum, heat, freshwater, and tracers with the
continuum fields take on a finite or averaged inter- overlying atmosphere, sea ice, rivers, etc. It is a
pretation. There is actually more than one one way primary region of importance for climate system
to interpret the relation between the continuum modeling. It is typically very well mixed in the
variables and those living on the numerical lattice. vertical through three-dimensional convective and
We introduce one method here, as it has achieved turbulent processes. These processes involve non-
some recent popularity in the ocean model literature. hydrostatic physics which requires very high
The finite volume method takes the flux form horizontal and vertical resolution (i.e., a vertical
continuum equations and integrates them over the to horizontal grid aspect ratio near unity) to ex-
finite extent of a discrete model grid cell. The re- plicitly represent. A parametrization of these
sulting control volume budgets are the basis for es- processes is therefore necessary in primitive
tablishing an algebraic algorithm amenable to equation ocean models which exploit the hydro-
computational methods. For example, vector calcu- static approximation. In this region, it is essential
lus allows for the parcel budget of a scalar field, such to employ a vertical coordinate that facilitates
as a tracer concentration described by eqn [5], to be the representation and parametrization of these
20 SCIENCE OF OCEAN CLIMATE MODELS

highly turbulent processes. Geopotential and required hybrid vertical coordinate methods remains
pressure coordinates, or their derivatives, are an active area in ocean model design.
the most commonly used coordinates as they
facilitate the use of very refined vertical grid spa-
cing, which can be essential to simulate the strong Ocean Climate Modeling
exchanges between the ocean and atmosphere, The use of physical ocean models to simulate the
rivers, and ice. ocean requires understanding and knowledge beyond
• Ocean interior. Tracer transport processes in the the fundamentals discussed thus far in this article.
ocean interior predominantly occur along neutral Most notably, we require information about boun-
directions. The transport is largely dominated by dary fluxes. There are two basic manners that ocean
large-scale currents and mesoscale eddy fluctu- models are generally used for simulations: as com-
ations. Water mass properties in the interior thus ponents of an ESM, whereby boundary fluxes are
tend to be preserved over large space and time- computed from atmosphere, sea ice, and river com-
scales (e.g., basin and decade scales). This prop- ponent models, based on interactions with the
erty of the ocean interior is critical to represent in evolving ocean; or in a stand-alone mode where
a numerical simulation of ocean climate. A po- boundary fluxes are prescribed from a data set, in
tential density, or isopycnal coordinate, frame- which case the uncertainties are huge due to sparse
work is well suited to this task, whereas measurements over much of the ocean. This un-
geopotential, pressure, and terrain-following certainty in observed fluxes greatly handicaps our
models have problems associated with numerical ability to unambiguously untangle model errors (i.e.,
truncation errors. The problem becomes more errors in numerical methods, parametrizations, and
egregious as the model resolution is refined, due to formulations) from flux errors.
the enhanced levels of eddy activity that pumps Correspondingly, the process of evaluating the fi-
tracer variance to the grid scale. Quasi-adiabatic delity of ocean simulations remains in its infancy
dissipation of this variance is difficult to maintain relative to the situation in atmospheric modeling,
in nonisopycnal models. where synoptic weather forecasts provide a stringent
• Ocean bottom. The ocean’s bottom topography test of model fidelity. Nonetheless, ocean obser-
acts as a strong forcing on the overlying currents vations, including boundary fluxes, are steadily im-
and so directly influences dynamical balances. In proving, with new observations providing critical
an unstratified ocean, the flow generally follows benchmarks for evaluating the relevance of ocean
lines of constant f/H, where f is the Coriolis par- simulations. Given the wide-ranging spatial and
ameter and H the ocean depth. Additionally, there temporal scales of oceanic phenomena, it is essential
are several regions where density-driven currents that observations be maintained over a wide network
(overflows) and turbulent bottom boundary layer in both space and time. In absence of this network,
(BBL) processes act as a strong determinant of we are unable to provide a mechanistic understand-
water mass characteristics. Many such processes ing of the observed ocean climate system.
are crucial for the formation of deep-water
properties in the World Ocean, and for repre-
senting coastal processes in regional models. It is
See also
for this reason that terrain-following models have
been developed over the past few decades, with Coupled Sea Ice-Ocean Models. Heat and
their dominant application focused on the coastal Momentum Fluxes at the Sea Surface. Heat
and estuarine problem. Transport and Climate. Ocean Carbon System,
Modeling of. Ocean Circulation: Meridional
The commonly used vertical coordinates introduced Overturning Circulation.
above can have difficulties accurately capturing all
flow regimes. Unfortunately, each regime is import- Further Reading
ant for accurate simulations of the ocean climate
Bleck R (2002) An oceanic general circulation model frame
system. To resolve this problem, some researchers
in hybrid isopycnic–Cartesian coordinates. Ocean
have proposed the use of hybrid vertical coordinates Modelling 4: 55--88.
built from combinations of the traditional choices. Durran DR (1999) Numerical Methods for Wave
The aim is to employ a particular vertical coordinate Equations in Geophysical Fluid Dynamics, 470 pp.
only in a regime where it is most suitable, with Berlin: Springer.
smooth and well-defined transitions to another co- Gent PR, Willebrand J, McDougall TJ, and McWilliams JC
ordinate when the regime changes. Research into the (1995) Parameterizing eddy-induced tracer transports in
 SCIENCE OF OCEAN CLIMATE MODELS 21

ocean circulation models. Journal of Physical McClean JL, Maltrud ME, and Bryan FO (2006)
Oceanography 25: 463--474. Quantitative measures of the fidelity of eddy-resolving
Griffies SM (2004) Fundamentals of Ocean Climate Models, ocean models. Oceanography d192: 104--117.
518pp. Princeton, NJ: Princeton University Press. McDougall TJ (1987) Neutral surfaces. Journal of Physical
Griffies SM, Böning C, Bryan FO, et al. (2000) Oceanography 17: 1950--1967.
Developments in ocean climate modelling. Ocean Müller P (2006) The Equations of Oceanic Motions,
Modelling 2: 123--192. 302pp. Cambridge, UK: Cambridge University Press.
Hirsch C (1988) Numerical Computation of Internal and
External Flows. New York: Wiley.
 GENERAL CIRCULATION MODELS
G. R. Ierley, University of California San Diego, La • An equation to express the principle of mass
Jolla, CA, USA conservation.
Copyright & 2001 Elsevier Ltd. • A heat equation, which describes the advection
(carrying by the fluid) and diffusion of temperature.
• A similar ‘advection–diffusion’ equation for salinity.
• An equation of state, which relates the pressure to
the density.
Introduction
Although it may, after the fact, sound obvious, it
A general circulation model (GCM) of the ocean is took scientists many years to appreciate the full role
nothing more than that – a numerical model that of rotation – which enters Newton’s law through a
represents the movement of water in the ocean. latitude dependent Coriolis force – in generating the
Models, and more particularly, numerical models, observed large scale circulation of ocean.
play an ever-increasing role in all areas of science; in Elements of these equations can become quite
geophysics broadly, and in oceanography specifically. complicated. For example, momentum in the upper
It was perhaps less the early advent of super- ocean is imparted by a complex, not yet fully
computers than the later appearance of powerful understood, process of wind–wave interaction. One
personal workstations (tens of megaflops and has to choose whether to represent the action of the
megabytes) that effected not only a visible revolution wind kinematically, which means that the spatial and
in the range of possible computations but also a temporal variability of the wind must be given be-
more subtle, less often appreciated, revolution in the forehand, or dynamically, in which case we must
very nature of the questions that scientists ask, and solve not only the set of equations above, but a
the answers that result. similar set that describes the simultaneous evolution
The range of length scales and timescales in of the atmosphere. Clearly the dynamical case is the
oceanography is considerable. Important dynamics, more ‘realistic’ of the two, but the price is a sub-
such as that which creates ‘salt fingers’ and hence in- stantially more involved computation.
fluences the dynamically significant profile of density It is issues such as these that force one to a choice
versus depth, takes place on centimeter scales, while that often pits understanding and intuition on the
the dominant features in the average circulation cas- one hand against verisimilitude and complete dy-
cade all the way to basin scales of several thousand namical consistency on the other. As the common
kilometers. Timescales for turbulent events, like waves aim of most large-scale models is to generate results
breaking on shore, are small fractions of seconds, of maximum realism, it usual that the models are
while at the opposite end, scientists have reliably frequently corrected, or ‘steered’, on the fly by ex-
identified patterns in the ocean with characteristic tensive use of data assimilation. It is not merely a
evolution times of order several decades.1 matter of slight refinement: no large-scale model
Most simply, a ‘model’ is no more than a math- (GCM) is yet sufficiently robust that it can be used
ematical description of a physical system. In the case for forecasting without considerable input of such
of physical oceanography, that description includes observationally derived constraints; in the oceanic
the following elements: component, for example, the need to force model
• The momentum equation (F ¼ mv_ , but expressed agreement at depth by continuously ‘relaxing’ the
in terms appropriate to a continuous medium), or solution to the smoothed data of the Levitus atlas (a
often in its place a derivative form, the ‘vorticity world-wide compendium of data from many sources,
equation’. smoothed and interpolated onto a regular grid). How
much of this fragility is because of explicit defects
(faulty ‘subgrid scale modeling’ and explicit omis-
1
It is useful to distinguish extrinsic evolution times, which span sions in the physics e.g., neglect of wave breaking)
geologic time, from intrinsic variation, which characterizes an and how much is due to discrete numerical imple-
isolated ocean and atmosphere, thus neglecting such secular in- mentation inconsistent with plausible continuous
fluences as orbital variation, change in the earth’s rotation rate, equations remains an open question.
variation in the solar constant, etc. Although not all causes of
variability have been identified, it is possible that even documented
Their limitations notwithstanding, large numerical
evolution over thousands of years may reflect the latter, intrinsic, models are one vital means by which we grapple with
variability. questions about global warming and a host of other

22
 GENERAL CIRCULATION MODELS 23

environmental issues that affect the way that both we instead we reach a tentative conclusion of not
and future generations will live. proven wrong (yet!). While it may stimulate con-
jecture, and have other worthy ends, the idea of
introducing artificial parameters for the express
purpose of manufacturing close agreement with
Models in Theory and Practice reality is nonetheless formally antithetical to the
Historically, computers were initially so limited that paradigm of testing independent, quantitative
the questions posed were often, in effect, slight ex- predictions.
tensions of preexisting analytic queries. To that ex- • Backward compatibility. As with confinement of
tent, such studies continued to conform (at least in form, the idea of compatibility achieves a purity in
principle) to what we may identity as four basic mathematics that is not to be expected of the
building blocks of most theories, which, in aiming to physical sciences. Each new bit of mathematics
describe physical reality become subject to con- must fit perfectly into the entire edifice of results
straints. Classical modes to which physical laws are already discovered (or invented, as you will).
found to conform generally2 include the following. Commonly, though not without exception, in the
physical sciences, newer theories are seen to en-
• Expression in quantity, extent, and duration. The
compass the older theories as special or limiting
language that offers itself as encompassing all
cases. We speak, for example, of the ‘classical
distinct physical conditions and all meaningful
limit’ as a means of recovering prerelativistic or
physical relations is fundamentally that of math-
prequantum results. Indeed, it is only in light of
ematics. As others before and since, the great
Einstein’s theory of special relativity that we can
mathematician Eugene Wigner too had a stab at
understand the limitations of Newton’s law, which
explaining what, in an eponymous essay, he
we now understand more fully as not a law, but a
termed ‘the unreasonable effectiveness of math-
limiting approximation. Oceanography has
ematics in the physical sciences’. It remains a
families of theories, each nesting one within the
conundrum.
next, like a series of oceanographic Matryoshka
• Confinement of form. This attains its purest ex-
dolls, the innermost of which is often the theory of
pression in the Platonic view that mathematical
‘quasigeostrophy,’ which dates from the 1950s.
entities are not invented, but discovered, and as
such have a prior, if not physical, existence. In
extension to physical theories, it corresponds to The unavoidable adoption and resulting sensitivity of
the belief that there is some true, ultimate, equa- GCMs to ad hoc parameterizations (e.g., subgrid
tion association with any natural phenomenon. scale modeling) or necessity set them apart from the
One important respect in which this idea must be traditional pursuit of the scientific method. This
tempered in application to fluids is the math- distinction was (presciently) appreciated at least by
ematical demonstration that a variety of different the early 1960s and it changes, or ought to change,
microscopic laws for interaction may all yield the one’s view of such models as rigorous arbiters of
same generic macroscopic law applicable to be- precise truths. And yet while it is true that numerical
havior at large space scales or timescales. If one’s experiments with GCMs are merely suggestive rather
aim is solely to understand the latter, then al- than truly predictive of future evolution of the ocean,
though for a given problem we might presume the sheer lack of experimental data, to say nothing of
that there is indeed a precise, if complicated, the lack of a control, means that theoretical ideas are
microscopic law, one’s effort might more profit- often assessed on the basis of their success in ex-
ably be spent understanding the passage to the plaining strictly numerical experiments.
large-scale limit. As computers became more powerful it was nat-
• Falsifiability. Implicit in the progress of science is ural to press for the most realistic model runs pos-
the idea that one makes and then tests hypotheses. sible. And, because the growth in computing power
But unlike in mathematics, in the physical sciences was increasingly realized through distributed3 as well
we cannot show the hypothesis is correct, only as mainframe (super), computing, the school of
that it is wrong. A hypothesis is never vindicated, ‘kitchen sink’ models, which started as a specialized
branch off the mainstream of oceanography – largely

2
We speak gently here, since to insist that those four are always
3
either necessary or sufficient would require that we introduce a Both virtually, through high speed and increasingly transparent
theory about theories: a metatheory. But we have no notion of how networks, and physically, through desktop workstations of con-
rigorously to evaluate such ideas! siderable power.
24 GENERAL CIRCULATION MODELS

limited in participation to those in close physical upon a not yet wholly secure foundation: it is still an
proximity to two or three central machines – became open research question whether the basic equation of
a powerful tributary in its own right: an autonomous fluid mechanics (the Navier–Stokes equation) is itself
discipline within oceanography, which naturally ‘globally well-posed’ in three dimensions.5 (It is
began to evolve its own criteria for relevance. widely believed to be so, but belief comes cheap.
A proof, however, is worth one million dollars(!) –
one of seven prizes in a competition recently an-
Limits on Numerical Models nounced by the Clay Mathematics Institute.)
We spoke of two sources of error common to large Even overlooking such foundational matters on
numerical models: difficulties in numerical imple- which mainly mathematicians would cavil about
mentation, and poorly modeled or unrepresented numerical models, the last three of the four principles
physical processes. In this section we consider spe- above are often violated in more apparent ways in
cific instances of each, starting with an abstract the application and development of present-day
mathematical point of view. But note that if we could models. We illustrate this divergence from traditional
with the wave of hand dispense with these two issues norms with a few representative examples to em-
(which one supposes are in principle tractable), the phasize the sometimes causal (not casual!) link be-
fact that we do not know the exact physical state of tween models and ‘reality.’ In delineating the borders
the ocean inevitably increases the uncertainty of the of the known, the unknown, and the unknowable, it
results. Moreover, even were we given that exact is important to discriminate between deduction and
state at some instant in time, the intrinsic and rationalization as competing processes for exploring
spontaneous genesis of disorder in such a physical and explaining those borders.
system must forever constrain our predictive power.
There are two key mathematical features of the • Although GCM simulations with a viscosity ap-
proaching that of water are at present inconceiv-
basic momentum (or vorticity) equation which bear
able, at least as a thought experiment it is worth
comment: conservation and dissipation. The non-
bearing in mind that those are the numerical re-
linear term is fundamental to the initiation and sus-
sults we would in principle compare against ob-
tenance of turbulence and by itself strictly conserves
servation to assess a given model.6 Short of that,
energy. (Other terms introduce explicit dissipation.)
GCMs use various formulations that ostensibly
In addition, in two dimensions the term conserves
mimic the dynamical effects of the unresolved
‘enstrophy’ (the square of the vorticity) and in three,
scales of motions. In the simplest instance, this
both ‘circulation’ and ‘helicity’ (the dot-product of
amounts to choosing a numerical viscosity several
velocity and vorticity). While one might hope for all
orders of magnitude larger than that of sea water.
such conservation properties to be preserved in nu-
But often the value is dictated by purely heuristic
merical implementations, some large models, often
numerical considerations: it is set at a threshold
those based on curvilinear – as opposed to Cartesian
value, any decrease below which leads to rapid
– coördinates – manage only to conserve energy.
Beyond the conserved quantities associated with
the nonlinear term, which include the energy and, in
4
general, the so-called ‘Noether invariants,’ there is a On dissipation, a deep, though perhaps insufficiently appreci-
ated, mathematical result is that the solution of a parabolic, dis-
more subtle property associated with the exact
sipative system quickly collapses onto a finite-dimensional
(continuous) equation: its associated ‘multi- ‘attractor’. This is remarkable. If you think of assigning a point to
symplectic geometry.’ Recent mathematical advances every one of N molecules of water in the ocean, and tracking the
make it possible for a discrete numerical model to velocity and position of each, the associated ‘phase space’ – just a
preserve such structure exactly, though as yet such record of that evolution – has dimension 6N. Because the mo-
mentum equation is derived on the basis that water is an infinitely
improvements have not been incorporated into any
divisible continuum, strictly we need to imagine that N approaches
working GCM. Is it quantitatively important that we infinity. Nonetheless, even in that infinite-dimensional phase space,
do so when basic fluid processes, such as convection, it remains true that the solution confines itself to only a finite, if
are as yet only crudely modeled? Until the experi- quite large, portion.
5
ment is tried, no one can say. But it is pertinent to Hadamard introduced the notion of ill-posedness of partial
note that a similar (that is, ‘symplectic’) refinement is differential equations. A problem is well-posed when a solution
exists, is unique, and depends continuously on the initial data. It is
critical for a numerical solution of sufficient accuracy
ill-posed when it fails to satisfy one or more of these criteria.
that one can decide whether planetary orbital mo- 6
There is a curious division among physical oceanographers as
tions are chaotic on astronomical timescales.4 to whether the large-scale flow we observe is, in the end, actually
Finally, strictly speaking, not only ambitious sensitive to the precise value of the viscosity of sea water. Pre-
models, but even more confined ‘process’ models, rest dictably, there are two camps: yes and no.
 GENERAL CIRCULATION MODELS 25

numerical blowup. It cannot be said to be satis- the outset, theorists generally do not have better al-
factory feature that a basic parameter is set not by ternatives to suggest. But one should always bear in
independent dynamical considerations but for mind the degree to which numerical simulations are
stability reasons, and those pertaining solely to the sensitive to these components, and seek independent
discrete form of the equations. (The solution to ways in which to constrain their parameters, in
the continuous equation would not blow up!) At isolated settings that test the limits of prediction
times, not only the coefficient, but the actual form against known measurements or, failing that, at least
of the diffusive operator is adjusted. As above, against fully converged, adequately resolved simu-
usually the motivation is intrinsically numerical, lations of a local or regional character. If, within the
so it is not surprising that a catalogue of the acceptable parameter range identified, it is found
various choices shows some, for example, that that the original model no longer gives adequate
create artificial sources (or sinks) of vorticity in large-scale predictions, then there are more basic
the flow. Others, subject to the given boundary problems to be addressed.
conditions, do not make mathematical sense in a
region where the fluid depth tends to zero (like
Atlantic City).
Summary
Oceanographers, unfortunately, do not have the From the numerical side, no computer improvements
luxury of extensive laboratory measurements from that can be seen on the horizon seem likely to make
which their dissipative parameterizations can be reasonably ambitious GCMs accessible to rigorous
calibrated. A program of direct observation in par- and extensive parametric and numerical exploration,
ticular regions where dissipation is thought to be a prerequisite to their complete understanding. From
significant is just getting under way. While such the mathematical side, it seems to be our funda-
measurements will help reveal deficiencies in present mental ignorance about turbulence that most se-
formulations, the largest GCMs will probably rely verely restricts the range of our grasp, leaving us with
on a solely heuristic approach for some time to come. an often painfully narrow range of computations to
which theoretical remarks can be significantly ad-
• An important, but numerically unresolved, pro-
dressed. For these structural reasons, the gulf be-
cess in the ocean is that of convection, which
tween theory and much numerical modeling will
typically occurs at small scales in, for example,
probably continue to widen for the foreseeable fu-
localized regions of intense surface cooling. The
ture, and thus there may grow to be—indeed some
overall thermal structure of the ocean is sensitive
would say it already exists—a division akin to C.P.
to this, a means by which ‘bottom water’ is
Snow’s ‘Two Cultures’.
formed; a cold, relatively less salty mass that
All the cautions about GCMs notwithstanding,
constitutes the deep Atlantic, for example. Be-
they have become an integral part of the study of
cause the horizontal resolution is too coarse to
physical oceanography. With due regard for the
encompass the sinking motions, various schemes
novel capacities and limitations of numerical models,
have been devised to mimic that effect. It has been
such scientific progress as we do make will more and
shown that one of the most common of these leads
more often hinge upon judicious computation.
paradoxically to unacceptable physical (and
mathematical) behavior as resolution is improved;
it has no verifiable correspondence to a realizable
Further Reading
physical process. The temptation with a model
that has been extensively tuned to give plausible
answers for other observables is to leave well The literature on ocean modeling is not yet productive of
enough alone. Unhappily, a model with one or definitive treatises, in large measure because the field is yet
young and rapidly evolving. Thus in lieu of textbooks or
more such elements whose limits are ill-defined or
similar references, the reader is directed to the following
nonexistent must inevitably produce end results series of articles.
whose errors are typically an opaque mix of ef- For some predictions on the perennially intriguing issue
fects: some physical, some numerical, some of what improvements in large-scale modeling may be
mathematical. In such circumstances, the program driven by plausible increases in computing speed with
of falsifiability of the physical components is apt massively parallel machines see
to be fatally compromised. Semtner A. Ocean and climate modeling Communica-
tions of the ACM 43 (43): 81–89.
The point is not that one should immediately dis- For a look back at the history of one of the single most
pense with all ad hoc parametrizations; excepting influential models in physical oceanography, see A.J.
those that are simply mathematically ill-posed from Semtner’s Introduction to ‘A numerical method for the
26 GENERAL CIRCULATION MODELS

study of the circulation of the World Ocean’, which Finally, for those readers desiring a more in depth
accompanies the reprinting of Kirk Bryan’s now classic appreciation of modeling issues and their implications for
1969 article of the title indicated. This pair appears back-to- specific features of the large scale circulation, consult the
back, beginning on page 149, in Journal of Computational careful review.
Physics, (1997) 135 (2). McWilliams JC (1996) Modeling the oceanic general
General readers may wish to consult the following circulation. In: Lumley JL, Van Dyke M. Read HL (eds)
succinct review, accessible to a broad audience: Annual Review of Fluid Mechanics Vol. 28, pp. 215–248
Semtner AJ (1995) Modeling ocean circulation Science Palo Alto, CA: Annual Reviews.
269 (5229): 1379–1385.
 OCEAN CIRCULATION: MERIDIONAL
OVERTURNING CIRCULATION
J. R. Toggweiler, NOAA, Princeton, NJ, USA The Cooling Phase – Deep-water
& 2009 Elsevier Ltd. All rights reserved. Formation
The most vigorous overturning circulation in the
ocean today is in the Atlantic Ocean where the upper
part of the Atlantic’s MOC carries warm, upper
ocean water through the Tropics and subtropics
toward the north while the deep part carries cold
Introduction dense polar water southward around the tip of Africa
and into the Southern Ocean beyond. The Atlantic’s
The circulation of the ocean has been traditionally MOC converts roughly 15  106 m3 s  1 of upper
divided into two parts, a wind-driven circulation that ocean water into deep water. (Oceanographers des-
dominates in the upper few hundred meters, and a
ignate a flow rate of 1  106 m3 s  1 to be 1 Sv. All
density-driven circulation that dominates below. The
the world’s rivers combined deliver c. 1 Sv of fresh
latter was once called the ‘thermohaline circulation’,
water to the ocean.)
a designation that emphasized the roles of heating,
The MOC in the Atlantic is often characterized as
cooling, freshening, and salinification in its pro-
a ‘conveyor belt’ or more generally as a continuous
duction. Use of ‘thermohaline circulation’ has all but
current or ribbon of flow. It is shown following a
disappeared among professional oceanographers.
path that extends through the Florida Straits and up
The preferred designation is ‘meriodional over-
the east coast of North America as part of the Gulf
turning circulation’, hereafter MOC. This usage re-
Stream. The ribbon then cuts to the east across the
flects the sense of the time-averaged flow, which
Atlantic and then northward closer to the coast of
generally consists of a poleward flow of relatively
warm water that overlies an equatorward flow of Europe. This is somewhat misleading because indi-
colder water at depth. It also reflects a recognition by vidual water parcels do not follow this kind of con-
oceanographers that most of the work done to drive tinuous path. The MOC is composed instead of
the circulation, whether near the surface or at depth, multiple currents that transfer water and water
comes directly or indirectly from the wind. properties along segments of the path. Individual
Like the thermohaline circulation of old, the MOC water parcels loop around multiple times and may
is an important factor in the Earth’s climate because recirculate all the way around the wind-driven gyres
it transports roughly 1015 W of heat poleward into while moving northward in stages.
high latitudes, about one-fourth of the total heat As the MOC moves northward through the tropical
transport of the ocean/atmosphere circulation sys- and subtropical North Atlantic, it spans a depth range
tem. Radiocarbon measurements show that it turns from the surface down to B800 m and has a mean
over all the deep water in the ocean every 600 years temperature of some 15–20 1C. During its transit
or so. Its upwelling branch is important for the through the Tropics and subtropics, the MOC be-
ocean’s biota as it brings nutrient-rich deep water up comes saltier due to the excess of evaporation over
to the surface. It may or may not be vulnerable to the precipitation in this region. It also becomes warmer
warming and freshening of the Earth’s polar regions and saltier by mixing with the salty outflow from
associated with global warming. the Mediterranean Sea. By the time the MOC has
The most distinguishing features of the MOC are crossed the 501 N parallel into the subpolar North
observed in the sinking phase, when new deep-water Atlantic the water has cooled to an average tem-
masses are formed and sink into the interior or to the perature of 11–12 1C. Roughly half of the water car-
bottom. Large volumes of cold polar water can be ried northward by the MOC at this juncture moves
observed spilling over sills, mixing violently with into the Norwegian Sea between Iceland and Norway.
warmer ambient water, and otherwise descending to Part of this flow extends into the Arctic Ocean.
abyssal depths. The main features of the upwelling The final stages of this process make the salty
phase are less obvious. The biggest uncertainty is about North Atlantic water cold and dense enough to
where the upwelling occurs and how the upwelled deep sink. Sinking is known to occur in three main places.
water returns to the areas of deep-water formation. The densest sinking water in the North Atlantic is

27
28 OCEAN CIRCULATION: MERIDIONAL OVERTURNING CIRCULATION

formed in the Barents Sea north of Norway where Newly formed water masses are easy to track by
salty water from the Norwegian Current is exposed their distinct temperature and salinity signatures and
to the atmosphere on the shallow ice-free Barents high concentrations of oxygen. The southward flow
shelf. Roughly 2 Sv of water from the Norwegian of North Atlantic Deep Water (NADW) is a prime
Current crosses the shelf and sinks into the Arctic example. NADW is identified as a water mass with
basin after being cooled down to 0 1C. This water a narrow spread of temperatures and salinities
eventually flows out of the Arctic along the coast between 2.0 and 3.5 1C and 34.9 and 35.0 psu,
of Greenland at a depth of 600–1000 m. The volume respectively.
of dense water is increased by additional sinking and Figure 1 shows the distribution of salinity in the
open ocean convection in the Greenland Sea north of western Atlantic which tracks the southward move-
Iceland. The dense water then spills into the North ment of NADW. Newly formed NADW (S434.9) is
Atlantic over the sill between Greenland and Iceland easily distinguished from the relatively fresh inter-
in Denmark Strait (at about 600-m depth) and mediate water above (So34.6) and the Antarctic
through the Faroe Bank Channel between the Faroe water below (So34.7). NADW exits the Atlantic
Islands and Scotland (at about 800-m depth). south of Africa between 351 and 451 S and joins the
As 0 1C water from the Arctic and Greenland Seas eastward flow of the Antarctic Circumpolar Current
passes over these sills, it mixes with 6 1C Atlantic (ACC). Traces of NADW reenter the Atlantic Ocean
water beyond the sills and descends the continental through Drake Passage (55–651 S) after flowing all
slopes down to a depth of 3000 m. The overflows the way around the globe.
merge and flow southward around the tip of The other major site of deep-water formation is
Greenland into the Labrador Sea as part of a deep the coast of Antarctica. The surface waters around
boundary current that follows the perimeter of the Antarctica, like those over most of the Arctic Ocean,
subpolar North Atlantic. A slightly warmer water are ice covered during winter and are too fresh to
mass is formed by open ocean convection within sink. Deep water below 500 m around Antarctica, on
the Labrador Sea. The deep water formed in the the other hand, is relatively warm (1.5 1C) and fairly
Labrador Sea increases the volume flow of the salty (34.70–34.75 psu). This water mass, known as
boundary current that exits the subpolar North Circumpolar Deep Water (CDW), penetrates onto
Atlantic beyond the eastern tip of Newfoundland. the relatively deep continental shelves around

0 37.0
34.2 36.0
35.0 35.0
34.6 34.9
34.4
35.0
34.6 34.9
34.7
2000

34.8
Depth (m)

34.9

4000 34.7 34.8

34.9

6000

60° S 40° 20° Eq 20° 40° 60° N

Latitude

Figure 1 North–south section of salinity down the western Atlantic from Iceland to Drake Passage. The salinity distribution has been
contoured every 0.1 psu between 34.0 and 35.0 psu to highlight NADW (34.9–35.0 psu).
 OCEAN CIRCULATION: MERIDIONAL OVERTURNING CIRCULATION 29

Antarctica where it is cooled to the freezing point. places but is not widespread. Attempts to directly
Brine rejection from the formation of new sea ice measure the turbulent mixing in the interior have
maintains fairly high salinities on the shelf despite shown that there may only be enough mixing to
the freshening effects of precipitation and the input modify perhaps 10% or 20% of the deep water
of glacial meltwater. Very cold shelf water (–1 1C, formed near the poles. There is no indication that
34.6–34.7 psu) is then observed descending the any deep water is actually upwelling to the surface in
continental slope to the bottom in the Weddell Sea. the warm parts of the Indian and Pacific Oceans.
Bottom water is also observed to form off the Adelie Vigorous mixing is found near the bottom, where it
coast south of Australia (B1401 E). is generated by tidal motions interacting with the
The volume of new deep water formed on the bottom topography. It is also found in the vicinity of
Antarctic shelves is not very large in relation to the strong wind-driven currents. Thus, the energy that is
volume of deep water formed in the North Atlantic, available to warm the deep waters of the abyss comes
perhaps 3–4 Sv in total. It however entrains a large mainly from the winds and tides.
volume of old Circumpolar Deep Water, as it sinks to It now appears that most of the deep water sinking
the abyss. Even with the entrainment, the deep water in the North Atlantic upwells back to the surface in
formed around Antarctica is denser than the NADW the Southern Ocean. The upwelling occurs within the
formed in the north. Antarctic Bottom Water channel of open water that circles the globe around
(AABW) occupies the deepest parts of the ocean and Antarctica. Figure 2 shows schematically how this is
is observed penetrating northward into the Atlantic, thought to work. The curved lines in the background
Indian, and Pacific Oceans through deep passages in are isolines of constant density (also known as iso-
the mid-ocean ridge system. pycnals). Their downward plunge to the north away
Small quantities of deep water are also observed to from Antarctica reflects the flow of the ACC out
form in semi-enclosed evaporative seas like the of the page in the center of the figure. Salty dense
Mediterranean and Red Seas. These water masses are water from the North Atlantic is found at the base of
dense owing to their high salinities. They tend to the plunging isopycnals. Westerly winds above the
form intermediate- depth water masses after exiting ACC (also blowing out of the page) drive the ACC
their formation areas in relation to the deep-water forward and also push the cold fresh water at the
and bottom-water masses formed in the North surface away from Antarctica to the north. Dense
Atlantic and around Antarctica. salty water from the North Atlantic is drawn upward
in its place. A mixture of the salty dense deep water
The Warming Phase – Upwelling and and the cold fresh surface water is then driven
northward out of the channel by the westerly winds
the Return Flow
and is forced down into the thermocline on the north
The upwelling of deep water back to the ocean’s side of the ACC.
surface was thought at one time to be widely dis- In this way, the winds driving the ACC continually
tributed over the ocean. This variety of upwelling remove water with North Atlantic properties from
was attributed to mixing processes that were hypo- the interior. Oceanographers are fairly certain that
thesized to be active throughout the interior. The something like this is happening because the winds
mixing was thought to be slowly transfering heat driving the ACC have become stronger over the last
downward, making the old deep water in the interior 40 years in response to global warming and the de-
progressively less dense so that it could be displaced pletion of ozone over Antarctica. The subsurface
upward by the colder and saltier deep waters form- water around Antarctica has become warmer, saltier,
ing near the poles. Since the main areas of deep water and lower in oxygen as upwelled water from the
formation are located at either end of the Atlantic, interior has displaced more of the cold fresh surface
and since most of the ocean’s area is found in the water, as shown in Figure 2. The thermocline water
Indian and Pacific Oceans, it stood to reason that the north of the ACC has also become cooler and fresher.
warming of the return flow should be widely dis- Numerical experiments with ocean general circu-
tributed across the Indian and Pacific. Schematic lation models show that much of the water forced
diagrams often depict the closure of the old thermo- down into the thermocline around the open channel
haline circulation as a flow of warm upper ocean eventually makes its way into the Atlantic Ocean
water that passes from the North Pacific through where it is converted again into deep water in the
Indonesia, across the Indian Ocean, and then around northern North Atlantic. Thus, the winds over the
the tip of Africa into the Atlantic. ACC, in drawing up deep water from the ocean’s
Observations made over the last 30 years point interior, have been shown to actively enhance the
instead to turbulent mixing that is intense in some formation of deep water in the North Atlantic.
30 OCEAN CIRCULATION: MERIDIONAL OVERTURNING CIRCULATION

S Westerly winds N
0m
Cold, fresh
Warmer,
saltier Lower
oxygen Cooler, fresher
Dense

Low oxygen
Antarctica
Salty, dense
water

2500 m
60° S 50° S 40° S

Figure 2 Schematic cross section of the ACC, showing how the winds driving the current in an eastward direction out of the page
also drive an overturning circulation. Stronger winds over the last 40 years have drawn more deep water to the surface south of the
ACC, which has caused the subsurface waters around Antarctica to become warmer, saltier, and lower in oxygen, and have produced
more downwelling to the north, which has made the thermocline waters north of the ACC cooler and fresher. Adapted from Aoki S,
Bindoff NL, and Church JA (2005) Interdecadal water mass changes in the Southern Ocean between 30 1 E and 160 1 E. Geophysical
Research Letters 32 (doi:10.1029/2005GL022220) and reproduced from Toggweiler RR and Russell J (2008) Ocean Circulation in a
warming climate. Nature 451 (doi:10.1038/nature06590).

The deep water drawn up to the surface around General Theory for the Meridional
Antarctica is colder than the surface water that Overturning Circulation
is forced down into the thermocline north of the
ACC. As this cold water comes into contact with the Figure 3 is a north–south section showing the dis-
atmosphere and is carried northward, it takes up tribution of potential density through the Atlantic
solar heat that otherwise would be available to warm Ocean. Most of the northward flow of the Atlantic’s
the Southern Ocean and Antarctica. The MOC MOC takes place between the 34.0 (B20 1C) and
then carries this southern heat across the equator 36.0 (B8 1C) isopycnals, that is, within the main
into the high latitudes of the North Atlantic where part of the thermocline, but some of the northward
it is released to the atmosphere when new deep water flow takes place among the denser isopycnals of the
is formed. The MOC thereby warms the North lower thermocline down to 36.6 g kg  1 (B1000 m).
Atlantic at the expense of a colder Southern The southward flow of NADW is located, for the
Ocean and colder Antarctica. Indeed, sea surface most part, below the 37.0 isopycnal.
temperatures at 601 N in the North Atlantic are on The isopycnals in the lower thermocline in Figure 3
average c. 6 1C warmer than sea surface temperatures are basically flat north of 401 S but rise up to the
at 601 S. surface between 401 and 601 S. The rise of the iso-
If the warming phase of the MOC involved a pycnals marks the eastward flow of the ACC, as
downward mixing of heat in low and middle shown previously in Figure 2. The relatively deep
latitudes, as once thought, the MOC would carry position of these isopycnals north of 401 S reflects the
tropical heat poleward into high latitudes. The mechanical work done by the winds that drive the
warming phase of the Atlantic’s MOC now seems ACC. The convergent surface flow forced by the winds
to take place in the south instead. This means that north of the ACC pushes the relatively light lower
the heat transport by the MOC through the thermocline water down in relation to the cold, dense
South Atlantic is directed equatorward and is op- water that is drawn up by the winds south of the ACC.
posed to the heat transport in the atmosphere. Numerical experiments carried out in ocean global
The net effect is a weakening of the global heat climate models (GCMs) suggest that all the isopycnals
transport in the Southern Hemisphere and a of the lower thermocline would be squeezed up into
strengthening of the heat transport in the Northern the main thermocline if the circumpolar channel were
Hemisphere. closed and the ACC eliminated.
 OCEAN CIRCULATION: MERIDIONAL OVERTURNING CIRCULATION 31

35.0 35.0 36.0

0 35.0
34.0
37.0 36.0 35.0 37.0
35.2 35.5
35.4
36.6
36.8

2000
37.0
Depth (m)

37.2

4000

6000

60° S 40° 20° Eq 20° 40° 60° 80° N

Latitude

Figure 3 Density structure of the Atlantic thermocline. Seawater density is referenced to a depth of 2000 m and has been zonally
averaged across the Atlantic basin (units g kg  1). Contours have been chosen to highlight the lower part of the thermocline. The
zonally averaged topography fails to capture the depths of the sills through the Greenland–Iceland–Scotland Ridge (651 N), which are
at 600 and 800 m.

The relatively deep lower thermocline puts rela- transformations in the north and south that convert
tively warm water just to the south of the sills be- light water to dense water and vice versa.
tween Greenland, Iceland, and the Faroe Islands in A different kind of theory is needed for the over-
the northern North Atlantic. This warm water is turning of the bottom water formed around
important because it sets up a sharp contrast with the Antarctica because the winds cannot play the same
cold water behind the sills, which ultimately drives role. The upwelling branch of this circulation is also
the flow of dense water over the sills and into the associated with the ACC as in Figure 2 but the
deep Atlantic. In this way, the winds and the ACC in upwelled water in this case flows poleward onto the
the south contribute as much, if not more, to the adjacent Antarctic shelves where it is cooled and
overflows in the north as the cooling that takes place sinks back into the interior. The winds in this case
at the surface. help cool the upwelled water by exposing it to the
The depths of the isopycnals in the lower ther- atmosphere and by driving the fresh Antarctic sur-
mocline lead one to a general theory for the MOC in face waters away but the strength of the bottom
the Atlantic. Deep-water formation in the North water circulation would appear to be limited by the
Atlantic converts relatively low-density thermocline rate at which old bottom waters can be warmed by
water into new deep water. It thereby removes mass mixing with the overlying deep water.
from the thermocline and allows the isopycnals of
the thermocline to be squeezed upward. The winds in
the south have the opposite effect as they draw deep Instability of the Meridional
water up to the surface and pump it northward into Overturning Circulation
the thermocline. They convert dense water from the
deep ocean into low-density thermocline water and Cooling of the ocean in high latitudes makes polar
cause the thermocline to thicken downward. In this surface waters denser in relation to warmer waters
way, the thermocline thickness reflects a balance at lower latitudes. Thus cooling contributes to a
between the addition of mass via winds in the south stronger MOC. The salinity section through the
and the loss of mass by deep-water formation in the Atlantic Ocean in Figure 1 gives one a superficial
north. The strength of the MOC should in this sense impression that salinification also makes a positive
be proportional to the thermocline thickness and the contribution to the MOC. This is actually not true.
32 OCEAN CIRCULATION: MERIDIONAL OVERTURNING CIRCULATION

The cycling of fresh water between the ocean and 30 1 E and 160 1 E. Geophysical Research Letters 32:
atmosphere (the hydrological cycle) results in a net (doi:10.1029/2005GL022220).
addition of fresh water to the polar oceans Broecker WS (1991) The great ocean conveyor.
which reduces the density of polar surface waters. Oceanography 4: 79--89.
Thus, the haline part of the old thermohaline Gnanadesikan A (1999) A simple predictive model for the
structure of the oceanic pycnocline. Science 283:
circulation is nearly always in opposition to the
2077--2079.
thermal forcing. The Earth’s hydrological cycle is Kuhlbrodt T, Griesel A, Montoya M, Levermann A,
expected to become more vigorous in the future Hofmann M, and Rahmstorf S (2007) On the driving
with global warming. This may weaken the processes of the Atlantic meridional overturning
MOC in a way which could be fairly abrupt and circulation. Reviews of Geophysics 45: RG2001
unpredictable. (doi:10.1029/2004RG000166).
NADW is salty because the upper part of the Manabe S and Stouffer R (1993) Century-scale effects of
MOC flows through zones of intense evaporation in increased atmospheric CO2 on the ocean–atmosphere
the Tropics and subtropics. If the rate at which new system. Nature 364: 215--218.
deep water is forming is relatively high, as it seems to McCartney MS and Talley LD (1984) Warm-to-cold
be at the present time, the sinking water removes water conversion in the northern North Atlantic
Ocean. Journal of Physical Oceanography 14:
much of the fresh water added in high latitudes and
922--935.
carries it into the interior. The added fresh water in Munk W and Wunsch C (1998) Abyssal recipes II:
this case dilutes the salty water being carried into the Energetics of tidal and wind mixing. Deep-Sea Research I
deep-water formation areas but does not erase the 45: 1977--2010.
effect of evaporation in low latitudes. Thus, NADW Orsi A, Johnson G, and Bullister J (1999) Circulation,
remains fairly salty and is able to export fresh water mixing, and production of Antarctic bottom water.
from the Atlantic basin. Progress in Oceanography 43: 55--109.
If the rate of deep-water formation is relatively low Rahmstorf S (1995) Bifurcations of the Atlantic
or the hydrological cycle is fairly strong, the fresh thermohaline circulation in response to changes in the
water added in high latitudes can create a low-salinity hydrological cycle. Nature 378: 145--149.
lid that can reduce or annihilate the MOC. There Rudels B, Jones EP, Anderson LG, and Kattner G (1994)
On the intermediate depth waters of the Arctic Ocean.
seems to be a critical input of freshwater input for
In: Johannessen OM, Muench RD, and Overland JE
maintaining the MOC. If the freshwater input is close (eds.) Geophysical Monograph 85: The Polar Oceans
to this threshold, the overturning becomes unstable and Their Role in Shaping the Global Environment,
and may become prone to wild swings over time. pp. 33--46. Washington, DC: American Geophysical
Coupled (ocean þ atmosphere) climate models Union.
from the 1990s projected that a fourfold increase in Russell JL, Dixon KW, Gnanadesikan A, Stouffer RJ, and
atmospheric CO2 would increase the hydrological Toggweiler JR (2006) The Southern Hemisphere
cycle sufficiently that the MOC might collapse. More westerlies in a warming world: Propping open the
recent coupled models are predicting that the warm- door to the deep ocean. Journal of Climate 19(24):
ing will, in addition, lead to stronger mid-latitude 6382--6390.
westerly winds that are shifted poleward with respect Schmitz WJ and McCartney MS (1993) On the North
Atlantic Circulation. Reviews of Geophysics 31: 29--49.
to their preindustrial position. This change in the
Toggweiler JR and Samuels B (1995) Effect of
westerlies puts stronger Southern Hemisphere west- Drake Passage on the global thermohaline circulation.
erlies directly over the ACC, which should make the Deep-Sea Research I 42: 477--500.
MOC stronger. Thus, the MOC could become Toggweiler JR and Samuels B (1998) On the ocean’s large-
stronger or weaker depending on whether the winds scale circulation near the limit of no vertical mixing.
or the hydrological cycle dominate in the future. Journal of Physical Oceanography 28: 1832--1852.
Toggweiler RR and Russell J (2008) Ocean Circulation in a
warming climate. Nature 451: (doi:10.1038/nature
See also 06590).
Tziperman E (2000) Proximity of the present-day
Antarctic Circumpolar Current. Ocean Circulation.
thermohaline circulation to an instability threshold.
Further Reading Journal of Physical Oceanography 30: 90--104.
Wunsch C and Ferrari R (2004) Vertical mixing, energy,
Aoki S, Bindoff NL, and Church JA (2005) Interdecadal and the general circulation of the oceans. Annual
water mass changes in the Southern Ocean between Reviews of Fluid Mechanics 36: 281--314.
 NORTH ATLANTIC OSCILLATION (NAO)
J. W. Hurrell, National Center for Atmospheric especially in the context of global climate change.
Research, Boulder, CO, USA This article defines the NAO and describes its rela-
Copyright & 2001 Elsevier Ltd. tionship to variations in surface temperature and
precipitation, as well as its impact on variability in
the North Atlantic Ocean and on the regional ecol-
ogy. It concludes with a discussion of the mech-
anisms that might influence the amplitude and
timescales of the NAO, including the possible roles
Introduction of the stratosphere and the ocean.
Simultaneous variations in weather and climate over
widely separated points on Earth have long been
noted in the meteorological literature. Such vari- What is the North Atlantic Oscillation?
ations are commonly referred to as ‘teleconnections’.
Like all atmospheric teleconnection patterns, the
In the extratropics, teleconnections link neighboring
NAO is most clearly identified when time averaged
regions mainly through the transient behavior of
data (monthly or seasonal) are examined, since time
atmospheric planetary-scale waves. Consequently,
averaging reduces the ‘noise’ of small-scale and
some regions may be cooler than average, while
transient meteorological phenomena not related to
thousands of kilometers away warmer conditions
large-scale climate variability. Its spatial signature
prevail. Though the precise nature and shape of these
and temporal variability are most often defined
structures vary to some extent according to the
through the regional sea level pressure field, for
statistical methodology and the data set employed in
which some of the longest instrumental records exist.
the analysis, consistent regional characteristics that
The NAO refers to a north–south oscillation in
identify the most conspicuous patterns emerge.
atmospheric mass with centers of action near Iceland
Over the middle and high latitudes of the northern
and over the subtropical Atlantic from the Azores
hemisphere, a dozen or so distinct teleconnection
across the Iberian Peninsula. Although it is the only
patterns can be identified during boreal winter. One of
teleconnection pattern evident throughout the year in
the most prominent is the North Atlantic Oscillation
the northern hemisphere, its amplitude is largest
(NAO). The NAO dictates climate variability from
during boreal winter when the atmosphere is dy-
the eastern seaboard of the USA to Siberia and from
namically the most active. During the months De-
the Arctic to the subtropical Atlantic. This wide-
cember through March, for instance, the NAO
spread influence indicates that the NAO is more than
accounts for more than one-third of the total vari-
just a North Atlantic phenomenon. In fact, it has been
ance in sea level pressure over the North Atlantic.
suggested that the NAO is the regional manifestation
A time series (or index) of more than 100 years of
of a larger scale (hemispheric) mode of variability
wintertime NAO variability and the spatial signature
known as the Arctic Oscillation (see below). Re-
of the oscillation are shown in Figures 1 and 21
gardless of terminology, meteorologists for more than
Differences of 415 hPa occur across the North At-
two centuries have noted the pronounced influence of
lantic between the two phases of the NAO. In the so-
the NAO on the climate of the Atlantic basin.
called positive phase, higher than normal surface
Variations in the NAO are important to society
pressures south of 551N combine with a broad re-
and the environment. Through its control over
gion of anomalously low pressure throughout the
regional temperature and precipitation variability,
Arctic. Because air flows counterclockwise around
the NAO directly impacts agricultural yields, water
low pressure and clockwise around high pressure in
management activities, and fish inventories among
the northern hemisphere, this phase of the oscillation
other things. The NAO accounts for much of the
is associated with stronger than average westerly
interannual and longer-term variability evident in
winds across the middle latitudes of the Atlantic onto
northern hemisphere surface temperature, which has
exhibited a warming trend over the past several
decades to values that are perhaps unprecedented 1
More sophisticated and objective statistical techniques, such as
over the past 1000 years. eigenvector analysis, yield time series and spatial patterns of
Understanding the processes that govern vari- average winter sea level pressure variability very similar to those
ability of the NAO is therefore of high priority, shown in Figures 1 and 2.

33
34 NORTH ATLANTIC OSCILLATION (NAO)

2
(Ln _ Sn)

_2

_4

_6
1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

Figure 1 Winter (December–March) index of the NAO based on the difference of normalized sea level pressure between Lisbon,
Portugal, and Stykkisholmur/Reykjavik, Iceland from 1864 to 2000. The average winter sea level pressure data at each station were
normalized by division of each seasonal pressure by the long-term mean (1864–1983) standard deviation. The heavy solid line
represents the index smoothed to remove fluctuations with periods o4 years.

SLP December _ March (hPa)

150˚W 150˚E

120˚W 120˚E

0
_2
0
90˚W 90˚E
0
_2
2

_2
_6
0
60˚W 2
2 60˚E
4

4
2
30˚W 30˚E

0˚

Figure 2 Difference in sea level pressure between years with an NAO index value 41.0 and those with an index value o  1.0 (high
minus low index winters) since 1899. The contour increment is 2 hPa and negative values are dashed.

Europe, with anomalous southerly flow over the The NAO is also readily apparent in meteoro-
eastern USA and anomalous northerly flow across logical data throughout the depth of the troposphere,
western Greenland, the Canadian Arctic, and the and its variability is significantly correlated with
Mediterranean. changes in the strength of the winter polar vortex in
Other documents randomly have
different content
have none of you while I go my own way to this."

Mr. Puddlebox took Mr. Wriford to the gate of the grounds, then went
back again in much trembling. An open window informed him of Mr.
Wriford's place of entry. He leant through to a sofa that stood handy, there
deposited the trinkets, and very softly shut the window down. When he
rejoined Mr. Wriford, fear's perspiration was streaming from him. "I've had
a squeak of it," said Mr. Puddlebox with simulated cheeriness. "Let's out of
this, and I'll tell you."

He walked Mr. Wriford long, quickly and far. While he walked he fought
again the battle that had been swift victory when he cast his bottle from
him; and in future days fought it again and met new tortures in each fight.

"Aren't you going to get any whisky?" asked Mr. Wriford when on a day,
pockets lined with harvest money, he noticed Mr. Puddlebox's abstinence.

"Whisky! Hell take such stinking stuff," cried Mr. Puddlebox and sucked
in his cheeks—and groaned; then put a hand in his tail-pocket and felt a
hard lump rolled in a cloth that lay where the whisky used to lie and said to
himself: "Two bottles—two bottles."

It was Mr. Puddlebox's promise to himself, and his lustiest weapon in his
battles with his desire, that, on some day that must come somehow, the day
when he should be relieved of his charge of Mr. Wriford, he would buy
himself two bottles of whisky and sit himself down and drink them. Into the
hard lump rolled in the cloth, and composing it, there went daily when his
earnings permitted it two coppers. When that sum reached eighty-four—two
at three-and-six apiece—his two bottles would be ready for the mere
asking.

Wherefore "Two bottles! Two bottles!" Mr. Puddlebox would assure

himself when most fiercely his cravings assailed him, and against the pangs
of his denial would combine luxurious thoughts of when they should thus
be slaked and fears of what might happen to his loony if he now gave way
to them.
 Much those fears—or the affection whence they rose—cost him in these
later days: swiftly their end approached. Much and more as summer passed
and autumn came sombrely and chill: swiftly their end as sombre day
succeeded sombre day, and they passed down into Cornwall and went along
the sombre sea. Village to village, through nature in decay that grey sky
shrouded, grey sea dirged: Mr. Puddlebox ever for tarrying when larger
town was reached, Mr. Wriford ever for onward—onward, on.

CHAPTER II

CROSS WORK

Ever for onward, Mr. Wriford—onward, onward, on!

Where, in the bright days, Mr. Puddlebox had taken the lead and
suggested their road and programme, now, in the sombre days, chill in the
air, and in the wind a bluster, Mr. Wriford led. He chose the roughest paths.
He most preferred the cliff tracks where wind and rain drove strongest, or
down upon the shingle where walking was mostly climbing the great
boulders that ran from cliff to sea. He walked with head up as though to
show the weather how he scorned it. He walked very fast as though there
was something he pursued.

Mr. Puddlebox did not like it at all. Much of Mr. Puddlebox's jolly
humour was shaken out of him in these rough and arduous scrambles, and
he grumbled loud and frequent. But very fond of his loony, Mr. Puddlebox,
and increasingly anxious for him in this fiercer mood of his.

There are limits, though: and these came on an afternoon wild and wet
when Mr. Wriford exchanged the cliff road for the shore and pressed his
way at his relentless pace along a desolate stretch cut into frequent inlets by
rocky barriers that must be toilsomely climbed, a dun sea roaring at them.
 "Why, what to the devil is it you're chasing, boy?" Mr. Puddlebox's
grumblings at last broke out, when yet another barrier surmounted revealed
another and a steeper little beyond. "Here's a warm town we've left," cried
Mr. Puddlebox, sinking upon a great stone, "and here's as wet, cold, and
infernal a climbing as I challenge you or any man ever to have seen. Here's
you been dragging and trailing and ripe for anything these three months and
more, and now rushing and stopping for nothing so I challenge the devil
himself to keep up with you."

"Well, don't keep up!" said Mr. Wriford fiercely. "Who wants you to?"

Mr. Puddlebox blinked at that; but he answered stoutly: "Well, curse me

if I do, for one."

"Nor me for another," said Mr. Wriford and turned where he stood and
pressed on across the shingle towards the next rocky arm.

Mr. Puddlebox sucked in his cheeks, felt at the hard lump in his pocket,
then followed at a little run, and caught Mr. Wriford as Mr. Wriford climbed
the further barrier of rocks.

"Hey, give us a hand, boy," cried Mr. Puddlebox cheerfully. "This is a

steep one."

Mr. Wriford looked down. "What, are you coming on? I thought you'd
stopped."

"You're unkind, boy," said Mr. Puddlebox.

Mr. Wriford, looking down, this time saw the blink that went with the
words. He jumped back lower, coming with reckless bounds. "I'm sorry," he
said. "I'm sorry. Look here, coming across this bit"—he pointed back to
their earlier stopping-place—"I felt—I felt rotten to think you'd gone."

"Why, that's my loony!" cried Mr. Puddlebox, highly pleased. "Come

down here, boy. Let's talk of this business."
 "But I wouldn't look back," said Mr. Wriford, "or come back. I've done
with that sort of thing."

"Why, so you have," said Mr. Puddlebox, rightly guessing to what Mr.
Wriford referred. "You can come down now, though, for I'm asking you to,
so there's no weakness in that. There's shelter here."

"I don't want shelter," said Mr. Wriford, and went a step higher and stood
with head and back erect where gale and rain caught him more full.

Mr. Puddlebox summoned much impressiveness into his voice. "Boy,"

said Mr. Puddlebox, "this is a fool's game, and I never saw such even with
you. Bring sense to it, boy. Tramping is well enough for fine days: winters
for towns. There's money to be found in towns, boy; and if no money,
workhouse is none so bad, and when we've tried it you've liked it and called
it something new, which is what you want. Well, there's nothing new this
way, boy. There's no work and there's no bed in the fields winter-time.
Nothing new this way, boy."

A fiercer drive of wind spun Mr. Wriford where he stood exposed. He

caught at a rock with his hands and laughed grimly, then stood erect again,
and pressed himself against the rising gale.

"Ah, isn't there, though?" he cried. "Man, there's cold and rain and wind,
and there's tramping on and on against it and feeling you don't care a damn
for it."

"Well, curse me, but I do," returned Mr. Puddlebox. "It's just what I do
mind, and there's no sense to it, boy. There's no sense to it."

"There is for me," Mr. Wriford cried. "It's what I want!" He turned from
fronting the gale. Mr. Puddlebox saw him measuring with his eye the height
where he stood from the ground, and called in swift alarm: "Don't jump!
You'll break your legs. Don't—"

Mr. Wriford laughed aloud, jumped and came crashing to his hands and
knees, got up and laughed again. "That's all right!" said he.
 "Boy, that's all wrong," said Mr. Puddlebox very seriously. "That's all of
a part with your rushing along as if it was the devil himself you chased; and
what to the devil else it can be I challenge you to say or any man."

Mr. Wriford took up the words he had cried down from the top of the
barrier. "It's what I want," he told Mr. Puddlebox. "Cold and not minding it,
and fighting against the wind and not minding it, and getting wet and going
on full speed however rough the road and not minding that. Cold and wind
and rain and sticking to it and fighting it and beating it and liking it—ah!"
and he threw up his arms, extending them, and filled his chest with a great
breath, as though he embraced and drunk deep of the elements that he stuck
to and fought and beat.

Mr. Puddlebox looked at him closely. "Sure you're liking it?" he asked,
his tone the same as when he often inquired: "Sure you're happy, boy?"

"Sure! Why, of course I'm sure. Why, all the time I'm thrashing along, do
you know what I'm saying? I'm saying: 'Beating you! Beating you! Beating
you!' and at night I lie awake and think of it all waiting outside for me and
how I shall beat it, beat it, beat it again when morning comes."

"Sit down," said Mr. Puddlebox. "I've something to say to you."

"No, I'll stand," said Mr. Wriford.

"Aren't you tired?"

"I'm fit to drop," said Mr. Wriford; and then with a hard face: "But sitting
down is giving way to it. I'll not do that. No, by God, I'll beat it all the
time."

Then Mr. Puddlebox broke out in exasperation and struck his stick upon
the shingle to mark it. "Why, curse me if I ever heard such a thing or knew
such a thing!" cried Mr. Puddlebox. "Beating it! I've told you a score time,
and this time I give it to you hot, that when you go so, you're spooked,
spooked to hell and never will be unspooked! 'Beating it, beating it, beating
it!' you cry as you rush along! Why, it's then that it is beating you all the
time, for it is of yourself that you are thinking. And that's what's wrong with
you, thinking of yourself, and has always been. And there's no being happy
that way and never will be. Think of some one else, boy. For God
Almighty's sake think of some one else or you're beat and mad for sure!"

Mr. Wriford gave him back his fierceness. "Think of some one else!
That's what I've done all my life. That's what locked me up and did for me.
I've done with all that now, and I'm happy. Think of some one else! God!"
cried he and snapped his fingers. "I don't care that for anybody. Whom
should I think of?"

"Well, try a thought for me," cried Mr. Puddlebox, relenting nothing of
his own heat. "I've watched you these four months. I've got you out of
trouble. Curse me, I've fed you and handled you like a baby. But for me
you'd like be lying dead somewhere."

"Well, who cares?" cried Mr. Wriford. "Not me, I don't."

"Ah, and you'd liker still be clapped in an asylum and locked there all
your days; you'd mind that. But for me that's where you'd be and where
you'll go, if I left you to-morrow."

Mr. Wriford cried with a black and angry face: "Well, if it's true, who
asked you to hang on to me? Why have you done it? If it's true, mind you!
For I've done my share. You've admitted that yourself. In the rows we've got
into I've done my share, and in the work we've done I've done more than my
share, once I've learnt the hang of it. Now then! That's true, isn't it? If
you've done so jolly much, why have you? There's one for you. Why?"

His violent storming put a new mood to Mr. Puddlebox's face. Not the
exasperation with which he had burst out and continued till now. That left
him. Not the jolly grin with which commonly he regarded life in general
and Mr. Wriford in particular. None of these. A new mood. The mood and
hue Mr. Wriford had glimpsed when, looking down from the barrier as Mr.
Puddlebox overtook him, and crying down to him: "I thought you'd
stopped," he had seen Mr. Puddlebox blink and heard him say: "You're
unkind, boy." Now he saw it again—and was again to see it before
approaching night gave way to following morn.
 Mr. Puddlebox blinked and went redly cloudy in the face. "Why?" said
he. "Well, I'll tell you why, boy. Because I like you. I liked you, boy, when
you came wretched up the Barnet road and thought there was one with you,
following you. I liked you then for you were glad of my food and my help
and caught at my hand as night fell and held it while you slept. Curse me, I
liked you then, for, curse me, you were the first come my way in many
years of sin that thought me stronger than himself and that I could be
stronger to and could help. I liked you then, boy, and I've liked you more
each sun and moon since. I've lost a precious lot in life through being what,
curse me, I am. None ever to welcome me, none ever to be glad of me, none
ever that minded if I rode by on my legs or went legs first in a coffin cart.
Then came you that was loony, that was glad of me here and glad of me
there, that asked me this and asked me that, that laughed with me and ate
with me and slept with me, that because you was loony was weaker than
me. So I liked you, boy; curse me, I loved you, boy. There's why for you."

This long speech, delivered with much blinking and redness of the face,
was listened to by Mr. Wriford with the fierceness gone out of his eyes but
with his face twisting and working as though what he heard put him in
difficulty. In difficulty and with difficulty he then broke out. "God knows
I'm grateful," Mr. Wriford said, his voice strained as his face. "But look at
this—I don't want to be grateful. I don't want that kind of thing. I've been
through all that. 'Thank you' for this; and 'Thank you' for that; and 'I beg
your pardon;' and 'Oh, how kind of you.' Man, man!" cried Mr. Wriford,
striking his hands to his face and tearing them away again as though scenes
were before his eyes that he would wrench away. "Man, I've done that thirty
years and been killed of it. I don't want ever to think that kind of stuff again.
I want just to keep going on and having nothing touch me except what hurts
me here in my body and not care a damn for it—which I don't. You're
always asking me if I'm happy, and I know you think I'm not. But I am.
Look how hard my hands are: that makes me happy just to think of that.
And how I don't mind getting wet or cold: that makes me happy, so happy
that I shout out with the gladness of it and get myself wetter. It's being a
man. It's getting the better of myself. You're going to say it's not. But you
don't understand. One man has to get the better of himself one way and one
another. With me it's getting the better of being afraid of things. Well, I'm
beating it. I'm beating it when I'm out here, tramping along. But when I'm
sheltering it's beating me. When you tell me—" He stopped, and stooping to
Mr. Puddlebox took his hands and squeezed them so that the water was
squeezed to Mr. Puddlebox's eyes. "There!" cried Mr. Wriford. "Grateful!
I'm more grateful to you. I'm fonder of you than any man I've ever met. But
don't tell me you're fond of me. I don't want that from anybody. When you
tell me that it puts me back to what I used to be. I'm grateful. Believe that;
but don't make me talk about it."

"I never did want you to," said Mr. Puddlebox. "Look here, boy. Look
how we begun on this talk. I told you to think of some one else, care for
some one else, and you broke out 'whom were you to care for?' and I gave
you, being cold and wet and mortal tired, I gave you 'For God Almighty's
sake care for me' and then told you why you should. Well, let's get back to
that. Care for me. Look here, boy. We were ten mile to the next village
along this devil of a place when we left the town. I reckon we've come four,
and here's evening upon us and six to go. Well, I can't go them, and that's
the end and the beginning of it. I'm for going back where there's a bed to be
had and while yet it is to be had, for they sleep early these parts. Wherefore
when I say 'for God Almighty's sake care for me,' I mean stop this chasing
this way and let's chase back the way we come. We'll forget what's gone
between us," concluded Mr. Puddlebox, reverting to his jolly smiles and
getting to his feet, "and I'll hate you and you'll hate me, since that pleases
you most, and back we'll get and have a dish of potatoes inside of us and a
warm bed outside. Wherefore I say:

"O ye food and warmth, bless ye the Lord: praise Him and magnify Him
for ever."

Mr. Wriford laughed, and Mr. Puddlebox guessed him persuaded once
again. But he set his face then and shook his head sharply, and Mr.
Puddlebox saw him determined. "No," said Mr. Wriford. "No, I'm not going
back. I'm never going back. If you want to know what I'm going to do, I'm
going to stay the night out here."

Mr. Puddlebox cried: "Out here! Now what to the devil—"

"I'd settled it," Mr. Wriford interrupted him. "I'd settled it when I thought
you'd gone back. There're little caves all along here—I saw one the other
side of these rocks. I'm going to sleep in one. I'd made up my mind when
you caught up with me. I'm going to do it."

Mr. Puddlebox stared at him, incapable of speech. Then cried: "Wet as

you are?"

"Wet as I am," said Mr. Wriford and laughed.

"Cold as it is and going to be colder?"

"Cold as it is and the colder the better."

"You'll stay alone," cried Mr. Puddlebox. "Curse me if I'll stay with
you."

"You needn't," said Mr. Wriford. "I'm not asking you to."

"But you think I'm going to," cried Mr. Puddlebox. "And you're wrong,
for I'm not. I'm going straight back, and I'm going at once, the quicker to
fetch you to your senses. I'm going, boy;" and in advertisement of his
intention Mr. Puddlebox began resolutely to move away.

Mr. Wriford as resolutely turned to the barrier of rocks and began to

climb.

"Come on, boy," called Mr. Puddlebox.

Mr. Wriford called back: "No. No, I'm going to stay. I'm going to see the
night through."

"You'll know where to find me," cried Mr. Puddlebox. "I'll be where we
lay last night."

Mr. Wriford's laugh came to him through the gathering gloom, and
through the gloom he saw Mr. Wriford's form midway up the rocks. "And
you'll know where to find me," Mr. Wriford called.

Mr. Puddlebox paused irresolutely and cursed roundly where he paused.

Then turned and stamped away across the shingle. When he reached the
rocky arm where first they had quarrelled he stopped again and again
looked back. Mr. Wriford was not to be seen.

"That'll go near to kill him if he stays," said Mr. Puddlebox. "And, curse
me, if I go back to him he will stay. I'll push on, and he'll follow me. That's
the only way to it."

They had spent the previous night in an eating-house where "Beds for
Single Men—4d." attracted wanderers. It was seven o'clock when Mr.
Puddlebox's slow progression—halting at every few yards and looking back
—at length returned him to it. He dried and warmed himself before the fire
in the kitchen that was free to inmates of the house.

"Where's your mate?" asked the proprietor. "Thought you was making
Port Rannock?"

"Too far," said Mr. Puddlebox; and to the earlier question: "He's behind
me. I'll wait my supper till he comes."

He waited, though very hungry. Every time the door of the kitchen
opened he turned eagerly in expectation that was every time denied.
Towards nine he gave up the comfortable seat he had secured before the
blaze and sat himself where he could watch the door. It never admitted Mr.
Wriford.

"What's the night?" he asked a seafaring newcomer.

"Blowing up," the man told him. "Blowing up dirty."

Mr. Puddlebox went from the room and from the house, shivered as the
night air struck him, and then down the cobbled street. Ten o'clock, borne
gustily upon the wind, came to him from the church tower as he turned
along the shore.

None saw him go: and he was not to return.

 CHAPTER III

WATER THAT TAKES YOUR BREATH

Mr. Puddlebox's landsman's eye showed him no signs of that "blowing

up dirty" of which he had been informed. A fresh breeze faced him as he
walked and somewhat hindered his progress; but a strong moon rode high
and lighted him; the sea, much advanced since he came that way, broke
quietly along the shore. "Why, it's none so bad a night to be out," thought
Mr. Puddlebox; and there began to change within him the mood in which he
had left the lodging-house. Seated there he had imagined a rough night, wet
and dark, and with each passing hour had the more reproached himself for
his desertion of his loony. Now that he found night clear and still, well-lit
and nothing overcold, he inclined towards considering himself a fool for his
pains.

An hour on his road brought change of mood again. The very stillness,
the very clearness that first had reassured him, now began to frighten him.
He began to apprehend as it were a something sinister in the quietude. He
began to dislike the persistent regularity of his footsteps grinding in the
deep shingle and to dislike yet more the persistent regularity of the breaking
waves. They rose about knee-high as he watched them, fell and pressed
whitely up the beach, back slowly, as though reluctant and with deep protest
of the stones, then massed knee-high and down and up again. Darkly on his
right hand the steep cliffs towered.

The monotony of sound oppressed him. He began to have an eerie

feeling as though he were being followed, and once or twice he looked
back. No, very much alone. Then his footsteps, whose persistent regularity
had wrought upon his senses, began to trouble him with their noisiness
upon the shingle. He tried to walk less heavily and presently found himself
picking his way, and that added to the eeriness, startling him when the loose
stones yielded and he stumbled.
 He approached that quarter where the shore began to be divided by the
rocky barriers that ran from cliff to sea. Then he apprehended what, as he
expressed it to himself, was the matter with the sea. It was very full. It
looked very deep. What had seemed to him to be waves rolling up now
appeared to him as a kind of overflowing, as though not spurned-out waves,
but the whole volume of the water welled, swelled, to find more room. The
breaking sound was now scarcely to be heard, and that intensified the
stillness, and that frightened him more. He began to run....

Mr. Puddlebox stopped running for want of breath; but that physical
admission of the mounting panic within him left him very frightened
indeed. He went close to the cliffs. Darker there and very shut-up the way
they towered so straight and so high. He came away from them, his senses
worse wrought upon. Then he came to the first of the rocky barriers that ran
like piers from the cliff to the sea, and then for the first time noticed how
high the tide had risen. When he came here with Mr. Wriford they had done
their climbing far from the cliff's base. Now the barrier was in great part
submerged. He must climb it near to the cliff where climbing was steeper
and more difficult. Well, there was sand between these barriers, that was
one good thing. Walking would be easier and none of that cursed noise that
his feet made on the shingle. With much difficulty he got up and looked
down upon the other side....

There wasn't any sand. Water where sand had been—water that with that
welling, swelling motion pressed about the shingle that banked beneath the
cliff.

Mr. Puddlebox said aloud, in a whisper: "The tide!" It was the first time
since he had started out that he had thought of it. He looked along the cliff.
From where he stood, from where these rocky piers began, the cliff, as he
saw, began to stand outwards in a long bluff. The further one went, the
further the tide would.... He carried his eyes a little to sea. Beneath the
moon were white, uneasy lines. That was where the sea swirled upon the
barriers. He looked downwards and saw the placid water welling, swelling
beneath his feet.

"The tide," said Mr. Puddlebox again, again in a whisper. He swallowed

something that rose in his throat. He ran his tongue around his lips, for they
were dry. He shivered, for the perspiration his long walk had induced now
seemed to be running down his body in very cold drops. He looked straight
above him and at once down to his feet again and moved his feet in
steadying of his balance: a sense of giddiness came from looking up that
towering height that towered so steeply as to appear hanging over him. He
looked along the way he had come; and he stood so close to the cliff-face,
and it bulked so enormously before him, that the bay he had traversed
seemed, by contrast, to sweep back immensely far—immensely safe.

Mr. Puddlebox watched that safety with unmoving eyes as though he

were fascinated by it. The longer he watched the more it seemed to draw
him. He kept his eyes upon one distant spot, half way along the bay and
high up the shore, and his hypnotic state presented him to himself sitting
there—safe. Still with his eyes upon it he moved across the narrow pier in
its direction and sat down, legs dangling towards the bay, in the first action
of descending. He twisted about to pursue the action, for he was a timid and
unhandy climber who would climb downwards facing his hold. As he came
to his hands and knees he went forward on them and looked across the fifty
yards of shingle-bank, the sea close up, that separated him from the next
pier of rocks. He was a creature of fear as he knelt there—a very figure of
very ugly fear, ungainly in his form that hung bulkily between his arms and
legs, white and loosely fat in his face that peered timorously over the edge,
cowardly and useless in his crouching, shrinking pose.

He said aloud, his eyes on the distant barrier: "I'm as safe there—for a
peep—as I am here. I can get back. Even if I get wet I can get back."

He shuffled forward and this time put his legs over the other side and sat
a while. Here the drop was not more than three feet beneath the soles of his
boots as they dangled. He drew them up. "If he's safe, he's safe," said Mr.
Puddlebox. "And if he's drowned, he's drowned. Where's the sense of—"

Something that floated in the water caught his eye. A little, round,
greyish clump. About the size of a face. Floating close to the shore. Not a
face. A clump of fishing-net corks that Mr. Puddlebox remembered to have
seen dry upon the sand when first he arrived here. But very like, very
dreadfully like a face, and the water rippling very dreadfully over it at each
pulsing of the tide. Floated his loony's face somewhere like that? Struggled
he somewhere near to shore as that? The ripples awash upon his mouth? His
eyes staring? Mouth that had laughed with Mr. Puddlebox these several
months? Eyes that often in appeal had sought his own, and that he loved to
light from fear to peace, to trust, to confidence, to merriment? Floated he
somewhere? Struggled he somewhere? Waited he somewhere for these
hands which, when he sometimes caught, proved them at last of use to some
one, stronger than some one else's in many years of sin?

Mr. Puddlebox slid to the shingle and ran along it; came to the further
barrier and got upon it; stood there in fear. Beyond, and to the next pier,
there was no more, between sea and cliff, than room to walk.

His lips had been very dry when, a short space before, looking towards
where now he stood, he had run his tongue around them. They were moist
then to what, licking them again, his tongue now felt. Cold the sweat then
that trickled down his body: warm to what icy stream fear now exuded on
his flesh. He had shivered then: now he not shivered but in all his frame
shook so that his knees scarcely could support him. Then it was merely
safety that he desired: now he realised fear. Then only safety occupied his
mind: now cowardice within him, and he knew it. Love, strangely, strongly
conceived in these months, called him on: fear, like a live thing on the rock
before him, held him, pressed him back. He thought of rippling water awash
upon that mouth, and looked along the narrow path before him, and licked
his arid lips again: he saw himself with that deep water, that icy water, that
thick water, welling, swelling, to his knees, to his waist, to his neck,
sucking him adrift—ah! and he looked back whence he had come and ran
his tongue again about his ugly, hanging mouth.

"I'm a coward," said Mr. Puddlebox aloud. "I can't come to you, boy," he
said. "I've got to go back, boy," he said. "I can't stand the water, boy. I've
always been terrified of deep water, boy. I'd come to you through fire, boy;
by God, I would. Not through water. I'm a coward. I can't help it, boy. Water
takes your breath. I can't do it, boy."

He waited as if he thought an answer would come. There was only an

intense stillness. There was only the very tiniest lapping of the water as it
welled and swelled: sometimes there was the faint rattle of a stone that the
sucking water sucked from the little ridge of pebbles against the cliff.
 Mr. Puddlebox looked down upon the water and spoke to it. The words
he spoke might have been employed fiercely, but he spoke them scarcely
above a whisper as though it were a confidence that he invited of the sea.
"Why don't you break and roar?" said Mr. Puddlebox to the sea, bending
down to it. "Why don't you break and roar in waves with foam? You'd be
more like fire then. There'd be something in you then. It's the dead look of
you. It's the thick look of you. Why don't you break and roar? It's the
swelling up from under of you. It's the sucking of you. Why don't you break
and roar?"

No answer to that. Only the aching stillness. Only the very tiniest, tiniest
lapping of the water as it welled and swelled: sometimes the tiny rattle of a
stone that from the ridge against the cliff the sucking water sucked.

In that silence Mr. Puddlebox continued to stare at the water. He stared at

it; and at its silence, and as he stared, and as silent, motionless, he continued
to stare, his face began to work as, in the presence of a sleeper, sudden
stealthy resolve might come to one that watched. Then he began to act as
though the water were in fact asleep. He looked all round, then he stepped
swiftly down to the little ridge. The pebbles gave beneath him and carried
his left foot into the water. He stood perfectly still, pressed against the cliff.
"Why don't you break and roar?" whispered Mr. Puddlebox. No answer. No
sound. He began to tread very cautiously towards the further pier, the palms
of his hands against the cliff, and his face anxiously towards the sea, and all
his action as though he moved in stealth and thought to give the sea the slip.
As he neared the barrier, so neared the cliff the sea. When but twenty yards
remained to be traversed the cliff began to thrust a buttress seaward, awash
along its base. "Water takes your breath," Mr. Puddlebox had said. A dozen
steps took him above his boots, and he began to catch at his breath as the
chill struck him. He opened his mouth with the intent to make these sobbing
inspirations less noisy than if drawn hissing through his teeth. He slid his
feet as if to lift and splash them would risk awakening the sleeping tide. He
was to his knees in it when he reached the rocks. Their surface was green in
slimy weed: that meant the tide would cover them. He got up, and on his
hands and knees upon the slime caught at his breath and peered beyond.

No beach was visible here: only water: perfectly still.

 It was a very short way to the next barrier, and of the barrier very short
what was to be seen. The buttress of the cliff pressed steadily out to what
was no more than a little table of rock, scarcely thicker above the surface
than the thickness of a table-top, then seemed to fall away. A trifle beyond
the table there upstood a detached pile of rock, rather like a pulpit and
standing about a pulpit's height above the water. That table—when it ran far
out along the shore—was where Mr. Puddlebox, looking back, had last seen
his loony stand. He remembered it, for he remembered the summit of the
pulpit rock that peered above it.

The idea to shout occurred to him. That low table seemed to mark a
corner. His loony might be beyond it. If he shouted— He did not dare to
shout. Here, more than before, the intensity of the silence possessed him.
He did not dare to break it. Here, with no beach visible, the water seemed
profoundly dead in slumber.

"Why don't you break and roar?" said Mr. Puddlebox. "Why don't you
—" he held his breath and crept forward. He lowered himself and caught his
breath. His feet crunched upon the shingle bed, the water stood above his
knees, and while the stones still moved where he had disturbed them he
stood perfectly still. When they had settled he began to move, sideways,
very slowly, his back against the cliff. Each sidelong step took him deeper;
at each he more sharply caught his breath. It seemed to him as though the
cliff were actually pressing him forward with huge hands. He pressed
against it with all his force as though to hold it back. It thrust him, thrust
him, thrust him. He was deep to his thighs. He was deep to his waist.
"Water takes your breath," Mr. Puddlebox had said. At each deepening step
more violently his breath seemed to be taken, more clutchingly had to be
recalled. He was above his waist. He stumbled and gave a cry and
recovered himself and began to go back; tried to control his dreadful
breathing; came on again; then again retreated. Now his breathing that had
been sobbing gasps became sheer sobs. He suddenly turned from his
sidelong progress, went backwards in two splashing strides whence he had
come—in three, in four, and then in a panic headlong rush, and as if he
were pursued clambered frantically out again upon the slimy rocks.
 As if he were pursued—and now, as if to sight the pursuit, looked
sobbing back upon the water he had churned. There was scarcely a sign of
his churning. Scarcely a mark of his track. Still as before the water lay
there. Still, and thick, and silent, and asleep, and seemed to mock his fears.

"Blast you!" cried Mr. Puddlebox, responsive to the silent mock. "Blast
you, why don't you break and roar?" He put a foot down to it and glared at
the water. "Why in hell don't you break and roar?" cried Mr. Puddlebox, and
flung himself in again, and splashed to the point at which he had turned and
fled, and drew a deep breath and went forward above his waist....

The cliff thrust him out and he was deeper; thrust again, and he was
above his waist. "Takes your breath"—he was catching at his breath in
immense spasms. The shore dropped beneath his feet and he was to his
armpits, the table of rock a long pace away. He was drawn from the cliff,
and he screamed in dreadful fear. He tried to go back and floundered
deeper. He was drowning, he knew. If he lost his footing—and he was
losing it—he would go down, and if he went down he never would rise
again. He called aloud on God and screamed aloud in wordless terror. The
tide swung him against the cliff and drew him screaming and clutching
along it. He stumbled and knew himself gone. His hands struck the table of
rock. He clutched, found his feet, sprang frantically, and drew himself upon
it. He lay there exhausted and moaning. When his abject mind was able to
give words to his moans, "O my Christ, don't let me drown," he said. "Not
after that, Christ, don't let me drown. O merciful Christ, not after that."

After a little he opened his eyes that had been shut in bewilderment of
blind terror and in preparation of death and that he had not courage or
thought to open. He opened his eyes. This is what he saw.

Beneath his chin, as he lay, the still, deep water. Close upon his right
hand the cliff that towered upwards to the night. A narrow channel away
from him stood the pulpit rock. The cliff ran sharply back from beside him,
then thrust again towards the pulpit; stopped short of it and then pressed
onwards out to sea. Its backward dip formed a tiny inlet over which,
masking it from the open sea, the pulpit rock stood sentinel. The back of the
inlet showed at its centre a small cave that had the appearance of a human
mouth, open. At low water this mouth would have stood a tall man's height
above the beach. A short ridge ran along its upper lip. In the dim light it
showed there blackly like a little clump of moustache. From its under lip,
forming a narrow slipway of beach up to it, there ran a rubble of stones as if
the mouth had emitted them or as if its tongue depended into the sea. The
corners of the mouth drooped, and here, as if they slobbered, the water
trickled in and out responsive to the heaving of the tide.

Mr. Wriford lay upon this slip. He lay face downwards. His arms from
his elbows were extended within the mouth of the cave. His boots were in
the water. His legs, as Mr. Puddlebox thought, lay oddly twisted.

CHAPTER IV

WATER THAT SWELLS AND SUCKS

Who is so vile a coward that one weaker than himself, in worse distress,
shall not arrest his cowardice? Who that has given love so lost in fear as not
to love anew, amain, when out of peril his love is called? Who so base then
not to lose in gladness what held his soul in dread?

First Mr. Puddlebox only stared. Water that takes your breath had taken
his. Water that takes your breath rose in a thin film over the rock where on
his face he lay, passed beneath his body, chilled him anew, and took his
breath again. He watched it ooze from under him and spread before him: lip
upwards where he faced it and ooze beneath his hands. Then gave his eyes
again towards the cave.

Who is so vile a coward? Mr. Puddlebox's teeth chattered with his body's
frozen chill: worse, worse, with terror of what he had escaped—God, when
that sucking water sucked!—fast, faster with that worse horror he besought
heaven "not after that" should overtake him. Who so vile, so base? Ah, then
that piteous thing that lay before his eyes! in shape so odd, so ugly—
broken? dead? Whom he had seen so wild, so eager? who child had been to
him and treated as a child? Who first and only in all these years of sin had
looked to him for aid, for counsel, strength? Who must have fought this
filthy, cruel, silent, sucking water, and fighting it have called him, wanted
him? Ah!

Who is so vile? "Loony," Mr. Puddlebox whispered. "Loony! Hey, boy!"

He only whispered. He did not dare a cry that should demand an answer
—and demanding, no answer bring. "Hey, boy! Loony!" He tried to raise
his voice. He dared not raise it. Anew and thicker now the water filmed the
rock about him. Here was death: well, there was death—that piteous
thing....

Then change! Then out of death life! Then gladness out of dread! Then
joy's tumult as one beside a form beneath a sheet should see the dead loved
move.

About the slipway, as he watched, he saw the swelling water, as if with

sudden impulse, swell over Mr. Wriford's boots, run to his knees, and in
response the prone figure move—the shoulders raise as if to drag the body:
raise very feebly and very feebly drop as if the oddly twisted legs were
chained.

Feebly—ah, but in sign of life! Revulsion from fear to gladness brought

Mr. Puddlebox scrambling to his feet and upright upon them. To a loud cry
there would be answer then! Loudly he challenged it. "Loony!" cried Mr.
Puddlebox, his voice athrill. "Hey, boy, what's wrong? I'm coming to you,
boy!"

It was a groan that answered him.

"Are you hurt, boy?"

There answered him: "Oh, for God's sake—oh, for God's sake!"

"Why, that's my loony!" cried Mr. Puddlebox in a very loud voice. "Hold
on, boy! I'm coming to you!"
 Excitedly, in excited gladness his terrors bound up, quickly as he could,
catching at his breath as his fears caught him, stifling them in jolly shouts
of: "Hold on for me, boy! Why, here I come, boy, this very minute!" he
started to make his way, excitedly pursued it.

"Hold on for me, boy!" The cliff along the wall of the inlet against which
he stood shelved downwards into the dark, still sea. "Here I come, boy!" He
went on his face on the table rock and with his legs felt in the water beneath
him and behind him. "Hold on for me, boy!" His feet found a ridge, and he
lowered himself to it and began to feel his way along it, his hands against
the cliff, above his waist the still, dark sea. "Here I come, boy! This very
minute!"

So he cried: so he came—deeper, and now his perils rose to fight what

brought him on. Deeper—the water took his breath. "Here I come, boy!"
Stumbled—thought himself gone, knew as it were an icy hand thrust in his
vitals from the depths, clutching his very heart. "I'm to you now, boy. Here
—" Terror burst in a cry to his mouth. He changed it to "Whoa!" He was
brought by the ridge on which he walked to a point opposite what of the
slipway before the cave stood dry. The ridge ended abruptly. He had almost
gone beyond it, almost slipped and gone, almost screamed.

"Whoa!" said Mr. Puddlebox. "Hold on for me, boy!" He took his hands
from the cliff and faced about where Mr. Wriford lay. Shaken, he felt his
way lower. God, again! Again his foothold terminated! Abruptly he could
feel his way no more. Like a hand, like a hand at his throat, the water
caught his breath. "Hold on for me, boy!" His voice was thick. "Hold on for
me, boy!" Clear again, but he stood, stood, and where he stood the water
swayed him. Here the cliff base seemed to drop. Here the depths waited
him. Facing his feet he knew must be the wall of the slipway. No more than
a long stride—ah, no more! If he launched himself and threw himself, his
foot must strike it, his arms come upon its surface where that figure lay.
Only a long stride. What, when he made it, if no foothold offered? What if
he missed, clutched, fell? He looked across the narrow space. Only that
spring's distance that figure lay, its face turned from him. He listened. The
silence ached, tingled all about him. Suddenly it gave him from the figure
the sound of breathing that came and went in moans.
 Who is so vile a coward? Swiftly Mr. Puddlebox crouched, nerved,
braced himself to spring. Ah, swifter thrust his mind, and bright as flame
and fierce as flame, as a flame shouting, flamed flaming vision before his
starting eyes. He saw himself leap. He saw himself clutch, falling—God, he
could feel his finger-nails rasp and split!—fallen, gone: rising to gulp and
scream, sinking to suffocate and gulp and writhe and rise and scream and
gulp and sink and go. Like flame, like flame, the vision leapt—upstreaming
from the water, shouting in his ears. Thrice he crouched to spring; thrice
like flame the vision thundered: thrice passed as flame that bursts before the
wind: thrice left him to the stillness, the sucking water, the sound of
moaning breath. A fourth time, a last time: ah, now was gone the very will
to bring himself to crouch!

He stood a moment, vacant, only trembling. His senses fluttered back to

him, and gone, so they informed him, something that before their flight had
occupied them. What? In his shaken state he was again a vacant space
searching for it before he realised. Then he knew. There was no sound of
breathing....

Trembling he listened for it, staring at the figure. Still; there was no
sound. Suddenly he heard it. Dreadfully it came. Feebly, a moaning
inspiration: stillness again—then a very little sigh, very gentle, very tiny,
and the prone figure quivered, relaxed.

Dead? Again, as on the table rock, afraid to call aloud, "Loony!" Mr.
Puddlebox whispered. "Hey, boy!"

No answer. Swelling about him came the creeping water, swayed him,
swelled and swayed again: high to his chest, higher now and moving him—
moving, sucking, drawing. Here was death: ah, well, wait a moment, for
there was death—that piteous thing face downwards there. He spoke softly:
"Hey, boy, are you gone?" The water rocked him. He cried brokenly, loudly:
"Loony! Are you gone, boy?"

Again, again, life out of death, joy's tumult out of fear!

He saw Mr. Wriford draw down his arms, press on his elbows, raise, then
turn towards him his face, most dreadfully grey, most dreadfully drawn in
pain.

Who so vile, so base?

Swift, swift revulsion to gladness out of dread. "Why, that's my loony!"

cried Mr. Puddlebox in a very loud voice.

Mr. Wriford said: "Have you come?"

"Why, here I am, boy!" He steadied his feet.

Very feebly, scarcely to be heard: "I don't see you."

"Why, there's no more than my nob to be seen, boy! I'm here to my nob
in the water." His feet were firm. He braced himself. "I'm to you, boy, and
I'm in the most plaguy place as I challenge any man ever to have been." He
crouched. "I've to jump, boy, and how to the devil—"

He launched himself. His foot struck the slipway bank—no hold!

Smooth rock, and his foot glanced down it! He had thought to spring
upward from what purchase his foot might find. It found none. Clutching as
he fell, he obtained no more than his arms upon the shingle of the slipway,
his chin upon it, his elbows thrusting deep, his fingers clutching in the
yielding stones.

"Loony!" Mr. Puddlebox cried. "Loony!"

He slipped further. He suddenly screamed: "Loony, I'm going! Christ,

I'm going!"

His face, in line with Mr. Wriford's, two arm's-lengths from it, was
dreadfully distorted, his lips wide, his teeth grinding. He choked between
them: "Can you help me, boy?"

Mr. Wriford was trying to help him. Mr. Wriford was working towards
him on his elbows, his face twisted in agony. As he came, "My legs are
broken," he said. "I'll reach you. I'll reach you."
 Eye to eye and dreadfully eyed they stared one upon the other. A foot's
breadth between them now, and now their fingers almost touching.

"I'm done, boy! Christ, I'm done!" But with the very cry, and with his
hand so near to Mr. Wriford's slipped again beyond it, Mr. Puddlebox had
sudden change of voice, sudden gleam in the eyes that had stood out in
horror. "Curse me, I'm not!" cried Mr. Puddlebox. "Curse me, I've bested it.
I've found a hole for my foot. Ease up, boy. I'm to you. By God, I'm to you
after all!"

Groan that was prayer of thanks came from Mr. Wriford. Fainting, his
head dropped forward on his hands. There was tremendous commotion in
the water as Mr. Puddlebox sprang up it from his foothold, thrashing it with
his legs as, chest upon the shingle, he struggled tremendously. Then he
drew himself out and on his knees, dripping, and bent over Mr. Wriford.

"I'm to you now, boy! You're all right now. Boy, you're all right now."

The swelling water swelled with new impulse up the shingle, washed
him where he knelt, ran beneath Mr. Wriford's face, and trickled in the
stones beyond it.

Mr. Puddlebox looked back upon it over his shoulder. He could not see
the table rock where he had lain. Only the pulpit rock upstood, and deep
and black the channel on either hand between it and the walls of their inlet.
He looked within the cave mouth before him and could see its inner face. It
was no more than a shallow hollowing by the sea. He looked upwards and
saw the cliff towering into the night, overhanging as it mounted.

He passed his tongue about his lips.

CHAPTER V
 WATER THAT BREAKS AND ROARS

In a very little while Mr. Puddlebox had dragged Mr. Wriford the three
paces that gave them the mouth of the cave and had sat him upright there,
his back against the cliff. Mr. Wriford had groaned while he was being
moved, now he opened his eyes and looked at Mr. Puddlebox bending over
him.

"Why, that's my loony!" cried Mr. Puddlebox very cheerfully. The flicker
of a smile rewarded him and from the moment of that smile he concealed,
until they parted, the terrors that consumed him. "Why, that's my loony!"
cried he, and went on one knee, smiling confidently in Mr. Wriford's face.
"What's happened to you, boy?"

Mr. Wriford said weakly: "I've broken my legs. I think both my legs are
broken." He indicated the pulpit rock with a motion of his head. "I climbed
up there. Then I thought I'd jump down. Very high and rocky underneath,
but I thought of it, and so I did it. I didn't land properly. I twisted my legs."

He groaned and closed his eyes. "Well, well," said Mr. Puddlebox,
holding his hands and patting them. "There, boy, there. You're all right now.
I'm to you now, boy."

"I suppose I fainted," Mr. Wriford said. "I found it was night and the tide
up to my feet. I began to drag myself. I dragged myself up and up, and the
tide followed. Is it still coming?"

"You're all right now, boy," said Mr. Puddlebox. "Boy, you're all right
now."

He felt a faint pressure from Mr. Wriford's hands that he held; he saw in
Mr. Wriford's eyes the same message that the pressure communicated. He
twisted sharply on his heels, turning with a fierce and threatening motion
upon the water as one hemmed in by ever-bolder wolves might turn to drive
them back.
 From where he knelt the water was almost to be touched.

II

Mr. Puddlebox got to his feet and stooped and peered within the cave.
The moon silvered a patch of its inner face. It gleamed wetly. He looked to
its roof. Water dripped upon his upturned face. The cave would fill, when
the tide was full. He caught his breath as he realised that, looked out upon
the dark, still sea, and caught his breath again. He stepped out backwards
till his feet were in the water and looked up the towering cliff. It made him
sick and dizzy, and he staggered a splashing step, then looked again. To the
line of the indentation that had seemed like a clump of moustache upon the
cave's upper lip, the cliff on either hand showed dark. Above that line its
slaty hue was lighter.

That was high-water mark.

He went a step forward and stood on tiptoe. The tips of his fingers could
just reach the narrow indentation—just the tips of his fingers: and sick
again he went and dizzy and came down to his heels and turned and stared
upon the dark, still sea.

Then he went to Mr. Wriford again and crouched beside him: took his
hands and patted them and smiled at him, but did not speak.

Mr. Wriford spoke. He said tonelessly: "Are we going to drown?"

"Drown?" cried Mr. Puddlebox in a very loud voice. "Why, boy, what to
the devil has drowning got to do with it? Drown! I was just thinking, that's
all. I was thinking of my supper—pork and onions, boy; and when to the
devil I shall have had enough, once I get to it, I challenge you to say or any
other man. Drown, boy! Why, these poor twisted legs of yours have got into
your head to think of such a thing! You can't be thinking this bit of a splash
is going to drown us? Why, listen to this, boy—" and with that Mr.
Puddlebox turned to the sea and stretching an arm towards it trolled in a
very deep voice:

"O ye sea of the Lord, bless ye the Lord: praise Him and magnify Him
for ever!

"That's all that bit of a splash is going to do," said Mr. Puddlebox very
cheerfully; "going to praise the Lord and going to damp our boots if we let
it, which, curse me, we won't. All we've got to think about is where we're
going to sit till the water goes back where, curse me, it should always be
instead of shoving itself up here. One place is as good as another, boy, and
there's plenty of them, but I know the best. Now I'm going to shift you back
a bit, loony," Mr. Puddlebox continued, standing upright, "and then we're
going to sit together a half-hour or so, and then I'm going to have my pork
and onions, and you're going to be carried to bed."

Very tenderly Mr. Puddlebox drew Mr. Wriford back within the cave.
"Now you watch me," said Mr. Puddlebox, "because for once in your life
I'm the one that's going to do things while you look on. There's only a pair
of good legs between us, boy, and that's ample for two of us, but, curse me,
they're mine, and I'm going to do what I want with them."

While in jolly accents he spoke thus Mr. Puddlebox was dislodging from
the floor of the cave large stones that lay embedded in the shingle and piling
them beneath the indentation that showed upon the cave's upper lip. He
sang as he worked. Sometimes "O ye sea" as he had trolled before;
sometimes "O ye stones;" sometimes, as he tugged at a larger boulder—

"O ye fearful weights, bless ye the Lord: praise Him and magnify Him
for ever!"

Always with each variation he turned a jolly face to Mr. Wriford; always
he turned from Mr. Wriford towards the sea that now had reached the
pedestal he was building a face that was grey, that twitched in fear.

"O ye whacking great stones, bless ye the Lord: praise Him and magnify
Him for ever!"
 Knee-high he built his pedestal, working furiously though striving to
conceal his haste. Now he stood in water as he strengthened the pile. Now
the water had swelled past it and swelled to Mr. Wriford's outstretched feet.
Now Mr. Puddlebox climbed upon the mound of stones and brought his
head above the narrow indentation above the cave. It showed itself to be a
little ledge. He thrust an arm upon it and found it as broad as the length of
his forearm, narrowing as it went back to end in a niche that ran a short way
up the cliff. There was room for one to sit there, legs hanging down;
perhaps for two—if two could gain it.

Mr. Puddlebox dropped back to the water and now dragged last stones
that should make a step to his pile. Then he went to Mr. Wriford.

"Now, boy," said Mr. Puddlebox very cheerfully. "Now I've got the
cosiest little seat for you, and now for you to get to it. You can't stand?"

"I can't," Mr. Wriford said.

"Try if I can prop you against the cliff."

He took Mr. Wriford beneath the arms and began to raise him. Mr.
Wriford implored: "Don't hurt me!" and as he was raised from the ground
screamed dreadfully. "Oh, God! Oh, God, don't, don't;" and when set down
again lay feebly moaning: "Don't! Don't!"

There immediately began the most dreadful business.

"Boy," said Mr. Puddlebox, "I've got to hurt you. I'll be gentle as I can,
my loony. Boy, you've got to bear it." He abandoned his pretence of their
safety, and for his jolly humour that had supported it, permitted voice and
speech that denied it and revealed the stress of their position. "Boy, the tide
is making on us. It's to fill this cave, boy, before it turns. There's slow
drowning waiting for us unless I lift you where I've found a place."

"Let me drown!" Mr. Wriford said. "Oh, let me drown."

The sea drove in and washed the cave on every side. Involuntarily Mr.
Wriford cried out in fear and stretched his arms to Mr. Puddlebox, bending
above him.

"Come, boy," said Mr. Puddlebox and took him again beneath the arms:
again as he was moved he cried: "Don't! Don't!"

"Boy," cried Mr. Puddlebox fiercely, "will you watch me drown before
your eyes?"

"Save yourself then. Save yourself."

"By God Almighty I will not. If you won't let me lift you you shall
drown me."

Then determinedly he passed his hands beneath Mr. Wriford's arms; then
resolutely shut his ears to dreadful cries of pain; then, then the dreadful
business. "Boy, I've got to hurt you. I'll be gentle, my loony. Bear it, boy,
oh, for Christ's sake bear it. Round my neck, boy. Hold tight. Bear it, boy;
bear it."

He carried his arms round Mr. Wriford's back, downwards and beneath
his thighs and locked them there. There were dreadful screams; but
dreadfully the water swelled about them, and he held on; there were moans
that rent him as they sounded; but he spoke: "Bear it, boy; bear it!" and with
his burden waded forth.

He faced from the sea and towards the pedestal he had built.

"Loony!"

"Oh, for God's sake, set me down."

"Now I've to raise you."

He began to press upwards with his arms, raising his burden high on his
chest.

"Wade out and drown me," Mr. Wriford cried. "If you've any mercy, for
God's sake drown me!"
 "You're to obey me, boy. By God, you shall obey me, or I'll hurt you
worse. Catch in my hair. Hold yourself up by my hair. High as you can. Up,
up!"

He staggered upon the steps he had constructed; he gained the pedestal

he had made. He thought the strain had become insupportable to him and
that he must fall with it. "Now when I lift you, boy, keep yourself up. I'll
bring you to my head and then set you back." He called upon himself
supremely—raised and failed, raised and failed again. "Now, boy, now!"

He got Mr. Wriford to the ledge and thrust him back; himself he clung to
the ledge and almost senseless swayed between his hands and feet.

Presently he looked up. "You're safe now, boy."

Mr. Wriford watched him with eyes that scarcely seemed to see: he
scarcely seemed to be conscious.

"I had to speak sharply to you, boy."

Mr. Wriford advanced a hand to him, and he took it and held it. "There
was nothing in what I said, boy."

He felt the fingers move in his that covered them. "I had to cry out," Mr.
Wriford said weakly. "I couldn't help it."

"You were brave, boy, brave. You're safe now. The water will come to
you. But you're safe."

"Come up!" said Mr. Wriford. "Come up!"

"I've to rest a moment, boy," Mr. Puddlebox answered him.

He held that hand while he stood resting. He closed his fingers upon it
when presently he spoke again. Now the sea had deepened all about, deep
to his knees where he stood. As if the slipway before the cave while it stood
dry had somehow abated its volume, it seemed to rise visibly and swiftly
now that this last barrier was submerged. All about the walls of the inlet
deeply and darkly it swelled, licking the walls and running up them in little
wavelets, as beasts of prey, massed in a cage, massing and leaping against
the bars.

"There's no great room for me beside you, boy," Mr. Puddlebox said and
pressed the fingers that he held.

"Come up," said Mr. Wriford. "Quickly—quickly!"

Mr. Puddlebox looked at the narrow ledge and turned his head this way
and that and looked again upon the sea.

III

Now, while he looked and while still he waited, the sea's appearance
changed. A wind drove in from seaward and whipped its placid surface.
Black it had been, save where the high moon silvered it; grey as it flickered
and as it swelled about the cliff it seemed to go. It had welled and swelled;
now, from either side the pulpit rock that guarded their inlet, it drove in in
steeply heaving mass that flung within the cave and all along the cliff and
that the cave and cliff flung back. It were as if one with a whip packed this
full cage fuller yet, and as though those caged within it leapt here and there
and snapped the air with flashing teeth.

"Now I'll try for it, boy," said Mr. Puddlebox. "These stones are shaking
under me."

Mr. Wriford withdrew his hand and with his hands painfully raised
himself a little to one side. The action removed his back from the crevice up
the cliff face in which it had rested. A growth of hardy scrub clung here, and
Mr. Puddlebox thrust forward his hand and pulled on it.

"Now I'll try for it, boy," he said again. He looked up into Mr. Wriford's
face. "There's nothing to talk about twixt you and me, loony," he said.
"We've had some rare days since you came down the road to me, boy. If this
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookultra.com