Oceanus

Oceanus

REPORTS ON RESEARCH FROM THE WOODS HOLE OCEANOGRAPHIC INSTITUTION

Vol. 39, No. 2 • Fall/Winter 1996 • ISSN 0029-8182

Oceans & Climate

^

The Ocean Conveyor Belt
Flows Around the World

PACIFIC

SAMW Subantarctic Mode Water

AAIW Antarctic Intermediate Water

RSOW Red Sea Overflow Water

AABW Antarctic Bottom Water

NPDW North Pacific Deep Water

AAC Antarctic Circumpolar Current

CDW Circumpolar Deep Water

NADW North Atlantic Deep Water
UPPER IW Upper Intermediate Water

IODW Indian Ocean Deep Water

One of the keys to understanding how and on what time scales the vast
volume of water in the ocean interacts with the atmosphere and modifies
Earth's climate is determining how water moves from the surface of the ocean
into the interior, how it returns from the depths, and how it flows between the
ocean basins. While many of the articles in this issue focus on the

movement of water within the North Atlantic, these
two recent figures from WHOI Senior
Scientist Bill Schmitz provide a
global synthesis of the present
understanding of the movement
of water between ocean basins
and across the depths.

The numbers in the top figure
are flow rates or transports in
units called Sverdrups (after Nor-
wegian oceanographer Harald U.
Sverdrup), which represent flow

at 1,000,000 cubic meters per second. The red arrows show flow
paths and rates in the shallow and intermediate depths. Green and blue
arrows and numbers show the paths and rates for the deep ocean and for bot-
tom flows, respectively. The figure at left provides a three-dimensional perspec-
tive, labeling the different water types moving along the pathways and adding
the color orange for the very salty, warm water that flows out of the Red Sea,
along with the color purple indicating near-surface circulations. These two
figures are part of what Bill Schmitz, who holds the W. Van Alan Clark, Jr.,
Chair for Excellence in Oceanography, calls his "final report," a summary (in a
somewhat speculative vein, he says) of what he has learned over the past 35
years about large-scale, low-frequency ocean currents. This two volume work is
being published as part of the WF1O1 Technical Report series.

Oceanus

REPORTS ON RESEARCH FROM THE WOODS HOLE OCEANOGRAPHIC INSTITUTION

Vol. 39, No. 2 • Fall/Winter 1996 • ISSN 0029-8182

Cover: R/V Oceania weathers a North Atlantic storm during a
March 1981 study of a warm core ring spawned by the Gulf Stream.
Inset: Researchers aboard R/V Endeavor (University ot Rhode Is-
land), an Oceania sister ship, wresde with a rosette water sampler
in the southern Labrador Sea during a spring 1991 investigation of
the origins of the deep western boundary current.
Urge photo by lames McCarthy. Harvard University Inset by Peter Undry. WHOI

Oceanus is published semi-annually by the Woods Hole
Oceanographic Institution, Woods Hole, MA 02543. 508-289-3516.
http //www whoi edu/oceanus

Oceania and its logo are « Registered Trademarks of the Woods
Hole Oceanographic Institution, All Rights Reserved.

A calendar-year Oceanus subscription is available for $15 in the
US, $18 in Canada. The WHOI Publication Package, including
Oceania magazine and Woods Hole Currents (a quarterly publication
for WHOI Associates and Friends) is available for a $25 calendar-
year fee in the US, $30 in Canada. Outside North America, the
annual fee for Oceanus magazine only is $25, and the Publication

Package costs $40 To receive the publications, please call

(toll free) 1-800-291-6458, or write: WHOI Publication Services,

P.O. Box 50145, New Bedford, MA 02745-0005.

To purchase single and back-issue copies of Oceanus, please
contact: lane Hopewood, WHOI-MS#5, Woods Hole MA 02543.
Phone: 508-289-3516. Fax: 508-457-2182

Checks should be drawn on a US bank in US dollars and made
payable to Woods Hole Oceanographic Institution.

When sending change of address, please include mailing label.

Claims for missing numbers from the US will be honored within

three months of publication; overseas, six months.

Permission to photocopy for internal or personal use or the

internal or personal use of specific clients is granted by Oceanus to

libraries and other users registered with the Copyright Clearance

Center (CCC), provided that the base fee of $2 per copy of the

article is paid directly to: CCC, 222 Rosewood Drive, Danvers, MA

01923 Special requests should be addressed to theOcomus editor

Oceans & Climate

Oceans & Climate 2

Tlie Ocean's Role in Climate & Climate Change
By Michael S. McCartney

If Rain Falls On the Ocean, Does It Make a Sound? 4

Fresh Water's Effect on Ocean Phenomena
By Raymond W. Schmitt

ALACE, PALACE, Slocum 6

A Dj'misrr of Free Floating Oceanographic Instruments
By Raymond W. Schmitt

Alpha, Bravo, Charlie... 9

Ocean Weather Ships 1940-1980
By Robertson P. Dinsmore

A Century of N. Atlantic Data Indicates Interdecadal Change 11

Surface Temperature, Winds, & Ice in the North Atlantic
By Clara Deser

North Atlantic Oscillation 13

By Michael S. McCartney

The Bermuda Station S 14

A Long-Running Oceanographic Show
By Terrence M. loyce and Lynne Talley

Sedimentary Record Yields Several Centuries of Data 16

Tlie Little Ice Age and Medieval Warm Period in the Sargiisso Sea
By Lloyd D. Keigwin

North Atlantic's Transformation Pipeline 19

It Chills and Redistributes Subtropical Water

By Michael S. McCartney, Ruth G. Curry, and Hugo F. Bezdek

Labrador Sea Water Carries Northern Climate Signal South 24

Subpolar Signals Appear Years Later at Bermuda
By Ruth G. Curry and Michael S. McCartney

Transient Tracers Track Ocean Climate Signals 29

By William I. lenkins and William M. Smethie, Jr.

New Data on Deep Sea Turbulence 33

Shedding Light on Verticil/ Mixing
By John M. Toole

Computer Modelers Simulate Real and Potential Climate 36
Combining Equations and Data Pushes Computers' Limits
By Rui Xin Huang and liayan Yang

The El Nino/Southern Oscillation Phenomenon 39

Seeking /te "Trigger" and Working Toward Prediction
By Lewis M. Rothstein and Dake Chen

1930

Editor: Vicky Cullen • Designer: Jim Canavan
Woods Hole Oceanographic Institution

Robert B. Gagosian, Director
Frank V. Snyder, Chairman of the Board of Trustees

lames M. Clark, President of the Corporation
Robert D. Harrington, Jr., President of the Associates

Woods Hole Oceanographic Institution is an Equal Employment Opportunity and Affirmative Action Employer

Printed on
recycled paper

OCEANUS

90N

60N

2

1.5

1

0.5

0

-0.5

-1

-1.5

-2

90S

120E

Annual surface
temperature
change in degrees
Centigrade for the
period 1975-1994
relative to 1955-
1974. This figure,
prepared for the
1996 Intergovern-
mental Panel on
Climate Change,
indicates that
Earth's surface has
been, on average,
warmer (predomi-
nating orange)
over the past 20
years compared to
the preceding 20
years. The cooler
blue areas show,
however, that the
warming has not
been universal.

60E

120E

Our thanks to

Senior Scientist

Robert A. Weller for

editorial assistance

with this issue.

Oceans & Climate

The Ocean's Role In Climate & Climate Change

Michael S. McCartney

Senior Scientist, Physical Oceanography Department

The past decade has brought rapid scientific
progress in understanding the role of the
ocean in climate and climate change. The
ocean is involved in the climate system primarily
because it stores heat, water, and carbon dioxide,
moves them around on the earth, and exchanges
these and other elements with the atmosphere.
Three important premises of the oceans and cli-
mate story are:

• The ocean has a huge storage capacity for
heat, water, and carbon dioxide compared to
the atmosphere.

• Global scale oceanic circulation transports heat,
water, and carbon dioxide horizontally over large
distances at rates comparable to atmospheric rates.

• The ocean and atmosphere exchange as much
heat, water, and carbon dioxide between them as
each transports horizontally.

The ocean and atmosphere are coupled — their
"mean states," evolution, and variability are
linked. Ocean currents are primarily a response
to exchanges of momentum, heat, and water
vapor between ocean and atmosphere, and the
resulting ocean circulation stores, redistributes,
and releases these and other properties. The at-

mospheric part of this coupled system exhibits
variability through shifts in intensity and loca-
tion of pressure centers and pressure gradients,
the storms that they spawn and steer, and the
associated distributions of temperature and water
content. Oceanic variability includes anomalies
of sea surface temperature, salinity,* and sea ice,
as well as of the internal distribution of heat and
salt content, and changes in the patterns and
intensities of oceanic circulation. These coupled
ocean-atmosphere changes may impact the land
through phases of drought and deluge, heat and
cold, and storminess.

One example of coupled ocean-atmosphere
variability is the El Nino/Southern Oscillation or
ENSO (see article on page 39). The appearance
of warm water at the ocean's surface in the east-
ern tropical Pacific off South America has a dra-

*Many of this issue's articles discuss the physical properties
of seawater. The density of seawater changes with tempera-
ture (measured in °C), salinity (measured in parts per
thousand or grams of salt per kilogram of water — typically
given without units, such as simply 34.9), and pressure.
The density of seawater (p) in kilograms per cubic meter is
close to and slightly larger than 1,000 kilograms per cubic
meter. "Potential density," (a), is the value of the relative
density if the seawater is brought to the surface without ex-
changing heat on its way up. This expression helps ocean-
ographers understand the water column's stability.

FALL/WINTER 1996

matic impact on weather and seasonal-to-
interannual climate. Considerable effort has
been dedicated to developing the ability to pre-
dict ENSO, including deployment and mainte-
nance of buoys and other observational systems
in the tropical Pacific and sustained attention to
improving models of ENSO. However, ENSO is
but one of the mechanisms by which the ocean
and atmosphere influence one another. Such
coupling occurs on many time scales, even over
centuries (see "Sedimentary Record" article on
page 16). There is growing interest among the
oceanographic community in developing a bet-
ter understanding of the ocean's role in climate
changes on decadal to centennial time scales,
and many of the articles in this issue focus on
such variability in the North Atlantic Ocean.

There are, as yet, no continuing observations
dedicated, as the observing network in the tropi-
cal Pacific is to ENSO, to monitoring, under-
standing, and predicting decadal climate variabil-
ity involving ocean-atmosphere interaction. Our
challenges are to learn from what observations
and modeling have been done and to develop
strategies for future work.

Sustained observations allow scientists to
detect climatic spatial patterns. For example, the
figure opposite shows interdecadal change in
land and sea surface temperatures. This figure is
taken from the 1996 Intergovernmental Panel on
Climate Change (IPCC) report, a huge effort of
the international climate research community to
assess Earth's climatic state every five years. The
predominating orange indicates that the earth's
surface has been, on average, warmer the past 20
years compared to the preceding 20 years. Signifi-
cant blue areas, principally over the oceans, show
that the warming has not occurred everywhere:
Large areas of the subpolar North Atlantic are
cold, sandwiched between
warm northern North
America and northern
Eurasia, and the North Pa-
cific is also cold, but with a
subtropical emphasis rather
than a subpolar emphasis.

The figure above right
puts a longer time perspec-
tive on the warming by
showing the hemispheric
and global average tempera-
ture over the past 135 years,
the rough limit of useful
sustained measurements.
These curves show the overall
global warming beginning
with the industrial age, but
note the roughly 60 year
oscillation this century, par-

04

5 o.o
i -02

^ -0.4
-0.6

04
0.2
00
-02

-04
-0.6

Southern Hemisphere

Globe

1860 1880

1900

1920

Year

ticularly in the north-
ern hemisphere, show-
ing steeper warming
trends 1910-1940/
1945 and 1975-1995.
Time series like these
lie at the heart of
controversies about
global warming as a
trend versus as a phase
of some mode of
"natural" climate
variability.

Continued sus-
tained measurements
of a broad array of
climate indicators will
eventually directly
answer key questions: Is the steep temperature
rise of the past 20 years the portent of a crisis: a
rise that will continue through the next century
and evolve into an increasingly major climate
perturbation? Or is the steep rise "just" a phase of
a natural oscillation of the climate system super-
imposed on a less severe warming? Or is the
entire warming trend of the past 135 years itself
just the warming phase of a still longer natural
oscillation? There is a preponderance of scientific
judgement, as carefully compiled and described
by the IPCC, that the answer will be somewhere
between the first two possibilities, and that this is
caused by human impact on the climate system.

This issue of Oceania emphasizes the North
Atlantic Ocean, but, to answer these scientific
questions, we must also take on the challenges of
filling in many sparsely sampled regions, build-
ing on the ENSO work in the Pacific and decadal
variability research in the North Atlantic, and
working toward understanding on a global basis.

1940 1960 1980 2000

Hemispheric and
global average tem-
perature for the
past 135 years.

Scientists aboard
R/V Knorr launch a
rosette water sam-
pler and conduc-
tivity/temperature/
depth instrument.
Much of the data
discussed in this
issue was collected
by such equip-
ment. Author
McCartney is the
fellow getting wet
at top left.

The Great Salinity
Anomaly, a large,
near-surface pool
of fresher-than-
usual water, was
tracked as it trav-
eled in the subpo-
lar gyre currents
from 1968 to 1982.

If Rain Falls
On the Ocean

Does It Make a Sound?

Fresh Water's Effect on Ocean Phenomena

Raymond W. Schmitt

Senior Scientist, Physical Oceanography

As with similar questions about a tree in the
forest or a grain of sand on the beach, it
may be hard to imagine that a few inches
of rain matters to the deep ocean. After all, the
ocean's average depth is around 4 kilometers and
only 1 to 5 centimeters of water are held in die
atmosphere at any one time. But it does matter, in
part because the ocean is salty. The effect of rain
diluting the salts in the ocean (or evaporation
concentrating them) can be greater than the effect
of heating (or cooling) on the density of seawater.

50

40

It also matters because rainfall and evapora-
tion are not evenly distributed across and among
ocean basins — some regions continuously gain
water while others continuously lose it. This
leads to ocean current systems that can be sur-
prisingly strong. The processes of evaporation
and precipitation over the ocean are a major part
of what is called "the global water cycle;" indeed,
by all estimates, they dominate the water cycle
over land by factors often to a hundred. The
addition of just one percent of Atlantic rainfall to
die Mississippi River basin would more than
double its discharge to die Gulf of Mexico.

As discussed previ-
ously in Oceanus, our
knowledge of the water
cycle over the ocean is
extremely poor (see the
Spring 1992 issue). Yet
we now realize that it is
one of the most impor-
tant components of the
climate system. One of
the significant pieces of
evidence for this comes
from a description of
die "Great Salinity
Anomaly" put together
by Robert Dickson
(Fisheries Laboratory,
Suffolk, England) with
other European ocean-
ographers. The Great
Salinity Anomaly
(GSA) can be character-
ized as a large, near-
surface pool of fresher
water that appeared off
the east coast of
Greenland in the late
1960s (see figure at
1 left). It was carried
5 around Greenland and
10° into the Labrador Sea

FALL/WINTER 1996

34

from Lazier, 1980
_J I

by the prevailing 35

ocean currents, in the
counterclockwise cir-
culation known as the
subpolar gyre. It hov- Q

ered off Newfound- 0^?

land in 1971 -72 and £

was slowly carried 5

back toward Europe in *°
the North Atlantic
Current, which is an
extension of the Gulf
Stream. It then com-
pleted its cycle and
was back off the east
coast of Greenland by 1964 1966

the early 1980s,

though reduced in size and intensity by mixing
with surrounding waters. The origin of the Great
Salinity Anomaly is thought to lie in an unusu-
ally large discharge of ice from the Arctic Ocean
in 1967. Its climatic importance arises from the
impact it had on ocean-atmosphere interaction
in the areas it traversed.

The GSA derives its climate punch from the
strong effect of salinity on seawater density, with
salty water being considerably denser than fresh
water. That is, these northern waters normally
experience strong cooling in the winter, which
causes the surface water to sink and mix with
deeper waters. This process, called deep convec-
tion (see figure below), is a way for the ocean to

Salinity in the Labrador Sea

from Ocean Weather Station Bravo

1968

1970

1972

Normal Ocean in Winter

large
heat loss

release heat to the atmosphere, heat that then
helps to maintain a moderate winter climate for
northern Europe. However, when the GSA passed
through a region, the surface waters became so
fresh and light that even strong cooling would
not allow it to convert into the deeper waters.
Thus, the deep water remained isolated from the
atmosphere, which could not extract as much
heat as usual from the ocean. The GSA acted as a
sort of moving blanket, insulating different parts
of the deep ocean from contact with the atmo-
sphere as it moved around the gyre. Its impact in
the Labrador Sea has been particularly well docu-
mented (see "Labrador Sea" article on page 24).
When the surface waters were isolated from deep
waters, they became

Ocean in Winter with
Fresh Surface Anomaly

Deep convection is a key component of the ocean's role in Earth's climate. Strong win-
ter cooling of surface waters causes them to become denser than water below them,
which allows them to sink and mix with deeper water. This process releases heat from
the overturned water to the atmosphere and maintains northern Europe's moderate
winter climate. The Great Salinity Anomaly interrupted this process as its pool of
fresher water prevented convection.

cooler. Changing sea
surface temperature
patterns can affect at-
mospheric circulation,
and may possibly rein-
force a poorly under-
stood, decades-long
variation in North At-
lantic meteorological
conditions known as
the North Atlantic Os-
cillation (see box on
page 13). For it is the
ocean that contains the
long-term memory of
the climate system. By
comparison, the atmo-
sphere has hardly any
thermal inertia. It is
3 difficult to imagine
| how the atmosphere
alone could develop a
regular decadal oscilla-
tion, but the advection
of freshwater anomalies
by the ocean circulation

Salinity as a func-
tion of time at 10
meters, 200 meters,
and 1,000 meters
depth as recorded
at Ocean Weather
Station Bravo (see
map on page 10) in
the Labrador Sea.
Deep convection is
possible when the
salinity difference
between shallow
and deep water is
small. This nor-
mally occurs every
winter. However,
from 1968 to 1971,
the presence of the
fresh, shallow,
Great Salinity
Anomaly prevented
deep convection.
Unfortunately,
Weather Station
Bravo is no longer
maintained. Scien-
tists will need to
use new technol-
ogy like the PAL-
ACE float (see Box
overleaf) in order
to reestablish such
time series. Such
data is essential for
understanding the
role of freshwater
anomalies in the
climate system.

OCEANUS

could be an important key to this climate puzzle.

Unfortunately, we have no ready means of
detecting freshwater pulses like the GSA. While
surface temperature can be observed easily from
space, surface salinity, so far, cannot. The salinity
variations important for oceanography require
high precision and accuracy, so there is no quick
and inexpensive method of measurement. We
have had to rely on careful analysis of sparse

historical records from mostly random and unre-
lated surveys gleaned from several nations to
piece the GSA's story together. But how many
other "near-great" salinity anomalies have we
missed because the signal was not quite large
enough? Is there a systematic way to monitor
salinity so that we know years in advance of an-
other GSA's approach?

In addition to variability within an ocean ba-

ALACE, PALACE, Slocum

A Dynasty of Free Floating Oceanographic Instruments

Autonomous diving floats have been developed by
Doug Webb of Webb Research, Inc. in Falmouth,
MA, in conjunction with Russ Davis of the
Scripps Institution of Oceanography. The Profiling Au-
tonomous LAgrangian Circulation Explorer (PALACE) is
a free float that drifts with the currents at a selected
depth, much like a weather balloon drifts with the winds.
At preset time intervals
(typically one or two
weeks) it pumps up a
small bladder with oil
from an internal reser-
voir, which increases its
volume, but not its
mass, and causes it to
rise to the surface. On
the way up it records
temperature and salinity
as a function of depth.
Once at the surface it
transmits the data to a
satellite system that also
determines its geo-
graphical position. The
drift at depth between
fixes provides an esti-
mate of the

"Lagrangian" velocity at
that time and place (as

Doug Webb was photographed on a catwalk above a test tank used to put

the Slocum glider through its paces,
opposed to "Eulerian"

measurements of the velocity past a fixed point. These
names derive from 18th century mathematicians who
originated these ways of looking at fluid flows).

The basic technology of the float has been used for
several years in the nonprofiling ALACE, which simply
provides velocity information. Hundreds of ALACES have
been successfully deployed in the Pacific and Indian
Oceans. A program to release a large number of PALACES
in the Atlantic is just getting underway.

The use of the ALACE as a platform for salinity mea-

surements is not without problems. The slow rising mo-
tion, and low power available, limit the type of sensor
that can be deployed. The problem of sensor drift due to
biological fouling may be severe in some regions, and
methods to prevent fouling are just being developed.
However, because the float spends most of its life in a
deep and climatically stable water mass, not subject to

near-surface atmospheric
variations, we should be
able to compensate for
any drifts.

But the fact that these
floats move around is
something of a drawback
if the objective is to
monitor ocean tempera-
ture and salinity. That is,
in the long run, we
would rather that they
stayed put and measured
the properties in one
place. Such a task could
be achieved if the float
were capable of gliding
_ horizontally and turning
I as it rose. The horizontal
~\ displacement achieved
- could be directed to
maintain one position,
with each excursion

compensating for the drift caused by ocean currents. With
Navy funding, Webb, Davis, and Breck Owens (WHOI)
are currently working on such a gliding float (see photo).
All these floats depend on batteries to power the elec-
tronic sensors, the pump that varies ballast, and the trans-
mitter that sends data to the satellite. The battery life is
around two years, depending on the frequency of profil-
ing and transmitting. One way to extend its life is to use
the ocean's vertical temperature differences to run a
simple heat engine. Doug Webb has another type of float

FALLWINTER 1996

sin, we would like to understand the large dillc-i-
ences in salt concentration among ocean basins,
(see figure on next page ) For example, the Pacific
Ocean is significantly fresher than the Atlantic
and, because it is lighter, stands about halt a
meter higher. This height difference drives the
flow of Pacific water into the Arctic through the
Bering Strait. The salinity difference between
these two major oceans is thought to be caused

by the transport of water vapor across Central
America: The trade winds evaporate water from
the surface of the Atlantic, carry it across Central
America, and supply rainfall to the tropical Pa-
cific. This water loss is the major cause of the
Atlantic's greater saltiness and its propensity to
form deep water. The extra rainfall on the Pacific
makes it fresher and prevents deep convection.
How does this atmospheric transport vary with

with such a propulsion system. It uses a waxy material
that expands when it melts at around 50 degrees, a tem-
perature the float encounters at several hundred meters
depth on each trip to and from the surface. This expan-
sion is used to store energy to pump ballast when needed.
Use of this "free" energy for propulsion reduces the load
on the batteries and extends the life of the float. The ther-
mal ballasting engine has been tested extensively in the
lab and recently deployed off Bermuda in a nongliding
float, where it performed over 120 depth cycles.

Doug Webb's dream is to marry the thermal engine
with the glider, and thus make a long-lived, roving (or
station-keeping) autonomous profiler possible. Years ago
he described the technical possibilities to the late Henry
Stommel, who developed a vision of how such an instru-
ment might be deployed in large numbers around the
globe (see Oceanus, Winter 1989/90). They called the
instrument Slocum, with the idea that it could circumnavi-
gate the globe under its own power, like New Englander
Joshua Slocum, the first solo sailor to perform that feat.
The Internet could allow scien-
tists to monitor Slocum data
from their home laboratories
around the world.

If we deploy enough
Slocums, their data should be
as valuable for predicting
global climate on seasonal to
decadal time scales as satel-
lites and weather balloons are
for forecasting the daily
weather. Indeed, one of
Slocum's key attractions is
that it is inexpensive enough
to deploy in large numbers.
Per-profile costs for both
temperature and salinity are
expected to be $50 or less,
once a mature system is oper-
ating— vastly cheaper than
anything possible using ships.
A globe-spanning array of
1,000 Slocums would cost less
than a new ship, yet provide
an unprecedented view into
the internal workings of the
global ocean. — Ray Schmitt

MIT/WHOI loint Program student Steve Jayne holds an ALACE
(Autonomous LAgrangian Circulation Explorer) float aboard
R/V Knorr during a 1 995- 1 996 (yes, Christmas at sea) cruise for
the World Ocean Circulation Experiment in the Indian Ocean.

About 1 day at surface

Recording temperature
and salinity as it rises

1,000m

Drifting
1 week

During a data collection and reporting cycle, a PALACE (Profiling Autonomous LAgrangian Circu-
lation Explorer) float drifts with the current at a programmed depth, rises every week or two by in-
flating the external bladder (recording temperature and salinity profiles on the way up), spends a
day at the surface transmitting data, then returns to drift at depth by deflating the bladder.

The average surface
salinity distribu-
tion in the global
ocean, as compiled
from many indi-
vidual ship mea-
surements, mostly
during this century.
The figure also
shows the approxi-
mate coverage ob-
tainable with an ar-
ray of about 1,000
Slocums or PAL-
ACES. These would
resolve the large
scale features of the
salinity field and
provide completely
new information
on its variability
with time. The ar-
ray would be an
early warning sys-
tem for the Great
Salinity Anomalies
of the future.

time? Since salinity is a good indicator of the
history of evaporation or precipitation, perhaps if
we had sufficient data, we could see changes in
the upper ocean salt content of the two oceans
that reflect variations in atmospheric transports.
How many years does it take for salinity anoma-
lies in the tropical Atlantic to propagate to high-
latitude convection regions and affect the sea-
surface temperature there? What is the impact on
the atmospheric circulation?

These and other climate problems will con-
tinue to perplex us until we make a serious at-
tempt to monitor salinity on large space and time
scales. One approach would be to maintain ships
in certain places to sample the ocean continually.
A modest effort along these lines was made after
World War II when weather ships were main-
tained at specific sites by several nations (see
following article). The data they collected provide
nearly the only long time-series measurements
available from deep-ocean regions. However, the
weather ships are all but gone; there is only one
now, maintained seasonally by the Norwegians.
Today's satellites provide information on ap-
proaching storm systems, but, unfortunately, they
cannot tell us what we need to know about ocean
salinity distributions.

It now appears that new technology will pro-
vide the key to the salinity monitoring problem,
at a surprisingly modest cost. The Box on pages 6
and 7 describes how we might obtain tempera-
ture and salinity profiles from data collected by
autonomous diving floats. It should be quite

feasible to deploy an array of these station-keep-
ing "Slocums" that would intercept and monitor
the progress of the "Great Salinity Anomalies" of
the future. In the next two years, a large number
of profiling ALACE (precursor to the Slocum)
floats will be deployed in the Atlantic in a pre-
liminary test of the general concept. In addition
to measuring temperature and salinity, Slocums
might some day measure rain. It turns out that
rain falling on the ocean does make a sound, and
work is underway to record that sound with hy-
drophones and develop algorithms to convert the
measured sound level to rain rates.The remaining
technical obstacles to development of a globe-
spanning array of station-keeping Slocums are
small. The only thing lacking is a strong societal
commitment to the support of such fundamental
research on the climate system of the earth.

This research was sponsored by the National Science
Foundation and the National Oceanic and Atmospheric
Administration's Climate and Global Change Program.

Mast of Ray Schmitt's career has been focused on very small-
scale processes in the ocean related to mixing by turbulence and
"salt fingers." Hoivevcr, he has been driven toward studies of the
global-scale hydrologic cycle by a desire to contribute to im-
proved weather and climate prediction, so that he can better
plan to take advantage of the rare good weather in Woods Hole.

,-.

FALL/WINTER 1996

Alpha, Bravo, Charlie

Ocean Weather Ships 1940-1980

Robertson P. Dinsmore

WHOI Marine Operations

The ocean weather station idea originated in
the early days of radio communications
and trans-oceanic aviation. As early as
1921, the Director of the French Meteorological
Service proposed establishing a stationary
weather observing ship in the North Atlantic to
benefit merchant shipping and the anticipated
inauguration of trans-Atlantic air service. Up to
then, temporary stations had been set up for
special purposes such as the US Navy NC-4 trans-
Atlantic flight in 1919 and the ill-fated Amelia
Earhart Pacific flight in 1937.

The loss of a PanAmerican aircraft in 1938 due
to weather on a trans-Pacific flight prompted the
Coast Guard and the Weather Bureau to begin
tests of upper air observations using instru-
mented balloons. Their success resulted in a
recommendation by Commander E. H. Smith of
the International Ice Patrol (and future Director
of the Woods Hole Oceanographic Institution)
for a network of ships in the Atlantic Ocean.
World War II brought about a dramatic in-
crease in trans-Atlantic air navigation, and in
January 1940 President Roosevelt established the
"Atlantic Weather Observation Service" using
Coast Guard cutters and US Weather Bureau ob-
servers. Most flights at this time were using south-

ern routes. On February 10, 1940, the 327-foot
cutters Bibb and Duane occupied Ocean Stations 1
and 2 — the forerunners of Stations D and E (see
chart on next page).

With the US entering the war, Coast Guard
cutters were diverted to anti-submarine duties,
and the weather stations were taken over by a
motley assortment of vessels ranging from con-
verted yachts to derelict freighters, mostly Coast
Guard operated. As trans-Atlantic air traffic in-
creased, so did the number of weather and plane
guard stations. The role of weather during the
Battle of Coral Sea and trans-Pacific flights re-
sulted in stations being set up in that ocean also.
At the service's peak, there were 22 Atlantic and
24 Pacific stations.

At war's end, the Navy intended to discontinue
weather ship operations, but pressure from sev-
eral sources resulted instead in establishment of a
permanent peacetime system of 13 stations.
These are shown on the next page, with the posi-
tions and operating nations listed in the accom-
panying table. Costs of the program were shared
by nations operating transoceanic aircraft.

A typical weather patrol was 21 days on-sta-
tion. A "station" was a 210-mile grid of 10-mile
squares, each with alphabetic designations. The
center square, which the ship usually occupied,
was "OS" (for "on-station"). A radio beacon

Coast Guard
Cutter Sebago was
photographed on
Station A in
lanuary 1949.

OCEANUS

Ocean Weather Stations
1 94O - 1 98O

Sta. Position

A 62W W; 33~J00' W
56"30'N;51'00'W
52"45' N; 3930' W
WOO' N; 4100' W
35°00' N; 48 00' W
36°00' N; 7000' W
6VOO'N;15°00'W
52*30' N; 20*00' W
45=00' W; 76 00' W
66*00' N; 02*00' E

B
C
D

E
H
I
)
K
M

Operator

U.S. & Wet/i.

U.S.

U.S.

U.S.

U.S.

U.S.

U.K.

U.K.

France

Norway

PACIFIC

Map shows the 13
permanent weather
stations established
in 1946 by the
United Nations
Civil Aviation Orga-
nization. Program
costs were shared
by nations operat-
ing transoceanic
aircraft. Letters
missing from the
alphabetical se-
quence were those
used for stations
occupied during
World War II but
not included in the
postwar weather
station program.

Weather balloons
were released from
weather ships every
six hours to gather
data from eleva-
tions as high as
50,000 feet.

transmitted the ship's location.
Overflying aircraft would check
in with the ship and receive
position, course and speed by
radar tracking, and weather data.
Surface weather observations
were transmitted every three
hours, and "upper airs" — from
instrumented balloon data —
every six hours. Using radiosonde transmitters and
radar tracking, balloon observers obtained air
temperature, humidity, pressure, and wind direc-
tion and speed to elevations of 50,000 feet.

Oceanographic observations were recom-
mended for weather ships almost from the start.
Beginning in 1945 and continuing to the end, US

Sta.

Position

Operator

N

30 N; 140 W

U.S.

P

50° N; 145" W

Canada

V

34 N; 164 f

U.S.

ships made bathyther-
mograph (B/T) observa-
tions that today consti-
tute the largest B/T
archive in existence.
Many specific, short-
term programs were
carried out with ocean-
ographers frequently
riding the ships. In
addition to serving as
weather reporters and
navigation aids,
weather ships occasion-
ally rescued downed
aircraft and foundering
ships. Dramatic weather
station rescues include
the Bermuda Sky
Queen in 1947 (Station

C), Pan-American 943 (Station N) in 1956, and
SS Ambassador (Station E) in 1964.

By 1 970, new jet aircraft were coming to rely
less on fixed ocean stations, and satellites were
beginning to provide weather data. In 1974, the
Coast Guard announced plans to terminate the
US stations, and, in 1977, the last weather ship
was replaced by a newly developed buoy. The
international program ended when the last ship
departed Station M in 1981.

Gipt. Dinsmore commanded the weather ship USCGC Cook
Inlet. During /»'» 28-year Coast Guard career, he seri'ed on four
North Atlantic weather ships and was weather ship program
manager before joining the WHO/ Staff in 1971. This article is
e\erpted from a text about twice this length. Interested readers
may request the longer account from the Oceanus office by
calling 508-289-3516 (email: oceanusmag@whoi.edu).

FALL/WINTER 1996

A Century

of North Atlantic Data
Indicates Interdecadal Change

Surface Temperature, Winds, & Ice in the North Atlantic

Clara Deser

Research Associate, University of Colorado

For hundreds of years mariners have re-
corded the weather over the world ocean.
Some 100 million marine weather reports
have accumulated worldwide since 1854, when
an international system for the collection of
meteorogical data over the oceans was estab-
lished. These reports include measurements of
sea surface temperature, air temperature, wind,
cloudiness, and barometric pressure. In the
1980s, the National Oceanic and Atmospheric
Administration (NOAA) compiled these weather
observations into a single, easily accessible digi-
tal archive called the Comprehensive Ocean-
Atmosphere Data Set. This important data set
forms the basis for our empirical knowledge of
the surface climate and its variability over the
world's oceans: One example of a variable sys-
tem is the phenomenon known as El Nino in the
tropical Pacific (see article on page 39). A major
challenge in climate research is to use these data
to document and understand the role of the
oceans in long-term — decadal and centennial —
climate change.

The figure at right shows the geographical
distribution of weather observations over the
oceans for three periods: 1880-1900, 1920-
1940, and 1960-1980. Before the turn of the
century, marine weather reports were largely
restricted to shipping lanes in the North Atlantic
and western South Atlantic. The North Pacific
was not well sampled until after World War II,
and the tropical oceans not until after about
1960; the southern oceans are still largely un-
measured. Due to the irregular sampling, we
focus on describing climate variations over the
North Atlantic back to the turn of the century.
Fortunately, the North Atlantic plays an impor-
tant role in world-ocean circulation.

Two parameters are of key importance to the
physical interaction between ocean and atmo-
sphere: sea-surface temperature and near-surface
wind. They control the rates of heat and momen-
tum transfer between the two media. The top
figure on page 12 displays the long-term average

distributions of sea-surface temperature and
near-surface wind over the North Atlantic. These
charts are based upon all available observations
since 1900. The prevailing westerly winds or
"westerlies" are a well-known feature of the wind
distribution. Sea surface temperatures are gener-
ally warmer in the East Atlantic than in the West
Atlantic at the same latitude, reflecting the mod-
erating influence of the Gulf Stream.

How have the wind and temperature distribu-
tions changed with time? A statistical technique
called empirical orthogonal function analysis
aids in identifying regions of coherent temporal

Geographical dis-
tribution of
weather reports
over the world.
Colored areas
show the average
number of weather
reports per month
in each 2° latitude
by 2" longitude
square over the
world oceans for
each of the time
periods indicated.
White areas indi-
cate there are no
reports.

OCEANUS

Average distribu-
tions of sea surface
temperature ("C)
(top) and near sur-
face wind climatol-
ogy (bottom) over
the North Atlantic
since 1900. The
longest wind arrow
corresponds to 8
meters per second.

change. The results of the statistical analysis point
to the area directly south and east of Newfound-
land as a site of pronounced sea surface tempera-
ture variability. The figure below shows the his-
tory of sea surface temperatures in this region
since 1900. There is a notable tendency for cold
and warm periods to be spaced approximately
one decade apart, as well as longer-term warming
and cooling trends that span several decades.

When the near-surface
wind field is analyzed
in a similar manner
(but independently
from the sea surface
temperatures), similar
decadal-scale oscilla-
tions and longer term
trends are evident. As
noted in the box on the
opposite page, this
basin scale pattern of
\ variability has been
| labeled the North At-
- lantic Oscillation.

What is the nature
of these decadal and
multi-decadal fluctua-
tions? Are they surface
signatures of oscilla-

tions inherent to the deep ocean circulation? Are
they global or confined to the North Atlantic?
What is the role of the atmosphere? There is
mounting evidence from mathematical models
that the North Atlantic Ocean's thermohaline
(heat and density driven) circulation may be-
have as a damped oscillatory system at decadal-
to-multidecadal frequencies, with the atmo-
sphere supplying the energy to maintain the
oscillations against dissipation. In order to test
the relevance of hypotheses generated from the
modeling work, further description of the ob-
served climate record is needed.

A composite picture of the decadal-scale varia-
tions can be formed by averaging all of the cold
(or warm) periods from the figure below left and
subtracting the long-term mean. The figure di-
rectly below shows such an "anomaly" compos-
ite of the cold events. When sea surface tempera-
tures to the east of Newfoundland are colder
than normal, the near-surface westerly winds are
stronger than normal. This relationship may be
indicative of positive feedback between atmo-
sphere and ocean: Stronger winds cool the ocean
surface by enhancing evaporation and heat loss,
while colder surface temperatures shift the lati-
tude of the storm track and prevailing westerlies
southward. Thus, the decadal swings in wind
and temperature may be a manifestation of a
coupled air-sea interaction process, in line with
recent modeling results. What determines the
time scale of the fluctuations, as well as their
amplitude, are unsolved issues at this time.

The decadal fluctuations in sea surface tem-
perature show an intriguing relation to the
amount of sea ice in the Labrador Sea, as the top
figure opposite shows. While information on sea
ice dates back only to 1953, it is evident that
each of the decadal swings of colder-than-nor-
mal temperatures was preceded by a period of
greater-than-normal amounts of sea ice. The
mechanism for this association is not well un-
derstood, although it is plausible that the cold,
stable water mass resulting from melting ice
could be carried by ocean currents into the re-

i surface temperatures for the region
ind east of Newfoundland since
•''•p.irtures from normal in de-

i il curve is a low-pass filtered
version I, curve, emphasizing fluctua-

tions li i .1 few years.

Composite ot decadal-scale cold events in the North
Atlantic using sea surface temperature and wind
anomaly patterns since 1900 Blue (red) contours indi-
cate colder (warmer) than normal sea surface tempera-
tures. The longest wind arrow is 1 meter per second.

FALLWINTSR1996

gion east of Newfoundland. Some researchers
have hypothesized a complex feedback loop
involving Arctic precipitation, runoff, salinity,
and ocean circulation to explain the decadal-
scale sea ice variations.

The lack of understanding of observed, long-
term climate events in the North Atlantic under-
scores the need for further research, particularly
in relating the deep ocean circulation to the
surface conditions. The work described by
Michael McCartney, Ruth Curry, and Hugo
Bezdek beginning on page 19 is one important
step in this direction.

This research was funded by a grant from the Atlantic
Climate Change Program of the National Oceanic and
Atmospheric Administration.

Clara Deser ii'as introduced to oceanography in l')<s i while ii
Research .-\sststant in U'HO/'s Physical Oceanography Depart-
ment She then went to the University of Washington to obtain
a Fh.D. in atmospheric sciences, and hai since kept her teet

History of sea ice
amounts in the La-
brador Sea in rela-
tion to sea surface
temperatures in the
North Atlantic
since 1953.

1954

I960

1970

1980

1990

WIT and her head dry at the Utui'ersity of Colorado at Boulder
Currently, she continues her intellectual pursuits on a part-time
basis u'hile raising two children with her husband lonathan.

North Atlantic Oscillation

The top two panels of the figure, sea level pressure
in millibars, show an example of regional shifting
climatic patterns. From work by leff Rogers of
Ohio State University, they show the high (+) and low
(-) extreme states of the North Atlantic Oscillation
(NAO). The regional atmospheric circulation over the
North Atlantic is normally characterized by a subpolar
high pressure cell centered near the Azores, and a subpo-
lar low pressure cell cen-
tered near Iceland and
Greenland. Between these
two centers the westerlies
blow from North America
towards Europe, while to
the north of the Icelandic
low, and to the south of
the Azorian high the winds
are easterlies. A characteris-
tic oscillation of the
strengths and positions of
these pressure centers
occurs interannually and
interdecadally. In the high
NAO state, the westerlies
are intense, and the cold
continental air they carry off northern North America is
warmed by heat liberated from the warm ocean waters
they blow across, and that warmed air flows across north-
ern Europe from the southwest. When the NAO is in its
low state, the Icelandic low pressure center is displaced
far to the south, off Newfoundland, and there is a high
pressure center over northern Greenland, causing cold

dry polar air to blow across northern Europe, and then
westward across the northern subpolar area towards
North America, warming on the way by the heat liberated
from the ocean to the overlying atmosphere. In this low
NAO state, northern Europe experiences much cooler
summers and more severe winters than in the high NAO
state, while Labrador experiences a milder climate. The
bottom two panels of the figure show that the winter

storm frequency patterns
for the two extreme states
of the NAO are quite dif-
ferent, with the northeast-
ern US experiencing more
Nor'easters during the low
NAO state than during the
high.

The differing winds and
^ the accompanying warmer
= or cooler periods for
| northern Europe and
i northern North America

'•- that occur when the NAO

&

" index marches from one
I extreme to the other over
~ periods of a decade or
more contribute significantly to the distribution of global
temperature change. Comparison of the NAO with a
similar climatic index known as the "Pacific-North
American" (PNA) index indicates that on decadal time
scales there may be coordinated variations throughout
the northern hemisphere or even the whole globe.

— Mike McCartne}'

Unfortunately,
Worthington's ef-
forts (photo at
right) were devoted
to a period of
minimum produc-
tion of this water.
In contrast to this
minimum period
in 1976 (denoted
by red curves of
temperature and
potential density),
a period of maxi-
mum climatologi-
cal production oc-
curred in 1964
(blue curves). In
both cases, the plot
shows the annually
averaged properties
for both calendar
years vs. pressure
to reduce eddy
noise. Note that
the underlying
thermocline at
pressures of more
than 600 decibars
is similar in both
periods: Changes
are not induced
from below.

The Bermuda Station S-
A Long-Running Oceanographic Show

Deeper Waters Show Warming Trend

Terrence M. Joyce

Senior Scientist, Physical Oceanography Department

Lynne Talley

Professor & Research Oceanographer,
Scripps Institution of Oceanography

A time series of hydrographic measure
merits was initiated at Bermuda in 1954
and continues to the present. It began
under the banner of the International Geophysical
Year (1957-1958) with the scientific support of
Henry Stommel of the Woods Hole Oceanographic
Institution and William Sutcliffe, director of the
Bermuda Biological
Station (BBS). The scien-
tists and personnel of
the originating institu-
tions have been the
most active participants
over the years, but the
data have been widely
used by the interna-
tional Oceanographic
community. While other
long time series of mea-
surements in the North
Atlantic began in asso-
ciation with weather
ships, (see "Alpha,
Bravo, Charlie" on page
9) only the Bermuda
measurements have a strong Oceanographic focus.

In recent years, large international programs
including the World Ocean Circulation Experi-
ment (WOCE) and the loint Global Ocean Flux
Study (IGOFS) have pro-
vided scientific justification
for continuation and expan-
sion of the multidisciplinary
Bermuda measurements. As
these programs begin to
wind down, it is important
to recognize the significance
of this time series study to
understanding of climatic
change: These measurements
are the principal source of
information about subsur-
face changes in the Sargasso
Sea and the subtropical gyre
over the past four decades.
In early years, the time series

During the cold winter of 1977, Val Worthington ventured
out aboard the National Oceanic and Atmospheric Admin-
istration's Researcher to study the formation of 18° Water,
one of the principal North Atlantic water masses, in the
northern Sargasso Sea. In the photo, Worthington is leaving
the "hero platform" after launching a Nansen bottle cast.

was denoted by the attribution "Panulirus," the
name of the small BBS research vessel used to
carry out the sampling. However, over the years,
several other research vessels have serviced the site,
and the present practice is to call the time series
Station S, in keeping with the convention for
many weather ship sites.

In the field of physical oceanography, Station S
is not optimally located: There are no major water
masses formed there, and the circulation of the
subtropical gyre and deep boundary currents only
peripherally affect the island. However, its location
in proximity to these
major North Atlantic
circulation features
make it ideal for deter-
mining larger-scale
changes in the basin as
they pass by. To make an
analogy with the study
of automobiles, a re-
searcher might visit
various factories to study
manufacturing and
design changes or just sit
by a busy highway and
observe the passing
traffic — Station S fits the
latter category very well.

In the short space

available, we wish to discuss some of the changes
observed at Bermuda, what they import for the
North Atlantic, and possible reasons for their
occurrence. We will only look at two "layers" in
the water column, the Eighteen-Degree Water and
the North Atlantic Deep Water.

Eighteen-Degree Water was first described by
WFIOI physical oceanographer Valentine
Worthington in 1959 as one of the major water
masses formed in the northern Sargasso Sea in late
winter: It is the principal type of subtropical water
found in the North Atlantic. It occurs at depths of
a few hundred meters at Station S and is character-
ized by a layer of nearly constant density having a
temperature of about 1 8° C. While this layer does
not form at the surface near Bermuda, it occurs just
below the depths of late winter mixed layers near
the island and is closely coupled to surface layers
found there. The Station S time series has been one
of the main barometers (or, more correctly, f/icr-
mometers) of changes in this water mass as it flows

FALL/WINTER 1996

hy the island from source regions to the northeast.

The thickness of Eighteen-Degree Water varies
by a factor of two over the course of the current 42-
year time series. Temperature and salinity (density)
changes also occur over time, and the layer tem-
perature is only approximately equal to 18°C. In
years when this Eighteen-Degree Water is produced
in large quantities, the surface salinity (but not
necessarily temperature) at Station S is high In
poor production years, the surface salinity is low.
Thus, long-term changes in this water mass seem
to be closely linked with processes that affect the
surface salinity High production periods seem to
be recur at intervals of approximately 12 to 14
years. Many of us recall when Worthington con-
vinced the funding agencies to mount a field study
of Eighteen-Degree Water, only to find that none
had formed that year. We can now see that the
climatological minimum of the signal at Bermuda
occurred during the mid 1970s when he went to
sea! Our study of the processes controlling this
variability has not provided a conclusive answer as
to why this periodicity occurs and how it is linked
to surface salinity — but not temperature — changes,
though we believe it must have some connection
with changes in atmospheric circulation and pre-
cipitation over the subtropical Atlantic Ocean.

At depths of 1,500 to 2,500 meters at Station S,
we find another clear signal that is not connected
with atmospheric changes over the subtropical
gyre. This is a slow increase in temperature over
time with a trend that is apparent in records that
date to the early 1920s when deep-water oceano-
graphic measurements were first made near Ber-
muda with accurate reversing thermometers. The
long-term trend of this temperature change is at a
rate of approximately 0.5° C
per century. It is one of the
clearest examples of a long-
term increase of ocean tem-
perature in the subsurface
ocean. That is not to say that
the annually averaged tem-
perature in this layer always
increases. In fact there are
decadal time-scale changes at
this depth too, with 1993
appearing to be the coldest in
20 years. Since waters at this
depth do not communicate
with the surface anywhere in
the subtropical gyre, it is the
subpolar gyre to the north
that is the most likely cause for the variability, if
not the trend. Our present studies indicate that
long-term changes in the production of Labrador
Sea Water are associated with the decadal variabil-
ity in the deep water at Station S. Since the Station
S bottom is at about 3,000 meters, this deep layer

800

1000
1200
1400

1600
u

1800
2000
2200
2400

10

Temperature :C

65 OO'W 64 45'W 64 30'W 64 15'W

The location of Station S is a short steam
southeast from St. Georges' harbor. The
smoothed bathymetry is plotted at one kilo-
meter depth increments.

is the deepest avail-
able in the time se-
ries. There appears to
be a lag of 5 to 6
years until the Labra-
dor Sea Water signal
appears at Bermuda
(see "Labrador Sea"
article on page 24).
Many of the cli-
mate studies that can
be undertaken using
the time series at
Bermuda are ongoing
and of greater value
as time goes on. This is related simply to the fact
that phenomena with time scales of a decade must
be studied with time series that span several
"events" in order to make any statistically signifi-
cant statements. The examples given above are
marginal in the statistical sense because even the
42-year data set is not long enough — we are al-
ways left looking at the most recent years of data
and wondering what will come next! For this
reason it is essential that the observations con-
tinue beyond the lives of some of the large field
programs like WOCE and JGOFS, which will end
in a few years. Though it is possible that techno-
logical advances may enable additional measure-
ments or more cost-effective methods, the Ber-
muda time series currently offers our principal
window into the climate of the subtropical gyre of
the North Atlantic.

The National Science Foundation has funded the Station S
work over the years.

Terr}' loyce was a member of the first class aiimitied to the MIT/
WHO/ loin! Program in Physical
Oceanography 1968. His first cniise,
with Henry Stommcl aboard
Atlantis II that same year, made a
port cat! in Bermuda, so his Ber-
muda connection extends far beyond
this article. Since completing his
doctorate in 1972, he has conducted
research at WHO/ except tor ,i six-
month stint in Germany. Though his
research interests have changed over
the years, there has always been a
strong thread ot seagoing work and
data analysis/interpretation. For the
past eight years, he has senvd as
director of the World Ocean Circula-
tion Experiment Hydrographic
Program Office based at WHO/.

Lynne Taliey was also ii WHOI/MIT
joint Program student, having begun

her graduate career as a large-scale observational oceanograplier
working on items like Eighteen-Degree Water using the Bermuda
record and Labrador Sea Water. She moved into theoretical
studies o/ unstable currents for her degree, completed in 1 982, but
several years after moving on through a postdoc and beginnings of
her research career she found herself squarely back in the interme-
diate and mode waters of the world, both literally and in print.

K 35.2 35.4

Salinity %.

Deep water
changes at Ber-
muda are illus-
trated by two con-
trasting years: 1959
(red curves) and
1987 (blue curves).
During the inter-
vening period, the
deep water warmed
up and became
saltier between
about 1,200
decibars (or ap-
proximately 1,200
meters) and the
bottom. At shal-
lower pressures,
eddy variability
(denoted by the
dotted lines on ei-
ther side of the an-
nual means) ob-
scures any
differences. The
long-term warming
trend at Bermuda
can be traced back
to 1922, when ac-
curate deep water
temperature mea-
surements were
made by the Dan-
ish ship Diimi //.

OCEANUS

Average summer-
time temperature
over six centuries in
the northern hemi-
sphere. Note gener-
ally cooler tem-
peratures between
1550 and 1900, the
period known as
the Little Ice Age.
(Data courtesy of
Raymond S. Brad-
ley, University of
Massachusetts).

Sedimentary Record Yields
Several Centuries of Data

Tlie Little Ice Age and Medieval Warm Period in the Sargasso Sea

-0.1

U1

e_

t
o

Lloyd D. Keigwin

Senior Scientist, Geology & Geophysics Department

New Englanders claim a birthright to
complain about the weather. As we note
that the summer of 1996 was coolest
and wettest in recent memory, most of us have
already forgotten that summer 1995 was unusu-
ally warm and dry. Such variability in weather is
normal, yet in historical times there have been
truly exceptional events. For example, 1816 is
known as the "Year Without a Summer."* During
that year, there were killing frosts all over New
England in May, lune, and August, (uly 1816 was
the coldest July in American history, and frosts
came again in September. Crop failure led to food
shortages through-
out the region.
Although the im-
mediate cause of
cooling has been
ascribed to the
volcanic eruption
of Tambora in
Indonesia the year
before, the Year
Without a Summer
occurred during a
time when weather
was generally more harsh than today. Persistently
harsher weather suggests a change in climate, and
the late 16th through the 19th centuries have
become known as the "Little Ice Age."

The Little Ice Age, and several preceding cen-
turies, which are often called the "Medieval
Warm Period," are the subject of controversy.
Neither epoch is recognized at all locations
around the globe, and indeed at some locations
there is clear evidence of warming while others
show distinct cooling. One author titled a paper:
"Was there a Medieval Warm Period, and if so
when and where?" Nevertheless, when data from
all Northern Hemisphere locations are consid-
ered, the annual average summer temperature
proves to be a few tenths of a degree lower dur-
ing the coldest part of the Little Ice Age in the
late 1500s and early 1600s. Various forcing

-0.3

1-0.4

-0.5

1800's

100

200
Years

mechanisms have been proposed for such
changes, including variation in the sun's energy
output, volcanic eruptions, and mysterious inter-
nal oscillations in Earth's climate system, but
none satisfy all of the data.

Natural climate changes like the Little Ice Age
and the Medieval Warm Period are of interest for
a few reasons. First, they occur on decade to
century time scales, a gray zone in the spectrum
of climate change. Accurate instrumental data do
not extend back far enough to document the
beginning of these events, and historical data are
often of questionable accuracy and are not wide-
spread geographically. Geological data clearly
document globally coherent climate change on

thousand-, ten thou-
sand-, and hundred
thousand-year time
scales, so why is the
record so confusing
over just the past
1,000 years? Second,
as humanity contin-
l ues to expand and
- make more de-
'- mands on our
planet, annual aver-
age temperature
changes of a degree could have considerable
social and economic impacts. Third, as there is
widespread agreement among climatologists that
changes due to human impacts on atmospheric
chemistry will eventually lead to global warming
of about two degrees over the next century, it is
important to understand the natural variability
in climate on the century time scale. Will the
human effects occur during a time of natural
warming or cooling?

Of several approaches to studying climate on
decadal to century time scales, here I will touch
on the study of long series of measurements
made at sea and the study of deep sea sediments.

•This phenomenon is described in Volcano Weather-
The Story of 1816, the Year Without a Summer by Henry
Stommel and Elizabeth Stommel (Seven Seas Press,
Newport, RI, 1983)

1600's

300 400

Before Present

500

600

FALL/WINTER 1996

Ordinarily, there is little overlap between these
two approaches. Reliable and continuous hydro-
graphic observations rarely extend back beyond
several decades, and deep sea sediments usually
accumulate too slowly to resolve brief climate
changes. I lowever, the northern Sargasso Sea is a
region where we have five decades of nearly con-
tinuous biweekly hydrographic data (see preced-
ing article), a long history of sediment trap col-
lections to document the rain of particles from
the sea surface to the seafloor, and exceptional
deep sea cores of sediment. The co-occurrence of
these three elements has led to one of the first
reconstructions of sea surface temperature for
recent centuries in the open ocean.

Oceanographically, Station S in the western
Sargasso Sea is important because tempera-
ture and salinity change there is typical
of a large part of the western North
Atlantic, and it is exclusively western
North Atlantic water that is trans-
ported northward and eastward by
the Gulf Stream. These are the wa-
ters that eventually cool and sink
in the Norwegian and Greenland
Seas, flowing southward to com-
plete a large-scale convection cell
that plays a fundamental role in
regulating Earth's climate.

In addition to long time series of hydro-
graphic data from Station S, the site is remark-
able for the long series of sediment trap data
collected by WHOI's Werner Deuser, beginning
in the 1970s. Those traps have recovered nearly
continuous samples of the seasonally changing
rain of particles that settle from surface waters to
the seafloor. An important component of those
particles is the calcium carbonate shells of
planktonic protozoans known as foraminifera.
There are about 30 species of foraminifera, or
"forams," and Deuser's investigations have es-
tablished the seasonal change in species abun-
dance and their stable isotope composition. We
now know from these studies that only one
species of planktonic foram, Globigerinoides
ruber, lives year-round at the surface of the Sar-
gasso Sea, and it happens to deposit its calcium
carbonate close to oxygen isotopic equilibrium
with seawater. This means that G. ruber is ideal
for reconstructing past changes in the tempera-
ture and salinity of Sargasso Sea surface waters,
as the figure at right illustrates.

Note that the average sea surface temperature
and salinity from near Bermuda display some
systematic variability on an annual average basis
since 1955. These changes reflect a decade-long
variability in the North Atlantic climate regime
that is known as the North Atlantic Oscillation
(see box on page 13). In this time series, the most

severe climate occurred
in the 1960s when
annual average sea
surface temperatures
were depressed about
half a degree by extreme
storminess in the west-
ern North Atlantic.
Cold, dry winds during
winter storms also
probably raised surface
ocean salinity in the
1 960s by promoting
increased evaporation.
If we had "annual aver-
age forams" from the

1960s, their oxygen isotope ratio would look
like the time series shown in purple. The
biggest climate change of the past five
decades could indeed be recorded by
the forams.

Long before I knew that G. ruber
was the best possible foram for
reconstructing sea surface tempera-
tures, I selected that species for my
stable isotope studies because of its
consistent abundance on the Ber-
muda Rise, in the northern Sargasso
Sea to the east of Station S. At the time
(the early 1980s), the Bermuda Rise was under
consideration as a possible site for burial of low
level nuclear waste, and it was necessary to know
how rapidly and continuously the sediment accu-
mulates. It turns out that because of the action of
deep ocean currents, fine-grained clay and silt
particles are selectively deposited there, resulting
in very high rates of sedimentation. And whether
samples are of modern or glacial age, G. ruber is
consistently present. Much of my work over the
past decade has documented the climate changes
that occur on thousand year time scales and are
preserved in foram isotope ratios and other data
from Bermuda Rise sediments.

Until recently, the available data from the
Bermuda Rise showed evidence of century- to
thousand-year climate change continuing right
up to about a thousand years ago, the age of the

36.25

36.50-
36.75-

£• 37.00-

c

5 37.25-
37.50

38.75-

Salinity

1950

1

1960

--0.75

— -Q

--0.50 £ 5

h

Shells of plank-
tonic animals
called formainifera
record climatic
conditions as they
are formed. This
one, Globigerinoides
niber, lives year-
round at the sur-
face of the Sar-
gasso Sea. The
form of the live
animal is shown
above, and its
shell, which is ac-
tually about the
size of a fine grain
of sand, at left.

1970 1980 1990

Year

2000

Bermuda Station S
hydrography
shows the oxygen
isotope ratio that a
foram would have
if it deposited its
shell in equilib-
rium with the an-
nual average sea
surface tempera-
ture and salinity
observed since
1954 at Stations
near Bermuda. The
large decrease in
sea surface tem-
perature and in-
crease in salinity in
the late 1960s was
caused by unusu-
ally unpleasant
weather those
years. (Tempera-
ture and salinity
data provided by
Terry loyce.)

OCEANUS

Since 1978, Scien-
tist Emeritus
Werner Deuser has
collected a nearly
continuous suite of
deep sediment trap
samples at the
Ocean Flux Pro-
gram site near Sta-
tion S. The Ocean
Flux Program traps
are shown follow-
ing recovery
aboard the Ber-
muda Biological
Station vessel
Weatherbird 11. The
traps were de-
ployed along a bot-
tom tethered
mooring at 500,
1,500 and 3,200
meters depths to
intercept particles
sinking through
the water column.
Deuser recently
passed the leader-
ship of the Ber-
muda time-series
program on to As-
sistant Scientist
Maureen Conte.

sediment at the tops of our cores. Because these
samples were recovered with large, heavy tools
that free fall into the seafloor, I suspected that
they might have pushed away sediments of the
last millennium without actually coring them. As
a test of this idea, we acquired a box core from
the Bermuda Rise (box cores penetrate the sea-
floor slowly and disturb surface sediments little)
and radiocarbon dated its surface sediment at the
National Ocean Sciences Accelerator Mass Spec-
trometry Facility located at WHOI. Results
showed that the sediment was modern, and addi-
tional dates were used to construct a detailed
chronology of the past few millennia. When tem-
peratures were calculated from oxygen isotope
results on G. ruber from the box core, and when
data were averaged over 50 year intervals, I found
a consistent pattern of sea surface temperature
change (see figure below). The core-top data indi-
cate temperatures of nearly 23 degrees, very close
to the average temperature at Station S over the
past 50 years. However, during the Little Ice Age
of about 300 years ago sea surface temperatures
were at least a full degree lower than today, and
there was an earlier cool event 26

centered on 1,700 years ago. Events
warmer than today occurred about
500 and 1,000 years ago, during
the Medieval Warm Period, and it
was even warmer than that prior to
about 2,500 years ago.

These results are exciting for a
few reasons. First, events as young
and as brief as the Little Ice Age
and the Medieval Warm Period
have never before been resolved in
deep sea sediments from the open
ocean. Because the Sargasso Sea
has a rather uniform temperature
and salinity distribution near the

25-

24-

23-

22-

21

surface, it seems that these events must have had
widespread climatic significance. The Sargasso
Sea data indicate that the Medieval Warm Period
may have actually been two events separated by
500 years, perhaps explaining why its timing and
extent have been so controversial. Second, it is
evident that the climate system has been warm-
ing for a few hundred years, and that it warmed
even more from 1,700 years ago to 1,000 years
ago. There is considerable discussion in the scien-
tific literature and the popular press about the
cause of warming during the present century.
Warming of about half a degree this century has
been attributed to the human-induced "green-
house effect." Although this is not universally
accepted, it is widely accepted that eventually
changes to Earth's atmosphere will cause climate
warming. The message from the Bermuda Rise is
that human-induced warming may be occurring
at the same time as natural warming — not an
ideal situation. Finally, building on the studies of
physical oceanographers and climatologists,
marine geologists and paleoclimatologists may
use the North Atlantic Oscillation as a model for
understanding North Atlantic climate change on
longer, century and millennial time scales.

This work was funded by the National Oceanic & Atmo-
spheric Administration's Atlantic Climate Change Program.

We encourage Oceanus authors to include a bit of humor in the
short biographies we request. Lloyd Keigwin claimed to be"a
humorless scientist" who doesn't like writing bios, so we asked
Eben Franks, a research assistant in Lloyd's lab, to provide some
information. Here's what Eben wrote: In addition to running a
demanding research program, Lloyd Keigu'in is also a Com-
mander in the Nav}' Reserve. Despite nearly 30 years of sea-
going experience he still finds himself subject to seasickness.
/Editor's note: This is not unusual among oceanographers!]
Lloyd has been deeply affected by episodes of the popular PBS
series "This Old House" and has spent 14 years (and counting)
demolishing two perfectly adijuate houses in the name of reno-
vation. His limited spare time is consumed with multifarious
projects ranging from attempting to convince the Naiy to
convert a nuclear sub for oceanographic research to casting
longing looks at the antique German and British spans cars
collecting dust in his burn.

Medieval
Warm Period

I | I I I I | I I I I | I I I FT r
500 1000 1500 2000

Years Before Present

2500

3000

Estimated sea surface temperature from Station S annual averages
and from Globigerinoides ruber shell oxygen isotopes averaged at
50-year intervals. Note that the range of sea surface temperature
variability on longer time scales is much larger than what has
been observed since 1954 at Station S.

FALL/WINTER 1996

North Atlantic's

Transformation Pipeline Chills

and Redistributes Subtropical Water

But It's Not A Smooth Process And It Mightily Affects Climate

Michael S. McCartney

Senior Scientist, Physical Oceanography Department

Ruth G. Curry

Research Associate, Physical Oceanography Department

Hugo F. Bezdek

Director, Atlantic Oceanographic and Meteorological

Laboratory, National Oceanic and Atmospheric

Administration

Warm and salty waters from the upper
part of the South Atlantic flow north-
ward across the equator and then
progress through the tropical and subtropical
North Atlantic to reach high latitudes. Beginning
with the intense northward flow of the Gulf
Stream off the East Coast of the United States,
these waters are exposed to vigorous cooling,

liberating considerable oceanic heat to the atmo-
sphere. This is the first stage of "warm water trans-
formation" within the North Atlantic, a process
that culminates in the high latitude production of
cold and fresh waters that return to the South
Atlantic in deep reaching currents beneath the
warm waters of the subtropics and tropics.

This article focuses on the part of this warm
water transformation that occurs northwards of
about 45° N, the subpolar realm of the North
Atlantic. Here the warm waters brought to the
area by the Gulf Stream flow eastward across the
basin and then sweep northwards in the eastern
Atlantic, continuing to cool, and freshening as
precipitation and continental runoff exceed
evaporation. This transformation occurs along

The pathways asso-
ciated with the
transformation of
warm subtropical
waters into colder
subpolar and polar
waters in the
northern North At-
lantic. Along the
subpolar gyre
pathway the red to
yellow transition
indicates the cool-
ing to Labrador
Sea Water, which
flows back to the
subtropical gyre in
the west as an in-
termediate depth
current (yellow).
In the Norwegian
and Greenland
Seas the red to
blue/purple transi-
tions indicate the
transformation to
a variety of colder
waters that spill
southwards across
the shallow ridge
system connecting
northern Europe,
Iceland, Green-
land, and northern
North America.
These overflows
form up into a
deep current also
flowing back to
the subtropics
(purple), but be-
neath the Labrador
Sea Water. The
green pathway also
indicates cold wa-
ters— but so influ-
enced by continen-
tal runnoff as to
remain light and
near the sea sur-
face on the conti-
nental shelf.

OCEANUS

1

Temperature ( "C)
of the deep mixed
layer near the end
of winter in the ar-
eas where that
depth exceeds 200
meters. It is based
on hydrographic
survey data re-
corded from 1957
to 1967.

Examples of the
variation of tem-
perature with
depth from the
stations used for
the figure above.

two distinct pathways. The Norwegian Current
carries part of the warm water flowing northward
past Ireland into the Norwegian and Greenland
(Nordic) Seas, while the subpolar gyre carries the
rest westwards towards and past Greenland to the
Labrador Basin.

The transformation from warm to colder water
is a multi-year process: Wintertime winds cool
the surface waters, causing them to convect or
vertically overturn and mix progressively more
deeply into the cooler waters beneath. This sea-
sonal overturning creates large volumes of verti-
cally homogenized water, called mode waters. In
summer, the sun heats the surface waters, form-
ing a cap of warmer water that effectively isolates
the mode water from contact with the atmo-
sphere. Surface cooling in the following winter
removes the cap, and reexposes the mode water,
which then undergoes another round of winter

Temperature (°C)

2 3 4 5 6 7 8 9 10 1 1 1 2 1 3 14 15

Labrador
Basin <

500

D
m
•a

1000

3
ro

1500

2000

cooling in which it gains more thickness. The
mode water thus cools and thickens progressively
through consecutive annual reexposures to the
atmosphere as it simultaneously flows counter-
clockwise around the subpolar gyre.

In winter data the seasonally exposed mode
waters form a smoothly varying ring of progres-
sively colder and more deeply convecting waters
(see figure at left). Estimates of flow speeds in the
subpolar gyre suggest a transit time of about a
decade for a parcel of water that enters the trans-
formation pipeline east of Newfoundland with a
temperature of 12° to 14°C, travels counterclock-
wise, and emerges from the pipeline in the Labra-
dor Basin at temperatures colder than 4° C. The
figure below left illustrates the great thickness of
the regional convection in this mode water ring
with temperature/depth profiles from the New-
foundland Basin where the ring begins, west of
Ireland in the northward flow of the eastern sub-
polar gyre, in the Irminger Basin east of
Greenland where flow is westward, and in the
Labrador Basin where the mode water ring ends.

The process of heat liberation from ocean to
atmosphere by water flowing along the pipeline
acts as a regional radiator, particularly for north-
ern Europe where the westerlies carry the heat
extracted from the ocean. There is considerable
evidence for interdecadal variability in this cli-
mate process. Our first evidence comes from the
Labrador Sea where deep wintertime convective
overturning constitutes the last stage of cooling
along the transformation pathway and vertically
homogenizes the water column to depths some-
times exceeding 2,000 meters, creating the so-
called Labrador Sea Water (coolest profile in
figure below left).

The following article discusses the history of
Labrador Sea Water (LSW). The figure on the
opposite page shows the LSW temperature record
overplotted with a smoothed version of the
North Atlantic Oscillation index, an expression of
the relative strength of the atmospheric westerlies
(see box on page 13). The LSW temperature his-
tory shows a long period of warming from the
1930s to 1971 followed by cooling from 1972 to
1993, culminating in the 1990s with the coldest,
freshest, and thickest LSW ever observed. This
cooling trend, however, was interrupted by a brief
warming in the late 1970s and early 1980s. Thus
the first piece of evidence for climate change in
the warm water transformation system is that the
system's end product, LSW, which in the figure
opposite maps in winter 1962 as 3.3°C, shows
interdecadal variability with temperatures as
warm as 3.5°C in 1970, compared to as cold as
3.1°C forty years earlier and as cold as 2.7°C
twenty-three years later.

Several agents interact to produce LSW. The

FALL/WINTER 1996

LSW within the Labrador Basin has a "residence
time." The total volume of LSW is only partially
replaced each year through the addition of trans-
formed warm water and export of LSW from the
Labrador Basin to the rest of the North Atlantic.
To visualize this, imagine the Labrador Basin as a
Jacuzzi. Water is supplied to the tub at some
temperature and at some rate, is well stirred and
mixed in the tub, and that mix is drained away at
a rate that matches the supply rate. The residence
time is the volume of the tub divided by the rate
of supply and is a measure of the time a given
parcel of water spends in the tub before going
down the drain, and thus is the average time the
water in the tub loses heat to the overlying colder
air. The temperature of the mixed water in the
tub depends on the temperature history of the
warm water being fed into the tub, the rate of
that water supply, the residence time, and the rate
of heat loss from the tub to the air above it. LSW
temperature similarly depends first on the tem-
perature history of the product emerging trom
the warm water transformation pipeline into the
Labrador Basin plus the rate of that flow, and
second on the history of the heat exchange be-
tween the Labrador Basin waters and the overly-
ing atmosphere and surrounding ocean (sort of
an uninsulated tub!). Other factors can influence
such histories: Ice atop the Jacuzzi alters how its
tub temperature evolves; similarly, sea ice and
upper ocean freshwater circulation can modify
convective cooling in the Labrador Basin.

The second kind of evidence for climate
change signals in the northern North Atlantic
comes from sea surface temperature (SST) data.
While such data ultimately represents just the
skin of the ocean, it has the advantage of much
higher data density in space and time than sub-
surface data, for it can be collected from ships of
opportunity without requiring a specialized re-
search vessel and in recent decades can be re-
motely sensed from satellites. SST is a quantity of
great importance in heat exchange with the over-
lying atmosphere. With some care in interpreta-
tion, it can be linked to conditions beneath the
sea surface. Clara Deser's article on page 11 dis-
cusses the overall regional northern North Atlan-
tic SST history. Subpolar and northern subtropi-
cal SSTs were overall anomalously warm in the
1950s and 1960s. Winter SST data reveals that
this general warmth involved warm SST anoma-
lies propagating with the oceanic circulation. The
figure on the next page documents the birth in
1951 of a warm winter SST anomaly east of New-
foundland (red patch) and traces its progression
around the subpolar gyre in subsequent winters
through 1968. In the late 1960s, a cold SST area
develops in the subtropics near 40° N, propagates
into the subpolar gyre with maximal extent in the

mid 1970s, fades around 1980, and reestablishes
in the mid 1980s. We call these SST "anomalies"
because they represent departures from the long-
term local temperatures.

Comparison of this interdecadal progression
of warm and cold SST anomalies to the LSW
temperature history shows provocative parallels:
The post-World War II LSW warming trend occurs
while the warm SST anomaly travels along the
transformation pathway. The LSW cooling period
beginning in 1971 coincides with substantial cold
SST areas in the subpolar gyre — but that general
coldness is interrupted by warmer SST anomalies

1940 1950

1960 1970

1980

1990

in 1980-81, when the LSW cooling trend was
also interrupted. Thus the second piece of evi-
dence for climate change in the warm water
transformation system: There are significant
interdecadal winter SST anomalies in the subpo-
lar gyre as a whole and in particular moving
along the warm water transformation pathway.
Periods of relatively warm SST anomalies along
the transformation pipeline correspond to peri-
ods when LSW warmed, and periods of relatively
cold SST anomalies to periods when LSW cooled.
These observations suggest that the supply of
transforming warm water to our Labrador Basin
Jacuzzi was running warmer in the 1950s and
1960s compared to the succeeding decades. To
further that idea, a link between SST anomalies
and the subsurface water undergoing transforma-
tion is needed.

The third kind of evidence for variability in
the northern North Atlantic climate system
comes from subsurface hydrographic data, which
allow us to link warm and cold winter SST phases
to variability in the mode water distribution. The
figure on page 23 shows temperature and salinity
difference fields at 400 meters in the northern
North Atlantic. This depth falls in the winter
convection range of the mode water along the
warm water transformation pathway (but below
the depth of the seasonal warming cap), and thus
periods of warm or cold 400-meter temperature
correspond to the array of isotherms of the top
figure on page 20 shifting counterclockwise or
clockwise, respectively. The fields in the figure on
page 23 are constructed by subtracting the tem-
peratures for one time period from the preceding
time period and mapping the difference. Our

The history of tem-
perature (circles
and black line) of
the Labrador Sea
Water convecting
in the central La-
brador Basin to
depths sometimes
exceeding 2,000
meters. This is
compared to the
interdecadal march
of an index of the
North Atlantic Os-
cillation, with the
reds indicating the
high index periods
of strong wester-
lies, and the blues
the low index peri-
ods of weak wester-
lies. See the North
Atlantic Oscillation
box on page 1 3 for
an explanation of
the high and low
states of the
"NAO".TheNAO
index plotted here
is formed from the
sea-level pressure
difference between
the subtropical
Azores high pres-
sure center and the
subarctic low pres-
sure center near
Iceland. The au-
thors thank lames
Hurrell (National
Center for Atmo-
spheric Research )
for the latest up-
dates of NAO in-
dex data.

OCEANUS

A time series of
maps derived from
winter SST mea-
surements from
1949 to 1996. Pro-
nounced warm
SST anomalies are
indicated by red
and pronounced
cold SST anoma-
lies by blue.

search for systematic warming/cooling is reason-
ably successful for the period after World War II,
but breaks down in more recent years because
less data was collected in the 1980s. Although
more measurements have been made in the
1990s, data generally takes several years to be-
come available from data archive centers. The
figure on page 23 shows that almost the entire
subpolar region at 400 meters warmed in 1958-
65 compared to 1950-57. Thus as the warm SST
anomaly traversed the transformation pipeline,
the deep winter convection temperatures were
warmer than usual.

This warming trend continued in the more
northern area through 1966-72, but reversed to
cooling in the upstream part of the transforma-
tion pipeline between Newfoundland and Ire-
land. We attribute this to an influx into the pipe-

line of abnormally cold subtropical waters —
which are visible in SST in 1969 to 1972 — and to
winter convection beginning to run abnormally
cold. In the final panel of the figure on the oppo-
site page we see that cooling has taken over al-
most everywhere, consistent with the continued
propagation of the cold SST disturbance along
the pipeline. Thus we find evidence that the sup-
ply pipeline of our oceanic Jacuzzi has slow
interdecadal variations of temperature, and that
its waiming and cooling trends are in phase with
the trends of the Jacuzzi tub water, the LSW.
So far, what we have described are several
aspects of interdecadal variability of upper ocean
signals and their linked behavior. Our fourth
kind of evidence for changes in the northern
North Atlantic returns to the LSW temperature
history and the overplotted NAO index recorded

FALL/WINTER 1996

60°N

50

40'

60'w

50'N

40'w

in the figure on the previous page.
This index registered high values
around 1950, declined to very low
values in the late 1960s, then rose to
very high values again in the early
1990s, indicating an interdecadal cycle
of the basic index of the North Atlan-
tic atmosphere's mid-latitude climatic
state. Dan Cayan of the Scripps Insti-
tution of Oceanography has shown
that a high NAO index tends to en-
hance liberation of heat from ocean to
atmosphere over the Labrador Basin
while a low NAO index tends to di-
minish it. Extrapolating his results to
this interdecadal time scale, the 1950s
and 1960s, with NAO index declining,
correspond to progressively reduced
cooling over the Labrador Basin. Thus
not only was our Jacuzzi supply pipe-
line running warm, but less heat was
escaping from the Jacuzzi tub, that is,
from the Labrador Basin to the overly-
ing atmosphere. Thus the weakening
westerlies reinforced the warming
trend for LSW. Conversely, in the pe-
riod of strengthening westerlies in the 1970s and
1980s, not only was the transformation pipeline
running cold, but the loss of heat from the ocean
to the atmosphere over the Labrador Basin was
progressively enhanced, reinforcing the LSW
cooling trend.

What we have described is a first step in this
sort of climate change work: establishing interde-
pendence in the climate system's physical proper-
ties. The second phase is even more difficult and
challenging than the first: How are those proper-
ties linked or physically coupled? Which "are in
charge," so to speak? Is the ocean a passive par-
ticipant, merely responding to atmospheric forc-
ing changes, or do feedbacks from the ocean to
the atmosphere force evolution of the atmo-
spheric system's climatic state? Certainly, there is a
large heat release to the atmosphere involved in
maintaining the mean state of climate in the
North Atlantic region, and it makes sense that
changes in the warm water transformation system
ought to lead to changes in the overlying atmo-
sphere. But the link between midlatitude SST
anomalies and their potential forcing of climate
change signals has been surprisingly elusive to
theory and modeling efforts, and is, in fact, one of
the primary unresolved issues in climate change
research. You might say it is our "missing link."

Continued measurements are essential to
further progress in unraveling the signals and
their underlying physics, both to monitor the
evolution of the system and to sharpen under-
standing of the physics of specific elements. Two

-2 0 -10 -02 02 10 2.0

Temperature Difference at 400db ( C)

-0 50 -0". 10 -0 02 0

Salinity Difference

02 010
at 400db (

of us (McCartney and Curry) took the
Institution's Research Vessel Knorr to the subpolar
gyre in fall 1996 to begin a concentrated two-year
international effort directed at understanding the
seasonal cycle of the warm water transformation
pipeline of the eastern subpolar gyre.

Mike McCartney came to the Institution in 1 973 with a me-
chanical engineering/ fluid mechanics degree, and the late Val
Worthington supeivised his on-the-job training in oceanography.
His interests in the formation of higher latitude water masses
and their subsequent circulation have taken him all over the
world's oceans, and he has been doing it long enough to person-
ally observe the climatic evolution of his favorite water masses.
His goal is to plan longer and more frequent crimes on his sail
boat, where, unlike on research vessels, fancy equipment can be
restricted to GPS navigation.

Ruth Curry came to Woods Hole in 1 980 as a volunteer at-sea
walchstander and "mud washer." Sixteen years — and many
cruises — later, the sea still holds its allure, but Ruth now sails as
a Chief Scientist, measuring change!' in ivater mass properties
and ocean circulation in search of pieces to the global climate
pu^le. Parenthood has changed her perspective somewhat —
(joins '" sea lised '" mean "° sleep, long hours, lousy food, and
unpleasant working conditions, she says. Now it means getting
more sleep than normal having to think only about work, and
getting all your meals cooked for you— a vacation!
Hugo Be^iek joined the Scripps Institution of Oceanography in
1970 following completion of a degree in physics. After four
years of projects in undenvater acoustics and air-sea transfer, he
became a program manager at the Office of Naval Research
(and ceased, he says, doing the actual work). In 1980, he
moved to another administrative position as Director of AOML,
a NOAA research lab that focuses on climate-related oceanogra-
phy and meteorolog}'. "During the last few years, " he reports, "I
have realized the error of my ways and have been struggling
mightily to redress past sins and return to honest work once
again. With the help of generous people such as my co-authors,
perhaps such an eventuality is possible."

Maps of tempera-
ture and salinity
changes for succes-
sive time periods at
a depth near 400
meters in the
northern North At-
lantic. Reds and
yellows indicate
that the later of the
two periods is
warmer or saltier,
while blues and
greens indicate the
later period is
cooler or fresher.
Gray areas indicate
small salinity and
temperature
change. White ar-
eas indicate a lack
of data.

Labrador Sea Water Carries
Northern Climate Signal South

Subpolar Signals Appear Years Later at Bermuda

Ruth G. Curry

Research Associate, Physical Oceanography Department

Michael S. McCartney

Senior Scientist, Physical Oceanography Department

Changes in wind strength, humidity, and
temperature over the ocean affect rates of
evaporation, precipitation, and heat
transfer between ocean and air. Long-term atmo-

Density Range of
Labrador Sea Water

spheric climate change signals are imprinted onto
the sea surface layer — a thin skin atop an enor-
mous reservoir — and subsequently communi-
cated to the deeper water masses. Labrador Sea
Water is a subpolar water mass shaped by air-sea
exchanges in the North Atlantic. It is a major
contributor to the deep water of the Atlantic, and
changes of conditions in its formation area can
be read several years
later at mid-depths in
the subtropics. Map-
ping these changes
through time is helping
us to understand the
causes of significant
warming and cooling
patterns we have ob-
served at these depths
in the North Atlantic
and links the subtropi-
cal deep signals back to
the subpolar sea surface
conditions.

Labrador Sea Water
(LSW) is the end-prod-
uct of the transforma-
tion process, described
in the preceding article,
that modifies warm and
saline waters through
heat and freshwater
exchanges with the
atmosphere. In the last
stage of this transfor-
mation, deep winter-
time convection occurs
in the Labrador Basin
between North America
and Greenland where

A: Structure of the North Atlantic basins is defined by bathymetric contours (2,000, 3,000, and 4,000 meters)
progressively shaded gray. The Mid-Atlantic Ridge separates the western and eastern basins. The dominant cur-
rents at mid depths (near 1,500 meters) are approximated by the pink lines (bringing relatively warm water
northward) and the blue lines (transporting cold waters southward).
B: Thickness in meters of the density layer corresponding to the Labrador Sea Water (LSW).
C: Temperature, in degrees centigrade, of a density surface in the middle of the LSW layer. These maps high-
light the geographical distribution of the Labrador Sea Water and the Mediterranean Outflow water masses
that mix to produce Upper North Atlantic Deep Water.

D: Temperature profiles from the Labrador Basin and the Mediterranean Outflow region (locations are shown
by x in panel A). Green shading denotes the density layer that is mapped in panel B for each profile.

strong westerly winds
cool the surface waters,
making them denser
than the underlying
deep water. Convection
occurs when the denser
surface waters sink and
mix with the deep water

FALL/WINTER 1996

to produce the cold, thick, and homogeneous
LSW water mass.

In the figure at left (top right panel) we use the
thickness of the LSW mass to indicate its source
region and the pathways along which it spreads.
In the North Atlantic, the LSW is very thick, but as
the circulation carries it away from the Labrador
Basin, mixing with other thinner water masses
progressively erodes its thickness. The intense
pink colors of the two righthand figures opposite
indicate thicknesses exceeding 2,000 meters in the
LSW formation area. Purple colors show some-
what thinner LSW spreading northeast into the
Irminger Basin east of Greenland, eastward via the
North Atlantic Current into the Iceland Basin, and
southwestward along the western boundary into
the subtropical basin.

In this, the cold, fresh and thick LSW contrasts
sharply to its neighboring North Atlantic water
masses of the same density. A cool, salty, but very
thin layer of Iceland-Scotland Overflow Water
occupies the northeast corner of the map as the
yellow-white colors on both thickness and tem-
perature maps. The warm, very salty, and thin
Mediterranean Overflow Water is represented by
the green (thickness) and yellow-red (tempera-
ture) tongues extending across the North Atlantic
from its source region at the Strait of Gibraltar.
Recirculations associated with the Deep Western
Boundary Current, the Gulf Stream, and the
North Atlantic Current (an extension of the Gulf
Stream) mix the LSW and Mediterranean waters,
creating the intermediate thicknesses (blue col-
ors) and temperatures (green colors) between the
two sources. The water mass that results from the
mixing of LSW and Mediterranean Overflow
Water is called the LIpper North Atlantic Deep
Water; it represents one of the major elements
exported into the South Atlantic as part of the
global conveyor belt (see inside front cover).

The 500-meter thickness contour roughly
separates the areas where LSW strongly influences
this layer from regions where Mediterranean
Overflow characteristics predominate and shows
that these LSW influences can be traced south-
wards along the western boundary all the way to
the tropics. Although the top of the LSW layer is
at the sea surface in its subpolar source region, in
the subtropics it is isolated from contact with the
atmosphere and occupies depths between 1,200
and 2,200 meters.

LSW properties — temperature, salinity, and
thickness — have changed significantly through
time, and continued measurements in the Labra-
dor Basin since the 1950s enable us to create the
time series shown in the figure above right. This
record shows a general warming from the 1930s
to 1971, followed by a cooling trend that persists
to the present. Thickness of the LSW layer is di-

Salinity Anomalies Occupy
Labrador Basin

o. 30-
E

• DOI«

1930 1940 1950 1960 1970 19

1930 1940 1950 1960 1970 1980 1990

-C
I-

1940 1950 1960 1970 1980 1990

rectly related to the intensity of wintertime con-
vection, with strong convection producing a thick
layer and weak convection resulting in a relatively
thin layer. Thick conditions in the 1930s, 1950s,
1970s, and 1990s indicate periods of strong con-
vection and loosely correlate to cooler, fresher
LSW conditions. Note the abrupt end to the
1950s and 1960s warming, increasing salinity,
and thinning with the onset of strong convection
in 1972. This cooling, freshening, and thickening
event, however, is interrupted by a period of weak
convection in the early 1980s. Then, by 1987, the
return of strong convection culminates in the
coldest, freshest, and thickest conditions ever
measured. The figure at far right contrasts vertical
temperature profiles from the warm and cold
phases of LSW to emphasize the extraordinary
cooling of the Labrador Basin's water column
over the past 25 years. Note that this cooling has
chilled the LSW beyond its previous cool state
more than 60 years ago, in the 1920s and 1930s.
Two factors principally determine LSW prop-
erty history: the strength of the winds and the
periodic appearance of freshwater anomalies at
the sea surface. The westerlies, which blow cold,
dry air from Canada across the Labrador Basin,
are a significant factor in determining the depth
of Labrador Basin wintertime convection. An
increase in wind strength removes more heat
from the surface waters and deepens the extent of
the convection. This also results in a cooler over-
all LSW, since the increased heat loss at the sea
surface is distributed downward as the water
column converts. The relative strength of the
westerlies is represented by the North Atlantic
Oscillation (NAO) index. (See NAO Box on page
13. The NAO index is defined in the figure cap-
tion on page 21.) Overplotted on the thickness
axis in the figure above, the NAO index (shaded

peralure CO

Left, time series of
Labrador Sea Water
properties in its
source region.
Thickness is the
vertical distance
(meters) between
two density sur-
faces that bracket
the Labrador Sea
Water. The North
Atlantic Oscillation
index has been
overplotted on the
thickness axis with
high index shaded
red and low index
blue. Years in
which surface sa-
linity anomalies
occupied the La-
brador Basin are
shaded green.
Above, depth pro-
files of temperature
for three different
years contrasting
the Labrador Sea
Water temperature
in its cool period
before World War
II (1935), the peak
of warming (1971),
and at its coldest
point (1993).

OCEANUS • 25

Temperature changes
recorded in the
1,500 to 2,500 meter
layer near Bermuda.
The top plots show
time series of Ber-
muda temperature
anomaly (red curve)
lagged by 6 years,
thickness of the La-
brador Sea Water
layer (blue curve) in
its formation area,
and temperature of
the L.SW core (green
curve). The lower
plot shows a lagged
correlation analysis
for Bermuda tem-
perature and Labra-
dor Sea Water (LSW)
thickness, which is
highest for lags of 5
to 7 years. The au-
thors' interpretation
is that subpolar
thickness anomalies
result in variability
of the volume of
LSW entering the
subtropics. A large
volume of LSW shifts
the balance of influ-
ence between LSW
and Mediterranean
Outflow towards
LSW. When the LSW
is thick, Bermuda
sees colder (and
fresher) conditions
about 6 years later,
while a thin LSW
source results in
stronger Mediterra-
nean Outflow influ-
ence and Bermuda
sees warmer (and
saltier) conditions
after about 6 vears.

red for high, blue for low) shows trends similar
to the LSW thickness: declining NAO index and
thinning LSW from the 1950s to 1970, a pulse of
strong westerlies and LSW thickening in the early
1970s followed by weak westerlies and thin con-
ditions in the late 1970s, then extremely strong
westerlies (high NAO index) and thick condi-
tions in the 1990s. Notice also in the figure on
the previous page that LSW temperature is warm-
ing during low NAO periods and cooling during
years of high NAO. Thus the atmospheric climate
signal becomes imprinted on the LSW thickness
and temperature.

The thin LSW layers of 1967-72 and the early
1980s correspond to buildups of extremely low
surface salinity conditions, which resulted in low
surface densities. Because this surface water re-
quired extraordinary cooling to make it denser
than the underlying water, it completely inhib-
ited convection in the Labrador Basin. These two
events, referred to as the "Great Salinity
Anomaly" and the "Lesser Great Salinity
Anomaly," eventually moved around the subpo-
lar gyre (see "If Rain Falls" on page 4). The first
Great Salinity Anomaly occupied the Labrador
Basin during a time of low NAO index when
weak winds would have reduced the convection
depths anyway. However, the Lesser Great Salinity
Anomaly hindered convection in the early 1980s,
which resulted in a thin LSW source despite
strong westerlies associated with a relatively high
NAO index. Both anomalies ended with a strong
increase in convection and a downward mixing
of the freshwater cap, which dramatically lowered
the LSW core salinity and temperature.

The subpolar LSW, carrying the imprinted
climate signals, enters the subtropics along two

Year at Bermuda
1960 1970 1980

1000

$ 1500

2500

Bermuda Temperature Anomaly
lagged by 6 years

1950

1960

1970 1980

Year at Labrador Basin

C -0.6
-t

Oi

° -0.4

c
o

1-0.2

0.0

4 6

Time Lag in Years

principal pathways: The deep western boundary
current transports LSW from the Labrador Basin
to the Caribbean Islands, and the Gulf Stream
and North Atlantic Currents carry it from the
western boundary out into the interior of the
ocean. The LSW in these flows is strongly stirred
and mixed by current and eddy action along
these pathways and in the ocean interior. In the
figure on page 24, the basin-scale, deep-water
properties thus represent a blending of the LSW
influence with other influences, principally the
Mediterranean Overflow Water. The resulting
blended water mass, known as the Upper North
Atlantic Deep Water (UNADW), exhibits a tem-
perature history that we can now relate to varia-
tions in LSW source properties. This link was
previously obscure because the subtropical
UNADW temperature signal is more strongly
influenced by the LSW thickness history than by
the LSW temperature history. Furthermore, the
time the ocean requires to transport and mix the
LSW into the subtropical LINADW introduces a
time delay to the link between these signals.
The temperature of the subtropical mid
depths (1,000 to 2,500 meters) has generally
warmed since the 1950s. The figure below left
(red curve) shows this warming trend using a
long time series measured at Bermuda and a
recent analysis of its thermal structure by Terry
Joyce and Paul Robbins (see "Bermuda's Sta-
tion S" on page 14). When a time lag is applied
to the Bermuda signal, its temperature is re-
markably similar to the LSW thickness (blue
curve), while the subpolar LSW temperature
signal (yellow curve) diverges after 1975. Corre-
lation between the subtropical temperature and
subpolar thickness signals is greatest at lags of

five to six years and
1990 2000 implies that when

subpolar convection is
strong — and the LSW
layer is thick — the
subtropics follow five
to six years later with
cooler temperatures at
mid depths. Con-
versely, weak convec-
tion and a thin LSW
-0.2 layer is followed by
warmer subtropical
temperatures approxi-
mately six years later.

To place this rela-
tionship into a geo-
graphic context, the
figure opposite maps
the thickness of the
LSW density layer in six
different time frames,

0.2

0.1

0.0

-0.1

1990

10

FALL/WINTER 1996

each spanning about 7 years
and chosen to represent
phases of LSW source varia-
tion. As noted above, the
thickness of the LSW layer in
its subpolar formation area
changes through time as the
convection intensity varies:
The Labrador Basin LSW
source is thick in the first two
time frames, extremely thin
in 1966-72 and 1980-86 (a
thick pulse in 1973-79 sepa-
rates these two periods), and
grows to extreme thicknesses
in the final time frame. Away
from the source (near the
western boundary east and
south of Newfoundland and
east of New England), the
layer is noticeably thin in the
1970s and 1980s, but ro-
bustly flooded with LSW in
the 1950s and 1990s. Over
the rest of the subtropics
(north of 10° latitude), the
LSW layer thickness changes
most in the fourth time
frame (1973-79) when the
Mediterranean Overflow
Water characteristics (green
colors) are extended north-
wards in the eastern basin
and westwards in the western
basin. Compare the areas
around the Azores and south
of Bermuda in each panel to
see this change.

In order to visualize the
impact of LSW temperature
and thickness anomalies (changes in temperature
and thickness) on the subtropics, the figure on
the next page maps the temperature and thick-
ness differences in one time frame compared to
the previous time frame for each period in the
figure at right. The patterns of anomalies show
large areas where thickness changes correspond
to temperature changes — where the layer thins,
temperatures grow warmer, and where the layer
thickens, temperatures are cooler— and delineate
where LSW exerts a strong influence. These pat-
terns also show consecutive instances where
temperature and thickness anomalies of one
color first appear in the Labrador Basin, rather
quickly move southward and eastward with the
western boundary current and North Atlantic
Current, and then, one time frame later, anoma-
lies of the same color appear in both the western
and eastern subtropical basins. The subtropical

Thickness of the
I,SW density layer
for six consecutive
time periods.

80:W 60:'W 40°W 20:'W

0 250

500 1,000 1,500 2,000

Thickness (meters) O1500= 34.62-34.72

2,500

deep water anomalies appear to lag behind the
subpolar LSW signal by five to seven years as the
lagged correlation of the Bermuda data suggests.

The subtropical temperature anomalies are
large compared to the subpolar temperature
anomalies. Because a signal weakens as it moves
away from its source, these subtropical signals
cannot be simply the advected subpolar tem-
perature anomalies. Rather, the time-delayed
subtropical response to LSW source variability
represents the slow adjustment of the subtropi-
cal deep water to the waxing and waning of LSW
strength so clearly visible in the figure above.
The thicker the LSW, the stronger its role in
mixing with Mediterranean Overflow water, and
this is manifested as an eastward and southward
erosion of the influence of the Mediterranean
Overflow Water on the subtropical deep water.
The time needed for the LSW to circulate and

OCEANUS • 27

THICKNESS DIFFERENCES

i°W 60°W 40° W 20 "W 0

TEMPERATURE DIFFERENCES

80°W 60°W 40°W 20"W 0°

:l

Nl

1000

300

200

100

1-100

-200

-300

1958-1965 compared to 1950-1957

1958-1965 compared to 1950-1957

1966-1972 compared to 1958-1965 1966-1972 compared to 1958-1965

60 N
50:'N

0.20

0.10

0.08

-0.08

40"N
30' N
20°N
10"N
0' CE^^^^^^^^^^K^^JE^^"^^^1' 1 1 n i n

1973-1979 compared to 1966-1972 1973-1979 compared to 1966-1972

1980-1986 compared to 1973-1979 1980-1986 compared to 1973-1979

1987-1994 compared to 1980-1996

1987-1994 compared to 1980-1996

Thickness difference fields (left column) and temperature difference fields
(right column) were constructed by subtracting thickness or temperature in two
consecutive time frames at each 1 -degree square in the North Atlantic. The
thickness represents the LSW density layer and temperature values are taken at a
density surface in the middle of that layer. Green-blue colors indicate layer
thickening and/or cooling in one time frame compared to the previous time
frame; yellow-red colors indicate layer thinning and/or warming.

mix into the subtropical basins results in a de-
layed appearance of the response. When the
LSW is thinner than normal, the Mediterranean
Overflow Water exerts more influence, and this
appears as a westward and northward extension
of the thin, warm, and salty characteristics.

Understanding the nature of the subtropical
temperature variations and knowing that the
subpolar convection has been extremely strong
from 1988 to 1995 enables us to predict that the
subtropical mid depths will continue to cool
through the 1990s. Tracing the extremely cold,
fresh, and thick signal that is now invading the
subtropics (quite pronounced in the bottom
panels of the figure on the next page) will pro-
vide us with valuable information concerning the
timing and geography of the complex mid-lati-
tude circulation system whose end product, the
Upper North Atlantic Deep Water, is exported to
the southern ocean.

Our WHOI colleague Bob Pickart (see
Oceanus, Spring 1994) has tracked the penetra-
tion of the extreme LSW along the deep western
boundary current and Gulf Stream system off
New England, and our University of Miami col-
leagues Rana Fine and Bob Molinari have recently
(summer 1996) sighted this extreme LSW signal
in the deep western boundary current off Abaco
in the Bahamas— one of the most exciting and
valuable results of their decade-long monitoring
program at that location.

We are planning a 1998 field experiment to
take advantage of this unique climate change
signal by measuring the subtropical western
basin's response to the LSW invasion at 24° N
and 15° N. Because of the time delay observed in
the subtropical response, we can be reasonably
confident that we will be in the right places to
measure this extreme LSW event, for we know
through the continued efforts of our Canadian
colleagues John Lazier and Allyn Clarke (Bedford
Institute of Oceanography) that the LSW source
continued to convert through the winter of 1995.
They report a cessation of deep LSW convection
in winter 1996. If that cessation is longer-lived
than a single anomalous winter event, then we
would expect it to appear as a subtropical climate
change signal in 2000 to 2002.

The authors' research is jointly funded by the National
Science Foundation-sponsored World Ocean Circulation
Experiment Program and the Climate and Global Change
Program of the National Oceanic and Atmospheric Admin-
istration. The authors thank Terry Joyce for collaboration
in producing the figure on page 26, James Hurrel (Na-
tional Center for Atmospheric Research) for providing the
most recent update of the NAO index data, and John
Lazier for providing the Labrador Basin data for recent
years and for his sustained effort for more than 30 years in
maintaining critical time-series measurements in the
hostile environment of the Labrador Basin.

28 * FALLA/VINTER 1996

Transient Tracers Track
Ocean Climate Signals

William I. Jenkins

Senior Scientist, Marine Chemistry & Geochemistry Dept.

William M. Smethie, Jr.

Senior Research Scientist, Lamont-Doherty Earth
Observatory, Columbia University

Transient tracers provide us with a unique
opportunity to visualize the effects of the
changing climate on the ocean. They trace
the pathways climate anomalies follow as they
enter and move through the ocean and give us
valuable information about rates of movement
and amounts of dilution. This knowledge is
important for developing ocean-climate models

to predict long term climate changes.

Humankind's activities have resulted in the
release of a number of globally distributed sub-
stances into the environment. These substances
enter the oceans, and, although they have little, if
any, impact on the environment, they travel
through and "trace" the biological, chemical, and
physical pathways of the ocean. The distributions
of these "tracers" change with time. For example,
isotopes created by atmospheric nuclear weapons
tests in the 1950s and 1960s were introduced in a
pulselike fashion, while atmospheric concentra-
tions of chlorofluorocarbons (CFCs), which

A bird's eye view of
the distribution of
tritium in the
North Atlantic. Pic-
ture yourself float-
ing a few hundred
miles above Nor-
way, looking
southwestward
down at the North
Atlantic. North
America is in the
top right corner of
the view,
Greenland to the
lower right, and
parts of Europe,
Great Britain, and
Africa are visible
on the lower left.
The spikes are
ocean islands. The
blue "blanket" is
the 1 Tritium Unit
isosurface (surface
of constant tritium
measured in 1981).
(One Tritium Unit
equals one tritium
atom to 10 '" hy-
drogen atoms.)
Underneath this
blanket lies water
that has not been
appreciably venti-
lated (in contact
with the atmo-
sphere) while wa-
ter above this level
has been ventilated
since the 1960s.

OCEANUS

threaten the earth's ozone layer, have been in-
creasing with time. We refer to these substances
in the ocean as "transient tracers" because their
distributions are evolving.

Transient tracers are valuable tools for study-
ing ocean climate. First, because they are new to
the ocean environment, they are indicators of
"ocean ventilation." Ventilation is the imposition
of atmospherically derived properties on water
masses. For example, waters in contact with the
atmosphere will have dissolved oxygen concen-
trations increased to equilibrium values with the
atmosphere. Providing their time history in the
atmosphere is known and the manner in which
they are transferred to the ocean is understood,
they can be used to construct and test models of
ocean ventilation and circulation. Observations
of their distributions in the ocean and time series
measurements of how they change with time are

2500

1970

1975

1980

A time series of tri-
tium in the Sar-
gasso Sea near Ber-
muda The plot of
tritium vs. depth
and time shows the
sudden arrival of
tritium at interme-
diate depths (1,000
to 1,500 meters) in
the late 1970s, and
at deeper depths
(2,000 to 2,500
meters) in the late
1980s. These
events correspond
to the onset of
cooling at these
levels, and signal
the arrival of newly
ventilated waters in
response to climate
changes farther
north.

powerful tools: They provide direct visualization
of climate changes, and they trace the pathways
along which ocean climate perturbations propa-
gate into the oceans. That is, changes in charac-
teristics and volumes of water masses due to
climate variations ultimately influence deeper,
more isolated regions of the oceans. How these
changes move to the deep ocean from regions of
contact with the atmosphere must be under-
stood. This process is an important mechanism
whereby the oceans couple to the atmosphere on
longer time scales, and probably plays a role in
determining the interannual to decadal variations
in global climate.

Observations of tracer distributions provide
information on processes that are very difficult
to observe any other way. Mixing and dilution,
for example, play a dominant role in the south-
ward transport of material along the deep west-
ern boundary of the North Atlantic. It has long
been known that newly ventilated North Atlantic

Deep Water travels southward in a concentrated
current, hugging the western edge of the Atlantic
basin. Although direct current measurements
indicate velocities of tens of centimeters per
second, the actual average propagation rate of
tracers down the western boundary is only one
or two centimeters per second. This is because
there is a tremendous amount of entrainment
and mixing associated with water recirculating
within the rest of the basin. The mixing slows the
progress of tracers and climate anomalies. Tran-
sient tracers are perhaps the only tools for mea-
suring the amount of interior exchange and
downstream propagation rates.

Tritium: the Cold War Legacy

It is said that every cloud has a silver lining,
and that seems to be true even if it is a mush-
room cloud. Although the atmospheric testing of
nuclear weapons re-
leased alarming
amounts of radioactive
debris into the environ-
ment, and caused un-
told damage to both
the environment and
human health, it also
provided oceanogra-
phers with some
unique tools to study
ocean circulation and
ventilation. We have
had the opportunity to
observe the entry of
these substances into
the oceans, and to see
how they are moved
around by physical,
chemical, and biological processes.

One of these bomb-produced tracers is tri-
tium, a radioactive isotope of hydrogen. There is
very little natural tritium in the world (the entire
global inventory would only weigh a few kilo-
grams!), but several hundred kilograms were
created in the hydrogen bomb explosions. This
tritium was injected into the stratosphere, almost
immediately oxidized to water, and fairly rapidly
rained out onto the surface of the earth. Since it
is chemically bound up as part of the water mol-
ecule (it is, after all, hydrogen), tritium is an
ideal tracer of the hydrologic system. Further,
since the bulk of the atmospheric weapons test-
ing was done in the northern hemisphere, most
of the tritium was deposited in the high latitude
northern hemisphere. Thus it is an ideal tracer of
ventilation and water mass formation in the
North Atlantic.

We can see this in the figure on the previous
page, which provides a bird's eye view of tritium

1985

/INTER 1996

distribution in the North Atlantic in 1981. Picture
yourself hovering somewhere over Norway, at an
altitude of a few hundred miles. You are looking
southwestward and downward on an isosurface
of tritium in the North Atlantic. An isosurface is
the two-dimensional analog of a contour line on
a map, shown here in a three-dimensional ocean.
The blue "blanket" you see in the picture is the
1 Tritium Unit isosurface, where we find a ratio
of 1 tritium atom to 100,000,000,000,000,000
hydrogen atoms. This isosurface corresponds to
about 5 or 10 percent of the maximum surface
water concentrations of tritium during the mid
1960s, when it was at its peak. Ail the water be-
neath this blanket has remained relatively iso-
lated from the tritium invasion, and, conversely,
all the water above this blanket has been at the
sea surface, exposed to the atmosphere, and thus
ventilated or otherwise involved in interaction
with the surface ocean in the 15 to 20 years be-
tween the bomb tests and this survey.

The blanket lies at about 500 to 1,000 meters
depth in the subtropics, but deepens to 1,500 to
2,000 meters just south of the Gulf Stream off the
New England coast. This is the effect of the Gulf
Stream recirculation, a tight gyre that effectively
ventilates the upper part of the ocean in this
region. There is also a fold extending southward
from this region at about 1,200 meters depth,
marking the intrusion of intermediate depth
waters toward the tropics. Most notably, however,
is the dramatic dive that the "blanket" takes to
the north, disappearing into the ocean floor. The
track along which this happens parallels the Gulf
Stream Extension/North Atlantic Drift. All the
waters north of this line have been ventilated to
the ocean floor on 10 to 20 year time scales. This
is a powerful statement regarding the time scales
of ocean ventilation, and has profound implica-
tions concerning how rapidly climatic variations
can propagate through the oceans.

We can see this in yet another way. The figure
at left is a plot of tritium vs. depth for a time
series near Bermuda from the late 1960s to the
late 1980s. The tritium data has been adjusted for
decay to the same date (January 1 , 1 981 ) to allow
us to more clearly see the time trends. The most
obvious trend is the downward propagation of a
tritium maximum (deeper red) from the surface
in the late 1960s to about 400 meters depth in
the late 1980s, a downward penetration rate of
18 meters per year. However, from the perspective
of climate change, the most important signal is
the sudden increase in tritium (green) at about
1,500 meters in the late 1970s and at 2,500
meters in the mid 1980s. Both of these features
correspond to the onset of significant cooling
events seen in the deep water at this station. The
correspondence between the sudden tritium

67 66 65 64

62

10 9

\

March 1991

0 50 100 150 200 250 0 50 100 150 200 250

3.0 2.5 20 16 14 12 1.0 0.9 0,8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

J p/como/es/

4500

5000

SO

Distance (kilometers)

increases and cooling offsets is highly suggestive
of significant changes in deep water ventilation at
those times. Indeed, if we could take another
"picture" of the tritium blanket shown in the
figure on page 29, we would see that it has been
pushed further downward and southward by a
climatic event.

Another Kind of Cold War Legacy

The development and manufacture of chlorof-
luorocarbons (CFCs) for use in refrigerators and
air conditioners (and later as spray can propel-
lants) seemed like a good thing at the time: CFCs
were easy to manufacture, nontoxic, chemically
inert, and stable. Production, use, and ultimate
release of CFCs into the atmosphere increased
annually in an almost exponential fashion from
their introduction in the 1930s. The unfortunate
influence of these compounds on the ozone
layer, however, has lead to international reduc-
tion in their manufacture and use. In 1990, the
US and 55 other nations agreed to end CFC pro-
duction by 2000. Meanwhile, however, oceanog-

Four chlorofluoro-
carbon (CFC) sec-
tions taken at vari-
ous times along
55° W south of the
Grand Banks. Note
the absence of any
significant CFC sig-
nal at the depth of
the Labrador Sea
Water (about 1,500
meters depth) in
1983, but the sud-
den flooding of
these depths with
CFCs in the later
sections, as newly
formed Labrador
Sea Water flows
around the Grand
Banks and into the
Sargasso Sea. These
changes corre-
spond to the tri-
tium increases seen
in the Bermuda
time series (see fig-
ure opposite).

raphers have found another "silver lining" in this
ecological cloud, which has permitted us to study
ocean ventilation and circulation: Waters that have
been in contact with the atmosphere in the past
few decades have taken up some of these com-
pounds, and hence have been labeled in a distinc-
tive way. Thus the distribution of CFCs in ocean
water provides us with important clues regarding
the pathways of newly formed water masses.

In the 1970s and early 1980s, there was not
much winter time convection and formation of
Labrador Sea Water. In fact, tracer sections (lines
of stations) taken across the Deep Western
Boundary Current to the south showed a decided
lack of newly venti-

A

2000 4000 6000 8000

Down Stream Distance (kilometers)

10000

The downstream
evolution of tri-
tium (upper panel,
in tritium units)
and tritium-he-
lium age (lower
panel, in years) vs.
distance in the
core of the deep
western boundary
current. Note the
approximately ten-
fold reduction in
tritium content in
the Deep Western
Boundary Current
core due to dilu-
tion with older,
surrounding deep
water, and the lin-
ear increase in age
downstream. The
age increase is con-
sistent with a
mean speed of
about 1.5 centime-
ters per second.

lated Labrador Sea
Water. This hiatus is
somehow related to
the complex interplay
between changing
climatic conditions in
the area and the fresh-
water outflow and
budgets of the Arctic.
The late 1980s and
early 1990s have her-
alded a dramatic
change in climatic
conditions in the La-
brador Sea. These
changes have resulted
in the production of a
large amount of Labra-
dor Sea Water, which is

now invading the ocean interior. You can see this
beginning to happen in the figure on the previ-
ous page, a time series of CFC sections made
along 55° W south of the Grand Banks. Labrador
Sea Water occurs in these sections at a depth of
about 1,500-2,000 meters (the middle heavy
dashed line). Notice that there was very little
CFC- 11 in this water in the 1983, although there
is a CFC tongue at a shallower level characteristic
of waters that are formed in the southeastern
corner of the Labrador Sea (the shallowest heavy
dashed line). Below the Labrador Sea Water core,
there is a weak but detectable CFC core in waters
characteristic of Denmark Straits Overflow water
(marked here by the deepest dashed line). Com-
bined, these three water masses form the Deep
Western Boundary Current system of the North
Atlantic, and are responsible for the southward
transport of newly ventilated waters.

However, in the 1990s, there is a sudden in-
crease in the amounts of CFCs in the Labrador
Sea Water core, as well as a steady increase in the
deeper core associated with waters from the Den-
mark Straits overflow. This increase is continuing
through the 1990s and is direct evidence of the

newly ventilated waters' arrival. This again is a
signature of the penetration of climatic anoma-
lies into the ocean interior.

While the 55° W sections capture the invasion
of the newly formed waters into the northern
Sargasso Sea, the pathway southward is not a
simple one. The Deep Western Boundary Current
is not a continuous ribbon of flow extending all
the way from the Grand Banks to the equator,
but rather a composite consisting of series of
interconnected gyres lined up along its path. A
fluid parcel that passes by the Grand Banks may
spend most of its time looping through these
gyres, and only part of its time in the Deep West-
ern Boundary Current. This is why the mean
propagation speed of tracers down the western
boundary is only 1 to 2 centimeters per second,
while velocities in the actual core of the Deep
Western Boundary Current are 10 to 20 centime-
ters per second.

The figure at left is a plot of tritium and tri-
tium-helium age in the core of the Deep Water
Boundary Current vs. distance downstream from
its origin. We see that the core becomes progres-
sively older (about 20 years in 10,000 kilometers)
corresponding to a speed of about 1.5 centime-
ters per second. Notably, the tritium concentra-
tion in the core decreases more than tenfold
downstream, partly due to decay, but largely due
to dilution and mixing with older, tritium-free
waters. This process of mixing is an important
mechanism for ventilating the abyssal ocean. It's
through this process that climate anomalies make
their way into the deep ocean.

Thus the transient tracers are telling us some-
thing very important about the propagation of
climatic changes into the deep ocean. They high-
light the pathways and give us the rates of move-
ment and dilution in the ocean. This information
is valuable because the ocean provides the long
term memory and feedback in the coupled
ocean-atmosphere-climate system, and is the key
to beginning to make long term predictions in
our ever changing climate.

The research discussed in this article was supported by
the National Science Foundation and the Office of Naval
Research.

Bill lenkins nailed life as a nuclear physicist but drifted into
environmental sciences out of a secret yearning to become a
forest ranger. Not having a good sense of direction, and tearing
bliicl; flies, however, he ended up as an oceanographer on Cape
Cod. He joined the WHO/ Chemistry Department (now the
Department of Marine Chemistry and Geochemistry) in 1974.

Bill Smethie's interest in oceanography began during childhood
summers spent at his grandfather's log cabin an the Virginia
side of the Potomac River. He embarked on his first oceano-
graphic cruise at age 7 when he attached a makeshift sail to his
inner tube ami set sail for the other side of the river. His doting
tuints prevented him from making it to the other side, but ever
since he has had a never-ending curiosity for what lies beyond
the horizon. He joined the Geochemistry Division of Lamont-
Doherty Geological Obseivatoiy in 1979.

L/WINTER 1996

New Data on Deep Sea Turbulence
Shed Light on Vertical Mixing

Rough Seafloor Topography Has Far-Reaching Effect

John M. Toole

Senior Scientist, Physical Oceanography Department

The global thermohaline circulation is basi-
cally a wholesale vertical overturning of
the sea, driven by heating and cooling,
precipitation and evaporation. (Changes in
temperature=f/!crmo, changes in salinity=/w/mi'.)
Bottom waters move equatorward from their
high-latitude regions of formation (the cold limb
of the circulation), upwell, and return poleward
at intermediate depth and/or the surface (the
warm limb). As the bottom waters are colder
than the overlying waters, this circulation is re-
sponsible for a large fraction of the ocean's
poleward heat transport. In addition, these flows
often redistribute fresh water, as the northward
and southward moving waters generally have
different salinities.

These oceanic heat and water transports play a
significant role in Earth's climate. The earth gains
heat from the sun at low latitude, and radiates
heat back to space about the poles. To maintain a
quasi-steady state, the ocean-atmosphere system
must carry heat from low to high latitude. At
mid-latitudes, where the poleward heat flux is
maximum, the oceanic and atmospheric contri-
butions are about equal. One component of the
atmospheric heat transport involves evaporation,
water vapor transport, and its subsequent con-
densation. Net north/south water vapor transport
in the atmosphere is balanced by liquid water
transport by rivers and ocean currents.

For almost 200 years, since the writing of
Count Rumford in 1797, there has been a basic
understanding of the cold limb of the thermoha-
line circulation. The combination of atmospheric
cooling, evaporation, and, in some cases, salt
rejection during the formation of sea ice causes
surface waters at high latitudes to become suffi-
ciently dense that they sink to the ocean bottom.
These newly formed deep waters subsequently
spread horizontally within the constraints of the
seafloor's bathymetry to renew the deep waters
found in the interiors of the world's oceans.
There are two principal formation sites for dense
bottom water: the Greenland and Norwegian
Seas of the northern North Atlantic Ocean, and

around the Antarctic continent, particularly
within the Weddell Sea. Together, these source
regions export some 20 to 30 million cubic
meters per second of bottom water to the other
ocean basins. (For comparison, the chiefly wind-
driven Gulf Stream, Kuroshio, and Agulhas Cur-
rents carry in excess of 100 million cubic meters
per second within horizontal circulations.)

The processes involved with the return limb of

The principal tool for work described in this article is the high resolution
profiler. It records temperature, salinity, pressure, and horizontal velocity 10
times per second on descent to the ocean floor, then returns to the surface. For a
detailed discussion of the instrument's development, see the Spring/Summer
1995 issue of Oceanm.

PACIFIC

Circulation sche-
matic of the
world's major wa-
ter masses (also see
inside front cover).
Of concern here
are the mixing pro-
cesses that modify
the bottom and
deep waters within
the cold-to-warm
limbs of the over-
turning circulation.

Subantarctic Mode Water
Antarctic Intermediate Water
Red Sea Overflow Water
Antarctic Bottom Water
North Pacific Deep Water
Antarctic Circumpolar Current
Grcumpolar Deep Water
North Atlantic Deep Water
Upper Intermediate Water
Indian Ocean Deep Water

the thermohaline circulation — the transforma-
tion of these bottom waters to lower density, and
their upwelling and eventual return to the high-
latitude cooling zones — are less well understood.
An upwelling of deep and bottom waters is be-
lieved to be fed by the continual supply of new
bottom water: Dense new waters intrude below
older waters and force them upwards. The bot-
tom water source strength of 20 to 30 million
cubic meters per second translates into a globally
averaged upwelling rate at mid-ocean depth of
about 3 meters per year. This upwelling has both
dynamical and thermodynamical implications.

To maintain a steady-state temperature distri-
bution in the face of this upwelling of cold water,
a compensating warming is required. This warm-
ing may be accomplished by internal mixing of
the deep ocean. Models exploring the thermody-
namic balance between the downward diffusion
of heat associated with mixing by turbulent ed-
dies and the upwelling of cold water were pub-
lished by Klaus Wyrtki (University of Hawaii)
and Walter Munk (Scripps Institution of Ocean-
ography) in the mid 1960s. At about the same
time Wyrtki's and Munk's papers appeared,
Henry Stommel, considering the dynamical ef-
fects of deep upwelling, proposed the existence of
abyssal gyre circulations involving poleward deep
flow in the ocean interiors fed by a series of west-
ern boundary currents. These boundary flows
ultimately connect to the high-latitude bottom
water formation sites. Twenty years later Frank
Bryan (National Center for Atmospheric Re-
search) published a study of an idealized, three-
dimensional ocean model showing a direct rela-
tionship between the intensity of the vertical
mixing and the strength of the thermohaline
overturning circulation. These theoretical ideas
linking diffusion, upwelling, and the deep cur-

rent systems have guided research on abyssal
circulation for the past three decades.

But how much vertical diffusion is there in
the oceans, and what processes sustain it?
Munk's application of his model to data from
the North Pacific Ocean required a downward
diffusive heat flux about 1,000 times larger than
that caused by molecular diffusion (the process
whereby differences in temperature or concentra-
tion of a dissolved substance are removed by the
random motion of molecules). More recent
studies concerning vertical diffusive heat fluxes
in semi-enclosed basins also required downward
diffusive heat fluxes thousands of times greater
than those due directly to molecular diffusion.
All of the researchers involved invoke turbulent
mixing as the mechanism supporting these large
diffusive heat fluxes.

Ocean turbulence is the focus of a subgroup of
physical oceanographers specializing in micro-
structure, that is, temperature and velocity struc-
tures occurring at spatial scales directly influ-
enced by seawater's molecular viscosity and
thermal diffusivity — typically around one centi-
meter. These scientists have extensively sampled
the upper ocean in recent years. Apart from the
surface layer (which is actively mixed by wind
and waves), the shallow ocean margins, and
highly sheared flows like the equatorial undercur-
rent, the microstructure data suggest turbulent
diffusive fluxes some ten times smaller than the
studies mentioned above. This seeming discrep-
ancy caused some to question the models used to
deduce the intensity of vertical diffusion from
microstructure data, and whether sufficient data
had been gathered to adequately describe ocean
microstructure. Relatively weak mixing in the
upper ocean away from boundaries was, however,
recently confirmed by a nontoxic chemical tracer
release experiment in the Northeast Atlantic led
by Jim Led well.

The apparent contradiction between micro-
structure-based and indirectly determined esti-
mates of vertical diffusion might actually reflect a
real difference with depth in the ocean. Part of
the problem is that the indirect estimates of verti-
cal mixing have been derived tor the deep ocean,
while the bulk of the microstructure observations
are from the top 1 kilometer of the ocean. Ray
Schmitt, Kurt Polzin, and I have recently ad-
dressed this issue with a series of cruises on
which we acquired full-ocean-depth profiles of
temperature and velocity microstructure. We find
evidence of enhanced turbulent mixing in the
deep ocean near the bottom, particularly in re-
gions where the bottom is rough. The zone of
enhanced mixing extends upward to several hun-
dred meters above the bottom, a span much
greater than that of the traditional bottom

34 «' FALL/WINTER 1996

boundary layer, a roughly 10-meter-
thick, vertically homogenized layer
that is maintained by bottom-gen-
erated turbulence. Our data also
show strong internal waves at these
sites, and we believe the enhanced
mixing is sustained by the breaking
of these internal waves, which are
both generated at and reflected
from the rough bottom.

These observations also docu-
ment striking horizontal patterns
in the turbulent mixing at depth.
Our current study (a joint micro-
structure-tracer experiment in
collaboration with Jim Ledwell)
is now underway in the Brazil
Basin, the region where Nelson
Hogg and colleagues inferred
significant vertical diffusion from
a heat budget for the bottom wa-
ters. In the interior of the basin
where the bottom is smooth, the
microstructure data imply turbu-
lent fluxes less than a tenth of
Hogg and colleagues' basin-aver-
aged value. In contrast, above the
rough flanks of the Mid-Atlantic
Ridge in the eastern third of the
basin, we deduce turbulent fluxes
greater than their figure.

We find that the horizontally
averaged turbulent heat flux for
our study region, based on the
microstructure data now in hand,
is in near accord with that derived
from the bottom water heat bud-
get. Our results suggest that vertical
diffusion in the deep ocean is
dominated by turbulent mixing
near rough bathymetric structures,
a refinement of Munk's hypothesis that it occurs
generally near the bottom. Greater average tur-
bulent fluxes may be achieved at depth than in
the upper ocean because a larger fraction of the
deep ocean is in close proximity to the bottom.
Spatially variable mixing in turn implies exist-
ence of horizontal circulations to distribute
modified waters from these mixing zones
throughout deep basins. Moreover, given the
dynamical links between mixing, upwelling, and
circulation, our findings hint that the deep gyres
predicted by Stommel might be highly distorted
in the real ocean.

The scientific community is just beginning to
document the intensity and patterns of mixing in
the ocean abyss. It is not surprising that mixing in
ocean climate models has so far been generally
taken as spatially uniform. Much work remains

,ent mixing in surface layer driven by win,
internal wavebreaking supports fhe deeper turbulent mixing

energy propagates away from peaks of
bathymetry as narrow beams of
internal waves

• possibly larger amplitudes and
more irregular motion where
'—ams intersect

little mixing in
ocean interior

tense mixing over
lugh bathymetry

to be done, both observational and theoretical, to
fully understand the role of turbulent mixing in
the ocean's thermohaline circulation.

The research discussed in this article was supported by the
National Science Foundation Initial development of the
High Resolution Profiler was supported by the Department
of Defense and the Office of Naval Research.

Attraction to the sea and ocean science began for lohn Toole
with a keen interest in sailing. He maintains an eclectic
research program at WHO/ that includes study of basin-scale
circulations and the processes of ocean mixing. Developing
understanding of the cold-to-warm limb of the thermohaline
circulation represents a synthesis of research supported by
grants from the National Science Foundation and the Office
of Naval Research. With WHO/ colleagues and his wife (and
chief foredeck crew), he also continues to campaign sailing
race courses through the summer, as research cruises and
meetings permit.

A schematic draw-
ing of turbulent
processes at work
in the ocean.

Computer Modelers

Simulate Real and Potential Climate,

Work Toward Prediction

Combining Equations and Data Pushes Computers' Limits

Rui Xin Huang

Associate Scientist, Physical Oceanography Department

Jiayan Yang

Assistant Scientist, Physical Oceanography Department

Although weather forecasting is accepted by
the public as part of daily life, oceanic
forecasting is not yet so advanced. There
are, however, successful examples of oceanic
forecasting — one is the newly developed skill to
predict El Nino/Southern Oscillation (ENSO)
events, largely due to improvements in ocean
modeling (see following article).

In 1982 and 1983, Eastern Australia and Indo-
nesia experienced the century's worst drought,
which led to devastation in agricultural regions
and rain forests, and even to loss of hundreds of

Authors Rui Xin
Huang, right, and

Jiayan Yang col-
laborate on ocean
process study and
climate prediction
models.

lives. These were just some regional impacts of
what is now known as the most severe ENSO
event on record, an anomalous warming of sea
surface temperature in the eastern Tropical Pacific
Ocean that occurs once every three to seven years.
The change in oceanic conditions associated with
ENSO is usually accompanied by atmospheric
shifts, and together these phenomena lead to
droughts in some parts of the world and flooding
in others. It is estimated that total worldwide
damage caused by the 1982-83 ENSO was more
than $10 billion.

Another climate variation, the North Atlantic

Oscillation (NAO), a shift of atmospheric pres-
sure fields between Iceland (65° N) and the
Azores (40° N) on decadal time scales, is known
to change weather conditions in Europe and
North America (see Box on page 13). ENSO and
NAO are just two examples of how natural cli-
matic fluctuations can dramatically affect the
world economy and our daily lives. Earth's cli-
mate changes ceaselessly, and it will surely con-
tinue to evolve, possibly in a more complicated
manner due to increasing atmospheric concentra-
tions of greenhouse gases. Thus there are pressing
reasons to improve our understanding of severe
climate variations, such as ENSO and NAO
events, and even to predict them before they oc-
cur so that the public can be informed and policy
makers can prepare for possible natural disasters.
Because the future is unobservable, we must rely
on numerical models for such forecasting.

Geologic studies of Earth's history show that
the world ocean has changed profoundly over
time. Modern observations indicate that there
have been noticeable changes in world ocean
circulation even during recent decades. As our
knowledge advances, so does our understanding
of the ocean's importance in the climate system.
Driven by wind stress as well as heat and freshwa-
ter fluxes, oceanic currents redistribute heat
across the globe and regulate our climate. The
ocean's enormous capacity to store heat also
buffers climate changes. Since the ocean and the
atmosphere exchange momentum, heat, and
fresh water across their interfaces, variation in
one fluid system can lead to changes in the other,
often in a chain reaction that amplifies initially
small deviations. Many climate phenomena, such
as ENSO, result from such interactions, which
can occur over a wide spectrum of time scales. So,
even though the weather can be forecast for a
week without considering oceanic circulation,
climate on time scales longer than a month must
include the ocean.

Temporal evolution and spatial variation of
the ocean are constrained by physical laws and
such external forcing fields as wind stresses and
surface buoyancy fluxes. The essence of climate

FALL/WINTER 1996

Satellite

Land-based

Forecasting

Facility

modeling is to integrate the dynamic equations
for these climate components forward in time,
starting with conditions that are often based on
actual observations. The basic idea of computer
simulation is to organize physical equations into
a net of grids arranged to cover a spatial domain,
such as the tropical Pacific Ocean for ENSO pre-
dictions, or the global oceans for carbon cycle
assessments, and then predict the climatic state at
each grid in the future based on its initial condi-
tion and its subsequent interactions with sur-
rounding grids as the conditions in each change.
Though the evolution of a climate event is
unrepeatable and beyond our control, computer
simulations can be repeated many times by vary-
ing the mathematic representation of conditions
and forces at work in each grid. Thus, numerical
models are very powerful tools for testing scien-
tific hypotheses and for examining important
climate processes.

One good example is a study by Frank Bryan,
who conducted numerical experiments in the
early 1980s while a graduate student at the Geo-
physical Fluid Dynamics Laboratory in Princeton,
N(. By running an idealized model for the Atlan-
tic Ocean, he showed that deep water formation
in the North Atlantic could be shut off by a
strong salinity perturbation in the subpolar ba-
sin. If such changes were to take place, the North
Atlantic's poleward heat flux would be substan-
tially reduced, and the European climate would
be remarkably less mild. His modeling results are

consistent with a paleoclimatic record that shows
deep water formation was interrupted about
12,000 years ago in an event known as the
Younger Dryas when melted glacial water flooded
the subpolar North Atlantic Ocean, resulting in
Northern Hemisphere cooling and even reduc-
tion in the deglaciation process.

The accuracy of a numerical model depends
on how well and how realistically it approxi-
mates boundary conditions, external forcing, and
key processes that govern the real climate system.
The ideal model would use a very fine spatial grid
and a small integration time step to resolve spa-
tial and temporal structures of all important
physical processes in the system. For instance,
mesoscale eddies, whirling parcels of fluid typi-
cally 40 to 200 kilometers wide in the subtropics,
play significant roles in redistributing momen-
tum, heat, salts, and other dissolved matter in the
ocean. Excluding such processes will definitely
lead to inaccurate model representations of the
oceans. However, most of the current climate
models use resolutions (spacing between two
adjacent grids) on the order of 100 kilometers or
coarser, too coarse to explicitly resolve the spatial
structures of mesoscale eddies.

Increasing the horizontal and vertical resolu-
tion of the models 10 times requires increasing
the number of grid points 1,000 times. In addi-
tion, the allowable time step for maintaining the
numerical calculation's stability will be at least 10
times smaller (the time step should be smaller

The authors envi-
sion a forecasting
center that would
receive near-real-
time data from a
variety of sources
by satellite trans-
mission for con-
stant updating of
ocean climate
models. The data
sources might in-
clude those shown
here. The surface
mooring transmits
meteorological
data as well as in-
formation from the
string of instru-
ments below it.
Slocum, described
on page 6, surfaces
once a week or so
at the top of its tra-
jectory to transmit
temperature and
salinity records
from ocean depths,
and SeaSoar data
on a variety of up-
per ocean charac-
teristics is beamed
from the ship. Data
collected by satel-
lites would include
sea surface eleva-
tion, wind stress,
temperature, and
perhaps salinity.

OCEANUS

than the grid size divided by the maximum veloc-
ity). Thus, we need a computer that is 10,000
times faster and has 1,000 times more memory.
For such high resolution, even the fastest com-
puters, such as the 128-processor CMS computer,
which reaches a speed of 3 gigaflops (3 billion
float point operations per second), cannot yet
accommodate a long-term global simulation.
Thus one key aspect of climate modeling involves
how to accurately represent important processes
that cannot be explicitly resolved due to model
resolution limits. This is called the subgrid-scale
parametrization.

Another important process that requires care is
the direction along which mixing occurs. In
many ocean general circulation models
(OGCMs), especially those formulated in fixed
spatial grids, mixing in a model grid is repre-
sented by some form of averaging its properties
with those in surrounding grids. In the real
world, mixing is most likely to occur along con-
stant density surfaces so that the mixing processes
do not work against the buoyancy. Recent work
by lames McWilliams and Peter Gent and their
colleagues at the National Center for Atmo-
spheric Research has significantly improved the
performance of ocean models that use spatially
fixed grids.

Rapidly developing computer technology allows
climate modelers to work at ever finer resolution
as they aim explicitly to model the structures and
temporal evolution of eddies. For instance, Albert
Semtner and his colleagues at the Naval Postgradu-
ate School in Monterey, California, have used an
OGCM with a horizontal resolution of about 10
kilometers to study circulation in the Arctic Ocean
and the Greenland and Norwegian Seas. Their
eddy-resolving model captures many observed
frontal structures, such as sharply defined features
often associated with strong jets, that coarse reso-
lution models have not been able to capture.

There are two major categories in ocean cli-
mate modeling, process studies and climate pre-
dictions. Process studies aim to understand im-
portant processes that operate the real climate
system, to identify mechanisms that give rise to
climate variations, or to explain particular pat-
terns observed in the real world. Such studies
often involve a hierarchy of models, from simple
ones to full, three-dimensional models. Climate
predictions, like numerical weather forecasts,
attempt to determine future climate states based
on available knowledge about how the climate
system evolves in time. Climate predictions al-
ways benefit from progress in process studies. For
instance, tremendous advances in process studies
of tropical air-sea interactions in the 1980s led to
great success in predicting ENSO events a year
ahead of time. Simple models play very active

roles in process studies. For example, the late
Henry Stommel used a very simple two-box
model to elucidate how the distinct difference
between air-sea fluxes of heat and fresh water lead
to multiple equilibrium states in the oceanic
thermohaline circulation. This seminal work laid
the foundation for our understanding of the sta-
bility and variability of the Atlantic overturning
circulation. More comprehensive models, which
can resolve mesoscale eddies and include ocean-
atmosphere interactions, have been used to verify
Stommel's work and to gain further understand-
ing of the oceanic thermohaline circulation.

Climate prediction models must prove they
can describe observed climate evolutions in the
past before they can be trusted for future predic-
tions. Therefore, modelers need observations to
calibrate and verify their models. Unlike atmo-
spheric data sets, oceanic observations have rela-
tively short records and sparse coverage in space.
New observing technologies, like satellite remote
sensing, acoustic tomography, and the long-lived
floats that Ray Schmitt describes on page 6, will
certainly expand the observing capacity and sup-
port climate modeling. A global observational
network is likely to be a combination of satellite-
borne instruments (which can measure sea sur-
face elevation, wind stress, temperature, and per-
haps salinity); automatic instruments, such as
buoys and data-transmitting floats; and tradi-
tional shipboard instruments. Data collected by
these instruments will be sent to a land-based
forecasting center, where the most powerful
supercomputers will merge current oceanographic
data and forecast oceanic conditions for the near
future. With the rapid advance of computer tech-
nology and our understanding of ocean physics,
oceanic forecasting will eventually become a
reality— perhaps early in the 21st century, marine
and climate forecasting will become routine.
The research discussed in this article was supported by the
National Science Foundation and the National Oceanic
and Atmospheric Amdinistration's Global Climate
Change Program.

Rui Xin Huang's primary research interest is large-scale oceanic
circulation, including wind-driven g}>res and thermohaline
emulation. When he was in high school, his dream was to
become an inventor like Edison. Through a long and winding
road, he came to Woods Hole, and found oceanography an
exciting field. He also likes swimming, gardening, and, above
all, pilules and games.

During his graduate student and postdoctoral years, liayan Yang
was interested mainly in tropical air-sea interaction. He decided
to take "a short break" away from the tropics to do a small,
high-latitude oceanography project when he was a postdoctoral
fellow at the University of California, Los Angeles. He has
stayed in high-latitude oceanography eivr since. He moved from
Los Angeles to Neu* England to get a bit closer to (though still
far way from) sea-ice margins. In his leisure time, he likes
hiking, swimming, biking, and karaoke.

38 • FALL/WINTER 1996

The El Nino/Southern
Oscillation Phenomenon

Seeking Its"Trigger" and Working Toward Prediction

Lewis M. Rothstein

Associate Professor, Graduate School of Oceanography,
University of Rhode Island

Dake Chen

Research Scientist, Lamont-Doherty Earth
Observatory, Columbia University

The El Nino/Southern Oscillation (ENSO)
phenomenon, an eastward shift of warm
water in the tropical Pacific and associated
effects on the atmosphere, is at the heart of glo-
bal interannual climate variability. The just com-
pleted, decade-long Tropical Ocean/Global At-
mosphere (TOGA) program was dedicated to
understanding and working toward predicting
ENSO by bringing together oceanographers and
atmospheric scientists in a coordinated observa-
tional and numerical modeling research pro-
gram. TOGA has not answered all the questions:
We have not uncovered the physical mechanisms
of the elusive ENSO "trigger" nor have our best
coupled air/sea numerical models been as suc-
cessful in predicting the rather irregular ENSO
signal of the 1990s as uhey were in predicting the
regular events of the 1 980s and hindcasting the
events of the late 1960s through the 1970s.

Prediction is the ultimate goal of ENSO re-
search. It is also the ultimate test for an ENSO
model and the theory underlying the model.
During the last decade, a number of forecast
models have shown predictive skills in both ret-
rospective and real time forecasting, and they are
now being used for routine ENSO prediction.
Nevertheless, the skill of even the best available
models is far from perfect, and there is still con-
siderable room for improvement in modeling,
observation, and forecasting techniques.

Factors that limit the current skill of ENSO
forecasts include:

• an inherent limit to predictability because of
the chaotic and random nature of the natural
system,

• model flaws such as oversimplified physics,

• gaps in the observing system, and

• flaws in the way the data is used (data assimila-
tion and initialization).

It seems likely that the inherent predictability
limit for ENSO is years rather than weeks or
months, though more theoretical study is needed
in this area. The observing system is improving,
but still far from satisfactory. Thus a challenge
facing the modelers is to improve model forecasts
by making the most reasonable and efficient use
of available data.

In the past few years much effort has been
devoted to assimilating various observational
data into die initial state of forecast models. The
most common approach is to improve die ini-
tial ocean conditions by assimilating observa-
tions of sea surface temperature, thermocline
(region of rapid temperature decline) deptJi, or
sea level into an ocean model prior to coupling
it with an atmosphere model. One problem

Time series of ob-
served (red) and
forecast (blue) El
Nino sea surface
temperature
anomalies. Fore-
casts with 0, 6, 12,
and 18 month
leads are shown in
different panels,
and the observed
anomalies are re-
peated from panel
to panel.

13
O

c

I

2
I

•E

I1"!"1!"1!"1!"1!1"!1"!1"!1"!"1!1"!"1!"

0 Month Lead

I"'!'"!'"!"'!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!'"!

72 74 76 78 '80 '82 '84 '86 '88 '90 '92 '94 '96

Year

OCEANUS

with this approach is that no attention is paid
to the ocean-atmosphere interaction during
initialization, so the coupled system may not be
well balanced initially and may experience a
shock when the forecast starts. A new initializa-
tion/assimilation procedure significantly im-
proves the predictive skill of one of our most
promising coupled models, which was con-
structed by Mark Cane and Steve Zebiak
(Lamont-Doherty Earth Observatory).

In the new methodology the model is initial-
ized in a coupled manner, using a simple data
assimilation scheme in which the coupled model
wind stress anomalies are "nudged" toward ob-

120 E

This diagram illus-
trates weather
pathways in the
North Pacific sub-
tropical/tropical
upper ocean and
the main horizon-
tal gyres and me-
ridional-vertical
cells of the region's
ocean circulation.

overpredicted, and the short warm episodes in
1993 and late 1994 are missed.

Although the predictive skill of this model is
most likely limited by its highly reduced physics,
the skill of more sophisticated coupled ocean-
atmosphere general circulation models presently
does not exceed that of the model described
above, at least in terms of the tropical Pacific sea
surface temperature. In order to predict the glo-
bal impact of ENSO, a two-tiered approach
appears to be reasonable: A physics-simplified,
coupled model is first used to predict tropical sea
surface temperature fields, and these fields are
then used as boundary conditions for a more-
complete-physics, global atmospheric
XModel..KurOShi0.... general circulation model to predict

the global distribution of atmospheric
disturbances. Scientists are rigorously
pursuing this kind of research.

A second area of progress concerns
improved understanding of the cou-
pling between different depths and
different regions of the ocean. A
popular ENSO paradigm that emerged
in the late 1980s was based on the
a observed, rather regular rythms of
f ENSO conditions during a span of 25
= years before 1990. The "delayed oscil-
~ lator" mechanism emphasizes east-
ward-propagating equatorial wave
processes, * westward-propagating off-
equatorial signals, and their asymmetric
reflections at eastern and western bound-
aries respectively. Despite the irregularities
in the 1990s ENSO, this wave propagation/
reflection paradigm is still compelling; it can
accomodate irregularities in the ENSO signal by
combining the tropical signal with longer-term
variability in the subtropics.

A number of studies have sought to under-
stand how tropical variability is linked to the mid
latitudes. Ocean circulation may provide the
links via several different pathways that are sum-
marized schematically above. These are not
simple, direct north-south flows; the existence of
vigorous zonal current systems complicate the
picture. In the upper layers of the ocean, up-
welled waters along the equator flow into the
subtropics, mainly through the mid-latitude
western boundary current (the Kuroshio). There
is an additional interior ocean pathway, through
the eastward Subtropical Countercurrent, that
more directly feeds subtropical sites where sur-

*Flow along the equator tends to be trapped there. The
Coriolis force, due to the earth's rotation, turns water that
flows south back to the north and water that flows north
back to the south. Because of this trapping, physical ocean-
ographers recognize the equator as a waveguide, where co-
herent signals or waves can be seen to propagate east-west
for long distances.

Model Eastward

Subtropical
Countercurrent

Re-circulating

Tropical

Gyre

300m

servations. The new procedure improves the
model's predictive ability as measured by a vari-
ety of statistical scores. It also eliminates the so-
called "spring prediction barrier," a marked drop
of skill in forecasts that try to predict across the
boreal spring, lound in many previous ENSO
forecast systems. The success of the new initial-
ization procedure is attributed to its explicit
consideration of ocean-atmosphere coupling,
and the associated reduction of initialization
shock and random noise.

As an example, the forecasts made by the im-
proved model are compared to observations in
the figure on page 39 in terms of the sea surface
temperature anomaly averaged over an area in
the eastern/central equatorial Pacific (5° S to 5° N
and 90° W to 150° W). The model is capable of
forecasting ENSO more than one year in advance.
The large warming and cooling events in the
1980s are particularly well predicted. However,
the model does a poorer job for the 1970s and
1990s: The 1976-77 event is largely

40 « FALL/WINTER 1996

face water moves deeper into the ocean. These
interior pathways are associated with a recirculat-
ing tropical gyre in and just helow the mixed
layer in the northeastern tropics. Below the mixed
layer, thermocline water from the suhtropics to
the tropics zigzags almost zonally across the
basin, succeeding in flowing toward the equator
only along zonally narrow, southward flowing
conduits. The low-latitude western boundary
currents serve as the main southward circuit for
the subducted (water moving from the surface to
depth), subtropical thermocline water.

A model constructed by the authors also indi-
cates important direct flow of thermocline water
through the ocean interior, confined to the far
western Pacific (away from the low-latitude west-
ern boundary currents) along 10° N. These south-
ward flowing waters are then swept eastward by
the North Equatorial Countercurrent, finally
penetrating to the equator in the central and
eastern Pacific. The water pathways in the sub-
tropical thermocline essentially reflect the surface
gyre circulation.

Along with our colleagues Ronghua Zhang
(University of Rhode Island) and Antonio I.
Busalacchi (NASA Goddard Space Flight Center),
we have examined the interannual variability of
these subtropical/tropical pathways and found
important propagating subsurface ENSO signa-
tures in the subtropical Pacific. There appears to
be continual movement of subsurface, basin-scale
anomalies that can then affect sea surface tem-
perature (SST) anomalies, especially in sensitive
regions where the thermocline is shallow. These
SST anomalies can then trigger coupled air/sea
interactions. A clear pattern of moving anomalies
is less obvious at the sea surface. The systematic
subsurface propagation is reminiscent of the
delayed oscillator: eastward along the equator,
westward off the equator with apparent further
propagation along the eastern and western
boundaries. Off the equator, subsurface propaga-
tion of anomaly patterns initiates an SST
anomaly in the North Equatorial Countercurrent
regions of the western Pacific, which then intensi-
fies and moves into the equatorial waveguide,
consistent with the mean water pathways found
above. We speculate that this could be a mecha-
nism for initiating coupled, air-sea interactions
that can begin to evolve as an ENSO event. The
cycling time of the subsurface anomaly patterns
may determine the ENSO's frequency. We look
forward to continuing our investigations to so-
lidify these assertions.

One challenge for the newly established Cli-
mate Variability (CLIVAR) program will be to
uncover the ENSO triggering mechanism and
enable intelligent design of a long-term ocean
and atmosphere monitoring system. CLIVAR is

MBI. WHOI LIBRARY

the oceanographic and atmospheric scientific
community's new program of climate prediction.
Its focus is on understanding the coupled air/sea
system's variability on seasonal-to-interannual-
to-interdecadal time scales for the purpose of
determining predictability, and then prediction.
Those observations would then feed into coupled
air/sea numerical models for the purpose of long
lead time forecasting, much like the present-day
weather forcasting systems. However, the
interannual ENSO signal does not exhibit a
simple ryhthm; there are clearly influences on
longer (decadal) time scales that need to be con-
sidered. There are clues as to what those signals
might be (for example, the North Atlantic Oscil-
lation— see page 13), but we are still in the early
stages of identifying these signals.

The natural system is not easily divided ac-
cording to time scales; it is a fully nonlinear sys-
tem. If we are to understand and eventually pre-
dict global interannual variability, we must not
limit ourselves to monitoring the air/sea system
over a few interannual cycles. Permanent moni-
toring systems are needed. It is the primary
charge of CLIVAR to help design such a monitor-
ing system while, at the same time, supporting
the evolution of the numerical prediction systems
that will issue the forecasts.

The authors' ENSO research is supported by the National
Oceanic and Atmospheric Administration, the TOGA
Program on Seasonal to Interannual Prediction, and the
National Aeronautics and Space Administration

Lew Rothstein started his career as a physical oceanographer on
the beautiful campus of the University of Hawaii. He is now a
professor at the University of Rhode Island and an editor of the
Journal of Geophysical Research. Dake Chen is also a physi-
cal oceanographer by training. He worked with Lew on various
tropical ocean models at the University of Rhode Island before
he joined the senior staff of the Lamont-Doherty Earth Observa-
tory last summer. Both of them are fond of building computer
models of the ocean and atmosphere, not only for scientific
research but also for fun.

Servicing an Auton-
omous Tempera-
ture Line Acquisi-
tion System
(ATLAS) mooring
of the Tropical
Ocean Global At-
mosphere (TOGA)
program's Tropical
Atmosphere-Ocean
(TAO) Array in the
Pacific Ocean.
ATLAS moorings
measure surface
winds, air tempera-
ture, relative hu-
midity, sea surface
temperature, and
subsurface temper-
ature to depths of
500 meters.

OCEANUS

1930

Woods Hole Oceanographic Institution

Woods Hole, MA 02543
508-457-2000