ENSO Science 222 1983a

16 December 1983, Volume 222, Number 4629

Oceanographic EveiDuring El Ni

Mark A. C

In July 1982 conditions in the easternequatorial Pacific were unremarkable;by October the sea-surface temperature(SST) was almost 5°C above normal andsea level at the Galapagos Islands hadrisen by 22 cm (1). The anomalies atdepth were even greater: a huge influx ofwarm water had increased the heat con-tent of the upper ocean at a rate thatexceeded the climatological surface heatflux by a factor of more than 3, and the

of fish and guano birdlocal economy (3). Sciserve the term El Nifno fic events (4). Althoughprompts a definition of ]of conditions off the 'coast, these changes arrectly to changes acrosscal Pacific and indirecthroughout the world'soceans.

Summary. El Ninio events, the most spectacular instances of intercin the ocean, have profound consequences for climate and the ocean1982-1983 El Ninio is perhaps the strongest in this century. El Ninhave followed a predictable pattern, but the recent event differsphysical oceanography of this El Nino is described and compared wevents.

SCIENCE

deviated markedly from the usual pat-tern. Thanks to the efforts of many scien-tists, particularly those in the EPOCSand PEQUOD programs (6), this event isthe most thoroughly documented one to

its date.Interest in the El Ninlo phenomenon

no has intensified with the recognition thatit is part of a global pattern of anomaliesin both the atmosphere and the ocean

-ane (7). Aspects of the atmospheric changes,notably the Southern Oscillation, wereidentified before the turn of the century,but an appreciation of their relation to

Is, crippling the the oceanic El Ninlo has come only in theientists now re- past 20 years (8). It is now generally felt,or these dramat- that the El Nifio/Southern Oscillationhistorical usage (ENSO) cycle involves an essential cou-El Nifio in terms pling between the atmosphere and theSouth American ocean, in which wind changes cause oce-re connected di- anic changes and changes in tropical SST-the entire tropi- affect atmospheric circulation (9). Thisctly to changes article addresses the oceanic part of theatmosphere and cycle.

To show how extraordinary the 1982-1983 El Nifio was, we will first describethe more typical El Nifio and the normal

annual variability oceanic variations. We will then brieflyecosystem. The consider the theoretical explanationso events usually that have been offered before discussingmarkedly. The the 1982-1983 event.

rith that of earlier

The Annual Cycle in the Tropical Pacific

thickness of the warm layer was nowgreater than all previously observed val-ues (2). Temperatures at the SouthAmerican coast were near normal, butwithin a month they too would risesharply. It was now obvious that whathad been labeled a warm event wouldturn out to be a major El Nifio, and anexceedingly odd one at that.

In January of a typical year a south-ward-flowing current brings warm wa-ters to the normally cold coast of Ecua-dor and Peru. The local fishermen namedthis current El Ninlo, in part because ofits proximity to Christmas and in part toacknowledge its benevolence: it oftencarries exotic flora and fauna from itsequatorial origins.At irregular intervals a catastrophic

version of El Nihto occurs. Massivewarming leads to widespread mortality16 DECEMBER 1983

El Nifio events have been documentedas far back as 1726 (5). On average theyoccur about once every 4 years, but theinterval between successive events hasbeen as short as 2 years and as long as 10(5). There are enough similarities amongthe different events to justify a commonname, and a conceptually useful pictureof the typical El Ninlo has emerged.Nevertheless, no two events are precise-ly alike with regard to amplitude, time ofonset, spatial characteristics, or biologi-cal consequences, and aficionados havebeen known to compare different eventsin a manner reminiscent of oenologistsdiscussing vintage years.

It is already clear that the 1982-1983El Nifio will be held in special regard.Not only was the amplitude of its ther-mal signal enormous, but the sequenceof the warming and the time of onset

Sea-surface temperature along theequator in the Pacific is warm in the westand cold in the east (Fig. lA). Thissurface picture also reflects the distribu-tion of oceanic heat content. Almosteverywhere in the ocean the surface wa-ters are well mixed by wind stirring.Along the equator in the Pacific thissurface mixed layer is usually 150 mdeep or deeper in the west, but it be-comes shallower to the east until it es-sentially disappears near the SouthAmerican coast. Sea level is also higherin the west. The trade winds, drivingcurrents westward along the equator,feed and maintain the buildup of excesswarm water on the western side.

Mark A. Cane is an associate professor of ocean-ography at the Center for Meteorology and PhysicalOceanography, Massachusetts Institute of Technol-ogy, Cambridge 02139.

1189

on

Janu

ary

5, 2

010

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

cf) Jul.

<°Apr.(l CH2-J Jan.l H\ 28/

O Apr. -AJan.Oct.

8

cmJul(u1) D 2

Apr.(4

t( H.

co Jul. (0)

a'Apr. (0)Jan. (0 )

Oct. (-1)

Jul. (-1) 0.5

Apr.

Jan.(0

~Oct.(_~Jul.L-1) ~

Figure 1A shows that there is very

little annual variation in the region west

of the date line, which is the largest pool

of very warm water in the world ocean.

Most of the variability in the tropical

Pacific occurs between the South Ameri-

can coast and 140°W from about 3°N to

15°S. This region is also far colder than

the mean for the tropics. Most of the

flow into it is relatively cold water

brought in from the south by the Peru

Current. The major outflow is in the west

through the South Equatorial Current,

which is driven westward by the prevail-

ing trade winds. Water flowing through

this region is strongly heated by the

atmosphere (10), increasing its tempera-

ture by several degrees before it reachesthe western Pacific.

Coastal and equatorial upwelling are

other sources for the cold SST in this

region. Winds along the South Americancoast are southerly, and the Coriolis

effect turns the surface currents off-

shore. The waters leaving the coast are

replaced by water from below. Similarly,easterlies induce equatorial upwelling

because the Coriolis effect turns the wa-ters poleward in both hemispheres, mak-

ing the surface flow divergent at theequator. The upward motions associated

with both forms of upwelling bring thethermocline, which in the tropics is just

beneath the mixed layer, nearer to the

surface. Turbulence in the surface mixed

layer tends to entrain water from below,mixing it rapidly enough to keep the

temperature in the mixed layer uniformwith depth. The tropical thermocline isparticularly sharp, with temperaturechanges as great as 10°C occurring in lessthan 50 m. Hence, differences in thermo-cline depth can result in drastic differ-

26 1 24 2 Fig. 1. Time-longi-

tude section of SST.3&Ni The section follows

26: 4 = the equator to 95°W,2

<< 4t(2 then follows the cli-

matological cold axis-o = to its intersection

with the South Ameri-can coast at 8°S (15).

, 2.0 (A) Mean climatology(4). The interval from

o= =$ 240 to 27°C is shadedH

to show the annualwarm tongue. (B)Composite El Nifioanomalies (15). El

Nifio year is year 0.(C) Anomalies from

1981 to 1983. Note thelarger contour inter-~~~~~val.

140° 1 00°W

ences in the temperature of the waterentrained into the surface mixed layer.

Beginning late in the boreal fall there isan annual warming in the usually coldeastern equatorial Pacific (Fig. IA). Inextratropical latitudes the annual cycleof SST variation is primarily a localthermodynamic response to seasonallychanging solar heating: temperatures risethrough spring and summer, reaching a

maximum in fall as the sun retreats equa-

torward and the net heat balance at theocean surface becomes negative. This isnot the case for the tropical Pacific SSTcycle. The eastern tropical Pacific isheated throughout the year (10), and thetemperature variations are primarily a

consequence of basin-wide ocean

dynamics rather than local thermody-namics. Variations in the surface heatflux are more a response to SST changesthan a cause.

No extant calculations allow a quanti-tative assessment of the possible dynam-ic influences on SST in the eastern equa-

torial Pacific, but there is enough infor-mation to indicate which mechanismsare significant. The southeast trades re-

lax, causing the flow through the PeruCurrent and South Equatorial Current toslow down. Since the rate of surfaceheating does not decrease, the surfacewaters become warmer. The weakeningof the winds also reduces both equatorialand coastal upwelling, diminishing thatsource of cold water.These local factors are aided by re-

mote influences. The waters to the westare always warm, and eastward advec-tion would lead to warming in the east. It

is notable that, during a normal year, theonly time when there is strong eastwardflow along the equator east of the Gala-

pagos is during the winter and earlyspring, when the warming occurs (11). Anumber of studies indicate that thesechanges are the ocean's response to theweakening of the easterly winds alongthe equator in fall and winter, especiallythe winds in the western and centralPacific (12). An additional component ofthis response is a deepening of the ther-mocline, especially along the equatorand the coast, so that water upwelled tothe surface is warmer than before.

The Canonical El Niiio

Unlike the Atlantic and Indian oceans,the magnitude of interannual variabilityin the tropical Pacific is as large as theannual signal. In many respects this vari-ability is bimodal in character, taking onone form in El Nifio years and anotherduring non-El Ninlo years (Fig. 2A) (13).

All El Nifios are different, but recentlya composite picture of the canonicalevent has emerged (14, 15). This com-posite is based on the fact that manyaspects of El Nihlo are closely linked tothe annual cycle (Fig. 1, A and B, andFig. 2). It summarizes our understandingof El Ninlo before the 1982-1983 event.

Prelude. There are stronger than aver-age easterlies in the western equatorialPacific for at least 18 months before astrong El Nifio event. These winds tendto move water from the eastern Pacific tothe west, and consequently sea level isunusually high in the west and low in theeast. At the same time the thermocline inthe west is deeper than average. SST isslightly warmer than average in the farwest and somewhat colder east of 160°E.

Onset. In the fall of the year precedingan El Ninlo there is already a warm SSTanomaly extending across the South Pa-cific between 15°S and 30°S, with anorthward extension across the equatorin the vicinity of the date line. In Sep-tember or October the easterlies begin todiminish along the equator west of thedate line. In response the sea level slopealong the equator begins to relax.

Event. Warming off the coast of SouthAmerica begins in December or January,building in magnitude from January toJune (Figs. 1 and 3). For the first severalmonths it is difficult to distinguish be-tween an El Ninlo and normal seasonalwarming. The anomaly peaks in April,May, or June. At the same time sea levelrises in a narrow region along the SouthAmerican coast and the thermocline be-comes deeper. There is also evidence fora sea level rise along the coast at least asfar north as San Diego (16). There isstrong southward flow at the coast. TheSST anomaly at the equator in the vicini-

on

Janu

ary

5, 2

010

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

ty of the date line persists throughoutthis period (Fig. iB). At this time thereare westerly wind anomalies along theequator from 100°W to 170°E. During the6 months or so after the peak SST at thecoast the warm anomaly spreads north-westward and then westward along theequator until it merges with the anomalyin the central Pacific. Warm water nowgirdles one-fourth of the globe. By fall,SST at the coast is only slightly abovenormal, although the colder isothermsare still significantly deeper than normal(17).The westerly wind changes associated

with El Ninlo reduce the strength of thewestward South Equatorial Current (18).Sea level falls in the west and rises in theeast (Fig. 2). In the 1976 El Nifio themass redistribution took place at an av-erage rate of 2.7 x 107 m3/sec, about halfthe strength of the South Equatorial Cur-rent (19).Mature phase. There is another warm-

ing at the coast beginning about Decem-ber and peaking early in the followingyear (Fig. 3). This time, however, thecoastal SST drops off sharply, perhapsbecoming even colder than normal byMarch. The positive SST anomaly re-mains in the central and eastern Pacificthrough the early part of the year (Fig.

20

10

0

-10

-20E0

0)

co 10

5

0

-5

-10

-15

iB). It disappears as the colder watersspread westward from the coast, reach-ing 140°W by about June and the dateline late in the year. During this periodthe winds relax toward their normal pat-tern and the westward sea level slope isreestablished.

Theory

A complete theory for El Ninlo mustacknowledge that SST changes influencethe atmospheric circulation and accountfor the two-way coupling between theatmosphere and the ocean. The narrowerperspective of the oceanographer isadopted here: we seek to understand theocean's response to prescribed meteoro-logical parameters. Although nonadia-batic processes such as surface heatingand wind stirring influence SST changes,we first consider the adiabatic processesthat alter the ocean's thermal structure.Many of the observed changes, notablythose in coastal sea level, seem largelyaccounted for by a linear theory basedon wind variations (20, 21) that empha-sizes the special role of the equatorialwave guide.The ocean's adiabatic response may

be analyzed as a sum of free and forced

15

10

5

0

-5

-10

10

5

0

-5

waves. For the time and space scales ofimportance in El Nifio events, only twotypes of wave motions are of possibleimportance: Rossby waves and equatori-al Kelvin waves. Away from the equatorthe principal dynamic balance in theocean is the geostrophic one betweenCoriolis and pressure gradient forces;this balance is characteristic of Rossbywaves and strongly constrains theirpropagation speeds. The vanishing of theCoriolis parameter at the equator allowsanother low-frequency wave form, theequatorial Kelvin wave. While theRossby waves generated by the windpropagate westward, the Kelvin wavecarries energy eastward. It is also veryfast: the gravest baroclinic (22) Kelvinwave can cross the Pacific in less than 3months. The most equatorially confinedRossby wave is three times slower, andthe Rossby wave speed decreases as thesquare of the latitude, so mid-latitudewaves would need decades to cross thePacific.The special properties of equatorial

motions are essential to the El Ninlophenomenon. A given change in thewind generates a stronger response atthe equator than elsewhere in the ocean,and equatorial waves are less susceptibleto the destructive influences of friction

J A J 0 J A J 0

Year- 1 El Nin8 year

J A JYear + 1

0 J A J 0 J A J 0 J A J 0Year - 1 El Nino year Year+ 1

Fig. 2. El Nifio signatures. (A) Sea level at Truk (152°E, 7°N) during indicated El Nifio events (top panel), for the composite El Nifno (bottom pan-el, continuous line), and for the annual mean in non-El Nifio years, 1953 to 1976 (bottom panel, dotted line). Note the similarity among El Nifnoevents and their collective difference from the semiannual cycle of non-El Nifio years. (B) Curves as in (A) but for sea level at Callao (79°W,12°S). In the eastern Pacific, El Ninlo events typically appear as an enhancement of the annual cycle.

on

Janu

ary

5, 2

010

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

and mean currents. We have noted thatthe prevailing equatorial easterlies pileup warm water in the west. A relaxationof the winds in the western or centralPacific excites packets of equatorial Kel-vin waves. There is strong observationalevidence (23) that the equatorial waveguide is effective enough to allow suchwaves to cross the Pacific to the easternboundary within a few months, raisingthe sea level there.

In principle, the local setup in re-sponse to alongshore winds can alsoalter the thermal structure at the coast,but in fact there is little change in thecoastal winds during El Nifio except atvery low latitudes (24). Hence, changesin the currents and in all aspects of thethermal structure, including sea leveldisplacement and thermocline depth, de-pend primarily on the amplitude of theincident Kelvin waves.The amplitude of the incident Kelvin

wave is determined by its initial value atthe western side of the Pacific plus theamount added by wind forcing as it prop-agates along the equator. Model calcula-tion (21) indicates that the latter is the

-02

20°N --0.2

20°8

BA March-Oaycb 2

20ON 0 QO

200

B August-October °':

200N .22_

principal influence. Further, the onlyforcing that matters is the zonal windstress within a few hundred kilometers ofthe equator. Figure 4 shows this forcingfor the gravest baroclinic mode as afunction of longitude and time (25). Thedashed lines show the path of a Kelvinwave. It is evident that for the compositeEl Nifio the primary cause of the rise insea level at the beginning of the El Niinoyear is the anomalous westerlies west ofthe date line. The second peak later inthe year is due to the more massivecollapse of the trades east of the dateline.The foregoing discussion is offered as

an explanation of the observed changesin sea level and thermocline depth; itdoes not account for SST anomalies.Data for the 1972 El Ninlo show that,averaged over the whole event, surfaceheating does not contribute to the warm-ing; in fact, because of increased evapo-ration the net flux anomaly is out of theocean (26). It is, of course, possible thatthere is a net surface heating part of thetime. Leetmaa (27) suggested that thismay be the case shortly after onset in a

0 Oh~~~~~~~~~~~~~~~,

200.20

CDcember-February -.02 -0'

100°E 1400 180° 1400 100°W

Fig. 3. Sea-surface temperature anomalies (°C) for the composite El Nifo (15). Contour intervalis 0.20C. (A) March, April, and May average during El Ninlo. (B) Average for the followingAugust through October. (C) Average for the following December through January.

1192

region of the eastern Pacific south of theequator (85°W, 50S), although the heatflux data are very uncertain. His studyalso indicated that some of the warmingresults from a slowing of the currents(28) through the area while the surfaceheating rate is undiminished. What Leet-maa demonstrated most forcefully, how-ever, is that the principal cause of thewarming is a rise in the temperature ofthe water flowing into the area studied.The likely source of this warmer water

is the South American coast. The initialwarming during an El Niino takes place atthe coast. This is true despite the factthat, since the alongshore winds do notweaken, coastal upwelling is unabated.However, after the equatorial Kelvinwave arrives at the coast the thermoclineis pushed down, with the result that thewater mixed into the surface layer iswarmer than before. A second and possi-bly more important factor is the advec-tion of warm water by the southward-flowing El Ninlo coastal current. Thiscurrent is another aspect of the Kelvinwave impinging on the coast. Thus thesurface warming is directly related to theadiabatic, remotely wind-driven physicsdiscussed above.There is both theoretical (29) and ob-

servational (1) evidence that eastern Pa-cific waters south of the equator neverreach the equator. Hence, different wa-ters are involved in the equatorial warm-ing, which at its peak extends some10,000 km from the South Americancoast to the date line (Figs. 1 and 3). Thewinds over this region are anomouslywesterly [Fig. 4; also see figure 4 ofRasmusson and Wallace (7)], so thewestward surface currents along theequator weaken, reducing the flow ofcolder water from the coast. (In a strongevent the currents may even reverse,carrying in warmer surface water fromthe west.) Since the surface heating rateremains high the slower moving watersbecome hotter than normal. At the sametime the flux of cold water from belowthe surface mixed layer is diminishedbecause the weakening of the local east-erlies reduces equatorial upwelling. Inaddition, the remaining upwelled wateris now warmer because remotely andlocally generated Kelvin waves act incoflcert with their subsequent reflectionsat the coast to depress the thermocline.

The 1982-1983 El Niiio

The wind anomalies that presage thetypical El Ninlo occur in fall, but theearliest signs of the 1982-1983 event ap-peared in spring. There were bursts of

SCIENCE, VOL. 222

on

Janu

ary

5, 2

010

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

westerlies in the vicinity of the date line(30), but in view of the great variabilityof the winds in this area their signifi-cance was uncertain. By May there wasa noticeable SST anomaly (Fig. IC). (Itsamplitude was small, but SST variabilityin this area is slight.) By July the windanomaly was sufficiently strong and per-sistent to make it clear that somethingunusual was afoot.However, none of the usual precur-

sors of an El Ninlo event were present.The easterlies had not been especiallystrong and there was no tendency forSST to be unusually low in the east andhigh in the west. Sea level had not built

up in the western Pacific and the thermo-cline was not unusually deep there (31).

In early summer, SST in the easternPacific was a bit above the climatologicalmean but well within the normal range ofinterannual variability (32). By August,warming in this area was substantial(Fig. iC) and the winds were highlyanomalous, with westerlies replacingeasterlies over much of the equatorialPacific. In a reversal of the usual El Ninlopattern (15), the mid-ocean war-ming didnot lag behind major anomalies at theSouth American coast (Figs. 3 and 5).

In the early part of the year, sea levelin the western Pacific followed its nor-

mal course, reaching a peak in March orApril and then falling off (33). In a typicalyear it rises again to a second peak in thefall (Fig. 2), but in 1982 it continued todrop throughout the year (Fig. 6). Thedrop-off was especially sharp at the endof June: sea level fell 12 cm at Rabaul(40S, 152°E) and 18 cm at Honiara (10°S,1620E).

In the mid-Pacific (Fanning Island inFig. 6), sea level was near normal intoJune and then began to rise rapidly: fromJune to September sea level at ChristmasIsland (2°N, 157°W) rose 25 cm; in anormal year it increases gradually untilDecember, with a total change of 10 cm.

20°N

00

200S

20°N

00

'2

September-November -

10 0o(\

200S1~March-May

1 00°E 1400 1800 1400 100°W

0111 , \ Fig. 4 (left). Forcing for the gravest baroclinic Kelvin wave, based on, , , , Ethe composite El Ninlo wind anomaly field (21). Dotted lines indicate

1400E 1800 1400 100°W the path of a Kelvin wave. The curve on the right gives the Kelvinwave amplitude at the eastern boundary. It is a measure of sea level change (or thermocline displacement) at the South American coast. Fig. 5(right). Sea-surface temperature anomaly maps for the 1982-1983 event (41); compare with Fig. 3. (A) September to November 1982. (B) March toMay 1983. Figure SC in Rasmusson and Wallace (7) shows December 1982 to February 1983.

20

0

20

E

a)

a)

Cz

0

-10

20

0

-10

20

0

-20Jan. Apr.

16 DECEMBER 1983

Jul. Oct. Jan.1982 1983

5oS Oo 5oN 5oS

100

00 _I II5 2

100~ ~ ~~5

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

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

_ _ s0.~~...0

E June 1981 N(= 00

ax 100

Qo 5oN 5oS 00 5oN 5oS Oo 5oN

300

500 L:

Fig. 6 (left). Sea level time series at selected tropical Pacific stations (42). Rabaul istypical of the western Pacific, Fanning of the central Pacific, and Callao of the SouthAmerican coast. Note that sea level falls in the west and rises in the east. The changesprogress eastward with time. Fig. 7 (right). Temperature sections (°C) along 85°W(43). Sections for 1981 and March 1982 illustrate normal conditions.

1193

+ 18

a)

12

Co.I-

en -6

a-)C @

E ,° 6( ZE ,a

0,

on

Janu

ary

5, 2

010

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

At Santa Cruz, 8000 km to the east in the DiscussionGalipagos, sea level began its rise inAugust, while at the coastal stations (forexample, Callao, Peru, in Fig. 6), sea

level began to rise in late September or

October.Subsurface effects were, if anything,

more unusual (Fig. 7). In November andDecember 1981 the center of the thermo-cline (the 20°C isotherm) at 85°W on theequator was centered at a depth of 40 m;a year later it was twice as deep (1).Similar depressions of the thermoclinewere observed throughout the easternhalf of the Pacific out to 160°W (34).From July to October, SST at 0°N,110°W rose about 2°C while the tempera-ture at a depth of 100 m increased by 8°C(1). This huge accumulation of warm

water in the eastern Pacific was appar-

ently sufficient to reverse the pressure

gradient along the equator (35). As a

result, the eastward-flowing EquatorialUndercurrent (36) began to disappearduring August at 159°W and did notreappear until January 1983 (35). To ourknowledge, this was the first time thatdirect measurement failed to show an

undercurrent in the central Pacific (37).These changes are consistent with the

idea that packets of Kelvin waves wereexcited by the wind changes in the cen-

tral and western Pacific and propagatedeastward along the equator to the SouthAmerican coast, raising sea level as theywent. This scenario was used to accountfor the canonical El Ninlo event, but in1982 the pattern of SST changes was

very different, probably because the tim-ing of the event was so different fromthat of the annual cycle.We noted above that the equatorial

and coastal warmings are distinct. In1982 the equatorial warming proceededas usual: surface currents changed fromwestward to eastward with passage ofthe Kelvin waves, carrying warmer wa-

ter into the region; at the same time thesuppression of the thermocline meantthat waters upwelled along the equatorwere warmer than before. El Ninloevents usually coincide with the annualcoastal warming, acting in concert withmean conditions to import warmer sur-face waters and depress the thermocline.The 1982_1983 event reached the coastin late summer, when the temperature isnormally headed toward its minimumvalue. Once the event took hold, howev-er, the results were dramatic: SST atCallao rose 2°C per month through thelast third of 1982 (38). In 1983 coastaltemperatures continued to rise throughthe normal warming period and beyond,exceeding climatological values by 6°Cor more by June.

1194

We have emphasized the unusual na-

ture of the 1982-1983 event, but in manyrespects it followed the typical El Nifiopattern. The canonical event begins witha weakening of the easterlies in the fall,at the time of the transition of the Asianmonsoon from its summer to its winterform. This leads to a winter warming atthe South American coast that thenspreads westward. After a turn towardnormalcy there is a second peak in thecharacteristic El Nifio signal (Fig. 2)stemming from a more massive collapseof the trades that begins in the spring (thetime of the other monsoon transition).The 1982-1983 event began with similarspring changes; the first phase was

missed.

This characterization is reinforced bychanges reported at the time of this writ-ing (July 1983): temperatures at the coasthave started to decrease toward normaland the wind system appears to be re-

turning to its normal pattern (39). How-ever, SST in much of the eastern equato-rial Pacific remains well above normal, a

reminder that this event is one of thestrongest ever recorded.While the 1982-1983 event was unusu-

al, it was not unprecedented. The very

weak event of 1963 had some similarcharacteristics, and there are sugges-

tions that the major events of 1930 and1941 followed a similar pattern [(7); fig-ure 11 of Rasmusson and Carpenter(15)]. In any case, analogies to pastevents are too imperfect and our under-standing of the phenomenon is too in-

complete to permit confident predictionof the behavior of El Ninios like the 1982-1983 event.The tropical locus of El Ninlo events

may be attributed to equatorial ocean

dynamics. However, although the same

physics governs all three tropicaloceans, El Ninlo is unique to the Pacific.We suspect this is a consequence of thesingular influence of tropical Pacific SSTon the atmospheric circulation. The at-mospheric circulation is largely poweredby the three tropical heat sources locatedover Africa, South America, and theAustralasian maritime continent. Thelast of these is by far the strongest, and,not being anchored to a landmass, isalso the least sedentary, migrating withthe changing seasons. It becomes amor-

phous during the transition periods of theAsian monsoon, and we speculate that itis vulnerable to anomalous influences atthese times. We further speculate thatSST variations of relatively small ampli-tudes in the warm western tropical Pacif-ic can induce excursions into the central

Pacific of the kind observed to precedeEl Nifio events (40). Throughout theevent, surface wind changes causechanges in SST in the manner we havedescribed, and these SST anomalies feedback on the atmosphere, inducing fur-ther wind changes, until the entireENSO cycle is played out.The 1982-1983 event is the best-ob-

served El Ninio to date, and analysis of itshould prove valuable in developing ourunderstanding of ocean physics and at-mosphere-ocean interactions. The mav-erick nature of this event is an additionalvirtue: we cannot do controlled experi-ments, so it is helpful to have nature doone for us. The event has already shownus that the first phase of the typical ElNifio is not essential; further study maygive us insight into the cause of El Nifioand lead to an ability to predict theseclimatically important events.Note added in proof: The return to-

ward normal has continued into October,but SST remains high. In the canonicalEl Ninlo, SST in the eastern Pacific de-creases very rapidly at the end of theevent, falling below the climatologicalvalue (Fig. 1B).

References and Notes1. D. Halpern, S. P. Hayes, A. Leetmaa, D. V.

Hansen, S. G. H. Philander, Science 221, 1173(1983).

2. J. Toole, Trop. Ocean-Atmos. News. No. 16(1983) (available from JISAO, AK-40, Universi-ty of Washington, Seattle 98195).

3. R. T. Barber and F. P. Chavez, Science 222,1203 (1983).

4. Scientific Committee on Oceanic Researchworking group 55 defined a major El Nifio asoccurring when SST at at least three of fivecoastal stations between Talara (5°S) and Callao(12°S) exceeds 1 standard deviation for four ormore consecutive months.

5. W. H. Quinn, D. 0. Zopf, K. S. Short, R. T. W.K. Yang, Fish. Res. Bull. 76, 663 (1978).

6. EPOCS (Equatorial Pacific Ocean Climate Stud-ies) is sponsored by the National Oceanic andAtmospheric Administration (NOAA); PEQUOD(Pacific Equatorial Ocean Dynamics) is spon-sored by the National Science Foundation.

7. E. M. Rasmusson and J. M. Wallace, Science222, 1195 (1983).

8. H. P. Berlage, Meded. Verh. Ned. Meteorol.Inst. 88, 1 (1966); J. Bjerknes, Tellus 18, 820(1966); Mon. Weather Rev. 97, 163 (1969).

9. S. G. H. Philander, Nature (London) 302, 295(1983).

10. S. Hastenrath and P. Lamb, Heat Budget Atlasof the Tropical Atlantic and Eastern PacificOceans (Univ. of Wisconsin Press, Madison,1978).

11. R. Lukas, thesis, University of Hawaii (1981).12. G. Meyers, J. Phys. Oceanogr. 9, 885 (1979); A.

Busalacchi and J. J. O'Brien, ibid. 10, 1929(1980).

13. G. Meyers, ibid. 12,.1161 (1982).14. This composite is based on the work of many

investigators, most notably K. Wyrtki (ibid. 5,572 (1975); ibid. 9, 1223 (1979); Mar. Technol.Soc. J. 16, 3 (1982)]. (The last of these is anontechnical account of the ENSO cycle.)

15. E. M. Rasmusson and T. H. Carpenter [Mon.Weather Rev. 110, 354 (1982)] give a completedescription of atmospheric and SST anomaliesduring El Nifio as well as many references.

16. D. B. Enfield and J. S. Allen, J. Phys. Ocean-ogr. 10, 557 (1980).

17. D. B. Enfield, in Resource Management andEnvironmental Uncertainty, M. H. Glanz andD. Thompson, Eds. (Wiley, New York, 1981).

18. K. Wyrtki, J. Phys. Oceanogr. 7, 779 (1977).19. __, ibid. 9, 1223 (1979).20. A. J. Busalacchi and J. J. O'Brien, J. Geophys.

Res. 86, 10901 (1981).

SCIENCE, VOL. 222

on

Janu

ary

5, 2

010

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

21. M. A. Cane (J. Phys. Oceanogr., in press)questions the adequacy of the linear theory.

22. The ocean circulation can be represented as thesum of an external (barotropic) mode, in whichthe entire vertical column responds as a unit,and an infinite number of internal (baroclinic)modes. Only the baroclinic response is impor-tant at the long time scales characteristic of ElNifio.

23. R. A. Knox and D. Halpern, J. Mar. Res. 40(Suppl.), 329 (1982); C. C. Eriksen et al., J.Phys. Oceanogr., in press.

24. D. Enfield, J. Geophys. Res. 86, 2005 (1981).25. Wind fields are composites from E. M. Rasmus-

son and T. H. Carpenter (15).26. C. Ramage and A. M. Hori, Mon. Weather Rev.

110, 587 (1982).27. A. Leetmaa, J. Phys. Oceanogr. 13, 467 (1983).28. The currents slow in response to a weakening of

the southeast trade winds.

29. P. Schopf, J. Phys. Oceanogr., in press.30. J. Sadler and B. Kilonsky, Trop. Ocean-Atmos.

Newsl. 16, 3 (1983).31. G. Meyers and J. R. Donguy, ibid., p. 8.32. Available climatologies differ by more than 1°C

in this area due to different averaging periods,data bases, and analysis techniques.

33. K. Wyrtki, Trop. Ocean-Atmos. Newsl. 16, 6(1983).

34. G. Meyers, private communication.35. E. Firing, R. Lukas, J. Sadler, K. Wyrtki,

Science 222, 1121 (1983).36. The Equatorial Undercurrent is a permanent

(with the exception noted) feature in the Atlanticand Pacific. It is a subsurface, eastward-movingcurrent at the equator with speeds often inexcess of 1 n/sec.

37. In May 1983 a westward jet was also found atnormal undercurrent depth at 95°W (S. Hayes,private communication).

38. R. L. Smith, Science 221, 1397 (1983).39. Spec. ClGm. Diagn. Bull. (15 July 1983) (avail-

able from Climate Analysis Center, NOAA,Washington, D.C. 20233).

40. The significance of these excursions was firstnoted by K. Wyrtki [Mar. Technol. Soc. J. 16, 3(1982)].

41. Courtesy of R. Reynolds.42. Island station data are courtesy of K. Wyrtki;

Callao data are from D. Enfield and S. P. Hayes[Trop. Ocean-Atmos. Newsl. 21, 13 (1983)].

43. A. Leetma, D. Behringer, J. Toole, R. Smith,ibid., p. 11.

44. I thank the many colleagues who reviewed anearly version of the manuscript. Special thanksare extended to those who generously contribut-ed unpublished data and to Lenny Martin forassistance in preparing the manuscript. Thiswork was supported by grant OCE-8214771from the National Science Foundation.

and named by Sir Gilbert Walker more

than a half-century ago (1). The primarymanifestation of the Southern Oscillationis a seesaw in atmospheric pressure atsea level between the southeast Pacific

Summary. The single most prominent signal in year-to-year climate variability is theSouthern Oscillation, which is associated with fluctuations in atmospheric pressure atsea level in the tropics, monsoon rainfall, and wintertime circulation over NorthAmerica and other parts of the extratropics. Although meteorologists have knownabout the Southern Oscillation for more than a half-century, its relation to the oceanicEl Niino phenomenon was not recognized until the late 1960's, and a theoreticalunderstanding of these relations has begun to emerge only during the past few years.The past 18 months have been characterized by what is probably the mostpronounced and certainly the best-documented El Nin'o/Southern Oscillation episodeof the past century. In this review meteorological aspects of the time history of the1982-1983 episode are described and compared with a composite based on sixprevious events between 1950 and 1975, and the impact of these new observationson theoretical interpretations of the event is discussed.

characterized by large, remarkably co-

herent climate anomalies over much ofthe globe. The pattern inherent in theseanomalies has been recognized gradual-ly, over a period of decades, as a resultof the collection and analysis of manydifferent climatic records; the recogni-tion process has been somewhat like theassembly of a global-scale jigsaw puzzle.Some of the pieces of this puzzle are

implicit in the Southern Oscillation, a

coherent pattern of pressure, tempera-ture, and rainfall fluctuations discovered

16 DECEMBER 1983

subtropical high and the region of lowpressure stretching across the IndianOcean from Africa to northern Australia.Other manifestations involve surfacetemperatures throughout the tropics andmonsoon rainfall in southern Africa, In-dia, Indonesia, and northern Australia(2, 3). When Walker's scientific contem-poraries expressed doubts concerningthese statistical relations because of thelack of a physically plausible mechanismfor linking climate anomalies in far-flungregions of the globe, he replied, "I think

the relationships of world weather are socomplex that our only chance of explain-ing them is to accumulate the facts em-pirically .. . there is a strong presump-tion that when we have data of thepressure and temperature at 10 and 20km, we shall find a number of newrelations that are of vital importance"(4).

Descriptive studies of the 1957-1958El Nifio event, based in part on routinemerchant ship data from the tropicalPacific, were instrumental in revealingthe link between El Ninlo and the South-ern Oscillation. The large-scale interac-tion between atmosphere and ocean wasconfirmed by retrospective statisticalstudies of past episodes (5). The emerg-ing unified view of the El Nifio/SouthernOscillation (ENSO) phenomenon is ex-emplified by Bjerknes's investigations(6) of the 1957-1958, 1963-1964, and1965-1966 ENSO episodes. These stud-ies were among the first in which satelliteimagery was used to define the region ofanomalously heavy rainfall over the dryzone of the equatorial central and east-ern Pacific during episodes of warm sea-surface temperature (SST), an aspect ofthe phenomenon that Walker apparentlywas unaware of. Bjerknes showed thatthese fluctuations in SST and rainfall areassociated with large-scale variations inthe equatorial trade wind systems, whichin turn reflect the major variations of theSouthern Oscillation pressure pattern.The linking of El Ninlo with the South-

ern Oscillation was viewed as evidencethat ocean circulation plays the role of aflywheel in the climate system and isresponsible for the extraordinary persist-ence of the atmospheric anomalies frommonth to month and sometimes even

Eugene M. Rasmusson is chief, DiagnosticsBranch, Climate Analysis Center, National Meteo-rological Center/National Weather Service, Wash-ington, D.C. 20233. John M. Wallace is a professorin the Department of Atmospheric Sciences anddirector of the Joint Institute for the Study of theAtmosphere and Ocean, University of Washington,Seattle 98195.

1195

Meteorological Aspects of theEl Ninfo/Southern Oscillation

Eugene M. Rasmusson and John M. Wallace

Each year various parts of the globeexperience regional climate anomaliessuch as droughts, record cold winters,and unusual numbers of storms. Butsome years, such as 1982 and 1983, are

on

Janu

ary

5, 2

010

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om