Bryan: Poleward Heat Transport by the Ocean 1137
References
Bennett, A.F., Poleward heat fluxes in Southern Phys. Oceanogr., 11, 1171-1193, 1981. Roemmich, D., Estimation of meridional heat Hemisphere oceans, •. Phys. Oceano•r., õ, Fuglister, F.C., Atlantic OceanAtlas of Temp- flux in the North Atlantic by inverse methods, 785-798, 1978. erature and Salinity Profiles and Data from •. Phys. Ocean., 10, 1972-83, 1980. Bernstein, R.L., and W.B. White, Meridional the International Geophysical Year of 1957- Sarmiento, J.L., and K. Bryan, An ocean transport eddy heat flux in the Kuroshio Extension 58, Woods Hole Oceanographic Inst., Atlas model for the North Atlantic. J. Geophys. Current, •. Phys. Oceanogr., 12, 154-159, Series 1, 209 pp. 1960. Res., 87(Cl), 394-408, 1982. 1982. Geor•i, D.T., and J.M. Toole, The Antarctic Sciremammano, F. Jr., The nature of the poleward Bryan, K., Measurements of meridional heat Circumpolar Current and the oceanic heat heat flux due to low-frequency current fluc- transport by ocean currents, J. Geophys. Res. and fresh water budgets, J. Mar. Res., tuations in the Drake Passage, •. Phys. 67, 3403-3414, 1962. (Vol.40, Suppl.), 183-197, 1982. Oceanogr., 10, 843-852, 1980. Bryan, K., Poleward heat transport by the ocean: Godfrey, J.S., and T.J. Golding, The Sverdrup Stommel, H., Asymmetry of interoceanic fresh- observations and models, Ann. Rev. Earth relation in the Indian Ocean, and the effect water and heat fluxes, Proc. Nat. Acad. Sci. Planet. Sci., 10, 15-38, 1982a. of the Pacific. Indian Ocean through-flow USA, 77(5) 2377-2381, 1980. Bryan, K., Seasonal variation in meridional on Indian Ocean circulation and on the East Stommel, H., and G.T. Csanady, Relation between overturning and poleward heat transport Australian Current, •. Phys. Oceanogr., 11, the T-S curve and global heat and atmos- in the Atlantic and Pacific Oceans: A model (6): 771-79, 1981. pheric water transports, J. Geophys. Res., study, J. Mar. Res., 40 (Supl. Vol.) 39-53, Hall, M.M., and H.L. Bryden, Direct estimates 85, (C1) 495-501, 1980. 1982b. and mechanisms of ocean heat transport, Stommel, H., and G. Veronis, Variational in- Bryan, K. and L.J. Lewis, A water mass model of Deep,Sea Res., 29, 339-360, 1982. verse method for study of ocean circulation,
the World Ocean, _ J. Geophys. Res., 84(C5) Hastenrath, S., Heat budget of tropical ocean Deep Se__aRe_•s.,28A, 1147-1160, 1981. 2503-2517, 1979. and atmosphere, •. Phys. Oceano•r., 10, Trenberth, K.E., Mean annual poleward energy Bryden, H.L., Poleward heat flux and conversion 157-170, 1980 transports by the oceans in the southern of available potential energy in Drake Hastenrath, S., On meridional heat transports hemisphere, Dyn. Atmos. Oceans, 4, 57-64, Passage, J. Mar. Res., 37, 1-22, 1979. in the World Ocean, •. Phys. Oceano•r., 12, 1979. Bryden, H.L., Ocean heat transport, Papers of 922-927, 1982. Vonder Haar, T.H., and A. Oort, New estimate the JSC/CCCOMeeting on Time Series of Ocean Hastenrath, S. and P. Lamb, Climatic Atlas of of annual poleward energy transport by Measurements, Tokyo, May, 1981, World Climate the Tropical Atlantic and Eastern Pacific Program, Publ. 21, IOC, Paris, •O, Geneva, northern hemisphereoceans, •. Phys. Oceanogr. Oceans, Madison: University of Wisconsin •, 169-172, 1973. 1982. Press, 105 pp, 1977. Weare, B., P.T. Strub, and M.D. Samuel, Annual Bryden, H.L., and M.M. Hall, Heat transport by Jung, G.H., Energy transport by air and sea. mean surface heat fluxes in the tropical currents across 25øN latitude in the Atlantic Section XIV of The Dynamic North, Book I, Ocean. Science, 207, 884-$86, 1980. Pacific Ocean, •. Phys. Oceanogr., 11, 705- published by Technical Assistant to Chief 717, 1981. Bunker, A., Computations of surface energy flux of Naval Operations for Polar Projects, 19 Wunsch, C., Meridional heat flux of the North and annual air-sea interaction cycles of the pp, 1956. North Atlantic Ocean, Mon. Weather Rev., Luyten, J. and H. Stommel, Recirculation re- Atlantic Ocean, Proc. Na__•t.Acad. of Sciences, 104, 1122-1140, 1976. USA, 77, 5043-5047, 1980. considered. J. Mar. Res., 40, (Suppl. Vol.) Wunsch, C., The North Atlantic general circula- Campbell, G.G., Energy transport within the 407-426, 1982. tion west of 50øWdetermined by inverse earth's atmosphere-ocean system from a Meehl, G.A., W.M. Washington, and A.J. Semtner, climate point of view, Ph.D. Thesis, Colorado methods, Re__y_v.Geophys. and Space Phys., 16, Experiments with a global ocean model driven 583-620, 1978. State University, Fort Collins, CO., 1981. by observed atmospheric forcing, •. Phys. De Szoeke, R.A. and M.D. Levine. The advective Wunsch, C., D. Hu, and B. Grant, Mass, heat, Oceanogr., 12, 301-312, 1982. salt and nutrient fluxes in the South Pacific flux of heat by mean geostrophic motions in Miller, J.R., G.L. Russell, and L-C Tsang, Ocean, J. Phys. Oceanosr. , (in press) 1983. the Southern Ocean, Deep-Sea Res. 28, 1057- Annual oceanic heat transports computed from WHst, G., and A. Defant, Scientific Results of 1085, 1981. an atmospheric model, D_•n. Atmos. Oceans, the German Atlantic of the Research Vessel Dobson, F., F.P. Bretherton, D.M. Burridge, J. 1982, in press. Crease, E.B. Kraus, and T.H. Vonder •aar, "Meteor" 1925-27, •, Atlas, Walter de Gruyter, Niiler, P.P., and W.S. Richardson, Seasonal Berlin Leipzig, 1936. The 'Case' Experiment: A Feasibility Study, variability of the Florida Current, J. Mar. World Climate Program, 22, World Meteor. Org. Res., 31, 144-167, 1973. Geneva, 95 pp. 1982. Oort, A.H., Glob81 Atmosphere Circulation Stat- Fu, L.L., The general circulation and meridional istics, 1958-1973, NOAAProf. Paper, 14, heat transport of the subtropical South U.S. •ov't. Printing Office, Washington,DC, (Received October •9, 1982; Atlantic determined by inverse methods, J. 323 pp, 1982. accepted January 6, 1983.)
REVIEWSOF GEOPHYSICSAND SPACE PHYSICS, VOL. 21, NO. 5, PAGES1137-1148, JUNE1983 U.S. NATIONALREPORT TO INTERNATIONALUNION OF GEODESYAND GEOPHYSICS 1979-1982
EQUATORIAL OCEANOGRAPHY
Mark A. Cane
Dept. of Meteorologyand Physical Oceanography,MIT, Cambridge,MA 02139
E.S. Sarachik
Center for Earth and Planetary Physics, Harvard Univ., Cambridge,MA 02138
Introduction stratification means that the ocean can respond strongly to basinwide winds on the climatically Interest and activity •in the equatorial important, and observationally accessible, oceans (defined arbitrarily as that part of the annual and interannual time scales. This reali- oceans within ten degrees of the equator) have zation has taken hold as the result of an inter- undergone a remarkable expansion in the last play among theory, modelling and observation. four years. The previous IUGG report (O'Brien, Linear wave ideas have provided a simple 1979) listed about one hundredreferences- the framework and commonlanguage to discuss a wide present one lists over two •hundred and fifty. range of equatorial phenomena. In particular, Among the many reasons for this growth, a pri- the equatorial Kelvin wave, which allows locally mary one is the realization of the rapid nature forced wind changes to be rapidly communicated of equatorial responses. The vanishing of the to the east of the forcing region, has been a Coriolis parameter in the presence of density particularly fruitful concept in equatorial oceanography. Linear and nonlinear numerical Copyright 1983 by the American Geophysical Union. models ranging from single layer shallow water models to full general circulation models have Paper number 3R0051. built upon these linear wave concepts to eluci- 0034-6853/83/003R-0051 $15.00 date the roles of stratification, mixing, and 1138 Cane and Sarachik: Equatorial Oceanography
non-linearity in the dynamics of a wide variety ticns of thermoc!ine depth, mid-ocean currents, of phenomena in the equatorial oceans. In addi- and boundary currents. We will then review a tion, extensive observations taken during field problem which has just begun to receive serious programs in all three oceans have become avail- attention, namely the factors that determine SST able in the last four years: the GARP Atlantic variability in equatorial oceans. We then dis- Tropical Experiment (GATE) in the Atlantic (see cuss the status of our understanding of the most the GATE Atlas: Duing, Ostapoff and Merle, 1980 spectacular manifestation of SST variability, and the GATE Supplements to Deep-Sea Research; the E1 Nino- Southern Oscillation phenomenon. Duing, 1980, and Siedler and Woods, 1980), the We proceed to discuss the problem of observing Indian Ocean Experiment (INDEX, see the August the surface winds, which we regard as a funda- 1, 1980 issue of •.l•Jl•); and the intensive mental factor limiting our understanding of the year 1979-1980 of the First GARP Global Experi- equatorial oceans. Finally, we will try to ment (FGGE, see McCreary, Moore and Witte, identify those current trends that we expect to 1981). In addition, preliminary theoretical and bear fruit by the time of the next IUGG report, observational work is regularly and efficiently and some directions for the more distant future. transmitted by the Tropical Ocean-Atmosphere Newsletter (edited by D. Halpern, JISAO, Univer- Thermocline Variations sity of Washington, Seattle). Interest in equatorial oceanography has The thermocline in equatorial oceans tends also intensified outside the oceanographic com- to be quite shallow, with an average depth of munity. All available data shows that the trop- about 100 m. Because of the action of the eas- ical ocean dominates ocean heat transport both terly trade winds, the mean thermocline tends to in the mean and in annual variations (see Bryan, be deeper in the western parts of the ocean and 1982, for a recent review) and is therefore a shallower in the east. It has proved illuminat- major component of the climate system. It is ing, over the last decade, to consider a model only within the last few years that a major role of the equatorial ocean consisting of a homo- for the equatorial ocean in interannual global geneous light fluid, of mean depth of order climate variability has been more than simply 100m, overlaying a heavier fluid. The interface suggested. Hotel and Wallace (1981) showed that between these fluids then represented the ther- periods of low Southern Oscillation, known to toocline, and to the extent that motions in the coincide with those periods of anomalous warm- fluid below the interface are small, the motion ings of the equatorial Pacific called E1 Nino, of the upper fluid satisfies the shallow water show correlation with anomalous winter climate equations. In the presence of steady zonal over the United States and that this anomalous winds, a no motion solution to the shallow water mid-latitude winter response is consistent in equations is •g' (h2)x = ß (x) where h is the its geographical pattern with having been forced depth to the interface and •(x) is the wind by atmospheric thermal anomalies associated with stress. The long-term mean Pacific thermocline the warm Pacific. While the correlations are does indeed seem to agree with this simple for- not large, and the atmospheric forcing mechan- mula, even though the Pacific in the mean has a isms by no means completely understood, the pos- wind-driven surface current and an equatorial sibility of predicting anomalous conditions over undercurrent, and therefore is far from motion- the U.S. during the winter following the onset less. The success of the simple shallow water of E1 Nino (e.g. Barnett, 1981a) seems to have model at modelling variations of thermocline galvanized both the meteorological and oceano- depth has been quite striking, but there is no graphic communities into a burst of research consensus on why it works as well as it does. into the origins of interannual sea surface tem- Certainly the currents that the shallow water perature (SST) variability in the Pacific and models predict are at odds with observations of into the mechanisms of atmospheric telecommuni- equatorial currents (see next section for more cations from equatorial to midlatitude regions. details). In order to usefully review U.S. progress Equatorial subsurface measurements have in equatorial oceanography in the limited space been taken infrequently in the Atlantic and available to us, we focus our primarily on the Pacific, and hardly at all in the Indian Ocean. low frequency response of upper equatorial oce- There is, however, enough climatological data in ans to forcing by the wind. We perceive this as the Pacific and Atlantic to at least give the the unifying theme of equatorial oceanography broad outline of seasonal variability. Meyers during the previous four years. A major (1979b) examined the climatological monthly development in this vein, unprecedented in variation of the 1•øC isotherm on the equator, large-scale oceanography, is the attempt to which is near the bottom of the thermocline in simulate ocean variability by forcing ocean the Pacific. He finds that the variability with models with real winds and verifying the model respect to the mean is relatively small and that response by comparing to oceanographic data the largest seasonal variability is in the records at selected points. While the ocean eastern Pacific. Tsuchiya (1979) examined the models used in these attempts have been highly geopotential gradient in the eastern Pacific and idealized and the quality of wind data dubious found it to be roughly in phase with the sea- at best, the initial results have been sonal zonal winds. Both of these results are encouraging. This attempt at simulation brings complementary to Horel's (1982) results which observation and theory into an intimacy rare in shows most of the seasonal variability of SST to oceanography and will, we hope, provide a con- be confined to the eastern Pacific. In the tinuing arena for testing our understanding of Atlantic, Merle (1980b, 1980c) finds that the dynamic and thermodynamic processes. seasonal variation of the 23øC isotherm (marking We begin this review by discussing varia- the upper part of the equatorial thermocline) is Cane and Sarachik: Equatorial Oceanography 1139
not small compared to the mean and that when the periodic response of the thermocline depends mean zonal winds across the equatorial Atlantic strongly on the winds within ten degrees of the are weak (in February-March), the thermocline equator and only weakly on winds beyond ten becomes almost flat. These results are in degrees. agreement with the previous results of Katz et Kindle (1979) and Busalacchi and O'Brien al (1977) for the zonal pressure gradient in the (1980) force a shallow water model of the equatorial Atlantic. Merle also finds that the Pacific with an approximation to the annual and variations in heat content are in phase with the semi-annual winds. They both find that the thermocline variations and that they are an semi-annual response of the thermocline in the order of magnitude larger than, and out of phase eastern Pacific is forced by the semi-annual with, the vertical heat fluxes through the sea winds in the central Pacific. Kindle (1979) surface. This implies that on a seasonal time achieves a better overall simulation of equa- scale at least, heat content variations are torial thermocline variability in the east pri- dynamically forced and that the surface heat marily, we believe, because Busalacchi and flux does not cause these heat content varia- O'Brien (1980) extended the winds observed tions but, rather, responds to the dynamically between _+Bdegrees and uniformly extended pole- induced SST variations accompanying the heat ward; as indicated earlier, the winds between B content changes. Further, by taking zonal aver- and 10 degrees do affect the equatorial ages across the entire Atlantic, Merle finds response. On the other hand, using mean monthly that there are annual heat content variations in winds over the entire Pacific, Busalacchi and the zonal average implying that heat is not sim- O'Brien achieve impressive agreement of the ply sloshed east-west, but must also be redis- meridional topography of the thermocline and its tributed meridionally. This latter result is observed annual variation with observation. consistent with a study by Katz (1981) of In the Atlantic, Busalacchi and Picaut dynamic topography in the western Atlantic show- (1982) and Cane and Patton (1982) force with ing that the entire dynamic topography pattern climatological monthly winds and compare with (consisting of a mean equatorial ridge at 4N and observations at Abijan on the northern coast of a trough at 10N between which geostrophically the Gulf of Guinea. Both are able to simulate flows the North Equatorial Counter Current) the annual variation well and both conclude, by flattens out in early spring and deepens in late dividing up the wind field geographically, that summe r. the response is due to the wind field across the Theoretical approaches to understanding entire ocean and not Just to wind changes in the thermocline variations have been of two major Western Atlantic as had been previously sug- types: those using idealized winds blowing over gested. There is no special contribution from shallow water models and those using approxima- the local winds at the Guinea coast. tions to the real winds blowing over shallow Finally, Busalacchi and O'Brien (1981) water models. The first type of calculation is force the linear shallow water model of the done to understand the basic dynamics of forced Pacific with a decade of observed winds subjec- response, while the second type is done to com- tively analyzed to fill gaps (Goldenberg and pare response to observations. O'Brien, 1981), and compare to a decade of tide An illustration of the first type of calcu- gauge observations of sea surface taken in the lation is Cane and Sarachik (1981) who forced a Galapagos. To the extent that motion is con- linear equatorial shallow water model with fined to a single baroclinic mode, sea surface periodically varying zonal winds of Gaussian variability is a proxy for themocline variabil- form in the meridional direction, and indepen- ity. The results show broad agreement with the dent of zonal coordinate. They solved for the sea surface interannual variability in that the response analytically over a wide range of model broadly reproduces the interannual cold parameters. The response for parameters charac- and warm periods. There are obvious differences teristic of annual forcing over the Atlantic however: the simulated variability seems to showed a remarkable (and perhaps fortuitous) have shorter time scales than observed and there agreement with many of the features of the are epochs of cold or warm water that the simu- annual Atlantic response as described by Merle lation misses completely. Considering that the (1980c). But more important, the calculation winds are of low quality, that the model uses a showed that response to periodic forcing dif- single vertical mode where perhaps two or more fered in fundamental ways from response to are called for, and that the verification data impulsive forcing. In particular, the response for themocline displacement is a proxy, we take to periodic forcing may be considered the stand- the results to be encouraging. ing sum of local and boundary responses, all at These attempts at simulation point to two the forcing frequency, and illustrated many of important needs that currently exist, and have the interference effects first predicted by to be met before any further progress can take Schopf, Anderson and Smith (1981). Furthermore, place. First, an accurate wind data set over it showed that the greater seasonality of the surface of the ocean over at least an annual response of the Atlantic versus the Pacific cycle, and second, a complete thermocline depth could not be attributed to the smaller size or data set simultaneous with the winds. Until both greater memory of the Atlantic (Philander, 1979) of these data sets are available, the confronta- since •memory • is a concept concept valid only tion between theory and experiment must neces- in initial value problems, not in periodic sarily be incomplete. It should be noted that response problems. The difference in the two field programs are due to be performed in response of the two oceans is attributable to 1983, the U.S. Program SEQUAL(Seasonal Equa- the greater seasonality of the winds over the torial Atlantic Experiment) and the French pro- Atlantic. The calculations also indicated that gram FOCAL (French Ocean Climate Atlantic exper- 1140 Cane and Sarachik: Equatorial Oceanography
iment) which have, as one of their goals, the response to periodic winds was explored sys- acquisition of simultaneous wind and thermocline tematically. For periods shorter than 10 days data set for the equatorial Atlantic over a com- there is little rectified motion, but as the plete annual cycle. period approaches 50 days intense eastward sur- face currents can develop. For periods up to Equatorial Currents 150 days there is great variability above the main thermocline (though not below) and an McCreary ( 1981 a, b) has proposed an intense undercurrent develops. For still longer interesting conceptual model for the steady periods the model ocean response is a succession equatorial undercurrent. The response of a of equilibrium states. (The numbers given are linear inviscid ocean to a zonal wind indepen- for a 5000 km basin and would increase linearly dent of the meriodional coordinate would be a with basin size.) Katz and Garzoli (1982) used motionless Sverdrup balance in which the wind these theoretical results to explain the stress is opposed by the pressure gradient observed seasonal variations of equatorial force. In McCreary's model the graver vertical currents in the Atlantic. The most striking modes are in such a Sverdrup balance, while for result is that most of the variation in tran- the higher modes the driving is balanced by the sport is due to the reversal of the surface dissipation of heat and momentum. The result is currents rather than an increase in undercurrent a zonal Jet -- an undercurrent-- centered in speed. the model thermocline. This work shows how As is true for their linear counterparts, friction and stratification can act in concert nonlinear models show current structure and to determine the characteristics of the under- amplitude to be highly dependent on vertical current. The form chosen for the vertical mixing processes. The calculation of Semtner viscosity and diffusion, however, is a function and Holland (1980) provides an interesting exam- of the stratification, making the two influences ple. It used a very low value for vertical eddy inseparable. McPhaden (1981) has proposed a viscosity and generated a great deal of horizon- linear model that uncouples the friction from tal mixing due to turbulent eddies. In fact, the stratification, though not the thermal dif- the eddy activity in this model appears to fusion. exceed what is observed in the real oceans while The models mentioned in the preceding para- the value used for vertical viscosity appears to graph both neglect nonlinear terms, which are be much lower than observations suggest (e.g. easily shown to be order one near the equator. Crawford and Osborn, 1981a). Pacanowski and The nonlinear studies of Cane (1979,1980) and Philander (1982) incorporated a more realistic, Philander and Pacanowski (1979,1981) have Richardson number dependent parameterization of emphasized the role of the meriodional circula- vertical mixing. They found that the vertical tion, especially the vertical advection of circulation and mixing at the equator are so momentum. In a model with two homogeneous lev- vigorous that surface heating is essential for els above the thermocline, Cane (1980) pointed maintaining the stratification on timescales of out the differing effects of this term with 100 days. Without this stratification a mixing easterlies and westerlies and used the results model will wipe out the shear between the sur- to explain the observed behavior of equatorial face and the undercurrent. Schopf and Cane currents in the Indian Ocean (also see McPhaden, (1982) found a similar result in a model that 1982b). Philander and Pacanowski (1979,1981) explicitly parameterizes the physics of the sur- extended these results to a multilevel strati- face layer. Our lack of understanding of verti- fied model. The papers referenced in this para- cal mixing processes remains a significant limi- graph combine these nonlinear ideas with linear tation on our ability to model the equatorial wave theory to explain the temporal behavior of circulation. the undercurrent and associated equatorial cir- culation. Cane (1979) showed that the linear Coastal Currents ideas explained the establishment of the pres- sure gradient, which is crucial for establishing The Somali Current has long held a particu- the meridional circulation and the undercurrent. lar fascination for those who work in equatorial In the nonlinear meridional wind case the oceanography. The observations taken during meridional advection of momentum destroys the FGGE have provided us with a clear description symmetry of the linear solution and results in a of the response of this current to the Southwest strong eastward Jet downwind of the equator. As Monsoon (of. Schott and Quadfasel, 1980,1982; a parcel of fluid moves meriodionally it picks Smith and Codispoti, 1980; Brown, Bruce and up energy from the wind and as it movesaway Evans, 1980; Bruce, Quadfasel and Swallow, 1980; from the equator its relative vorticity changes Bruce, Fieux and Gonella, 1982; Evans and Brown, to compensate the change in planetary vorticity. 1981). The picture that emerges shows two anti- The result is an eastward Jet downwind of the cyclonic gyres, one turning offshore at about equator and strong upwelling on the upwind side 10N and the other at about 4N. The latter is (Cane, 1979; also see Philander and Pacanowski, observed to migrate northward beginning in the 1981b). The multilevel model of Philander and late summer, eventually coalescing with the more Pacanowski (1979) allows the vertical structure northerly gyre which remains approximately fixed of modes and currents to differ. They added the in space. (This behavior is shown most clearly important result that the adjustment of the in the sequence of pictures of the thermal front upper ocean is via the second baroclinic mode, at the northern edge of the gyres given in the gravest mode which is very nearly trapped Brown, Bruce and Evans, 1980). Explanations for above the thermocline. various aspects of the observed behavior have In Philander and Pacanowski (1981a) the appeared in the literature, but there is no Cane and Sarachik: Equatorial Oceanography 1141
coherent synthesis as yet. The more northerly Philander (1979) and McPhaden and Knox (1979) gyre has been attributed to the local wind did theoretical studies of the influence of an stress curl (which is unusually intense at this undercurrent-like mean flow on a shallow water location; cf. Wylie and Hinton, 1982 or Schott system. There have been no studies which con- and Fernandez-Partagas, 1981 ). The calculations sidered the influence of a current system with a of Cox (1979) indicate that such a gyre will complex vertical structure. form in response to the northward strengthening Among the interesting observations reported of the longshore winds. Philander and Delacluse over the past four years are those of deep equa- (1982) propose that the turn-off latitude of the torial jets. Beneath the thermocline in all southerly gyre is set. by the interior flow: in three tropical oceans there are equatorially response to a northerly wind an eastward jet confined currents with vertical scales of will develop at about 4N (cf. Cane, 1979) and several hundred meters and speeds of several the western boundary current will adjust to sup- tens of cm/sec down to depths of several ply the needed mass flux. Another possible (and thousand meters. (For examples in the Atlantic somewhat similar) explanation is contained in see Horigan and Weisberg, 1981, or Weisberg and the study of cross equatorial flows given by Horigan, 1981; for the Indian Ocean, Eriksen, Anderson and Moore (1980). Neither study 1980 or Luyten, 1982; for the Pacific, Leetmaa explains the northward migration of this and Spain, 1981, or Eriksen, 1981). The data feature. However, Cox (1979) was able to simu- presently available is not sufficient to permit late such behavior in a numerical experiment in a conclusive characterization of the suite of which the wind along the coast was relaxed. equatorial motions, but a speculative picture is Philander and Delacluse (1982) contrasted beginning to emerge. Equatorial waves exist at the behavior at the western boundary in response all frequencies: there is no cutoff at a to southerly winds (viz. the Somali Current) nonzero inertial frequency. It appears that the with that at an eastern boundary. At both boun- deeper motions exhibit spectral peaks at annual daries the surface flow is in the direction of (Eriksen, 1981) and/or semiannual periods the wind, but at the eastern side the sea sur- (Luyten and Roeromich, 1982), periods for which face sets up so that the pressure gradient there is an obvious source. (At present, it is opposes the wind stress, driving a poleward not clear how much of the deep motions are undercurrent (also see McCreary, 1981b and Phi- quasi-steady). There is also evidence, espe- lander and Yoon, 1982). This is clearly evident cially in the upper õ00m or so, for peaks at in observations taken off the coast of Peru periods in the range from 16 to B0 days (e.g. (Brockmann et al, 1980; Brink, Halpern, and Duing and Hallock, 1980). The most likely Smith, 1980). Theory implies that the eastern explanation is that these motions result from boundary is an extension of the equatorial instabilities of the mean currents (Philander, waveguide, and that the reflection of a Kelvin 1978; Cox, 1980). It is possible that some wave incident on the boundary at the equator smooth spectral form will characterize motions should be evident at all latitudes along the at shorter periods, but it is also likely that coast. The observational study of Enfield and significant peaks related to peaks in the winds Allen (1980) suggests that this indeed the case. will appear (Garzoli and Katz, 1981). Allen and Romea (1980) and Romea and Allen It would be a helpful simplification if the (1982) have considered the modification of the deep Jets could all be described as a sum of coastal response due to the presence of the con- just a few types of wave motions; e.g. Kelvin tinental shelf. waves, mixed Rossby-gravity waves and the gravest Rossby waves. Eriksen's analyses Equatorial Waves and Deep Jets (1980,1981) indicate that this is not the case: the observations can only be described by a sum Given their central role in theoretical over many wave types. equatorial oceanography, it is natural that observationalists would seek evidence for the Sea Surface Temperature Variability existence of equatorial waves in the data record. As an overall characterization of these The equilibrium temperature at the tropical efforts it may be said that a number of papers sea surface is determined by a complicated have made a strong case, but that no data set is interaction between solar radiation, infrared complete enough to identify more than a very few radiation and evaporation and is basically of the characteristic signatures of such waves. determined by atmospheric processes (Sarachik, An early effort is that of Weisberg, Horigan and 1978). There are now only three ways to • Colin (1979), who present evidence for the SST at a point: by changing the vertical heat existence of a mixed Rossby-gravity wave in the flux into into the ocean across the sea surface Atlantic. The most striking example is the Kel- by purely atmospheric processes; by changing vin wave event observed in the Pacific in 1980 advection of heat into that point by a variety (Knox and Halpern, 1982). A disturbance in sea of purely oceanic processes; and by increases level crossed the Pacific from the dateline to in mixed layer depth (on an ocean whose tempera- the Galapagos at a speed of 2.9 m/s without ture decreases with depth) forced by increased dispersing. These characteristics imply that wind stirring. A fundamental distinction among it should be identified as a packet of first these three methods of changing SST is that in baroclinic mode Kelvin waves. Most of the work the first method vertical heat fluxes through on equatorial waves has ignored the effects of the sea surface cause the SST change while in mean currents, but there are a few exceptions. the second and third method, all else being Hallock (1980) and Weisberg (1980) attempted to equal, the SST change is dynamically caused in account for mean flow influences in GATE data; such a way that the vertical heat fluxes through 1142 Cane and Saraqhik: Equatorial Oceanography
the surface tend to restore the equilibrium and determined. In an investigation concentrating therefore oppose the SST change. Sometimes all on improving the temperature structure in this things are not equal - positive feedbacks can model, Pacanowski and Philander (1981), find arise, for example, in eastern oceans when cool- that a Richardson number dependent vertical mix- ing of the SST by ocean dynamics leads to ing of momentum, and a vertical heat flux into stratus cloud formation by advective atmospheric the ocean are both necessary factors in obtain- processes thereby reducing the solar radiation ing realistic thermocline temperatures. reaching the surface and reducing rather than A recent numerical model by Schopf and Cane increasing the vertical heat flux into the (1•82) has explicitly coupled a mixed layer to a ocean. All three methods of changing SST act in low vertical resolution dynamical model. This various places in equatorial oceans. We will model effectively replaces the directly wind discuss the observations of SST changes and the driven, but fixed depth, upper layer in the status of the models recently developed to model of Cane (1•7•) with a thermally active and address the processes that change SST. variable depth mixed layer: it therefore consti- In the central Indian Ocean, where tempera- tutes the simplest model that not only contains ture gradients are weak, advective temperature enough dynamics to simulate the equatorial changes are small and SST changes are caused undercurrent but also has all the basic mechan- almost entirely by heat fluxes through the sea isms (albeit in simplified form) that affect SST surface: simple one dimensional mixed layer variability. Initial tests with this model models driven by fluxes of heat and momentum at indicate a strong assymmetry between upwelling the surface seem adequate for describing SST and downwelling: upwelling lowers SST directly variability on time scales up to the annual by vertical advection while downwelling leads to (McPhaden, 1982). In the central Pacific, no direct SST change but does make it possible Stevenson and Niiler (1982) find that two-thirds for SST to slowly change by entrainment at the of the heat storage in the mixed layer can be base of the mixed layer. This assymmetry explained by heat fluxes through the sea sur- clearly indicates that on short time scales face, the remaining third presumably accountd there need be no relation between thermocline for by advective affects. The eastern Pacific variability and SST variability. Further exper- exhibits a large and complex annual cycle with iments with the model (Schopf and Harrison, cold water appearing in northern summer and pro- 1•82) indicate that the existence of SST signa- pagating westward with time (Hotel, 1982). tures associated with downwelling Kelvin signals Niiler indicates that 60% of the variance of are advectively produced, and are therefore storage in a very large area (encompassing that very much a function of the pre-existing mean part of the Pacific showing large annual varia- state of the ocean. For some mean wind condi- tion, 6 N-6 S, 76 W-140 W) can be accounted for tions, they showed that a downwelling Kelvin by vertical heat fluxes through the surface wave was able to induce SST changes similar in although finer spatial scale budgets have not amplitude and form to the initial coastal and year been made. In the eastern Atlantic, the equatorial warming occurring during E1 Nino annual signal is dominated by a cold summer episodes in the Pacific. While these initial tongue of water that seems dynamical in origin results are encouraging, deficiencies in the in that the coldest SST's appear at times of parameterization of vertical surface heat smallest heat content (i.e. shallowest thermo- fluxes, taken proportional to the difference of cline) and are marked by increase• of the heat SST and a fixed atmospheric temperature, are flux into the cold tongue (Merle, 1980). clearly apparent. A numerical model that has the capacity to As thermally active equatorial ocean models deal with all three types of SST change must begin to be developed, it appears as though the have enough vertical resolution to resolve the most difficult issue remaining to be addressed mixed layer, adequate dynamics so as to be able is that of the surface boundary condition for to model the wave, advective, and diffusive vertical heat fluxes. The heat flux into the processes that affect SST, and an adequate ocean is influenced in an essential way by enough representation of the atmosphere so as to advective and convective processes in the atmo- get reasonable heat fluxes through the sea sur- sphere which are in turn influenced by sea sur- face. One layer models, which may work well at face temperature. The usual type of boundary simulating thermocline depth changes, are condition that parameterizes heat flux by the totally inappropriate for simulating SST difference between some fixed atmospheric tem- changes, except in those places and on those perature and the calculated SST is inadequate in time scales where thermocline variability is a that it takes no account of the role of the suitable proxy for SST variability. ocean in determining the atmospheric tempera- Philander has been using a sixteen layer ture. Perhaps a correct surface condition can general circulation type ocean model forced by only be applied by coupling a full atmospheric wind stresses, but not heat fluxes, at the sea model to the ocean. In any case, it appears surface (Philander and Pacanowski 1980, 1981 ). that progress in modelling SST variations may While carrying no explicit mixed layer, the require a deeper appreciation by oceanographers non-uniform vertical resolution is chosen to be of the nature of the interaction between the finest near the surface and since there are no ocean and the atmosphere. surface heat fluxes, temperature changes are dynamically determined. In a study of reponse The E1 Nino- Southern Oscillation Phenomenon to meridional wind variations, Philander and Pacanowski, 1981, find patterns of SST distribu- The Southern Oscillation is an oscillation tion remarkably like those in the eastern Atlan- of surface pressure primarily between the broad tic where SST is believed to be dynamically low pressure area over the "maritime continent" Cane and Sarachik: Equatorial Oceanog•aphy 1143
centered on Indonesia and the center of the westerlies in this warm patch. On the coast of South Pacific high in the southeast Pacific, but Peru, warm anomalies are still small but have also with global manifestations. A convenient begun to grow. By the April of the E1 Nino index of the SO is the anomalous pressure year, the anomalies are large at Peru (in phase difference (seasonal cycle removed) between the with the normal seasonal cycle), warm water has Southern Pacific high and the Indonesian low. propagated westward halfway across the Pacific, It is known that this index, the SOI, oscillates and the warm patch on the dateline seems to have irregularly with periods roughly from 3 to 5 expanded eastward. In this April of the EN years and that periods of lowest SOI coincide year, the SPCZ is anomalously far north, the with anomalously warm water across the equa- ITCZ moves southward onto the equator to lie torial Pacific from the coast of South America over warm water, and positive precipitation to beyond the dateline, a distance comprising of anomalies and westerly wind anomalies appear more than a quarter of the circumference of the over the equatorial warm water. By September of earth. This phenomenon of the interannual vari- the EN year, the warm dateline anomaly and the ation of Pacific SST, the anomalies beginning at westward propagative warm anomaly have merged to the coast of Peru in phase with the normal sea- give warm SST anomalies across the entire equa- sonally warm water in March but then propagating torial Pacific all the way to Indonesia. There westward into the open Pacific, is called E1 are now very large westerly anomalies in the Nino. Because the causal relationship between wind all across the Pacific but strongest in the the pressure fluctuations of the SO and the warm western Pacific west of 160W. The SPCZ has water anomalies of E1 Nino remains unclear, we stayed far north, the ITCZ has stayed far south will sometimes lump both phenomena together into and they meet roughly at the dateline rather a single acronym, ENSO. than near Indonesia. Precipitation shows large While scenarios for ENSO due to BJerknes increases near the dateline with decreases over and Wyrtki have been in existence for many Indonesia, all indicating that the vast thermal years, the lack of data over vast regions of the heat source, that would usually be over the mar- Pacific have required so much averaging over itime continent centered about Indonesia has space and time in order to obtain smooth moved eastward to the dateline. By January of results, that spatial and temporal relationships the year following E1 Nino, these trends have have become obscured. While still plagued by a reached their peak and a return to normal is sparsity of data, Rasmussen and Carpenter (1982) indicated by anomalously cold water everywhere assembled new SST and wind sets and composited to the south of 20 S. Cold anomalies then the data for six E1 Nino periods, those of 1951, appear at the coast of Peru and begin propagat- 1953, 1957, 1965, 1969 and 1972. Because the ing westward along the equator, and the E1 Nino sequence of events for each individual E1 Nino cycle has come to a close. was so tightly coupled to the seasonal cycle, it The scenario presented is indeed compli- was found possible to composite the data by cated and we have gone into it in some detail months during the E1 Nino years but even here, because at our present level of understanding, data sparsity first required that 3 month aver- it is not clear which parts are essential and ages be taken. This paper gives the first which are not. Wyrtki (1975,1979,1982) has detailed description of the sequence of events offered hypotheses as to which elements are in the ENSO phenomenon over the entire tropical essential to E1 Nino. He notes that in the year Pacific basin and therefore is a major milestone preceding an E1 Nino event the easterlies over on our road to understanding it. the Pacific are anomolously strong, raising sea We may summarize their description of the level in the western Pacific and lowering it in composite E1 Nino as follows. As early as the the east (see especially Wyrtki, 1979). He pro- May of the year preceding E1 Nino, the South poses that the weakening of these winds gen- Pacific high begins its anomalous weakening and, erates downwelling equatorial Kelvin waves that since the low pressure region over Indonesia propagate to the coast of South America and exhibits no anomally until several months later, depress the thermocline there. Enfield (1981c) the weakening of the SOI is initially affected shows that during the 1972 E1 Nino the thermo- only by the South Pacific high. By September of cline off Peru was normal in November-December the year preceeding E1 Nino, the most striking 1971, but had lowered substantially by anomalies seem to be connected to the South February-March 1972. In the 1975 paper Wyrtki Pacific high with slightly warmer SST everywhere indicates that the weakening of the winds in the South of 20 S (the analyzed region extends only central Pacific is the critical initiator of to 30 S), anomalously warm SST and anomalous these events. However, the data presented by convergence to the south of the normal position Rasmussen and Carpenter makes it clear that the of the South Pacific convergence zone (SPCZ), winds in this region do not relax until after and some weak indication of anomalously strong the warming at the coast has already begun to equatorial easterlies to the west of the date- spread westward. Busalacchi and O'Brien (1981), line. Thus, in this antecedent September, it in the simulation using real winds referred to appears as if the SPCZ, that convergence zone earlier, identified the relaxation of the winds extending from Indonesia southeastward across west of the dateline as the precursor to the the Pacific, has moved anomalously far south. coastal warming and suggested that the subse- By contrast, in the antecedent December, the quent collapse of the easterlies in the central SPCZ has moved anomalously far north, warm SST Pacific helps maintain the deep thermocline at anomalies remaining everywhere south of 20 S the coast throughout the E1 Nino year. accompanied by anomalous westerlies, but now a In summary, Wyrtki (1982) views E1 Nino as warm anomally has appeared on the equator at the a vast sea-saw. The ocean is preconditioned by dateline with indications of slight anomalous the stronger, high SOI easterlies, causing warm 1144 Cane and Sarachik: Equatorial Oceanography
water to pile up in the west. Their subsequent larger than the measurement errors. Even when relaxation generates Kelvin waves that carry the this accuracy can not be obtained, useful water back to the east. Wyrtki regards the results are still assured if the errors in the preconditioning as essential; this is a bit puz- wind field force errors in the response small zeling because according to theory the ocean compared to the amplitude of the response. The dynamics should respond to the any changes in required accuracy of the wind field varies from the winds. problem to problem and has not as yet been quan- There are as yet no theories that account tified - an obvious way to do this would be by for the entire sequence of events that comprise series of numerical simulations. ENSO. It is generally felt that only the ocean Currently, surface winds over the ocean are has inherent timescales long enough to account routinely collected only by selected island sta- for the four year quasi-periodicity of the tions, by ships of opportunity, and by satellite cycle. McCreary (1982) has constructed the measurements of the motion of low level clouds first model to make this idea concrete. It pro- and interpolating down to the surface. All of duces a periodic ENSO signal with the period set these methods suffer from inherent inaccuracies by the cross Pacific travel times of an eastward and from spatial inhomogenieties that render it Kelvin wave plus a westward higher order Rossby impractical to construct a basin-wide surface wave. In its present form the model does not wind field by any single technique. On rare account for the variable spacing of E1 Nino occasions, enough research platforms are placed occurrences or the apparent relation of E1 Nino in the ocean to enable the construction of a to the seasonal cycle, but its most serious limited area surface wind field directly from shortcoming is the schematic nature of the measurements; this was accomplished during the atmospheric component and its unrealistic GATE experiment, for example Krishnamufti and response to changes in the model ocean. Krishnamufti, 1980. That the locking of ENSO to the seasonal Ship of opportunity measurements of the cycle may be a clue to the entire sequence of surface wind have been archived for over sixty ENSO was suggested by Philander (1983). Based years and this makes it possible to construct on a complete documentation of the seasonal climatological monthly wind fields over entire cycle of SST and winds over the Pacific by Hotel ocean basins (e.g. Hastenrath and Lamb, 1979; (1982), Philander argues that the warm water Hellerman, 1980) but even here, coverage is ade- that normally appears near Peru early in the quate only in limited regions. Since the qual- year is particularly susceptible to positive ity of the individual wind measurements that go feedback due to air-sea interactions. He argues into the climatological average is not high, that a warm anomaly off Peru produces enhanced errors rapidly accumulate in those regions where westerlies at the western edge of the westward ships rarely venture. Furthermore, the propagating normal seasonal warming which in representativeness of the monthly wind field turn enhances the warm water. While this is an depends on the degree of interannual variability argument, rather than a model, it contains the present in the wind field and on possible high intriguing suggestion that it is E1 Nino that frequency aliasing (Harrison and Luther, 1982). causes the SO rather than the other way around. Only recently has any attempt been made to esti- As we see it, the current situation is as mate the errors inherent in the construction of follows. There is an indication that the EN climatological monthly wind fields (Hellerman, response is in the winds (Busalacchi and 1982). O'Brien, 1981) and that the precursors to the The situation with respect to the construc- Peruvian warm water is the relaxation of the tion of actual, rather than climatological, sur- easterlies west of the dateline. Further, this face wind fields is still more unsatisfactory relaxation is connected to the appearance of a since the number of data available during any pool of warm water near the dateline (Rasmussen particular month is very limited. Subjective and Carpenter, 1982), but no one has yet analysis by meteorologists trained to recognize explained where this pool of warm water came wind patterns has been used (Goldenberg and from, or why so much of the antecedent warm O'Brien, 1981) to spatially interpolate the anomalies appear south of 20 S or why the South monthly Pacific ship wind field onto a regular Pacific High starts weakening. The ability to grid for the years 1961-1970 but the quality of continuously monitor SST and winds over the the resultant wind field remains unknown. Cloud entire South Pacific, presumably by satellite, winds have been show to give reasonable time seems a prerequisite to answering these ques- averaged surface winds (Halpern, 1979) in some tions. regions but seem to fail in others (Garzoli, et al., 1982). Perhaps the best that can be done Surface Winds at the moment is to use all the sources of sur- face winds real-time assimilated into numerical Since the equatorial thermocline and the weather forecast models which are then corrected equatorial currents respond primarily to changes afterwards when additional data, too late to in the surface wind stress, and since SST varia- have been included in the real time assimila- bility can additionally depend on the heat flux tions, become available. This method is still into the ocean which in turn depends in an limited by the relatively small amount of data essential way on the surface wind, it is clear available and by their spatial inhomogeneity but that in order to understand the forced response would improve slowly as the forecast models we must know the surface winds to some degree of improve. Lacking detailed comparisons with in accuracy. Ideally, we would like to know the situ measurements, it is at present impossible winds to an accuracy such that errors in the to assess the accuracy that even this best wind field would force errors in the response no method can provide. Cane and Sarachik: Equatorial Oceanography 1145
The capability for almost synoptic surface logical data undergo more detailed and searching wind measurements that satellite instruments analyses, we expect that these models containing could provide, especially the scatterometer (see a single homogeneous layer above a delta func- the papers in Gower, 1981 and the Vol. 8?, April tion themocline will no longer prove adequate BO, 1982 issue of J. Geophys. Res.)is an excit- and that models with more resolution in the ing possibility for the future. A basin-wide vertical will have to be invoked. surface wind field of known accuracy would be a In emphasizing purely equatorial processes, major boon to equatorial oceanographers. we have not given enough attention to some basic larger scale questions which we expect to begin +- • •--.-.,,• •,,,tn= the next four veers. In Summary and Outlook particular the entire question of the connection of equatorial circulations with the gyre scale The previous four years have seen notable circulation of the major ocean basins remains progress in the use of simple shallow water obscure. Sub-categories of this question models for describing the response of the ther- involve the relation of the eastern boundary mooline (and, by proxy, sea level) to the forc- currents to the equatorial current- ing by large scale wind stresses (e.g. Busalac- countercurrent system, the role of upwelling in chi and O'Brien, 1980, 1981). We noted that eastern boundary current dynamics and thermo- these models are not adequate to describe equa- dynamics, and the sources of water and salinity torial currents because the averaging above the for each component of the equatorial current- thermocline is too severe to allow the appropri- countercurrent system and the eastern and ate non-linearities to develop. A simple divi- western boundary currents that connect to them. sion of the homogeneous layer above the thermo- In addition, some very basic descriptive equa- cline into two levels, one communicating torial oceanography still remains to be done. directly with the wind stress and one responding For example, we still lack a climatology of the to the pressure changes induced by thermocline South Equatorial Countercurrent and of the sub- tilt, provides a model that describes much of surface countercurrents and the seasonal the variability of both thermocline and currents variations of most components of the equatorial in equatorial regions. We are aware of no case current-contercurrent system still remains to be where upper ocean current variations described described in all three oceans. by such a simple model disagree qualitatively During the next four years, we expect much with those simulated by ocean general circula- emphasis on theoretical investigations of the tion models containing many levels in the verti- mechanisms responsible for SST variability, and cal. on field investigations of SST variability on These simple models are clearly inapplica- annual and interannual time scales. A variety ble to the description of the complicated of models will be developed, both linear and motions observed below the thermocline. If non-linear, having increased resolution in the these deep motions are forced from the surface, vertical and containing thermodynamically active a crucial issue that needs to be addressed is upper levels. We expect to see these models the rate of energy leakage through the thermo- used for simulations of not only thermocline cline down into the deeper ocean. Current esti- depth and structure and the associated currents, mates are that dissipation rates below the ther- but also of the concommitent SST variations. mocline are small, implying that the energy flux Many new data sets will become available through the thermocline is also small. Hence from ongoing and planned field programs in the most of the energetic motion forced at the sur- equatorial oceans: EPOCS (Equatorial Pacific face is trapped above the thermocline. This Ocean Climate Studies) in the eastern Pacific; leads us to believe that we can model the essen- Tropic Heat in the central Pacific; SEQUAL(Sea- tial features of upper ocean variability without sonal Equatorial Atlantic Experiment) and FOCAL considering the coupling to the deeper ocean (French Ocean Climate Atlantic Experiment) in below. Just why this is so is not altogether the Atlantic. We expect a continuing and clear. There has been some work related to this deepening interest in the problem of the E1 Nino issue by Philander (1978) and McCreary (see phenomenon theough the ENSO (El Nino- Southern McCreary, Moore and Picaut, 1982), but m• Oscillation) program and the E1 Nino Rapid remains to be done on the influence of variable Response experiment, both under the overall stratification and realistic dissipation. The aegis of TOGA (Tropical Ocean - Global Atmo- effects of mean currents, notably the equatorial sphere). We expect that satellite remote sensing undercurrent and the vigorous meridional circu- and telemetry will play an increasingly impor- lation associated with it, on the vertical pro- tant role in the execution and design of these pagation of waves has not yet been considered. and future field programs. Additional questions remain about the Finally, we expect oceanographers to become degree of realism needed in modelling the stra- more and more interested in the general role of tification in and above the thermocline. Prel- the ocean in climate and increasingly aware of iminary results already indicate that the the problems of heat transport in the oceans. description of sea level variability requires This will require additional study of the gen- more than a single baroclinic mode and that the eral nature of couplings at the air-sea inter- thickness of the thermocline itself undergoes face and the particular question of how to variations on seasonal time scales. McCreary parameterize vertical heat fluxes through the (1981a,b,) has long emphasized the need to con- ocean surface. As our capacity to make measure- sider high order modes in order to model the ments in the equatorial oceans increases, the vertical structure of the current system. As requirement for more accurate surface winds more ocean data are collected, and as climato- becomes more pressing. 1146 Cane and Sarachik: Equatorial Oceanography
We close by noting that the three major Acknowledgements: This work was supported by oceans differ in size, geography, and wind forc- NASA Grant NGR 22-009-727 and NSF Grant 0CE- ing. It will be through the comparative study 79220•6 at MIT and by NASAGrant NGL 22-007-228 of equatorial processes in all three oceans that at Harvard. We wish to thank Dr. D.E. Harrison our understanding of equatorial oceanography for critically reading the paper and D.F. Fran- will be most profoundly increased. zosa for carefully preparing the manuscript.
References
Allen, J.S. and R.D. Romea,On coastal trapped mulation and analytic results, •. Mar. •es., Duing, W., F. Ostapoff and J. Merle, eds. • waves at low latitudes in a stratified ocean, 3_•, 233-252, 1979. •. Fluid Mech., 98, 555-585, 1980. calOceano•ranh¾ofthe Tronical Atlantic Dur- Cane, M.A., The response of an equatorial ocean JJIgGATE, University of Miami Press, 117 p., Anderson, D.L.T. and D.W. Moore, Cross- to simple wind stress patterns: II. Numerical 1980. equatorial inertial jets with special results., •. Mar. Res.,•?, 253-299, 1979. Duing, W., ed., Ecuatorial andS-Scale Oeeanog- relevance to very remote forcing of the Somali Cane, M.A. and E.S. Sarachik, Forced bareclinic current, DeeD-SeaRes., 26A, 1-22, 1979. ranhy, Supp. II to Deep-Sea Res.,26A, Pergamon ocean motions, III: The linear equatorial Press, pp. 356, 1980. Angell, J.K., Comparison of variations in atmos- basin case,•. Mar. •es., 3/, 355-398, 1979. pheric quantities with sea surface temperature Elliott J.A. and N.S. Oakey, Average microstruc- Cane, M.A., On the dynamicsof equatorial ture levels and vertical diffusion for Phase variations in the equatorial eastern Pacific, currents, with application to the Indian Non. Wea. Rev., 10•, 230-243, 1981. III, GATE, Oceano•raohyandSurfac9 Layer Ocean, DeeD-Sea,es., 27A, 525-544, 1980. MeteoroloRvinthe•/• Scale, Supp. I to Bacastow, J.A., J.A. Adams, C.D. Keeling, D.J. Cane, M.A. and D.W. Moore, A note on low- Moss, T.P. Whorf and C.S. Wong, Atmospheric Deep-Sea Res., 26A, eds. G. Sledlet and J.D. frequency equatorial basin modes,•. Phys. Woods, PergamonPress, 273-294, 1980. carbon dioxide, the Southern Oscillation, and Oceano•r., 11, 1578-1585, 1981. the weak 1975 E1 Nine, Science, 210, 66-68, Egger, J., G. Meyers and P. B. Wright, Pressure, Cane, M.A. and E.S. Sarachik, The response of a wind and cloudiness in the tropical Pacific 1980. linear bareclinic equatorial ocean to periodic Barnett, T.P. and W.C. Patzert, Scales of ther- related to the southern oscillation, Mon. Wea. forcing, •. Mar. Res., •9, 651-693, 1981. Rev., 10q, 1139-1149, 1981. mal variability in the tropical Pacific, •. Cane, M.A. and Y. DuPenhoat, The effect of Phys. Oceanogr., 10, 529-540, 1980. Enfield, D.B. and J.S. Allen, On the structure islands on low frequency equatorial motions, and dynamics of monthly mean sea level Barnett, T.P., Statistical prediction of North •. Mar. Res., •0 1981. American air temperatures from Pacific predic- anomalies along the Pacific coast of North and Cane, M.A. and R.J. Patton, A numerical model South America, •. Phys. Oceanogr., 10, 557- tors, Men. Wes. Rev., 109, 1021-1040, 1981. for low frequency equatorial dynamics, 1982. 588, 1980. Barnett, T.P., Statistical relations between Cane, M.A. and R.J. Patton, Sources of coastal ocean/atmosphere fluctuations in the tropical Enfield, D.B., Thermally driven wind variability upwelling in the Gulf of Guinea, 1982. in the planetary boundary layer above Lima, Pacific, •. Phys. Oceano•r., 11, 1043-1058, Cane, M.A. and E.S. Sarachik, Seasonal heat 1981. Peru, •. OeoDhvs. •es., 86, 2005-2016, 1981. transport in a forced equatorial bareclinic Behringer, D.W. and H. Stommel,Annual beat gain Enfield, D. B., Annual and nonseasonal variabil- model, •. Phys. Oceanogr., 1983. ity of monthly low-level wind fields over the of the tropical Atlantic computed from subsur- Chiu, W-C. and A. Lo, A preliminary study of the face ocean data, •. Phys. Oeeanogr., 1•, southeastern tropical Pacific, Men. Wes. Rev., possible statistical relationship between the 109, 2177-2190, 1981. 1393-1398, 1981. tropical Pacific sea surface temperature and Boyd, J.P., The nonlinear equatorial Kelvin Enfield, D.B., E1 Nine Pacific eastern boundary the atmospheric circulation, Men. Wes. Rev., response to interannual forcing, in Resource wave, •. Phys. Oceano•r., 10, 1-11, 1980. 10'7, 18-25, 1979. Boyd, J.P., Equatorial solitary Waves. Part I: Mana=ementand Environmental Uncertainty, eds. Chiu, W-C., A. Lo, D.H. Weidler, Jr., and D. M.H. Glantz and J.D. Thompson,John Wiley & Rossby solirons, •. Phys. Oceanogr., 10, Fulker, A study of the possible statistical 1699-1718, 1980. Sons, Inc., 213-254, 1981. relationship between the tropical Pacific sea Eriksen, C.C., An equatorial transoct of the Boyd, J.P. and Z.D. Christidis, Low wavenumber surface temperature and atmospheric circula- instability on the equatorial beta-plane, Geo- Indian Ocean, •. Mar. Bes., •?, 215-232, 1979. tion, Men. Wes. Rev., 109, 1013-1020, 1981. Eriksen, C.C., Evidence for a continuous spec- •k•. •es. Let., 3, 769-772, 1982. Clarke, A.J., On the generation of the seasonal Brink, J.H., D. Halpern and R.L. Smith, Circula- trum of equatorial waves in the Indian 0cean, coastal upwelling in the Gulf of Guinea, •. •. Geonhys. •es., 85, 3285-3303, 1980. tion in the Peruvian upwelling system Near 15 •. •es., 8•., 3743-3751, 1979. S, •. Geonhys.•es., 85, 4036-4048, 1980. Eriksen, C.C., Deep currents and their interpre- Clarke, A.J., The reflection of equatorial waves tation as equatorial waves in the western Brink, K.H., A comparisonof lon• coastal from oceanicboundaries, submitted to •. Phys. trapped wave theory with observations off Pacific Ocean, •. Phys. Oceanogr., 11, •8-70, Oceano•r., 1982. 1981. Peru, •. •Jl•. Oceano•r., 12, 897-913, 1982. Clarke, R.A., Changesin the upper ocean within Eriksen, C.C., Equatorial wave vertical modes Brockmann,C., E. Fahrbach, A. Huyer and R.L. the C-scale array during Phase III, GATE-1 Smith, The polewardundercurrent along the observed in a Western Pacific island array, •. Oceano•ranh•andSurfaeeLa•erMeteorolo•vin •. Res., 8?, 1982. Peru coast: 5 to 15 S, DeeD-SeaRes., 27A, the•/• Scale, Supp. I to Deep-SeaRes. 26A, 847-856, 1980. Ericksen, C.C., C-eostrophic equatorial deep eds. 6. Siedler and J.D. Woods,Pergamon Press, jetc, •. Mar. Res., _ZLQ.,SueD, 143-157, 1982. Brown, O.B., J.G. Bruce and R.H. Evans, Evolu- 115-127, 1980. tion of sea surface temperature in the Somali Evans, R.E. and 0.B. Brown, Propagation of ther- Cochrane, J.D., F.J. Kelley, Jr., and C.R. mal fronts in the Somali current system, basin during the southwest monsoonof 1979, 01ling, Subthermooline countercurrents in the Solenos, 209, 595-597, 1980. Deco-Sea,es., 28A. 521-527, 1981. western equatorial Atlantic Ocean, •. Phys. Flohn, H., Oceanic upwelling as a key for abrupt Brown, O.B., Observation of long period sea sur- Oceanogr., 3, 724-738, 1979. face temperature variability during GATE, climatic change, •. Met. Sec. Janan, 60, Cox, M.D., A numerical study of Somali current 268-273, 1982. Ecuatorial and A-Scale Oceanography, Supp. II eddies, •. Phys. Oeeanogr.,R, 311-326, 1979. to Deep-Sea Res., 26A, ed. W. Duing, Pergamon Garzoli, S. and E.J. Katz, Observations of Cox,M.D., Generationand propagationof 30-day inertia-gravity waves in the Atlantic from Press, 103-124, 1978. wavesin a numerical model of the Pacific, •. Bruce, J.G., Eddies off the Somali coast during inverted echo sounders during FGGE,•. Phys. •Jl•. Oeeano•r., 10, 1168-1186, 1980. Oceanogr. 11, 1463-1473, 1981. the southwest monsoon, •. Geonhvs. Res., 8• Cox, M.D., A numerical study of surface cooling 7742-7748, 1979. Garzoli, S.L., E.J. Katz, H.-J, Panitz and P. processes during summer in the Arabian Sea. in Speth, In situ wind measurementsin the equa- Bruce, J.G., D.R. Quadfasel and J.C. Swallow, Monsoon Dynamics, eds. J. Lighthill & R. P. Somali eddy formation during the commencement torial Atlantic during 1979, Oceanolo•ica Pearce, CambridgeUniversity Press, 529-540, Acta, •, 281-288, 1982. of the southwest monsoon, 1978, •. Geephys. 1981. Gent, P., Standing equatorial wave modes in •es., 856654-6660, 1980. Crawford, W.R., Pacific equatorial turbulence, Bruce, J.G., M. Fieux and J. Gonella, A note on bounded ocean basins, •. Phys. Oeeano•r., 3, •. Phys. Ooeano•r. 12, 1137-1149, 1982. 653-662, 1979. the continuance of the Somali eddy after the Crawford, W.R. and T. R. 0shorn, Microstructure Gent, P.R. and A.J. Semtner, Jr., Energy trap- cessation of the Southwest monsoon, Oceanolo- measurements in the Atlantic equatorial under- ping near the equator in a numerical ocean •ieaActa, •, 7-9, 1982. current during GATE, Eouatorial andS-Scale model, •. •]LV_•. Oceano•r., 10, 823-842, 1980. Bruce, J.G. and W.H. Beatty, III, Observations Oeeano•ranhy, Supp. II to Deep-Sea Res., 26A, of the diurnal heating in the western equa- ed. Walter Duing, Pergamon Press, 285-308, Gent, P.R., Forced standing equatorial ocean torial Indian Ocean, Ooeanolo•iqsAeta, •. 1980. wave modes, •. Mar. Res., •9, 695-709, 1981. 137-139, 1982. Crawford, W.R. and T. R. Osborn, Energetics of Gibbs, R.J., Wind-controlled coastal upwelling Bryan, K., Poleward heat transport by the ocean: the Atlantic equatorial undercurrent, Eeua- in the western Atlantic, Deep-Sea,es., 27, Observations and models. Ann. Rev. Earth torial and A-Scale Oceano•ranhv, Supp. II to 857-866, 1980. Planet. S_9JE., 15-38, 1982. Deep-Sea Res.,26A, ed. Walter Duing, Pergamon Goldenberg, S.B., and J.J. O'Brien, Time and Bunker, A.F., Trends of variables and energy Press, 309-323, 1980. space variability of tropical Pacific wind fluxes over the Atlantic Ocean from 1948 to Crawford, W.R. and T.R. Osborn, The control of stress, Men. Wes. Rev., .l•, 1190-1207, 1981. 1972, Men. Wes. •ev., 108, 720-732, 1980. equatorial ocean currents by turbulent dissi- Gower, J.F.R., ed., •.•from Sncae, Busalacchi, A.J. and J.J. O'Brien, The seasonal pation, Science, 212, 539-540, 1981. Plenum Press, pp. 978, 1981. variability of the tropical Pacific, •. Phys. Duing, W. and Z. Halleck, Equatorial waves in Green, K. and V.T. BuOhwald, Interior shelf Ooeano•r., 10, 1929-1952, 1980. the upper central Atlantic, GATE-2, Ecuatorial waves on an equatorial - plane, •. Phys. Busalacchi, A.J. and J.J. O'Brien, Interannual andS-Scale Oeeano•ranh¾, Supp. II to Deep-Sea Oceanogr., 3, 1299-130•, 1979. variability of the equatorial Pacific in the Res. 26A, 26, ed. W. Duing, PergamonPress, Grose, P.L., Transition in water mass properties 1960's, •. Geonhvs.Res., 86, 10901-10907, 161-178, 1980. 1981. between 200 and 500 meters depth in the GATE Duing, W., R. L. Molinari and J. C. Swallow, B-array, in Oceano•r•nhyandSurfac9 Laver Busalacchi, A.J. and J. Picaut, Seasonal varia- Somali current: evolution of surface flow, Meteorolo•k-V in the]•/• Scale, Supp. I to bility from a model of the tropical Atlantic, Science, 209, 588-589, 1980. Deep-Sea Res., 26A, eds. 6. Siedler and J.D. •. Phys. Oeeano•r., 1982. Duing, W., and A. Leermss, Arabian Sea cooling: Woods, Pergamon Press, 147-160, 1980. Cane, M.A., The response of an equatorial ocean A preliminary heat budget, •. Phys. Oeeano•r., to simple wind stress patterns: I. Model for- Halleck, Z., On wind-excited, equatorially 10, 307-312, 1980. trapped waves in •he presence of mean Cane and Sarachik: Equatorial Oceanography 1147
currents, in Ecuatorial and A-Scale Qceano•ra- face in the equatorial Atlantic, J. Mar. Res., MaCteary, J.P., Jr., A model of tropical ocean- •hv, Supp. II to Deep-Sea Res., 26A, ed. W. 3•, 53-63, 1981. atmosphere interaction, Yon. Wes. Rev., 110, Duing, PergamonPress, 261-284, 1980. Katz, E.J. and S. 6arzoli, Response of the 1982. Halpern, D., Surface wind measurements and low- western equatorial Atlantic ocean to an annual MaCteary, J.P., J. Pieaut and D.W. Moore, Effect level cloud motion vectors near the intertrop- wind cycle, J. Mar. Res., •0, SUDD., 307-327, of annual remote forcing in the eastern tropi- cal Atlantic, J. Phys. Oceanogr., 1982. ical convergence zone in the central Pacific 1982. ' Ocean from November 1977 to March 1978, Mon. Keshavamurty, R.N., Response of the atmosphere McPhaden, J. and R.A. Knox, Equatorial Kelvin Wea. Rev., 107, 1524-1534, 1979. to sea surface temperature anomalies over the and inettic-gravity waves in zonal shear flow., J. Phys. Oceano•r., 9, 263-277, 1979. Halpern, D., A Pacific equatorial temperature equatorial Pacific and the teleconnections of section from 172 E to 110 W during winter and the southern oscillation, J. Atmos. Sci., 3•, McPhaden, M.J., Continuously stratified models spring 1979, Deco-Sea Res., 27A, 931-940, 1241-1259, 1982. of the steady-state equatorial ocean, J. Phys. Oceano•r., 11, 337-354, 1981. 1980. Kindle, J.C., Equatorial Pacific ocean variabil- Halpern, D., Variability of near-surface ity - seasonal and E1 Nino time scales, Ph.D. McPhaden, M.J., Variability in the central equa- torial Indian Ocean - Part I: Ocean dynamics, currents in the Atlantic north equatorial Dissertation, Dept. of Oceanography, Florida countercurrent during 6ATE, J. Phys. Ocean., State University, pp. 139, 1979. J. Mar. Res., •Jl, 157-176, 1982. 10, 1213-1220, 1980. Knox, R.A., Time variability of Indian Ocean MePhaden, M.J., Variability in the central equa- torial Indian Ocean, Part II: Oceanic heat and Harrison, D.E. and D.S. Luther, Observing long- equatorial currents, DeeD-Sea Res., 28A, 291- turbulent energy balances, J. Mar. Res., •Jl, period fluctuations of surface winds in the 295, 1981. tropical Pacific: Preliminary results from Knox, R.A. and D. Halpern, Long range Kelvin 403-q19, 1982. Island data, submitted Yon. Wes. Rev., 1982. wave propagation of transport variations in MaPLaden, M.J. and A.J. Semtner, The effects of mean shear flow on equatorial basin modes, J. Hastenrath, S. and P.J. Lamb, Climatic Atlas of Pacific ocean equatorial currents, J. Mar. the Indian Ocean, Part I: Surface climate and Res., qO, Suoo., 329-339, 1982. Phys. Oceano•r., 12, 795-804, 1982. atmospheric circulation, The University of Eraus, E.B. and H.P. Hanson, Air-Sea interaction Meehl, G.A., Characteristics of surface current flow inferred from a global ocean current data Wisconsin Press, 1979. as a propagator of equatorial ocean surface Hastenrath, S. and P.J. Lamb, Climatic Atlas of temperature anomalies, J. Phys. Oceano•r., set, J. Phys. Oceano•r., 12, 538-555, 1982. the Indian Ocean - Part II: The oceanic heat 1982. Merle, J. and J.Le Floch, Cycle annuel moyen de la temperature dans les couches superieures de budget, The University of Wisconsin Press, Krishnamufti, T.N., and R. Krishnamufti, Surface l'Ocean Atlantique intertropical, Oceancivics 1979. meteorology over the 6ATE A-scale, in Ecua- Acta, 1, 271-276, 1978. Hastenrath, S., Heat budget of tropical ocean torial and A-Scale Oceano•raohv, Supp II to Merle, J., M. Fieux and P. Hisard, Annual Signal and atmosphere, J. Phys. Oceano•r., 10, 159- Deep-Sea Res. ,26A, ed. W. Duing, Pergamon and interannual anomalies of sea surface tem- 170, 1980. Press, 29-61, 1980. Hastenrath, S. and P.J. Lamb, On the heat budget Lamb, P.J. and A.F. Bunker, The annual march of perature in the eastern equatorial Atlantic Ocean, Equatorial and A-Scale Oeeano_•raohv, of hydrosphere and atmosphere in the Indian the heat budget of the north and tropical Ocean, J. Phys. Oceano•r., 10, 694-708, 1980. Atlantic Oceans, J. Phys. _O_•ano_•r., 1982. Supp. II to Deep-Sea Res., 26A, ed. W. Duing Hayes, S.P. and H.B. Milburn, On the vertical Lass, H.U., V. Bubnov, J.M. Huthnance, E.J. PergamonPress, 77-101, 1980. structure of velocity in the eastern equa- Katz, J. Meinke, A.deMesquita, F. Ostapoff Merle, J., Variabilite thermique annuelle et torial Pacific, J. Phys. Oceano•r., 10, 633- and B. Voiturez, Seasonal changes of the zonal interannuelle de l'ocean Atlantique equatorial 635, 1980. pressure gradient in the equatorial Atlantic Est. L'hypothese d'un •E1 Nino• Atlantique, Hayes, S.P. and C.J. Powell, Vertical wave during the FGGEyear, Oceanolo•ica Acta, 1982. Oceanoloeica Acta, •, 209-220, 1980a. number spectra of temperature fine structure Lau, K-M.W., A numerical study of tropical Merle, J., Seasonal variation of heat-storage in in the Equatorial Pacific, J. Geonhvs. Res., large-scale air-sea interaction, J. Atmos. the tropical Atlantic Ocean, •l• 85. 4029-4035, 1980. .•L., •6, 1467-1489, 1979. Acta, lq80, .•,, 455-•3, 1980b. Hayes, S.P., Vertical fine structure observa- Lau, K-M.W., Oscillations in a simple equatorial Merle, J. Seasonal heat budget in the equatorial tions in the eastern equatorial Pacific, J. climate system, J. Atmos. Sci., •8, 248-261, Atlantic Ocean, J. Phys. Oceano•r., 10, 464- G•oohvs. Res., 86, 10983-10999, 1981. 1981. •69, 1980c. Hayes, S.P., J.M. Toole and L.J. Mangum,Water Leetmaa, A. and H. Storereel, Equatorial current Meyers, 6., On the annual Rossby wave in the mass and transport variability at 100 W in the observations in the western Indian ocean in tropical North Pacific Ocean, J. Phys. equatorial Pacific, 1982. 1975 and 1976, J. Phys. Oceano=r., 10, 258- Oceano•r., 9, 663-674, 1979. Hayes, S.P., A comparison of geostrophic and 269, 1980. Meyers, 6., Annual variation in the slope of the measured velocities in the Equatorial Under- Leetmaa, A., H.T. Rossby, P.M. Saunders and P. 14 C isotherm along the Equator in the Pacific current, J. Mar. Res., •Jl, SMD•t., 219-229, Wilson, Subsurface circulation in the Somali Ocean, J. Phys. Oceano•r, 9, 885-891, 1979. 1982. current, Science, 209, 590-592, 1980. Meyers, 6., Do Sverdrup transports account for Hellerman, S., Charts of the variability of the Leetmaa, A., J.P. McCreary, Jr., and D.W. Moore, the Pacific north equatorial countercurrent?, wind stress over the tropical Atlantic, in Equatorial currents; observations and theory, J. Geonhvs. Res., 85, 1073-1075, 1980. Ecuatorial and A-Scale •, Supp. II in Evolution of Physical Oceano•raohv, ads. Meyers, 6., Interannual variation in sea level to Deep-Sea Res. 26A, ed. W. Duing, Pergamon B.A. Warren and C. Wunsch, 18q-196, MIT near Truk Island - A bimodal seasonal cycle, Press, 63-75, 1980. Press, 1981. J. Phys. Oeeano•r., 12, 1982. Hellerman, S., Normal monthly windstress over Leetress, A. and P.F. Spain, Results from a velo- Meyers, 6., W. White and K. Hasunuma,Annual the world ocean with error estimates, J. Phys. city transect along the equator from 125 to variation in baroelinie structure of the Oceano•r., 1982. 159 W, J. Phys. Oceano=r., J_l, 1030-1032, northwestern tropical Pacific, Oeeano•r. troo. Hisard, P., Observation de responses de type 1981. 1, 59-69, 1982. <
terization of vertical mixing in numerical 111, 1982. 2279-2292, 1979. models of tropical o•eans, J. Phys. Oceanomr., Ripa, P., Resonant interactions of equatorial Weare, B.C., P.T. Strub and M.D. Samuel, Annual 11, 1443-1451, 1981. waves in a one layer model I. Non local mean surface heat fluxes in the tropical Pazan, S.E., and G. Meyers, Interannual fluctua- triads and triads of waves with the same Pacific ocean, J. •J•. 9_•., 11, 705- tions of the tropical Pacific wind field and speed, 19 82. 717, 1981. the southern oscillation, Mon. Wes. Rev., 110, Ripa, P., Resonant interactions of equatorial Weare, B.C. and P. Ted Strub, Annual mean atmos- 587-600, 1982. waves in a one layer model II. General clas- pheric statistics at the surface of the tropi- Perkins, H., Low-frequency forcing of the tropi- sification of resonant triads and double cal Pacific ocean, Mon. Wes. Rev., 10q, 1002- cal atlantic ocean under ITCZ during GATE, in triads, 1982. 1012, 1981. OceanograPhY and Surface Laver Meteorology •n Roemmich, D., The balance of geostrophis and Weare, B.C., E1 Nino and tropical Pacific ocean the •/C Scale, Supp. I to Deep-Sea Res., 26A, Ekman transports in the tropical Atlantic, J. surface temperatures, J. Phys. Oceano=r., 12, eds. G. Sledlet and J.D.Woods, Pergamon Press, Phys. Oceano•r., 1982. 17-27, 1982. 225-236, 1980. Romea, R.D., and J.S. Allen, On forced coastal Wears, B.C. and J.S. Nasstrom, Examples of Philander, S.G.H., Instabilities of equatorial trapped waves at low latitudes in a stratified extended empirical orthogonal function ana- zonal currents II, J. Geoohvs. •es., ocean, J. Mar..•..0.., ..a,.•, 369-401, 1982. lyses, Mon. Wes. Rev., 110, 481-504, 1982. 3679-3682, 1978. Romea, R.D., and J.S. Allen, On vertically pro- Weisberg, R.H., A. Horigan and C. Colin, Equa- Philander, S.G.H., Nonlinear coastal and equa- pagating coastal Kelvin waves at low lati- torially trapped Rossby-gravity wave propaga- torial jets, J. Phys. Oceanocr., 9, 739-747, tudes, J. •J•. •ceano•r. submitted, 1982. tion in the Gulf of Guinea, J. Mar. Res., 1979. Sarachik, E.S., Tropical sea-surface tempera- 67-86, 1979. Philander, S.G.H., Upwelling in the Gulf of ture: An interactive one-dimensional Weisberg, R.H. and A.M. Horigan, Low-frequency Guinea, J. Mar. •es., 33., 23-33, 1979. atmosphere-ocean model, ]lYJX. Atmos. and Oce- variability in the equatorial Atlantic, J. Philander, S.G.H., Equatorial waves in the pres- ans, 2, 455,469, 1978. Phys. Oceano=r., 11, 913-920, 1981. ence of the equatorial undercurrent, J. •J•. Siedler, G. and J.D. Woods, eds., Oceano•ranhv Weisberg, R.H., Equatorial waves during GATE and Oceanomr., i, 254-262, 1979. and Surface Laver Meteorology •n the •/C_ their relation to the mean zonal circulation, Philander, S.G.H., Variability of the tropical Scale, Supp. I to Deep-Sea Res., 26A, Pergamon E•uatorial and A-Scale Oceano=ranhv, Supp. II oceans, Dvn. of Atmos. and Oceans, •[, 191-208, Press, pp. 294, 1980. to Deep-Sea Res., 26A, ed. W. Duing, Pergamon 1979. Schopf, P.S., The role of Ekman flow and plane- Press, 179-198, 1982. Philander S.G.H., and R.C. Pacanowski, The gen- tary waves in the oceanic cross-equatorial Weisberg,R.H., L. Miller, A. Horigan•nd J.A. eration of equatorial currents, J. Geoohvs. heat transport, J. Ph•s. Oceano•r., 10, 330- Knauss, Velocity observations in the equa- Res., 85, 1123-1136, 1980. 341, 1980. torial thermocline during GATE, Eouatorial and Philander, G. and Walter Duing, The oceanic Sehopf, P.S., D.L.T. Anderson and R. Smith, A-Scale •.9•, Supp. II to Deep-Sea circulation of the tropical Atlantic and its Beta-dispersion of low-frequency Rossby Waves, Res., 26A, ed. W. Duing, Pergamon Press, 217- variability, as observed during GATE, in Eoua- Dvn. of Atmos. and Oceans, 5, 187-214, 1981. 248, 1980. torial and A-Scale •, Supp. II to Sehopf, P.S. and M.A. Cane, On equatorial dynam- Williams, R.G. and R.E. Dennis, Variability of Deep-Sea Res. ,26A, ed. Walter Duing, Pergamon ics, mixed layer physics and sea surface tem- the upper ocean mass field in the eastern Press, 1-27, 1980. perature, J. Phys. Oeecnoir. submitted, 1982. tropical Atlantic during GATE, Oceano_•ranhv Philander, S.G.H., The equatorial undercurrent Sehopf, P.S. and D.E. Harrison, On equatorial and Surface Laver Meteorolo• •n the •/C revisited, Ann. Rev. Earth Planet. Sci., 8, Kelvin waves and E1 Nino: I. Effects of sim- Scale, Supp. I to Deep-Sea Res., 26A, eds. G. 191-204, 1980. ple mean wind fields, 1982. Sledlet and J.D. Woods, Pergamon Press, 129- Philander, S.G.H., The response of equatorial Sehott, F. and D.R. Quadfasel, Development of 146, 1980. oceans to a relaxation of the trade winds, J. the subsurface currents of the northern Somali Witte, J., Physical processes in the equatorial of .•..,,%. Oceano•,r., 11, 176-189, 1981. current gyre from March to July 1979, Science, regions of the oceans: A bibliography, Philander, S.G.H. and R.C. Pacanowski, The oce- 20•, 593-595, 1980. Nova/NYIT Press, pp. 73, 1980. anic response to cross-equatorial winds (with Sehott, F. and J. Fernandez-Partagas, The onset Worthem, A. and G. Mellor, Turbulence closure application to coastal upwelling in low lati- of the summer monsoon during the FGGE 1979 model applied to the upper tropical ocean, tudes), Tellus, •, 201-210, 1981. experiment off the East African coast: A com- Oceano=rauhv and Surface Laver Meteorology •n Philander, S.G.H. and R.C. Pacanowski, Response parison of wind data oclicoted by different the •/C Scale, Supp. I to Deep-Sea Res., 26A, of equatorial oceans to periodic forcing, J. means, J. GeoDhvs. Res., 86, 4173-4180, 1981. eds. G. Sledlet and J.D. Woods, Pergamon Oeoohvs. •es., 86, 1903-1916, 1981. Schott, F. and D.R. Quadfasel, Low frequency Press, 237-272, 1980. Philander, S.G.H., and J.-H. Yoon, Eastern boun- fluctuations off Somalia during the 1979 sum- Wylie, D.P. and B.B. Hinton, Some statistical dary currents and coastal upwelling, J. Phys. mer monsoons, J. •J•XA. Oceano=r., 1982. characteristics of cloud motion winds measured Oceano•r. 12, 862-879, 1982. Schott, F. and D.R. Quadfasel, Variability of over the Indian ocean during the summer mon- Philander, S.G.H., and P. Delecluse, Coastal the Somali current system during the onset of soon, Mon. Wea. Rev., 10q, 1810-1812, 1981. currents in low latitudes (with application to the southwest monsoon, 1979, J. •J•. Wylie, D.P. and B.B. Hinton, The feasibility of the Somali and E1 Nino currents), Deed Sea Oceno=r., 1982 ß estimating large-scale surface wind fields for •es., to appear, 1982 Semtner, A.J., Jr., and W.R. Holland, Numerical the summer MONEXusing cloud motion and ship Philander, S.G.H., E1 Nino - Southern Oscilla- simulation of equatorial ocean circulation. data, Bound-Laver Met., 21, 357-367, 1981. tion phenomena, Nature, 1983. Part I: A basic case in turbulent equili- Wylie, D.P., B.B. Hinton and K.M. Millerr, A Preller, R. and J.J. O'Brien, The influence of brium, J. Phys. Oceano_•r., 10, 667-693, 1980. comparison of three satellite based methods bottom topography on upwelling off Peru, J. Schaffer, G., On the upwelling circulation over for estimating surface winds over oceans, J. Ph•s. Oceano_•r., 10, 1377-1398, 1980. the wide shelf off Peru: 1. Circulation, J. Aool. Met., 20,, 438-449, 1981. Quadfasel, D.R., Low frequency variability of Mar. Res., •Jl 293-314, 1982. Wylie, D.P. and B.B. Hinton, The wind stress the 20 C isotherm topography in the western Smith, S.L. and L.A. Codispoti, Southwest mon- patterns over the Indian Oceans during the equatorial Indian Ocean, J. Geoohvs •es., 87, soon of 1979: Chemical and biological summer monsoon of 197 9., J. Phys. Oceanocr. 1990-1996, 1982. response of Somali coastal waters, Science, 12, 186-199, 1982. Quadfasel, D.R. and F. Schott, Water-mass dis- 209, 597-600, 1980. Wyrtki, K., E1 Nino - The dynamic response of tributions at intermediate layers off the Stevenson, J.P. and P. Nillet, Tropical heat the equatorial Pacific ocean to atmospheric Somali coase during the onset of the southwest flux and storage during the FGGE shuttle forcing, J. Phys. Oceanocr., 5, 572-584, 1975. monsoon, 1979, J. Phys. Oceano_•r., 1982. experiment, 1982. Wyrtki, K., The response of sea surface topogra- Ramage, C.S., S.J.S. Khalsa and B.N. Meisner, Swallow, J.C., Observations of the Somali phy to the 1976 E1 Nino, J. •J•. 0usahoer., The central Pacific near-equatorial conver- current and its relationship to the monsoon •t, 1223-1231, 1979. gence zone, J. Geonh_v••es., 86, 6580-6598, winds, Moonscon Dvn•ies, eds. J. Lighthill Wyrtki, K., E. Firing, D. Halpern, R. Knox, O.J. 1981. and R.P. Pearce, Cambridge University Press, MeNally, W.C. Patzert, E.D. Stroup, B.A. Taft Ramage, C.S. and A.M. Hori, Meteorological 444-452, 1981 ß and R. Williams, The Hawaii to Tahiti shuttle aspects of E1 Nino, Mon. Wes. Rev., 10q, Toole, J.M., Anomalous characteristics of equa- experiment, Science, 211, 22-28, 1981. 1827-1835, 1981. torial thermohaline finestructure, J. Phys. Wyrtki, X., An estimate of equatorial upwelling Rammasson, E.M. and T.H. Carpenter, Variations Oeecnoir., 11, 871-876, 1981 ß in the Pacific, J. Phys. Casanoir., 11, in tropical sea surface temperature and sur- Tsuchiya, M., Seasonal variation of the equa- 1205-1214, 1981. face wind fields associated with the southern torial zonal geopotential gradient in the Wyrtki, K., Eddies in the Pacific north equa- oscillation/E1 Nino, Mon. Wes. Rev., 110, eastern Pacific Ocean, J. Mar. Res., 33., 399- torial current, J. Phys. Casanoir., 12, 7•6- 354-384, 1982. 407, 197 9. 749, 1982. Rasmusson, E.M., and T.H. Carpenter, The rela- Tsuchiya, M., The origin of the Pacific equa- Wyrtki, K, The southern oscillation, ocean- tionship between eastern equatorial Pacific torial 13 C water, J. Phys. paeanoar., 11, atmosphere interaction and E1 Nino, Mar. Tech. sea surface temperatures and summer monsoon 794-812, 1981. Soe. J., 16, 3-10, 1982. rainfall over India, May, 1982. Verstraete H-M., J. Pieaut and A. Morliere, Wyrtki, K. and G. Eldin, Equatorial upwelling Reed, R.J. and R.M. Lewis, Response of upper Atmospheric and tidal obsevations along the events in the central Pacific, J. Ph•s. ocean temperatures to diurnal and synoptic- shelf of the Guinea Gulf, • and A- Oceano•r., 12, 984-988, 1982. scale variations of meteorological parameters Scale Oceano_•raohv, Supp. II to Deep-Sea Res., Yoon, H-G., Effects of islands on equatorial in the GATE B-Scale area, • and 26A, ed. W. Duing, Pergamon Press, 343-356, waves, J. Geoohvs. Res., 86, 10913-10920, Surface Laver Meteorology •n the •/C_ Scale, 1980. 1981 ß Supp. I to Deep-Sea Res.26A, eds. G. Siedler Walsh, J.J., R.W. Whirledge, W.E. Esaias, R.L. Yoon, J-H. and S.G.H. Philander, The generation and J.D. Woods, PergamonPress, 99-114, 1982. Smith, S.A. Huntsroans, H. Santander and B.R. of coastal undercurrents, J. pecanoar. Rips, P. and S.P. Hayes, Evidence for equatorial De Mendiola, The spawning habitat of the Peru- Jaoan, in press, 1982. trapped waves at the Galapagos Islands, J. vian anchovy, En_•raulis •in_•ens, DeeD-Sea •es. Geoohvs. Res., 86, 6509-6516, 1981. 27A, 1-27, 1980. Ripa, P., Nonlinear wave-wave interactions in a Wears, B.C., A statistical study of the rela- one-layer reduced-gravity model of the equa- tionships between ocean surface temperatures (Received October 29, 1982; torial •-plane, J. Phys. Oeecnoir., 12, 97- and the Indian monsoon., J. Atmos. Sci., •6, accepted January 10, 1983.)