All information below was supplied by the workshop participants. Participants' names are in boldface; some participants are co-authors of more than one poster. Abstracts are listed in alphabetical order by presenter. Additional authors names are included below the titles.

Helene Banks
Hadley Centre for Climate Prediction and Research, Meteorological Office, Bracknell, UK

Indian Ocean Circulation and Heat Budget in a Coupled Climate Model

The circulation in the Indian Ocean of a coupled climate model (HadCM3) is examined. The model circulation shows a strong subtropical gyre in the surface, mode and intermediate waters, with Indonesian throughflow waters increasing the outflow of surface and intermediate waters. The circulation of deep and bottom water is much weaker than observational estimates with 3.5 Sv flowing northward below 2000 m across 32°S compared with estimates of O (11-28 Sv). The Indonesian throughflow in HadCM3 has a magnitude of 24 Sv, which we show to be attributable to both the windstress curl (driving a near surface core?) and the bottom pressure torque (driving a deep core?).

The heat budget is analysed and compared with observational estimates. The flux divergence over the Indian Ocean is O (-0.1 PW), which is considerably less than observational estimates. We consider two sensitivity experiments changing the throughflow. As the throughflow is reduced, the heat flux divergence over the Indian Ocean increases (as expected) but the sensitivity is less than 0.02 PW per Sv. When the net barotropic flux is zero (but baroclinic exchange is allowed), the increase in the flux divergence is mainly achieved by an increase in the magnitude of the over-turning of the Indian Ocean, with flow below 2000 m across 32°S increasing from 3 Sv to 10 Sv.

Swadhin Behera
Frontier Research System for Global Change, Tokyo, Japan

Variability in the Indian Ocean
S.K. Behera^1,2, P.S. Salvekar², and T. Yamagata^1,3
1 Institute for Global Change Research, Tokyo, Japan
² Indian Institute of Tropical Meteorology, Pune, India
³ Dept of Earth and Planetary Physics, Graduate School of Science, University of Tokyo

The seasonal and interannual variability in the dynamic and thermodynamic characteristics of the Indian Ocean (north of 30°S) are investigated using results of a 10-year model simulation. A two-layer, moderately complex numerical ocean model is used to simulate the variability. The model is forced by the mean monthly surface heat flux and momentum flux. The flux fields are derived from National Center for Environment Prediction (NCEP) reanalysis data sets for the period 1982-91.

In general, the upper layer currents are well simulated in both the simulations. The phase and amplitude of the time series of SST anomalies compare well with that of the observed SST anomalies. Interannual variability in the surface heat forcing plays a key role in producing the realistic model SST anomalies. The influence of interannual wind is only limited to regions near to the coast in the Arabian Sea during the Indian summer monsoon season and a small region in the central Indian Ocean (south of 5°S) from November through January. Warm anomalies of the order of 0.5°C are found to cover the whole equatorial region during 1982-83, 1987-88, and 1990-91 in both the simulations and the observation. These periods coincide with the El Niño events of the equatorial Pacific Ocean.

Harry Bryden
Southampton Oceanography Centre, UK

Observations of the Agulhas Current at 32°S
Harry L. Bryden¹ and Lisa M. Beal^2
1 Southampton Oceanography Centre
² Lamont-Doherty Earth Observatory, Columbia University

The full depth velocity field of the Agulhas Current, the western boundary current of the southwest Indian Ocean, has been directly measured using a Lowered Acoustic Doppler Current Profiler (LADCP). Fifteen combined CTD O₂/LADCP stations were occupied across the current at 32°S off the east coast of South Africa in February-March 1995, as part of the WOCE Indian Ocean project. The deep velocity structure of the Agulhas Current was found to be very different than previously described using geostrophic estimates. In particular, LADCP results reveal a V-shaped pattern for the level of no motion across the Agulhas Current and an Agulhas Undercurrent is observed flowing equatorward below 800 m depth directly beneath the surface core of the south-westward flowing Agulhas Current.

Careful comparisons of direct and geostrophic velocities suggest that the velocity structure of the Agulhas Current is essentially geostrophic below about 200 to 300 m depth, where differences are generally less than estimated errors. Above these depths the LADCP-measured shears and geostrophic shears exhibit differences and velocities can diverge. The near-surface geostrophic departures are most likely explained by a combination of flow curvature and sampling biases. Since the geostrophic and directly measured shears are well matched away from the surface, a depth-averaged fit between geostrophic velocities and the LADCP measurements exhibits small standard deviations, and yields a geostrophic volume transport for the Agulhas Current of 73 Sv, just 3% less than the direct (LADCP) estimate. The most recent full depth estimate in the literature is from Toole and Warren (1993) who estimate a volume transport of 85 Sv. The baroclinic velocity structure from Toole and Warren and from this work are similar indicating that the difference in the transport estimates is due to the choice of geostrophic reference level and ultimately to the presence of the previously unobserved Agulhas Undercurrent.

Combined CTD/LADCP measurements allow transports of distinct water masses to be directly estimated. The Undercurrent primarily carries North Atlantic Deep Water and Antarctic Intermediate Water equatorward. Close to the continental slope, the upper part of the Undercurrent carries a small amount of Red Sea Water equatorward, but over the entire section the transport of Red Sea Water is poleward.

Teresa Chereskin
Scripps Institution of Oceanography

Surface Layer Circulation and Transports at 8.5°N in the Arabian Sea during June and September 1995
T. K. Chereskin, W. D. Wilson¹, H. L. Bryden², and J. Morrison^3
1 Atlantic Oceanographic and Meteorological Laboratory, Miami FL
² Southampton Oceanography Centre, U.K.
³ North Carolina State University - Raleigh

Two hydrographic/acoustic Doppler current profiler (ADCP) transects were made across 8.5°N in the Arabian Sea during June and September 1995 as part of the World Ocean Circulation Experiment's one-time survey and repeat hydrography programs. The observations document the early and late stages of the 1995 Southwest Monsoon, for which peak winds occurred in July. The surface layer velocity was dominated by the shallow (< 250 m), intense (maximum velocity > 150 cm/s), anticyclonic circulation at the western boundary (the Great Whirl). Between June and September, the Great Whirl intensified, extended further east, and freshened. The basin-integrated surface layer ageostrophic transport was consistent with Ekman tranport estimated from the wind. Ship winds were consistent with other wind estimates (e.g., FNOC NOGAPS) and with wind climatologies (Hellerman-Rosenstein, Josey).

The Ekman transport in June was about 18 Sv southward; it was partially compensated by northward geostrophic transport relative to 300 m. The ageostrophic flow was confined to a very shallow layer (about 50 m), and the surface layer was strongly stratified, with a maximum salinity layer at depths between 50 and 70 m. The Ekman layer temperature (transport-weighted temperature) was not significantly different from the surface, but the Ekman layer salinity was about 0.3 psu higher than the surface.

The Ekman transport in September was about 8 Sv southward, and the surface layer geostrophic transport was also southward. The ageostrophic velocity penetrated much deeper in September (to about 160 m), and the thermocline was correspondingly deeper. The salinity maximum layer extended to the surface in mid-basin, possibly advected from the northern Arabian Sea. The Ekman layer temperature was about 1.5°C colder than the surface, and the Ekman layer salinity was about 0.3 psu fresher than the surface value.

Christina Coatanoan
Woods Hole Oceanographic Institution

Mixing of Water Masses and Biogeochemical Processes in the Eastern Indian Ocean
C. Coatanoan^1,2, N. Metzl³, M. Fieux⁴, and B. Coste^2
1 Woods Hole Oceanographic Institution
² Laboratoire d'Oceanographie et de Biogeochimie, Campus de Luminy, Marseille
³ Laboratoire de Physique et Chimie Marine, Université Paris
⁴ Laboratoire d'Oceanographie Dynamique et de Climatologie, Université Paris

A multiparametric approach is applied in the eastern Indian Ocean to seasonally analyze the mixing coefficients of water masses and to calculate, in the water column, the quantitative fraction due to the biological processes. The data were measured during two cruises of the Java Australia Dynamic Experiment program (JADE) carried out in August 1989 (SE monsoon) and February-March 1992 (NW monsoon). Seven sources have been identified in the studied area for the 200-800 m layer: the Subtropical Indian Water (STIW), the Central Indian Water (CIW), the modified Antarctic Intermediate Water (m-AAIW), the Indonesian Surface Water (ISW), the Intermediate Indonesian Water (IIW), the Arabian Sea-Persian Gulf Water (AS-PGW), and the Arabian Sea-Red Sea Water (AS-RSW). The selected conservative tracers are potential temperature, salinity and oxygen for the seasonal approach. The nutrient data, available for the first cruise (1989), have been introduced for the biological approach. The mass conservativity and positive mixing coefficients are considered in the analysis as constraints.

The seasonal results indicate an increase of the AS-RSW contribution during the NW monsoon whereas the contribution of the south Indian waters decreases. No large changes are observed for the Indonesian waters. In the biological study, the approach consists to take into account unknowns, for the nonconservative tracers (oxygen, phosphate and nitrate), which represent contents, for each tracer, due to biological processes. The biological fractions, calculated for each nonconservative tracer, show relations (-O₂/P : -178.4 and -O₂/N : -6.9), which are close to the Redfield ratios. Furthermore, these fractions are compared to large-scale primary production estimates. The results indicate a correlation between the primary production in the surface layer and the biological fractions in depth calculated from the analysis.

Kathleen Donohoe
University of Hawaii

Velocity Measurements from Acoustic Doppler Current Profiling on WOCE Hydrographic Program Indian Ocean Sections
K.A. Donohue, E. Firing, P. Hacker, and J. Hummon

An extensive set of current measurements from two types of acoustic Doppler current profilers (ADCP) was made in addition to salinity, temperature, and chemical tracer surveys throughout the Indian Ocean as part of the WOCE Hydrographic Program (WHP). A shipboard ADCP provided high-resolution sections of the upper 300 m along the cruise track. At each CTD station, a self-contained ADCP, lowered with the CTD (LADCP), yielded a full-depth velocity profile.

The direct-velocity measurements are particularly valuable where currents are strong, such as high latitudes, the western boundary of the basin, and along the meridional ridges that constrain near-bottom flow. Viewed in conjunction with the hydrographic measurements, the ADCP data not only provide a complementary view of the circulation but also reveal structures not well repre-sented by the hydrography. Three examples of interesting structure observable in the ADCP data are shown. First, along the western end of I2 near Mombasa at about 4°S, a strong anticyclonic eddy is centered at 1000 m, with meridional velocity components exceeding 40 cm/s both to the south and to the north. The eddy coincides with an oxygen minimum and a salinity maximum indicative of Red Sea Water. Second, LADCP measurements from three cruises off the coast of South Africa during the first half of 1995 show a consistent vertical structure of the Agulhas Undercurrent: maximum equatorward velocities near 30 cm/s, low velocity shears, and equator-ward flowing intermediate waters with Red-Sea Water influence. The Agulhas Undercurrent would be poorly represented by a level of no motion geostrophic estimate: the level of no motion would be placed below the influence of Red Sea Water in order to send this water poleward (away from its source). Third, the I5W LADCP section across the Mozambique Basin shows the expected northward deep western boundary current along the Mozambique Plateau, but also shows a nearly barotropic recirculation occupying most of the basin. Simple integration of the LADCP transport below the sill depth to the north (3000 m) results in a net transport of 0.3 Sv.

Bruno Ferron
Massachusetts Institute of Technology

Toward an Indian Ocean WOCE Synthesis

The Indian Ocean general circulation is estimated by fitting the MIT Ocean General Circulation Model to the annual mean climatological hydrography and surface forcing, using the model and its computer-generated adjoint. Open boundary conditions are implemented to the west of the Indonesian Archipelago and near 30°S. The approach simultaneously optimizes the initial conditions of the hydrographic fields, surface fluxes, and the open boundary conditions (temperature, salinity, and horizontal velocities).

In a second part, we present the ongoing work directed at estimating the general circulation of the Indian Ocean in a dynamically consistent way by quantitative combination of a time-dependent general circulation model (GCM) with World Ocean Circulation Experiment (WOCE) hydrographic data, altimetry, XBTs, wind, heat and freshwater surface flux data.

Rana Fine
Rosenstiel School of Marine and Atmospheric Science, University of Miami

Circulation and Ventilation of the Indian Ocean at the Time of WOCE
Rana A. Fine, John L. Bullister, William M. Smethie, Jr., Mark J. Warner, Dong-Ha Min, and Ray F. Weiss

Chlorofluorocarbon (CFC) data from the United States participants in WOCE will be presented. The age information inherent in the CFCs will be used with hydrographic data to constrain water mass circulation and effective spreading rates of high latitude Southern Hemisphere and Marginal Seas waters into the tropical Indian Ocean. Bottom waters entering the Indian Ocean in the Mascarene Basin at 20°S transported above blank level CFCs in 1995. Throughout the Indian Ocean, there were measurable CFCs from the surface at least through the level of Antarctic Intermediate Water (AAIW). The equatorward spreading of the youngest AAIW can be seen in the elevated concentrations of CFCs in the western basins of the Indian Ocean. Elsewhere concentrations at that level are lower in waters from the Red Sea and Indonesian throughflow, and AAIW from the southwestern Pacific. Whereas, the Subantarctic Mode Waters (SAMWs) that lie above have highest CFC concentrations emanating from the southeastern part of the South Indian subtropical gyre, and they form a broad subsurface maximum in the central portions of the gyre. There is a strong front in SAMWs and the rest of the thermocline at the equatorward boundary of the subtropical gyre. The pathways for equatorward transport of thermocline waters out of the gyre appear to be via the western boundary current into the Arabian Sea on time scales of less than 20 years. However, they undergo substantial modification in the equatorial band.

Albert Fischer
MIT/Woods Hole Joint Program

Atmospheric Forcing and Upper Ocean Response in the Arabian Sea during 1994-95: Observations and Comparisons with Climatology and Models

Accurate, year-long time series of meteorology and upper ocean response were collected from an array of surface and subsurface moorings deployed in the north-central Arabian Sea from October 1994 to October 1995. Two major periods of mixed layer deepening, associated with the NE and SW monsoons, were observed. The NE monsoon mixed layer deepens primarily in response to convection, while during the SW monsoon wind-driven shear accomplishes the deepening. Bulk fluxes were calculated from the observed meteorology, and large differences were found between the observations and older climatologies, while the SOC climatology for 1980-95 was close to the buoy monthly means. Comparison of the observations with NCEP and ECMWF model surface products show that the models provide realistic surface winds, but fail to replicate other observed surface meteorology or to produce realistic heat fluxes. Estimates of the heat flux due to horizontal advection made from the moored array show that horizontal advection is a major component of the heat budget during the monsoon seasons. We seek to better understand the effect of high-frequency (diurnal-scale) buoyancy and wind forcing on the mixed layer and their contribution to mixing processes, and to better understand the basin-scale effects of high-frequency mixed layer dynamics through interaction with the interior flow.

Arnold Gordon
Lamont-Doherty Earth Observatory, Columbia University

Northeast Indian Ocean, North of the Throughflow: The Bay of Bengal
Arnold L. Gordon and Claudia Giulivi (LDEO)

As with all of the ocean basins, tropical heating of Indian Ocean water must be eventually vented to the atmosphere. In the Atlantic and Pacific Oceans this can be accomplished by advective heat transport to the higher cold northern and southern latitudes. The Indian Ocean does not have this luxury as Asia blocks an ocean route to the northern subpolar regions. Tropical heating not used to drive enhanced evaporation, fueling the Asian Monsoon, must somehow find its way into the Agulhas Current as part of the meridional circulation pattern. Exactly how the two great northern embayments, the Arabian Sea and Bay of Bengal "deal" with this is not clear. There is also their contrasting freshwater budgets: the Arabian Sea is a desert; the Bay of Bengal is reminiscent of an estuary. The required hydrological balance may be dealt with [through] zonal exchange within the equatorial currents.

The Indonesian throughflow crosses the Indian ocean near 12°S within the south equatorial current. The throughflow stream acts to isolate the thermocline and intermediate water masses of the monsoon regime to the north from those of the subtropical regime to the south. North of the throughflow, during the winter monsoon I-9n data show saline upper thermocline water from the Arabian Sea protruding eastward between 5°S and 5°N, embedded within a counter clock-wise circulation gyre (Hacker et al. 1998). North of this equatorial flow is the low salinity thermocline of the Bay of Bengal. The low salinity extends downward to the 12°C isotherm (near 300 m), cooler water than produced at the winter sea surface. It is suggested that downward flux of low salinity surface water is a product of the tidally active eastern Bay of Bengal. For waters cooler than 12°C the saline water mass source is from the Arabian Sea (with possible input from subtropical south Indian Ocean and Red Sea). The oxygen is low throughout the Bay of Bengal water column below the surface layer, however, in the southern Bay of Bengal and in the equatorial zone, centered at 11°C (380 db), there is a weak oxygen maximum, coupled with a salinity maximum. This feature may be a remnant of the SAMW from the south Indian Ocean, that spreads to the region north of the throughflow along the western boundary of the tropical Indian Ocean. The thermohaline and mass zonal exchange between the Arabian Sea and Bay of Bengal, and the meridional exchange across the Indonesian plume, is a fundamental climate question.

A curiosity is the Bay of Bengal benthic boundary layer (BBL). Strong near bottom gradients of nutrients and oxygen are found in a rather limited area in the western Bay of Bengal. The Arabian Sea also displays a nutrient rich BBL. Edmond et al. 1979 and Broecker et al. 1980 first identified this feature using GEOSECS stations. The BBL nutrients follow Redfield ratios; oxidation of bottom sediments derived from surface productivity is assumed. While the Arabian Sea has strong upwelling that supports high surface productivity, the Bay of Bengal upwelling is much weaker. High productivity may not be supported by Ekman induced upwelling, but rather by the enormous amount of dissolved and particulate material that enters the Bay of Bengal with the river water.

Makassar Strait Transport: Initial Estimate Based on Arlindo Results
Arnold L. Gordon and R. Dwi Susanto (LDEO)

Makassar Strait is the primary pathway of the Pacific to Indian Ocean transport called the Indones-ian throughflow. The transport was measured as part of the Indonesian USA Arlindo program, by two moorings deployed within the Labani Channel, a deep constriction near 3°S. Both moorings were operative from December 1996 to February 1998; the westernmost mooring operated until July 1998. The moorings were deployed during a weak La Niña phase. An El Niño condition began in March 1997, becoming extreme during 1997 summer and fall, relaxing in early 1998. The Makassar thermocline depth and transport reflect the phases of ENSO. Thermocline isotherms were deeper during La Niña, when the warm pool of the tropical Pacific, with its relatively high sea level and deep thermocline, is pressed up against the Pacific entrance to the Indonesian seas. Shallower thermocline occurs during El Niño when the warm pool shifts eastward in the Pacific, reducing its accessibility to the Indonesian seas. ENSO effects are also seen in velocity data. As El Niño takes hold, deeper instruments recorded reduced speeds, as the through-flow shoals with the thermocline. The 1997 average Makassar Strait throughflow transport is 9.3 Sv, assuming the flow in the upper 200 m equals the flow measured by the 200-m moored current meter. Other models for the surface flow range from 6.7 Sv (zero surface flow) to 11.3 Sv (thermocline shear is extrapolated to the sea surface). Comparison of the time series overlap months of December to February also reflect the ENSO effect: the transport during El Niño months of December 1997 to February 1998 average 5 Sv, while La Niña months of December 1996 to February 1997 average 12 Sv, a 2.5-fold difference. The Makassar transport determined from the Arlindo data are at the higher end of estimates based in Timor Sea and Indian ocean studies.

Catherine Goyet
Woods Hole Oceanographic Institution

Anthropogenic CO₂ Concentrations along WOCE I1
C. Goyet, C. Coatanoan, G. Eischeid (Woods Hole Oceanographic Institution)

As part of a cooperative effort of the Joint Global Ocean Flux Study (JGOFS) and of the World Ocean Circulation Experiment (WOCE) program, we have measured total CO₂ (TCO₂) and total alkalinity (TA) along WOCE cruise I1 in the North Indian Ocean. The measurements were performed on board R/V Knorr in September - October 1995. The primary purpose of this work is to understand the penetration of anthropogenic CO₂ along these ocean sections. Here we present a novel approach to the calculation of anthropogenic CO₂ in the ocean based upon the fundamentals of water-sources mixing.

The data show large spatial variations in surface seawater of both total CO₂ (up to 50 µmol/kg) and total alkalinity (up to 40 µmol/kg). The variations are mainly associated with physical processes characterized by water masses of different temperature and salinity. The contrasts between the sections across the Arabian Sea and the Bay of Bengal emphasize the large property differences between the two ocean basins. Multiparametric analyses on the data clearly show the relative contributions of different water sources in each of the ocean sections. The mixing coefficients calculated from the multiparametric analyses are further used to quantify anthropogenic CO₂ concentrations in each water source. The results indicate that the surface water-sources contain 47.8, 42.1, and 50.4 µmol/kg in the Gulf of Aden, the Arabian Sea and the Bay of Bengal respectively. In the surface waters there is slightly more anthropogenic CO₂ across the Bay of Bengal than across the Arabian Sea. In contrast, anthropogenic CO₂ has penetrated significantly deeper in the Gulf of Aden than in the Arabian Sea and in the Bay of Bengal.

Peter Hacker
University of Hawaii

Bay of Bengal Circulation: Observations and Model Results

Velocity and property observations were made along lines I8N and I9N as part of the WOCE Hydrographic Program expedition during February and March 1995, the time of the northeast monsoon. Upper ocean observations document vigorous circulation features including zonal currents, cross-equatorial flows, and recirculation eddies. These circulation features are respons-ible for the transport and stirring of the fresher Bay of Bengal water and the saltier western Indian Ocean water. The observations are qualitatively compared to the output of several high-resolution numerical models. The more realistic models may provide a useful temporal context of seasonal and annual evolution for the interpretation of the WOCE snapshot. In addition, the observations provide details of the spatial structure of temperature, salinity and velocity. These details probably need to be realistically simulated in upper ocean and coupled air-sea interaction models.

Gregory Johnson
NOAA / Pacific Marine Environmental Laboratory, Seattle WA

Flow of Bottom and Deep Water in the Amirante Passage and Mascarene Basin
Gregory Johnson, David Musgrave¹, Bruce Warren², Amy Ffield³, Donald B. Olson^4
1 University of Alaska - Fairbanks
² Woods Hole Oceanographic Institution
³ Lamont-Doherty Earth Observatory, Columbia University
⁴ Rosentiel School of Marine and Atmospheric Science, University of Miami

In the Indian Ocean, the Amirante Passage is the sill through which relatively cold, fresh, oxygen-rich, and nutrient-poor bottom water spreads northward into the Somali Basin from the Mascarene Basin. The passage is also a conduit through which relatively warm, salty, oxygen-poor, and nutrient-rich deep water spreads south. Previous estimates for northward transport of bottom water in the passage have been made from station pairs and sections without benefit of tracer measurements. Previous estimates of southward transport of deep water are scarce. Three hydrographic sections were made a across the passage in 1995 and 1996 as part of the World Ocean Circulation Experiment (WOCE). Two WOCE sections were also made perpendicular to the western boundary in the Mascarene Basin, just south of the passage. The geostrophic shear field is used with the salinity, dissolved oxygen, and silica distributions to select a range of zero-velocity surfaces (ZVSs) on potential isotherms for 1.0 to 1.1°C (hence a range of geostrophic transports) for which the flow direction is consistent with the tracer distributions. Objective mapping is used to obtain flux estimates below the deepest common level of station pairs. Estimates in the Mascarene Basin result in a bottom water volume transport from 2.5 to 3.8 Sv northward toward the passage below the ZVSs and a deep-water transport between the ZVSs and 2.5°C from 11.6 to 6.4 Sv southward. Estimates within the passage result in transport from 1.0 to 1.7 Sv northward for the bottom water and from 8.6 to 3.8 Sv southward for the deep water.

Yoshiteru Kitamura
Meteorological Research Institute, Tsukuba, Japan

Modeling the Indian Ocean Variability Using Wind Stress from JMA Operating Analysis

Interannual variabilty of the upper Indian Ocean from 1984 to 1995 is reproduced with an OGCM using JMA operational wind stress. The dominant two modes of the basin-wide vertically averaged temperature (VAT) over 300 m are coincident with those derived from an XBT line bewteen Java and Australia (Meyers, 1996). The 1st mode has a large correlation with the Pacific ENSO, while the 2nd mode has an exteremly strong peak in 1994. Disappearance of the Wyrtki jet seems to be accounted for by the development of the 2nd mode. The model fails to reproduce accurate sea surface heat balance. It requires about 28 W/m² (averaged over >15°S) additional cooling to avoid the SST increasing trend.

David Legler
Center for Ocean Atmospheric Prediction Studies, Florida State University

Variability of FSU Surface Fluxes

A variety of surface flux products have been produced and analyzed at FSU. Monthly mean surface pseudo-stress fields for the Indian ocean are available for 1971 to present. A more comprehensive set of monthly mean surface fields for 1961-1989 indicates significant interannual variability. New objective analysis and data adjusting techniques are being implemented to provide an improved suite of products for the similar time period. Daily gridded NSCAT wind products highlight intramonthly variability. Particularly interesting are the patterns in the southern hemisphere where in-situ data are sparse. The FSU WOCE DAC is still seeking surface meteorology data from several Indian ocean cruises and encourages PI's to make these data available at the Indian ocean workshop.

Mark Luther
Department of Marine Science, University of South Florida

Modeling the Circulation of the Upper Indian Ocean

We propose to continue analysis and application of our Indian Ocean circulation model in conjunction with data from the WOCE and ONR/JGOFS field programs. The model is a 3.5 layer reduced gravity model forced by observed atmospheric fields. The model domain covers the Indian Ocean basin north of 30°S at a resolution of 1/12 degree. The improved mixed layer formulation included in this model will enable a better understanding of the upper ocean processes responsible for the observed oceanic heat gain, subsequent water mass modification, and high primary production observed in the Arabian Sea during the southwest monsoon. The large volume of shipboard and moored observations recently taken during the ONR, JGOFS, and WOCE field programs provide an excellent opportunity to test, verify, and improve the mixed layer formulation. In turn, the model can be used to interpret the observations in a consistent dynamical framework to better understand the physics governing basin scale circulation and heat budget, mixed layer structure, and seasonal to interannual variability. James O'Brien and David Legler at the Florida State University are providing the atmospheric fields needed to force the model. We will continue to make model output available to the research community via the internet by anonymous ftp and the World Wide Web (http://ompl.marine.usf.edu).

Yukio Masumoto
Dept of Earth and Planetary Science, University of Tokyo

Forced Rossby Waves in the Southern Tropical Indian Ocean
Yukio Masumoto and Gary Meyers

Seasonal and interannual variation of the upper southern tropical Indian Ocean (STIO) is described by harmonic and EOF analysis of the depth of the 20 C isotherm (D20) derived from XBT data and from an OGCM. The harmonic analysis shows a band of large annual amplitude between 8°S and 20°S extending across the STIO in both the XBT and model data. The annual phase shows a steady westward propagation in both model and observations. It turns out that the Ekman pump-ing on a large scale over the open ocean and the westward propagation of long, nondispersive, baroclinic Rossby waves are responsible for generating much of the structure in the amplitude and the phase of the annual signal in the STIO. The EOF analysis of the interannual anomaly of D20 on two XBT lines and in the model results suggest an oscillation in the zonal tilt of the thermocline across the STIO. The generation of the interannual anomalies in the model are also due mainly to the wind stress curl integrated along the Rossby wave trajectories in the STIO.

Mary Cait McCarthy
Scripps Institution of Oceanography, University of California - San Diego

Potential Vorticity and Its Implications for Variability
Mary Cait McCarthy and Lynne D. Talley

Potential vorticity has been mapped on neutral density surfaces in the Indian Ocean using WOCE CTD data and historical bottle data. As a dynamical quantity, potential vorticity and its gradients are important to information propagation. As a tracer, potential vorticity helps identify water masses and indicates their circulations. Subantarctic Mode Water and Antarctic Intermediate Water both have strong potential vorticity signals which are relatively meridionally homogeneous and which coincide with depths influenced by the wind-driven gyre. At mid-depths the potential vorticity field is beta-dominated, while deep water shows structure which deviates both from beta and from topographic potential vorticity. The region of homogenized potential vorticity surrounded vertically by regions with much higher meridional gradients suggests that Rossby waves will travel faster there than predicted by standard theory. Estimates are given for the increased phase speeds.

Julie McClean
Naval Postgraduate School, Monterey CA

A Suite of High Resolution Global Ocean Models: Descriptions and Analyses in the Indian Ocean
Julie McClean, Mathew Maltrud, Robin Tokmakian, Wieslaw Maslowski, and Albert Semtner

Details of a suite of recently completed high resolution global ocean model runs are presented. Each of these simulations has been run as part of a complementary effort to develop very realistic high resolution models for climate studies. Essential to the ongoing development of these models are performance assessments that are achieved through comparisons with high quality data sets such as those collected during WOCE. Here each of the runs are described along with preliminary results focusing on the Indian Ocean.

The Los Alamos National Laboratory Parallel Ocean Program (POP) 1/6-degree model was restarted at the end of 1992 from an earlier integration (Maltrud et al., 1998) and was run through 1997 to obtain the most suitable output for data comparisons. The model is forced with ECMWF wind stresses and by Barnier monthly mean surface heat fluxes obtained from climatological ECMWF analyses. At the locations of the WOCE current meters and across key passages such as the Indonesian through flow, very high-frequency output was saved over the full depth of the water column. Model trajectories, originating at the start positions of WOCE floats, were initiated every 10 days and were then integrated forwards in time as part of the model run. Trajectories from the Indian Ocean will be shown.

The two other models: a fully global 1/3-degree version of POP on the displaced North Pole grid and the 1/4-degree Semtner and Chervin model known as the Parallel Ocean Climate Model (POCM), are forced with daily wind stresses, heat and salt fluxes for the period 1979-1997. The forcing consists of the ECMWF reanalysis products for 1979-1993 and ECMWF operational data for 1994-1997. The 1/3-degree POP includes a mixed layer (KPP) and the Sandwell and Smith (1997) topography. Once interpolated to the model grid, the topography was checked carefully to ensure reasonable flow through key passages. Preliminary results will include the mean and variability of mass and heat transports through the Indonesian seas and fluxes across across co-located WOCE sections. Mixed layer variability will be examined using EOFs and mixed layer depths will be compared with available data.

Stephen P. Murray
Coastal Studies Institute, Louisiana State University

Moored Observations of Mass and Salt Transport, Red Sea to Indian Ocean
S.P. Murray¹ and W. Johns^2
1 Louisiana State University
² University of Miami

Exchange flow between the Red Sea and Gulf of Aden-Indian Ocean through the Bab el Mandab Strait was measured continuously from June 1995-January 1997. ADCP, conventional current meters, and temperature-salinity chain moorings allow an unprecedented look at the magnitude and seasonal evolution of the inflow layer from the Gulf and the high salinity outflow layer from the Red Sea. Timing, structure, and evolution of the summer mid-depth intrusion of cold, low salinity water into the Red Sea from the Gulf is measured for the intrusion cycles of 1995 and 1996. We find the deep outflow is strong in June 1995. From July to September, deep outflow persists but is attenuated. The dominant summer feature, cold low salinity intermediate layer intrusion, persists for three months and carries cold nutrient-rich water to the Red Sea. Salinity transport computations allow estimates of basinwide evaporation rates. Winter regime begins in September, is fully developed by November, and continues to March 1996. Lower layer speeds are 0.8-1.0 m/sec; upper layer is 0.4-0.6 m/s. At maximum exchange in February, outflow transport reaches 0.7 Sv. Recent computations of evaporation rates and salt flux balances are consistent with layer transports made from velocity measurements only.

Kerstin Petuhov
Institut für Meereskunde, Universität Kiel, Germany

Distribution of Characteristic Hydrographic Parameters during the SW Monsoon 1997 along the Drift Trajectory
K. Petuhov and J. Waniek

The aim in the hydrographic investigation of the JGOFS - cruise Sonne 120 in the western Arabian Sea was to find a filament within the coastal upwelling of Oman and follow it with a drifter. We want to examine the structures and characteristics of the formed cold filaments during the period of the SW monsoon, when the main wind stress is concentrated into a jet (Findlater Jet) over the central Arabian Sea. Changes in wind stress and direction cause a high level of variability in the structures of the filaments.

Our poster gives an overview of the first results of the hydrographic data collected between June 12 and July 12, 1997. It illustrates the distributions of characteristic parameters like temperature, salinity, density, and fluorescence in some horizontal and vertical plots along and across the drift trajectory.

Later we plan to analyse the upper water masses during this time period. The T/S diagram of the CTD stations along the drifter shows obviously that the salty Arabian Sea Water (ASW) is already mixed with upwelled low salinity water masses. Correlation of the fluorescence signal with temperature and calculation of the mixed layer depth are helpful results for planktological processes. Analysis of the drifter velocity data show that the drifter moved at first to the NE with a mean velocity of 6 cm/s in the filament. The movement was accompanied by strong, fast changes in the temperature signal at scales smaller than 10-20 nm. These do not result from changes of the floating depth but are caused by the variability of the temperature field only.

Olaf Plaehn
Institut für Meereskunde, Universität Kiel, Germany

Ventilation and Circulation in the Arabian Sea
O.Plaehn¹ and M. Rhein^2
1 University of Kiel, Germany
² University of Rostock, Germany

A CFC data set from the Arabian Sea, collected during several cruises in April, June/July, and August/September 1995 as well as in January 1998, will be presented. Simulation of a mixed layer model including air-sea gas exchange processes was made to get a better knowledge of the ventilation of the mixed layer.

The correspondence between measurements and model results shows the strong influence of the meteorological forcing on the observed variations apart from strong advectiv currents. At the intermediate depth range CFC concentrations and salinities are simulated by a box-model to analyze the net transfer in the northwestern Arabian Sea. Transport calculations and estimates of the mean age were made for several intermediate water masses. An extremly large CFC-12 signal was found in the PGW, first observed in the Gulf of Oman three years ago, while in 1998 the signal was found near Socotra. This special feature was used to track and quantify the PGW in the northern Indian Ocean.

Joerg Reppin
Institut für Meereskunde, Universität Kiel, Germany

Variability of Currents and Tansports in the Equatorial Indian Ocean South of Sri Lanka
J. Reppin, F. Schott and J. Fischer (Institut für Meereskunde Kiel, Germany)
D. Quadfasel (Institut für Meereskunde Hamburg, Hamburg, Germany)

The zonal circulation south of Sri Lanka is an important link for the exchange of water between the Bay of Bengal and the Arabian Sea. The poster shows results from the WOCE ICM8 array along 80.5°E. Results from a first array of three moorings from January 1991 to March 1992 are used to investigate the Monsoon Current regime. Measurements from a second array of six current meter moorings which was deployed between 45°S and 5°N from July 1993 to September 1994 are used to investigate the annual cycle and interannual variability of the equatorial currents at this longitude.

Nils H. Rix
Institut für Meereskunde, Universität Kiel, Germany

Implications on Variability and Heat Transport from Regional Indian Ocean GCM Results

A series of sensitivity experiments conducted with a high resolution basinscale GCM of the Indian Ocean sector, with open boundary conditions at the eastern and southern boundary, suggest that the spatial and temporal distribution of mesoscale model variability, if compared to SSH variability as seen by an altimeter, is strongly dependent on wind forcing and model configuration details. Interannual as well as seasonal quasi-realistic mesoscale variability can only be adequately modeled with appropriate forcing.

Sensitivity studies with changes in boundary conditions also show that meridional overturning and heat transport are a function of these boundary conditions. Although a deep overturning circulation can be forced through the southern boundary, the necessary changes in boundary data render a strong deep overturning, that would contribute significantly to the annual mean meridional heat transport in the whole basin, unlikely.

Investigation of the dependence of the meridional heat transport on wind forcing indicates that the wind is primarily responsible for the meridional heat transport north of 10°S. The interannual heat transport variability of annual mean heat transport shows a range of variability that encompasses almost all available annual mean profiles of meridional heat transport that have so far been compiled from observational and model data by various methods and authors.

The main mechanism of annual mean heat transport in the model is an overturning circulation that is primarily wind driven and is mainly confined to the upper 1000 m of the water column.

Andreas Schiller
CSIRO, Australia

Modeling Interannual Variability in the Indian Ocean

A global ocean general circulation model (GFDL MOM2) is used to simulate interannual variability for the period January 1987 to March 1998. The model is coupled to the Kleeman and Power (1995) atmospheric boundary layer model to produce reliable surface heat fluxes. The model is driven by monthly mean net downward shortwave radiation from the NCEP reanalysis project (Kalnay et al., 1996) and by monthly mean FSU wind stresses (Legler et al., 1989). Poleward of the 30°latitude belt, the wind stresses have been blended into Hellerman and Rosenstein (1983) seasonal mean wind stresses. The OGCM is validated with observations based on expendable bathythermograph (XBT) data and satellite altimetry data (SSH).

Our ocean model is able to reproduce the major observed space- and timescales of interannual variability in the Indian Ocean. Away from the (unresolved) boundary currents, correlation coef-ficients along the XBT sections for simulated and observed anomalies of SSH (0.5) and D20 (0.6) are significantly higher than those for SST anomalies (0.3-0.4). This suggests that data from satellite altimetry, although valuable for complementary model validation, do not contain sufficient information to provide an accurate picture of near-surface ocean (thermo-)dynamics, in particular for SST. It emphasizes the need for improved flux products of wind stress and heat fluxes.

Friedrich Schott
Institut für Meereskunde, Universität Kiel, Germany

Monsoon Circulation of the Western Arabian Sea and Northern Somali Current
F. Schott, J. Fischer, L. Stramma, and J. Reppin

The monsoon response of the western Arabian Sea was one of the central objectives of the German WOCE Program "Tropical Oceans". The poster summarizes the field work based on data from the ICM7 array across the northern Somali Current and five ship surveys with R/V Meteor (April, June and August 1995) and R/V Sonne (August 1993 and January 1998). Special emphasis of the summer monsoon studies is on the development of the Great Whirl, the exchange of Somali Current water masses with the interior of the Arabian Sea, including the role of the Somali-Socotra passage, and the depth penetration of the monsoon response.

Janet Sprintall
Scripps Institution of Oceanography, University of California - San Diego

Some Recent Observations of the Indonesian Throughflow as It Enters the Indian Ocean
Janet Sprintall, Jackson Chong, Susan Wijffels, Nan Bray, Susan Hautala

The Indonesian throughflow is widely acknowledged as a highly significant component of the global ocean thermohaline circulation and heat budget, and therefore represents an important component of the earth's climate system. Recent interest has led to efforts in understanding the magnitude and temporal variability of ITF. However, most of these studies have been based on numerical models, as it has only been in recent years that the political process in Indonesia has made it possible to collect comprehensive oceanographic measurements in Indonesian waters.

A collaborative group of principal investigators from the United States and Australia have focused their measurement efforts on the "outflow straits", documenting the throughflow as it leaves the interior Indonesian seas and enters the Indian Ocean through the Indo-Australian Basin. The high quality data set is composed of a shallow pressure gauge network, underway ADCP and CTD surface measurements, a current meter mooring on the South Java coast, WOCE repeat hydrographic and XBT sections, and sea surface altimetry. This poster describes these data sets, and outlines initial efforts to:

1. determine the seasonal variability of the surface throughflow transport as it enters the Indian Ocean, using the WOCE repeat hydrographic sections.
2. partition the upper layer transport among the various Indonesian straits, spanned by the pressure gauge network
3. identify the sources of the two forms of intraseasonal (60-day) energy : discrete throughflow eddies, and intraseasonal changes in sealevel found along the Indonesian coastal waveguide, and
4. examine how the South Java Current alters the water mass structure in the Indo-Australian basin.

Hiroaki Ueda
Meteorological Research Institute, Tsukuba, Japan

Interactive System of Ocean and Atmosphere over the Indian Ocean in Terms of Seasonal Evolution of Summer Monsoon

It has been emphasized that large-scale differential heating between continent and the surrounding ocean is fundamental importance for establishment of the Asian summer monsoon. However, we have revealed that the ocean also plays a consequential role in seasonal evolution of the "oceanic monsoon" over the Indian Ocean and the western subtropical Pacific. First, the seasonal change process of the South Asian monsoon is investigated in the aspect of air-sea interaction. Second, we examined how interannual variation of the Indian summer monsoon is modified by the ocean. In the end, weekly-scale statistical relationships between the seasonal change of SST and surface wind speed are investigated.

P.N. Vinayachandran
Frontier Research System for Global Change, Tokyo

The Southwest Monsoon Current east of Sri Lanka
P. N. Vinayachandran,Yukio Masumoto, and Toshio Yamagata
Instititute for Global Change Research, Frontier Research System for Global Change, Tokyo
Japan Department of Earth and Planetary Physics, University of Tokyo, Japan

The general eastward flow in the north Indian Ocean during summer, which is called the Southwest Monsoon Current (SMC), flows eastward south of India, turns around Sri Lanka and enters the Bay of Bengal. The intrusion of the SMC into the Bay of Bengal is studied using the XBT observations along the shipping route between Sri Lanka and Malaca Strait, TOPEX/POSEIDON sea surface height anomalies, and an ocean general circulation model.

The intrusion appears first as a broad northward shallow (confined to the upper 200 m) flow in the central part of the Bay of Bengal during May. As the season advances it moves westward, intensifies and becomes narrow. The mean seasonal (May-September) transport of the SMC into the Bay of Bengal is about 10 Sv. The zonal variation of the geostrophic velocity across 6°N calculated using the XBT data compares well with that from TOPEX/POSEIDON altimeter data. However the SMC in the XBT data is faster (40 cm/s) than in the altimeter data or the numerical simulation (25 cm/s). Harmonic analysis of the depth of 20 C isotherm together with a simple forced Rossby wave model demonstrates that the SMC east of Sri Lanka is forced by both Ekman pumping in the Bay of Bengal and Rossby wave radiation associated with the spring Wyrtki jet in the equatorial Indian Ocean.

Roxanna Wajsowicz
University of Maryland, College Park

Interannual Variability in Monsoon Rainfall from an Oceanographic Perspective

In considering interannual variability in Indian monsoon rainfall, a fundamental question is where does the extra water come from (go to) in strong (weak) years? There are several possibilities:

1. the atmospheric circulation forces more (less) of the incident moisture to precipitate,
2. the moisture accumulation path is broader (narrower) and possibly encompasses new sources (sinks),
3. the path is the same, but there is increased (decreased) E-P along the route.

Possibilities (ii) and (iii) have been investigated using the reanalysed COADS, da Silva et al. (1994) and simple models. The paths are very similar to the climatological one, except at the southern end, where in strong (weak) years, the spiral out from the Mascarene High was broader (narrower). There were significant large-scale positive (negative) evaporation rate anomalies in strong (weak) years over the Southern Indian Ocean, but nothing of substance over the Arabian Sea. Further investigation of the source of these evaporative anomalies gives that their spatial patterns strongly resembled those of the corresponding qs-qa anomalies rather than the windspeed anomalies. Also, it is qs more so than qa, which provides the anomaly. The seasonal cycle heat budget is investigated with the help of a simple model to further understand how these anomalies arise.

Linda Waterman
CSIRO, Australia

Indian Ocean Circulation in a Regional Indian Ocean and a Global Ocean Circulation Model

The properties of a regional Indian Ocean general circulation model and the Indian Ocean sector of a global general circulation model will be validated. This poster will focus on some aspects of vertical mixing in the respective models, and their impact on the large scale water masses. Aspects of a hybrid vertical mixing scheme will be highlighted, and will include a general discussion of physical processes yet to be included in such models.

Yuzhu You
Institut für Meereskunde, Universität Kiel, Germany

Epineutral and Dianeutral Circulation of the Indian Ocean

This poster assembles the updated results from several studies carried by this author in the main thermocline, intermediate and deep waters and offers a pre-WOCE condition of the circulation in the Indian Ocean. It is hoped that these results can be examined and compared with those from WOCE data analysis. The epineutral and dianeutral circulation of the Indian Ocean is derived from water mass mixing model by using temperature, salinity, oxygen and nutrient data, and from water mass transformation and dianeutral velocity equations. It is based on the water mass spreading paths, mixing patterns, flow streamlines from acceleration potential mapped on neutral surface, and the dianeutral velocity calculation. The dianeutral velocity itself is contributed by several processes, cabbeling, thermobaricity, turbulent mixing and double diffusion. For the deep water, the epineutral and dianeutral circulation gives a mechanism for the North Atlantic Deep Water replacement, i.e. the transformation of Circumpolar Deep Water from the Southern Ocean to a upward flow in the northern Indian Ocean. All these results are concluded to several schematics of seasonal thermocline, intermediate and deep water circulation.