IP2A

IP2a

Development of Stratification and Gravity Currents in an Inverse Estuary (Persian Gulf)

A.A. Bidokhti

Institute of Geophysics, Tehran University, Iran

The development of stratification and gravity currents, and their breakdown due to tidal and wind forcing, in the Persian Gulf is described. Excessive evaporation from such a semi-enclosed water body and onset of cold season produces stratification. From temperature and salinity observations, it is shown that at the start of cold period, strong gravity current is channelled parallel to the Gulf axis, through the straight of Hormoz, with speeds of up to 0.4 m/s. After an initial inertial stage, a quasi-geostrophic frictional cyclonic circulation develops in the Gulf which can be important for the transport of marine pollution and salt.

Critical conditions for the weakening of the gravity currents by tidal and wind stirrings (based on Richardson number) are examined. They show that the effect of wind is small compared to the effects of tidal mixing. Similar processes have been observed in the Spencer Gulf of Australia (Lennon 1987).

The life-time of the gravity current is about 5 months, and it has an average thickness of 18 m and it travels and distance of about 150 km from the straight of Hormoz. When entering the open sea it casades down the continental shelf and produces a baroclinic eddy in the entrance of the Gulf.

IP2b

Subsurface Eddies in the Andaman Sea during October 1996

Y. Sadhuram, P.V. Narvekar, V.V.S.S. Sarma, P.N.M. Sastry and M.V. Subrahmanyam

National Institute of Oceanography, Regional Centre, Visakhapatnam, India
National Institute of Oceanography, Dona Paula, Goa, India

CTD data in the Andaman Sea is very sparse. We have collected CTD data at 30 stations in the Andaman Sea from 24th October to 7th November 1996 during the cruise 118 of ORV Sagar Kanya. During a CTD operation at 10ƒ30’N, 93ƒ15’E on 31 October, we noticed abnormal variations in temperature and salinity in the layer 90-120 m deep. CTD profile during the downcast (at 06:30 hrs IST; cast-I) indicated a temperature of 22.86ƒC and salinity 34.25% at 100 m while during the upcast (at 08:30 hrs IST; cast-II), extraordinary decrease in temperature (18.69ƒC) and increase in salinity (34.60%) were noticed at that depth. Between the cast-I and cast-II the isopycnal depth (s q =24) varied from 120 m to 90 m and sound velocity reduced by 10 m/s. An unexpected minimum in dissolved oxygen (34 µmole/l) and a peak in silicate (36 µmole/l) were also noticed at this location. CTD profiles taken at 0930 hrs IST (cast-III), however, showed a temperature of 20.22ƒC and salinity of 34.55% at 100 m. Unfortunately this phenomenon came to our notice while processing the data after the vessel had moved away from the location. We therefore re-occupied the location on 5th November 1996 and conducted time series CTD observations. We did not notice any abnormal variation in the temperature of salinity values this time. The values averaged over 20 hrs observational period (19.9ƒC and 34.54% respectively) were very close to those observed in cast-III on 31st October. Spatial distributions on 100 m depth contours of temperature, salinity and s q , plotted after taking the ship’s drift into the account, suggest a possibility of warm core and cold core eddies of smaller dimensions at 10ƒ30’N, 93ƒ15’E. Another warm core eddy centred at 13ƒ20’N, 93ƒ09’E hugging the coast is also noticed. Further analysis of the data is under progress, which will throw more light on the origin and mechanisms of these eddies.

IP2C

Ventilation and Aging of Water Masses in the Southern Indian Ocean Derived from Hydrographic, CFM and Oxygen Data

Johannes Karstensen and Matthias Tomczak

Institut fuer Meereskunde an der Universitaet Hamburg, Germany
Flinders Institute for Atmospheric and Marine Sciences, Flinders University, Adelaide, Australia

We study the spreading and mixing of Indian Central Water (ICW) and Subantarctic Mode Water (SAMW) in the south eastern Indian Ocean. ICW and SAMW are distinct water masses with different formation mechanisms but similar temperature/salinity characteristics. In the eastern south Indian Ocean, particularly south of Australia, the formation of SAMW may be the major ventilation mechanism for the lower thermocline.

We use a new method based on a combination of Optimum Multi Parameter analysis and CFM/oxygen mixing analysis to determine the age of the two water masses. This method allows the determination of individual water mass ages in regions of mixing and therefore enables us to follow water mass movement in greater detail than other methods, which give only the combined pseudo-age of the water mass mixture. We define the age of a water mass as the time a water parcel needs to travel from its source region, where it receives its individual tracer characteristics, to the point of observation. This age indicates pathways of water masses, which differ from circulation pathways because the age is determined by advective and diffusive processes.

We apply the method to section I5 of the World Ocean Circulation Experiment (WOCE) where ICW and SAMW meet and mix. We use regions of nearly pure ICW and SAMW to determine the biogeochemical parameters required by the method, ie the oxygen consumption rate and the Redfield ratios. The results confirm values given by other authors.

In the eastern part of the south Indian Ocean SAMW dominates the oceanic thermocline; it is found to be about 5 years old before it begins to mix with ICW circulating in the subtropical gyre. While most SAMW joins the equatorward gyre movement of the south eastern Indian Ocean, some of it propagates westward as a result of turbulent diffusive mixing, reaching 55ƒE after 15 - 20 years. Pure ICW is present only in the lower thermocline of the region 40ƒ - 55ƒE and is 3 - 5 years old before beginning to mix with SAMW. An increase of ICW age with depth is in good agreement with its formation through subduction. It takes ICW some 25 - 30 years to reach 110ƒE.

IP2d

Acoustic Thermometry in the Indian Ocean

Andrew M. Forbes and John D. Penrose

CSIRO Division of Marine Research, Hobart, Australia
Centre for Marine Science and Technology, Curtin University, Perth, Australia

For the past seven years the Australian CSIRO Division of Marine Research, together with collaborators at Curtin University in Western Australia have participated in the development of long-range acoustic thermometry. The dependence of sound speed on water temperature, one factor involved in the existence of the deep sound channel, has motivated a program, the Acoustic Thermometry of Ocean Climate (ATOC), to use long range sound travel times to measure ocean temperature on basin scales. A prototype network was established in the Pacific Ocean in late 1995, which has now been routinely gathering travel time data (and thereby, path averaged temperature of the ocean's interior) for eighteen months.

The Indian Ocean has, in basin scale terms, a unique suite of thermodynamic and climate characteristics, in part because it is alone amongst the major ocean basins to be bounded to the north at tropical latitudes. The influence of the Indian Ocean on Australia's climate is acknowledged but only partially understood. Variability of the continental climate is likely to be linked closely to variability of the ocean climate, but the current state of Indian Ocean monitoring is poor compared with atmospheric monitoring by terrestrial weather stations. New observations are needed to improve our understanding of the physical processes at work in the Indian Ocean that may impact our climate, and to provide data that may be used to constrain and improve the reliability and predictive capability of numerical models of Australian climate variability.

Acoustic thermometry may offer an effective new method of monitoring the temperature structure and heat content of the Indian Ocean. At a March '96 SCOR symposium on Acoustic Monitoring of the Global Ocean, held at the CSIRO Division of Marine Research in Hobart, participants from a number of nations expressed the desire to instrument the Indian Ocean as the next phase of acoustic thermometry, to be called the Indian Ocean Climate Initiative (IOCI).

This paper outlines the background to ATOC, illustrates the success of the technique with the first year's results from the Pacific and describes key features of the planning for the IOCI program.

ip2e

helium-3 distribution across the pacific throughflow, eastern indian ocean

P. Jean-Baptiste, M. Fieux and A.G. Ilahude

CEA-Saclay, DSM/LMCE, Gif-sur-Yvette, FRANCE
LODYC, CNRS/ORSTOM, University of Paris VI, Paris, FRANCE
LIPI, J1. Pasir Putih 1, Jakarta, INDONESIA

Water exchange between the Pacific and the Indian oceans via the Indonesian archipelago is of particular importance because of its involvement in global thermohaline circulation. Geostrophic calculations show that mean transport is mainly concentrated in the first five hundred meters of the water column (Fieux et al., 1996a,b). However, long-term current measurements in the Timor Strait (Molcard et al., 1996) show that velocities increase again below 500 m, suggesting that deep water exchanges could also be significant. We investigate the exchange of intermediate and deep water between the two oceans by studying the helium isotopes distribution in the Eastern Indian Ocean. ³He is enriched 7 and 10 fold with respect to the atmosphere in submarine hydrothermal emanations. It is injected into deep ocean waters where it offers a unique tracer for deep circulation in a depth range in which hydrographic indicators generally show near uniform distributions. The Pacific basin is by far the most enriched in ³He, mainly released in great quantity on the East Pacific Rise but also in the back-arc spreading centres of the Western Pacific. Similarly, ³He excesses are observed in the Indian Ocean, mainly in the western part of the basin, over the Indian ridge system from the Gulf of Aden to Rodriguez Triple Junction. However the magnitude of the ³He anomaly in the Indian Ocean is only nearly half of that in the Pacific. This interocean ³He gradient therefore provides a tool for tracing the exchange of intermediate and deep water between the two oceans.

The helium isotope data were obtained as part of JADE (Java Australia Dynamic Experiment), a cooperative project between Indonesian and French laboratories designed to study the Pacific-Indian throughflow during both Southeastern and Northwestern monsoons (in August and February respectively). This joint project started in August 1989 with the JADE 89 cruise, going from the coast of Australia to Bali, followed by a series of stations on a Bali-Timor section, at the outflow of the Indonesian Seas, 12 miles of a Lombok-Sumba-Dana-Roti line. The same route was repeated almost identically in February-March 1992, during the NW monsoon, to compare data in the two opposite monsoonal regimes. In the north, it was possible this time to carry out hydrographic work much closer to the continental shelf and on the main sills, particularly in the Timor Channel (between Roti and the Australia shelf), east and west of Savu Island and in the Sumba Strait (between Sumba and Flores).

The study of the ³He distribution shows that the South Banda deep waters, between 1000 and 1400 m, leave the Banda basin over the sills in the straits of Sumba, Savu, Roti and Timor, and enter the equatorial Indian Ocean. The permanence of the ³He in both opposing monsoonal regimes, and the only weak dilution of the ³He signal between the straits and the Australia-Bali section hundreds of kilometres apart, suggest that this deep outflow is not negligible. In the Timor Sea, ³He/⁴He measurements confirm that the bottom flow through the Timor Trench below 1400 m originates in the Indian Ocean.

Molcard et al., J.Geophys.Res., 101, 12411-12420 (1996)
Fieux et al., J.Geophys.Res., 101, 12421-12432 (1996a)
Fieux et al., J.Geophys.Res., 101, 12433-12454 (1996b)

ip2f

intermediate water CIRCULATION, ventilation, mixing and transformation in the indian ocean

Yuzhu You

Laboratoire d’OcÈanographie Physique, MusÈum National d’Histoire Naturelle, Paris, FRANCE
(Currently at: Institut f¸r Meereskunde, Kiel, GERMANY)

Pre-WOCE hydrographic data including temperature, salinity, dissolved oxygen and nutrients were used in a water mass mixing model of three water masses: Antarctic Intermediate Water (AAIW), Indonesian Intermediate Water (IIW) and Red Sea Intermediate Water (RSIW) that includes the influence of the Persian Gulf Intermediate Water (PGIW). Water-mass mixing ratios were plotted on six closely spaced neutral density surfaces and several cross sections. The intermediate water circulation and ventilation of the Indian Ocean could then be inferred from those patterns of quantitative water-mass mixing proportions. Equatorward AAIW is advected with the subtropical gyre and transits to the north through the western boundary region. The inflow of IIW into the eastern Indian Ocean forms a front as a result of mixing between AAIW and IIW, blocking the equatorward flow of AAIW into the eastern Indian Ocean. There is evidence that about 10-20% of AAIW spreads to the equatorial Indian Ocean through the western boundary but not beyond 5ƒN. IIW shows a westward tongue zonally crossing the Indian Ocean towards Africa from where it spreads to the south joining the Agulhas eddy. In the eastern Indian Ocean, IIW largely contributes to the Bay of Bengal. The path of RSIW southward is clearly shown in the western boundary. It also largely contributes to the Bay of Bengal but about 10% less than that contributed by IIW.

Dianeutral mixing and water mass transformation are examined on the neutral surfaces. south of the Antarctic frontal zone, AAIW is clearly indicated by a diffusive tongue in meridional Turner Angle sections. As it moves equatorward, transformed AAIW is characterised by a tongue of double stable water extending great distances towards the equator. An area-mean net dianeutral upwelling velocity of 0.11x10^-7 m s^-1found north of 32ƒS across the lowermost neutral surface, meant a net upward volume transport of 0.6 Sv. This weak but net upwelling transport roughly corresponds to a net downwelling transport of 0.5 Sv across the same neutral surface south of 45ƒS. However, much stronger upward dianeutral transport of 1.4 Sv is found across the uppermost neutral surface north of 32ƒS, while south of 45ƒS a net downwelling transport across the uppermost neutral surface shows as large as 5.4 Sv. That the downwelling transport across the uppermost surface is 10 times larger than across the lowermost surface south of 45ƒS has a strong hint for the equatorward transport of AAIW between the uppermost and lowermost neutral surfaces.

ip2g

seasonal variations of thermocline CIRCULATION and VENTILATION in the indian ocean

Yuzhu You

Laboratoire d’OcÈanographie Physique, MusÈum National d’Histoire Naturelle, Paris, FRANCE
(Currently at: Institut f¸r Meereskunde, Kiel, GERMANY)

Two seasonal hydrographic data sets, including temperature, salinity, dissolved oxygen and nutrients, are used in a mixing model which combines cluster analysis with optimum multiparameter (OMP) analysis to determine the spreading and mixing of the thermocline waters in the Indian Ocean. The mixing model comprises a system of four major source water masses, which were identified in the thermocline through cluster analysis. They are, Indian Central Water (ICW), North Indian Central Water (NICW) interpreted as aged ICW, Australasian Mediterranean Water (AAMW) and Red Sea Water (RSW)/Persian Gulf Water (PGW). The mixing ratios of these water masses are quantified and mapped on (a) four isopycnal surfaces which span the thermocline from 150 to 600 m in the northern Indian Ocean, (b) two meridional sections along 60ƒE and 90ƒE, and (c) two zonal sections along 10ƒS and 6ƒN.

The mixing ratios and pathways of the thermocline water masses show large seasonal variations, particularly in the upper 400-500 m of the thermocline. the most prominent signal of seasonal variation occurs in the Somali Current, the western boundary current, which appears only during the Summer Monsoon. The northward spreading of ICW into the equatorial and northern Indian Ocean is by way of the Somali Current centred at 300-400 m on s_q=26.7 isopycnal surface during the SW Summer Monsoon and of the Equatorial Counter Current during the NE Winter Monsoon. More ICW carried into the northern Indian Ocean during the SW Summer Monsoon is seen clearly in the zonal section along 6ƒN. NICW spreads southward through the western Indian Ocean and is stronger during the NE Winter Monsoon. AAMW appears in both seasons but is slightly stronger during the summer in the upper thermocline. The westward flow of AAMW is by way of the South Equatorial Current and slightly bends to the north on s_q=26.7 isopycnal surface during the SW Summer Monsoon, indicative of its contribution to the western boundary current. Outflow of RSW/PGW seems effectively blocked by the continuation of strong northward jet of the Somali current along the western Arabian Sea during the summer, giving a rather small contribution of only up to 20% in the Arabian Sea.

A schematic summer and winter thermocline circulation emerges from this study. Both hydrography and water-mass mixing ratios suggest that the contribution of the water from the South Indian Ocean and from the Indo-Pacific throughflow controls the circulation and ventilation in the western boundary region during summer. However, during winter, the water is carried into the eastern boundary by the Equatorial Counter Current and leaks into the eastern Bay of Bengal, from where the water is advected into the northwestern Indian Ocean by the north Equatorial Current. The so called East Madagascar Current as a southward flow occurs only during the summer as suggested by both hydrography and water-mass mixing patterns from this paper. During the winter (austral summer) the current seems reversed to a northward flow along the east of Madagascar, somewhat symmetrical to the Somali Current in the north.

ip2h

The Pacific Influence on the Circulation of Intermediate Waters of the Indian Ocean

Young-Hyang Park and Lucien Gamberoni

Laboratoire d'OcÈanographie Physique, MusÈum National d'Histoire Naturelle, Paris, FRANCE

Analysis of the climatological Levitus data indicates the three source water masses at intermediate depths (around 1000 m) in the Indian Ocean: the Antarctic Intermediate Water (AAIW), the Red Sea Water (RSW), and the Banda Intermediate Water (BIW) of Pacific origin. The data are projected on the neutral surface of AAIW core and the mixing ratios between these water masses are quantified by an optimal multiparameter analysis. The most original result is the unexpected role of the BIW in mixing and modification of intermediate waters in the whole Indian basin. Its major contribution is in the equatorial band where its mixture with the RSW, with an almost equal mixing ratio, forms the so-called "Equatorial Water" which finally fills the entire Bengal Bay. The freshest and most oxygenated AAIW enters into the subtropical gyre of the south Indian Ocean through the east of Kerguelen, in a region centered at 90ƒE. It circulates anticyclonically within the gyre and its northernmost influence is limited at about 10ƒS just north of Madagascar. The Mozambique Channel is the "melting pot" of the three water masses; the mixing ratio is 30 to 60% for the AAIW, 30 to 40% for the BIW, and 20 to 30% for the RSW, with the AAIW influence increasing rapidly to the southern limit of the channel. This mixture extends further southward into the Agulhas Current region along the south-east coast of South Africa. The frequently reported interleaving of saltier water with less saline, ambient AAIW of the region is not uniquely due to the RSW influence but also due to the BIW influence, with nearly equal importance of 10 to 20%. These new findings highlight the unforeseen influence of the Pacific, via the BIW, on the circulation of intermediate waters throughout the Indian basin.

IP2i

UPWELLING/DOWNWELLING OVER THE ARABIAN SEA DURING SOUTH WEST MONSOON SEASON

Ram P. Sree, M.M. Ali, R. Sharma and K.G. Reddy

Space Applications Centre, Ahmedabad, INDIA

Using ERS-1 Scatterometer measurements of NRCS values and ECMWF analysed wind fields, upwelling/downwelling phenomenon over the Arabian Sea has been computed. The wind field obtained from ECMWF data well matches with that of the wind speeds derived from the Scatterometer, in particular over the Somali region. When the radar backscatter signatures are low, wind speeds are also low.

The complete reversal of the wind system with the change of the monsoon presents a unique opportunity to study the occurrence of seasonal changes of welling phenomenon associated with wind regimes. The monthly mean welling for the South West monsoon is calculated during June to September. During the South West monsoon, the south equatorial current branches near the African Continent and its northern branches forms the strong Somali current. The Somali current continues to flow northward and is very strong during this period. In the open waters the drift is easterly. The coastal circulation of the Arabian Sea remain clockwise and is strengthened.

Offshore welling computations reveal a strong upwelling regime near the Somali coast from June to September which is mainly due to the Somali Jet prevailing during the monsoon season. The strong upwelling present near the Somali coast in the monsoon months reduced after the withdrawal of the southwest monsoon and when the winds reverse the direction, downwelling is observed in this region. In the other parts of the region both upwelling and downwelling are observed depending on the wind velocities.

IP2j

On the Subsurface Salinity Maximum in the Indian Ocean

V.A. Sosnin

Pacific Oceanological Institute, RUSSIA

As a rule, subsurface salinity maximum (SSMax) in the Indian ocean is always associated with a horizontal advection of water mass properties. However, an analysis of the seasonal variability of SSMax characteristics allows us to propose an alternative point of view on its origin. Consider SSMax not as outside water mass but as temporal phenomenon which depends on local processes of atmosphere-ocean interaction.

There are two regions in the ocean where SSMax occurs in different ways according to the sign of fresh water balance (FWB). In the regions with negative FWB the salinity maximum appears and disappears during the year and in this case it is a sequence of seasonal salinity variability on the sea surface. Properties of the SSMax (the depth and salinity value) depend on local values of FWB.

In the regions with positive FWB the SSMax is a permanent feature but its properties depend on the values of FWB too. Our point of view is that climatic displacement of the border of the regions with different FWB determines the temporal scales of appearance and disappearance of this phenomenon in different areas of the ocean.

IP2L

Annual Cycle of the depth of the 20ƒC isotherm in the southern tropical Indian Ocean

Yukio Masumoto and Gary Meyers

Graduate School of Science, University of Tokyo, JAPAN
CSIRO Division of Marine Research, Hobart, AUSTRALIA

Seasonal variation of the upper southern tropical Indian Ocean (STIO) is described by an annual harmonic of the depth of the 20ƒC isotherm derived from XBT data and from an ocean general circulation model. Both of the results show a band of large annual amplitude between 8ƒS and 20ƒS extending across the STIO. The observed maximum amplitude appears at 11ƒS, 90ƒE with the secondary maximum at 15ƒS, 55ƒE. The maximum amplitude simulated in the model is located at 11ƒS, 65ƒE. The annual phase shows a steady westward propagation in both results.

The generation of the annual wave is discussed in terms of Ekman pumping and the westward propagation of long, nondispersive, baroclinic Rossby waves. The Ekman pumping on a large scale over the open ocean strongly modifies waves radiating from the eastern boundary and generates much of the structure in the amplitude and the phase of the annual signal in the STIO.

IP2m

Studying the monsoon circulation in the Indian Ocean using altimeter data

Vibeke E. Jensen, Paul Samuel and Ola M. Johannessen

Nansen Environmental Remote Sensing Center, Bergen, Norway

The oceanic variability in response to the monsoon system in the northern Indian Ocean has been studied using Geosat, ERS-1 and TOPEX/POSEIDON altimetry data. We focus on the Somali Current area and the western part of the Bay of Bengal, as the most prominent variations occur in this area, and because most in situ observations are from these areas.

Mesoscale eddies with a diameter of at least 500-600 km and maximum amplitudes of 42 cm are observed during the monsoon periods. The propagation and life time of eddies in the study area are in good agreement with those reported in previous investigations using in situ data and numerical models.

Rossby waves are identified in the Indian Ocean. At equator they are observed to have a westward propagating speed of 28 km/d, and away from equator the speed decrease.

IP2n

INTRA-ANNUAL VARIABILITY OF THE INDIAN OCEAN CIRCULATION

Benny N. Peter and Keisuke Mizuno

Cochin University of Science and Technology, Cochin, India
National Reseasrch Institute of Far Seas Fisheries, Shimizu, Japan

The correlation between long term mean heat content and dynamic heights were determined from the temperature profiles of the Indian Ocean. The dynamic heights and the heat content are significantly correlated in the upper 400m of the Indian Ocean, except in the Indonesian throughflow region and the northern part of the Arabian Sea.

The dynamic topography obtained by this method is clearly depicting the seasonal circulation patterns of the Indian Ocean. The South Equatorial Current and the South Indian Ocean Current are more well developed in the annual signal. The North Equatorial Current is comparatively weak. The bimonthly distribution displays seasonal changes of the currents. Strong Indonesian throughflow is observed in August. The seasonal changes in the flow patterns of the Leewin Current off Western Australia shows in the bimonthly maps.

The time-longitude sections of dynamic height anomaly along different latitudes enable us to bring out the annual and semi annual signals of circulation. The semi-annual signal of the Kelvin wave is clearly identified at the equator. Away from the equator, the westward propagating Rossby wave signal is obtained and is very strong along 4ƒN and S. Along 12ƒS an annual signal of westward propagating Rossby wave generating from the Indonesian throughflow region is distinguished. The phase speed of the Rossby wave obtained at 12ƒS is less than 0.2m/s whereas 0.4m/s speed is found at 4ƒN and S. The equatorial Kelvin wave is propagating at a greater speed of about 1m/s.

IP2P

The Indonesian Throughflow and its impact on the upper Indian Ocean

A. Schiller, J.S. Godfrey, P. McIntosh and G. Meyers

CSIRO Division of Marine Research, Hobart, Australia

The Indonesian Throughflow allows warm Pacific water to flow into the Indian Ocean. Strong tidal flow in the Indonesian region causes enhanced mixing of deep and surface waters, modifying the water masses on their way to the Indian Ocean. As a consequence, the modified SST and energetic heat exchanges influence the extent of atmospheric convection and monsoon rain in the eastern Indian Ocean.

We use a global ocean general circulation model (Pacanowski, 1995) with a tropically-enhanced grid resolution and a level formulation of the Chen et al. (1994) mixing scheme. The model is forced with seasonal and interannual boundary conditions using the monthly mean Florida State University "pseudostresses" for 1985-1990 (Stricherz et al., 1992), blended poleward of 30ƒ into Hellerman and Rosenstein (1983) seasonal mean wind stresses. We use an explicit heat flux formula (similar to the one used by Seager et al., 1988) together with a damping term to compensate for heat flux errors (Godfrey and Schiller, 1997). Sea surface salinity is restored to observations. Incorporating a parameterization of tidal mixing in the Indonesian region leads to improved SST and heat flux patterns there.

The simulated tidal mixing increases the heat flux into the ocean in this region by about 50 watts/m². The water mass structure in the upper Indian Ocean and the way in which it is modified by the Indonesian Throughflow and tidal mixing is described. Improvements in mixed layer depth due to altering the solar penetration formulation and factors controlling SST in the Indian Ocean are also discussed.

IP2q

On the role of seasonal versus annual mean winds in setting the Indian Ocean heat flux pattern

Linda J. Waterman and J. Stuart Godfrey

CSIRO Division of Atmospheric Research, Aspendale, AUSTRALIA
CSIRO Division of Marine Research, Hobart, AUSTRALIA

Earlier authors have found that the annual mean pattern of heat flux into the Indian Ocean is quite well reproduced when the model is driven by the annual mean winds and relaxed to observed annual mean SST; the (much larger) seasonal cycle of winds is omitted.

Here we repeat this experiment in a model of the Indian Ocean, and also report results from the converse run, in which the seasonal cycle of winds is retained and the model is relaxed towards observed mean seasonal SST's, but the annual mean winds are omitted. Comparison of the three runs - with full winds, with annual mean winds only, and seasonal mean winds only - allow us to infer the relative roles of the annual mean and seasonal mean winds in maintaining the long-term mean heat flux into the Indian Ocean.

ip2r

Circulation in the Arabian Sea using
carbon-14

B.L.K. Somayajulu, R. Bhushan, S. Krishnaswami and P.V. Narvekar

Physical Research Laboratory, Ahmedabad, INDIA
National Institute of Oceanography, Dona Paula, Goa, INDIA

Extensive measurements of radiocarbon in the water column of the central and eastern Arabian sea and overhead air have been made to characterise the water mixing parameters and air-sea exchange rate of CO₂. During 1993-95, nine stations were occupied between 6-22ƒN and 58-74ƒE. The radiocarbon measurements were made following conventional beta counting technique by extracting CO₂ from 100l seawater, converting it to benzene and assaying in a low background liquid scintillation counter for ¹⁴C activity. The sampling depths were decided by a CTD cast so as to cover the principal water masses of the region. In addition to ¹⁴C, other parameters viz. temperature, salinity, oxygen and SCO₂ were also measured onboard the ship. Among the stations sampled, there were the GEOSECS stations 416, 417 and 418 which were reoccupied after a gap of 16-17 years with a view to determine the variations in both the natural and bomb components of ¹⁴C and to use them to derive mixing parameters.

The D¹⁴C of the surface (£5m) waters ranged from 31±5‰ to 63±5‰ and showed no particular geographical variation/trend. At three close-by stations (12±0.5ƒN; 67±1ƒE) where 2-3 samples were measured from the surface 100m, the D¹⁴C values were uniform at 51±5‰ in one profile whereas in the other two there was a decrease of 25 to 48‰ between 5m and 75m. The D¹⁴C values of the mixed layer were higher during GEOSECS than of the corresponding reoccupied stations by ~20%. The bomb ¹⁴C inventory during the reoccupied GEOSECS showed an increase by few percent. Deeper penetration of the bomb-¹⁴C activity by vertical mixing over the past 17 years i.e. since GEOSECS appears to be the logical explanation for these changes in D¹⁴C and its inventory.

For the deep water data (between 900m and 3000m depths), the vertical advection velocities (w) with corresponding diffusivities (K) have been derived using a one-dimensional model. The values of w and K are a few m/yr and a few cm²/s respectively. An account of these measurements will be presented.

IP2s

A Comparison Between Somali and Arabian Upwellings Based on Physical and Biological Indicators

Oldemar Carvalho Junior and Matthias Tomczak

The Flinders University of South Australia,
School of Earth Science, AUSTRALIA

This paper examines the two upwelling regions located at the Somali and Arabian coast of the Indian Ocean. The analysis is based on physical and biological indicators such as temperature, salinity, net-down-fresh-water-flow, pseudostress wind and chlorophyll. Time diagrams are used to describe the evolution of temperature, salinity and chlorophyll through the year.

The upwelling at the Somali coast is related to the formation of two gyres. Previous analysis of temperature in the area speculated that the cold upwelled water from the Somali current does not penetrate northward due to the presence of the warm surface water from the Gulf of Aden. When comparing the two upwelling regions the one located at the Arabian coast presents higher salinities than the one at the Somali coast.

A multiple regression statistical analysis is performed between temperature and salinity variables and pseudostress wind, net-down-fresh-water-flow and Ekman pumping parameters in order to establish the relative importance for each variable. Temperature and salinity values related to each predictor are then plotted in a five square degree grid.

The dynamics of the two upwellings are addressed according to the relationship between the variables. Sea surface temperature (SST), sea surface salinity (SSS) and chlorophyll are prescribed from the predictor variables and reflect monthly mean oceanographic characteristics over the study area. Based on SST and SSS variation, associated with the predictor variables, the significance of local and remote forcing effects on the upwellings is shown.

IP2t

Current Structure Across the Arabian Sea at 8.5ƒN During June 1995

Robert L. Molinari, W. Douglas Wilson and Amy Ffield

NOAA/Atlantic Oceanographic and Meteorological Laboratory, USA
Lamont-Doherty Earth Observatory, Palisades, New York, USA

As part of the World Ocean Circulation Experiment repeat hydrography program, the NOAA ship MALCOLM BALDRIGE occupied the western portion of I1 (8.5ƒN) during June, 1995. Observations included hydrography from a CTD and direct velocity from both a hull-mounted and a lowered Acoustic Doppler Current Profiler. WOCE station spacing was followed along the transect. The upper layer current structure was consistent with the early stages of the Southwest Monsoon. On the western boundary, "The Great Whirl" was observed, with speeds greater than 50 cm/s confined to the upper 50 m. Transport of the upper 50 m of this feature was 7 Sv. Offshore, flow above 50 m is generally to the south and on the average about 20 cm/s. A transport per unit width profile across the section shows that this predominantly Ekman transport is the upper layer component of a shallow meridional overturning cell. Approximately half the Ekman transport is compensated for by northward flow between 50 and 300 m. Intermediate current structure can be characterized by a series of alternating flows. The amplitude of this wavelike structure was about 5 cm/s and the wavelength about 4 degrees of longitude. Below 3000 m, the velocities show a bottom intensified southward flow along the western boundary with a transport of 5 Sv. East of this boundary flow to the Carlsberg Ridge, flow was in general to the north with a net transport of 19 Sv. West of the Carlsberg Ridge in the Arabian Basin, flow is predominantly to the south with a net transport below 3000 m of 13 Sv. Thus, a portion of the northward flow in the Somali Basin at these depths turns westward north of the section but the largest part turns east and returns east of the Ridge. The Arabian basin water mass characteristics below 3000 m are different than those in the Somali basin, indicating water mass modification in the northern reaches of the Arabian Sea.

IP2u

Nutrient Relationships in the Northern Indian Ocean

P. Chapman, S. Doney and D. Olson

U.S. WOCE Office, Texas A&M University, College Station, USA
NCAR, Boulder, USA
University of Miami, Miami, USA

During the U.S. WOCE cruises in the Indian Ocean during 1995, samples were taken in both the Bay of Bengal and the Arabian Sea. Nitrate:phosphate relationships suggest two distinct source water regions in the north and south Indian Ocean and an intermediate mixing zone between about 12ƒN and the South Equatorial Current. The northern region is affected by nitrate reduction (<12 µmol/kg) in the Arabian Sea and by the inflow of water from the Persian Gulf having an anomalously low N:P ratio. Nitrate reduction in the Bay of Bengal is very much less than in the Arabian Sea. The data and known current patterns suggest that most of the water in the northern Bay of Bengal is derived from the Arabian Sea.

The region in the Arabian Sea where nitrate reduction takes place occurs only north of 12ƒN. It remains uncertain whether nitrate reduction in this region occurs locally, or whether it occurs elsewhere and the reduced nitrate is then advected into the area. The fact that no free sulphide has been found in the region despite oxygen concentrations near zero suggests the latter. We believe that most of the nitrate reduction occurs along the shelf of the Indian west coast, where low levels of oxygen are found above muddy reducing sediments, and that the freshly-reduced water is advected into the center of the northern Arabian Basin, from where it is only slowly replaced.

IP2w

Relationship between the seasonal wind field and the interannual variation of the satellite-derived air-sea turbulent heat flux in the Indian Ocean

Masanori Konda, Norihisa Imasato and Akira Shibata

Division of Earth and Planetary Sciences, Graduate School of Science, Kyoto University, Kyoto, Japan
National Space Development Agency of Japan/Earth Observation Research Center, Roppongi, Tokyo, Japan

We investigated the interaction of the variations at different time scales of the turbulent air-sea heat flux and sea surface temperature (SST) in the Indian Ocean, where both the seasonal and the interannual variability of the thermal correlation between the sea and the air are remarkable. The monthly turbulent heat flux and SST used in this study are obtained by satellite data, ranging from July 1987 to December 1994. We focused on the thermal correlation between the sea and the air on a line along which the seasonal wind is the strongest in the Indian Ocean. In accordance with the seasonal reversal of the wind direction, anomalies of humidity and air temperature on this line are strongly affected by the atmospheric mass in the North hemisphere in the boreal winter, and the South hemisphere in the boreal summer. On the other hand, SST anomaly on this line are not correlated with the local air-sea thermal interaction. Therefore, the interannual variations of atmospheric and oceanographic conditions of temperature and humidity along the course of the strong seasonal wind are correlated more closely with the advective transport than with the local air-sea interaction. It is inferred from these facts that the air-sea turbulent heat flux anomaly in the tropical Indian Ocean is decided not by the local thermal interaction but by the remote effect of the atmospheric anomalies in the mid latitude advected by the seasonal wind.

IP2x

Surface Layer Circulation and Water Mass Distribution at 8.5ƒN in the Arabian Sea During June and September 1995

T.K. Chereskin, W.D. Wilson, H.L. Bryden, A. Ffield
and J. Morrison

Scripps Institution of Oceanography, La Jolla, USA
Atlantic Oceanographic and Meteorological Laboratory, Miami, USA
Southampton Oceanography Centre, Southampton, UK
Lamont-Doherty Earth Observatory, Palisades, New York, USA
North Carolina State University, Rayleigh, USA

Two hydrographic/acoustic Doppler current profiler (ADCP) transects were made across 8.5ƒN 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. The Ekman transport in June was about 18 Sv southward; it was partially compensated by northward geostrophic transport relative to 300 m. The Ekman transport in September was about 8 Sv southward, and the surface layer geostrophic transport was also southward.

The vertical structure of the ageostrophic velocity, the stratification, and the water mass properties were quite different between the two occupations. The ageostrophic flow was confined to a very shallow layer in June (about 50 m), and the surface layer was strongly stratified, with a maximum salinity layer at depths between 50 and 70 m. The ageostrophic velocity penetrated much deeper in September (to about 160 m), and the thermocline was correspondingly deeper. In September the salinity maximum layer extended to the surface in mid-basin, possibly advected from the northern Arabian Sea.

IP2y

Seasonal variations of mass and heat in the Indian Ocean estimated from observations and numerical modeling

C. Perigaud, F. Melin, J. McCreary and P. Delecluse

Caltech/JPL, Pasadena, USA
Nova Southeastern University, Dania, USA
LODYC, Paris, FRANCE

Climatological winds and air-sea fluxes data sets are used to force a 2 and 1/2-layer model over the Indian Ocean. Hydrographic profiles over 1980-1995 derived from Smith (1995) provide climatological data sets that are used for comparison and analysis of the model simulations. The change of heat content in the upper ocean over 400m compares well in amplitude and phase with the simulated one; the largest discrepancies show up on the eastern side of the equator, where data (model) indicate a change of 200W/m² (50W/m²), and in the western region, at the center of the cyclonic gyre south of the equator, where data (model) indicate a change of 10W/m² (250W/m²). Averaged zonally, the meridional heat transport calculated from hydrographic data and from the model air-sea fluxes agrees to some extent with the simulated one. The phase is particularly well reproduced by the model. Between May and October, the model simulates a southward cross-equatorial heat transport of 1.3 Petawatts (PW), and during the rest of the year, the northern ocean replenishes its heat content at a rate of 0.9 PW. The corresponding figures for the observed estimates are 0.9 PW and 0.5 PW respectively. We are presently analyzing OGCM simulations obtained with the OPA model over the Indian Ocean with the same climatological forcings. Our objective is to understand the reasons for the similarities and discrepancies between the various observed and simulated estimates of cross-equatorial heat transport.

IP2z

Satellite Altimetry of the South Indian Ocean

Karen J. Heywood and Y.K. Somayajulu

School of Environmental Sciences, University of East Anglia, Norwich, UK
National Institute of Oceanography, Dona Paula, Goa, India

Sea level anomalies from the TOPEX/POSEIDON and ERS-1 altimeters have been used to study the South Indian Ocean (10ƒN to 40ƒS, 30ƒE to 120ƒE). We have used two years of data (October 1992 to September 1994) to study both the properties of the Rossby waves, and the characteristics of the eddy kinetic energy field.

Baroclinic Rossby wave propagation is seen at all latitudes outside the near equatorial region, and is particularly prominent at 15ƒS. Fast Fourier Transform techniques permit the determination of the wavelength and period of the single spectral components, while the Radon Transform enables evaluation of the phase speeds of the propagating signals. The speeds of the observed signals are compared with a revised theory of Rossby wave propagation which includes the baroclinic east-west mean flows (Killworth, Chelton and de Szoeke, submitted to Journal of Physical Oceanography, 1996). A greater agreement between oberved speeds and predicted speeds is found than with the classical Rossby wave theory. The effects of mid ocean ridges on the Rossby wave properties are investigated.

Changes in eddy kinetic energy field are associated with changes in the monsoonal circulation. Differences between southwest and northeast monsoons are discussed in the light of both the different ocean circulation and the wind forcing. Each basin of the South Indian ocean is investigated separately to determine its seasonality.

IP2aa

Diapycnal Diffusivity Distributions in the Upper Oceans

Huai-Min Zhang and Lynne D. Talley

Graduate School of Oceanography, University of Rhode Island, USA
Scripps Institution of Oceanography, University of California, San Diego, USA

Theoretical and modeling efforts have shown that diapycnal mixing is central to the dynamics of the ocean thermocline and the deep ocean. How to accurately parameterize mixing proceeses caused by eddy activities is one of the major problems in the numerical modeling of the climate (coupled atmosphere-ocean) system because of the lack of knowledge in mixing rates in the ocean. Limited available estimates for diapycnal eddy diffusivity range from O(0.1) to O(100) cm²/s depending on location as well as the estimation methods. In this report we present the mean diapycnal mixing rates as functions of potential density surfaces in the upper ocean in an effort to determine their geographical variations. The diapycnal and diathermal diffusivities are estimated from buoyancy and heat budgets for water volumes bounded by isopycnals and isotherms, air-sea surface and coastlines. The requirement of closed isopycnals relegates the computation to shallow waters generally above 200m. Comprehensive analysis is given to the Indian Ocean, with an extended global general description. Generally diapycnal diffusivity is large (of O(1cm²/s)) near sea surface and in tropics, and decreases poleward and with depth (of O(0.1cm²/s) for isopycnals outcropping at about 40ƒN and 40ƒS and with depths around 150 m in mid ocean).

For the Indian Ocean, the Bay of Bengal and the NE Indian Ocean has the freshest and lightest water, and diapycnal diffusivity remains at about 1.3cm²/s between 20.2 and 22.0 in density anomaly. Further poleward (south) and at greater depth, it decreases from 0.9 to 0.5cm²/s for isopycnals between 23.0 and 25.0 in density anomaly. A parallel calculation for isotherms (unlike density distribution, the warmest water straddles the equator) shows that diathermal diffusivity decreases poleward and with depth. For waters bounded by 25.0 in density anomaly, diapycnal diffusivity is 0.1cm²/s for the Indian-Pacific system (at depth around 170m) and is 0.2cm²/s for the Atlantic (at depth around >120m). Diapycnal diffusivity across 27.5 density anomaly surface (at depth around 400m) is calculated to be 0.2 cm²/s globally. Calculation for the Atlantic shows that runoff from the Amazon plays an important role in the western tropical Atlantic. Diapycnal diffusivity for 23.0 density anomaly surface is only 0.05cm²/s when only air-sea buoyancy flux is considered, but it increases to 0.4cm²/s when the runoff is added.

IP2bb

Meridional Mass and Heat Transports of the Indian Ocean

Tong Lee and Jochem Marotzke

Jet Propulsion Laboratory, Pasadena, USA
Department Earth, Atmospheric & Planetary Sciences, MIT, Cambridge, USA

Climatological hydrography, surface heat and freshwater fluxes, and wind fields are used together with 3 years of altimetry in an attempt to infer the climatological annual mean and seasonal cycle of meridional mass and heat fluxes of the Indian Ocean. The transports are calculated from the solutions of a general circulation model (GCM), which is fitted to the monthly averaged observations while interannual variability is demanded to be small. The GCM's adjoint is used to optimally estimate initial temperature, salinity, surface fluxes, and heat and salt exchanges with the rest of the World Ocean.

Meridional overturning and heat transport display large seasonal variations, with maximum amplitudes of 18 Sv and 2 PW, respectively. North of 20ƒS, meridional mass transport reverses sign with the seasons at all depths. The variations are mainly due to Ekman transports (monsoon reversal in the northern, changing easterlies over the southern Indian Ocean) and external mode transports (Somali current). That the dynamics of the seasonal variability is linear is further supported by the weak rectification effects. The large heat transport variations are balanced mainly by storage changes and not by surface flux variations, except south of 10ƒS. Inclusion of altimetry increases the estimated seasonal transport variabilities, but inconsistencies near the Indonesian throughflow arise, probably because of the short duration of altimetry or the model's lack of an open boundary.

The annual mean meridional overturning is vigorous (14 Sv) only above 500 m and predominantly wind-driven. Deep overturning is weak, in contrast to the extant analyses of the hydrographic sections at 18ƒS and 32ƒS. An attempt is made to understand causes of these discrepancies; plausible candidates are (a) the GCM lacks a representation of mass exchanges with the rest of the World Ocean; (b) both model and climatology have inadequate resolution to represent important density gradients; (c) the transports inferred from the sections do not represent time-averages; and (d) errors in reference level velocities inferred from the sections.

IP2cc

Preconditioning of the Indian Ocean by the Indonesian Throughflow on Interannual Timescales

Roxana C. Wajsowicz

Department of Meteorology/JCESS, University of Maryland, College Park, USA

In his data analysis of connections between the strength of the Indian Monsoon (measured by rainfall) and Warm and Cold Events in the Southern Oscillation, Meehl (1987) found large negative SST anomalies developed as the year of a strong Indian Monsoon progressed. Hasenrath & Greischer’s (1993) data analysis of the monsoonal heat budget of the combined ocean-atmosphere system confirms earlier analyses that the southern Indian Ocean, south of 10ƒS, is the major source of Monsoon water vapour. However, a cause-and-effect relationship between the SST anomalies, evaporative flux anomalies and strength of the Indian Monsoon has yet to be established. Further, interannual wind-stress variations alone cannot account for the development of the observed large SST anomalies.

Interestingly, the southernmost latitude of the Java-Sumatra peninsula is about 10ƒS, and so the Indonesian throughflow exerts its influence over the latitudes described above. A global version of NASA/GSFC’s multi-layer, reduced-gravity model, developed for ENSO prediction, is used to investigate the mechanisms by which upper layers of the Indian Ocean are ‘pre-conditioned’ by variations in the warmth and freshness of the Indonesian throughflow, so enabling larger SST anomalies to develop in response to the interannual wind-stress variations. (The model has been extended to prognose both temperature and salinity, and permit the prescription of the natural surface boundary condition of net fresh-water flux).

Results from a series of experiments illustrating the need for a pre-conditioning mechanism, wave and horizontal advective processes involved in the possible mechanisms, and the different temporal and spatial scales, will be described.

IP2dd

Structure and Transport of the Agulhas Current

Harry L. Bryden, Lisa M. Beal and Michael N. Tsimplis

Southampton Oceanography Centre, Empress Dock,
Southampton, UK

From a hydrographic section combining CTD and LADCP measurements across the Agulhas Current southeast of South Africa during February-March 1995, the poleward transport of the Agulhas Current is estimated to be 70 Sv, about 15 Sv less than other recent estimates. The smaller transport is due to the presence of an equatorward flowing undercurrent, clearly indicated by the LADCP observations, beneath the Agulhas Current. This Agulhas Undercurrent has an equatorward transport of 7 Sv; it was directly observed a second time in April 1996; and there are indications that it was also present in October 1987. Combining the smaller Agulhas Current transport with the 1987 transindian hydrographic section across 32ƒS results in a smaller poleward heat transport across 32ƒS and a smaller net overturning circulation in the Indian Ocean. Year-long current meter measurements appear to confirm the smaller Agulhas Current transport and the presence of an Agulhas Undercurrent.

IP2ee

THE HEAT BALANCE AND SEA SURFACE TEMPERATURE OF THE INDIAN OCEAN: REGULATION AND VARIABILITY

P. Webster, J. Loschnigg, K. Sahami and C.S. Willett

Program in Atmospheric and Oceanic Science, University of Colorado, Boulder, USA

The regulation of sea surface temperature (SST) in the tropical Pacific Ocean warm pool has been the subject of considerable debate for the last few years. The debate surrounds the relative roles of evaporation increases and solar radiation reductions by clouds in modulating the SST and keeping it within observed bounds. However, the Indian Ocean offers a counterpoint to the Pacific Ocean situation. The SST in the North Indian Ocean has a much stronger annual cycle than the Pacific warm pool and by May-June is the warmest body of water on the planet. Through the spring the SST increases at a rate of about 1ƒC/month. However, during this period, the North Indian Ocean is virtually cloudless and the net heat flux into the ocean is between 100-150 W/m² or 5-10 times that of the Pacific Ocean warm pool. Using mixed-layer models, it is estimated that the observed North Indian Ocean warming is a factor of three smaller than expected from the observed heat flux. That is, with the observed heat flux the North Indian Ocean should heat at a rate of 3ƒC/month. Clearly, cloud-radiation feedbacks cannot be the regulating device as the region remains cloudless until the onset of the monsoon. The purpose of the study is to determine what are the physical processes that regulate the North Indian Ocean SST.

The annual cycle of the heat budget of the Indian Ocean is calculated using the McCreary, et al., (1993) 2 1/2 level model. The model is forced using satellite observations of the surface heat budget and the winds from numerical reanalysis products. The characteristics of the model annual cycle are validated using the independent satellite altimeter data. The analysis confirms results from annually averaged calculations of McCreary, et al., (1993) that the heat balance in the North Indian Ocean is accomplished by substantial and reversing heat transports across the equator and changes in storage. The reason for the slow increase of SST in the North Indian Ocean results from a strong southward heat flux. Thus, more so than in the western Pacific Ocean, oceanic dynamics play a crucial role in the regulation of SST and the balancing of the heat budget.

Through Lagrangian trajectory calculations the role of transients in accomplishing the cross-equatorial heat transports is calculated and compared with the annually averaged calculations of McCreary et al., (1993). The results of the analysis are similar as expected. Even though there is very large subannual variance in the Indian Ocean the calculations by the two methods are within 5% of each other. However, the modes of transport are, of course, accounted for differently.

Finally, the very long term variability of the SST of the Indian Ocean is considered. Like the ENSO phenomena, there is considerable interdecadal variability. In times of low ENSO variability, it seems that the interannual Indian Ocean SST variability plays a more significant role in influencing the South Asian rainfall than has been the case during the last two decades.

IP2ff

Observations of Surface Heat Advection in the Indian Ocean

Cynthia S. Willet, Pearn P. Niiler and Peter J. Webster

Program in Atmospheric and Oceanic Sciences, University of Colorado, Boulder, USA
Scripps Oceanography Institute, University of California,
La Jolla, USA

Sea surface temperature data and drifting buoy data are combined to study the surface advective processes of heat in the Indian Ocean. Spatial temperature gradients (dT/dx + dT/dy) from satellite sea surface temperature measurements are combined with the surface velocity data (U) to estimate monthly, seasonally, and yearly averaged surface heat advection. The drifter data also provides a bulk temperature that is used to verify the satellite measured temperatures. The rate of change of heat advection is then compared to contemporaneous changes in local air surface temperatures to investigate possible forcing or feedback influences with the monsoon system. The results are compared with similar analyses from the Pacific Ocean and also with the results of the McCreary et al 2 1/2 level ocean model.

IP2gg

The Subtropical Indian Ocean Response to Monsoonal Variability

Cynthia S. Willett and Peter J. Webster

Program in Atmopspheric and Oceanic Sciences, University of Colorado, Boulder, USA

Ocean waves, observed in sea surface height (SSH) anomaly fields from satellite altimetry, are found throughout the tropics reflecting back and forth across the Indian Ocean and impacting Indian Ocean circulation.

McCreary et al. [Prog. in Oceanogr, 31, 1993] in their study of Indian Ocean dynamics find evidence of both upwelling Kelvin and downwelling Rossby waves influencing circulation and mixed layer depth (MLD) throughout the north subtropical Indian Ocean. Bruce et al. [JGR, 99, 1994] documented the existence of the Laccadive High, an anticyclonic feature that forms near the southern tip of India during the northeast monsoon. This feature is observed as a positive anomaly in a preliminary analysis of satellite altimetry. These waves and wave-like features (i.e. Laccadive High) are important dynamical processes in the subtropical Indian Ocean because of their impact on the ambient ocean structure (i.e. circulation, MLD, and heat fluxes).

The preliminary analysis of the TOPEX SSH anomalies (October 1992 through November 1995) between 15ƒN and 15ƒS shows an intriguing number of waves with varying wave lengths, amplitudes, phase speeds and direction. The most prominent SSH variations are annual and westward propagating. These features are found north of 6ƒN and south of 6ƒS, and between 105ƒE and 75ƒE, and between 75ƒE and 50ƒE. These features are propagating slower than the linear Rossby wave speed and the non-dispersive baroclinic Rossby wave speed. The temporal and spatial appearance of these features seem to be synchronized with the Asian-Australian Monsoon system. These height variabilities may be atmospherically forced (i.e. locally modulated by evolving winds) or enhanced by instabilities in the ocean structure, or the eastern-most variation may be an extension of a wave projected from the Pacific Ocean through the Indonesian throughflow.

The nature of these features, as detected in satellite altimetry, are analyzed (i.e. wave type, frequency, temporal variations in amplitude) through decomposition with wavelet analysis. Then, the wind stress fields, sea surface temperature fields, and coincident satellite SSH anomalies are studied with a correlation analysis to estimate air-sea interactions including momentum exchanges (and possible lags), and heat fluxes. The SSH analysis is also used to validate an intermediate ocean model forced independently by surface heat budgets obtained using satellite data and wind stress from numerical reanalysis products.

IP2hh

Interannual Variation of the Southern Tropical Indian Ocean

Gary Meyers and Yukio Masumoto

CSIRO Division of Marine Research, Hobart, Australia
University of Tokyo, Tokyo, Japan

The TOGA/WOCE XBT Programme established several XBT lines using Volunteer Observing Ships in the Indian Ocean starting in 1983. The coverage on low density lines was typically 12 sections per year with 90 miles between drops, and on frequently repeated lines as much as 18 sections per year with 60 miles between drops. The XBT data from lines IX1 (Fremantle to Sunda Strait), IX12 (Fremantle to Red Sea) and IX22 (Port Hedland to Japan) are used to characterise large scale, interannual variability of baroclinic structure and currents in the eastern half of the Southern Tropical Indian Ocean (STIO). The generation of the variations by wind forcing is studied using ocean models.

Using data near the eastern boundary from IX1 and IX22, EOF analysis shows that the dominant mode of variation off the northwest coast of Australia is an El NiÒo signal, apparently entering the region from the Pacific Ocean. A second mode is strongest on the coast of Indonesia. It is a response to winds over the equatorial Indian Ocean, and is consequently called here the Indian Ocean mode. It has a shorter time scale than the El NiÒo signal. Variation in the central STIO on IX12 is dominated by the Indian Ocean mode. EOF analysis does not show an El NiÒo signal, expected to propagate from the coast of Australia. A global ocean general circulation model forced by winds for the period 1990 - 1993 produces variations in the STIO that are dominated by the Indian Ocean mode. Preliminary results indicate that wind stress curl over the eastern STIO substantially modifies extraequatorial baroclinic variations propagating to IX12.

IP2ii

Seasonal variation of transport between the Red Sea and the Indian Ocean

Stephen P. Murray, William Johns, Abraham Fisseha, Ali DouAbul and Mohammed Al-Safani

Coastal Studies Institute, Louisiana State University, Louisiana, USA
RSMAS/MPO, University of Miami, Miami, Florida, USA
Resources and Environment Division, Ministry of Marine Resources, ERITREA
Department of Oceanography, University of Sana’a, YEMEN

The Bab al Mandab Strait linking the Red Sea and Indian Ocean via the Gulf of Aden provides the pathway for Red Sea deep water to spread through the Indian ocean thermocline. Indirect estimates of the transport of Red Sea water through the Bab al Mandab Strait suggest an annual mean transport of 0.33 Sv (Siedler, 1968), varying from approximately 0.6 Sv in winter to nearly zero in later summer (Patzert, 1974). A mooring program to measure the annual cycle of volume and salt transport and its fluctuations began in May 1995 with deployment of 14 moorings, inlcuding 3 moored ADCPs and 12 SeaCats.

The ADCP and SeaCat moorings have clearly observed the annual cycle of volume and salt transport. Our records begin in May 1995, with a two-layer exchange expected for wintertime conditions (Siedler, 1968). Deep outflow of salinity in excess of 40 psu extend up to 75 m above the bottom with speeds reaching 80 cm/sec. Surface layer inflow is about half that strength.

By early July, the two-layer winter regime ceases and is replaced by a three-layer flow structure with weak (and occasionally vanishing) deep outflow in the lowest 25 m, a thick intermediate inflow layer, and outflow in the near surface layer (approximately the upper 20 m). The surface outflow is attributed to a seasonal reversal in winds over the southern Red Sea to north westerly during the Indian Ocean summer monsoon season (June-September). The intermediate layer is a massive intrusion of cold (18ƒ C) low salinity (36-
36.5 psu) water from the Gulf of Aden over the 40-130 m depth range from June to October.

Around the beginning of October, this three-layer regime gives way again to the two-layer winter regime which builds to full strength by November. Maximum exchange occurred in February with lower layer outflow speeds greater than 1.2 m/s and inflow speeds of approximately 0.6 m/s. Near the end of the record, in late March 1996, there is weakening of the two-layer flow which could possibly signal an early transition to the summer regime in 1996, or be related to a shorter term fluctuation. Throughout the record there are fluctuations in inflow/outflow intensity on time-scales of about one month that are likely related to variations in surface forcing. Transport computations are compared to the indirect estimate of Siedler (1968) and Patzert (1974).

IP2jj

Dynamic and Thermodynamic Effects of the Indonesian Throughflow on the Indo-Pacific Basin

Ragu Murtugudde, James Beauchamp and A. Busalacchi

Code 970/Bldg 22, NASA/GSFC, Greenbelt, USA
Laboratory for Hydrospheric Processes, Hughes STX, USA
Laboratory for Hydrospheric Processes, NASA/GSFC, Greenbelt, USA

Most model studies of seasonal-to-interannual variability of the tropical Pacific are conducted with the Indonesian Throughflow (ITF) closed. We investigated the interannual variability of the ITF and its influence on the thermodynamics and the dynamics of the Indo-Pacific basin. Since most of the ITF is supposed to occur in the upper 400 m, we ignore topographic effects in the region and employ a reduced gravity, primitive equation, ocean GCM coupled to an advective atmospheric mixed layer. Simulations are carried out over the Indo-Pacific basin (50S-45N) with the ITF closed and open.

Interannual simulations for 1980-1995 are carried out with interannual winds but climatological cloudiness, solar radiation and precipitation data. The seasonality of the model ITF and its interannual variability agree very well with the available data and other model estimates. The model also reproduces seasonal-to-interannual variability in the Indian and Pacific Oceans. Opening of the ITF leads to cooler SSTs in the Pacific with a shallower thermocline and warmer SSTs in the Indian Ocean with a deeper thermocline. However, the largest thermodynamic and dynamic signals occur below the surface. Detailed analysis of the simulations with and without the ITF also show the effects of the ITF on the size of the warm pool, surface currents, the equatorial undercurrent and the eastern equatorial Pacific thermocline. Significance of the effects of the ITF, possible consequences of excluding the ITF will be discussed along with the variability of the ITF along various pathways in the model. Large scale seasonal-to-interannual forcing of the ITF in this model will be compared to other models and data.

The Indo-Pacific warm pool is also a region of the tropics where salinity effects could potentially be very important. Model experiments with active salinity and constant salinity in each layer show that considerable difference occur in the ITF if active salinity is not allowed in the model. Analysis of salinity effects on model transport through the Indonesian Channel and also on model dynamics and thermodynamics in the Indo-Pacific region will be presented.

Interannual variability of the ITF with active and constant salinity and the correlation with ENSO will be discussed.