I.4. The Southern Tropical Indian Ocean

The Indian Ocean is a predominantly southern hemisphere ocean, large areas of which are relatively unexplored, compared to the more travelled northern hemisphere basins, resulting in a scarcity of both oceanographic and meteorological data. It is therefore useful to use numerical models in order to increase our understanding of the southern tropical Indian Ocean, through the comparison of model fields with available observational data.

The dominant feature of the observed southern hemisphere circulation in the Indian Ocean is the subtropical cyclonic (clockwise) gyre comprised of the westward flowing South Equatorial Current (SEC) at 12°S to the south, the South Equatorial Countercurrent to the north, and the East African Coastal Current in the west. This gyre is the response of the ocean to the mean basin-wide wind stress curl distribution and exhibits most of the features of a classical mid-latitude gyre, modified by its proximity to the equatorial wave-guide and by the large seasonal variability in the wind fields. As described previously, the presence of the equatorial wave guide allows energy to propagate away from the western boundary in the form of equatorial Kelvin waves and mixed Rossby-gravity waves. The low-latitude of this gyre allows energy input by the winds across the interior to propagate westward much more rapidly than in a mid-latitude western boundary region.

I.4.1. The South Equatorial Current links to the South Equatorial Current

The SEC seperates the clockwise tropical gyre from a counterclockwise subtropical gyre to the south. The SEC roughly follows the line of zero wind stress curl (Le Blanc, 1996). Seasonal variability in the latitude range of the SEC is generated to the east of 100°E by the annual cycle in the wind stress curl and propagates westward as a Rossby wave at approximately 0.1 m.s-1 (Woodberry et al. 1989). These waves have already been mentioned in the previous chapter concerning the Indonesian throughflow. Morrow and Birol (1997) investigated seasonal and interannual variability in the south-eastern Indian Ocean with the aid of TOPEX/POSEIDON (T/P) altimeter data for the 3 year period 1993-1995. They observe a most important annuam signal centred around 10-15°S, with maximum amplitude of 15 cm occuring at 90°E. This signal corresponds to annually-forced Rossby waves, previously described by Périgaud and Delecluse (1992) using GEOSAT data for the years 1988-1989 and a shallow-water tropical model. Morrow and Birrol (1997) notice however that " it is not well understood why the maximum amplitude occurs at 90°E - where the wind stress curl is weakest. ". The influence of the Indonesian throughflow in triggering these features might be worth investigating. Woodberry et al. (1989) note that " the structure of the SEC does not depend crucially on the imposition of an Indo-Pacific throughflow [in the numerical model], although details of the flow in the eastern basin most likely are affected by the throughflow. "

These waves are partially blocked by the banks along the Seychelles-Mauritius Ridge at 60°E. In March through June, a small clockwise eddy forms to the east of the Seychelles-Mauritius Ridge between 10° and 12°S, most likely due to non-linear interactions among the incoming long Rossby waves and the reflected short Rossby waves. By late July, this eddy has become small enough in diameter that it is advected through the gap in the Seychelles-Mauritius Ridge at 12° to 13,5°S and continues toward the west. This advective oscillation appears to be the source of the 70-day oscillations seen at the east coast of Madagascar ( Woodberry et al., 1989).

I.4.2. The East African Coastal Current

Part of the SEC passas through the gap between 12° and 13,5°S and on to the coast of Madagascar, while substantial portion flows to the south around the Seychelles-Mauritius Ridge, and then toward the west between 17° and 19°S. The flow that passes through the gap flows around Cape Amber, at the northern tip of Madagascar and continues westward to the African coast, where it feeds the East African Coastal Current (EACC). The EACC runs northward throughout the year between latitudes 11°S and 3°S, with surface speeds exceeding 1m.s-1 in northern summer, during the southwest monsoon.

The EACC has an important role in linking the boundary currents north of Madagascar and at the equator. Though the SEC splits into a north-going and a south-going branch, the latter feeding the Mozambique Channel, it seems that most of the water in the upper 300 dbar of the northern branch of the SEC goes into the EACC (Swallow et al., 1991). In northern winter (the northeast monsoon), the EACC meets the southward flowing Somali Current near 3°S, and together they form the Equatorial Countercurrent (ECC), at the surface.

Mysak and Mertz (1984) found a 40- to 60- day oscillation in the longshore currents at the African coast between the equator and 5°S. Quadfasel and Swallow (1986) reported 50-day oscillations in current meter records off the northern tip of Madagascar. 50-day oscillations are also found in numerical shallow-water model results forced by monthly mean winds (Kindle and Thompson, 1989 ; Woodberry et al., 1989). Since the shortest period resolved in the wind forcing is 60days, the 50-day variability cannot be due to atmospheric forcing. Kindle and Thompson (1989) and Schott et al. (1988), conclude that the oscillations are due to internal instabilities in the ocean. Horizontal shear (barotropic) instability throughout the region may lead to the formation of eddies at a period of 40 to 50 days since they are the only dynamical insatbility mechanism in the shallow-water model used by Woodberry et al. (1989).

It is still unclear why the period of 50 days is predominant. It may be that this period is a natural one for the system. It is worth noting however, that 40- to 50-day oscillations were reported in the tropical atmomsphere by Madden and Julian (1972) and found in winds over the western Indian Ocean by Mertz and Mysak (1984). Since the hypothesis of wind-forcing for explaining the presence of 40- to 50-day osciallations is not acceptable, it may be possible that the oceanic oscillations force those in the atmosphere through convection processes. COADS' Sea Level Pressure (SLP) data show a maximum of variability in the western Indian Ocean around Madagascar, through empirical orthogonal function (EOF) analysis (Le Blanc, 1996).

This region is therefore an important one since anomalies within the seasonal cycle might be reflected on the African coast at the equator and trigger eastward Kelvin or Yanai waves. Thus a strong signal would quickly travel across the basin and might perturbate the Indonesian coastal circulation in the form of a coastal trapped Kelvin wave. Numerical tests with a shallow-water model show that most of the energy of a Kelvin wave reaching the Indonesian coast at the equator, is propagating southward into the Indo-Pacific throughflow (personal results). This leads us to the study of interannual variability of the Indian Ocean which will be the subject of the following chapters.

I.4.3 The Mozambique Channel and the Agulhas Current

links to the Agulhas Current

A unique aspect of the region is that the Mozambique, East Madagascar, and Agulhas Currents all run against the prevailing wind direction during the Southwest Monsoon season. Winds near southern Africa are westerly to southwesterly throughout the year.

The Agulhas, south of 30°S is one of the strongest currents of the world ocean with a mean speed of 1.6m.s-1 . It shows little seasonal variation but peak speeds can exceed 2.5m.s-1 . The contribution of the Mozambique Current to the Agulhas Current is comparatively small ; the East Madagascr Current is the more important source for the Agulhas Current. The East Madagascar Current is the southern branch of the South Equatorial Current which seperates as it reaches the eastern coast of Madagascar (Tomczak & Godfrey, 1994).