Warm waters with temperature exceeding 28.5ﾟ C are present in the vicinity of the equator with layer thickness (100m) towards north of equator (Fig. 10). Towards south, the surface temperature decreased to about 22ﾟ C at the southern most latitude (25ﾟ S). The presence of a meso-scale cold core eddy in the Bay of Bengal is evident from the upper layer (400m) XBT thermal structure north of 5ﾟ N. Its salinity structure, however, is not known. While vertical gradients of temperature are strong in the thermocline in the region north of 5ﾟ S, they are weak with diffused thermocline south of 8ﾟ S. Interestingly, large vertical excursions of isotherms from depths as deep as 700m to shallower levels are noticed at the southern latitudes.
The theta-S structure at the stations between 5ﾟ N and the equator (Fig. 12) clearly shows the presence of three watermasses in the upper 450m followed by a bottom watermass below depths of 3500m.
The subsurface low salinity watermass (34.75 PSU) is located at 24.0 sigma-theta around 130m at 5ﾟ N. The salinity of this watermass increases gradually towards the equator. Interestingly, intrusion of low salinity watermass around 26.0 sigma-theta is noticed at 2.5ﾟ S.
While the upper level saline waters have the origin in the Arabian Sea and flow towards east, the source of less saline subsurface waters at depths of 130m deserves explanation and further exploration. Towards this, the vertical distributions of nutrients and oxygen are examined to find a clue. This has shown that the above less saline waters are also characterized by higher silicates, phosphates and oxygen. This suggests that flow into this region is associated with the run-off or fresh water discharge. The flow through the Bay of Bengal having the origin in the Andaman Sea as reported earlier by Sarma et al., (1986) deserve a revisit. Possibly the waters from Malacca Straits through the NEC flowing from east to west during northern winter season (Sverdrup et al., 1942; Wyrtki, 1961) could also be a candidate. However, the presence of high salinity waters at the near surface depths indicates the weakening of the NEC in the upper 100m by January (eastern transect) due to weakening of the northeasterly winds in this region. The surface circulation also supports the weak westward flowing NEC. The flow between 3ﾟ N and 1ﾟ S is mainly towards southwest and tends to become westward at the western transect (south of Sri Lanka). Between 1ﾟ S and 5ﾟ S, an eastward flow associated with ECC which turns northeastward east of 85ﾟ E can be seen. In this region, the nearsurface high salinity (35.4 PSU) watermass centered at 23.0 sigma-theta is brought in to wards east by thr ECC that prevails during the northern winter season following Wyrtki (1973). This ECC brings with it the high salinity waters of the Arabian Sea towards east which compares well with Sverdrup et al. (1942).
The vertical section of sound speed from Cape Leewin to Madras along with the computed intensity plot are shown in Fig. 13. The variation in the depth of the SOFAR channel and the increased intensity in its vicinity are clearly seen from these diagrams. Two acoustic paths identified by the researchers at CSIRO, Hobart & N. I. O, Goa emanating from a point source situated off Pt. Leeuwin, southwest Australia, reach the Indian Ocean shores-Madras and Southern Peninsular India (Fig. 14). In order to obtain and select suitable experimental site(s) for better signal reception utilizing Geodesic path and refracted geodesicpath, algorithms were developed. Ray trajectories for a rangedependent environment have been worked out apart from travel time and signal strength estimations. The climatic data collected and processed at standard depths available at one degree intervals and the CTD and XBT profiles taken along Madras-Perth-Madras transects were used for calibration of the geodesic acoustic paths connecting the Indian shores to southwest Australia which is crucial for the deployment of any equipment for long term monitoring.
