The Bay of Bengal receives a large influx of freshwater from precipitation and river discharge. Outflow of excess freshwater and inflow of saltier water is required to prevent the bay from freshening. Relatively fresh water flows out of the bay along its boundaries and inflow of saltier water occurs via the Summer Monsoon Current (SMC), which flows eastward from the Arabian Sea into the bay. This saltier water, however, slides under the lighter surface water of the bay. Maintaining the salt balance of the bay therefore demands upward mixing of this saltier, subsurface water. Here, we show that an efficient mechanism for this mixing is provided by upward pumping of saltier water in several bursts during the summer monsoon along the meandering path of the SMC. Advection by currents can then take this saltier water into the rest of the basin, allowing the bay to stay salty despite a large net freshwater input.

We present data from three acoustic Doppler current profilers (ADCPs) moored off Cuddalore (12 • N), Kakinada (16.5 • N), and Gopalpur (19 • N) on the continental slope of the western Bay of Bengal and one mooring on the slope of the northern bay (89 • E, 19 • N; referred to as being located at Paradip). The data were collected during May 2009 to March 2013 and the observations show that the seasonal cycle, which includes the annual cycle, the semi-annual cycle, and a peak around 120 days, dominates the observed variability of the East India Coastal Current (EICC). Spectral analysis suggests that the 120-day peak dominates the seasonal variability at Paradip and is strong at Gopalpur and Kakinada. The annual cycle is coherent along the western boundary of the bay, i.e., the east coast of India, but with significant phase differences between moorings. At the semi-annual and 120-day periods, the alongshore coherence is weaker. Intraseasonal variability is weaker than the seasonal cycle, particularly at Cuddalore and Paradip, and it exhibits seasonality: the strongest intraseasonal variation is during spring (February-April). Peaks around 12 and 20-22 days are also seen at Gopalpur, Kakinada, and Cuddalore. A striking feature of the currents is the upward phase propagation, but there are also instances when phase propagates downward. The much lower vertical phase speed in the top ∼100 m at Cuddalore leads to a distinct undercurrent at this location; at other locations, the undercurrent, though it exists often, is not as striking. During spring, however, the EICC tends to flow poleward (eastward) at Cuddalore, Kakinada, and Gopalpur (Paradip) over the top ∼300 m, which is the maximum depth to which observations were made. The cross-shore component of the EICC is much weaker than the alongshore component at Cuddalore and, except for a few bursts during spring, at Kakinada and Gopalpur. It is only at Paradip, on the slope of the northern boundary, that significant cross-shore flows are seen during spring and the summer monsoon (June-August) and these flows are seen to be associated with eddy-like circulations in the altimeter data. We use the ADCP data to validate popular current data products like OSCAR (Ocean Surface Currents Analyses Real-time), ECCO2 (Estimating the Circulation and Climate of the Ocean, Phase II), and GODAS (Global Ocean Data Assimilation System). The OSCAR currents at Paradip match the observed currents well, but the correlation is much weaker at the other three locations. Both ECCO2 and GODAS fair poorly, particularly the latter because its variability in this boundary-current regime is extremely weak. Though it performs badly at Paradip, ECCO2 does capture the observed variability on occasions at the other locations.

5The Bay of Bengal (BoB) plays a fundamental role in controlling the weather systems that make up the South Asian summer monsoon system. In particular, the southern BoB has cooler sea surface temperature (SST) that influence ocean-atmosphere interaction and impact on the monsoon. Compared to the southeast, the southwestern BoB is cooler, more saline, receives much less rain, and is influenced by the Summer Monsoon Current (SMC). To examine the impact of these features on the monsoon, the BoB Boundary Layer Experiment (BoBBLE) was jointly undertaken by India and the UK during June

We present current data from acoustic Doppler current profilers (ADCPs) moored on the continental slope off the west coast of India. The data were collected at four locations (roughly at Kanyakumari, Kollam, Goa, and Mumbai) extending from ∼ 7 • to ∼ 20 • N during 2008-2012. The observations show that a seasonal cycle, including an annual cycle, is present in the West India Coastal Current (WICC); this seasonal cycle, which strengthens northward, shows considerable interannual variability and is not as strongly correlated along the coast as in climatologies based on ship drifts or the altimeter. The alongshore decorrelation of the WICC is much stronger at intraseasonal periods, which are evident during the winter monsoon all along the coast. This intraseasonal variability is stronger in the south. A striking feature of the WICC is upward phase propagation, which implies an undercurrent whose depth becomes shallower as the season progresses. There are also instances when the phase propagates downward. At the two southern mooring locations off Kollam and Kanyakumari, the cross-shore current, which is usually associated with eddy-like circulations, is comparable to the alongshore current on occasions. A comparison with data from the OSCAR (Ocean Surface Currents Analyses Real-time) data product shows not only similarities, but also significant differences, particularly in the phase. One possible reason for this phase mismatch between the ADCP current at 48 m and the OSCAR current, which represents the current in the 0-30 m depth range, is the vertical phase propagation. Current products based on Ocean General Circulation Models like ECCO2 (Estimating the Circulation and Climate of the Ocean, Phase II) and GODAS (Global Ocean Data Assimilation System) show a weaker correlation with the ADCP current, and ECCO2 does capture some of the observed variability.

No abstract