Three drifters were deployed in the northeastern Arabian Sea at ($69.18^{0}\rm{E}, 19.77^{0}\rm{N}$). They were released almost at the same time on 29 November 2016. The three drifters, initially moved poleward along the direction of the west India coastal current. The distance between any two drifters is less than 5 km for initial 8 days. The drifters veered apart when they moved along the edge of an anti-cyclonic or cyclonic circulating loop, thereby increasing the distance between them. Within a period of six months, the three drifters were in different directions and the distance between them was more than 600 km.

We describe the variability of the West India Coastal Current (WICC) during October 2008 to October 2018 using data from ADCP (acoustic Doppler current profiler) moorings deployed on the continental slope off the west coast of India. The four moorings are deployed off Mumbai ($\sim 20^{0}\rm{N}$), Goa ($\sim 15^{0}\rm{N}$), Kollam ($\sim 9^{0}\rm{N}$), and Kanyakumari ($\sim 7^{0}\rm{N}$). This 10-year data set allows us to attach a statistical significance to the conclusions drawn by Amol et al. (2014) on the basis of four years (October 2008–October 2012) of ADCP data. The longer data set confirms the earlier finding that intraseasonal variability in the 30–90-day band dominates the variability of the WICC at all locations and that this intraseasonal variability peaks during the winter monsoon. The annual cycle (300–400 days) is strong and statistically significant at all locations. The phase propagates upward for the annual cycle and this phase difference is seen in the relative phases of both, the ADCP currents at 25 and 48 m as well as the 48 m ADCP and satellite-derived currents. The intra-annual (100–250 days) and intraseasonal currents show instances of both upward and downward phase propagation. The alongshore wavelet coherence is high on seasonal time scales between adjacent mooring locations and several instances of high coherence are seen even on intraseasonal time scales. Data gaps off Goa and Kanyakumari restrict the significant wavelet power to the ADCP records off Kollam and Mumbai, and the coherence analysis shows that the WICC off Kollam leads Mumbai on seasonal scales. The direction of the alongshore WICC is, however, largely determined by the direction of the significantly larger intraseasonal component. Though the climatological seasonal cycle over the whole record does show the canonical equator ward flow during the summer monsoon (June–September) and poleward flow during the winter monsoon (November–February), the scatter around the daily mean is very high.The data show that the WICC may flow in either direction on a given day of the year, with this unpredictability of direction being stronger off Kollam, where the $1-\sigma$ band of the daily mean alongshore WICC shows that it can flow in either direction in most months. The seasonality is stronger off Mumbai, where the width of the $1-\sigma$ band is less. The decade-long continuous record off Kollam and Mumbai shows that the sub-annual along shore WICC at both locations is significant and is comparable to or stronger than the annual component.The cross-shore sub-annual current is also strong off Kollam and is seen to be associated with eddy-like circulations.

We describe the variability of the East India Coastal Current (EICC) during 2009–2018 using data from ADCP (acoustic Doppler current profiler) moorings deployed on the continental slope in the western Bay of Bengal. The four moorings are deployed off Gopalpur ($19.5^{0}\rm{N}$), Visakhapatnam ($\sim 18^{0}\rm{N}$), Kakinada ($\sim 16^{0}\rm{N}$), and Cuddalore ($\sim 12^{0}\rm{N}$) on the Indian east coast. The longer data record allows us to attach a statistically more robust basis to the conclusions drawn by Mukherjee et al. (2014) on the basis of four years (2009–2013) of ADCP data. The data confirm that the seasonal cycle dominates the variability of the EICC. The amplitude of the annual band varies over the time series. In the intra-annual band, the variability switches between the semi-annual and 120-day bands off Gopalpur, Visakhapatnam and Kakinada, but the semi-annual band is stronger than the 120-day band off Cuddalore throughout the time series. Upward phase propagation is common in the seasonal bands, but downward phase propagation is common in the intra-annual band of Cuddalore during the summer and winter monsoons, leading to stronger undercurrents there. Off Cuddalore, even the annual EICC appears as a shallow current. In contrast, the EICC appears as a deep flow of Gopalpur, Visakhapatnam, and Kakinada particularly during the spring inter-monsoon. This deep flow is evident at these locations even in the intraseasonal (30–90-day) band; the longer data set suggests, however, that the intraseasonal variability does not necessarily peak during spring. The annual EICC is coherent along the coast, but it is only the semiannual band that shows a comparable coherence between Kakinada and Cuddalore: in the 120-day and intraseasonal bands, the EICC decorrelates along the coast. Wavelet analysis suggests significant variability at sub-annual periods. The sub-annual EICC exceeds $20 cm s^{-1}$ on many occasions, but it too decorrelates along the coast. The long ADCP record allows us to confirm the dominance of seasonality in the EICC regime in a robust fashion; the data show that the EICC tends to flow in its canonical poleward (equatorward) direction during spring (winter). This dominance of seasonality enhances the predictability of the EICC.

We describe the variability of the East India Coastal Current (EICC) during 2010–2018 on the outer continental shelf using data from four ADCP (acoustic Doppler current profilers) moorings deployed off Gopalpur ($\sim19.5^{\circ}\rm{N}$), Visakhapatnam ($\sim18^{\circ}\rm{N}$), Kakinada ($\sim16^{\circ}\rm{N}$), and Cuddalore ($\sim12^{\circ}\rm{N}$) on the east coast of India. In general, the shelf EICC mirrors the slope EICC for the annual and semi-annual cycles, but the shelf-slope coherence is weaker and patchy for the 120-day and intraseasonal bands. The seasonal cycle, which consists of the annual, semi-annual, and 120-day bands, dominates the observed variability. The amplitude of the annual cycle varies over the time series. In the intra-annual band, variability tends to switch between the semi-annual and 120-day bands, but the former dominates throughout the time series off Cuddalore, the southernmost location. The EICC appears as a shallow current in all period bands, including the seasonal cycle, off Cuddalore, but even the intraseasonal EICC appears as a deep current at the other three locations. A wavelet analysis shows seasonal variation of the wavelet power in the intraseasonal band, suggesting that the amplitude of intraseasonal variability itself varies with season, but there is no clear seasonal pattern. As on the continental slope, the annual and semi-annual components are coherent along the coast, but alongshore coherence is weak at shorter time scales. Upward phase propagation is evident for the seasonal cycle at all locations, but downward phase propagation, seen on the slope off Cuddalore, is evident on the shelf as well. The 500-day low-pass filtered shelf EICC is not weak and the sub-annual variability is comparable to that on the slope. The long ADCP record allows us to confirm the dominance of seasonality in the EICC regime in a robust fashion; the data show that the EICC tends to flow in its canonical poleward (equatorward) direction during spring (winter). This dominance of seasonality enhances the predictability of the EICC.