Estimating Vertical Velocity in the Equatorial Indian Ocean

In contrast to the relatively steady easterly trade wind regimes of the tropical Pacific and Atlantic Oceans that favor upwelling and the development of a sea surface temperature (SST) cold tongue along the equator, the Indian Ocean is characterized by energetic seasonally reversing monsoon winds and downwelling favorable mean westerlies along the equator. This downwelling circulation and its variability have profound consequences for the general circulation of the ocean, climate variability, marine biogeochemistry, ecosystem dynamics, and pelagic fisheries. Yet, there have been very few studies estimating vertical velocity from direct observations in the Indian Ocean.

To address this problem, we use horizontal velocity time series from the Research Moored Array for African-Asian-Australian Monsoon Analysis and Prediction (RAMA) in the central equatorial Indian ocean to estimate vertical velocities based on vertical integration of the continuity equation for incompressible flow. The estimates are computed in the upper 200 m of the water column over a 5-year period over 2008-2013 from an array of acoustic Doppler current profiler (ADCP) moorings centered at 0°, 80.5°E. Mean downwelling was observed from the surface to 80 m, driven primarily by meridional convergence in the surface layer and secondarily by convergence of zonal flow along the equator. Vertical velocity near the surface is negatively correlated with local zonal wind stress, consistent with near surface flow variations governed by Ekman dynamics. In the thermocline, mean divergence was observed consistent with geostrophic flow in balance with a zonal pressure gradient set up by the mean surface westerly wind stress. This divergence was fed by both near surface downwelling and deeper upwelling from below. SST was not significantly correlated with near surface vertical velocity at this location in a way that would suggest vertical advection was important in the surface layer heat balance. This result indicates that other processes (e.g., surface heat fluxes and lateral advection) must control SST variability at this location.