Impact of along-track altimeter sea surface height anomaly assimilation on surface and sub-surface currents in the Bay of Bengal

In this study, altimeter based sea surface height anomaly (SSHA) observations from the Jason-3 altimeter are assimilated in a high resolution Indian Ocean model and the impact of assimilation is analysed in the Bay of Bengal region. The study uses a 3-D primitive equation numerical ocean model forced with analysed atmospheric data to generate ocean state for a particular period, in this case from the year 2018. To study the impact of SSHA assimilation, two sets of model outputs are generated, one in which no SSHA is assimilated (Ctrl-R) and the other in which along-track SSHA from the Jason-3 satellite is assimilated (Assim-R) using Ensemble Optimum Interpolation technique. The impact of assimilation is primarily analysed on ocean surface currents. For this purpose, satellite-derived sea surface currents and buoy based vertical current profiles are used to evaluate Ctrl-R and Assim-R. It is found that there is a profound improvement in ocean current simulation after assimilation of SSHA. Mesoscale eddies that were weak and misplaced in Ctrl-R have now improved in Assim-R in terms of both strength and position. Sub-surface current simulation from Assim-R exhibits observed variability in most of the evaluation period, which was absent in the Ctrl-R. Computed statistics show much improved (15%-35%) ocean current simulations in Assim-R at all the depths. Particle trajectory analysis has been performed and it is seen that assimilated model velocity fields faithfully simulate the observed trajectory of tracers. Power spectra of simulated SSHA from both Assim-R and Ctrl-R were compared with spectra from satellite SSHA and it is found that after assimilation the slopes of the simulated spectra become much closer to slope of spectra from the observations. These results indicate that the along track assimilation of SSHA improves the model analysis and thereby provides accurate initial conditions for model forecast.