Impact of assimilation of SCATSAT-1 data on coupled ocean-atmospheric simulations of tropical cyclones over Bay of Bengal

This study examines the impact of assimilation of the surface winds obtained from SCATterometer SATellite-1 (SCATSAT-1) in predicting the tropical cyclones over the Bay of Bengal using a coupled ocean-atmospheric model. Three sets of numerical experiments are conducted for six cyclones during post-monsoon (VARDAH, GAJA, PHETHAI) and pre-monsoon cyclones (MORA, FANI, AMPHAN). The first experiment, ‘CONTROL’, is conducted with the Weather Research and Forecasting - Ocean Mixed Layer (WRF-OML) model initialized using Global Forecasting System analysis and ocean initial conditions obtained from the HYbrid Coordinated Ocean Model (HYCOM) model. The second experiment, ‘PREPBUFR’, is conducted by assimilating the National Center for Environmental Prediction (NCEP) prepared BUFR observations with the WRF-OML model and three-dimensional variational assimilation method. Further, the 'SCATSAT' experiment is conducted as PREPBUFR, but additionally, the SCATSAT-1 surface wind vectors are assimilated. Our results of the simulated tracks from three experiments suggest that CONTROL and PREPBUFR simulations exhibit faster translation speed and more track deviations than the India Meteorological Department (IMD) observations. The realistic representation of low-level cyclonic vortex through the assimilation of SCATSAT-1 winds seems to produce positive feedback to both track and intensity, producing a significant improvement in predicting intensity and marginal enhancement on simulation of track and translation speed. The assimilation of winds further improved the representation of different life cycles of the storms as seen in IMD. The analysis of air-sea parameters, in terms of mixed layer deepening, sea surface temperatures, and air-sea flux exchanges, suggests that the response of air-sea feedback is strong in SCATSAT compared to PREPBUFR and CONTROL. Overall, the assimilation of SCATSAT surface winds improved the WRF-OML performance on the prediction of track and intensity, upper ocean response, primary and secondary circulations of tropical cyclones, and the rainfall distributions.