Role of regional and global datasets in the simulation of intense tropical cyclones over Bay of Bengal region in a convection-permitting scale

The efficacy of global and regional datasets on the prediction of extremely severe cyclonic storms over the Bay of Bengal (BoB) was evaluated using the Weather Research and Forecasting (WRF) model in a double-nested domain with a 4 km finer resolution on three different datasets, namely FNL, ERA-Interim, and Indian Monsoon Data Assimilation and Analysis (IMDAA). The initial cyclonic vortex, the vertical profile of horizontal wind speed, and relative humidity from different datasets were assessed to evaluate the initial structure and validated with IMD best-fit track data. The model results highlight that simulations with FNL data predict tracks and intensities more accurately for the majority of cyclonic storms compared with the IMDAA and ERA-Interim datasets. Simulations with FNL data exhibit the least mean track errors of 70, 126, 121, and 204 km for days 1–4, respectively. Additionally, the mean wind error of five extremely severe cyclonic storms (ESCSs) using FNL data is approximately 9.3, 4.6, 7.7, and 10.9 m/s, respectively, from day 1 to day 4. It is observed that the regional reanalysis of IMDAA datasets outperformed the forecast of several parameters such as maximum surface wind speed, central sea level pressure, and rainfall for the ESCSs Fani and Sidr. The FNL dataset overpredicted the amount of 24-h accumulated rainfall compared with the ERA-Interim and IMDAA datasets, whereas the IMDAA dataset performed better with lower values of root mean square error (148 mm/day), standard deviation (124 mm/day), and higher correlation (0.68) with the TRMM dataset. Model predictions highlight that the regional dataset IMDAA performs better in predicting rainfall magnitude compared with the global dataset due to the added assimilation of numerous local observations. The regional dataset could be improved by exploring large-scale circulation features and their significant role in predicting the track, intensity, and landfall location of the tropical cyclones.