Storm surge prediction improvement using high resolution meso-scale model products over the Bay of Bengal

The Bay of Bengal (BoB) basin is regarded as one of the most cataclysmic basins in the world from a tropical cyclone (TC) destruction point of view. The major part of the devastation is attributed to seawater inundation over several kilometers inland due to storm surge. Thus, storm surge prediction with advanced lead time can contribute to minimizing damage and loss. Therefore, an attempt has been made to improve the storm surge prediction with a longer lead time using the high-resolution mesoscale model outputs. Weather Research Forecasting (WRF) and Hurricane WRF (HWRF) have been used to simulate 8 TCs over BoB and the model track and intensity both in terms of pressure drop and 10 m maximum wind speed have been provided as input to produce surge height using IIT Delhi dynamical storm surge model. The models’ reliability has been verified by analyzing different initial condition simulations of all TCs selected. 96–24 h forecast length with 12 h interval, prior to landfall have been used in this study. HWRF shows better overall predictability for the track, intensity as well as landfall; however, both models are capable of producing reliable results with 3–4 days of lead time. Using these model outputs in the surge model, the predicted surges for each TC are presented. The statistical analysis of the surge predictions using different intensity inputs shows that both models can generate reliable surge prediction when compared to the Indian National Centre for Ocean Information Services observed surge heights. The skill of the models in predicting storm surge shows that the HWRF wind input has the highest skill followed by ARW wind, HWRF pressure drop, and ARW pressure drop. Thus, this study suggests that the early warning of TCs by IMD should include the surge prediction from these highly reliable mesoscale model products with 96–72 h lead time in order to mitigate catastrophic loss associated with storm surge.