Sensitivity of physical schemes on simulation of severe cyclones over Bay of Bengal using WRF-ARW model

Gauging appropriate physical parameterization schemes for any numerical weather prediction model is indispensable for obtaining high accuracy in tropical cyclone forecasting. In this study, combinations of five microphysics, three cumulus convection, and two planetary boundary layer (PBL) schemes are investigated with respect to track, intensity, and time of landfall to determine an optimal combination of physical schemes of the weather research and forecasting (WRF) model (version 4.0) with advanced research WRF (ARW) core. All sensitivity experiments are carried out by taking the initial and boundary conditionsfrom the National Centers for Environmental Prediction Global Forecast System (NCEP-GFS). The simulated track, intensity, and landfall time are compared with the Indian Meteorological Department (IMD) observations. The sensitivity experiments reveal that the KF cumulus is performing better with YSU PBL along with WSM6, Ferrier (new eta), and Thompson microphysics for the track (position and time), and intensity with the least errors. Furthermore, we examined the performance of the model with the above combination of schemes on four severe landfalling cyclones (Bulbul, Hudhud, Aila, and Sidr). The root mean square error (RMSE) for central pressure gives the least value in the range of 0.4 to 8 hPa and 0.2 to 3.7 ms −1 for maximum surface wind (MSW) during landfall with YSU-KF- Ferrier combination. The equivalent potential temperature shows strong vertical mixing up to 500 hPa in the case of YSU-KF-Ferrier, which enhances the formation of warm-core, which further explains the intensity of cyclones. Overall, the track, intensity, and rainfall forecasts for the extreme cyclones considered in this study are consistent with IMD observations using YSU PBL, KF cumulus convection, and Ferrier microphysics.