Sensitivity analysis of the simulation of tropical cyclone Amphan using the WRF model

Abstract

The sensitivity of twenty-four physics option combinations in the Weather Research and Forecasting (WRF) model is investigated through simulations of the track, intensity, and landfall of tropical Cyclone Amphan. Simulation accuracy is evaluated using data from an international best track archive. Track performance is assessed using root mean square error (RMSE), while intensity is evaluated by comparing the minimum sea level pressure (MSLP) and the maximum wind speed measured at a height of 10 m above sea level. The results indicate that the combination of the New Tiedtke cumulus scheme, University of Washington planetary boundary layer scheme, Dudhia shortwave scheme, Rapid Radiative Transfer Model longwave scheme, and the WRF Single-Moment 6-Class microphysics scheme produces the most accurate track simulation, with an RMSE of 72 km, a mean absolute MSLP error of 27 hPa, and a mean absolute wind speed error of 22 ms⁻¹. Replacing the microphysics scheme with the Thompson microphysics scheme, while keeping the other physics options unchanged, results in a slightly higher track error but improves the landfall simulation, yielding the lowest landfall position error (43 km) and a time error of 1 hour and 32 minutes. These results indicate a trade-off between track and intensity accuracy and landfall prediction. Overall, the findings contribute to improved understanding of WRF physics configuration choices, support the development of more reliable cyclone forecasting frameworks, and highlight the need for better representation of intensity-related processes.