Evaluation of WRF Microphysics for Port-Specific Cyclone Forecasting over Bangladesh

Abstract

This study simulates cyclonic storms Hamoon and Remal using Weather Research and Forecasting (WRF) configuration to determine the most suitable microphysics schemes for early warning over seaports of Bangladesh. The North Indian Ocean region lacks localized, high-resolution forecasting, which affects port operations and coastal preparedness. WRF was configured with nested domains at 9km and 3km resolution using WSM6, WSM7, and SBU-YLin schemes for 168-hour and 96-hour simulations. Five-variables (surface wind, precipitation, sea-level pressure, temperature, and vertical relative humidity) were evaluated using MAE, MB, RMSE, and track errors. GFS provided boundary conditions, while validation used ERA5, GPM rainfall, BMD and BAF station data, and RSMC best track observations. Results show strong agreement between the simulations and observations, confirming the reliability of model. WSM6 and WSM7 performed best for Hamoon at Chattogram Port, whereas SBU-YLin produced better short-range forecasts for Remal at Mongla Port, where 96-hour wind RMSE reached 6.8kmh⁻¹ (Hamoon, WSM6) and 6.2 kmh⁻¹ (Remal, SBU-YLin). Therefore, 96-hour forecasts better capture track, wind, and pressure, while 168-hour runs sometimes better represent precipitation trends. Future work should incorporate additional physics, data assimilation, and more cyclone cases to strengthen localized-forecasting and enhance port safety and coastal resilience in Bangladesh.

The Dhaka University Journal of Earth and Environmental Sciences, Vol. 15(1), 2026, P 63-78