Bias-correction and dynamical downscaling strategy to improve the prediction of extreme weather events on extended range

This study aims to document the strengths and shortcomings of forecast strategy being developed based on dynamical downscaling for better extended range prediction of extreme weather events. One heavy rainfall event and two cyclone cases are selected to dynamically downscale the global forecasts of Extended Range Prediction (ERP) system using Weather Research and forecasting (WRF) model at 9km for Indian domain. The ERP outputs are bias-corrected and fed to WRF as lateral boundary conditions to minimize migrated errors from parent model via boundary conditions. Results show bias-corrected downscaled ERP is more efficient in predicting extreme weather with 10-12 days lead time.