An improved method for dynamical extended-range forecasting of persistent severe rainfall based on the coupling of MPAS-A and regional models

Abstract

Persistent severe rainfall (PSR) events are highly hazardous and cause significant disasters. This study focuses on extended-range forecasting of PSR. On the basis of the regional Weather Research and Forecasting model (WRF), the coupling of the Model for Prediction Across Scales-Atmosphere (MPAS-A) and the Dynamic Extended Forecasting framework for WRF (referred to as MPAS-EXT) was developed, enabling high-precision dynamic extended forecasts up to 20 days in advance. The study simulates the PSR event from June 18 to 21, 2022. Results show that the MPAS-A model provides a more accurate simulation of the large-scale weather systems during this PSR event, and the 500 hPa weather pattern simulation is more conducive to the occurrence of the PSR event. Regarding the location and intensity of the PSR event, MPAS-EXT provides more accurate simulations for both the rainfall location and intensity. Specifically, the model shows superior performance in simulating areas with precipitation exceeding 50 mm. Moreover, MPAS-EXT offers a more comprehensive and clear depiction of the rainband structure of the entire event, the simulations of water vapor flux and wind fields are closer to the results of ERA5. Through Threat Score (TS) and Fractions Skill Score (FSS) evaluations, MPAS-EXT has higher TS than the other experimental groups for 0–100 mm precipitation, for precipitation >50 mm, the FSS values increase as the neighborhood radius expands. Calculations of Root Mean Square Error (RMSE) and Total Energy Error (TEE) further reveal that MPAS-EXT reduces bias accumulation during the integration and minimizes total environmental energy errors.