Constrained by atmospheric internal variability and complexity of precipitation processes, current numerical models demonstrate limited skill for subseasonal precipitation. This study develops a U-Net-based correction model for subseasonal rainfall in summer over China, incorporating precipitation, atmospheric circulation variables, and topographic features from the regional Climate–Weather Research and Forecasting (CWRF) model. Results indicate that the U-Net model systematically corrects biases over 6 mm day^-1 on average, extending prediction skill from the lead time of 2 pentads (10–14 days) to 5 pentads (25–29 days). The U-Net model provides substantial forecast improvement in Central China, South China, and Southwest China while maintaining the high prediction skills of the CWRF model over North China. Interpretability analysis further demonstrates topography, 200-hPa zonal wind and 500-hPa geopotential height (500-hPa temperature) collectively contribute over 45%, with underlying mechanism also discussed. This study confirms the value of a hybrid dynamical–statistical approach in enhancing regional model prediction performance, providing a novel methodology for subseasonal precipitation prediction.

An unprecedented persistent heavy precipitation occurred in Henan province during 17—22 July 2021, causing huge economic losses. Currently, extreme precipitation forecasting is still a hotspot and a difficult issue in sub-seasonal climate prediction research. Regional climate models provide a new way to further improve sub-seasonal precipitation forecasting in China with finer spatial resolution and better parameterization of physical processes compared to that of the global models. This study uses the regional Climate-Weather Research and Forecasting model (CWRF) nested with the China Meteorological Administration Climate Prediction System version 3 (CMA_CPSv3) to improve prediction capabilities for this persistent heavy precipitation event. It is shown that the spatial distribution, magnitude, and forecast accuracy of precipitation predicted by CWRF are improved compared to that predicted by CMA_CPSv3. Although both models underestimate the amount of precipitation, the CWRF forecasts larger accumulated precipitation and spatial distribution of precipitation is more consistent with observation. CWRF forecasts initialized on 26 June and 29 June are better than that of CMA_CPSv3 on the same initial dates. The CWRF significantly improves the forecast of low-level wind fields and low-level jets in East Asia compared with the CMA_CPSv3. The CWRF is particularly effective in improving the simulation of directions of low-level jets and water vapor fluxes, allowing water vapor to converge on the windward slopes of mountain ranges and providing favorable water vapor conditions for precipitation. The CWRF better forecasts the water vapor flux convergence and ascending motions over Zhengzhou, and all these improvements lead to higher precipitation forecasting skill of CWRF.