The study focused on improving the accuracy of weather forecasts for hydrological purposes in the Beijiang River Basin. Researchers used a method called ensemble pro-processor to enhance the Global Ensemble Forecast System and Climate Forecast System version 2. By evaluating different post-processing schemes, they found that considering cumulative forecast precipitation and incorporating observed data led to more accurate predictions. The integrated forecast combining GEFS and CFSv2 after post-processing was significantly better for medium- and long-term forecasts. This research demonstrates how post-processing techniques can enhance the accuracy of hydrological forecasting in various time scales.