Study on Drought Prediction System and Its Application in Northern China Based on Regional Numerical Model

North China is affected by steep terrain, complex circulation and strong land-atmosphere interaction, resulting in frequent droughts. In recent years, extreme drought events in this region have increased under the background of global warming, threatening the sustainable development of social economy. In order to promote the prediction level of extreme drought events, it is urgent to improve the simulation performance of regional numerical models for drought factors such as precipitation and temperature. Therefore, the localization and improvement of RegCM5.0 are studied in this paper, and the drought prediction system of RegCM5.0 is developed. The improved scheme includes updating the surface type data, modifying the thermal conductivity parameterization model and optimizing the cumulus convection scheme. The results show that the bias of precipitation is reduced by 50% and the 2m air temperature by 1℃. The correlation coefficient of precipitation, 2m air temperature is increased from 0.61 to 0.72 (α = 0.001) and from 0.83 to 0.88 separately. The correlation coefficients of drought index such as SPI and SPEI can reach above 0.50 (α = 0.05). Case test shows that the prediction system accurately forecast the extreme drought events in summer of 2020 and extreme heat and drought events in 2024. The study has important value for improving the early warning level of extreme drought events in northern China.