Study on characteristics of heavy precipitation and model prediction correction under typical topography in Guizhou

Based on the hourly ground precipitation observations from the national and regional stations in Guizhou Province from 2016 to 2020, the spatial and temporal characteristics of heavy precipitation in the typical terrains of Guizhou were analyzed in this study. For typical terrains, the Beipan River was selected as the typical representative of Guizhou's valley terrain, while Leigong Mountain and Fanjing Mountain were chosen as the typical representatives of Guizhou's high mountain terrain. The three-hour precipitation forecast datasets during 2019-2020 provided by the European Centre for Medium-Range Weather Forecasts (ECMWF) were used, and a frequency matching method was applied to objectively correct the heavy precipitation forecast in the above three regions. The results are as follows. (1) Heavy precipitation in the three regions mostly occurs in June. The heavy precipitation frequency increases in Beipan River and Leigong Mountain year by year, while the increase in Fanjing Mountain is not significant. The heavy precipitation in the two mountainous regions mainly occurs during the night and early morning. While two heavy precipitation peaks are found in the Beipan River basin, one in the afternoon and one at night. (2) The distribution of heavy precipitation in the three areas is somewhat related to the terrain. The frequent and concentrated areas of heavy precipitation in the Beipan River are mainly located in the low-lying areas in the northern part of the basin, with that being in the south of Leigong Mountain and on the eastern side of Fanjing Mountain. Considering the slope of the terrain, the distribution of rainfall shows similar characteristics. In the regions of Beipan River and Fanjing Mountain with slopes less than 20°-30°, and the regions in Leigong Mountain with slopes less than 40°-50°, the frequency of heavy precipitation increases with the slope. While in other regions, the frequency of heavy precipitation decreases with the slope. (3) Objective corrections of heavy precipitation in these regions are applied using the frequency matching method. It is found that the frequency adjustment method has a certain effect in reducing the over-forecast of precipitation when the precipitation amount is less than 8.2 mm in Beipan River, less than 10.3 mm in Leigong Mountain, and less than 7.3 mm in Fanjing Mountain. When precipitation is greater than the above thresholds, the frequency matching method can significantly improve the levels of the predicted precipitation. After the correction, the overall performance of heavy precipitation prediction in the three regions has improved. A noticeable increase in the frequency of heavy precipitation prediction and comparable spatial distributions between the predictions and observations can be found, with the best improvement observed in the Beipan River.