多源融合降水产品在重庆强降水过程中的应用评估OA

This study evaluates the performance of two multi-source fusion precipitation products,ART_1 km and CLDAS,based on hourly precipitation data from 2144 surface meteorological stations and 31 heavy rainfall events in Chongqing in 2023.The results indicate that the ART_1 km product has a distinct advantage in monitoring precipitation processes,with a correlation coefficient as high as 0.990(95%CI:0.987-0.993)compared to observed values,which is significantly superior to the CLDAS product's 0.915(p＜0.01).The mean absolute error(MAE)and root mean square error(RMSE)of ART_1 km are 0.62 mm and 2.86 mm,which are 87%and 67%lower than those of CLDAS,respectively.For different intensity levels of precipitation,ART_1 km outperforms CLDAS in both threat score(TS)and bias(BIAS).In particular,for heavy rain and above,CLDAS shows a distinctly low TS score and obvious negative BIAS.Cross-regional comparisons reveal that in regions with a topographic complexity index(TCI)about 0.8,the ART_1km product's TS score for heavy rain only decreases by 0.02,while the CLDAS product decreases by 0.15,indicating that high-resolution products have a significant"compensation"effect on topographic errors.The case study of heavy rainfall events in Chongqing confirms that in observation-sparse valleys and karst trough valleys,the ART_1 km product can partially replace ground observations,providing a"space-for-precision"solution for similar terrains.Our research results suggest that in complex terrain regions,the ART_1 km product should be prioritized for heavy rain center localization and flash flood warnings;for observation blind spots,it is recommended to combine mobile rain gauges or X-band radar for sparse-dense collaborative observation.High-resolution multi-source fusion precipitation products are crucial for enhancing the responsiveness of disaster warning systems and the efficiency of disaster management.Therefore,it is essential to fully leverage their complementary advantages in meteorological monitoring and disaster warning to provide robust technical support for relevant decision-making.