Improving Radar QPE for Convective Precipitation Through Vertical Profile of K DP Correction Based on GPM‐DPR 3D DSD Retrievals

Abstract Vertical variation of convective precipitation causes significant errors in ground‐based radar quantitative precipitation estimation (QPE). To address this issue, we analyze 11 years (2014–2024) of vertical profiles of equivalent S‐band specific differential phase () and reflectivity (). is calculated from drop size distribution (DSD) parameters, while is converted from Ku‐band measurements from the global precipitation measurement mission dual‐frequency precipitation radar (GPM‐DPR), for consistency with operational S‐band radar measurements. A climatological vertical profile of (VPSDP) correction model is developed, incorporating precipitation intensity, environmental freezing level height, and sensor geometry differences. The model correction performance for ‐based QPE is evaluated using 25 convective precipitation events over eastern China in 2021, with comparison to the established vertical profile of reflectivity (VPR) correction model for ‐based QPE. Results show that the VPR correction improves ‐based QPE more substantially than the VPSDP correction improves ‐based QPE, due to the greater inherent vertical variation error in . Before correction, both ‐ and ‐based QPE data show systematic underestimation, with relative mean bias (RMB) values of −0.29 and −0.33, respectively. After correction, RMB values improves to −0.17 for both methods. The VPR correction decreases relative mean absolute error (RMAE) from 0.47 to 0.41 and root mean square error (RMSE) from 11.29 to 9.89 mm, while the VPSDP correction reduces RMAE from 0.44 to 0.39 and RMSE from 10.57 to 9.39 mm. Overall, VPSDP‐corrected QPE demonstrates higher accuracy than VPR‐corrected QPE, with the performance advantage increasing at higher rainfall intensity.