A Real-Time Radar QPE Error Correction for Convective Precipitation Using Long-Term TRMM-PR and GPM-DPR Observations

Radar quantitative precipitation estimation (QPE) plays an important role in precipitation forecasting and flood early warning. Due to the vertical variability of precipitation, accurate radar QPE remains an ongoing challenge as the radar observations at high altitudes are typically used to estimate surface precipitation. The common underestimation error for convective precipitation has been recognized, but effective solutions for operational use are still lacking. In this study, a real-time vertical profile of reflectivity (VPR) correction algorithm is proposed to address this issue. The long-term observations of convective precipitation vertical structure over 24 years provided by space-borne radars (SRs), that is, the tropical rainfall measuring mission precipitation radar (TRMM-PR) and the global precipitation measurement mission dual-frequency precipitation radar (GPM-DPR), are used to generate climatological convective VPRs. The vertical structure variability of convective precipitation due to local precipitation characteristics, precipitation intensity, and environmental freezing level height is taken into account. In addition, the differences in frequency and sampling strategy between spaceborne radars and ground-based radars (GRs) are considered in the construction of the climatological convective VPRs, which can be used to correct radar QPE errors in real time. The VPR correction algorithm is evaluated using 28 typical convective precipitation events in China. The validation results show that the systematic underestimation error of the radar QPE for convective precipitation can be effectively reduced after the VPR correction. The relative mean bias (RMB), which is typically less than -0.3, is increased to between -0.2 and 0, and the root mean square error (RMSE) is significantly reduced.