Radar Reflectivity Assimilation Method Based on Adaptive Hydrometeor Phase Identification Algorithm Using Dual‐Polarization Radar and Its Impacts on Severe Weather Prediction

Abstract The indirect radar reflectivity assimilation method, which utilizes hydrometeor retrieved from radar reflectivity, is straightforward, computationally efficient, and beneficial for improving severe weather forecasting. However, this method is susceptible to retrieval errors due to inherent uncertainties when deriving multiple hydrometeor species from reflectivity alone. To address this, this study introduces a blended retrieval approach within the WRFDA framework, enhancing hydrometeor retrieval accuracy by incorporating dual‐polarization radar measurements. Specifically, the proposed method first estimates the individual hydrometeor contribution factors to the total reflectivity using two existing retrieval schemes. These initial estimates are then refined by integrating dominant hydrometeor species identified from dual‐polarization radar observations through a hydrometeor classification algorithm (HCA). This physically constrained optimization redistributes the hydrometeor contribution factors, leading to improved hydrometeor analyses for assimilation. Results from four convective cases demonstrate that the new scheme with the dual‐polarization‐based HCA produces more physically realistic hydrometeor analyses, leading to superior forecast accuracy. For the Anhui case, EXP_HCA (the experiment using the HCA‐restrained blending retrieval scheme) significantly improved the 3‐hr precipitation forecasts (25‐mm threshold), increasing the probability of detection and success ratio by approximately 0.09 and 0.035, respectively, over EXP_BD (the experiment using the background‐hydrometeors‐dependent retrieval method). In the Nantong gale case, EXP_HCA achieved a higher threat score (TS) for maximum surface wind (17.2 m s −1 threshold), surpassing EXP_T (the experiment using background‐temperature‐based retrieval method) and EXP_BD by 0.131 and 0.134. Additionally, the quantitative verification of the precipitation forecasts for two cases in Jiangxi provided further evidence of the effectiveness of the blending scheme.