A Novel Approach to Assimilate ZDR Observations with an Ensemble Kalman Filter Data Assimilation System

The mean diameter update (MDU) approach based on an ensemble-based radar data assimilation system is developed to assimilate differential reflectivity (Z_DR). This approach transforms the background variables to update, rainwater mixing ratio (q_r) and rainwater total number concentration (N_Tr), to mass-weight mean diameter (D_m) in order to generate more correction through the high correlation between simulated Z_DR and D_m. In addition, different variable transformation strategies are investigated to examine the performance of the MDU approach. A series of assimilation experiments are conducted with two different double-moment microphysics parameterization schemes for a real squall line case, in which intense reflectivity (Z_H) is characterized by numerous small raindrops. The results show that the MDU approach can further reduce the analysis errors of both Z_H and Z_DR at the latter cycles. The improvement of the Z_DR analysis errors corresponds to the negative D_m difference between the experiments with and without employing the MDU approach. Through the statistical analysis, it is found that the number of grid points with large (small) D_m decreases (increases). Without normalizing q_r, the MDU approach not only improves the Z_DR analysis but also maintains the intensity of the Z_H analysis. Besides, the performance of the quantitative precipitation forecast (QPF) is further improved with the implementation of the MDU approach. In conclusion, an appropriate transformation of the background variables to update can make better use of Z_DR observations to illustrate the analysis microphysical states and improve the accuracy of the QPF after assimilation.