An innovative approach to Z_DR data assimilation using an ensemble Kalman filter: a proof-of-concept study

Differential reflectivity (Z_DR) observations from dual-polarimetric radars are directly related to the mean diameter of raindrops. To leverage this one-to-one relationship, the present study develops a mean diameter update (MDU) approach based on a local ensemble transform Kalman filter radar data assimilation system for Z_DR assimilation, which enables the explicit updating of the mass-weighted mean diameter (D_m) of raindrops. A series of assimilation experiments using both pseudo and real radar observations is conducted to examine the feasibility of the MDU approach. The results of the scalar assimilation experiment indicate that explicitly updating D_m by the assimilated pseudo Z_DR observation further enhances the accuracy of the D_m analysis. The single-pseudo-observation assimilation experiment reveals that the MDU approach effectively leverages the strong correlation between simulated Z_DR and D_m, and more corrections are propagated to the grid points near the pseudo Z_DR observation. In the observation system simulation experiments (OSSEs) for the Mei-Yu front case, assimilating pseudo Z_DR observations with the MDU approach reduces analysis errors of simulated Z_DR and rainwater state variables in each cycle. The forecast performance is the most favorable in terms of the forecast skill score and the rainfall probability when the forecast is initiated with more accurate microphysical states obtained by the MDU approach. Regarding the real-observation experiments, the analysis and forecast results are consistent with those of the OSSEs. The application of the MDU approach to both pseudo and real observations confirms its ability to exploit the intrinsic relationship between Z_DR and D_m, improving the accuracy of analyses and forecasts.