Direct Assimilation of Dual-polarization Radar Using the Hydrometeor Background Error Covariance in the CMA-MESO Model and its Sensitivity to the Microphysics Scheme

Numerical models play an important role in convective-scale forecasting, and dual-polarization radar observations can provide detailed microphysical data. In this study, we implement a direct assimilation operator for dual-polarization radar data using the hydrometeor background error covariance (HBEC) in the China Meteorological Administration MESO-scale weather forecasting system (CMA-MESO, formerly GRAPES-MESO) and conducted assimilation and forecasting experiments with X-band and S-band dual-polarization radar data on two cases. The results indicate that the direct assimilation of dual-polarization radar data enhanced the microphysical fields and the thermodynamic structure of convective systems to some extent based on the HBEC, thereby improving precipitation forecasts. Among the sensitivity tests of microphysical parameterization schemes, including the LIUMA scheme, the THOMPSON scheme, and the WSM6 scheme (WRF Single-Moment 6-class), we find that the greatest improvement in the equivalent potential temperature, relative humidity, wind, and accumulated precipitation forecasts occurred in the experiment using the WSM6 scheme, as the distribution of solid precipitation particles was closer to the hydrometeor classification algorithm from the dual-polarization radar observations in our cases.