Evaluation of Two Microphysics Schemes in Operational CMA-MESO for Simulating a Cold Vortex-Induced Heavy Rainfall

Comparative evaluation of cloud microphysical schemes has always been a focal point of research. Existing cloud microphysical schemes still face uncertainties. This study focuses on a cold vortex-induced heavy rainfall event in Northeast China in July 2023, using the China Meteorological Administration Mesoscale Model (CMA-MESO) operational model to conduct comparative retrospective experiments with the single-moment microphysics scheme WSM6 (namely the Weather Research and Forecasting model Single Moment 6-class scheme) and the double-moment scheme called LIUMA. The goal is to evaluate the forecasting capabilities of the schemes for cold vortex-induced heavy rainfall. The results show that both schemes could generally reproduce this heavy rainfall event; however, the LIUMA scheme’s precipitation time series exhibited a correlation coefficient of 0.75 and a root mean square error (RMSE) of 0.67 mm h⁻¹, which are closer to observations compared to WSM6’s values of 0.70 and 1.15 mm h⁻¹, respectively. It is found that WSM6 overestimated precipitation intensity more significantly. Additionally, the raindrop size distributions (RSDs) from LIUMA are more consistent with observations. The net latent heat peak in WSM6 is 2.2 × 10⁻⁴ K s⁻¹, higher than that of LIUMA of 2.0 × 10⁻⁴ K s⁻¹. The stronger latent heating in WSM6 enhanced dynamic effects, leading to more vigorous convection and resulting in overestimated precipitation intensity. The mixing ratios of ice-phase and liquid-phase hydrometeors simulated by LIUMA are higher than those by WSM6, with overestimation of ice-phase particles being particularly pronounced. This may be one of the reasons why LIUMA simulated radar reflectivity is significantly stronger than the observations. The main differences between the two schemes lie in the representation of ice-phase processes and the interactions between ice-phase and liquid-phase particles, and a more detailed evaluation of these processes will require advanced cloud observation techniques. The results obtained from this study help better understand the operational CMA-MESO model performance in forecasting heavy rainfall and benefit its further improvement.