Verification and Improvement of Cloud Microphysics Parameterization Schemes in WRF Model for Simulation of a Deep Convection System

In this study, strong precipitation generated by the landing of typhoon Fung-Wong in 2008 was simulated using WRF WDM6 double-moment cloud microphysics scheme. The model simulations were evaluated by statistical methods using cloud top brightness temperature (TBB) and precipitation radar reflectivity data from MTSAT-1R, TRMM satellite, and the Satellite Data Simulator Unit. To reduce the difference between the simulation results and satellite observations, sensitivity experiments were conducted in relation to autoconversion rate of cloud water to rain, ice nucleation, terminal velocity, and slopes of snow and graupel. The results show that the precipitation, strong convective cloud system, and convective columnar radar echo simulated by the WDM6 scheme are in agreement with observations, however, showing stronger results in some areas. The WDM6 scheme simulation produced more shallow convective clouds and underestimated the frequency of convective cloud systems. The radar echoes simulated by different cloud types are all strong, and the vertical distribution of the convective cloud radar echoes is close to observation. The experimental results show that modifying the production rate for the autoconversion of cloud water to rain in the WDM6 scheme effectively improves the simulation results. Simultaneously, it is found that the concentration of initial cloud droplets affects the production rate for the autoconversion of cloud water to rain and ultimately affects the simulation results of the cloud structure and radar reflectivity. A high concentration of initial cloud droplets will worsen the simulation result.