Influence of cloud microphysical schemes on CMA-GD model prediction of a warm-sector heavy rainfall in South China

Abstract

This work evaluates the influence of cloud microphysical schemes on prediction of a warm-sector heavy rainfall with an operational modeling system of CMA-GD in Guangzhou Regional Meteorological Center (GRMC). The heavy rainfall is produced by a MCS occurred over Pearl River Delta in South China on May 21, 2020. Four cloud microphysical schemes (including WSM6, WDM6, THOMP and LIUMA) are investigated to understand their impacts on structure and evolution of rainfall system. Results show that the WSM6 over-predicts the 24-hour accumulated rainfall, while the other three schemes underestimate the rainfall. In general, these deviations of total rainfall are mainly caused by generated rainfall during mature stage of MCS. Four schemes all underestimate precipitation during this period, but the deviation is the least in WSM6 and WDM6 schemes. As far as both WSM6 and WDM6 schemes, quantitative verification shows that the threat score (TS) and the false alarm ratio (FAR) as well as the proportion of convective/stratiform precipitation in the WSM6 single-moment scheme are superior to those in the WDM6 double-moment scheme. Investigation of vertical distributions of precipitation particles and the associated thermodynamic response in the environment shows that compared with THOMP and LIUMA, WSM6 and WDM6 simulate more ice and snow in the upper level and more rain and cloud water in the low troposphere as MCS evolves into mature state. With the top-heavier heating and the strongest upward motion lasting longer time, WSM6 may lead to stronger dynamical feedback to large-scale environment compared with THOMP. All these reveal that WSM6 is the most accurate scheme simulating this warm-sector rainstorm and the importance of more accurate simulation on the evolution and structure of precipitation.