Impact of different microphysics and cumulus parameterizations in WRF for heavy rainfall simulations in the central segment of the Tianshan Mountains, China

Abstract

With the continuous deepening of hydrological simulations in the alpine regions, high-resolution rainfall data is urgently needed as driving data for distributed hydrological models. Therefore, the objective of this study is to evaluate the performance of the WRF model for the accumulated rainfall simulations in the central segment of the Tianshan Mountains. The WRF model is configured with triple nesting of 27, 9, and 3 km for 28 experimental setups using four microphysics schemes (Morrison, WSM6, Goddard, and Thompson) and seven cumulus convection schemes (Kain-Fritsch, Betts-Miller-Janjic, Grell-Freitas, Grell-3, KF-CuP, New SAS, and Grell-Dévényi). The performance of these WRF configurations for two typical heavy rainfall simulations are first assessed via comparisons between simulation and observation; then, its influence on the simulations is analyzed. The results show that 1) There are significant differences in the rainfall area simulated by 28 combinations; 2) The 28 experimental setups show that higher snow crystal content is consistent with more rainfalls. Only the WSM6 possesses the mechanism to adjust snow and ice content with the temperature, serving as the most suitable microphysics scheme for rainfall simulation; 3) From the perspective of the accumulated rainfall and its large value distribution, the advantage of the Grell-3 possesses the mechanism for the settlement effect extended to adjacent grids, making it the most suitable cumulus convection scheme for the simulation. Overall, the WRF model presents a strong capacity for rainfall simulation in complex terrain. Statistical investigations on various WRF setups identify suitable microphysics and cumulus convection schemes to predict heavy rainfall in the Tianshan Mountains accurately.