Abstract:
This study uses the global numerical weather forecast product from GFS (Global Forecast System) of NCEP (National Centers for Environmental Prediction) as the initial field for model forecasting. The regional mesoscale forecast system, China Meteorological Administration Mesoscale Model (CMA-MESO) version 5.1 (formerly GRAPES_MESO), is used to analyze a rainstorm event that occurred in the Sichuan Basin during 3–5 September 2021. Five sets of experiments are designed, using three different resolutions (1 km, 3 km, and 10 km) and two cloud microphysical parameterization schemes (WSM6 and Thompson) for numerical simulations. The results indicate that: (1) The simulated rain belt generally aligns with the actual situation, but discrepancies in the timing, area, and intensity of heavy rainfall are observed. As the precipitation threshold increases, the TS score decreases, whereas the bias amplitude increases, leading to higher rates of false and missed reports. (2) No significant difference is observed in the simulation results of water vapor flux between the same resolution cumulus parameterization scheme and different microphysical schemes. The simulated rainstorm regions across all five experiment groups exhibit strong updrafts, and the simulation intensity increases as the resolution increases. (3) The simulation results for liquid particles are similar across different cloud microphysics schemes at the 1-km resolution. However, the results for solid particles show notable differences. (4) When the cumulus parameterization scheme is applied at the 3-km resolution, a large deviation in the simulation results of the heavy precipitation center is found. Overall, the experimental simulation results of the precipitation process indicate that under high-resolution conditions, the Thompson scheme’s saturation adjustment scheme is slightly better than that of the WSM6 scheme. Moreover, the 1 km_thompson configuration provides the most accurate depiction of rain belts, yielding the best precipitation simulation.