Abstract: Increasing attention has been directed toward the impact of the uncertainty of the physical processes on prediction accuracy. This study utilizes the Weather Research and Forecasting Model (WRF v4.4) model, and it considers a rainstorm in South China from May 21 to 22, 2020, as an example to quantify the uncertainty among the microphysics process, boundary layer process, and cumulus convection process, which are closely related to precipitation. Further, it aimed to compare the discrepancy among different parameterization schemes in the same physical process using multivariate analysis of the variance method and Tukey’s test. Furthermore, the impact of the differences in microphysics schemes on the precipitation simulation error was analyzed in detail. The results show that the microphysics scheme is the most significant for precipitation simulation and model prediction, and the interactions between different physical processes cannot be overlooked. On this basis, the optimal parameterization scheme combination, which is determined through the Taylor skill score, is selected as the WSM7 scheme for the microphysics process + the YSU scheme for the boundary layer process + the GF scheme for the cumulus convection process. From the perspective of hydrometeors and microphysics conversion processes in different microphysics schemes, hail plays a significant role in the simulation of extreme precipitation. The change in raindrops mixing ratio may be attributed to the melting term of ice particles. The evaporation of rain affects the intensity of the cold pool through latent heat absorption, and it then affects the subsequent precipitation propagation. Furthermore, it directly affects the distribution of heavy precipitation.