Based on satellite observations and the community radiative transfer model (CRTM), this paper evaluates the simulation results of three cloud microphysical schemes (Morrison, Thompson, and WD6) aiming at a heavy precipitation event in the middle reach of Yangtze River. The heavy precipitation once occurred from 29 June to 1 July 2016 and was brought by the clouds moving eastward from the Tibetan Plateau. The results show that compared to the high-resolution fused satellite precipitation observations, the three cloud microphysical schemes perform very well as a whole. They all successfully simulate the beginning, developing and the maturing stages of the precipitation event, of which the performance of Morrison scheme is the best. However, the three schemes have greater uncertainties in simulating the solid hydrometeor. Relative to the ERA5 cloud cover data and the brightness temperature observation from Himawari-8, the three schemes overestimate the high cloud cover but underestimate the mid-low cloud cover. Compared to the satellite-retrieved cloud products, it is found that the downward shortwave radiations simulated by the three schemes are smaller in the precipitation area due to the smaller contents of cloud ice and cloud water as well as the larger cloud effective radius. Furthermore, the initial field water vapor used in WRF model simulation is less than the GPS atmospheric precipitable water in the precipitation area. The error of extreme precipitation simulation is the result of the combined effects of thermodynamics, dynamics, and cloud physics. These findings could serve as a reference for the evaluation and improvement of extreme precipitation simulation.