Abstract: A cold cloud system dominated by stratoclouds from January 5 to 7, 2017 is simulated by the GRAPES_Meso model using two cloud microphysics schemes (Morrison and WDM6) to investigate the impacts of different cloud microphysical schemes on the cloud microet alphysical process and cloud radiation process. The results indicate that the performance of Morrison double-moment cloud microphysics scheme including the warm and cold cloud physical processes is better than the WDM6 scheme. The simulated cloud water path, effective raet aldius of cloud ice and cloud optical thickness by the Morrison scheme are larger than those by the WDM6 scheme, while the simulated effecet altive radius of cloud water by Morrison is smaller than WDM6. The Morrison scheme is more accurate than WDM6 for simulating the effective radius of cloud water (cloud ice), cloud optical thickness, cloud water path, and the horizontal distribution of shortwave radiation on the ground. The effective radius of cloud water (cloud ice) simulated by the Morrison scheme is larger (smaller) than that by the WDM6 scheme. This leads to that the ratio of cloud water and snow simulated by the Morrison scheme is larger than that by the WDM6 scheme, while the raet altio of rainwater and cloud ice and hail simulated by Morrison is obviously smaller than that by WDM6. The shortwave radiation simulated by the Morrison scheme is smaller than that by the WDM6 scheme, more consistent with the CERES satellite retrievals. The different microphyset alical processes represented in the two schemes also affect the processes for latent and sensible heat at the surface, which results in that the simulated sea level temperature and radiative flux with the Morrison scheme are smaller than those with the WDM6 scheme. These impacts of the two different schemes are significantly greater over the land than over the ocean.