Abstract: Previous studies show that the cloud microphysical process affects the precipitation, as well as the intensity, internal structure and evolution process of tropical cyclones. Thus, the rational description of the cloud microphysical process is crucial. And the correctness of the cloud microphysical process is the basis of the high-resolution model in simulating precipitation and fine-scale structure of typhoon. The two-moment bulk microphysics scheme developed by Chinese Academy of Meteorological Sciences (CAMS) is a mixed phase two-moment cloud microphysics scheme, which can simulate cloud microphysics processes in different weather systems. However, whether it can be applied to the simulation of tropical cyclones is still uncertain.Four numerical experiments of typhoon Usagi (2013) are conducted by using the Weather Research and Forecasting (WRF) model with Chinese Academy of Meteorological Sciences two-moment microphysics scheme (CAMS). The simulated track, intensity, cloud microphysics and rainfall are compared with the observed typhoon best track dataset and satellite observations to evaluate performances of CAMS microphysics scheme and investigate the possible impacts of cloud microphysical processes on Typhoon Usagi. To overcome the overestimation of snow content in control experiment (CTRL), three sensitivity experiments are designed:Modifying coefficients of snow particle mass and falling velocity (EXP1), using the typical oceanic cloud droplet parameter (EXP2), and including changes in both EXP1 and EXP2 (EXP3). It shows that the snow content is significantly reduced in the EXP1 and EXP3 due to the increased rate of accretion of snow by graupel and the slightly reduced snow mass flux, and the content of whole ice-phase hydrometeors are also reduced. The rapid intensification process in the early stage of typhoon Usagi is well captured in EXP2 and EXP3 owing to the better simulated CAPE in eye region, and their intensity and track are also better than those in CTRL. Although the hourly precipitation rate in each experiment is generally stronger, the spatial distribution of precipitation in EXP3 is more consistent with the observation. As a result, modifying coefficients of snow mass and falling velocity as well as using the typical oceanic cloud droplet concentration in CAMS microphysics will significantly reduce the snow content and improve the simulated track, intensity and precipitation. These results could not only provide ideology for improving the cloud microphysical parameterizations in simulating typhoon, but also improve the understanding of cloud microphysics impacts on typhoon process and help improving the cloud microphysical parameterization schemes in simulating typhoons.