Using Hybrid variational-ensemble data assimilation method directly assimilate both Microwave Temperature Sounder (MWTS) data and Microwave Humidity Sounder (MWHS) data of FY-3A satellite in the case of the historical Beijing "7.21" extreme precipitation event in 2012, and compared to the three-dimensional variational assimilation method. Results show that both assimilation methods can improve the simulated precipitation results,but the Hybrid variational-ensemble data assimilation method is better than the three-dimensional variational assimilation method,which can effectively reduce the false heavy rainfall, and improve the simulation of the location of heavy rainfall center. The result of the SAL quantitative verification method for precipitation forecast shows the same conclusion. Comparing the root mean square error between the initial field and the reanalysis data obtained those two methods, the results of the the Hybrid variational-ensemble data assimilation method is smaller than the three-dimensional variational assimilation method,which means the analysis field of the Hybrid variational-ensemble data assimilation method is closer to real situation than that of the three-dimensional variational assimilation method. Both assimilation methods can change the physical information in the initial field, but the increments produced by those two methods are different in both magnitude and distribution. Three-dimensional variational assimilation method affects larger region in the simulation area for wind field,temperature field and relative humidity field than that of the Hybrid variational-ensemble data assimilation method. Increments obtained by the three-dimensional variational assimilation method are nearly homogeneous and isotropic,while increments obtained by the Hybrid variational-ensemble data assimilation method are inhomogeneous,anisotropic and "flow-dependent". Hybrid variational-ensemble data assimilation method makes the distribution of the initial field more reasonable, and it can improve the simulation of precipitation.