In order to reveal the deviation characteristics of severe convective storm and mesoscale convective system (MCS, in satellite infrared channel), we employ the conventional observation and the unconventional observation (radar and satellite data) to analyze the simulation performance of CMA-MESO model in simulating the extremely severe rainstorm that occurred in Henan Province from 19 to 21 July 2021 by using traditional and new spatial verified methods. Besides, we also investigate the causes for the model deviations from the perspectives of water vapor, momentum, trigger and maintenance mechanism of precipitation. The results show that, firstly, the model can well capture the shape of rain band and duration of weak echoes as well as the evolution trend of intensity and area of MCS in the early and late stages of the primary precipitation process. Secondly, the model deviations mainly lie in that the model underestimated the intensity of precipitation, failing to predict the extreme hourly precipitation at Zhengzhou Station and the evolution trend of hourly precipitation of main rain band. Besides, it seriously underestimated the durations of convective storms and severe convective storms. Furthermore, the model missed the sharply increasing of MCS area in the afternoon and the simulated MCS was displaced westward and northward relative to the observation. Thirdly, the model deviation was mainly owing to the incorrect simulation of water vapor. The vertical distribution of simulated water vapor was not reasonable, and the simulated water vapor transports from Typhoon In-Fa and Typhoon Cempaka were both weak. In fact, the weak low-level jet and insufficient ultra low-level easterly jet pulsation directly led to the insufficient transport of simulated water vapor. Moreover, the model’s failure in forecasting the stably maintaining mesoscale convergence line on the surface near Zhengzhou Station, in conjunction with the deficiency of atmospheric instability and underestimation of unstable convective available potential energy, made the development of the simulated convection not strong enough and finally caused the inadequate simulated precipitation.