Abstract: Based on reanalysis data from a global numerical model of National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR),data from European Centre for Medium-Range Weather Forecasts (EC) and Shanghai Meteorological Bureau-WRF ADAS Rapid Refresh System (SMB-WARMS),and other conventional meteorological observational data,the forecast deviation of extreme precipitation event for a warm-region in Hubei province on July 5-6,2016 was analyzed.The results showed that the low-level wind field predicted by the EC model moves westward faster than the real situation,resulting in westward biases in the predicted heavy precipitation area.The forecast deviation in precipitation intensity and falling area during this event is mainly due to lack of awareness of the dynamic triggering effect of the shear line at 925 hPa and its influence on the dry layer in the upper troposphere,the rapid increase of water vapor and its extremity,and the evolution of Mesoscale Convective System (MCS).Although there is a deviation between the 72 h,48 h,and 24 h precipitation forecast and the real precipitation,the 12 h precipitation area in the SMB-WARMS model can be adjusted in advance,which can be used as a good indicator to the modification of precipitation prediction.