Abstract: Based on the daily data of daily rainfall,precipitation forecast of 0~24 hours in the southwest center - wrf adas real-time modeling system(SWC-WARMS) and the intelligent grid forecast model of the Central Meteorological Observatory(NWGD) on middle and lower reaches of the Jinsha River from April to October 2019, we analyzed the cumulative monthly areal rainfall,the monthly extreme value distribution of daily areal rainfall and the characteristics of rainstorm frequency distribution. The results are as follows:(1) The monthly accumulated areal rainfall and heavy rainfall are mainly from June to September. (2) A total of 22 rainstorms occurred between April and October 2019. The area with the highest number of heavy rains is area A,and there is no heavy rain in area D.(3)The SWC-WARMS is better than the NWGD in prediction of monthly accumulated areal rainfall in area C,D and E,and in the prediction of light rain and moderate rain in area Aand C. (4)The Mean Absolute Error(Ea),Fuzzy Grading(MP) and Treat Score(TS) results show that Intelligent grid forecast model of the Central Meteorological Observatory better than the numerical forecast model of the southwest region to forecast the areal rainfall in the middle and lower reaches of Jinsha River as a whole.