Abstract: A typhoon inverted trough rainstorm on August 9 to 10, 2012 in northeast Jiangsu is studied in this paper. The WRF mesoscalemodel is used for the numerical simulation of this storm. GTS observation data and GPS sounding profiles are assimilated in the initial field torun three groups of three-dimensional variational data assimilation experiments. Detailed analysis was carried out to the simulated data. Toobjectively analyze the impact of assimilating these data on precipitation in the experiments,TS score and the precipitation forecast deviationswere analyzed. The results showed that (1) assimilation of GTS observational data can improve the precipitation area, (2) the GPS soundingprofile assimilation does not effectively improve the simulation results, and (3) assimilating the above two kind of data can effectively improvethe forecast of precipitation area and intensity.