Abstract: Based on the background field of the AREM (Advanced Regional Eta-coordinate Model) IC (Initial Condition) analysis, Doppler radar data are integrated into the Local Analysis Prediction System (LAPS) for a heavy rainstorm event in Beijing on 21 July 2012. To study impacts of the radar data integration and cold or hot start on the heavy rainfall prediction of AREM model forecast, the difference of the initial fields and the difference of the simulation results under the cold and hot start are compared before and after the integration of Doppler radar data. The results show that:the initial field increases meso-scale information and vapor intensity after the radar data are ingested into the LAPS, and the vapor transportation in the east is also increased. Additionally, the center of the rain belt and the precipitation intensity forecasts are improved. Radar data integration and hot start can improve spin up of AREM model, and increase the upward current velocity, further improve the intensity of the precipitation.