A new generation mesoscale WRF forecast model and its 3DVar assimilation system were used to conduct assimilation research of Doppler radar data on a squall line event on June 14, 2009 in Jiangsu province. Based on strict quality control on the radar data, a series of sensitivity experiments about optimized adjustment of scale factors and assimilation frequency were designed and carried out. It is shown that the optimized adjustment of scale factors and assimilation frequencies can lead to the better performances of initial conditions. Specifically, the appropriate scale factors setting can effectively improve the distribution characteristics of 700 hPa wind, 850 hPa temperature and composite reflectivity in initial conditions, then further improve the short-term precipitation forecast and vertical structure configuration of storm. Moreover, the higher assimilation frequency makes the distribution of composite reflectivity in initial conditions closer to the observation and makes short-term rainfall prediction more accurate.