The characteristics of the mesoscale numerical forecasting system developed by the Linyi Meteorological Bureau, the interpretation of the model forecast products, and the application of radar data in the model system are introduced. The evaluation of forecasting results was conducted from June to October 2005. The system displays considerable capability in shortrange weather forecast for predicting precipitation and temperature. The system is considered to be valuable for improving the quantification, timeliness, and accuracy of weather forecasting. By using the selfdeveloped SDAF software package in the application of forecasting products, the methods for forecasting heavy rainfall, air quality, extreme temperature, dryhot winds, etc. are generated. The Doppler radar echo intensity is used to modify the humidity field in numerical model, and the forecasting results are improved obviously. The primary experiments on the retrieval of the wind vector fields from the Doppler radial velocity to modify the initial forecast fields are carried out, and some conclusions are obtained.