Abstract: In order to timely put out forest fires and improve the efficiency of forest fire monitoring, by using radar data, meteorological sensors of temperature, precipitation, relative humidity, wind speed and other information, through information layer, feature layers and data processing from decision makers, radar source information model of forest fire monitoring was developed. We used the model to investigate forest fires instance data from 2015 to 2017 in Yueyang. Results show forest fire probability based on radar level values of multi-source information fusion. The model was used for real-time monitoring of fires within the radar monitoring range of Yueyang in 2018. The model was found to efficiently monitor forest fires and the fire hit rate reached 83%.