Method and device for predicting lightning stroke forest fire danger based on data driving model

The invention discloses a lightning stroke forest fire danger prediction method based on a data-driven model, and the method comprises the steps: obtaining historical forest fire records of a designated region and a designated time period, screening out the historical forest fire records of fire caused by lightning, and taking the historical forest fire records as historical lightning stroke forest fire records; associating the three-dimensional lightning monitoring data, the meteorological element data, the satellite remote sensing data and the digital elevation data to historical lightning forest fire records; constructing a forest fire danger positive and negative sample set; screening input features; and constructing a gradient boosting decision tree model based on the screened input features and the forest fire danger positive and negative sample sets. The invention also discloses a prediction device. According to the method, multi-source data related to lightning forest fire danger is fused, and a decis