Anthropogenic grassland fire susceptibility dataset in Qinghai Province

Combined with NDVI, relative humidity, population density, distance from residential areas, distance from roads, slope, aspect, annual average temperature, human footprint index, distance from rivers and other data at a resolution of 1km in Qinghai Province from 2000 to 2020, a maximum entropy model is constructed to simulate and predict the susceptibility of man-made grassland fires in Qinghai Province.

To improve the spatial partitioning accuracy, the Kvamme gain statistical model was used to validate the partitioning accuracy simulated by the maximum entropy model. The AUC value of the model accuracy was 0.978, and the G value was close to 1, indicating that the constructed artificial grassland fire susceptibility evaluation model can effectively identify high-risk areas where artificial grassland fires occur. The dataset has a resolution of 1km and is in GeoTIFF file format, where 1 represents low susceptibility areas, 2 represents medium susceptibility areas, and 3 represents high susceptibility areas.