The fire situation during the dry season of Thailand, in the last 10 years, has become more severe. The Tad Sung Forest Park area has reported the intensity of wildfires for the past 7 years. This research has applied the geographic weighted regression (GWR) model to generate a spatial relationship analysis model for wildfires. This research aims to create a spatial model to analyze the risk of hazardous areas against wildfire and to analyze the factors that affect forest fire risks in order to protect against wildfires. The service area (SA_LY) model was obtained through the first approach. The wildfire-GWR results of the study showed that the model can predict at the R² level greater than 82% and varies according to the sub-area boundaries. Factors affecting the acceleration of wildfires are (positive coefficient) the digital elevation model (DEM), normalized burn ratio (NBR), land surface temperature (LST) and (negative coefficient) normalized difference vegetation index (NDVI), slope and aspect. In addition, the distance from the road factor has little effect on wildfire intensity in most areas. The results of the research are used to create a risk-sensitive map of wildfires through surveillance by importing the independent variable factors in the model and using it as a prototype of the same kind of space.