Abstract
Forest fires are among the most hazardous natural phenomena, with their frequency and intensity increasing under current climate change conditions. In Kazakhstan, comprehensive spatial assessment of natural fire hazard remains insufficiently studied. This study proposes a multi-factor integrated model for assessing natural fire hazard in the Kostanay region based on remote sensing (RS) data and geographic information system (GIS) technologies. The model incorporates vegetation spectral indices (NDVI, NDMI, EVI), topographic parameters, and meteorological variables. All factors were normalized, and their weights were determined using an expert-based approach. Spatial analysis was performed in the ArcGIS environment. The results allowed for zoning the territory of the Kostanay region according to levels of natural fire hazard. It was found that a significant portion of the area falls within moderate-high and high-risk zones. The highest concentration of fire occurrences was observed in alluvial plains and areas dominated by pine forests. Verification results showed that more than 80 % of recorded fires are located within high and very high hazard zones. The findings can be used for wildfire risk assessment, identification of high-risk areas, and improvement of fire monitoring and management systems.