The increasing severity of wildfires calls for data-driven regional strategies for prevention and early intervention. This paper presents a diagnostic approach that combines artificial intelligence (AI), satellite data (mainly from Copernicus/Sentinel), and machine learning to assess and forecast wildfire risk in Pleven Province, Bulgaria. The study focuses on two key dimensions of regional diagnostics: geographic location and natural landscape. It introduces a Potential Fire Risk Index (PFRI), built from satellite-based indicators such as NDVI, NDMI, Land Surface Temperature (LST), Fire Weather Index (FWI), and dry forest biomass density (DBDI). Using five years of data, the model identifies spatial patterns of vegetation stress and soil moisture deficit, providing early insights into high-risk zones. The findings highlight potential wildfire hotspots beyond the 2024 events and offer evidence-based recommendations for prevention policies, including AI- powered early warning systems. This study demonstrates how advanced geospatial technologies can support sustainable territorial governance in climate-sensitive regions.
Miroslav Mihaylov (Wed,) studied this question.