Assessing Wildfire Risk at National Level Using a Pan-European Approach Under a Changing Climate

Wildfires are becoming an increasingly severe challenge, threatening ecosystems, economies, and communities worldwide. As climate change intensifies, understanding how wildfire risk changes across different landscapes is crucial for developing effective prevention and management strategies. This study presents an innovative approach to assessing wildfire susceptibility and hazard by combining machine learning with climate data at both a broad pan-European scale and detailed national levels. By first modeling wildfire susceptibility across Europe, this research provides a large-scale perspective that captures regional climate and environmental variability. This pan-European model then serves as the foundation for more refined national-level assessments, improving the accuracy of wildfire hazard predictions. The study introduces a structured 12-class hazard framework, integrating climate data and vegetation types to better capture wildfire risk under different conditions. The results prove how key climate variables—such as temperature, humidity, and wind speed—play a crucial role in shaping fire risk, though their influence varies across different countries. By integrating advanced machine learning techniques with high-resolution climate data, this research provides a valuable tool for understanding wildfire risk. The framework developed here offers insights that can support efforts to mitigate fire damage and improve long-term resilience in fire-prone regions.