A Human Behavior Wildfire Ignition Probability Index for Application to Mainland Portugal

Wildfire ignitions are often linked to environmental and climatic factors, but human behavior plays a critical role, particularly in rural southern Europe. However, tools to quantify the probability of human-caused ignitions are lacking. This study addresses this by developing a human behavior wildfire ignition probability index focused on mainland Portugal, a region historically vulnerable to wildfires. Statistical analyses, including multicollinearity checks and a Generalized Linear Model, were used to analyze ignition data, while geospatial analyses estimated the ignition probabilities for 2021 and 2022. Inputs included human activity indicators, land use types, and proximity to residential roads. The resulting probability maps identified high-risk areas, particularly in forested zones and near residential roads. These maps closely aligned with documented human-caused ignitions, confirming the model’s reliability. The index is a robust tool for identifying high-risk areas and has significant potential to improve fire prevention strategies by targeting the most vulnerable regions. Future research should explore its integration into forecasting systems for real-time fire prevention and response strategies as well as its adaptation to other regions with similar wildfire risks.