Temperature and Vegetation Stress Drive Mediterranean Wildfire Risk More Than Drought: Spatial Analysis of Chlef, Algeria Using the ICRI-SA Composite Fire Risk Index

Semi-arid regions of the Mediterranean are at the forefront of climate change, facing accelerated warming and increased ecosystem stress that directly amplify wildfire risk. This study provides a comprehensive, multi-decadal (1990–2024) analysis of climate change in the Chlef wilaya, Algeria, and quantifies its relationship with the regional fire regime. We employ advanced time-series analysis, including Mann-Kendall tests and breakpoint detection, to identify significant climatic shifts. A suite of climate and drought indices (FWI, SPI, PDSI, Aridity Index) are calculated and their predictive power for fire occurrence is rigorously evaluated using machine learning models. Our analysis reveals a statistically significant warming trend of + 0.043°C per year (p < 0.001), culminating in a total increase of + 1.45°C over the period. A major climatic breakpoint was detected in 2019 , after which the mean annual temperature abruptly increased by + 1.38°C . While precipitation trends were not statistically significant, machine learning models identified mean annual temperature as the single most powerful predictor of fire activity (XGBoost Importance: 50.3% ), outperforming all complex drought indices. Further modeling incorporating vegetation stress (NDVI) and topography achieved near-perfect predictive accuracy (CV-AUC = 0.998), with NDVI being the dominant factor (52.2% importance). This research demonstrates that in this semi-arid system, long-term warming and its impact on vegetation health are more direct and potent drivers of fire risk than short-term precipitation deficits. The findings provide critical, data-driven evidence for adapting fire management strategies to a new, hotter climatic reality.