Forest fire dataset for European mountain regions (2000–2025)

This dataset supports the analyses presented in the article “Climate-driven upward spread of forest fires in European mountain regions” by Mirela Beloiu, Tomoki Loeillot, Verena C. Griess, Dimitris Poursanidis, and Fanny Petibon, published in Nature Communications. It integrates a range of spatial, climatic, ecological, and anthropogenic data to assess changes in forest fire activity across eight major European mountain regions between 2000 and 2025, with a particular focus on elevation shifts and their climatic drivers. Integrated Data Sources and Their Use 1. Forest Fire Perimeters Source: European Forest Fire Information System (EFFIS) Period: 2001–2025 Spatial Resolution: ~250 m (MODIS-based) with Sentinel-2 contributions in recent years Filtering Criteria: Fires ≥30 ha, maximum elevation of the fire ≥800 m, with ≥50% forest cover (based on CORINE Land Cover) Use: Core dataset for quantifying fire occurrence, size, and elevation. Spatial polygons were used to extract biophysical and climatic characteristics, and to derive three fire subsets: D800: Fires occurring above 800 m with ≥50% forest cover (n = 6,255) Z90: Fires above the 90th percentile of annual fire elevation T10: Top ten highest-elevation fires per year 2. Vegetation and Fuel Types Source: CORINE Land Cover (Kosztra et al., 2019) Use: Assigned dominant fuel types to fire polygons: coniferous, broadleaf, mixed forest, or transitional woodland/shrub. Fuel type was used as a categorical predictor in models and for exploring fuel-specific fire behavior by elevation. 3. Elevation Data Source: Copernicus Digital Elevation Model (DEM) Resolution: 100 m Use: Determined elevation of fire polygons (min, max, mean) and fire elevation trends. Also used to delineate high-elevation thresholds and quantify elevation shifts over time. 4. Climate Variables Source: TerraClimate (Abatzoglou et al., 2018) Resolution: 4 km, resampled to 1 km Period: 2001–2024 Variables Used: Actual Evapotranspiration (AET) Precipitation (PPT) Soil Moisture (SM) Wind Speed (WS) Vapor Pressure Deficit (VPD) Temperature (T) Climatic Water Deficit (DEF) Potential Evapotranspiration (PET) Palmer Drought Severity Index (PDSI) Use: Climate anomalies were calculated by comparing annual values to the 2001–2022 mean. These anomalies were extracted within fire perimeters and included in statistical models to identify climate drivers of burned area and fire elevation. 5. Human Pressure / Ignition Proxy Source: Global Human Footprint (Mu et al., 2022) Resolution: 1 km Years Used: 2000, 2018 Use: Used as a static proxy for ignition pressure within fire perimeters. Included as a covariate to control for human influence in fire models. 6. Biogeographic and Mountain Boundaries Source: Kapos et al. (2000); European Environmental Agency Use: Delineated the eight major mountain regions analyzed: Iberian Mountains, Pyrenees, Apennines, French and Swiss middle mountains, Alps, Carpathians, Balkans, and Turkish mountains. These were used for stratified analysis and to account for regional fire regimes in models. Data Applications Quantifying trends in fire elevation and high-elevation fire expansion. Identifying a threshold (~1400 m) where fire behavior shifts from fuel- to climate-limited. Modeling the effects of vegetation, elevation, climate anomalies, and human footprint on fire size and frequency. Classifying mountain regions based on fire frequency and elevation persistence (e.g., rare vs. reoccurring high-elevation fires). Included in the Repository Cleaned and filtered fire datasets: 20260309_fires_6225_71var.xlsx -with 71 observations and 6,255 fires (the Excel version contains one page with the explanation of each column name), 20260309_fires70var_5608_no0101.xlsx - 5,608 fires - it does not contain the fires on 1st January from 2000-2005. Shapefile with the forest fires > 800 m: fires_6225_71var.shp Mountain regions group from Europe: Mt_regions_grp.shp R scripts for data processing, statistical modeling, and visualization of figures 1, 2, 3, and 4, as well as the supplementary materials 5. Documentation (README) and metadata