Satellite-Derived Wildfire Dynamics in Bulgaria (2001–2022): Trends, Climate, and Topographic Context

Abstract

Understanding wildfire dynamics and their climatic and topographic drivers is essential for effective fire management in Southeastern Europe. This study presents the first long-term assessment of wildfire trends across different land-cover types in Bulgaria, providing a methodological framework that is applicable across the wider Balkan region using the FireCCI51 burned area dataset (2001–2022). Burned area patterns were analyzed in relation to temperature, relative humidity, precipitation, the Ångström Index, Vapor Pressure Deficit (VPD), and topographic conditions. Burned area declined across most land-cover types, particularly croplands and forests. Croplands accounted for 96.4% of the total burned area (4.9 Mha), while forests, grasslands, shrublands, and wetlands represented 3.6%. Burned area was positively correlated with temperature (r = 0.37–0.53) and VPD (r = 0.40–0.53), and negatively with relative humidity (r = −0.48 to −0.55) and the Ångström Index, highlighting the importance of atmospheric dryness for wildfire activity. Seasonal climate-fire analyses further indicated that burned-area variability was more strongly associated with climatic conditions during the peak fire season than during the pre-fire season, particularly through lower relative humidity and higher vapor pressure deficit values. Topographic analysis revealed clear partitioning across land-cover classes. Cropland and wetland fires were concentrated in flat, low-elevation landscapes (mean altitudes ~117 m and ~40 m; slopes ~3.36% and ~2.04%), while forest and grassland fires occurred in higher, steeper terrain (mean altitudes ~351 m and ~315 m; slopes ~14.25% and ~9.76%). Shrubland occupied an intermediate niche (mean altitude ~287 m; slope ~9.43%), statistically overlapping with both forest and grassland categories, and acting as a transitional zone between lowland agricultural plains and mountainous ecosystems. Fire activity showed a clear seasonal pattern, with the highest burned areas occurring in August, coinciding with maximum VPD and minimum Angström index values. These results highlight the important role of atmospheric dryness and temperature variability in shaping wildfire activity in Bulgaria and demonstrate the usefulness of long-term satellite burned-area datasets for regional fire monitoring that can be applied not only in Bulgaria but also across the Balkan region.