Unravelling Global Fire Regimes: Seasonal Mapping and Multi-Variable Characterization of Fire Activity
Keywords: Fire Regimes, Spatio-Temporal Patterns, Clustering, Satellite Remote Sensing
Abstract. Wildfires are fundamental ecological processes that shape terrestrial ecosystems and influence atmospheric dynamics. This study presents a global assessment of fire frequency, intensity, and burned extent over a 24-year period (2001-2024) using multi-source satellite observations. Spatio-temporal variability of forest fire activity and the critical role of seasonality in modulating fire behaviour has been captured in a grid of 0.25°. Our results reveal clear regional and hemispheric patterns wherein southern and central Africa, northern Australia, and parts of South America consistently exhibit the highest fire activity, with strong spatial and seasonal variability. In contrast, the Northern Hemisphere remains relatively stable, with lower fire occurrence and limited changes over time, aside from modest seasonal fluctuations in certain regions. Seasonal dynamics are especially pronounced in the tropics, reflecting variations in climatic drivers and fuel availability. To explore interactions among fire parameters, we conduct grid-level correlation analyses, revealing strong positive associations among frequency, intensity, and extent in tropical and subtropical regions. These relationships weaken or decouple in temperate and boreal zones, highlighting the influence of seasonal climate and vegetation dynamics. Building on these insights, clustering-based classification was used to delineate global fire regimes based on the combined behavior of the three parameters. The resulting maps reveal distinct spatial configurations and temporal evolution, with dynamic regime shifts across much of the Southern Hemisphere and comparatively stable regimes in the north.
