Fire Regimes are conceptually useful to land managers and are qualitatively understood, but few quantitative techniques exist for empirically delineating geographic regions whose wildfire spatial and temporal characteristics, re-visitation frequency, and intensities are similar. We used thermal “hotspot” data collected globally by the two MODIS sensors during their 17-year orbital history. These data were used in an unsupervised multivariate geographic clustering process on a parallel supercomputer in order to statistically produce a quantitative discrimination of different Fire Regimes globally, including identification of similar regimes across hemispheres. We included both human-caused fires and wildfires, classifying both types of Fire Regimes empirically.
To appropriately address opposing seasonal juxtaposition across northern and southern hemispheres, we developed a special transformation of fire dates, based on latitude and temporal proximity to solstices and equinoxes, which allows statistical discrimination of, say, “summer” fires, regardless of the calendar month or hemisphere in which they occurred. This date transform permits recognition of similar fire seasonality in both northern and southern hemispheres. Representation of day-of-year as sine/cosine pairs allows the clustering algorithm to recognize burn dates that are seasonally grouped, even when they bridge the end of the calendar year.
Using 21 hotspot characteristics describing within-year seasonality, across-year return frequency, size and intensity, we produced global maps statistically discriminating the planet's most-different 10, 20, 50, 100, 500, 1000 and 3000 global Fire Regimes. Using principal component analysis to produce statistical colors, we also visualized the degree of similarity and graphically identified similar and different fire characteristics.
Results/Conclusions
Geographically distant locations which share similar Fire Regime characteristics were found, including Fire Regimes spanning across hemispheres. Regularly occurring human-caused agricultural Fire Regimes were also identified globally. Mirrored symmetrical latitude patterns are visible in each hemisphere, but latitude alone is insufficient alone to explain global Fire Regime patterns. Pure primary statistical colors, which show within-year seasonality, are primarily found in temperate zones, but mixtures of primary colors are seen in the torrid zone, where fire seasonality is less marked. Fire Regimes having two distinct annual peaks of fire frequency were the most common globally, followed by areas having 3 peaks per year. Bi-modal Fire Regimes typically have fire occurrence peaks both before and following the growing or rainy monsoon season. Locations sharing similar global Fire Regimes may have similar ecological effects and impacts from fire, and similar management knowledge and successful adaptation strategies might be borrowed, shared, or adopted.