Skip to Main Content
U.S. Forest Service
Caring for the land and serving people

United States Department of Agriculture

Home > Search > Publication Information

  1. Share via EmailShare on FacebookShare on LinkedInShare on Twitter
    Dislike this pubLike this pub
    Author(s): Gregory K. Dillon; Matthew H. Panunto; Brett Davis; Penelope Morgan; Donovan S. Birch; William M. Jolly
    Date: 2020
    Source: Gen. Tech. Rep. RMRS-GTR-415. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 107 p.
    Publication Series: General Technical Report (GTR)
    Station: Rocky Mountain Research Station
    PDF: Download Publication  (18.0 MB)


    Burn severity is the ecological change resulting from wildland fires. It is often mapped by using prefire and postfire satellite imagery and classified as low, moderate, or high. Areas burned with high severity are of particular concern to land managers and others because postfire vegetation, soil, and other important ecosystem components can be highly altered. In this study, we developed Random Forest statistical models describing the occurrence of high burn severity across the contiguous United States. We divided our work into 17 regions in the western United States and 8 regions in the eastern United States, and further subdivided them by forest and nonforest vegetation settings, resulting in 50 separate models. Our predictor variables represented prefire vegetation, topography, and fuel moisture, and our response variable was satellite-derived burn severity from the U.S. interagency Monitoring Trends in Burn Severity program. We created statistical models from 6,663 fires in the western United States (1984–2007) and 5,295 fires in the eastern United States (2000–2013). We then mapped our model predictions as a Severe Fire Potential (SFP) index ranging from 0 to 100. Our Random Forest models performed reasonably well (area under the receiver operating characteristic curve [AUC] = 0.71–0.93), though independent validation of our spatial predictions was less promising (AUC = 0.50–0.79). We analyzed spatial distributions of our SFP predictions, as well as satellite-based burn severity observations, across mapping regions and broad landform classes. We present observations about the spatial distribution of high severity fire, the variables influencing burn severity, and suggestions for applications of our SFP data that we believe managers and scientists interested in burn severity will find useful. This is the first national effort to understand where and why fires burn severely. Our consistent set of burn severity maps, field observations, and predictor variables together move us toward a better understanding of the “ecological footprint” of fires in forests and rangelands across the contiguous United States.

    Publication Notes

    • You may send email to to request a hard copy of this publication.
    • (Please specify exactly which publication you are requesting and your mailing address.)
    • We recommend that you also print this page and attach it to the printout of the article, to retain the full citation information.
    • This article was written and prepared by U.S. Government employees on official time, and is therefore in the public domain.


    Dillon, Gregory K.; Panunto, Matthew H.; Davis, Brett; Morgan, Penelope; Birch, Donovan S.; Jolly, William M. 2020. Development of a Severe Fire Potential map for the contiguous United States. Gen. Tech. Rep. RMRS-GTR-415. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 107 p.


    burn severity, fire severity, fire ecology, landscape, remote sensing, RdNBR, dNBR, Composite Burn Index

    Related Search

    XML: View XML
Show More
Show Fewer
Jump to Top of Page