A polygon-based modeling approach to assess exposure of resources and assets to wildfire

Authors

Matthew P. Thompson, USDA Forest ServiceFollow
Joe Scott, Pyrologix, LLC
Jeffrey D. Kaiden, University of Montana
Julie W. Gilbertson-Day, USDA Forest Service

Document Type

Article

Date of this Version

2013

Citation

Nat Hazards (2013) 67:627–644, DOI 10.1007/s11069-013-0593-2.

Comments

U.S. government work.

Abstract

Spatially explicit burn probability modeling is increasingly applied to assess wildfire risk and inform mitigation strategy development. Burn probabilities are typically expressed on a per-pixel basis, calculated as the number of times a pixel burns divided by the number of simulation iterations. Spatial intersection of highly valued resources and assets (HVRAs) with pixel-based burn probability estimates enables quantification of HVRA exposure to wildfire in terms of expected area burned. However, statistical expectations can mask variability in HVRA area burned across all simulated fires. We present an alternative, polygon-based formulation for deriving estimates of HVRA area burned. This effort enhances investigations into spatial patterns of fire occurrence and behavior by overlaying simulated fire perimeters with mapped HVRA polygons to estimate conditional distributions of HVRA area burned. This information can be especially useful for assessing risks where cumulative effects and the spatial pattern and extent of area burned influence HVRA response to fire. We illustrate our modeling approach and demonstrate application across real-world landscapes for two case studies: first, a comparative analysis of exposure and area burned across ten municipal watersheds on the Beaverhead- Deerlodge National Forest in Montana, USA, and second, fireshed delineation and exposure analysis of a geographically isolated and limited area of critical wildlife habitat on the Pike and San Isabel National Forests in Colorado, USA. We highlight how this information can be used to inform prioritization and mitigation decisions and can be used complementarily with more traditional pixel-based burn probability and fire intensity metrics in an expanded exposure analysis framework.