Russell A. Parsons1, Robert E. Keane1, and Matthew G. Rollins1

1Present Address: USDA Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory, P.O. Box 8089, Missoula, MT 59807. E-mail: rparsons@fs.fed.us Phone: (406) 329-4872 Fax: (406) 329-4877

Wildland fire management requires fire regime maps that accurately characterize historical frequency, severity, and pattern of fire at multiple scales. Coarse scale fire regime maps provide national scale context for planning and allocation of resources, while mid and fine scale fire regime maps provide the greater detail needed to implement landscape treatments that mimic historical disturbance regimes. Three main strategies have been used to develop predictive fire regime maps. Knowledge-based strategies construct rules to classify fire regimes based on topography and vegetation. Statistical strategies build predictive surfaces from fire history data and other spatial data. Simulation strategies use landscape dynamics simulation models to simulate fire spread and vegetation change over long periods; fire regime maps are then summarized over the entire simulated temporal range. We illustrate the application of each strategy with an example landscape in the Rocky Mountains, compare the outputs, and address some of the potential pitfalls of each approach with a focused sensitivity analysis on each strategy. Based on our results, we present some guidelines as to which strategy is most appropriate under different scenarios. We conclude with recommendations for an improved method that integrates elements of the different strategies and with which we expect that fire regimes can consistently and accurately be mapped over large areas and at multiple scales.