Simulating Long-Term, Inter-annual Fire Regime Variability in the Continental United States: Applications to Current and Future Conditions

Previous research has shown that wildfire occurrences, at sufficient spatial and temporal scales, exhibit Self-Organizing Criticality (SOC) behavior from which modeled frequency-area curves can be derived. The resulting curves have been previously employed to show variations in fire regimes at the ecoregion scale in the Contiguous United States (CONUS). With a contemporary, 24-year wildfire event database, we model frequency-area curves and transform them to cumulative probability distributions to simulate annual inventories of wildfire events across the CONUS in time, space, and by fire size. We propose techniques for distributing these events within each ecoregion based on geographic and biophysical criteria. We explore the application of these simulations to modeling possible future fire emissions scenarios in a photochemical grid model (PGM), as well as examining temporal changes in the modeled curves, using successive portions of the database, that may be useful for projecting changes in fire activity by ecoregion in the near future. Challenges and potential limitations to these applications are explored.