Beyond average: a new approach to calculating fire regime departures applied to Western United States forests

Background: Changes in climate and vegetation, in combination with fire exclusion, are altering and homogenizing fire regime attributes compared to historical conditions. Fire regime changes are commonly quantified using departure metrics based on differences in measures of central tendency (i.e., means) between time periods. These metrics can mischaracterize complex changes to fire regime attributes because the distributions underlying these attributes are often not well described by parameters. Results: We developed a non-parametric index of fire regime departure that quantifies distributional changes to fire regime attributes between time periods using the Earth Mover’s Distance. We used this departure metric to compare fire frequency and burn severity between historical (~1600–1880) and contemporary (1985–2021) time periods in western US forests. In addition, we compared the proposed metrics with a standard suite of measures of central tendency. Departure metrics based on measures of central tendency reported lower relative departures within frequent fire forests and higher relative departures within infrequent fire forests than the EMD-based method. We found that 89% of western US forests are experiencing less frequent and more severe wildfires than historical baselines. Large departures are associated with increased human land-use intensity, and landscapes prioritized by the Wildfire Crisis Mitigation plan are on average, more departed than non-priority landscapes. Conclusions: This proposed method captures facets of fire regime departures that metrics based on measures of central tendency cannot. These new metrics can aid the evaluation and targeting of treatments to restore historical fire regimes and manage the resilience of fire-prone landscapes.