Divergent biophysical responses of western United States forests to wildfire driven by eco-climatic gradients

Abstract

Understanding vegetation recovery after fire is critical for predicting vegetation-mediated ecological dynamics in future climates. However, information characterizing vegetation recovery patterns after fire and their determinants over large geographical extents is limited. This study uses Moderate Resolution Imaging Spectroradiometer (MODIS) leaf area index (LAI) and albedo to characterize patterns of post-fire biophysical dynamics across the western United States (US) and further examines the influence of topo-climatic variables on the recovery of LAI and albedo at two different time horizons, 10 and 20 years post-fire, using a random forest model. Recovery patterns were derived for all wildfires that occurred between 1986 and 2017 across seven forest types and 21 level III ecoregions of the western US. We found differences in the characteristic trajectories of post-fire vegetation recovery across forest types and eco-climatic settings. In some forest types, LAI had recovered to only 60 %–70 % of the pre-fire levels by 25 years after the fire, while it recovered to 120 %–150 % of the pre-fire levels in other forest types, with higher absolute post-fire changes observed in forest types and ecoregions that had a higher initial pre-fire LAI. Our random forest results showed very little influence of fire severity on the recovery of both summer LAI and albedo at both post-fire time horizons. Post-fire vegetation recovery was most strongly controlled by elevation, with faster rates of recovery at lower elevations. Similarly, annual precipitation and average summer temperature had significant impacts on the post-fire recovery of vegetation. Full recovery was seldom observed when annual precipitation was less than 500 mm and average summer temperature was above the optimal range, i.e., 15–20 °C. Climate influences, particularly annual precipitation, were a major driver of post-fire summer albedo change through its impact on ecological succession. This study provides quantitative measures of primary controls that could be used to improve the modeling of ecosystem dynamics post-fire.