Investigation of Post-fire Evapotranspiration Rate Changes and Recovery using Remotely-Sensed SSEBop Data at the CONUS Scale

Abstract:

Evapotranspiration (ET) accounts for a significant portion of regional water budgets in much of the fire-prone western United States. Even small changes to ET rates can translate to meaningful shifts in runoff patterns and makes forecasting the direction and magnitude of wildfire-induced ET alteration of critical importance. We use 1-km surface energy balance ET depth estimates from the Operational Simplified Surface Energy Balance model for the conterminous United States (CONUS) to evaluate post-fire changes in burned areas from 2006 to 2015, including approximately 130,000 pixels and 5,000 individual fires. K-means clustering is used to group pixels by response in terms of percent change in the ratio of evaporation to precipitation (ETP), and independent topographic, locational, land cover, geologic, and fire-specific variables are studied to identify inter-cluster patterns for the first five post-fire years. Relationships between ETP modification to independent variables are tested using correlation coefficients and ANOVAs. The southwestern United States exhibited the largest actual post-fire reductions in ETP. Burned areas in the northwest saw similar negative percent changes but had smaller pre-fire ETP. The largest decreases in ETP were correlated with higher burn severities and greater percentages of pre-fire shrub/scrub cover and less deciduous and evergreen cover. Smaller ETP changes occurred in pixels with more deciduous and evergreen forest and wetland and lower burn severities. Intra-fire ETP response variability was also high in the western United States, with the largest ETP reductions occurring in areas where the percent burned and burn severities were highest. Eight percent of the pixels evaluated exhibited increased ETP in the first two post-fire years, and all clusters approached pre-fire ETP over the five-year post-fire study period. Results from this work highlight the variability in spatial and temporal patterns in post-fire ETP and will help inform downstream water managers of their vulnerability to wildfire-induced changes to water budgets.