Estimating changes in streamflow attributable to wildfire in multiple watersheds using a semi-distributed watershed model

Abstract

More than half of water supply in the western United States is sourced from forested lands that are increasingly under wildfire risk. Studies have begun to isolate the effects of wildfire on streamflow, but they have typically used coarse temporal resolutions that cannot account for the numerous, interconnected watershed processes that control the responses to rainfall events. In this study, we employed a method to isolate fine-scale (daily) effects of fire. Wildfire effects were estimated as the difference between measured post-fire streamflow and unburned scenarios of post-fire streamflow simulated by a hydrologic model calibrated to pre-fire conditions. The method was applied to track hydrologic recovery after wildfires in six burned watersheds across the western United States: North Eagle Creek, NM (2012 Little Bear Fire), Lopez Creek, CA (1985 Las Pilitas Fire), City Creek, Devil Canyon Creek, East Twin Creek, and Plunge Creek, CA (2003 Old Fire). All six watersheds experienced prolonged increases of post-fire streamflow, with the most consistent changes occurring during periods of low streamflow. Following 6 years of increased streamflow, Lopez Creek experienced 6 years of reduced streamflow before returning to the pre-fire hydrologic regime. North Eagle Creek and the four watersheds affected by the Old Fire continued to have elevated streamflow 9 and 18 years post-fire, respectively, without returning to the pre-fire hydrologic regime.