Can panel studies help untangle the effects of wildfires on low flows in the western United States?
Abstract
The recent coincidence of increased wildfires with severe drought in the western United States has sparked discussions regarding the effects of wildfires on water supply, including at sites without long streamflow records. Recent studies have focused on generalizing post-fire streamflow responses at individual sites and developing cross-sectional (regional) models that explain post-fire responses at these sites as a function of basin characteristics over fixed time periods. In contrast, panel studies, which simultaneously consider variations in hydrologic responses over time and across sites, can incorporate very short streamflow records spanning disparate periods. However, such panel studies have not been previously designed to analyze the effects of wildfires on low flows.
Due to a relative paucity of gauged basins with substantial areas burned at or above moderate severity, we first develop a panel model that predicts annual time series of low-flow metrics not subject to wildfire impacts. We then evaluate the extent to which two extensions of this model can predict post-wildfire low-flow responses in ungauged locations: (1) one that evaluates predictions of post-fire low-flow metrics using post-fire vegetation indicators, such as the Relative Differenced Normalized Burn Ratio (RdNBR), and (2) one that predicts the likelihood of observing low-flow increases or decreases at various post-wildfire time intervals after adjusting for weather differences between pre and post-wildfire periods. We compare our findings with recent studies that have identified regions in the western US where wildfires have tended to increase and decrease low flows. Finally, we discuss the benefits and challenges of our panel study design.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFMGC52D..04H