Effects of high elevation wildfires on peak snow water equivalent in California's Sierra Nevada headwater regions
Year: 2026
Type: article
Abstract: Increases in high‐elevation wildfires in California’s Sierra Nevada region (SNR) have raised concerns about potential impacts on snowpack, a major source for the state’s water supply. Quantifying fire-induced changes in snow storage remains challenging in mountainous regions with complex topography and large climatic and ecological gradients, due to limited field observations and process models. This study leveraged high-resolution Airborne Snow Observatory (ASO) lidar surveys (2023-2024) to assess the impacts of 2020-2021 wildfires on peak snow water equivalent (SWE) across six SNR watersheds. We assembled >32,000 matched burned-unburned 50-m pixel pairs with similar pre-fire environmental conditions and assessed how changes in peak SWE (ΔpeakSWE: burned - unburned) varied with burn severity, pre-fire forest structure, meteorological variables, and terrain above the snowline. Across matched pairs, post-fire peak SWE differences varied in both magnitude and direction, depending on yearly weather (e.g., wet vs. normal year) and local environmental setting. The largest positive burned–unburned difference occurred in dense forests with ≥75% canopy loss at intermediate elevations in 2023, where burned areas had 13.0 cm higher peak SWE on average (12% relative to the matched unburned mean of 105.1 cm); this difference was substantially smaller in 2024. The spatially cross-validated extreme gradient boosting (XGBoost) machine learning model showed terrain provided the strongest predictive information, followed by vegetation structure. SHapley Additive exPlanations (SHAP) and a shallow decision tree further showed that sites with lower heat exposure, locally lower terrain positions, and intermediate elevations generally were associated with higher peak SWE in burned areas compared to matched unburned areas, whereas taller pre-fire canopies and exposed, high-energy terrain were associated with lower peak SWE. These results demonstrate wildfire effects on peak SWE are strongly context dependent and provide an assessment framework for anticipating where high elevation wildfires may increase or reduce mountain snow storage.
Source: Environmental Research Letters
Language: English
Topic: Fire effects on ecosystems
Subfield: Global and Planetary Change
Field: Environmental Science
Domain: Physical Sciences
Keywords: snowpack, burn severity, canopy loss, XGBoost, SHAP
SDGs: Climate action, Life on land
Open Access status: gold
Estimated APC paid: $3,045
Award: G9001368683
OpenAlex ID: W7226005822
