EVALUATING THE INFLUENCE OF VEGETATION PARAMETERS IN A SNOW ENERGY AND MASS BALANCE MODEL IN WILDFIRE IMPACTED LANDSCAPES

Abstract

High-elevation landscapes that accumulate deep, persistent snow have been increasingly impacted by larger and more intense wildfires in recent decades. Following wildfire, the snow mass and energy balance is altered, leading to changes in peak snow water equivalent (SWE), faster snow melt rates, and earlier snow disappearance dates. Despite the added complexity that post-fire areas present for water forecasting, few studies have assessed snow model performance in post-fire locations relative to in-situ observations. This study evaluates iSnobal, a widely-used physics-based model, over five winters following the 2020 Cameron Peak Fire burn scar (Colorado, USA) with a focus on vegetation parameters. I updated the LANDFIRE vegetation dataset used in the model setup to reflect post-fire conditions and tested the model sensitivity to each individual parameter. Snow disappearance dates improved by up to 45 days relative to the default run, bringing modeled dates within a median of -2 days (± 5 days) across stations and years. However, average daily net radiation differences of up to 47 W/m2 (113% of observed value) were found between updated model runs and measured radiation at weather stations in burned areas. I also used published vegetation burn severity thematic classes to assign vegetation parameters, which led to improved agreement in High and Moderate burn severity forested areas later in the melt season. Finally, modeled a set of endmember 20-year post-fire vegetation recovery scenarios, which demonstrated limited SWE changes between excellent and poor recovery, highlighting the long-term, transformative impact of these disturbances on snow processes. This work emphasizes the importance of vegetation parameters in iSnobal and the need for accurate parameterizations for post-fire snow water resource modeling.