Harpold, Adrian A 1 ; Molotch, Noah P 2 ; Stamper, Brooke E 3
1 University of Colorado
2 University of Colorado
3 University of Colorado
Snowmelt is the primary water source for human needs and ecosystems services in much of the Western U.S. Regional warming is expected to hasten snowmelt disappearance by increasing snowmelt rates and reducing snow water equivalent (SWE). The soil water budget mediates the effects of changing snowmelt patterns on vegetation water availability and runoff, but soil moisture response is difficult to measure and predict. The goal of this research therefore, is to develop a framework for evaluating the sensitivity of soil water availability to changes in snowmelt across a mosaic of climate, vegetation, and soils characteristic of Western U.S. mountain ecosystems. In this presentation three overlapping analyses that address this goal will be described: 1. Variable snowmelt and soil moisture response across three mixed-conifer forests, 2. Soil moisture response across hundreds of sites with variable physiography, and 3. Quantifying the importance of snowmelt to soil water availability in the Northwest U.S. First, I will demonstrate the complex spatiotemporal patterns of snowmelt in three mixed-conifer Critical Zone Observatory (CZO) sites. Despite differential snowmelt patterns with respect to forest position across the sites, we show that the timing of peak soil moisture (PSM) response is nearly coincident with the day of snow disappearance (DSD). Second, we expand this analysis to cover 260 snow telemetry (SNOTEL) stations to investigate the synchronicity of snowmelt and soil moisture response. We found that that the DSD explained 47% of the variability in PSM timing, with 68% of station-years having PSM-DSD deviations within 2 weeks. Interestingly, station-years in the Cascade and Sierra Nevada ecoregions tended to have PSM preceding DSD by 3 to 7 days, which could not be explained by SWE accumulation or seasonal precipitation. As a final analysis, we explored a subset of 48 SNOTEL and Soil Climate Analysis Network (SCAN) stations with soil water retention information. Focusing in the Northwest U.S. we found that the timing of snow disappearance was a stronger predictor of the duration of water stress than annual or seasonal precipitation. Together, our results imply that earlier snow disappearance will lead to earlier peak soil moisture and increase the likelihood and duration of soil water stress late in the growing season. We observed little influence of post-snowmelt rainfall on reducing growing-season water stress, which does not bode well for predicted trends towards earlier snowmelt and more intense summer rainfall for much of the Western U.S. Our datasets provide motivation and geographic constraints on applying remote sensing measurements of snow disappearance (e.g. MODIS) as a surrogate for peak soil moisture timing. Future research opportunities exist to verify remote sensing soil moisture measurements (e.g. SMAP) and land-surface model soil moisture predictions with these datasets.