A Simple Framework for Quantifying Warming-based Snowpack Declines at the Landscape Scale

Thursday, 18 December 2014
Christopher Tennant1, Benjamin T Crosby1, Sarah Godsey2, Robert Van Kirk3 and DeWayne Derryberry4, (1)Idaho State University, Geosciences, Pocatello, ID, United States, (2)Idaho State University, Geosciences, Idaho Falls, ID, United States, (3)Humboldt State University, Mathematics, Arcata, CA, United States, (4)Idaho State University, Mathematics, Pocatello, ID, United States
The amount and timing of water release from snowpack sets the potential and pace for human and ecological systems residing in or near snowy topography. Warming will likely increase the snowline elevation and reduce the amount of the landscape that develops a significant snowpack. Because mountain catchments are topographically complex, understanding their sensitivity to warming and snowpack loss remains challenging. Previous research indicates that elevation is a dominant control on broad-scale snow accumulation. Thus, at the intermediate watershed scale (50 km2 < drainage area < 500 km2), a catchment’s elevation distribution should be a robust predictor of its susceptibility to snowpack loss. We present a simple framework that captures the climate and elevation-based sensitivity of mountain catchments to warming-driven snowpack loss. Framework development is based on the characterization of ~3,200 watershed elevation distributions in the northern Rocky Mountains, U.S. and is coupled with estimates of increase in snow water equivalent (SWE) with elevation for a variety of climatic types. Potential catchment-wide snowpack loss is modeled across a range of warming scenarios using a hypsometric approach and a sigmoidal function that can be easily adjusted to reflect warming-based alterations in how SWE increases with elevation. Monte Carlo simulations and regression analyses for 20,000 synthetic elevation distributions suggest that location, scale, and shape parameters that quantify distribution characteristics can be used to model SWE loss. Our model shows that similar amounts and patterns of SWE loss can occur even when catchments are centered in different elevation bands, suggesting that simple multi-parameter estimates of sensitivity may prove useful. We identify critical thresholds where different elevation distribution types exhibit unique responses to warming and discuss possible implications for mountain forests and carbon and nitrogen cycling.