Snow exerts a primary control on global hydroclimate and serves as a critical water resource for over 1 billion people globally. Despite its importance to both human and natural systems, there are myriad uncertainties that complicate the representation of snow accumulation and melt processes in land surface models. For example, it is not well understood whether snowpack cold content—the energy deficit that must be satisfied before melt can begin—develops primarily as a function of new snowfall or a negative surface energy balance. I will present work using both ground observations and validated model output showing that new snowfall is the dominant source of cold content development at two sites in the Colorado Rocky Mountains. I will also discuss how the increased snowpack cold content of the higher, colder alpine site resists the effects of climate warming on snowmelt timing and rate. Another issue complicating model representation of snowpack processes is the treatment of precipitation phase partitioning. Many models use a spatially uniform threshold and/or range to discriminate between rain and snow, despite observations showing significant spatial variation in the air temperature at which rainfall and snowfall are equally likely. I will detail how relative humidity controls spatial variation in the rain-snow air temperature threshold and how the choice of a precipitation phase method introduces significant uncertainty to simulated snowfall fraction across the Northern Hemisphere. The final portion of my presentation will focus on how recent advances in remote sensing from NASA’s Airborne Snow Observatory enable improved validation of snow model output in the Tuolumne River Basin of California’s Sierra Nevada. I will discuss the issues involved with forcing a high-resolution snow model over a large simulation domain and how such issues can be resolved using information from both near-surface meteorological measurements and a gridded climate dataset. Ultimately, improved representation of snowpack processes across spatial scales will lead to better predictions of snowmelt rate and timing, which has implications for both water resources management and simulations of future hydroclimatic states.
Speaker: Keith S. Jennings, PhD Candidate, NASA Earth and Space Science Fellow, University of Colorado Boulder
Host: Erica Woodburn