Interannual Variability of Snowpack and Spring Season Hydroclimatology in the Southwestern United States

Temperature projections for the Southwest United States suggest that the region’s already marginal mountain snowpack will decline significantly during this century. A decrease in snowpack would affect land surface-atmosphere interactions as mediated by post-snowmelt changes in spring soil moisture and the partitioning of turbulent surface fluxes of latent and sensible heat. In this study, we examine observed interannual variability of snowpack and related hydrologic variables in the new NASA North American Land Data Assimilation System Phase 2 (NLDAS-2) dataset, to infer the changes in surface energy and water budgets that might be expected from future snowpack decreases. NLDAS-2 uses observations from multiple sources to drive several land surface models at 1/8 degree resolution; we use output from the Mosaic land surface model. Results from trend and variability analysis of 1979-2009 NLDAS-2 data show that post-snowmelt soil moisture in the mountainous southwestern U.S. depends principally on spring snowpack and surface temperature, rather than on total seasonal precipitation. We diagnose surface energy budgets for low-snow and high-snow years. We find that the impacts of decreased snowpack persist into the warm season through the partitioning of sensible and latent heat fluxes, such that the Bowen Ratio increases substantially in the spring season following low-snow years. In years of high snow, upward latent heat flux between April and mid-June is 5 to 30 W/m2 higher than in years of low snow. Understanding how the effects of snowpack decline persist into the warm season through soil moisture and turbulent fluxes has implications for warm season precipitation prediction, through altering the availability of local surface moisture for evaporation and through proposed land surface-mediated monsoon feedbacks.