Cycles of snow accumulation and melt are primarily driven by precipitation and temperature, and the impact of these drivers can be influenced by factors including elevation and canopy cover. While previous studies used limited point measurements or relied on model simulations, here we use the University of Arizona daily 4-km snow dataset over the contiguous United States from 1982 to 2017 to investigate snowpack trends and drivers. We explore 1 April snow water equivalent (SWE) as a function of October–March mean temperature and cumulative precipitation. We find that precipitation is more important than temperature in determining year-to-year SWE variability, and significant trends in SWE over Western CONUS are largely driven by temperature at low and mid elevations and by both precipitation and temperature at high elevations. Large differences in the SWE seasonal cycle between “forested” and “non-forested” pixels are caused by differences in precipitation and temperature associated with these pixels; the direct effect of tall versus short vegetation is small, in contradiction with prior modeling and point measurement-based studies that highlighted the effects of forest cover on both snow accumulation and melt. Further studies are needed to resolve this issue.
