Regional trends in snowmelt timing for the western United States throughout the MODIS era

• Published in: Physical geography 2022-05, Vol.43 (3), p.285-307
• Main Authors: O'Leary, Donal S., Hall, Dorothy K., DiGirolamo, Nicolo E., Riggs, George A.
• Format: Article
• Language: English
• Subjects: drought; MODIS; phenology; remote sensing; Snowmelt; spring
• ISSN: 0272-3646; 1930-0557
• DOI: 10.1080/02723646.2020.1854418

Snowmelt controls important physical and ecological processes and is widely expected to occur earlier under most climate change scenarios. The western United States (US) is dependent on seasonal snowpack for water resources and recreation, and this diverse landscape is likely to experience continued changes to snowmelt timing that will differ across the domain. In this study, we use NASA's cloud gap-filled snow-cover maps (MOD10A1F and MYD10A1F) to detect trends in snowmelt timing for hydrologic years 2001-2018. We find that 7.04% of the snow zone (i.e., areas with 10+ years of snow throughout the study period) of the western US is experiencing statistically significant (α = 0.05) trends of earlier snowmelt, while 2.62% of the snow zone is experiencing significant trends of later snowmelt. Analyses at the ecoregion level reveal regional trends, with many southwestern ecoregions experiencing large areas of dramatically earlier snowmelt (e.g., 19.84% of the Mojave Basin and Range). Interestingly, the North Cascades and the Northern Rockies ecoregions have substantial areas of later snowmelt (5.45% and 4.99%, respectively). Our work builds upon previous estimates of snowmelt timing to identify an overall trend of earlier snowmelt while highlighting the high spatial variability in snowmelt trends throughout the western US.