The future of snow is in space: using ICESat-2 data to better understand the controls of terrain parameters on snow distribution at the watershed scale

Abstract:

Seasonal snowpack accounts for ~70% of the water supply in the western United States, and measuring snow accumulation and ablation remotely has long been a stated goal of NASA. Aerial and terrestrial Lidar systems are promising tools for the mapping and study of seasonal snowpack; Lidar is precise, has dense spatial coverage, and can map sub-meter changes in elevation due to snow accumulation even in regions with vegetation cover. The 2018 launch of ICESat-2, a spaceborne Lidar system, has the potential to enable much more spatially extensive lidar mapping of seasonal snow than previously possible when paired with snow-free surface elevation data. However, using ICESat-2 to map seasonal snow accumulation in the middle latitudes is difficult because ICESat-2 doesn’t have exact spatial repeat coverage outside of the polar regions. In preliminary work, ICESat-2 data products also struggle in steep and highly varied terrain characteristic of much of the mountainous regions in the middle latitudes. Here we explore the use of data from ICESat-2 to measure seasonal snowpack in two watersheds in southwest Idaho: Dry Creek and Reynolds Creek. Seasonal snow depths are synthesized with terrain parameters calculated from existing Lidar-derived digital terrain models (DTMs) of the watersheds in order to better understand controls on the spatial distribution of snowpack. Accumulation and ablation measurements from the ICESat-2 ATL08 Land and Vegetation Height product are compared to in situ and aerial lidar data. While the in situ and aerial lidar measurements are more precise and accurate, they do not span as diverse terrain as the ICESat-2 data, limiting the applicability of terrain parameters derived from these data to more expansive regions.