Item: ASSIMILATING LIDAR AND IN-SITU COMMUNITY SNOW MEASUREMENTS TO IMPROVE SNOWPACK ESTIMATES ACROSS THE WESTERN UNITED STATES
-
-
Title: ASSIMILATING LIDAR AND IN-SITU COMMUNITY SNOW MEASUREMENTS TO IMPROVE SNOWPACK ESTIMATES ACROSS THE WESTERN UNITED STATES
Proceedings: Proceedings, International Snow Science Workshop, Whistler, BC, Canada, 2026
Authors:
- David F. Hill [ Oregon State University, Corvallis, OR, USA ]
- Zachary Butler [ Oregon State University, Corvallis, OR, USA ]
- Katreen Wikstrom Jones [ Alaska Department of Geological & Geophysical Surveys, Fairbanks, AK, USA ]
- Gabriel Wolken [ Alaska Department of Geological & Geophysical Surveys, Fairbanks, AK, USA ] [ University of Alaska Fairbanks, Fairbanks, AK, USA ]
Date: 2026-09-28
Abstract: Seasonal snowpack supplies a large fraction of water resources across the Western United States, making accurate estimates of snow distribution and melt timing critical for water resource management. However, substantial uncertainty persists due to sparse observations and model limitations, particularly in complex terrain. This study integrates high-resolution airborne light detection and ranging (lidar) data with in-situ observations from SNOTEL, Snow Courses, and the Community Snow Observations project (CSO) to improve snowpack estimation across three critical basins: Lake Tahoe, California; the Central Cascades, Oregon; and Summit County, Colorado. Using SnowModel simulations for water years 2022–2024, we evaluate the impact of different data assimilation strategies. This includes spatially distributed data (lidar-derived snow depth) only, point-based observations (SNOTELs, Snow Courses, and CSO) only, and combined approaches, on model performance in representing snowpack distribution. We further assess how lidar acquisition frequency and the spatial resolution of the input lidar influence model skill to identify minimum data requirements for reliable snowpack estimation. Results indicate that assimilating a snow depth raster at 50 meters yields the most accurate representation of snowpack across all study regions. However, comparable performance is achieved using only point-based observations, a resampled lower resolution snow depth raster, and when combining all available datasets. These findings suggest that point-based and lower-resolution spatially distributed assimilation offers a cost-effective alternative in data-limited settings while still providing meaningful improvements in model performance. Overall, assimilating available snow observations substantially enhances snowpack representation, particularly in complex high-elevation terrain. This research provides practical guidance for optimizing snow data assimilation strategies and improving estimates of snowpack characteristics, with direct implications for water supply forecasting and resource management across the Western United States.
Object ID: ISSW2026_P1.21.pdf
DOI: https://doi.org/10.15788/1790098737
Language of Article: English
Presenter(s): David Hill
Keywords: snowpack modeling, lidar, data assimilation, community science
Page Number(s): 644 - 651
-