Lasers vs. Lasers: An intercomparison between lidar datasets, GNSS, and snow-probe transects from NASA's SnowEx campaign.

NASA's SnowEx campaign at Grand Mesa, CO collected a suite of ground control observations - including Terrestrial Laser Scanning (TLS), manual snow-probe transects, and differentially-corrected Global Navigation Satellite System (GNSS) observations - to quantify the accuracy of airborne-lidar derived surface elevation and snow depths from the Airborne Snow Observatory (ASO). Each ground control method has a range of uncertainty in its ability to measure snow depth or elevation. ASO data were compared to the suite of observations to determine whether the differences in elevation or snow depth were within the associated uncertainties. Differences in median snow depth values between ASO and TLS were around 6 cm (5%), and both contained similar spatial patterns, where differences between standard deviations were around 4 cm (3%). Furthermore, differences in snow depth between ASO and TLS were within the uncertainty of both lidar datasets and results were insensitive to the gridding resolution. Similarly, within a TLS site, median snow depth values from the snow-probe transects agreed with the median snow depth values from ASO and TLS to within 3-13 cm (2-18%). However, the snow-probe transects underestimated the standard deviation at TLS sites with significant spatial variability, as the difference in standard deviation between ASO and the snow-probe could be as large as 40 cm (41%). These results demonstrated the utility of ASO data and indicate that ASO can be used as a spatial snow depth dataset to validate various remote sensing techniques, improve our understanding of snow depth's spatial variability, evaluate hydrologic models, and ultimately improve water forecasting in snow dominated watersheds. Using lessons learned from this intercomparison, an outline of suggestions for future ground validation campaigns of airborne sensors is provided.