Evaluation of snow depth retrievals from ICESat-2 using airborne laser-scanning data

• DOI: 10.5194/tc-17-2779-2023
• Publication Date: 2023-07-13T06:57:02Z
• Authors: César Deschamps-B...; Simon Gascoin; David Shean; Hannah Besso; Ambroise Guiot; Juan Ignacio Lópe...
• Keywords: Environmental sciences; QE1-996.5; 550; 13. Climate action; [SDU.STU.HY] Sciences of the Universe [physics]/Earth Sciences/Hydrology; GE1-350; Geology; [SDU.STU.HY]Sciences of the Universe [physics]/Earth Sciences/Hydrology; 15. Life on land; 01 natural sciences; 0105 earth and related environmental sciences

Abstract. The unprecedented precision of satellite laser altimetry data from the NASA ICESat-2 mission and the increasing availability of high-resolution elevation datasets open new opportunities to measure snow depth in mountains, a critical variable for ecosystem and water resource monitoring. We retrieved snow depth over the upper Tuolumne basin (California, USA) for 3 years by differencing ICESat-2 ATL06 snow-on elevations and various snow-off digital elevation models. Snow depth derived from ATL06 data only (snow-on and snow-off) offers a poor temporal and spatial coverage, limiting its potential utility. However, using a digital terrain model from airborne lidar surveys as the snow-off elevation source yielded a snow depth accuracy of ∼ 0.2 m (bias) and precision of ∼ 1 m (random error) across the basin, with an improved precision of 0.5 m for low slopes (< 10∘), compared to eight reference airborne lidar snow depth maps. Snow depths derived from ICESat-2 ATL06 and a satellite photogrammetry digital elevation model have a larger bias and reduced precision, partly induced by increased errors in forested areas. These various combinations of repeated ICESat-2 snow surface elevation measurements with satellite or airborne products will enable tailored approaches to map snow depth and estimate water resource availability in mountainous areas with limited snow depth observations.