Monitoring Snow Depth with ICESat-2 at mid-latitudes: A Synergistic Multi-Scale Framework Integrating Ground-Airborne-Satellite Observations in Northern Xinjiang, China

Accurate observation of seasonal snow depth (SD) across spatial scales remains a major challenge in mid-latitude regions, particularly over complex terrain where sub-footprint heterogeneity and scale mismatch strongly affect satellite-based retrievals. Although ICESat-2 has demonstrated high potential for SD estimation in high-latitude regions, its performance in mid-latitude areas is constrained by the limited availability of snow-free digital elevation models (DEMs) with centimeter-level vertical accuracy and by the scarcity of reliable ground-based validation due to ground-track shifting.

To address these challenges, we established a multi-scale “ground-airborne-satellite” synergistic observation framework within a controlled study area in northern Xinjiang, China. To reconcile spatial scale mismatches among the different observational platforms, UAV-LiDAR data were employed as a validated intermediate-scale bridge (RMSE = 6.03 cm against in-situ measurements). Based on this framework, we conducted an error propagation analysis to quantify ICESat-2 SD uncertainty under varying terrain conditions.

Results indicate that ICESat-2 achieves excellent accuracy over flat, open terrain (slope < 5°), with an RMSE of 6.69 cm. In contrast, over complex sub-footprint terrain combining steep slopes and artificial structures, SD deviations increased substantially, ranging from -30 to +60 cm, reflecting the strong influence of sub-footprint terrain heterogeneity on SD retrieval. Across the entire study area, ICESat-2 maintains robust overall performance, yielding a total RMSE of 15.61 cm.

This study demonstrates the feasibility of accurate ICESat-2 SD retrieval in mid-latitude regions and emphasizes the critical influence of sub-footprint terrain complexity on SD uncertainty. The proposed multi-scale observational framework provides a transferable approach for interpreting satellite-derived snow products and for improving the representation of snow processes across scales.