| 08 Apr 2026
Machine Learning and Explainable AI for Sentinel-1–Based Snow Depth Estimation: Insights from Three Different Mountainous Regions
Abstract. Mountain snow plays an important role in regulating freshwater resources and climate processes, making accurate monitoring of its depth and distribution essential for water resource planning and climate modelling. However, monitoring snow depth (SD) in mountainous regions remains challenging due to the extreme topographic and climatic conditions. Synthetic Aperture Radar (SAR)-based remote sensing is particularly useful in such regions due to its high spatial and temporal resolution and its ability to penetrate deep snow. This study focuses on improving Sentinel-1 C-band SAR backscatter-based SD estimation, which supports large-scale, continuous monitoring but faces limitations in vegetated, shallow, or wet snow conditions. To address these limitations, the present study developed a machine learning (ML) framework that integrates change detection-based backscatter indices, direct backscatter variables, and auxiliary datasets to improve SD estimation over the existing ML models. The framework is designed for implementation using preprocessed Sentinel-1 data from Google Earth Engine, thereby reducing computational requirements and enabling scalability across regions and time periods. It is implemented across three different mountainous regions, the Colorado Rocky Mountains, the European Alps and the Indian Western Himalayas, each representing diverse characteristics. Results demonstrate that the developed ML model significantly outperforms existing approaches in all regions. As the physical scattering mechanisms underlying the backscatter-based methods are not yet fully understood, explainable AI techniques are employed to interpret model behaviour and investigate how different input variables contribute under varying conditions. The analysis reveals distinct patterns across seasons, snow types, vegetation conditions, and regions, providing physical insights into C-band backscatter-based SD estimation.
This manuscript presents a machine-learning framework for snow-depth estimation in mountainous regions using Sentinel-1 C-band SAR backscatter variables, change-detection-based backscatter indices, and several auxiliary predictors. The method is applied to the Colorado Rocky Mountains, European Alps, and Indian Western Himalayas. The authors report improved performance relative to the existing C-Snow and C-RISE products, and use SHAP-based explainable AI to interpret regional and condition-dependent model behaviour. The manuscript explicitly states two main objectives: developing an improved scalable SAR-based snow-depth estimation method and gaining better understanding of the physical mechanisms underlying C-band backscatter-based snow-depth estimation.
While the empirical prediction results are promising in some regions, I do not think the manuscript provides a sufficiently significant scientific contribution for publication in The Cryosphere. In particular, the second stated objective is not achieved. The explainable AI analysis describes the behaviour of the trained ML model, but it does not establish physical mechanisms linking snow depth and C-band backscatter. Moreover, the model appears to rely strongly on auxiliary variables such as snow cover duration, elevation, day of water year, snow class, and altitude group. The authors’ own SHAP analysis shows that SCD is the most influential variable across all regions, with DEM also highly important, whereas the backscatter variables are secondary and condition dependent.
Therefore, the manuscript reads mainly as an applied machine-learning retrieval exercise rather than a contribution to radar snow physics or cryospheric process understanding. The current framing overstates the physical insight provided by the analysis. I recommend rejection. A substantially reframed and redesigned study might be suitable elsewhere as an applied ML paper, but in its current form I do not think it meets the scientific standard expected for The Cryosphere.