The superiority of the Adjusted Normalized Difference Snow Index (ANDSI) for mapping glaciers using Sentinel-2 multispectral satellite imagery

Accurate monitoring of glaciers’ extents and their dynamics is essential for improving our understanding the impacts climate environmental changes in cold regions. The satellite-based Normalized Difference Snow Index (NDSI) has been widely used mapping snow cover glaciers around globe. However, snow-covered areas using existing indices remains a challenging task due to incapabilities separating snow, glaciers, water. This study aimed evaluate new index apply machine learning algorithms improve accuracy glaciers. A based on satellite data from Sentinel-2 was tested, which we call Adjusted (ANDSI). ANDSI (besides NDSI) with five different algorithms, namely Artificial Neural Network, C5.0 Decision Tree Algorithm, Naive Bayes classifier, Support Vector Machine, Extreme Gradient Boosting, map performance evaluated against ground reference data. Four glacierized regions countries (Canada, China, Sweden, Switzerland-Italy) were selected as sites proposed ANDSI. Results showed that outperformed original NDSI, classifier best overall Kappa among classifiers majority cases. NDSI yielded results an average (around) 91% 95% glacier across all models demonstrates serves superior improved method accurately