Remote sensing tools for monitoring supraglacial debris cover features and their fluctuations

High mountain environments are characterized by highly glacierized, complex, dynamic topography, often exhibiting a heterogeneous debris mantle comprising ponds and exposed ice cliffs, associated with differing ice ablation rates. These has been an increased interest in understanding these supraglacial surface features, i.e., the formation and expansion of supraglacial ponds and implications for glacier hydrology and glacier-related hazards, notably glacier lake outburst flood (GLOF) events. Until recently, supraglacial debris surfaces and their features have received less attention compared to mapping of debris-cover extents due to methodological challenges posed by their ephemeral nature. As a result, they remain poorly quantified in global glacier databases including the Global Land Ice Measurements from Space (GLIMS) and the Randolph Glacier Inventory (RGI). Furthermore, remote sensing studies used to generate these datasets generally rely on traditional “whole pixel” image classification techniques, which do not allow decomposition of a pixel into constituting materials. In this talk I summarize the state-of-art remote sensing techniques to characterize supraglacial features, such as debris material, ice cliffs, supraglacial ponds and vegetation. I particularly highlight the potential of spectral unmixing routines multi-temporal Landsat and Sentinel data combined with high-resolution multispectral imagery to quantify the composition of debris cover at multiple scales across the Himalaya with an emphasis on supraglacial ponds. I summarize the current strengths and limitations of these methods and discuss steps needed such as automation and open-source tools.