Skip to Main Content
Supraglacial meltwater streams and lakes that form each summer across large expanses of the Greenland Ice Sheet (GrIS) ablation zone have global implications for sea level rise but remain one of the least studied hydrologic systems on Earth. Remote sensing of supraglacial streams is challenging owing to their narrow width (~1-30 m) and proximity to other features having similar visible/near-infrared reflectance (lakes and slush) or shape (dry stream channels, crevasses, and fractures). This letter presents a new automated “spectral-shape” procedure for delineating actively flowing streams in high-resolution satellite imagery, utilizing both spectral and pattern information. First, a modified normalized difference water index adapted for ice ( NDWIice) enhances the spectral contrast between open water and drier snow/ice surfaces. Next, three NDWIice thresholds are used to mask deep-water lakes and discern open water from slush, in concert with a multipoints fast marching method to rejoin resulting stream fragments. Comparison of this procedure with manual digitization for six WorldView-2 images in southwestern Greenland demonstrates its value for detecting actively flowing supraglacial streams, particularly in slushy areas where classification performance dramatically improves (85.2% success) versus simple threshold methods (52.9% and 59.4% success for low and moderate thresholds, respectively). While a simple threshold approach is satisfactory for areas known to be slush free, the procedure outlined here enables comprehensive stream mapping across the GrIS ablation zone, regardless of slush conditions and/or the presence of similarly shaped glaciological features.