A Novel Approach for Automatic Glacial Lake Type Identification in the Parlung Zangbo Basin via Glacially Fed Lake Subdivision

Glacial lakes in the Third Pole are critically important for climate change and ecological environments. Classifying different types of glacial lakes has become increasingly crucial for dynamic lake monitoring and glacial lake outburst flood (GLOF) assessment. However, automatic identification of glacial lake types still faces numerous challenges. This study developed an automatic classification scheme for glacial lakes by integrating the longest glacier centerline with glacier retreat zones and glacial meltwater flow paths. The scheme comprehensively considers the spatial relationship between glacial lakes and their parent glaciers, as well as dam properties, enabling accurate derivation of key parameters for each glacial lake, including lake type, number of supply glaciers, total area, and the length of inflow channels from parent glaciers. Applying the proposed rule-based classification scheme to 1429 glacial lakes, integrated from eight glacial lake inventories, revealed that the Parlung Zangbo Basin (PLZB) contains 13 supraglacial lakes, 41 ice-contact lakes, 521 glacier-proximal lakes, 235 glacier-distal lakes, and 619 non-glacially fed lakes. The classification scheme is sensitive to changes in glacier extent and can accurately identify non-glacially fed lakes within 10 km of glaciers. Furthermore, this study refined the classification of non-contact glacier-fed lakes into “glacier-proximal” and “glacier-distal” categories, providing a more detailed basis for assessing dam stability and glacial influence, thereby contributing to future large-scale susceptibility assessments of GLOF events.