Glacial lake formation probability mapping in the Himalayan glacier: A probabilistic approach

Glacial lakes are critical zones for future flooding and events such as glacial lake outburst floods, a threat to infrastructures and human life. A recent catastrophe (Chamoli flash flood) indicates its complex nature. Therefore, cross-linking of lake formation processes is essential. Our study describes the probability of glacial lake formation on a pixel basis for the Himalayan glaciers using logistic regression. Here, we used three causative factors, viz., climate, glacial and topographic, to determine how a future trigger of climate changes may affect the glacial and topographic parameters, which may generate new lakes. A total of 10,056 Himalayan glaciers were considered here for glacial lake formation probability mapping. Logistic regression analysis exhibits that precipitation, surface texture, slope, plan curvature, temperature and profile curvature yielded maximum weightage accordingly towards lake formation probability with an accuracy of 79% in receiver operating characteristics curve (ROC). In our results, central Himalayan glaciers were found to be more prone to lake formation, with 5.73% of pixels coming under a higher probability zone (>0.8 probability), followed by eastern (4.45%) and western (1.82%) parts. The lower part of the clean glacier and the higher part of the ablation area for highly debris-covered glacier gave a higher probability of lake formation. In the case of medium debris-covered glaciers, the lake formation probability decreases with an increase in distance from the end of the ablation area.

Automated summary

This study examines the probability of glacial lake formation on the Himalayan glaciers by analyzing climate, glacial, and topographic factors. The results show that central glaciers are more prone to lake formation.