GLOFs pose significant risks in CKNP, home to 608 glaciers, including Baltoro. Previous research struggled to integrate upstream, downstream, environmental, and anthropogenic factors due to data heterogeneity. This study addresses these challenges with a two-stage framework combining machine learning, fuzzy logic, and AHP. Upstream and downstream variables were examined using multi-source datasets like SRTM DEM, ICIMODIS Inventory, and Landsat-generated glacier borders. Fuzzy Posterior Probabilities for GLOF classification were generated by applying fuzzy logic and AHP weighting. The results indicate among thirteen models, GBM performed best (AUC = 0.83) and trend analysis from 2010 to 2024 reveals moderate changes in GLOF susceptibility, with moderate-risk lakes showing notable area growth. This approach enhances GLOF susceptibility assessments by combining upstream and downstream factors, offering a scalable solution for climate adaptation, hazard management, and sustainable planning in glacierized areas.
Abbas et al. (Sun,) studied this question.