MULTI SENSOR MAPPING AND ANALYSIS OF GLACIAL LAKES AND GLOF RISKS IN THE GHIZER DISTRICT OF THE HINDU KUSH HIMALAYAN RANGE

Abstract

The Northwesterner Gilgit Baltistan’s, Ghizer district is highly susceptible to glacial lake outburst floods (GLOFs), due to climate change. Nearly 24 GLOFs have occurred in this area in the last ~200 years, demonstrating the growing recurrent nature of these incidents. Taking this into account, the assessment of risks associated with GLOFs was investigated in this study. In the first phase, all glacial lakes in the region were identified, and changes between 2000 and 2023 were mapped using moderate resolution satellite images (Landsat). To map built-up and agriculture areas, Landsat's lower resolution limits its use in such complex topography. Therefore Sentinel-2 data was used, images from 2016 to 2023 were classified using random forest (RF) classifier. A total of 617 glacial lakes covering ~31.67 km2 of area were mapped in 2023. Since 2000, ~88 glacial lakes have appeared, showing an increasing trend in the number of lakes. In the second phase, categorization, and susceptibility to GLOFs were assessed using multi-criteria decision analysis (MCDA). Grass GIS tool, r.avaflow was used to generate GLOFs simulations based on parameters such as friction, density, release area, travel time and two travel time scenarios i.e., 1800 and 3600 seconds for four of the high-weighted glacial lakes. The results showed that the glacial lake near Darkut village, Yaseen valley pose a significant threat to downstream communities, while two other lakes in Gupis valley will have moderate effect on the infrastructure and agriculture. No significant threat is posed by the glacial lake of Punyal valley.