Semi-automated mapping of glacial lakes – a study in Sikkim Himalayas, India

In the Himalayas, rising number of glacial lakes with increase in area of existing lakes is a matter of serious concern and correct mapping is an important step towards regular monitoring of these glacial lakes. This study used Double Window Flexible Pace Search method to identify an optimal threshold of Normalized Difference Water Index value on multi-spectral Sentinel-2 image of 31st October 2020. A threshold value of 0.30 is obtained that resulted in identification of 419 glacial lakes with an estimated total area of about 29.86 ± 6.85 km2. The results are found to be in close agreement with previously published research article on Sikkim Himalayas and boundaries of classified lakes are validated by overlaying on high resolution PlanetScope dataset for the same date. Hence, this study presents an effective and simple way for mapping of glacial lakes that may further be extended for regular monitoring of glacial lakes in the region.