Supraglacial lake evolution on Tracy and Heilprin Glaciers in northwestern Greenland from 2014 to 2021

Yefan Wang,Shin Sugiyama
DOI: https://doi.org/10.1016/j.rse.2024.114006
IF: 13.5
2024-01-29
Remote Sensing of Environment
Abstract:Supraglacial lakes form on the Greenland ice sheet during the melting season as meltwater accumulates in topographic depressions on the surface. They have the potential to increase ablation by reducing albedo and enhance ice flow when water drains to the bed. Despite recent progress in understanding the surface hydrology of the Greenland ice sheet, the seasonal and year-to-year variations of supraglacial lakes remain poorly understood. Here, we employed a machine learning method trained with random forest classifiers on Landsat 8 and Sentinel-2 imagery within Google Earth Engine for detecting supraglacial lakes. We applied this method to two large marine-terminating glaciers (Heilprin Glacier and Tracy Glacier) in northwestern Greenland to investigate the distribution, inter-annual and seasonal evolutions of supraglacial lakes between 2014 and 2021. Over the eight study years, cumulated maximum lake extent on Heilprin Glacier (22.84 km 2 ) was approximately three times larger than that of Tracy Glacier (7.60 km 2 ) despite the similarity in the glacier areas (654 km 2 and 540 km 2 ). The lake extent exhibited significant interannual variability controlled by summer PDD (sum of positive degree days), and the upper limit of lake formation coincided approximately with ELA (equilibrium line altitude). Greater lake areas were observed in 2016 and 2019, but obviously below-average areas in 2017–2018. The variabilities were pronounced in the region above 800 m a.s.l., particularly at Tracy Glacier. On the seasonal scale, lakes begin forming in early- to mid-June and disappear toward the end of August. The lake formation was triggered by the initiation of runoff in the regions below 800 m a.s.l., while it coincided with the onset of surface melt at the higher elevations due to the presence of ice/snow-buried lakes. Cascading rapid lake drainage events were observed in the mid-elevation areas (400–800 m a.s.l.) of Heilprin Glacier, which induced an abrupt decrease in the lake area at the beginning of July in the years of 2019 and 2020. Our study demonstrates the potential of the machine learning method to analyze supraglacial lakes in dense satellite images. The results imply inland expansion of supraglacial lakes under warming climate in Greenland and a possible interaction between draining lake water and glacier dynamics.
environmental sciences,imaging science & photographic technology,remote sensing
What problem does this paper attempt to address?