Evaluation of supraglacial lake depth estimation techniques using Sentinel-2, ICESat-2, TanDEM-X, and in situ data, along with an analysis of rapid drainage events over Northeast Greenland

Supraglacial lakes are dynamic hydrological features that play a significant role in surface mass balance estimations. These lakes act as conduits for surface and subglacial runoff, the amount of which varies quite strongly based on the local surface temperature, rainfall, and snowpack thickness. Additionally, supraglacial lakes tend to undergo impactful events called rapid drainages, during which a crack opens at the lake bed, draining the meltwater to the glacier bed within hours or days. Not only does this contribute to glacier mass loss and freshwater influx into the ocean, but it can also cause temporary glacier speed-ups due to the reduction of friction on the glacier bed. Currently, the influencing factors involved in triggering these rapid drainages are minimally understood. This research firstly focuses on the comparison of several lake depth estimation methods in order to be able to accurately monitor seasonal lake development and quantify meltwater volumes. Secondly, the temporal and spatial variances in rapid drainages in Northeast Greenland, specifically over Zachariae Isstrom and Nioghalvfjerdsfjorden (79N Glacier), are evaluated in order to understand underlying causes and to quantify the amount of meltwater lost through them. 

Several established and novel supraglacial lake depth estimation methods are evaluated in this research, comprising of (1) a radiative transfer model based on Sentinel-2 data, (2) an empirical equation derived from ICESat-2 lake crossings and Sentinel-2 data, (3) an empirical equation derived from in situ sonar data gathered in Northeast Greenland and Sentinel-2 data, and (4) TanDEM-X elevation data. These four methods are directly compared on five supraglacial lakes in Northeast Greenland, highlighting the advantages and limitations of each method. Furthermore, the three methods based on Sentinel-2 imagery are applied to the peak melt dates in Northeast Greenland over the 2016 to 2022 melt seasons to understand seasonal variations. Finally, individual lakes are tracked throughout the seven melt seasons to allow for a detailed assessment of rapid drainage occurrences in the region. Overall, insight into the behavioral patterns and influencing factors involved with the rapid drainages of supraglacial lakes and the amount of meltwater lost from them has been gained.