Comprehensive Mapping and Climatic Assessment of Supraglacial Lakes in Dronning Maud Land, Antarctica

Anthropogenic climate change is real, impacting various components of the Earth. Surface temperatures have reached unprecedented levels in the recent past, ice sheets are rapidly losing mass, and their contributions to global sea level are rising. The future is uncertain, posing key challenges to ecosystems and human livelihood. A major contributing factor in this uncertainty is how ice sheets, particularly the Antarctic ice sheet – the largest reservoir of ice on Earth – will respond to climate change and affect sea levels and global climate. Developing accurate models that parametrize ice sheet behavior and response remains a major scientific challenge, requiring a thorough understanding of all associated processes and feedback mechanisms. This necessitates comprehensive knowledge of processes like marginal supraglacial lake evolution, which have been identified as critical precursors to ice shelf instability through mechanisms such as ice shelf flexure, hydrofracture, and collapse. Recent studies have documented the widespread presence of Antarctic supraglacial lakes and outlined their potential implications. However, much remains to be achieved to clearly identify the driving factors behind the spatiotemporal variability of lakes and their evolution to better understand their impacts. This thesis aims to address this gap by adopting a regionfocused approach in Dronning Maud Land, East Antarctica. Using the complete archive of Landsat 8 and Sentinel-2 optical imagery between 2014 and 2021, we map supraglacial lakes across marginal Dronning Maud Land. We then perform a climatic assessment using reanalysis climate data (ERA5 and ERA5-Land) and simulations from a high-resolution regional climate model (Modèle Atmosphérique Régional - MAR) to evaluate relationships between ponding, climatic conditions and environmental factors. In Paper I, we present the results of multi-seasonal supraglacial lake mapping and depth estimation, covering seven melt seasons from 2014/15 to 2020/21. Our findings reveal significant interannual variability in lake extents across all seven lake regions, which remains consistent throughout Dronning Maud Land during both high and low melting/ponding years. While climatic assessments using ERA5 and ERA5-Land data were inconclusive, highresolution 5 km simulation from the MAR provided some key insights. The lake-climate assessment highlighted the role of downslope winds, solar and longwave radiation, and precipitation in driving seasonal ponding extents. The higher resolution of MAR simulations more effectively captured the influence of local topography on climate, a limitation of coarser reanalysis datasets, allowing for a more accurate representation of near-surface atmospheric conditions. Additionally, we identified a strong correlation between the Southern Annular Mode (SAM) and lake extents in Dronning Maud Land, underscoring the influence of largescale atmospheric teleconnections on meltwater generation and ponding in the region. In Paper II, we investigate the role of climatic and environmental factors on the spatial distribution of lakes in Dronning Maud Land. The mapping exercise presented in Paper I revealed large variability in regional lake extents. In the follow-on paper, we found that spatial distribution of lakes is controlled by locally-driven downslope katabatic and foehn winds. We discussed mechanisms by which non-summer winds precondition the surface for summer melting, contributing to summertime ponding. This study presents the first documentation of foehn winds in Dronning Maud Land, in addition to providing finer mapping of katabatic winds, characterizing their spatial extent, frequency and warming effects on coastal ice-sheet surfaces. Finally, we identified other ice shelves in the region that are exposed to wind-induced warming and are potentially vulnerable to future meltwater ponding. The findings from this thesis provide novel insights into Antarctic supraglacial lake dynamics through a detailed study of Dronning Maud Land. By adopting a regionally comprehensive approach and utilizing high-resolution climate simulations, this study uncovered previously unrecognized dynamics that are crucial for understanding the future of surface hydrology in Antarctica. These insights have the potential to enhance model simulations of surface hydrology, thereby improving predictions of future ice shelf stability and contributing to a better understanding of the ice sheet responses to climate change.