Research Projects
Example of a rapid supraglacial lake drainage event:
drain_lake.mp4What happens to Greenland's supraglacial lakes each season?
Along the margins of the Greenland Ice Sheet, lakes form on the surface during the summer months. These lakes can drain rapidly over a few hours or days through cracks in the ice, delivering water to the base of the ice sheet and influencing ice velocity. At the end of the summer, remaining surface meltwater refreezes, or can sometimes remain liquid, insulated by snowfall and buried beneath the surface. In this work, we develop a new machine learning method to automatically categorize lakes that drain, refreeze, or become buried based on the lake's evolution in satellite imagery.
Machine learning and remote sensing for more accurate snow depth measurements
Seasonal snow plays a crucial role in society and understanding trends in snow depth and mass is important for making informed decisions about water resources and adaptation to climate change. Here we introduce a machine learning approach to enhance satellite-based snow depth estimation over the European Alps. We integrate Sentinel-1 SAR imagery, optical snow cover observations, and topographic, forest cover and snow class information to provide snow depth estimates at 100 m resolution.
iceland_poster.pdfIce shelf firn evolution
Firn air content, or the amount of air space in the upper ice sheet layers, is important because it can hold meltwater that might otherwise run off into the ocean or pond on the surface. In this project we build a firn emulator which uses near-surface temperature, wind speed and precipitation to predict firn air content across Antarctic ice shelves. Using our emulator, we predict ice-shelf firn air content throughout the 21st century under different forcing from the Climate Model Intercomparison Project suite of global climate models.
How well does CESM2 represent Antarctica's surface climate?
Global climate models are essential for understanding future changes in the Earth's climate system. In this project we investigate the global climate model CESM2 and its representation of Antarctica’s near-surface climate. We compare CESM2 output of near-surface temperature, wind speed, melt, surface mass balance, and incoming long and short wave radiation with re-analysis models, satellite observations, and in-situ weather station observations.
fig03.pdfGreenland Ice Sheet buried lakes
Buried lakes are an under-studied component of ice sheet hydrologic networks. In this work we use Sentinel-1 SAR imagery to automatically detect buried lakes at a continent-wide scale, providing a database of buried lakes across the Greenland Ice Sheet over two years, and investigate the climatological drivers behind their formation.
Observations of a buried lake drainage
In-situ observations of lake drainage events are rare. In this study we combine in-situ (ground penetrating radar and GPS) and remote (satellite-derived elevation change and Sentinel-1 backscatter change) observations to tell the story of a buried lake drainage event in East Antarctica.
What does the inside of a buried lake look like?
BNM_SubsurfaceMelt_NoSub.mp4Video credits: Stef Lhermitte