funded under the Our recent simulations of Greenland Ice Sheet mass loss have shown that rates of ice loss in the 21st century will be much higher than during the past 11,700 yrs.
However, this work extrapolated results from southwest Greenland to the entire ice sheet and made some simplifying assumptions about the last 12,000 years of climate. Our new project is expanding the domain to the entire ice sheet, improving the climate data to account for more climate proxies such as leaf wax and diatoms, constraining the ice sheet with more geochronological data from cosmogenic nuclide dating, and doing more to capture the uncertainties present in the analysis. Our role at UM is quantifying the uncertainty. We are currently using neural networks to emulate the ice sheet model so that we can determine the sensitivity of the ice thickness to uncertainty in the climate.
