Modeling 2.7 Million years of Greenland ice sheet fluctuations using high Arctic temperature reconstructions

Understanding the behavior of the Greenland ice sheet (GrIS) during the Pleistocene is necessary for predicting its contribution to future sea-level rise. However, competing evidence for and against dramatic ice loss on Greenland during recent interglacial stages complicates our understanding of the longer-term history of the GrIS. Here we use continuous climate reconstructions from the Pliocene-Pleistocene transition until present day, including a terrestrial reconstruction from Lake El'gygytgyn, Arctic Northeast Russia, to drive a hybrid shallow-ice/shallow-shelf three-dimensional thermomechanical ice-sheet model of the GrIS. The recently developed Lake El'gygytgyn climate record is unique because it provides the only continuous reconstruction of continental Arctic paleoclimate since the late Pliocene, and this is the first time the record has been used to drive an ice-sheet model on these timescales. We use an ensemble approach to account for uncertainty in the climate forcing and evaluate simulations by comparing the predicted cosmogenic nuclide inventory around Greenland's margin and beneath the GrIS to existing measurements. Using basal processes to constrain ice sheet history also allows us to compare the total erosion predicted by the model to subglacial observations of erosion and offshore Pleistocene-age sediment packages. Our results have implications for understanding the initiation and persistence of the GrIS in response to the super-interglacials, glacial-interglacial cycles, and long-term cooling that characterize Pleistocene climate evolution. We show that compared to other available climate reconstructions, the terrestrial record from Lake El'gygytgyn suggests that Greenland was nearly ice-free for recent interglacials, including MIS 7 and MIS 11. We use our results to identify regions where new data would have the greatest impact on constraining model performance and therefore the greatest potential to advance our understanding of GrIS history.