New Insights from 10-Be on the Deglaciation of Greenland: Comparison of a Continent-Wide Database with Climatic Forcings and Model Results

Abstract:

We present a database of all 10-Be ages dating Greenland ice-sheet (GIS) retreat, enabling (1) better understanding of late deglacial through mid-Holocene behavior of the GIS and (2) independent verification of a prominent ice-sheet model. Several regions in southern and southeastern Greenland show instantaneous deglaciation within uncertainty, suggesting retreat from the coast to the present ice margin occurred over several centuries or less. Regions in western Greenland indicate slower deglaciation, with retreat taking up to several millennia. In three locations (Scoresby Sund, Johannes Jensen Land, and Kangerlussuaq), early deglaciation to the coast (by 15 ka) is observed. This variability may result from regional variation in climate or bedrock topography, leading to different rates of ice retreat. A comparison of the overall distribution of 10-Be ages with temperature records from ice cores and the ocean reveals deglaciation in Greenland is particularly sensitive to ocean warming from 15-8 ka, although atmospheric warming at the time also contributed to ice retreat. Peak deglaciation occurred at 12-11 ka during peak ocean warmth, with a secondary peak at 8-7 ka. The Huy2 ice-sheet model of the last deglaciation of the GIS shows deglacial peaks at similar times (10.5 ka and 7.5 ka) as the 10-Be ages. However, the earlier deglaciation in 10-Be ages is not reflected in the model, and nearly all 10-Be ages are older than their respective ice-sheet model ages. This may be due to systematic inheritance within 10-Be samples, but tight clustering of 10-Be ages suggests these represent actual deglaciation ages instead of random inheritance. Alternately, the model may under-predict deglaciation age due to its relatively coarse resolution, preventing it from including finer-scale topography which may significantly influence ice sheet behavior. The model also does not include ocean temperature forcing, which may have been an important forcing in driving GIS retreat when the ice margins were marine terminating. Future ice-sheet modeling should include oceanic effects to test this hypothesis.