Abstract. The last deglaciation offers an unique opportunity to understand the climate – ice sheet interactions in a global warming context. In this paper, to tackle this question, we use an Earth system model of intermediate complexity coupled to an ice sheet model covering the Northern Hemisphere to simulate the last deglaciation and the Holocene (26–0 ka BP). We use a synchronous coupling every year between the ice sheet and the rest of the climate system and we ensure a closed water cycle considering the release of freshwater flux to the ocean due to ice sheet melting. Our reference experiment displays a gradual warming in response to the forcings, with no abrupt changes. In this case, while the amplitude of the freshwater flux to the ocean induced by ice sheet retreat is realistic, it is sufficient to shut down the Atlantic meridional overturning from which the model does not recover within the time period simulated. However, with reduced freshwater flux we are nonetheless able to obtain different oceanic circulation evolutions, including some abrupt transitions between shut-down and active circulation states in the course of the deglaciation. The fast oceanic circulation recoveries lead to abrupt warming phases in Greenland. Our simulated ice sheet geometry evolution is in overall good agreement with available global reconstructions, even though the abrupt sea level rise at 14.6 kaBP is underestimated, possibly because the climate model underestimates the millenial- scale temperature variability. In the course of the deglaciation, large-scale grounding line instabilities are simulated both for the Eurasian and North American ice sheets. The first instability occurs in the Barents-Kara seas for the Eurasian ice sheet at 14.5 kaBP. A second grounding line instability occurs circa 12 kaBP in the proglacial lake that formed at the southern margin of the North American ice sheet. With additional asynchronously coupled experiments, we assess the sensitivity of our results to different ice sheet model choices related to surface and sub-shelf mass balance, ice deformation and grounding line representation. While the ice sheet evolutions differ within this ensemble, the global climate trajectory is only weakly affected by these choices. In our experiments, only the abrupt shifts in the oceanic circulation due to freshwater fluxes are able to produce some millenial-scale variability since no self-generating abrupt transitions are simulated without these fluxes.

中文翻译：

---

气候和冰盖从最后一个冰期到工业化前的冰盖演变-气候耦合模型

摘要。最后一次冰消作用为了解气候提供了独特的机会-全球变暖背景下的冰盖相互作用。在本文中，为了解决这个问题，我们使用中等复杂程度的地球系统模型以及覆盖北半球的冰盖模型来模拟最后一次冰消和全新世（26-0 ka BP）。我们每年在冰盖和其他气候系统之间使用同步耦合，并考虑到由于冰盖融化而向海洋释放淡水通量，因此我们确保水循环封闭。我们的参考实验显示了对强迫的逐渐升温，并且没有突然变化。在这种情况下，冰盖退缩引起的淡水通向海洋的幅度是现实的，足以关闭大西洋子午线翻转，使模型无法在模拟的时间内恢复。然而，通过减少淡水通量，我们仍然能够获得不同的海洋环流演变，包括在冰消冰过程中，关闭和活跃环流状态之间的某些突然转变。快速的海洋环流恢复导致格陵兰岛突然变暖。尽管低估了14.6 kaBP的突然海平面上升，但我们模拟的冰盖几何形状演化与可用的全球重建总体上很好地吻合，这可能是因为气候模型低估了千分之一摄氏度的温度变化。在冰消过程中 模拟了欧亚和北美冰盖的大规模接地线不稳定性。第一次不稳定性发生在14.5 kaBP的欧亚冰盖的Barents-Kara海中。在北美冰盖南缘形成的冰湖中，第二条接地线的不稳定性发生在12 kaBP左右。通过其他异步耦合实验，我们评估了结果对与表面和子架质量平衡，冰变形和接地线表示有关的不同冰盖模型选择的敏感性。尽管在这个整体中冰盖的演化有所不同，但全球气候轨迹只受到这些选择的微弱影响。在我们的实验中