Diverse models, diverse interglacial results? Sea ice physics versus model forcing

Abstract

<p>Different IPCC-CMIP6 climate models give diverse results when run under common interglacial forcing. The mid-Holocene and the Last Interglacial are the two time periods from which we have the most results. The diversity is particularly true for the Arctic, where sea ice physics plays a key role. Whilst scientists have known for more than twenty years that summer temperatures in the Arctic during the Last Interglacial – the warm period around 127,000-128,000 years ago - were around 4°C above those of today (from lake, peat, and marine core data), the cause of this warmth puzzled scientists until 2020. Until 2020, it was thought this Last Interglacial warmth may have been driven by Arctic vegetation changes.  We present an analysis of a variety of CMIP6 model simulations run during the Last Interglacial.  Only one model simulates a fully sea ice-free Arctic during the summer –it includes an advanced representation of melt ponds in the sea ice model. Melt ponds are shallow pools of water which form on the surface of Arctic sea ice. We find that the inclusion of melt ponds within models is likely crucial for understanding Last Interglacial sea ice loss and Arctic warmth, and touch on the relationship between Arctic sea ice changes in warm climate and on high equilibrium sensitivity CMIP6 models. Alongside the impact of different physics in the models, we also consider forcing aspects including the impact of the meltwater from deglaciation and top-of-the-atmosphere radiation (orbital).</p>