Modeling insights into arctic climate and cryospheric characteristics during past interglacials

Abstract

Ongoing anthropogenic climate change is accelerating the loss of both land ice from the Greenland ice sheet (GrIS) and sea ice in the Arctic, underscoring the urgency to improve our understanding of the mechanisms threatening these critical climatic systems. Utilizing a modeling-centric approach, this thesis examines the past warm interglacials Marine Isotope Stage 5e (MIS-5e; ca. 130-115 ka) and MIS-11c (ca. 430-395 ka). MIS-11c in particular is regarded as enigmatic, with very warm boreal summer temperatures for a long period suggested by proxy records despite relatively weak solar forcing. However, time-slice simulations of this period suggest that orbital forcing conditions actually reduce atmospheric heat fluxes over Greenland and act as a negative feedback to further warming. Climate forcing data from these MIS-11 simulations were then used to conduct an ensemble of GrIS simulations using an ice model with a focus on identifying key uncertainties. Constraining these results using ice-core data suggests that the GrIS conservatively contributed at least 3.2 to 4.6 m of sea-level rise during MIS-11. Moreover, the lapse-rate methodology that was applied for correcting surface temperatures exhibited the greatest influence on these results, and a data-based spatially and temporally varying lapse rate performs best. Finally, the behavior of summer Arctic sea ice was examined in both MIS-5e and MIS-11c, with ice-free conditions only being achieved in early MIS-5 (127 ka). Summer ice levels exhibited a strong inverse relationship to Integrated Summer Energy in both interglacials, and interannual variability of ice bore a strong relationship to the mean-state area. Specifically, variability maximized for intermediate ice-areas, with both high- and low-ice states exhibiting limited variability. This suggests that the present century will see increasingly variable summer ice pack as mean-state ice levels approach those seen in the “intermediate” time-slice simulations.