Sensitivity of Last Interglacial sea-level high stands to ice sheet configuration during Marine Isotope Stage 6

Estimates of peak global mean sea level (GMSL) during the Last Interglacial (LIG, ∼129-116 ka) based on geological sea level high-stand markers require a correction for the contaminating influence of glacial isostatic adjustment (GIA). This correction is obtained by calculating the viscoelastic response of the Earth to changes in the ice and ocean load prior to and following the LIG. While ice retreat over the last deglaciation is relatively well constrained, changes in ice cover prior to the LIG are more uncertain. We investigate the sensitivity of numerical predictions of GIA during the LIG to variations in the geometry of pre-LIG ice cover and the timing of the deglaciation into the LIG, with a particular focus on Marine Isotope Stage (MIS) 6 (∼190-130 ka). We demonstrate that reconstructing the pre-LIG ice history by replicating the last glacial cycle back in time, rather than using ice volume approximations based on oxygen isotope records, can introduce errors in LIG high-stand predictions of ∼5 m at sites on the peripheral bulge of major ice complexes, and up to ∼2 m at far-field sites. We also demonstrate that predictions of LIG sea level are more sensitive to the geographic distribution of ice cover during MIS 6 than previously recognized. Adopting simulations which vary the relative size of Late Pleistocene ice cover over North America and Eurasia can yield a change in predicted high-stand elevations of ∼5 m in both the near and far field of northern hemisphere ice sheets. This far-field sensitivity arises, in part, from the reorientation of Earth's rotation axis during MIS 6, which in turn drives sea-level changes with a distinct geographic signature. In future work we will apply the insights gained here to re-evaluate the observed geographic variability in geological high-stand markers of LIG age and estimates of GMSL based upon them.