The impact of 3-D Earth structure on far-field sea level following interglacial West Antarctic Ice Sheet collapse

Abstract

Prior to inferring ice sheet stability from past interglacial sea-level records, these records must first be corrected for the contaminating effects of glacial isostatic adjustment (GIA). Typical GIA corrections, however, neglect variability in the signal that may be introduced by Earth's 3-D rheological structure. We predict sea-level changes due to a collapse of the West Antarctic Ice Sheet (WAIS) over an idealized 6 kyr-duration interglacial using four viscoelastic Earth models. Two of these are 3-D viscosity models inferred from seismic tomography. The third is a 1-D (i.e., depth varying) viscosity model that is equivalent to the spherically averaged “background” viscosity profile adopted in both 3-D Earth models. The fourth is a 1-D model that has a higher upper mantle viscosity but still falls within the class of models inferred from independent global GIA studies. We find that the discrepancy between 3-D and 1-D Earth model calculations of sea level in the far field of the melt zone is of order 0.3 m or less, with the 1-D Earth models producing higher sea level than the 3-D simulations. This value is 10% of the global mean sea-level (GMSL) rise associated with modeled ice sheet collapse by the end of the model interglacial (∼3 m) and a similar fraction of far-field sea-level changes. However, the value is a significantly larger fraction (∼60%) of the geographically variable (i.e., non-GMSL) component of the far-field sea-level signal due to GIA associated with modeled WAIS collapse (±0.5 m). Neglecting lateral variations in Earth structure in modeling the response to excess melting of WAIS during the interglacial compounds any error introduced by neglecting such structure in predictions of interglacial sea-level change driven by the preceding glacial cycle.