Item

Abstract

Changes in sea level and vertical land motion associated with glacial isostatic adjustment (GIA) are embedded in paleo and geodetic data sets used to constrain the past and future evolution of the Greenland ice sheet. Thus, understanding of ice sheet evolution goes hand in hand with our ability to simulate the GIA signal accurately. We aim to improve the accuracy of Greenland GIA simulations by interrogating regional geophysical data sets to determine better 3-D models of Earth structure in this region. We use a self-consistent Bayesian joint inversion framework (LitMod) to constrain lithosphere and shallow mantle properties and their uncertainty from multiple data sets. The inversion results indicate a high sensitivity to the input seismic dataset, so we incorporate a new, regional high-resolution surface wave dataset based on the two-station interferometry method. In terms of simulating GIA, a key inversion output is the regional temperature field. We sub-sample a high variance set of 25 temperature fields to define 25 models of the lithospheric thickness (LT) and 50 models of sub-lithosphere viscosity structure (using two different scalings). Our results indicate that viscosity and LT vary, respectively, by 3-4 and 2 orders of magnitude across Greenland. So the predicted GIA signal shows significant differences compared to simulations based on the more traditional 1-D (spherically symmetric) viscosity models. We will present results based on these 50 Earth models and two different ice history models and compare them to geological reconstructions of relative sea-level change and GNSS observations of vertical land motion.