Improved GIA correction yields larger Antarctic mass loss.

Abstract:

New regional ice deglaciation models have recently been developed to match a variety of geologic, glaciological, and geodetic observations. In Antarctica, these new models exhibit a smaller East Antarctic ice loss since the Last Glacial Maximum, and hence yield a smaller GIA correction to the Antarctic estimates than those predicted by ICE5G. These revised models yield less negative ice mass losses when using GRACE data. Although these new models represent a significant advance in Antarctic GIA modeling, there are still large uncertainties associated with them. One of the large uncertainties is due to the fact that in the East Antarctica interior, the GIA reconstruction is poorly constrained by observations. These new models assume a monotonic decrease in loading in the last 5000 years. We examine the impact of this assumption on the GIA estimates and how a more realistic non-monotonic loading scenario could impact the results and the GRACE ice mass estimates. We use GRACE in combination with output products from the Regional Atmospheric Climate Model (RACMO) and recent studies of deglaciation history to derive an improved GIA correction, which include a non-monotonic loading scenario, is consistent with available geological and geodetic constraints and reconstruction of recent climate history. We find a larger correction, which implies larger losses of the Antarctica ice sheet by about 70 Gt/yr.