Comparisons among contemporary glacial isostatic adjustment models

Abstract

We compare 14 glacial isostatic adjustment (GIA) models from different authors to assess how well the GIA effect could be removed from space geodetic data, such as those of the Gravity Recovery and Climate Experiment (GRACE) satellites, to study climate related and other mass changes within the Earth system. Direct comparisons of model outputs confirm unacceptably large discrepancies among models. Here we use two approximate relations within any GIA model without rotational feedback discovered by Wahr et al. (1995), i.e., an uplift–geoid relation and an uplift–gravity relation, to test the models for internal consistency. We find that all models without rotational feedback fit the relations reasonably well with only one exception, which still appears to fit the relations, but not as well. However, some models with rotational feedback fit the relations reasonably well, while some do not. As a result, the discrepancies arising from differences in formulations or computational methods, especially rotational feedback formulations, appear to dominate over those arising from differences in ice history and Earth rheology. We conclude that the accuracy and consistency of GIA models, especially agreement of results obtained using different formulations and computational methods, need to be substantially improved to full exploit contemporary space geodetic data, such as GRACE data, to enhance the constraints on ice-sheet mass balance and the mass component of global sea-level change.