Impact of Heterogeneous Mantle Viscosity on the Glacial Isostatic Adjustment Correction for GRACE(‐FO): Implications for Surface Mass Trends

Abstract

Accurate interpretation of time‐variable gravity from GRACE (‐FO) requires correction for glacial isostatic adjustment (GIA), which has a gravitational signal similar in magnitude to that of contemporary hydrological and cryospheric mass change. Traditional GIA models used to correct space gravimetry data assume a radially symmetric mantle viscosity structure, ignoring the 3D structure evident in seismic tomography. We quantify the sensitivity of GRACE‐derived surface mass trends to laterally heterogeneous mantle viscosity by computing a suite of global GIA models using the finite‐element code CitcomSVE. We utilize 3D viscosity fields derived from tomography and two commonly used ice‐loading histories, systematically increasing heterogeneity about their associated radial viscosity profiles. GIA solutions are converted to GRACE‐compatible Stokes coefficients and equivalent water height to recompute GRACE mass trends. While global mean ocean and Greenland ice mass trends are weakly affected, regional terrestrial water storage and ocean mass patterns exhibit large and spatially coherent changes. 3D GIA corrections lead to significantly more drying conditions in Canada and slightly more drying conditions throughout most land‐masses, can lead to a reduction in the long‐term rate of ice mass loss in Antarctica, and can reverse the ocean mass trend in key ocean basins. Models combining elevated lower‐mantle viscosity with strong lateral heterogeneity eliminate problematic GRACE ocean anomalies in the eastern North Atlantic and potentially resolve the disagreement between GIA‐ and geodynamics‐derived estimates of mantle viscosity. Our results demonstrate that realistic 3D viscosity is essential for quantifying and reducing GIA‐related uncertainty in GRACE interpretations, improving confidence in inferred surface mass changes.