Investigating the possibility of East Antarctic ice mass loss as an explanation for GPS-derived observations of horizontal motion

GPS sites in the Transantarctic Mountains operating under the Antarctic Network (ANET) component of the Polar Earth Observing Network (POLENET) record crustal motion in response to glacial isostatic adjustment (GIA). Observed horizontal motions are towards former ice mass centers in West Antarctica, opposite to the expected and modelled pattern of deformation due to GIA. The disagreement between observed and predicted surface deformation suggests modification to one or both primary GIA model inputs, ice history and earth properties models, is needed. 1D GIA models for Antarctica utilize radial earth models, yet mantle viscosity mapped by seismology indicates a strong boundary in earth properties between East and West Antarctica. GPS-derived horizontal crustal motions are consistently near-perpendicular to the boundary and a gradient in the magnitude of motion across the boundary is observed, with velocities increasing from the stronger, East Antarctic side, to the weaker, West Antarctic side. The spatial correlation between horizontal crustal displacements and modeled viscosity values suggests a causal relationship, and offers an explanation for the mismatch with 1D GIA models that do not incorporate lateral variation. Alternatively, we investigate the possibility of East Antarctic ice mass loss as an explanation for the discrepancy between observed and predicted surface deformation. Ice history scenarios invoking removal of ice mass from the Wilkes Subglacial Basin are coupled with a range of 1D earth models, and a comparison between predicted and observed motions made. Results suggest that East Antarctic unloading may explain both the magnitude and direction of observed motions for some regions. Perhaps more significantly, surface displacements located within an "interference zone" between the West Antarctica ice mass center and our postulated Wilkes ice mass center are strongly matched for a variety of earth model combinations, supporting the case for ice unloading from the Wilkes Subglacial Basin since the Last Glacial Maximum.