Gravitational signature and apparent mass changes in Amundsen Embayment caused by low viscosity GIA model constrained by rapid bedrock displacement

Abstract:

The Amundsen Embayment sector of West Antarctica is experiencing some of the fastest sustained bedrock uplift rates in the world. These motions, recorded by the Antarctic GPS Network (ANET), cannot be explained in terms of the earth’s elastic response to contemporary ice loss, and the residues are far too rapid to be explained using traditional GIA models. We use 13 years of very high resolution DEM-derived ice mass change fields over the Amundsen sector to compute the elastic signal and remove it from the observed geodetic time series. We obtain a very large residual - up to 5 times larger than the computed elastic response. Low or very low mantle viscosities are expected in this area based on existing heat flow estimates, seismic velocity anomalies, thin crust, and active volcanism, all of which are associated with geologically recent rifting. We hypothesize that the rapid crustal displacement manifests a low viscosity short-time-scale response to post- Little Ice Age ice mass changes, including ice losses developed in the last decade or so. A plausible ice history for the last hundred years is made by using the actual measurements from 2002 to 2014, and 25% of the present day melting rate before 2002. We then simulate and fit the bedrock displacement – both vertical and horizontal - with a spherical compressible viscoelastic Earth model having a low viscosity shallow upper mantle. We show that we can constrain the shallow upper mantle viscosity very well and also explain most of the signal (amplitude and direction) by using 2 x10^18 Pa s. However we are not able to precisely constrain the thickness of the lithosphere (the preferred thickness is more than 50 km, quite thick for that region) or ice history. By using our preferred set up (earth model + ice history) we compute the GIA gravitational signature and convert it in equivalent superficial water density (see figure) that can be directly used to correct the mass changes observed by GRACE.
For the Amundsen sector the apparent trend of mass changes caused by the GIA fast mantle flow is more than 15 Gt/yr, which is 11.5% of to the -130 Gt/yr altimetry derived trend. Therefore, previous GRACE derived mass changes in Amundsen sector have been systematically underestimated.