Mass Balance of East Antarctic Ice Sheet: Reconciling ICESat Altimetry with GRACE Gravimetry and Long-term Ice History

Mass balance results for the East Antarctic (EA) ice sheet range from gains of 136 Gt/yr in 1992-2009 from ERS-radar and ICESat-laser altimetry to gains of 45 Gt/yr in 2003-2009 and 70 Gt/yr in 2009-2016 from GRACE gravimetry. The principal residual uncertainty lies in the corrections: 1) for vertical motion of the underlying bedrock for altimetry and 2) the change in gravity caused by crustal motion and mantle flow for gravimetry. Both of these corrections are derived from numerical models of the Earth's crust and the underlying mantle flow (e.g. Ivins et al., 2013; Whitehouse et al., 2012) that calculate the response to the ice loading/unloading history. The correction to mass changes calculated from changes in ice-sheet surface elevation is simply the modeled vertical motion (dB/dt) of the bedrock times the relative density (0.91) of ice. The local change in gravity sensed by the satellite is a combination of the local change in ice thickness and the sub-ice mass of the Earth's crust and underlying mantle, for which a correction commonly called the Glacial Isostatic Adjustment (GIA) is modeled. Over long times, the primary GIA factor is viscous flow of the deep mantle with density 5.5. Therefore, the ratio of the corrections for gravimetry and altimetry tends to be the ratio of their densities (6 to 1). We calculate the spatial distribution of the ratio from two Earth models. Using the spatial distribution of mass changes from GRACE and ICESat for their 2003-2009 overlap and the corrections ratio, we calculate the additional dB/dt to bring GRACE and ICESat derived mass changes into agreement. Preliminary results using a spatially-averaged ratio of the modeled correction for EA indicate convergence of GRACE and ICESat at respective additional mass gains of 141 Gt/yr and 24 Gt/yr corresponding to an average error in dB/dt of 2.53 mm/yr (additional downward motion). For comparison, the conclusion (Zwally et al., 2015) of an average 15.9 mm/yr long-term ice growth in EA (i.e. a new finding of ice loading generally not included in the loading/unloading history in Earth models) would imply a slightly larger additional dB/dt of - {15.9 mm/yr x 1/6} = - 2.65 mm/yr. Other ice-history findings indicating long-term ice growth in EA since the LGM include accumulation increases in ice cores and Holocene glacier advances in the Dry Valleys.