Item

Abstract

We estimate the ice-mass loss in West Antarctica from the Gravity Recovery and Climate Experiment (GRACE) using the following approach. First, we predict the gravity-field change over Antarctica arising from present-day ice-mass changes, mainly derived from satellite altimetry, and the ongoing glacial-isostatic adjustment (GIA), which is computed by subjecting a viscoelastic earth model to a thermomechanical reconstruction of Antarctica's late-Pleistocene glaciation. We filter the prediction according to the current GRACE resolution using the Wiener optimal filter, and confine the spectral range of Stokes potential coefficients to that feasible for this regional investigation. We then use this prediction to appraise the temporal linear trends estimated from 4 independent GRACE gravity-field releases based on the Wiener optimal evaluation approach. This evaluation uses linear convolution filtering and the subsequent addition of multiple GRACE signals in order to minimize (in a least-squares sense) the difference between the prediction and the combined GRACE signal. We select the release identified as being in best agreement with the predicted gravity-field change over Antarctica and correct the data for GIA-induced gravity-field change. Finally, we adjust the prediction by modifying the mass balance of drainage basins mainly in West Antarctica using the L1- norm minimization criterion. We demonstrate that this approach, which uses a priori information about the behaviour of the ice sheet, allows improved estimates of regional ice-mass change to be obtained from GRACE.