Towards a Realistic and Exhaustive Error Estimate of Volume Balance of the Antarctic Ice Sheet from Radar Altimetry
Flament, Thomas1; Rémy, Frédérique2
1LEGOS, FRANCE; 2LEGOS / CNRS, FRANCE
Radar altimetry has strong assets as a method for estimating volume change of the Antarctic ice sheet: it provides regular and dense spatial coverage and its record is now long, since the beginning of the 1990's. The use of an along-track processing scheme further improves spatial resolution of the time series. But converting the along-track measurements of elevation change to continental volume change is not straightforward. We focus on the overall estimate of the error on volume change; we do not tackle the problem of firn densification and conversion from volume to mass change. Here we propose to investigate several contributions to the error and estimate their impact on the volume balance of the ice sheet: sparse spatial coverage at lower latitudes, poor performance of classic altimetry in coastal/mountainous regions and varying influences of corrections of the radar wave penetration depth, interpolation schemes and relocation methods of the gridded elevation trend estimate. The spatial coverage of reliable measurements in particular is a problem: coastal regions features the largest elevation change signals but are the least well sampled one, both because of the ground track pattern and the radar performance over steep terrain. The along-track processing allows determining a decorrelation radius and estimating an appropriate interpolation distance. A better understanding of the radar penetration depth is gained by comparing Envisat measurement with ICESat measurements. On this matter, we hope to obtain further progress from measurements in the Ka-band (35.75 GHZ) by the AltiKa altimeter onboard SARAL. We will propose a volume change estimate for the Envisat period with an error estimate as complete as possible.