Interannual variations of the mass balance of the Antarctica and Greenland ice sheets from GRACE

We propose a new estimate of the mass balance of the West/East Antarctica and Greenland ice sheets from GRACE for the recent period (July 2002 March 2005) and compute the corresponding contribution to the global mean sea level. We use new GRACE geoid solutions provided by the Groupe de Recherche en G\ eod\ esie Spatiale (GRGS/CNES), at the resolution of \~ 400 km and sampled at 10-day interval. In the three regions, significant interannual variations are observed, which we approximate as linear trends over the short time span of analysis. Over Greenland, an apparent total volume loss of 119 +/- 10 cu km/yr water is observed. For the Antarctica ice sheet, a bimodal behaviour is apparent, with volume loss amounting to 88 +/- 10 cu km/yr water in the West, and increase in the East amounting to 72 +/- 20 cu km/yr water. These GRACE results are affected by land hydrology contamination and glacial isostatic adjustment (GIA) of the solid Earth since last deglaciation. We correct for both land hydrology contamination (using a global hydrological model) and GIA using the ICE-4G model for Greenland and the IJ05 model for Antarctica. Corrected for both land hydrology contamination and GIA, GRACE volume rates are - 129 +/- 15 cu km/yr, - 107 +/- 23 cu km/yr and + 67 +/- 28 cu km/yr for Greenland, West Antarctica and East Antarctica respectively. In terms of sea level rise, the GRACE-based ice sheets contributions are + 0.36 +/- 0.04 mm/yr for Greenland, + 0.30 +/- 0.06 mm/yr for West Antarctica and - 0.19 +/- 0.07 for East Antarctica for the time interval of study. The total Antarctica contribution to sea level over this short time span is thus slightly positive (+ 0.11 +/- 0.09 mm/yr). The ice sheets together contribute to a sea level rise of 0.47 +/- 0.1 mm/yr. The results reported here are in qualitative agreement with recent estimates of the mass balance of the ice sheets based on GRACE and with those based upon other remote sensing observations. Due to the very short sampling time span for which the GRACE data are available, it is not yet possible to distinguish between interannual oscillations and long-term trend associated with climate change.