DEM DEVELOPMENT AND PRECISION ANALYSIS FOR ANTARCTIC ICE SHEET USING CRYOSAT-2 ALTIMETRY DATA

Digital elevation models (DEMs) are of fundamental importance to many geoscientific and environmental studies in Antarctic and due to relatively poor coverage by ground based surveys, the main data source for constructing an Antarctic DEM is satellite altimetry. The newest operating satellite-borne altimeter with ice applications is the ESA satellite CryoSat-2, which was launched in April 2010. CryoSat-2 provides altimetry data up to a latitude of 88°S/N, which is a significant improvement to previous satellite-borne altimeters. Based on two full cycles of CryoSat-2 observations acquired between December 2012 and January 2015, we derived a new DEM for the Antarctic Ice Sheet. The accuracy of generated DEM depends largely on the interpolation method adopted and five widely used interpolation methods were compared using the Cross Validation method. The Kriging method yielded better estimates for the Antarctic Ice Sheet and was adopted when constructing the final DEM. For the CryoSat-2 LRM data we followed an iterative approach to correct for the surface slope, and the slope correction was applied to each data point using the relocation method. Data gap beyond the latitudinal limit of the CryoSat-2 mission (88°S) was filled by contour data from Antarctic Digital Database (ADD). Finally, we present a new Antarctic DEM with a pixel size of 1 km×1 km. The accuracy of the final DEM was assessed by ICESat, IceBridge and GPS data and compared with four widely used Antarctic DEMs namely Bamber 1 km DEM, ICESat DEM, RAMPv2 DEM and JLB97 DEM. The results show that the CryoSat-2 DEM has an uncertainty of 0.73±8.398 m. The vertical accuracy of the DEM is better than 1 m at domes, better than 4 m for the ice shelves, better than 10 m for the interior ice sheet and over 150 m for the rugged mountainous and coastal areas.