A Digital Elevation Model of the Greenland Ice Sheet based on Envisat and CryoSat-2 Radar Altimetry

Abstract:

With the launch of the first radar altimeter by ESA in 1992, more than two decades of radar altimetry data are now available. Therefore, one goal of ESA's Ice Sheet Climate Change Initiative is the estimation of surface elevation changes of the Greenland Ice Sheet (GrIS) based on ERS-1, -2, Envisat, CryoSat-2, and, in the longer term, Sentinel-3 data. This will create a data record from 1992 until present date. In addition to elevation-change records, such data can be processed to produce digital elevation models, or DEMs, of the ice sheets. The DEMs can be used to correct radar altimetry data for slope-induced errors resulting from the large footprint (e.g. 2–10 km for Envisat vs. 60 m for ICESat laser altimetry) or to correct for the underlying surface topography when applying the repeat-track method. DEMs also provide key information in e.g. SAR remote sensing of ice velocities to remove the interferograms' topographic signal or in regional climate modeling.

This work focuses on the development of a GrIS DEM from Envisat and CryoSat-2 altimetry, corrected with temporally and spatially coincident NASA ICESat, ATM, and LVIS laser data. The spatial resolution is 2 x 2 km and the reference year 2010. It is based on 2009 and 2010 data, the 2009 data adjusted to 2010 by accounting for the intermediate elevation changes. This increases the spatial data coverage and reduces data errors. The GIMP DEM has been corrected for negative elevations and errors in the north, and used to constrain the final DEM.

The recently acquired observations and increased data coverage give a strong advantage to this DEM relative to previous models, based on lower-resolution, more temporally scattered data (e.g. a decade of observations or only ICESat data, limited to three annual 35-day acquisition periods). Furthermore, as surface changes occur continuously, an up-to-date DEM is necessary to correctly constrain the observations, thereby ensuring an accurate change detection or modeling process.