Application of stratified kriging to ERS-1 and ERS-2 radar altimeter data to investigate ice elevation changes in the Lambert Glacier/Amery Ice Shelf system, East Antarctica

Sea level rise due to climate change is sufficiently well established. Sea level change is affected by a variety of different factors. One of the main factors is the change of the accumulation and melt rate on polar ice sheets which varies the polar ice mass balance. About 90% of the polar ice masses reside in Antarctica which makes the Antarctic mass balance highly relevant for predictions of sea level rise. The area of interest in this thesis is the Lamber Glacier/Amery Ice Shelf system. Its catchment area encloses about 10% of East Antarctica. Therefore it is representative for a significant part of the Antarctic ice sheet. Elevation changes of the ice surface indicate ice mass balance changes. Conventional mass balance studies do not account for the spatial behaviour of these changes. With the methods presented in this thesis it is possible to show the spatial distribution of mass balance changes. The elevation of the ice surface can be measured over wide areas in short time periods using satellite radar altimetry. To determine significant elevation changes time series of about five to ten years are necessary. In this thesis radar altimeter data of ERS-1 and ERS-2 over a time period from 1992 to 2002 are used. To get accurate elevation measurements geophysical, atmospherical, and instrumental height corrections are applied. These corrections still comprise inaccuracies and diverse factors that affect measurement quality can not be corrected for. This limits the applicability of these data. The change of elevation within the Lambert-Gletscher/Amery Eisschelf (East Antarctica) is calculated with 35 day time resolution for digital elevation models with 3 km grid spacing. These elevation models are calculated by a special interpolation technique (stratified kriging) based on the radar altimeter data. The area of interest consists of several morphological homogenious regions that are divided by steep slopes or ridges (discontinuities). In contrast to other interpolation techniques these discontinuities can be accounted for by stratified kriging. Using this method calculation of the spatial correlation (variography) and interpolation are accomplished separately within each region. Neighbouring regions are connected afterwards. The delineation of units is accomplished by the analysis of gaps between data as well as by a split moving window algorithm. To diminish the influence of a global spatial trend during interpolation bicubic smoothing splines are used fo