Height change detection in Antarcita using satellite altimetry data and Kriging / Kalman filtering techniques

Abstract

Studies of the response of ice sheets to climate change require data sets with high accuracy and uniform ice-sheet coverage. The most common technique used in analyzing satellite altimetry data to study height change in the ice sheets is the dH/dt technique based on the cross-over geometry. In this thesis two alternative techniques to cross-overs are studied and applied to all ERS radar and ICESat laser altimetry data to study height change in Antarctica. The first technique, block kriging, uses all available data to build a statistical model of the elevation field. Results of height rate of changes dh/dt ill Antarctica for the years 1995-2001 produced using block kriging and cross-over analysis are compared. In the Amery Ice Shelf and in the West Antarctic coastal area and near latitude -81°N, the difference in dh/dt between the two methods are statistically significant. A second technique combines kriging and Kalman filtering to allow for time evolutions of the height change rate and other parameters used in the description of the surface elevation field. An application of the technique to laser altimetry data from the current NASA's ICESat mission shows the potential of the technique in detecting height change.

(cont.) In addition, the method can potentially characterize surface elevation in Antarctica at small horizontal scales of the order of the laser footprint size. The quality of laser altimetry data at the present is not sufficient for height change detection at the accuracy level of 2cm/yr required for studying mass balance in Antarctica.