Topography from land radar altimeter data: Possibilities and restrictions

Abstract

Whilst satellite radar altimetry has been widely utilised over both ocean and ice surfaces for topographic mapping, applications over land have received relatively little attention. This is in part due to the complex nature of echoes returned from rapidly varying topographic land surfaces, which can cause an altimeter to generate erroneous range estimates. One approach to improving these data is to retrack using a single retracker, and construct a spatial average of heights obtained to give an estimate of mean orthometric height. This paper presents results obtained using an alternative approach: reprocessing returns at all levels of complexity through an expert system, which chooses one from a series of ten reprocessing algorithms based on an analysis of the return waveform shape. The selected algorithm then recalculates the range to surface, and hence derives an orthometric height. Utilising this approach with the geodetic mission dataset from ERS-1 has generated over 100 million height points with a near-global distribution. This paper presents selected results from this research using ERS-1 geodetic mission data together with ERS-1 and ERS-2 35 day data to demonstrate the accuracy to which orthometric heights can be determined, using global crossover analysis and comparison with ground truth. The paper illustrates applications of these data including validation and error correction of Digital Elevation Models, and discusses use and limitations of direct mapping with altimetry.