Deriving Greenland Ice Sheet Properties from TanDEM-X Mission Data

Abstract:
This paper presents for the first time the results of the classification of the different snow facies of the Greenland ice sheet based on interferometric radar data. The approach is based on the exploitation of X-band TanDEM-X interferometric synthetic aperture radar (SAR) acquisitions. Large-scale mosaics of radar backscatter and volume decorrelation, derived from the interferometric coherence, are used as input data set for applying a classification algorithm based on the c-means fuzzy clustering. The unique data set provided by TanDEM-X is particularly suitable for this analysis due to the single-pass bistatic acquisition mode which does not suffer from temporal decorrelation. The located snow facies can then be related to the well known dry snow, upper and lower percolation, and wet snow/ablation zones. The obtained results have been verified with external snow melt data and independent multi-temporal TanDEM-X data. The proper location of the different facies finally allows to estimate the X-band penetration depth into the ice sheet from the TanDEM-X volume decorrelation mosaics.