Spatial texture in AirSAR images of the Greenland ice sheet

This work presents studies of the spatial structure (texture) of SAR images of ice sheets and its role in their interpretation and classification. JPL, polarimetric AirSAR data of the southwestern part of the Greenland ice sheet is analysed using first order statistics, the power law semivariogram (PSV), and the grey level cooccurrence matrix (GLCM) techniques to determine the extent to which areas of wet snow and ice facies can be distinguished. First order statistics are not able to discriminate these classes. The PSV and GLCM techniques work at P band, especially in the case of the cross-polarized data.<
>