Potential of interferometric SAR for classification of land surfaces

Coherence maps obtained from repeat-pass interferometric SAR can be used to provide information on properties of land surfaces together with information on the backscattering coefficient and its temporal variation between the two acquisitions. Different effects degrading the coherence are described and discussed. A coherence texture measure is introduced. Properties derived from a coastal area including boreal forest, snow covered fields, marshlands and fast ice affected e.g. by wind and temperature changes close to 0°C are reported for different baselines and temporal delays. It is illustrated how the coherence information is sensitive to different properties than the backscattering coefficient and its temporal change and combining the information will improve classification of land surfaces