Experimental study on the phase statistics of GBSAR imagery

Abstract

Synthetic Aperture Radar (SAR) interferometry is a well-established remote sensing technique that compares the phase of two or more complex SAR images for highly accurate measurement of displacements. Recently, the ground-based interferometric SAR (GBSAR) has proven to be a powerful tool for monitoring land deformation such as landslides, glacier movements and volcanic hazards. By using GBSAR, it is possible to detect subtle changes in the order of a fraction of the radar wavelength. However, the accuracy of the interferometric phase measurement is greatly affected by the degree of decorrelation between the SAR images that have been acquired from slightly different times or at different positions. The major contributors of phase decorrelation include radar receiver noise and temporal scene decorrelation. In this paper, the phase statistics of a typical GBSAR instrument are studied. The closed-from probability density function of the phase statistics is first derived for random media in the presence of homogeneously distributed scatterers and when strong scatterers are present within a resolution cell. The phase deviations are analyzed and compared with the experimental results obtained from a GBSAR instrument in an aneachoic chamber and at outdoor test sites.