Effect of clouds on ERS SAR interferograms applied to land subsidence detection

We analysed ERS-1, 2 Synthetic Aperture Radar (SAR) interferometric data pairs from the Kanto Plains, which includes Tokyo, Japan. Fringe features that were independent of actual land deformations were found from the interferograms obtained. A SPOT HRV image acquired 18 min after the SAR data provides strong evidence for cloud perturbation of SAR data, because it reveals patterns of cloud distribution that are similar to the fringe features in the interferograms. GMS-5 visible-channel images indicate the direction of cloud movement, which supports the presence of cloud-induced phase delays, but the spatial resolution of GMS-5 data is not fine enough for comparison with the SAR data. Other interferograms indicate the land subsidence patterns that agree with known levelling processes. The extraction of information on land subsidence is possible from the area where the atmospheric effects are small. The results of this study suggest that optical sensor imagery with a spatial resolution equivalent to that of the SAR imagery is required for the evaluation of cloud-induced perturbations in radar signals.