Application of L-band differential SAR interferometry to subsidence rate estimation in reclaimed coastal land

The subsidence rate in a reclaimed coastal land has been estimated by using JERS-1 L-band SAR two-pass differential interferometry. Measurement of subsidence rate during reclamation is difficult to obtain as a two-dimensional subsidence map based on sparsely distributed field measurements is required. Owing to the severe temporal decorrelation induced by frequent soil loading and intense deformation gradients, the L-band was found to be effective for observing subsidence in reclaimed land. We evaluated the accuracy of the estimated subsidence rate using field measurements obtained by a magnetic probe extensometer from 42 ground stations. A two-dimensional subsidence map was generated from seven qualified pairs. The correlation coefficient R between the two-dimensional radar measurements and the in situ data was 0.87 with a rms error of 1.42 cm. Two interferometric pairs obtained from an adjoining JERS-1 path were also used to verify the results. Independent estimations from the two different JERS-1 paths correlated each other with correlation coefficients R of 0.97 and 0.80. The main sources of the error were the reference DEM errors and additional phase noises calculated from phase fluctuation at stable points. The error from these two sources was +/- 1.27 cm. The estimated maximum subsidence was about 60 cm over 352 days. The results demonstrate that L-band differential SAR interferometry is a useful tool in geological engineering applications.