Synthetic aperture radar (SAR) interferometric baseline parameters form important input for SAR interferometry. In this paper, a nonlinear error model is established for the SAR interferometric baseline and parameterized as a polynomial based on the natural nonlinearity of the orbit of a satellite. Unlike conventional models, the proposed model takes into account the nonlinear part of the baseline error. A theoretical derivation is performed based on the imaging geometry of interferometric SAR, and the results of the analysis show that the parameters of the nonlinear baseline error model can be obtained from the relationship between the orbit, the nominal baseline, the baseline error, and the residual interferogram phase. A sample data set from the Japanese Earth Resources Satellite-1 (JERS-1) L-band SAR is used to validate the proposed model, and the results indicated that the compensation of the residual interferogram phase of the test data is superior to that provided by conventional models.

Interferometric SAR is an emerging earth observation technique, especially useful in cartography and surface subsidence survey, Differential SAR Interferometry (DInSAR) provides an alternative and complementary method to investigate the land subsidence phenomena, which can provide two dimensional deformations on the line of sight (LOS) of radar on areas of thousands square kilometers. In order to obtain accurate surface subsidence, precise orbit must be provided so as to compensate the reference surface phase and topographic phase. This paper proposed a new method for SAR interferometric baseline rectification specifically for SAR satellites those have no precise orbits. The method based on the imaging geometry of interferometric SAR, during the processing we setup a coordinate system called RXA, where R is a vector directing from the scene center to the corresponding position on the orbit, A is the vector of satellite velocity, X is normal to both R and A to construct a right hand coordinate system. There are two steps in our baseline rectification process, which are perpendicular baseline rectification and parallel baseline rectification, and the singular value of phase frequency was removed by the DIA method. In our study two InSAR pairs are processed, the results show the new proposed method could compensate not only the linear part of the orbit error but also nonlinear part of the orbit error.

European Remote Sensing satellites ERS-1 and ERS-2 synthetic aperture radar (SAR) data are used to analyze terrain deformation in the area of Wieliczka, Poland using interferometric techniques. Terrain deformations in Wieliczka are due to (i) salt mining, (ii) catastrophic suffusion caused by water inflows to the mine and (iii) landslides in the southern part of the region. To discriminate between different deformation regimes differential SAR Interferometry (D-InSAR) and persistent scatterers interferometry (PSI) are used. A comparison with published information based on leveling data is performed. It can be concluded that rapid subsidence caused by catastrophic water inflow cannot be identified with InSAR techniques. However, both D-InSAR and PSI retrieve information about slow subsidence due to mine convergence. These data match with subsidence maps constructed from leveling data. A limited amount of coherent scatterers were identified on landslides. Some scatterers can be related to buildings with clear damage caused by landslide movement.