A study on PS-InSAR and D-InSAR for DDM determination in hilly areas in Taiwan

The phase information of the SAR(Synthetic Aperture Radar) image is used to evaluate a large-scale DDM (Digital Displacement Model) in a short time by D-InSAR (Differential Interferometric SAR). However, there are many practical problems to be solved somehow, especially in Taiwan, due to dense vegetation cover and quick variation on weather which would drop coherence. PS-InSAR (Permanent Scatterers Interferometric SAR) adopts numerous multi-time SAR images to find out permanent scatterers having high coherence. Taking these permanent scatterers to be unwrapped with multi-dimension algorithm can evaluate successive displacement along a time series. There is dense vegetation cover in hilly areas in Taiwan. The quality of DDM determined by D-InSAR is not good because the coherence would be critically dropped in these hilly areas. However, some permanent scatterers, not adopted by D-InSAR, can be applied by complex algorithm of PS-InSAR. Taking these permanent scatterers to be unwrapped with multi-dimension algorithm can evaluate successive displacement along a time series. The accuracy of PS-InSAR can achieve mm level better than cm level of D-InSAR, in which the cities and towns are often the computation areas. PS-InSAR is usually used to study the subsidence with micro-variation. The 921 earthquake occurred in central Taiwan is an experimental case. The ERS-1 and ERS-2 images are used to operate PS-InSAR and D-InSAR. The DDMs determined by PS-InSAR and D-InSAR are compared and some practical problems are analyzed. In this paper, PS-InSAR with the improved method for selecting candidates of permanent scatterers can estimate displacements on permanent scatterers, which can't be done by using a general method.