Coseismic Deformation Monitoring Using Image Matching

D-InSAR is a new technology that was developed and widely used in surface deformation monitoring in recent years. But for large deformation caused by earthquake, D-InSAR is always unusable due to signal saturation or serious decorrelation. Image matching technology can be a very good complementary for D-InSAR. The imgae matching technology firstly obtains pixel offset through the precise co-registration between the master and slave image by the cross-correlation method, and then, according to the relationship between offset and deformation, extract the useful deformation. This paper will focus on the image matching technology and its application on monitoring the earthquake deformation.Firstly, this paper introduces the principle and theory of SAR image matching, including SAR images co-registration and the procedure of offset estimation. Secondly, to resolve the drawback of traditional methods for removing orbit offset, we present a new data processing chain of SAR image matching, which is simpler and does not need ground control points (GCPs). We applied this new chain to the 2008 Ms 8.0 WenChuan earthquake and 2010 Ms 7.1 YuShu earthquake and got satisfied results. Finally, in order to compare the SAR image matching and the optical image matching technology, we made a primary research on optical image matching. According to the automatic and precise co-registration of optical satellite images, we conducted an experiment by using two SPOT images, and obtained the NS and EW horizontal deformation caused by 1999 Hector Mine earthquake in California.