TOPS Model Data InSAR Deformation Monitoring And Registration Method

Landslide is a kind of recurrent,widespread and dangerous geological disaster,nearly two-thirds of regions are prone to landslides in China.With the changes in the natural environment and the impact of human activities the frequency of landslides is increasing,and the damage caused by landslides is increasing.In recent years,landslide monitoring technology has become the focus of people's research.For Guizhou,the problems of complex climate,vegetation and geological conditions have made it difficult to monitor the deformation of the landslide.The current landslide monitoring methods may have problems with high intensity and cost,insufficient monitoring accuracy and the low timeliness.Interferometric Synthetic Aperture Radar(InSAR)technology is an effective means of landslide monitoring with its all-day,all-weather,high-resolution and continuous space coverage.Sentinel-1 data is widely used in monitoring geological hazards because of its advantages of wide coverage,short revisit period,strong timeliness and free.Special attention is given to the coregistration of the Terrain Observation by Progressive Scans(TOPS),which poses strict requirements regarding the accuracy of better than 0.001 of the pixel.The complex mountainous terrain in Guizhou is easy to appear shadows and overlaps on SAR images,which brings great challenges to landslide monitoring.To solve these problems,this paper studies the method for TOPS images coregistration and the process of coregistration in Time series InSAS,and the technology of multi-data source combined landslide monitoring in mountainous area of Guizhou.Accordingly,the main research contents and achievements in the paper are as follows:(1)The basic principles and key technologies of InSAR are sorted out,the compositions of the interference phase are analyzed,and the phenomenon of layovers and shadows on the SAR image is analyzed from the image features.The principle,process and applicability of the two time series analysis methods PS-InSAR and SBAS-InSAR are analyzed.(2)According to the TOPS features of Sentinel-1,this paper studies the high-accuracy overall azimuth coregistration technology.The first coregistration approach consists on a geometric prediction of the shifts using precise orbit information and an external DEM and ICC method.Then we use ESD to eliminate phase jumps and achieve high-accuracy coregistration.Aiming at the temporal decorrelation caused in the processing of time-series InSAR,a coregistration processing chain for TOPS mode images is proposed.(3)We use ALOS PALSAR2,ascending and desending Sentinel-1 data to detect the potential landslides.ALOS-2's high resolution,high coherence and Sentinel-1A's high temporal resolution are complementary to each other.Through Sentinel-1A ascending and desending data,the monitoring voids caused by overlaps and shadows are effectively remedied,and some potential landslides in Guizhou monitoring area are successfully detected.