Monitoring Landslides Deformation Along Three Gorges Reservoir Bank With Multi-source SAR Data

Three Gorges area which has the special terrain distribution is a region with lots of geological disasters.After the completion of the Three Gorges Dam,accompanying with the rising of water level and the influence of human activities,some ancient landslides in the area had been activated which caused more potential hazards.The emergence of landslides poses a huge threat to the shipping of the Yangtze River and the life and property of people living along the Yangtze River.Traditional monitoring methods such as leveling have high accuracy but low monitoring efficiency,which is not suitable for a large range of monitoring scene.As a new radar imaging technology,synthetic aperture radar interferometry(In SAR)has gradually been applied to landslide detection and monitoring in different scenes with the launch of a new generation of high-resolution SAR satellites and the accumulation of available data.In this paper,Terra SAR-X satellites and Radar SAT-2 satellite images covering the Badong and Wushan town areas in 2019 were collected,At the same time,the Sentinel-1A satellite data sequence covering Huangtupo landslide was also collected,The D-In SAR and time series analysis results of radar line-of-sight deformation were carried out for some typical landslides along the river in this area.Gongjiafang,Baiheping,and Huanglashi landslides were analyzed using D-In SAR technology.Huangtupo landslide was monitored by SBAS-In SAR time series analysis technology using Sentinel-1A data.The main contents are as follows:(1)The progress of radar interferometry and its application in the field of landslide deformation monitoring were summarized.The deformation results of Gongjiafang,Baiheping and Huanglashi landslides in the differential interferogram of ascending and descending images were analyzed.And we compared the results of deformation under the different conditions,such as,coherence and different viewing angles,slopes and aspect angles.According to the distribution of the spatial and temporal baselines of the ascending and descending images,we choose the short baseline or ultrashort baseline interferogram images as analysis emphatically,and the motion trends and stability of the three landslide bodies in 2019 were analyzed.(2)Based on the sentinel-1A satellite's 084 ascending orbit image data collected from August 2016 to March 2020,In SAR timing series analysis technology was usedto obtain the deformation results of Huangtupo landslide surface.According to Huangtupo's geological structure,5 major deformation regions(A,B,C,D,E)on 4sub-landslide bodies were taken as the analysis objects,16 PS feature points were selected from which the cumulative deformation trend on different sub-regions was compared and analyzed regionally.The results show that region A,which is located at the posterior boundary of Linjiang No.2 landslide and the front boundary of substation landslide,and region C,which is located at the posterior boundary No.1landslide,Garden spot landslide and the west boundary of substation landslide,had a large deformation rate,while region E,which is located at the rear boundary of substation landslide,had the lowest deformation rate.(3)The projection conversion formula from radar line-of-sight(LOS)to slope direction of landslide is derived.Based on DEM data of 30 m resolution,the slope direction distribution of Huangtupo landslide body were preliminarily obtained,combining the imaging geometric parameters of the Sentinel-1A satellite,the distribution map of conversion factor(cos?)between the line-of-sight and slope directions of the Huangtupo Landslide was calculated,and then,we carried out quantitative statistical analysis about conversion factor(cos?).(4)Based on the rainfall grid data from China National Meteorological Information Center,the monthly rainfall value on the Huangtupo landslide was obtained by interpolation.The correlation between rainfall and landslide variables was analyzed,and further quantitative statistical analysis was conducted.preliminary correlation analysis results show that rainfall was not the main factor causing Huangtupo landslide deformation.