Research On Radar Remote Sensing Landslide Hazard Identification And Deformation Monitoring In Yanyuan Region

Yanyuan county,which is located in the south section of Hengduan mountain area,is one of the most vulnerable areas in China's geological environment.Many faults in the area are developed vertically and horizontally,the lithology is broken,the Quaternary loose accumulation is widely covered,the terrain is steep and the height difference is disparate.Landslides,as the main type of disaster in the region,have become more frequent in the construction of infrastructure for surface reconstruction,resulting in loss of local personnel and property.Geotechnical displacement is the most intuitive feature of landslide.The conventional monitoring methods of surface displacement are often time-consuming and laborious,restricted by the external environment,can not meet the needs of large area landslide detection and high precision monitoring.Spaceborne radar time series interferometry can extract the features of wide-band surface deformation rate at mm/a levels,which is beneficial to slow creep landslide identification and monitoring.SBAS-InSAR technology,as one of the representatives of time series analysis technology,can eliminate the effects of space-time decoherence and atmospheric delay in traditional interferometry.Limited by the complex surface environment and radar imaging mode,this technology still has a series of problems that need to be researched in the application of regional recognition and multi-dimensional,high-precision monitoring of hidden dangers in mountainous areas.Therefore,based on the research of surface interference characteristics and SBAS-InSAR technology in Yanyuan County,this paper selects the wide-band single-look complex number products of Sentinel-1A data.Considering that the image registration accuracy is better than sub-pixel and the effect of phase deviation on the measurement accuracy in the interference processing,using fractional step registration,assisted by external data,to increase the reliability of the interference results,achieved a good landslide detection effect.Subsequently i carried out two-dimensional landslide deformation monitoring experiments on typical landslide mass,studied the landslide evolution characteristics and formation mechanism.(1)Using the RI index and the spatial decoherence model of slope and vegetation,the detection blind area and poor sensitive area of the study area are extracted.The applicability and constraints of appling the Sentinel-1A ascending track data to identify landslide in this area are systematically analyzed,and the difficulties inSBAS-InSAR data processing and corresponding solutions and strategies are discussed.(2)Using the data of 23 periods Sentinel-1A data that the time span is about 2 years,the SBAS-InSAR monitoring of Yanyuan surface is carried out,and the results of visibility and coherence are preserved.The region along radar uplift and subsidence and the corresponding deformation characteristics are obtained,and the deformation rate of Yanyuan surface is obtained.According to the relationship between radar observation direction and slope,the sliding rate range of landslide along slope is-147.86?0 mm/a.(3)The threshold of landslide motion is studied,the deformation points caused by suspected landslide are obtained,the spatial analysis of deformation points is carried out,and the 20 unstable slopes,58 new landslide points and 32 old landslide points with slow or slower deformation are quickly identified in combination with the remote sensing image characteristics and the development law of landslide historical data.The result has been verified by historical data,which proves the effectiveness of the technology.(4)Based on the analysis of the time and space characteristics of the landslide deformation development in glass village,the deformation characteristics of the landslide in vertical direction and sliding upward are extracted by fusion of the SBAS-InSAR treatment results of the landslide.The evolution process of the landslide is briefly analyzed according to the deformation-time curve,which provides the basis for the study of landslide mechanism and monitoring and early warning.