Research And Application Of Potential Landslide Hazard Identification Method Based On Multi-Source Remote Sensing

As one of the main kinds of geological disasters,landslides are widely distributed and destructive,which cause huge losses of people and property every year.In the alpine and canyon areas,the frequent landslides are caused by strong tectonic activities,complex stratigraphic lithology and changeable climatic conditions.Therefore,early identification and hazard assessment of landslide features are an important part of disaster prevention and mitigation.The continuous development of remote sensing technology provides more diversified technical means for geological disaster investigation.In this paper,through multi-source remote sensing information,which is combining InSAR deformation measurement technology with optical remote sensing,with high-resolution UAV images as supplementary data,the identification method and application technology process of landslide hidden danger are explored.On August 3,2014,a Ms6.5 earthquake occurred in Ludian,Yunnan,which triggered a large number of secondary collapse and landslide.Under the triggering action of the Ludian earthquake,the early ancient landslides may have reactivation or increased activity intensity.Focusing on the nature and intensity of mountain activity on both sides of NiuLan River basin,the annual average surface deformation rate was obtained by using Sentinel-1 data to identify potential landslide sites.Multi-temporal satellite and UAV high-resolution images will further analyze the interpretation and deformation characteristics of landslide hazards,as well as quantitative and qualitative observation and research of landslide hazards.Through the analysis of the spatial distribution characteristics of the identified landslide hidden dangers,combined with the regional geographical and geological conditions,the risk assessment of the regional landslide hidden dangers is realized.The main results of the paper are as follows:(1)The remote sensing identification features and methods of landslide hazards are summarized and sorted out.Based on the landslide elements and combined with the topography and geological structure background of NiuLan River Basin,the interpretation signs and factor characteristics of landslide hazards in remote sensing images are analyzed,including the regional surface deformation characteristics obtained from SAR images.And the morphological structure,fracture development,tonal change,hydrological phenomenon and other features of landslide in optical image.(2)InSAR technology was used to analyze the time series deformation of the study area.SBAS-InSAR processing was performed on sentinel-1 images of 121 KMT T128 and 110 KMT T062,respectively,to obtain the annual average surface deformation rate diagram along the line of sight of regional satellites.Combined with the optical image,the abnormal surface deformation caused by human engineering activities,land subsidence,physical weathering and other factors is excluded,and the hidden danger point of obvious surface deformation in the region was obtained.(3)Based on the hidden danger optical image interpretation signs extracted from the landslide elements,the regional hidden danger of landslide is interpreted,and the hidden danger of landslide is analyzed from the perspective of deformation characteristics and shape trace characteristics.The complementary results of SAR and optical image interpretation can make up for the defects of single remote sensing method in terms of scale and Angle.Field data were collected for key areas with potential landslide hazards.Ortho images of UAV were collected for an area of about 150 square kilometers,and 449 LiDAR images were collected for Hongshiyan Hydropower station.UAV high-precision monitoring provides data for the review and analysis of landslide hidden danger points extracted by optics and SAR,and obtains details of morphological characteristics of key deformation areas.(4)Based on the identification,quantitative analysis and spatial distribution of landslide hidden danger points,regional landslide risk analysis is carried out.Elevation,gradient,slope direction,stratum lithology and fault distance,distance from drainage for impact factors,the information model for regional landslide hazard risk assessment for the regional risk zoning map,the majority area in the middle of danger,concentrated distribution in high-risk groups on both sides of the steep terrain NiuLan River valley area.(5)With NiuLan River Basin as an example,using multi-source remote sensing techniques,including optical satellite,radar,satellite,aerial drones,LiDAR technology applications,from multi-source data acquisition,data processing,information processing,modeling analysis and the quantitative identification and assessment of landslide,etc.,to build the multi-source remote sensing technology of landslide hazard identification,technical scheme and the working process It is proved that multi-source remote sensing is necessary to identify and evaluate the hidden danger of landslide.This paper summarizes the technical process and method of comprehensive remote sensing investigation of large-scale landslide hazards in alpine and canyon areas,and provides reference for regional disaster prevention and mitigation.