Research On Landslide Identification Method Of Datong County In Qinghai Based On SBAS-InSAR Technology

China is a country with a vast territory and frequent geological disasters.There are many kinds of geological disasters,widely distributed and harmful.Landslides are a common geological hazard with a wide and relatively concentrated geographical feature.The study area is located in the transitional zone between the western end of the Loess Plateau and the Qinghai-Tibet Plateau.It is one of the most fragile areas in China.With the continuous transformation of surface and surface layers of human activities in the region,geological disasters such as landslides occur frequently.Surface deformation is the most direct manifestation of landslides.Generally,the deformation of the landslides is very slow in the early stage.Although it does not directly pose a threat to humans,it will cause the building to crack and even become unstable and slippery as it is slowly deformed.Caused significant losses.SBAS technology can eliminate the effects of time-space disassociation and atmospheric effects in D-InSAR technology,and can acquire millimeter-level deformation information,which is very suitable for monitoring the slow surface deformation of long-term sequences.However,how to successfully identify landslides in complex mountainous areas with SBAS technology is still under investigation.Based on the Sentinel-1A data,the surface deformation monitoring and early identification of landslides were carried out based on SBAS-InSAR technology using Datong County,Xining City,Qinghai Province,which is located in the transitional zone between the western end of the Loess Plateau and the Qinghai-Tibet Plateau.The DEM data and satellite parameters were used to analyze the study area from the topography and RI index respectively.The results were basically consistent,and the good visibility area obtained by the RI index and the low sensitivity area obtained by the terrain were retained.Deformation information.Based on the results of the visibility analysis,the unreliable deformation points of the forest land,shrub land and river surface are screened out.Since the moving direction feature of the landslide is roughly similar to the slope direction before the landslide occurs,we further convert the deformation rate of the obtained radar line of sight direction to the maximum slope direction,and study the motion threshold of the deformation point;Statisticalanalysis was carried out on the obtained credible deformation points,and a total of142 cold spot areas and 137 hot spot areas,namely 279 potential landslide areas,were obtained.Combined with the surface deformation information obtained by SBAS technology,the image characteristics of the landslide in the study area and the impact factors,279 potential landslide areas were further analyzed,and 144 potential landslide points were finally identified.According to the survey conducted by the Qinghai Geological Survey in the economic belt along the Qinshui River in Qinghai Province in April 2018,21 of the 144 potential landslide points are unstable slopes,22 are new landslide points,and 14 are old landslides.87 is a non-landslide point.Finally,two typical landslides in the study area were selected for analysis.The research results show that SBAS technology is very feasible for the monitoring and early identification of large-scale terrain complex mountain landslides.It can predict the location of landslides on the regional scale and accurately define the landslide range.