ArcGIS-based Spatial Prediction Of Landslide Susceptibility In Changbai County

Changbai Korean Autonomous County is located in the area with frequent geological disasters in the northeast of China and the southeast of Jilin Province.The geological disasters are inseparable from the intensification of economic construction and intense human production activities.The geological ecological environment of Changbai County is getting worse and worse,and various environmental problems are more prominent.In this paper,the landslide in Changbai County is taken as the research object.Based on the geological conditions of Changbai County,the present characteristics,triggering factors and location characteristics of geological disasters in Changbai County are studied,and the influencing weight factors of landslide susceptibility are discussed and graded and quantified.Based on Arc GIS platform,information method,logistic regression method and decision tree C4.5 algorithm are used to evaluate the landslide susceptibility in Changbai County.The center result of this paper are as follows:(1)There are 549 landslide geological disasters in Changbai County.The scale of various types of disasters is relatively small.Landslide geological disasters mainly occur in the road area built by cutting slopes artificially in Changbai County,mainly in the section from Changbai County to Qidaogou along Yalu River National Highway,the section from Changbai County to Songjiang in S302,and other sections with steep terrain and towering cliffs.(2)Twelve influencing weight factors of landslide geological disaster susceptibility are selected,including surface elevation,slope inclination,slope plane bending rate,slope profile bending rate,relative position with roads,rivers,faults,rock and soil classification characteristics,vegetation index(NDVI),terrain humidity index(TWI)and precipitation,which are analyzed according to disaster points and various influencing weights(3)Based on Arc GIS platform,using the information method,logistic regression method and decision tree C4.5 algorithm,the landslide geological disaster prone degree zoning of the study area was carried out,and the landslide prone degree zoning map of each method was generated.ROC curve is used to test the accuracy of the zoning evaluation results of landslide geological disaster susceptibility by three methods.