Landslide Susceptibility Mapping Based On Landslide Classification And Weighted Information Value

Xining is located in the transitional zone between the Loess Plateau and the Qinghai-Tibet Plateau with strong tectonic activities,with fragile ecological environment and frequent landslides,resulting in huge property losses.For this reason,it is of great significance for local disaster prevention and reduction and planning and construction to carry out the evaluation of landslide susceptibility in the study area.On the basis of the actual field investigation,the paper first classifies the landslides in the study area,and summarizes the development and distribution characteristics of all kinds of landslides and their main influencing factors;finally,based on the landslide classification and weighted information model,the vulnerability of landslides is evaluated.The main conclusions are as follows:1.The field investigation shows that there are obvious differences in the development scale and distribution characteristics of different types of landslides in the study area.Accordingly,according to the composition of sliding materials,landslides in the study area can be divided into three types: loess landslides,mudstone landslides and gravel landslides.1)In terms of development scale,the development scale of mudstone landslides in the study area is the largest,followed by gravel landslides,and the development scale of loess landslides is the smallest.2)In terms of the distribution of landslides,Field investigation and GIS analysis show that loess landslides in the study area are mainly distributed in Quaternary loess and loess strata with a slope of 24?44° and slope height below 100 m,mudstone landslides are mainly distributed in Paleogene and Neogene mudstone strata with a slope of 16?36° and a relative height difference of 300 m and 400 m,and gravel landslides are mainly distributed in the alluvial layer with a slope of 24?40° and a relative height difference of 200 m and 300 m.2.Different types of landslides in the study area have different influencing factors,and the same influencing factors also have different action mode and intensity for different types of landslides.Among them,topography is the prerequisite for the occurrence of any type of landslide,and rainfall and human engineering activities are the inducing factors of landslide.on this basis,different types of landslides are developed under the influence of different factors.3.Based on the weighted information model and landslide classification,the vulnerability evaluation of landslides in the study area is carried out,and the statistical method and ROC curve method are used to verify the evaluation effect.1)The high incidence area of mudstone landslides is the largest,which is mainly distributed in the front of low mountains and hills strongly affected by human engineering activities;the high incidence area of loess landslides is medium and relatively scattered,and the poor topographic conditions on both sides of valleys and valleys in the hills may occur;the area of high risk area of gravel landslides is the smallest,but the distribution is concentrated,which is mainly distributed in the front of hills dominated by avalanche layers.2)Mudstone landslides and loess landslides may occur in areas with poor topographical conditions,strata and lithology dominated by mudstone and covered with thin layer of loess.mudstone landslides and gravel landslides may occur in hilly front high and steep slopes dominated by Paleogene mudstone.Loess landslides and gravel landslides have little spatial correlation and rarely occur in the same area.3)On the whole,the area of high incidence area of landslide is mainly distributed in the junction of hills and plains,and the area of high incidence of landslides in the north is larger than that in the south,and the intersection area of the three types of landslides is very few,indicating that the possibility of three types of landslides in the same area is very low.4)The evaluation accuracy of the two methods is 85.23% and 0.844 respectively,which shows that the vulnerability evaluation method based on landslide classification and weighted information model is feasible.