Landslide Susceptibility Evaluation Of Lulang-Tongmai Section In Sichuan Tibet Highway Based On GIS

The research area is located in the southeast of Qinghai-Tibet Plateau,the geological environment is complex and the tectonic activity is strong,which makes the landslide disaster very develop.This paper introduces the geological environment background of the research area by collecting the background data of the basic geological environment of the research area.Based on Google Earth software,landslide disasters are identified and analyzed,as well as the factors affected by landslide disasters.Seven evaluation factors,such as elevation,lithology,distance from river,distance from road,distance from fault zone,vegetation and rainfall,were selected,and based on Arc GIS platform,the information quantity model,logic regression model and spatial logic regression model were selected to evaluate the landslide susceptivity of the study area.The main conclusions are as follows:(1)The landslide disaster in the study area was interpreted according to the remote sensing interpretation mark of the landslide disaster,and a total of 189 landslide disasters were interpreted in the study area,with a distribution area of 83.97km~2.Landslide disasters in the study area were mainly down-layer landslides,and the scale of landslides was dominated by medium and large-scale landslides.In the five sections of Dewu-Reba,Reba-Layue,Layue-Polong,Polong-Tangdui and Qucun-Suotong,the development is more concentrated.(2)Based on the selection principle and correlation analysis of evaluation factors,seven evaluation factors,such as elevation,formation rock,distance from river,distance from road,distance from fault zone,vegetation and rainfall,were selected,and the landslide proneness analysis was carried out by using information quantity model,logic regression model and spatial logic regression model.The accuracy test of the evaluation results of the three models is carried out,and the accuracy of the three models is the spatial logic regression model,the logical regression model and the information quantity model from high to low.which shows that the development of landslide disaster in the study area has considerable spatial effect.(3)The results of the landslide prone evaluation of the spatial logic regression model were used as the final evaluation results,which showed that the area of the very high prone area was 179.04km~2,the area of the high prone area was 125.64 km~2,the area of the medium prone area was 117.77 km~2,and the area of the low prone area was147.61 km~2.The very high-prone areas and high-prone areas are mainly distributed in the southwestern part of the study area,mainly along rivers,on both sides of roads and near fault zones.The medium and low prone areas are mainly distributed in areas with higher elevation,less human engineering activities and high vegetation coverage.