Engineering Application of Landslide Susceptibility Mapping in Linear Engineering Optimization

Abstract

Regional landslide susceptibility mapping has an important role in guiding linear engineering optimization. Based on the geographic information system (GIS) platform and R language MaxEnt software, in this research, the China National Highway 109 New Line Expressway (Beijing Sixth Ring Road-City Boundary Section) was taken as an example. Using the maximum entropy (MaxEnt) model, with ten evaluation indicators, including the elevation, slope, aspect, relief, topographic wetness index (TWI), roughness, lithology, distance to road, normalized difference vegetation index (NDVI), and land use type, a landslide susceptibility mapping was completed to support linear engineering optimization. The ROC value of the prediction model was 82.1%, and the %LRclass was 2.25, which strongly demonstrated the reliability of the landslide susceptibility mapping results. Then, the percentages of two roads with different landslide probability levels were calculated. It was found that only 10% of the new line was in a very high class, which was much lower than 26% of the existing line. This study proved that the regional landslide susceptibility mapping can be used to support optimization of the construction plan of the new line and provided a broader basis for decision-making for similar linear projects in future engineering projections.