Landslide Susceptibility Analysis on the Vicinity of Bogotá-Villavicencio Road (Eastern Cordillera of the Colombian Andes)

Landslide occurrence in Colombia is very frequent due to its geographical location the Andean mountain range, with a pronounced orography, significant geological complexity and an outstanding climatic variability. More specifically, study area around Bogotá-Villavicencio road central sector of Eastern Cordillera one regions highest concentration phenomena, which makes priority. An inventory detailed analysis 2506 landslides has been carried out, five basic typologies have differentiated: avalanches, debris flows, slides, earth flows creeping areas. Debris avalanches occur mainly metamorphic materials (phyllites, schists quartz-sandstones), areas sparse vegetation, steep slopes lower sections hillslopes; meanwhile, creep Cretaceous lutites, crop/grass lands, medium low lower-middle hillslopes. Based on this analysis, landslide susceptibility models made for different methods (matrix, discriminant random forest neural networks) input factors. The results are generally quite good, average AUC-ROC values above 0.7–0.8, machine learning most appropriate, especially forest, selected number factors (between 6 8). degree fit (DF) usually shows relative errors than 5% success higher 90%. Finally, integrated map (LSM) shallower deeper types movements. All LSM show clear zonation as consequence control susceptibility.