Feasibility of using UAV in surveying and 3D modeling of landslides (case study of Ardabil province)

Investigating areas geologically prone to landslides is essential in all construction projects, especially roads. The importance of this issue even doubles if the area in question includes multiple landslide cases. Various factors are effective in the occurrence of landslides, including geological and lithological factors, soil thickness and texture, loose rocks, vegetation cover percentage, topography, slope amount, weather, and groundwater. In this research, a landslide study is conducted based on drone observations and numerical modeling to specify the important slopes prone to landslides. Two unmanned aerial vehicles (UAVs) equipped with cameras and sensors, along with a set of high-resolution images, were used to identify and assess the surface of landslide-prone areas. First, digital surface models were prepared based on the amount of slopes, surface features of the land, the topographical profiles, and a three-dimensional (3D) model of the area with the help of drone data. Then, the geological characteristics of the area were investigated using the borehole data. After physical and analytical modeling of slope failure, these data were interpreted using specialized software packages: (1) Rocscience SLIDE for numerical analysis by limit equilibrium method and (2) and (3) PIX4Dmapper and Agisoft Photoscan for 3D modeling. Eventually, the prone areas and sliding mechanism were identified using the analytical models. The results revealed the high accuracy and ability of the proposed method in 3D and numerical modeling of landslides. Overall, this method allows for determining the landslide's dynamics, expansion, and surface.