The Use of Lidar Data and VHR Imagery to Estimate the Effects of Tree Roots on Shallow Landslides Assessment

Abstract

The study of geo-hazards has been benefited from the technological advances in the field of Remote Sensing (RS) techniques as the ALS (Airborne Laser Scanners) Systems with Very High Resolution (VHR) cameras. Recently, the LiDAR (Light Detection and Ranging) is an active sensor technique used for a variety of geoscientific applications including slope monitoring to retrieve ground surface displacements at high spatial resolution. Additionally, LiDAR has been widely used in order to collect high-resolution information on forests structure for the determination and characterization of vegetation cover due its ability to capture multiple returns and to reach the ground, even in forested areas, allowing the generation of Digital Terrain Models (DTMs) for the estimation of forest variables. In this paper, a LiDAR dataset and VHR imagery from aerial survey was used in the southwest zone of Medellín City-Colombia where the most frequent landslides are shallow and triggered by rainfall. Slopes with gradients up to 30% on residual soils characterize the study area, having about of 30% of forest cover consisting predominantly of Eucalyptus and Coniferous forests. For the estimation of the tree roots effects on the shallow landslides assessment on a natural slope, interpolation processes were developed from the LiDAR 3D point cloud, obtaining DTMs of 1 m-pixel. Additionally, orthophotos with the same spatial resolution were acquired in the aerial campaign. The proposed workflow was implemented on a GIS platform, and considers the extraction of the tree heights by generating a Canopy Height Model (CHM), while for the delineation of the tree crown a process of image segmentation was developed. Once the vegetation has been characterized using LiDAR products and dendrometric relationships, the Limit Equilibrium Method (LEM) was used to evaluate slope stability considering the effect of vegetation (trees). The results indicate that the proposed workflow allows to obtain adequate stability indicators for the estimation of tree roots contribution and additionally, this RS technique allows saving resources in this kind of analysis.