Abstract:

Investigation of high-level and highly concealed landslides is challenging and difficult to be reached by manpower, with low efficiency and high risks. The use of aerial remote sensing technologies such as drone optics and airborne LiDAR can effectively overcome these problems, and has been extensively applied to today’s research. However, currently most of those research and applications focus on macroscopic investigations and interpretations, with only limited research on refined identifications and measurements of the development factors, such as landslide cracks and landslide walls. In order to further refine the interpretation features of the landslide elements based on drone optics and airborne LiDAR recognition, and improve the comprehensive application of this technology, this study selected the Wenchuan Longxigou high-level landslide in the active area of earthquake in western Sichuan as an example. The methods include drone digital photogrammetry, airborne LiDAR distance measurement, interpretation comparison and spatial measurement, as well as on-site investigation and verification. In total, twenty eight tension cracks, two shear cracks, and eight steep slopes were extracted. Two landslide walls, five leading edge boundaries, and five secondary deformation zones were divided based on the arrangement and combination of the landslide elements into two secondary sliding zones and four deformation zones. The validations are highly consistent with the indoor interpretations, proving the reliability and accuracy of this proposed method. The research results summarized and compared the differences in color tone, texture, and spectrum of landslide elements on different data, explored the collaborative extraction method of ‘color tone texture map’ of landslide elements using unmanned aerial vehicle optics and airborne LiDAR, and elaborated on the fine identification processes of high-level landslides by ‘interpreting elements first and then zoning and combining’. The results provide technical references for the fine identification and prevention of other high-risk high-level landslides, and have a significant value in the applications.

Key words: UVA, airborne LiDAR, high level landslide, element identification, Longxigou of Wenchuan