The Application of Emerging Monitoring Technologies on Very Slow Vegetated Landslides

Geohazard monitoring is becoming increasingly important alongside increased expectations for the protection of the public. Technological advances in the field of remote monitoring and instrumentation has allowed for an economically efficient means of data collection. Traditional methods of instrumentation have often required expensive, time consuming, and intrusive monitoring methods using permanent instrumentation, such as slope inclinometers and shape-acceleration-arrays. These traditional methods often require site access for large borehole equipment and drilling and installation crews with advanced training. Modern technologies can collect information over large spatial extents, and forms of data which would be impossible or extremely expensive to obtain using traditional methods. In this thesis, the use of differential global positioning systems (GPS), terrestrial light detection and ranging (LiDAR) laser scanning, and unmanned aerial vehicle (UAV) photogrammetry for improving landslide monitoring is analyzed. The use of these technologies is well proven, but there are technical limitations of these technologies regarding landslide velocity and vegetation. This thesis focuses on investigating these limitations, methods in which these limitations can be overcome, and the knowledge we were able to obtain from application of these technologies to a Very Slow (As defined by Cruden and Varnes 1996), vegetated landslide. This work is performed with the aid of a study site, called the Chin Coulee landslide, in Southern Alberta, Canada. The Chin Coulee landslide is a large deep-seated, Very Slow, vegetated landslide, and provided a challenging testbed for the study of the limitations of these modern technologies. It was found that differential GPS systems work well in slow moving conditions, although short-term, month-to-month data sampling would be inadequate for accurately representing landslide movement. Water conditions in and around the site, including overland erosional flow, internal groundwater flow resulting in seepage along the slope, and in the case of Chin Coulee, the reservoir elevation, are relevant to landslide movement and vary throughout the year. To fully understand landslide movement, it is recommended that at the very least, a full calendar year study be performed, with 2-3 years of study often being more appropriate for fully understanding the mechanisms which lead to instability. As only one year of monitoring was available, it is difficult to identify the true impact events have on landslide stability. Limit equilibrium analysis shows support for reservoir drawdown decreasing landslide stability on Chin Coulee, with up to 8% reduction in factor of safety from 1.06 to 0.98 being observed during a historical critical drawdown scenario. Application of terrestrial LiDAR to slow-moving, vegetated landslides posed several challenges, most notably the detectable limit of movement. During slow moving conditions, without extended monitoring periods, movement will often be under the detectable level of movement, referred to as level of detection (LOD). The required duration between scans depends heavily on site and scanning conditions. Scans performed on highly vegetated sites from long distances will increase the LOD. In the case of Chin Coulee, it was not possible to bring the LOD below 50 – 70 mm. Due to the slow movement rate this LOD necessitated a monitoring window of roughly 12 months. UAV photogrammetry was used for feature tracking of erosional channels and headscarp locations for comparison to historical information collected in 1998. Identification of increased erosion channel growth and headscarp movement was possible. A novel application of photogrammetry was the creation of a 3D model based on air photos collected in 1982 following landslide initiation. Change detection using this 3D model and a current day LiDAR scan was performed to observe the evolution of the landslide over the past 36 years. This helped to support the proposed failure mechanism for the Chin Coulee landslide. UAV-based change detection was performed using two photogrammetry models of Chin Coulee. UAV photogrammetry was limited on Chin Coulee due to the inability of photography to penetrate vegetation. The achievable LOD for this change detection was calculated at roughly 90 mm. UAV photogrammetry-based change detection appeared to show exaggerated movement in some regions, suggesting model inaccuracy. Identification of the limitations of these modern technologies is an important step for adoption into the field of geotechnical engineering. Due to these limitations these technologies are not yet suitable for all conditions and purposes but provide strong monitoring options when viable.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.