Predicting landslide stability, runout, and failure velocity at Cook Lake landslide, Wyoming
Abstract
Proper understanding of risk is predicated on accurate characterization of hazard. Landslide hazards are common, but are difficult to characterize without a full site investigation, which are often unavailable. This thesis develops a framework for obtaining inexpensive field and laboratory data to characterize three major components of landslide hazard: current stability, runout distance (L), and failure velocity. The framework is designed to analyze deep-seated, soil landslides, and is evaluated using data from the Cook Lake, Wyoming landslide. The framework consists of four tools. Two of these are flowcharts that predict both catastrophic runout and velocity. In line with runout predictions, the third tool estimates the ratio of landslide drop height to runout length, H/L, from measurements of planimetric area. The final tool utilizes back analysis and forward modeling to evaluate a slope’s stability and its sensitivity to hydrologic changes, relying on multiple analyzed sections and sensitivity analysis in lieu of detailed subsurface data. The two flowcharts were developed by synthesizing literature about landslide geometry, local topography, and pore-water pressure. The H/L estimation tool correlates the planimetric area of well-characterized landslides to landslide mobility. The stability modeling tool uses fieldwork, direct shear testing, and limit equilibrium analyses to characterize the behavior of the geologic materials. Comparison of the backward model, which evaluated conditions that triggered a 1997 landslide at Cook Lake, to the forward model revealed that portions of the landslide are more unstable than they were in 1997. Estimates of H/L for a future reactivation range from 0.17 to 0.25. The runout flowchart also indicates that the landslide is likely to exhibit H/L > 0.1, though unexpected contractive soil behavior might cause long runout. The velocity flowchart indicates that the landslide is not likely to move faster than 1.8 m/hr. The combination of these slower velocities and shorter runout indicators would allow evacuation of a nearby campground at Cook Lake during a future reactivation. Other similarly vulnerable sites can benefit from these tools.Rights
Copyright of the original work is retained by the author.Collections
Related items
Showing items related by title, author, creator and subject.
-
Development of a landslide risk rating system for small-scale landslides affecting settlements in Guatemala CityMany settlements in the Guatemala City Metropolitan Area (GCMA) face significant possibilities of fatalities due to their location in steep ravines that are subjected to periodic large-scale landslides. Since the housing in the at-risk areas is relatively low-cost, it is typically cost-prohibitive to mitigate the risk to an acceptable level. Thus, permanent relocation is the only truly viable option to ensure the long-term safety of everyone. However, there are several economic and social obstacles impeding successful implementation of a relocation program. Still, there are many other landslide risk reduction techniques (such as retaining walls, community drainage systems, and alert systems) implemented by government organizations and non-profit groups. These techniques are helpful in landslide risk reduction (LRR), but residents are only partially involved in the entire process. Therefore, increasing residents’ education and ability to better understand their level of landslide risk will help with LRR. Residents can integrate and collaborate with the government organizations and non-profits implementing mitigation techniques and of even greater benefits, the education and ability for residents to understand their landslide risk can provide additional avenues for LRR not otherwise achievable. The purpose of this research is to develop a landslide-risk-rating-system (LRRS) that can be used by trained residents to better understand their risk (similar to other landslide or rockfall hazard-rating systems commonly used by department of transportation organizations). The focus of this LRRS is only on small-scale landslides (typically the size of a house or less) because evaluating the risk of large-scale landslides is too complicated to be done by trained non-technical experts. The LRRS asks questions related to landslide risk that can be used to calculate a landslide risk score to indicate the relative level of risk. The LRRS was created by reviewing published literature documenting other landslide rating systems and incorporating similar factors correlated with landslide risk. Then, forty sites were visited in the GCMA to inventory the factors at houses that are vulnerable to landslides in order to evaluate which factors were most useful for predicting the relative risk. The predicted risk scores were compared to ranked risk scores estimated by the author to ensure the results were valid. Statistical analysis identified which of these factors best-predicted landslide risk. These factors include slope angle, slope height, strength of slope material or material type, aperture of cracks, spatial impact, largest probable landslide volume, largest probable percentage of the living area that could be impacted from a landslide, and total person-hours a living area is occupied per day. Future work should focus on the transformation of the tool into a more user-friendly format for use by residents, the implementation process, and monitoring plan.
-
Mapping and analysis of the Rio Chama landslide and evaluation of regional landslide susceptibility, Archuleta County, ColoradoRecent landslides, such as the West Salt Creek landslide in Colorado and the Oso landslide in Washington, have brought to light the need for more extensive landslide evaluations in order to prevent disasters in the U.S.. The goal of this research is to characterize and map the Rio Chama landslide, evaluate conditions at failure, predict future behavior, and apply these findings to create a regional susceptibility model for similar failures. Based on the classification scheme proposed by Cruden and Varnes (1996), the Rio Chama landslide is an active multiple rotational debris slide and flow complex with observed activity since 1952, located near the headwaters of the Rio Chama River in south-central Colorado. Site reconnaissance was conducted in 2015 and 2016 and coupled with laboratory testing of samples and limit equilibrium stability analysis. A hierarchical heuristic model using an analytic hierarchy process was applied to evaluate the susceptibility of the region to failures similar to the Rio Chama landslide. Weights were assigned to parameters based on their influence on landslide susceptibility, and weighted parameters were combined to produce a regional susceptibility map. The causative factors in order of most to least contribution to susceptibility are: slope angle, lithological class, surface waterbody density, stream density, slope aspect, profile curvature, and land use. The regional model accurately and consistently identifies zones as having low, moderate, and high susceptibility to failures similar to the Rio Chama landslide. Discussion of the causative factors and their impact on susceptibility, as well as the implications of each susceptibility zone is presented. The regional model is intended to identify areas susceptible to failures similar to the Rio Chama landslide, allowing appropriate preventative action to be taken.
-
Evaluation of shallow landslide-triggering scenarios through a physically based approach: a case study from Bulathsinhala area, Sri LankaTransient Rainfall Infiltration and Grid-based Regional Slope-stability analysis (TRIGRS) is a regional, physically based stability model which could be applied to predict shallow landslides. It is important to evaluate the accuracy of TRIGRS for the prediction of landslide locations using actual events before use the TRIGRS model for further applications. This study presents the application of TRIGRS for Bulathsinhala area, Kaluthara in south western part of Sri Lanka where the number of shallow landslides occurred on 26th May 2017 and many of those events transitioned into damaging and killing debris flows. A back analysis of that landslide event was executed to authenticate the model by using different methods and techniques for the definition of the input parameters. Reliability of the model was evaluated through comparison with the 2017 landslide inventory in the particular area and it was revealed that most of the actual landslides were occurred in the predicted area (FS<1) of the model. In order to quantify the effectiveness of the model, an index was proposed in the study called LRclass (landslide ratio for each predicted FS class). The obtained values of the LRclass index realize the trustworthiness of the model which indicates the considerably higher value (60%) for the lowest stability class. With this particular manner, the output of the study could be used to implement more reliable land use management and development plans and resettlement procedures. Further the TRIGRS model is advantageous for susceptibility mapping and landslide flow path analysis, particularly when linked with various advanced applications using GIS spatial functions.
