Landslides
Combining geological mapping and instrumental observations for characterizing debris-flow source areas at Gadria, Italy
Project framework
Massive, abrupt sediment inputs – typical of debris flows and large floods occurring in steep channels dramatically transform the morphology of mountain areas. The sediment supply to the channel network is typically episodic, controlled by the interaction between geomorphic conditions and hydrological processes (Benda and Dunne, 1997). In the Alps, one of the largest, most recent perturbations of sediment supply has been the Last Glacial Maximum, which dramatically altered landslide activity and sediment yields during the postglacial period (e.g., Savi et al. 2014). In the Vinschgau-Venosta Valley (eastern Italian Alps, Figure 1a), intense debris-flow activity altered sediment continuity along the Adige River during the last deglaciation, imposing large-scale bed aggradation, confluence migration, and channel obstructions with the formation of temporary lakes and fan-delta systems (Brardinoni et al., 2018).
Understanding the effect of debris flows and bedload events on channel morphology, as well as quantifying sediment yield, is of paramount relevance for hazard assessment and design of mitigation measures. Climatic change represents a further challenge for recognizing current and future trends in sediment transport and thus design of adequate control structures. However, the identification of the source areas and the quantification of the sediment yield at catchments scale are difficult tasks that need both a deep knowledge of the geology and multi-parametric observations. Thus, long-term instrumental monitoring through catchment-scale sensor networks can provide precious information, especially if coupled with high-resolution topographical surveys. The measurement of rainfall permits linking the occurrence of the sediment fluxes to their most common triggering factor. The Gadria catchment (eastern Italian Alps, Figure 1) offers the opportunity to understand the main processes driving sediment supply at the catchment scale thanks to the detailed, ongoing monitoring activities (Coviello et al., 2021).
The main goal of this project is to investigate debris-flow source areas at Gadria in terms of (i) geology, (ii) sediment availability and (iii) degree of instability.
Supervisors: Velio Coviello (unibz / CNR IRPI, Italy), Francesco Comiti (unibz, Italy), Andrea Manconi (ETH)
Surface deformation and rock slope failure events at Moosfluh, VS
Project framework
The slow, long-term evolution of rock slopes is punctuated by failures producing rapid mass wasting. However, the spatial and temporal relationships governing the progressive development of rock slope movements towards failure are still poorly understood. Slope movements associated to deep-seated rock slope instabilities occur throughout the Alps and range generally from 1 to 10 cm/year (Crosta et al., 2013). At some locations, the surface displacements suddenly increase to higher values and the rock mass starts to degrade, tensile fractures form and the frequency of rock fall activity increases (Petley, 2004). A deep-seated slope instability located in the vicinity of the Great Aletsch glacier, in the area called “Moosfluh”, has shown during the past 20 years evidences of a slow but progressive increase of surface displacement. The moving mass associated to the Moosfluh rockslide affects an area of about 2 km2 and entails a volume estimated in the order of 150-200 Mm3 (Glueer et al., 2019) In the late summer 2016, an unusual acceleration of the Moosfluh rockslide was observed. In the period September-October 2016 maximum velocities have reached locally ~1 m/day, and such a critical evolution resulted in an increased number of local rock failures and caused the generation of several deep tensile cracks, hindering the access to hiking paths visited by tourists (Glueer et al., 2020). The surface deformation and the rock fall events have continued to be sustained also in the following years and are now back to the levels before the crisis.
Objectives
The main goal of this project is to investigate spatial and temporal correlations between surface displacement and rock slope failures at the scale of the Moosfluh slope. During the acceleration phase, rock fall events take place at different locations of the landside body, and involve different volumes. We aim at comparing/correlating surface deformation and processes observed with ground based remote sensing techniques with the occurrence, the location and the size of rock failure in order to better characterize the kinematic evolution of a failing rock slope over space and time.
Supervisor: Dr. Andrea Manconi
Rock slope deformation analysis in the Southern Swiss Alps
Project framework
Assessing the catastrophic potential of these phenomena is of vital importance in Alpine areas, due to their abundance, size and interaction with valuable elements at risk. Structural countermeasures are often not feasible or very expensive in these contexts, due to the size and the complex spatial and temporal evolution of the affected slope. A deep understanding of the geological and geomorphological predisposing factors, of the mechanisms and constraints on the evolution of such slow slope instabilities is thus required to set up scenarios and implement successful mitigation strategies. The occurrence of large rock slope failures in the Southern Swiss Alps is characterized from a high temporal variability, and governed by a combination of factors such as the shape of the valley, the slope morphology, the glacial history (thickness of the ice on the slope during the last glaciation and age of slope exposure after deglaciation), and the geological, structural and geomechanical local framework of each site.
Surface deformation is a key indicator to capture spatial and temporal changes and spatial behavior of these phenomena. Among all the available in-situ and remote-sensing monitoring techniques, satellite-based radar interferometry (DInSAR) allows nowadays to identify and map ground deformations, maximizing the spatial and temporal coverage at a relatively low cost. In Ticino and Graubünden Cantons, Switzerland, where several large rock slope failures have occurred in the past, active surface deformation is currently observed with DInSAR at several slopes. Instabilities included in this territory share a similar geological and morphostructural context, so, the different development time shown by phenomena appears to be affected mainly by the paleo-climatic (glacial and periglacial history) and/or specific geomechanical characteristics of each site.
Objectives and Methods
We aim at investigating one or more rock slope instabilities located in the Southern Swiss Alps and showing currently signs of active deformation. In this framework, it is possible to set up several MSc projects, each one with specific targets depending on the site. This will define also the combination of methodologies to be used, which will be discussed in the early phases of the project, by considering also the specific interests of the student. Possible sites are Cima del Simano (TI), Casaccia (GR), Peccia (TI), and Monte Crenone (TI). The work will span from field investigations to data collection and analysis, and possibly include also numerical modeling.
Supervisors: Prof. Christian Ambrosi, Dr. Cristian Scapozza and Alessandro De Pedrini (SUPSI), Dr. Andrea Manconi (ETH and GAMMA Remote Sensing), Dr. Tazio Strozzi (GAMMA Remote Sensing)
Analogue Modelling of Large Landslides
Introduction
Landslides are a natural hazard that kill about 4,000 people every year and cause billions of dollars in damage worldwide. New techniques to understand landslide mechanisms and quantify landslide risk can reduce losses from landslides. In many large landslides in rock and soil, two key questions must be answered to assess risk: 1) what is the volume of material involved in the failure? 2) What is the subsurface mechanism governing movement? This project addresses both of these questions using an analogue modelling approach.
Large landslides have distinctive 3D surface displacement fields, which are a function of their volume and movement mechanism. Several authors have recently proposed methods to use this field data to estimate the volume and movement mechanisms of large moving landslides, based on inverting subsurface characteristics from surface displacements. Validation of these methods is difficult, due to a lack of subsurface data. The goal of this project is to run analogue laboratory experiments to understand the relationship between surface and subsurface displacements.
Objectives
This project aims to design and test the feasibility of a new laboratory-scale analogue modelling approach for accurately measuring the surface displacements that result from a given landslide mechanism and volume. Subsurface movement mechanism and volume of landslides depends on their kinematic mode, which is controlled by factors such as slope geometry, planes of weakness and material type. The effect of all these factors on surface and subsurface displacements will be explored in the present project.
No special skills are required for this project, however an interest in landslides, laboratory work and processing of image data is considered an asset.
Supervisors: Dr. Jordan Aaron
Investigating traces of subaerial rockfalls in Swiss perialpine lakes
Introduction
It is documented that rockfalls have entered various Swiss perialpine lakes since de-glaciation. A part of them have even caused lake tsunamis (impulse waves) (e.g. Huber, 1982; Fuchs and Boes, 2010). Currently, an interdisciplinary project, funded by the Swiss National Science Foundation and the Swiss Federal Office for the Environment aims to better understand the key concepts of lake tsunamis from generation to inundation using Swiss lakes as field laboratories. Previous limnogeological studies on perialpine lakes have focused on the investigation of tsunamis generated by subaqueous mass movements and the respective hazard (Schnellmann et al., 2002; Kremer et al., 2012; Hilbe and Anselmetti, 2015; Strupler et al., 2018), not by impulse waves that are caused by subaerial mass movements (e.g. Fuchs and Boes, 2010). In order to better assess the rockfall-induced tsunami hazard on Swiss perialpine lakes, a solid knowledge on the spatiotemporal occurrence of past rockfall deposits in the lake basins is crucial.
Objectives
The proposed MSc Thesis is aimed at constructing a consistent database of rockfall deposits in Swiss lakes. High-resolution bathymetric and reflection seismic data allow the investigation of these rockfall deposits. Important research questions to be investigated include:
- What is the recurrence rate of subaerial rockfalls in Swiss perialpine lakes since de-glaciation?
- How large are the documented rockfall deposits? Are small rockfalls more likely to occur than large rockfalls?
Supervisors: Dr. Katrina Kremer, Dr. Michael Strupler
Fabric of Stable and Unstable Rockslide Basal Rupture Planes
Introduction and research questions
Assessments of the future evolution of unstable rockslides, for example at Brienz (GR), is of utmost importance for the people living in a runout area and the corresponding authorities responsible for the management of such landslide risks. The evolution of large unstable rock slopes can theoretically include (i) a stabilization, (ii) a continuation of slow creep movements, possibly connected to a progressive decay of the heavily fragmented rock mass into a long lasting series of rock fall events, or (iii) an acceleration of a large rock mass volume with the subsequent formation/transition into a very mobile and rapid rock avalanche (Figure 1). The main goal of this research project is related to the analysis of causes leading to the formation of very rapid rock avalanches at a few selected sites.
Many factors can influence the temporal evolution of rock slope instabilities. In theory the problem can be separated into an analysis of slope stability and an assessment of landslide dynamics, i.e. the landslide velocity evolution for a stage, when the slope has become unstable and unbalanced forces drive landslide acceleration (Alonso et al. 2010). When back-analyzing the stability of a failed rock slope, the resulting critical shear strength is normally much higher than what is required for the formation of a long runout of a very rapid rock avalanche. This implies that shortly after slope failure a substantial strength loss in the landslide has to occur in order to generate large unbalanced forces and a rapid acceleration (Hungr 2007). When only considering the basal rupture plane of a rock slope instability (not forces at lateral margins or inside the slide body) the following mechanisms could be responsible for significant strength loss:
- Breakage of rock bridges through shear or tensile rupture propagation
- Surface roughness reduction through shearing of asperities
- Shearing of clays
- Frictional heating
Frictional strength of the basal sliding plane is strongly depended on fabric, mineralogy and environmental conditions (saturation, pore pressure, temperature, stress).
In this project we will investigate the link between sliding mechanisms and fabrics at a few key sites in the Alps, based on detailed macro- and microstructural analyses of samples from basal shear surfaces of creeping rockslides and rock avalanches. Sliding plane materials are expected to have very similar fabrics and properties as creeping (aseismic) and seismic tectonic fault rocks: While most creeping tectonic faults are composed of thick gauges and breccia (Figure 2), earthquake ruptures often take place along very narrow ultraclasites or mirror like shear planes (Figure 3). The stability and velocity dependence of friction at the rupture plane with the potential for velocity weaking and hardening is a key for the understanding the potential dramatic accelerations of unstable rock slopes.
Supervisors: Prof. Simon Löw (main supervisor), Dr. Michael Plötze
Collecting and Analysing Timelapse Point Clouds of Moving Debris Flows
Introduction
Debris flows are extremely rapid landslides that can travel for long distances. Understanding and predicting debris flow motion requires knowledge of the mechanisms which lead to their catastrophic behavior. Three such mechanisms, longitudinal sorting, entrainment and liquefaction, will be investigated in this project. These mechanisms are shown schematically in Figure 1.
Longitudinal sorting leads to a concentration of coarse boulders at the flow front, which results in significant flow resistance. Liquefaction occurs in fine grained material behind the flow front, and the resulting liquefied slurry can build large flow depths behind the front, resulting in damaging surges. Entrainment involves the incorporation of path substrate into to flow, resulting in an increase in volume (e.g. Hungr, 2000; Iverson et al., 2000; Kaitna et al., 2016).
Although these mechanisms have been previously studied, many of their aspects are poorly constrained at present. High quality field data collected from moving debris flows would help to understand these mechanisms. The goal of this project is to collect and analyse such field data.
Objectives
The objective of this project is to apply novel sensor technology, developed for robotics and driverless cars, in order to measure flow attributes of moving debris flows. This dataset will then be analysed to provide insights into longitudinal sorting, liquefaction and entrainment.
Study area
The debris flow monitoring station will be installed in the Illgraben catchment, located in Valais and shown on Figure 2. Multiple debris flows occur in this catchment every year, and it features an extensive debris flow monitoring network which the new installations will complement.
Supervisor: Dr. Jordan Aaron
How does rain drive a dragon?
Short description
Rock slope failure often shows a progressive manner such that numerous small and intermediate-scale instability phenomena appear before the occurrence of a large-scale catastrophic event. For example, on 15th May 2012, a massive rockslide carrying a large volume (~210,000 m3) of debris occurred near the village of Preonzo in the Swiss Alps. Field monitoring and observational data indicate that before this avalanche, the slope continuously exhibited creeping behaviour over a long period of time as well as many instability features such as smaller-sized rockslides and countless rockfalls. During the slope destabilisation process, it exhibited episodic displacement rate accelerations often in coincidence with rainfall events, indicating a significant impact of rainwater infiltration on the progressive weakening of the slope. As the rock slope approaches catastrophic failure, the nonlinear interactions among its constitutes at different scales become very significant and the positive feedbacks may lead to large-scale coherent collective behaviours dominating the entire system, e.g. basal rupture plane formation. Such crises and extremes, called “dragon-kings” (a double metaphor for an event that is both extremely large and of a unique origin), tend to live beyond other smaller events. The objective of this MSc thesis is to understand the mechanisms that govern progressive rock slope failure leading to the emergence of dragon-kings and catastrophic failure, with a specific focus on the analysis of the potential correlation between slope responses and rainfall events. The research aims to provide quantitative insights into the mechanisms underpinning the observed complex slope phenomena in the field.
Supervisors: Dr. Qinghua Lei, Prof. Simon Löw, Prof. Didier Sornette
Application of the hydro-mechanical landslide triggering model STEP-TRAMM in test region for Landslide Early Warning Systems
Motivation
Rainfall induced shallow landslides are hazardous processes in mountainous regions that are difficult to predict due to the progressive nature of the triggering process at various scales and the spatial heterogeneities of soil hydro-mechanical properties. To mitigate damage caused by this ubiquitous process in steep areas, the federal office of the environment (FOEN) plans to develop and test an early warning system for rainfall induced landslides. As a test region, the ‘Napf’ area (~300 km2) was chosen. Most of existing early warning systems are based on rainfall thresholds, linking rainfall duration and intensity with the probability of landslide occurrence. These thresholds omit the role of the antecedent water content defining the amount of rainfall requested to cause potentially unstable conditions. The spatial distribution of the antecedent soil water content and the partitioning of the rainfall into surface and subsurface flow must be known to quantify load and soil strength patterns within slopes. This spatial distribution can be deduced from hydrological models. However, the spatial distribution of modeled water content and related soil strength and slope stability calculations does not reproduce real landslide patterns resembling random point patterns within a large catchment area. The contrast between the (large) simulated areas with similar stability index and the rare occurrence of observed landslides is related to the nature of the triggering mechanism involving progressive failures within a heterogeneous system. The evolution of a local failure to a mass release event (or to stabilization by stronger regions within a slope) depends on the complex interaction between constituting elements. These interactions and progression of damage are implemented in the model STEP-TRAMM that was successfully applied in the past to reproduce landslide patterns in small catchments of the test region ‘Napf’. However, these simulations were conducted with high spatial resolution (1 – 2 meters) that are too detailed to be applied at the scale of an entire region. Recently, the model was tested for larger areas with coarse spatial resolution for other areas to reproduce catastrophic events (Macao, 2017) and the effect of deforestation in various regions around the globe. Based on these studies at larger spatial scale, the application of STEP-TRAMM to simulate landslide patterns for a range of rainfall events in entire test area ‘Napf’ seems feasible.
Supervisors: Dr. Peter Lehmann, Physics of Soils and Terrestrial Ecosystems, ETH Zurich;Dr. Jordan Aaron, Engineering Geology, ETH Zurich; Adrian Wicki, Mountain Hydrology and Mass Movements, WSL; Dr. Manfred Stähli, Mountain Hydrology and Mass Movements, WSL.