ID12: Climate Change Impacts on Alpine Mass Movements

Abstract ID 607 | Date: 2022-09-14 10:10 – 10:20 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Mcardell, Brian W. (1); Hirschberg, Jacob (1); Munch, Jessica (2); Meyrat, Guillaume (2); Bartelt, Perry (2)

Sediment cascade problems are characterized by a general lack of knowledge of the initial and boundary conditions. Additionally, the timing of past events can only rarely be constrained, and future events cannot be predicted with certainty. While sophisticated 3D multi-phase slope stability and mass-movement runout models may include accurate descriptions of the dominant processes in each part of the sediment cascade, such models are computationally expensive and consequently only a few simulations can be performed in the time available for engineering analyses. This limits the ability to accurately predict runout, especially when considering multiple scenarios describing the initial and boundary conditions. E.g. a difference of a factor of two in the initial volume of a rock avalanche may dramatically influence the runout of the landslide and formation of subsequent debris flows. Considering various plausible volume-, friction-, ice- and water-content scenarios may necessitate many 10’s of simulations to constrain the maximum plausible runout distance of a complex sediment cascade event such as at Piz Cengalo in Switzerland in 2017 or Chamoli in India in 2021.

Post event calibration of an historical rock avalanche is relatively straight forward. In general, one assumes that the initial mass instantaneously fails and disintegrates into a flowing landslide, and then the friction is typically calibrated to best match observed runout patterns. Prediction prior to an event is a much more challenging task, especially considering that the initial volume and mode of failure (e.g. one coherent landslide, or a series of retrogressive failures) cannot be accurately predicted.

The additional complexity of coupling several single-process models, or using single models with internal process transitions (e.g. the transformation of a landslide to a debris flow), without detailed knowledge of initial and boundary conditions, further reduces the ability to provide meaningful predictions within a reasonable amount of time. Therefore, we advocate the use of computationally inexpensive 2D depth-averaged flow models for runout prediction including entrainment of sediment, water, ice, and snow along the flow path, and separate models for predicting subsequent down-slope effects such as debris flows.

In this contribution we describe our efforts to develop a modelling framework, whereby estimates of sediment availability along the flow path are coupled with a simple hydrological model to estimate the water content of the deposits. These models now include rock, ice, snow and water phases, while including other relevant processes such as phase-changes (ice-water) and the formation of an air-blast.

Abstract ID 301 | Date: 2022-09-14 10:20 – 10:30 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Kenner, Robert (1); Mott, Rebecca (1); Bartelt, Perry (1); Bazargan, Mohsen (1); Cicoira, Alessandro (1); Gaume, Johan (1,2); Hirschberg, Jacob (1); Kyburz, Michael (1); Lehning, Michael (1); Mcardell, Brian (1); Sovilla, Betty (1); Weber, Samuel (1)

The rock slope instability at Spitze Stei located in the Swiss Alps, near the village Kandersteg, might trigger one of the largest cascading processes chain in the recent history of the European Alps. The geological disposition causes a long-term retrogressive rock slope instability, whose past activity is assumed to have been influenced by climate change. In case of failure, up to more than 12 Mio m³ will impact an older rockslide deposit and may also affect the nearby Öschinensee lake, a popular tourist attraction. Immediate debris flow activity, affecting populated areas downstream, has to be expected following a rock slope failure.

We analyse possible process chains initiated at this site. This analysis will start from a high-resolution characterization of the meteorological-climatological forcing which includes both, short-term extreme weather and long-term climatological changes. Snow cover, precipitation and long-term temperature evolution are crucial factors for the initiation and progression of cascading mass wasting processes.

The destabilisation mechanism of the rock slope, considering the role of changing permafrost, will be described by a thermo-mechanical model. Laboratory experiments on rock samples collected at the instability are used to determine rock mechanical parameters. A kinematical model of the slope provides the geometrical information on the instable rock compartments. Resulting failure scenarios will be used as an input to simulate the dynamics of the following rock avalanche using two different modelling approaches (RAMMS & MPM), including entrainment of snow, and (variably saturated) sediments. The MPM approach will be further used to assess the potential impacts of the rock avalanche on the lake.

The project will also investigate the conditions leading to the initiation of debris flows following a rock avalanche event. This analysis considers the potential sediment liquefaction during the impact on all sediment deposits along the flow path. Possible debris flows scenarios will be simulated using RAMMS, with a special focus on the contribution of sediment entrainment under different weather conditions such as degree of saturation, streamflow, or snow cover along the flow path.

Our investigations are embedded within the CCAMM project cluster Cascading Processes, which addresses the hazard risk caused by process chains in alpine regions. We want to improve knowledge on the initiation, dynamics, and controlling factors of sediment cascades related to rapid mass movements. Hereby, the example of the Spitze Stei rock instability is used to establish exemplary analysis concepts and models that can be used to anticipate future cascading events.

Abstract ID 263 | Date: 2022-09-14 10:30 – 10:40 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Ortler, Marcel (1); Brauer, Achim (2); Fabbri, Stefano C. (3); Kowarik, Kerstin (4); Kueck, Jochem (5); Reschreiter, Hans (4); Strasser, Michael (1)

Intramountainous regions are highly vulnerable to climatic changes, global warming and natural hazards with their cascading effects. To holistically understand past environmental changes, frequencies and impact of natural hazards, it is crucial to extend our limited observational and instrumental data with long high-resolution archives (e.g. archeological sites, swamps or lake sediments). The Hallstatt-Dachstein region represents an alpine environment, where the interconnections of geohazards, climate change and evolution of human-environment relations can be tracked over the last 7,000 years.

The study site Lake Hallstatt is located within the center of the UNESCO World Heritage Cultural Landscape Hallstatt-Dachstein/Salzkammergut, Austria, a region with one of the oldest histories of human salt mining worldwide. The Hipercorig Hallstatt History (H3) project reveals a high-resolution sedimentary archive with two parallel cores (core A: 41m, core B: 51m) to study geo-hazards (e.g. earthquakes, debris flows or floods), climate changes and anthropogenic imprints.

We present a lake stratigraphy based on non-destructive core logging data, visual core and lithofacies description and age modelling using 14C dating, incorporated with borehole logging (of hole B, magnetic susceptibility and natural gamma spectrum) and Core-Log-Seismic-Correlation. The core logging involves (i) x-ray computed tomography, (ii) multi-sensor-core-logger data (gamma density, magnetic susceptibility and color spectrophotometry), and (iii) xrf-scanning data. The stratigraphic succession comprises at least four major mass-movement deposits, six >1 m thick turbidite deposits and multiple >5 cm thick flood deposits during the Holocene.

The Holocene and Late Pleistocene stratigraphic record of the H3 long-cores will enable the first integration of archaeological studies covering 7,000 years of human salt mining, including documented prehistoric large catastrophic landslides (dated to: ~1061 BCE and ~662 BCE), mass movements or heavy perception. Also, the H3 core will allow to better assess prehistoric mitigation strategies of natural hazards (e.g. river diversion), showing exceptional high resilience of the local prehistoric community towards natural hazards. This will improve our understanding of the early development and environmental imprint of one of the oldest cultural landscapes worldwide.

Abstract ID 734 | Date: 2022-09-14 10:40 – 10:50 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

De Vugt, Lotte (1); Zieher, Thomas (1,2); Moreno, Mateo (3,4); Steger, Stefan (3); Rutzinger, Martin (1)

For the hazard and risk assessment of landslides in mountainous regions, it is important that the underlying models are reliable. Current methods of landslide susceptibility assessment, especially dynamic physically-based slope stability modelling approaches, still face the problem of time-consuming calibration for new study areas and new hydro-meteorological scenarios. Calibration is necessary with each new assessment, as the geotechnical and hydrological parameters of the soil and bedrock are usually not available in full detail and vary significantly over different locations. In addition to this, the model has to be calibrated to account for generalizations, as most physically-based models can only approximate the actual processes behind slope stability. Recent studies on physically-based landslide modelling investigated the use of input parameter ensembles for their models to account for the naturally occurring variability of soil properties within a study area and reduce the time needed to calibrate the models. Comparison of these approaches with calibrated single input combination approaches showed a significant increase in the performance of such models. Current research activities focus on the question whether the use of parameter ensembles improves the spatial transferability of a model and how a well-performing ensemble could be found in a computationally efficient way. This study, framed within the PROSLIDE project, aims to answer these questions by comparing the performance of a model and a parameter ensemble from an existing study in Vorarlberg, Austria with a new spatial slope stability assessment using this ensemble in the Passeier Valley, Italy. The main goal is to find a robust parameter ensemble that can be used to accurately predict landslide susceptibility under different hydro-meteorological scenarios and at different locations. A secondary goal is to elaborate a method that reduces the time needed for defining such an ensemble. The study builds upon the model TRIGRS (Transient Rainfall Infiltration and Grid-Based Regional Slope-Stability Analysis), which is one of the most commonly used physically-based models for dynamic slope stability assessments. The implementation of the method developed in this study enables faster assessment of landslide susceptibility in new areas and under different hydro-meteorological scenarios. Further research on improving the performance of physically-based landslide models, should investigate how this method can be used in combination with geotechnical or soil maps to implement ensembles with spatially varying input parameters. Additionally, future research should investigate how these new methods can be applied to more complex landslide models that better approximate the processes behind landslides, e.g., STARWARS/Probstab.

Abstract ID 794 | Date: 2022-09-14 10:50 – 11:00 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Hormes, Anne (1); Ostermann, Marc (2); Plörer, Matthias (3); Amabile, Anna Sara (2); Vecchiotti, Filippo (2)

Deep-seated gravitational slope deformations (DSGSDs) commonly encompass entire slopes from the crest areas to the valley floor and are often accompanied by a series of subsequent processes such as debris flows, landslides, and rockfalls. In most cases, a DSGSD is dissected into several slabs that may have different dimensions and deformation rates. Especially in glacially oversteepened valleys, spontaneous failure events occur in areas relatively close to the valley bottom.

On Christmas Eve in 2017, a rock avalanche with a volume of about 117,000 m³ occurred in the Vals valley in Tyrol, burying the state road underneath and nearly reaching inhabited houses. The detachment area is located in the lower section of a DSGSD, the Windbichl DSGSD, that covers at least 4.3 km² and extends over a height difference of about 1000 meters. Since 2018, the province of Tyrol has been monitoring the detachment area of the Vals rock avalanche using various methods (periodic TLS tachymetric continuous measurements, extensometers), which resulted in about 10-15 mm/year horizontal deformation in the WSW direction and about 14 mm/year of horizontal deformation in the SSW direction with a vertical component of about -4 mm/year.

About 1.5 km east of the 2017 rock avalanche is the Horlicher Wand, a steep to overhanging rock face up to 450 meter high. The Horlicher Wand is also part of the Windbichl DSGSD. In an integrated approach, we analyzed the hazard potential of the entire Windbichl DSGSD with special focus on the Horlicher Wand.

The current state of deformation of the Windbichl DSGSD was assessed by analyzing 2D InSAR Sentinel-1 (2017-2019) in combination with detailed geomorphological mapping. For drone photogrammetry of the Horlicher Wand (2021), semi-automatic kinematic analysis of potential release areas on the rock face was validated with field data and runout simulations with RAMMS: rockfall.

InSAR data indicate vertical displacements of 0.8 cm/year and horizontal displacements of 1.5 cm/year. The 3D model of the rock face shows several unstable and overhanging parts susceptible to rockfall processes. Semi-automatic kinematic analysis identified wedge failure and flexural toppling/falling as main processes, which were also confirmed by field data. To investigate climate impacts, such as extreme precipitation on DSGSD movement patterns, longer InSAR observation time series and a basic engineering geology understanding of the underlying processes must first be investigated to filter possible climatic triggering and the geologic antecedent conditions.

Abstract ID 867 | Date: 2022-09-14 11:00 – 11:10 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Voelk, Maria Sophia (1,2); Ortner, Gregor (1,2,3); Bruendl, Michael (1,2); Christen, Marc (1,2)

Various studies suggest that changes in the climate system such as temperature rise and extreme precipitation events may influence gravity-driven hazards. Within the WSL research program “Climate Change Impacts on Alpine Mass Movements”, we develop a framework to model mass movement hazards and risk altered by climate- and socio-economic changes in Switzerland.

For rockfall under current climate conditions, we use a high-resolution terrain model, a soil classification layer, an algorithm for automatic detection of potential release points and a forest layer as input for the RAMMS::LSHIM Large Scale Hazard Indication Mapping method to develop hazard indication maps for scenarios with a 30y, 100y and 300y return period.

To address possible changes on rockfall due to climate change, we additionally consider data of the CH2018 climate change scenarios, information on permafrost degradation at the release areas and altered forests in the transit zones to show the likely influence on rockfall run-out.

These hazard indication maps are taken as input into risk assessment models to produce risk maps, which depict spatial and temporal changes of rockfall risks based on the combination of hazard, exposure and vulnerability information. These maps can support decision making processes and will allow to define risk adaptation measures considering climate change impacts.

Abstract ID 613 | Date: 2022-09-14 11:10 – 11:20 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Hermle, Doris (1); Gaeta, Michele (2); Krautblatter, Michael (1); Mazzanti, Paolo (2,3); Keuschnig, Markus (4)

Optical remote sensing analyses of mass movements are required for future alpine safety. High spatiotemporal UAS (unmanned aerial system) data can be employed, using digital image correlation (DIC), to derive ground motion. This enables to investigate the evolution of mass movements and relate them to influencing processes caused by climage change impacts.

This study compares the effective detection and monitoring potential of image registration techniques of the area–based phase correlation algorithm, implemented in COSI–Corr, and the intensity–based dense inverse search optical flow algorithm, performed by IRIS, for UAS data. The dataset consists of seven high accuracy orthophotos of 0.16 m resolution acquired between 2017 and 2021. We studied mass wasting processes of the Sattelkar complex landslide, situated in a steep, glacially–eroded, high–alpine cirque (2,130-2,730 m asl), Austria. The cirque infill is characterised by massive volumes of glacial and periglacial debris, remnants of a dissolving rock glacier and rockfall deposits. The latter are continuously fed by low magnitude high frequency rockfalls from the surroounding headwall of granitic gneiss. Since 2003 there is an increase of dynamic processes and between 2012-2015 rates up to 30 ma^-1 were observed. After ongoing heavy precipitation in August 2014 a 170.000 m³ debris flow was triggered. It is assumed that high water (over)saturation causes spreading and sliding of debris on the glacially smoothed bedrock floor.

Displacement calculations from both algorithms provide knowledge about the extent and internal zones of the mass movement and are qualitatively supported by manually traceable boulders (<10 m). For phase correlation excessive ground motion and surface changes limited the signal to 12 m because of decorrelation and ambiguous displacement vectors. In contrast, optical flow returns more coherent displacement rates with no upper motion limit but some underestimated zones. Increases of motion in our displacement calculations and acceleration curves can be associated to observations of high precipitation in summer 2020 as well as the strong precipitation event in August 2021. We show that UAS data provides trustworthy, relative ground motion rates for moderate velocities, thus enabling us to draw conclusions regarding internal landslide processes.

Abstract ID 579 | Date: 2022-09-14 11:20 – 11:30 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Moos, Christine; May, Dominik; Dorren, Luuk; Schwarz, Massimiliano

Forests in mountain regions play a crucial role in protecting settlements and infrastructure against natural hazards. At the same time, more frequent and severe disturbances due to climate change may lead to a temporary reduction in the provision of the protection service of forests. In this study, we analysed the long-term dynamic of the protective effect of forests against rockfall after severe disturbance events for different rockfall and site characteristics. We therefore modelled stand growth of the three dominating tree species F. sylvatica, P. abies and A. alba using long-term data of the Swiss National Forest Inventory (SNFI). We then quantified the potential energy dissipation capacity of forests as a function of stand growth based on a rapid rockfall assessment tool. Finally, we analysed the evolution of the protective effect for varying rockfall dispositions. The results show that the potential energy dissipation capacity of the stands mainly depends on the altitude, slope and the species. While a high protective effect against small blocks (0.1-0.5 m³) is regained already in the first 20-30 years after a disturbance, the recovery of the protective effect takes substantially longer for larger blocks (> 1m³). After ~200 years of stand growth, the protective effect decreases again for small blocks because of a limited number of tree stems for sufficient impacts. The study allows for the quantification of the long-term dynamics of the protective effect of forest against rockfall for a wide range of site and rockfall dispositions. This is of particular importance in the face of increasing disturbance frequency and severity due to climate change.

Abstract ID 492 | Date: 2022-09-14 11:30 – 11:40 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Chmiel, Małgorzata (1,2); Walter, Fabian (1); Belli, Giacomo (3); Marchetti, Emanuele (3)

In recent years, the rock slope near Spitze Stei has exhibited elevated displacement rates that exceed 10cm per day, which suggests a growing instability of 20 million m³. At the Spitze Stei slope, first-hand observations of rock falls from the slope terminus have been linked to the recently increased slope dynamics. The deformation of the Spitze Stei slope is heterogeneous: the difference in lateral velocity induces the formation of rock compartments separated by deep-reaching fractures. Moreover, borehole imaging and geological mapping show the presence of degrading permafrost and suggest active glide planes and shear zones that contribute significantly to slope displacement.

Climatic factors (such as freezing-thawing cycles, rain, and temperature changes) can cause mechanical changes within the slope and possibly at gliding planes that affect slope stability. Seismic waves propagating within the slope are sensitive to these mechanical changes, and we can measure the changes through seismic interferometry. At the same time, mass movements, such as slope failures, generate both seismic (when impacting the Earth surface) and infrasonic (primarily when the air is pushed aside by an accelerating mass) waves. We can use these seismo-infrasonic signals to detect in time and space the mass-movement events and characterize their source mechanism.

Here, we aim to understand how climatic factors affect the Spitze Stei slope. To this end, we use seismology to analyze sub-surface processes at the slope and detect mass movements with seismic and infrasonic measurements. In October 2021, we installed three seismometers directly at the slope that continuously recorded seismic noise. We also installed an infrasound array ~1500 m in front of the slope terminus to strengthen mass-movement detection.

We first show the initial results of a joint analysis of seismic and infrasonic signals of slope failure activity at the Spitze Stei slope. In particular, we investigate the differences between seismic and infrasonic data features (including waveform characteristics and spectral content). This analysis is a first step towards compiling mass movement time series as the Spitze Stei slope through machine learning methods. We also use ambient seismic noise to analyze sub-surface processes through seismic interferometry and derive the resonance frequency of rock compartments. Spitze Stei gives a perfect opportunity to study the preparation phase of slope instability and investigate precursory mechanical damage in a slope impacted by climatic triggers.

Abstract ID 788 | Date: 2022-09-14 11:40 – 11:50 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Kholiavchuk, Dariia; Zelenchuk, Volodymyr; Pasichnyk, Mykola

Recent climate change in the Ukrainian Carpathians has contributed to the intensity and reoccurrence of geomorphological processes. Latter has in turn changed the vertical distribution of mountain ecosystems. Forest ecosystems together with air temperature and precipitation characteristics serve as apparent indicators of the processes. Accordingly, the study aims to identify spatiotemporal features of snow avalanche processes, landslides, rockfalls using bioindication in the focus areas of the Chornogora and Borzhava massifs in the Ukrainian Carpathians, hydrometeorological data from two weather stations Play and Pogegevskaya in the years of 1961-2015 and evidence from the local residents.

Assessment of eight focus areas in the Chornohora massif and two in the Borzhava massifs are provided based on the surveys of the Bystrets local residents and the detected deformations and damages of forest formations. The most intensive events of snow avalanches were distinguished in 1977, 1995, 1998 and 2001. These avalanche events have caused the greatest changes in the forest ecosystems of the central part of the Chornohora.

The ecosystem changes impacted by snow avalanches involve the replacement of forest ecosystems with shrubs and the loss of valuable ecosystems. Other geomorphological processes linked to the forest ecosystem destabilization are landslides and rockfalls studied in the field expeditions in 2019-2021. The latter contributes to the climate change interpretation. An increase in precipitation and temperature extremes are found the dominant hydroclimatic triggers of slope instabilities in the area. In particular, Major snow avalanches are associated with the warm spells from the Mediterranean or local radiation warming of the surface air layer.

Identified damaged coniferous formations have become the raw data for the ongoing dendrogeomorphological studies. They involve dendrogeomorphological reconstruction of slope processes together with synoptic analysis aimed at the climate change interpretation of the hazard events in the years 1961-2015.

Abstract ID 538 | Date: 2022-09-14 13:30 – 13:40 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Rom, Jakob; Haas, Florian; Heckmann, Tobias; Betz-Nutz, Sarah; Altmann, Moritz; Fleischer, Fabian; Becht, Michael

Debris flows play a big role in the morphodynamic systems of high mountain regions as well as in the interaction with other geosystems (e.g. biosphere). To get a better understanding of the future development of debris flow dynamic, it is crucial to gain a good understanding of the development of the dynamic in the past. However, it is difficult to obtain a sufficient debris flow record in high mountain areas, as for example historical text chronicles do often not provide detailed information for high altitude regions above settlement places and dendromorphological studies are only possible in forested areas.

Within this study, we establish a record of slope-type debris flows since the end of the Little Ice Age for the Grastal and the Zwieselbachtal, two side valleys of the Horlachtal, which is located within the Stubai Alps in Tyrol, Austria. Here, aerial and terrestrial photographs are limited to the period since 1947. Therefore, a detailed debris flow record for the timespan between 1947 and 2020 could be established based on a multitemporal analysis of available aerial photographs of twelve different time steps. For the timeframe between the end of the Little Ice Age and the first area-wide aerial photographs in 1947, we used lichenometric methods to date old debris flow deposits. Lichenometry is based on the even and consistent growth rate of the thalli of the lichen Rhizocarpon geographicum, which is a widespread species within the High Alps. During an extended field survey in summer 2021, we measured the diameter of lichen thalli at 105 different sites in the study areas. With the help of the collected data, we established a calibration curve using the thalli of 52 lichen sites from known historical glacier extents as well as debris flow deposits, which could be dated by the comparison of aerial photographs. Using this calibration curve, we were able to date 53 different debris flow deposits in the study area using the Five-Largest-Lichen approach and therefore extent the debris flow record to the period prior to 1947.

Abstract ID 881 | Date: 2022-09-14 13:40 – 13:50 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Acharya, Anushilan (1); Steiner, Jakob (2); Watanabe, Teiji (3)

The cryosphere of the HKH region not only sustains the livelihoods of billions of people residing downstream through its capacity to store water but also holds potential for hazards. One of these hazards, avalanches, so far remain poorly studied as the complex relationship between climate and potential triggers is poorly understood due to lack of long-term observations, inaccessibility, severe weather conditions, and financial and logistic constraints. In this study, available literature and data are reviewed to identify gaps and propose future directions of research and mitigation that should be undertaken to further strengthen the resilience of mountain communities against this hazard.

23 major avalanches with more than 10 fatalities in a single event are identified in four countries of the HKH region since 1972-2022. Afghanistan has the highest recorded avalanche fatalities (708) followed by Nepal (565), India (560) and Pakistan (278). Additionally, 349 people lost their lives while climbing on one of the 8000 m peaks located in the HKH range. Although fatalities are significant, and local long-term impacts of avalanches may be considerable, so far, no adaptation or mitigation measures exist in the region. Development of hazard zonation maps based on archive and modeling, development of regional data server and information hub, and the involvement of local authorities and communities in disaster risk reduction and management in collaboration with researchers, academics, local governments, scientists, and other relevant stakeholders should be considered to minimize the gaps. There is a necessity for extensive and long-term avalanche studies to better understand avalanche mechanisms, processes, and societal impacts in order to plan for sustainable development and resiliency in the mountainous areas of the HKH region.

To show the potential of relatively straightforward mapping approaches, potentially accessible to local stakeholders, and to understand the existing hazard situation in an exemplary area, an avalanche susceptibility map of the Langtang Valley using the Multi-Criteria Decision Analysis (MCDA)-Analytical Hierarchy Process (AHP) model in a GIS environment has been applied. Seven different terrain parameters including slope, curvature, terrain roughness, elevation, aspect, ground cover, and snow depth have been considered to derive avalanche a susceptibility map. This in turn is compared with tracked avalanches from remote sensing approaches and accounts of the local population.

Abstract ID 858 | Date: 2022-09-14 13:50 – 14:00 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Ortner, Gregor (1,2,3); Bründl, Michael (1,2); Kropf, Chahan M. (3,4); Yves, Bühler (1,2); Bresch, David N. (3,4)

Several recent studies show that changes in the climate system, such as temperature increase and extreme precipitation events, strongly influence gravity-induced hazards. As part of the Climate Change Impacts on Alpine Mass Movements research program, we are developing a framework for modeling mass movement risk as it is altered by climate and socioeconomic factors. In a first approach, we modeled avalanche risk in central Switzerland for the current climate situation and three avalanche hazard scenarios. For each of these scenarios, we considered different 3-day increases in snow depth for avalanche formation derived from meteorological stations. For the modeling, we applied the RAMMS::LSHIM Large Scale Hazard Indication Mapping algorithm, which combines an automatic delineation of potential release areas from a high-resolution terrain model with a forest layer to represent the spatial distribution of avalanche impacts for each of the selected scenarios.
To model potential climate change impacts on avalanche hazard, we use downscaled data from CH2018 climate change scenarios. We compare the stations used for the first approach with the stations used for downscaling. A medium and a extreme snowfall scenario are derived from the CH2018 data, for the first- and the second half of the century, respectively. In order to determine the snow depth distribution for the future scenarios we use the multi-purpose snow and land-surface model “SNOWPACK”. The resulting altered avalanche hazard situation is simulated using the RAMMS::LSHIM method, and risks are analyzed with the CLIMADA probabilistic Python-based risk assessment platform. High-resolution building layers are used to identify monetary values and assign vulnerabilities. The results are risk maps depicting changes in avalanche risk based on the combination of hazard, exposure and vulnerability information. These maps enable the evaluation of appropriate risk management options, thus contributing to decision support and highlighting areas where climate change adaptation measures may be required.

Abstract ID 427 | Date: 2022-09-14 14:00 – 14:10 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Dai, Xiaoru (1); Schneider-Muntau, Barbara (1); Krenn, Julia (2); Zangerl, Christian (3); Fellin, Wolfgang (1)

The Ludoialm is located in the municipality of Münster in Tyrol, Austria. The landslide is situated in the Northern Calcareous Alps which forms the regional geological framing. Quaternary fluvial-glacial sediments form the uppermost layer of the landslide which were deposited on marly sediments. The remarkable acceleration/reactivations of the Ludoialm landslide occurred probably due to the intensive snow melting in early April 1967 and also in February 1999. The material loss during this reactivation phase is approximately 486,000 m³ in the landslide area, obtained by GIS analysis. Until today, it is assumed that the landslide still moves at a low rate of activity. Although a temporal relationship between meteorological events and slope displacement has been obtained, the hydro-mechanical coupled processes responsible for the initial landslide formation and the ongoing movement characteristics have yet to be identified.

This research work provides a comprehensive analysis of the triggering factors of this landslide from the geological and geotechnical perspectives. The geotechnical strength parameters are determined on the basis of laboratory analyses. A representative cross-section of the landslide area is selected for the 2D numerical investigation. We use the simplified pre-failure geometry of the cross-section as the calculation model, aiming to simulate the initial slope failure process by finite element limit analysis (FELA) and strength reduction finite element method (SRFEM). The location of the sliding zone obtained by the calculation is compared with the geological model based on field observations.

In this case study, the high groundwater table due to the extreme snow melting event provided a very unfavorable factor for the slope stability. This triggering factor is well confirmed by the computation.

Abstract ID 151 | Date: 2022-09-14 14:10 – 14:20 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Schlögl, Matthias (1,2); Heiser, Micha (1); Scheidl, Christian (1); Fuchs, Sven (1)

In recent years, losses due to torrential flooding (fluvial sediment transport, debris floods, and debris flow) have been increasing in the Eastern European Alps. While a considerable number of studies on flooding along large foreland rivers are available, comprehensive investigations linking exposure and mitigation to the frequency of torrential flooding in mountain headwater catchments are still outstanding. Repeatedly, however, effects of climate change and settlement growth have been postulated as root causes for the development of loss. In the context of hazard mitigation, however, it remains unclear to which degree loss dynamics can be attributed to these causes. We addressed this question based on a record of approximately 12,000 events covering the period between 1962 and 2017, a database containing roughly 120,000 mitigation structures, an inventory of the building stock and 15 climate indices related to hazard triggering conditions. While the indices of triggering precipitation and the number of exposed buildings increased steadily, frequency, magnitude and seasonality of damage-inducing torrential flooding did not show clear trends. This contradiction was attributed to a compensatory effect of the increasing number of technical mitigation structures. Maintaining these structures is of paramount importance to counteract future effects of climate change on the magnitude and frequency of events and the increasing demand for land development in hazard-prone areas.

Abstract ID 302 | Date: 2022-09-14 14:20 – 14:30 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Kenner, Robert (1); Gischig, Valentin (2); Gojcic, Zan (3); Quéau, Yvain (4); Kienholz, Christian (5); Figi, Daniel (6); Thöny, Reto (6); Bonanomi, Yves (7)

Lidar measurements and UAV photogrammetry provide high-resolution point-clouds well suited for the investigation of slope deformations. Commonly, multitemporal point clouds are registered within a stable reference frame and compared using either cloud-based methods such as C2C, C2M or M3C2 or raster-based methods such as difference DEMs or feature tracking algorithms. All these state-of-the-art methods underestimate the absolute displacement rates of objects and struggle to exploit the full information content hidden in point clouds.

Using three examples of large-scale slope instabilities located in Switzerland, which are actively monitored for reasons of hazard prevention, we go beyond absolute displacement rates when analysing point clouds acquired by terrestrial laser scanning. We used the new method F2S3, enabling direct 3D point cloud comparison in combination with a varying relative referencing of the points clouds. This allows us to analyse every spatial component of the movement, such as spatially highly resolved displacement angles.

Furthermore, we compensate the main displacement of a moving rock compartment using adapted referencing strategies. In simple words, we push the displaced rock compartment back into its original position. This enables us to compare the shape of the rock compartment before and after the displacement. These secondary deformation signals are valuable indicators of the type of displacement process. We can thus identify differences in kinematic behaviour of individual rock compartments, highlight active shear planes within moving rock masses and define the kinematic process driving the slope displacements.

In the case of rock slides, the directions of the 3D displacement vector field, calculated with F2S3, reflect the shape of the underlying sliding planes. Using methods for surface integration, known from applications like shape from shading, we can model the shape of sliding planes. The shape and location of the sliding planes provide crucial information, such as estimations of destabilised rock volumes. All this information significantly contributes to process understanding at the sites investigated and thus supports decision making in hazard management.

Abstract ID 482 | Date: 2022-09-14 14:30 – 14:40 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Viaggio, Stefania; Iacopino, Alessandro

Risk assessment of rain-induced landslide over large areas is quite challenging due to the complexity of the phenomenon. Standard methods of 3D slope stability analysis (i.e., LEM, FEM or FDM) cannot be efficiently applied over extended areas with high resolution, in addition the monitoring changes in soil saturation is a key aspect in the landslide analysis. However, there is no simple relationship between the water content into the soil and the hydraulic conditions of the slopes at the depths at which the landslides develop, so the knowledge of the actual soil moisture should be monitored. Hence, to create a warning system for monitoring and forecasting landslide susceptibility to measured/forecasted rainfalls, LAMP (LAndslide Monitoring and Predicting) system has been developed (Bovolenta et al., 2016). In the frame of the INTERREG-ALCOTRA project called AD-VITAM, LAMP has been applied to several sites located in the Alpine territory on the Italian-French border.

LAMP is based on a physical based Integrated Hydrological Geotechnical (IHG) model (Passalacqua et al., 2016) fed by a low-cost and self-sufficient Wireless Sensor Network (WSN), allowing quasi-real time analyses.

The IHG model is designed to describe the response to susceptibility to landslides (specifically debris and earth slides) of a few square kilometres, typically at scale 1:5.000. Implemented in GIS environment, the modelling is completely 3D, the spatialization being made possible through appropriate data interpolation and extrapolation methods from in situ investigations and geotechnical surveys (Passalacqua et al., 2013). The site characterization requires the knowledge of piezometric measurements, stratigraphy, physical parameters, soil strength and permeability.

The model is fed by monitoring data (rainfall, temperature, soil water content) which vary both in space and time, considering the wetting condition of the soil and the water table oscillation in the previous days.

Soil moisture data are measured using capacitance sensors that allow an extremely fast response and a little request of maintenance. The use of low-cost sensors in landslide areas can allow the monitoring of large territories, but appropriate calibration is required. Multiple installations (along vertical alignments) of capacitance sensors are placed in the nodes of the monitoring network. They provide real-time water content profiles in the shallow layers (typically in the upper meter) of a slope. The integration of such monitoring system in the IHG model may be useful for the analysis and prediction of landslides triggered by rainfalls and could be of real support in risk management.

Abstract ID 688 | Date: 2022-09-14 14:40 – 14:50 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Kronenberg, Marlene (1); Neukom, Raphael (1,2); Huggel, Christian (2); Muccione, Veruska (2); Bebi, Peter (3,4); Bottero, Alessandra (3,4); Caviezel, Andrin (3,4); Ringenbach, Adrian (3,4); Salzmann, Nadine (3,4)

Cumulative climatic extreme events pose a substantial risk to society and nature, as they can propagate through various socio-economic systems via process cascades. Adaptation to future climate requires estimations of the likelihood and possible combined impacts of cumulating meteorological/climatic extreme events. Due to the very rare occurrence of low probability events, such estimations remain challenging.

Currently, only a few approaches are available to quantitatively model the manifold cascading effects that may propagate through natural and societal systems after the occurrence of combined climatic extremes (e.g. drought, windstorm). In a pilot study conducted for the Swiss Federal Office for the Environment (FOEN) we adapted methods from the field of civil protection and used expert knowledge to develop impact storylines and estimate probabilities and magnitudes of adverse effects of extreme events on society and ecosystems.

We developed an extensive expert survey to estimate the feasibility of a combined drought event and subsequent cascading hazards leading to the loss of the protective function of forests in the southern Swiss Alps. Twenty-nine experts from science, administration and practice provided quantitative estimates of drought thresholds and damage probabilities induced by two consecutive very dry and warm seasons. Results suggest that the probability of the “no harm” case (i.e. protective function retained) decreases from around 70% for single extreme events (drought, windstorm or bark beetle infestation) to less than 20% for combined effects of drought, windstorm and bark beetle infestation.

A follow up study currently focusses on evaluating the forest protective function against gravitational natural hazards under future climatic conditions and disturbance regimes for a case study in the canton of Grisons, Switzerland. We aim to directly implement and extend the cascading hazard scenarios based on the above-mentioned semi-quantitative approach into an avalanche/rockfall model allowing for an in-depth quantitative screening. Here, we present the results of the survey and report on the progress and challenges of the modelling study.

Abstract ID 762 | Date: 2022-09-14 14:50 – 15:00 | Type: Oral Presentation | Place: SOWI – Lecture hall HS3 |

Mayer, Stephanie; Van Herwijnen, Alec; Schweizer, Jürg

Numerical snow cover models allow simulating snow stratigraphy using meteorological input data from automatic weather stations, numerical weather prediction or climate models. To assess avalanche danger for short-term forecasts or with respect to long-term trends induced by a warming climate, modeled snow stratigraphy has to be interpreted in terms of mechanical instability. While instability indices describing the mechanical processes of dry-snow avalanche release have been implemented into snow cover models, there exists no readily applicable method that combines these metrics to predict snow instability.

We developed a novel, machine-learning based method to assess snow instability from snow stratigraphy simulated with SNOWPACK. Employing a data set of observed and corresponding simulated snow profiles, we trained and validated a random forest (RF) classification model based on observed snow instability. To develop the model, we manually compared 742 observed snow profiles with their simulated counterparts, which included selecting a simulated layer corresponding to the observed rutschblock failure layer. We used the observed stability test result and an estimate of the local avalanche danger to construct a binary target variable (stable vs. unstable). The final RF classifier aggregates six explanatory variables describing the weak layer and the overlying slab into the output probability of instability (P_unstable). Our model can be applied to every layer of a simulated snow profile, which offers the possibility of detecting the weakest layer and assessing its degree of instability with one single index, the maximum of P_unstable. We then simulated future snow stratigraphy using downscaled climate scenarios in the region of Davos, Switzerland and applied our RF model to comprehensively assess changes in snow instability due to climate change.