Dr. Fabian Walter

The properties of laser signals are affected by deformation of the optical fibre through which they are transmitted. While this deformation dependence is undesirable in telecommunication, it can be exploited for the construction of novel seismic sensors that fill a niche in data acquisition where traditional seismometer arrays would be difficult to deploy. This includes densely populated urban centers, the oceans, volcanoes and the Earth's polar regions. These notes complement a presentation on recent methodological developments and applications in fibre-optic seismology. The first part is focused on the use of distributed fibre-optic sensing in cryosphere research, and specifically the investigation of the internal structure and seismicity of glaciers and ice sheets. The second part is dedicated to recent advances in integrated fibre-optic sensing, with emphasis on novel measurement principles and sensitivity.

Ice streams are major regulators of sea level change. However, standard viscous flow simulations of their evolution have limited predictive power owing to incomplete understanding of involved processes. On the Greenland ice sheet, borehole fiber-optic observations revealed a brittle deformation mode that is incompatible with viscous flow, over length scales similar to the resolution of modern ice sheet models: englacial ice quake cascades that are unobservable at the surface. Nucleating near volcanism-related impurities that promote grain boundary cracking, the ice quake cascades appear as a macroscopic form of crystal-scale wild plasticity. A conservative estimate indicates that seismic cascades are likely to produce strain rates that are comparable in amplitude with those measured geodetically, providing a plausible missing link between current ice sheet models and observations.

Distributed Acoustic Sensing (DAS) technology repurposes fiber optic cables (FOCs) into seismic arrays, offering unprecedented dense strain/strain-rate measurements. The metre-scale virtual sensor spacing is typically unattainable with standard seismological equipment. Consequently, DAS provides an extraordinary amount of suitable data for seismic monitoring applications. However, intrinsic characteristics of this technology, such as signal axial polarisation, coupling inhomogeneities, or sensitivity to site conditions, can affect seismic phase amplitudes and their coherence, potentially reducing the number of useful measurement points. To gain a deeper understanding on the relative importance of these phenomena, this study analyses 'real data' from various seismic events recorded by shallow-horizontal DAS deployments. Thus, we take advantage of the pool of different array dimensions and geometries to avoid biased observations. We focus on the spatial variability of P-wave amplitudes, signal-to-noise ratios and waveform correlation, ideally mimicking the usage of absolute and differential arrival times for seismological monitoring purposes. We observed significant amplitude variations, which cannot be fully explained by signal polarisation along the FOC. Additionally, waveform correlation often exhibits a complex and faster decay with interchannel distance. These findings suggest the importance of avoiding 'blind' usage of shallow-horizontal DAS arrays and emphasise the need for case-dependent data selection/weighting procedures.

Distributed acoustic sensing (DAS) technology enables the detection of waves generated by seismic events, generally as uniaxial strain/strain rate time-series observed for dense, subsequent, portions of a Fibre Optic Cable (FOC). Despite the advantages in measurement density, data quality is often affected by uniaxial signal polarization, site effects and cable coupling, beyond the physical energy decay with distance. To better understand the relative importance of these factors for data inversion, we attempt a first modelling of noise patterns affecting DAS arrival times for a set of seismic events. The focus is on assessing the impact of noise statistics, together with the geometry of the problem, on epicentral location uncertainties. For this goal, we consider 15 'real-world' cases of DAS arrays with different geometry, each associated with a seismic event of known location. We compute synthetic P-wave arrival times and contaminate them with four statistical distributions of the noise. We also estimate P-wave arrival times on real waveforms using a standard seismological picker. Eventually, these five data sets are inverted using a Markov chain Monte Carlo method, which offers the evaluation of the relative event location differences in terms of posterior probability density (PPD). Results highlight how cable geometry influences the shape, extent and directionality of the PPDs. However, synthetic tests demonstrate how noise assumptions on arrival times often have important effects on location uncertainties. Moreover, for half of the analysed case studies, the observed and synthetic locations are more similar when considering noise sources that are independent of the geometrical characteristics of the arrays. Thus, the results indicate that axial polarization, site conditions and cable coupling, beyond other intrinsic features (e.g. optical noise), are likely responsible for the complex distribution of DAS arrival times. Overall, the noise sensitivity of DAS suggests caution when applying geometry-only-based approaches for the a priori evaluation of novel monitoring systems.

Effective use of the wealth of information provided by Distributed Acoustic Sensing (DAS) for mass movement monitoring remains a challenge. We propose a semi-supervised neural network tailored to screen DAS data related to a series of rock collapses leading to a major failure of approximately 1.2 million m³ on 15 June 2023 in Brienz, Eastern Switzerland. Besides DAS, the dataset from 16 May to 30 June 2023 includes Doppler radar data for partially ground-truth labeling. The proposed algorithm is capable of distinguishing between rock-slope failures and background noise, including road and train traffic, with a detection precision of over . It identifies hundreds of precursory failures and shows sustained detection hours before and during the major collapse. Event size and signal-to-noise ratio (SNR) are the key performance dependencies. As a critical part of our algorithm operates unsupervised, we suggest that it is suitable for general monitoring of natural hazards.

Observations of glacier melt and runoff are of fundamental interest in the study of glaciers and their interactions with their environment. Considerable recent interest has developed around distributed acoustic sensing (DAS), a sensing technique which utilizes Rayleigh backscatter in fiber optic cables to measure the seismo-acoustic wavefield in high spatial and temporal resolution. Here, we present data from a month-long, 9 km DAS deployment extending through the ablation and accumulation zones on Rhonegletscher, Switzerland, during the 2020 melt season. While testing several types of machine learning (ML) models, we establish a regression problem, using the DAS data as the dependent variable, to infer the glacier discharge observed at a proglacial stream gauge. We also compare two predictive models that only depend on meteorological station data. We find that the seismo-acoustic wavefield recorded by DAS can be utilized to infer proglacial discharge. Models using DAS data outperform the two models trained on meteorological data with mean absolute errors of 0.64, 2.25 and 2.72 m³ s^-1, respectively. This study demonstrates the ability of in situ glacier DAS to be used for quantifying proglacial discharge and points the way to a new approach to measuring glacier runoff.

Am 14. April 2024 um 06:56 Uhr ereignete sich in der Westflanke des Piz Scerscen (Berninamassiv, Engadin, Graubünden) auf einer Höhe von 3640 m ü. M. ein grosser Bergsturz. Rund 5.5 Millionen m³ Fels und Gletschereis lösten sich, stürzten auf den darunter liegenden Vadret da Tschierva und rissen in der Sturzbahn eine dicke Schneedecke und grosse Mengen Gletschereis mit sich. Dieses Gemisch aus Gestein, Eis und Schnee floss über das Gletscherplateau, durch die darunterliegenden Eisfälle und zwischen den Seitenmoränen des Gletschers hinunter und erzeugte eine grosse Staubwolke, bevor es sich auf rund 2100 m ü. M. im Val Roseg, fast 6 km vom Anriss ablagerte. Das Ereignis forderte glücklicherweise keine Opfer und verursachte keine Infrastrukturschäden. In diesem Bericht beschreiben wir die Bedingungen, die zum Zeitpunkt des Ereignisses herrschten und nutzen die verfügbaren Informationen, um die Prozesse zu analysieren, die vor und während des Ereignisses stattfanden.

While earthquakes are well-known to trigger surface hazards and initiate hazard cascades, whether surface hazards can instead trigger earthquakes remains underexplored. In 2018, two landslides on the Tibetan plateau created landslide-dammed lakes which subsequently breached and caused catastrophic outburst floods. Here we build an earthquake catalog using machine-learning and cross-correlation-based methods which shows there was a statistically significant increase in earthquake activity (local magnitude ≤ 2.6) as the landslide-dammed lake approached peak water level which returned to the background level after dam breach. We further find that ~90% of the seismicity occurred where Coulomb stress increased due to the combined effect of direct loading and pore pressure diffusion. The close spatial and temporal correlation between the calculated Coulomb stress increase and earthquake activity suggests that the earthquakes were triggered by these landslide hazard cascades. Finally, our Coulomb stress modeling considering the properties of landslide-dammed lakes and reservoir-induced earthquakes globally suggests that earthquake triggering by landslide-dammed lakes and similar structures may be a ubiquitous phenomenon. Therefore, we propose that earthquake-surface hazard interaction can include bidirectional triggering which should be properly accounted for during geological hazard assessment and management in mountainous regions.

Surface hazards can form hazard cascades which expand their reach. However, our understanding of their complex dynamics and ability to mitigate their impacts remain limited. In 2018, two landslides dammed the Jinsha River in the Tibetan plateau and formed landslide-dammed lakes. Subsequent dam breaches prompted the evacuation of >120,000 people. An early warning system for floods using a regional seismic network has been proposed on the basis of catastrophic floods having been detected ∼100 km away, with seismic energy proportional to discharge. Surprisingly, we find that this catastrophic outburst flood was undetectable beyond a few kilometers, with peak seismic energy preceding peak discharge. We propose that river channel stability also controls seismic energy generation and should be considered for accurate monitoring of catastrophic floods. In contrast, we find that the various processes during dam breach can be well-characterized seismically further away and provide warning ∼60 min before discharge exceeds monsoon flood levels. We also show that numerical modeling of dam breaches which typically lacks in situ measurements can benefit from incorporating seismic data as constraints. Finally, we show that this event drastically increased sediment fluxes ∼670 km downstream for years and may significantly reduce the capacity of hydropower plants. Our results reveal ways to improve early warning of catastrophic outburst floods and the need to consider surface hazards' long-term impact when managing infrastructure in mountainous regions.

Seismic instruments placed outside of spatially extensive hazard zones can be used to rapidly sense a range of mass movements. However, it remains challenging to automatically detect specific events of interest. Benford's law, which states that the first non-zero digit of given data sets follows a specific probability distribution, can provide a computationally cheap approach to identifying anomalies in large data sets and potentially be used for event detection. Here, we select vertical component seismograms to derive the first digit distribution. The seismic signals generated by debris flows follow Benford's law, while those generated by ambient noise do not. We propose the physical and mathematical explanations for the occurrence of Benford's law in debris flows. Our finding of limited seismic data from landslides, lahars, bedload transports, and glacial lake outburst floods indicates that these events may follow Benford's Law, whereas rockfalls do not. Focusing on debris flows in the Illgraben, Switzerland, our Benford's law-based detector is comparable to an existing random forest model that was trained on 70 features and six seismic stations. Achieving a similar result based on Benford's law requires only 12 features and single station data. We suggest that Benford's law is a computationally cheap, novel technique that offers an alternative for event recognition and potentially for real-time warnings.

Forecasting hanging glacier instabilities remain challenging as sensing technology focusing on the ice surface fails to detect englacial damage leading to large-scale failure. Here, we combine icequake cluster analysis with coda wave interferometry constraining damage growth on Switzerland's Eiger hanging glacier before a 15,000 m³ break-off event. The method focuses on icequake migration within clusters rather than previously proposed "event counting." Results show that one cluster originated from the glacier front and migrated by 13.9(±1.2) m within 5 weeks before the break-off event. The corresponding crevasse extension separates unstable and stable ice masses. We use the measured source displacement for damage parametrization and find a 90% agreement between an analytical model based on damage mechanics and frontal flow velocities measured with an interferometric radar. Our analysis provides observational constraints for damage growth, which to date is primarily a theoretical concept for modeling englacial fractures.

Rapid process cascades can lead to destructive debris flows. Identifying and characterizing the processes conditioning debris-flow occurrence will strongly contribute to the mitigation of debris-flow hazards. In recent years, the rock slope near “Spitze Stei”, in the Kandersteg region, Switzerland, has exhibited elevated displacement rates exceeding 10 cm per day, suggesting a growing instability up to 20 million m³. The accumulated sediments at the bottom of the Spitze Stei slope are mobilized as debris flows by melting snow and heavy summer precipitations. Here, we use seismology combined with an intelligent algorithm to automatically detect rockfall and landslides at the Spitze Stei rock slope. These mass movements act as primary sediment sources delivering sediments to a debris-prone channel. Our initial results quantify mass movement activity before two debris flow events that occurred in 2022 and identify their triggers. Such analysis can contribute towards mitigating debris flow hazards and extending warning time, especially for debris flows triggered by factors other than precipitation.

Glacier seismology is a valuable tool for investigating ice flow dynamics, but sufficient data acquisition in remote and exposed glaciated terrain remains challenging. For data acquisition on a highly crevassed and remote outlet glacier in Greenland we developed self-sufficient and easily deployable seismic stations, "SG-boxes". The SG-boxes contain their own power supply via solar panel, a three-component omni-directional geophone and a GNSS receiver. The SG-boxes can be deployed and retrieved from a hovering helicopter, allowing for deployment in difficult terrain. To assess their performance we conducted a field test comparing the SG-boxes to established on-ice geophone installations at Gornergletscher in Switzerland. Moreover, data from a first SG-box deployment in Greenland were analyzed. The SG-boxes exhibit consistently higher noise levels relative to colocated conventional geophones and a correlation between noise levels, wind and air temperature is found. Despite their noise susceptibility, the SG-boxes detected a total of 13,114 Gornergletscher icequakes over 10 days, which is 30% of the total number of icequakes detected by conventional geophone stations. Hence, even in sub-optimal weather conditions and without additional noise reduction measures, the SG-boxes can provide unique and valuable data from challenging glaciated terrain where no conventional seismic installations are possible.

Avalanches and other hazardous mass movements pose a danger to the population and critical infrastructure in alpine areas. Hence, understanding and continuously monitoring mass movements are crucial to mitigate their risk. We propose to use Distributed Acoustic Sensing (DAS) to measure strain rate along a fiber-optic cable to characterize ground deformation induced by avalanches. We recorded 12 snow avalanches of various dimensions at the Vallée de la Sionne test site in Switzerland, utilizing existing fiber-optic infrastructure and a DAS interrogation unit during the winter 2020/2021. By training a Bayesian Gaussian Mixture Model, we automatically characterize and classify avalanche-induced ground deformations using physical properties extracted from the frequency-wavenumber and frequency-velocity domain of the DAS recordings. The resulting model can estimate the probability of avalanches in the DAS data and is able to differentiate between the avalanche-generated seismic near-field, the seismo-acoustic far-field, and the mass movement propagating on top of the fiber. By analyzing the mass-movement propagation signals, we are able to identify group velocity packages within an avalanche that propagate faster than the phase velocity of the avalanche front, indicating complex internal structures. Importantly, we show that the seismo-acoustic far-field can be detected before the avalanche reaches the fiber-optic array, highlighting DAS as a potential research and early warning tool for hazardous mass movements.

In view of a national Landslide Early Warning System (LEWS) for Switzerland, a pilot study is currently running in the Napf-region (central Switzerland), a hilly semi-forested area (186 km²) with a substantial number of shallow landslide observations in the past decades. The study includes a comprehensive soil wetness monitoring, a test of a numerical model to simulate distributed landslide triggering, as well as the application of Distributed Acoustic Sensing (DAS) at the scale of a steep slope. We show that the applied tools provide a promising basis for a systematic observation, assessment and warning of critical landslide situations. At the same time, this work also suggests that the soil physical conditions can and need to be addressed in an adequate way to correctly rate the spatial and temporal criticality for landslide triggering.

Accurate debris-flow modelling depends on the ability to simulate surging and pulsing behaviour. However, our understanding of these phenomena is starved of observational constraints. Here we propose the use of seismic measurements, which resolve the arrival of coarse-grained roll wave fronts in debris flows at Illgraben, Switzerland. Roll waves likely play a key role in flow pulses but are typically only observed with point measurements like depth gauges. We compare in-torrent force plate measurements with near-torrent seismic records and discuss how these data can test existing roll wave theories.

Steinschläge, Lawinen, Murgänge und ähnliche Massenbewegungen erzeugen elastische Wellen im Boden und in der Luft. Diese Signale können mit Seismometern und Infraschallmikrophonen aufgezeichnet und mit geeigneten Algorithmen detektiert werden. Dies bietet Perspektiven für die Überwachung von und Warnung vor gravitativen Naturgefahren, da die Aktivität von Sturzprozessen vor einem katastrophalen Abbruch häufig erhöht ist. Da die Installation von Seismometern und Infraschallanlagen relativ einfach und günstig ist, könnte diese Methode in der Zukunft eine wichtige Rolle bei der Hangüberwachung spielen. Hier stellen wir eine Pilotstudie an der Hanginstabilität «Spitze Stei» im Kanton Bern vor. Dieser Hang bewegt sich mit bis zu 10 cm pro Tag und die Region hat im Holozän Bergstürze von hunderten Millionen Kubikmetern produziert. Obwohl es derzeit keine Anzeichen für unmittelbar bevorstehende grosse Abbrüche gibt, bietet die Steinschlag- und Rutschaktivität gute Voraussetzungen für eine Testüberwachung mit Seismometern und Infraschallmikrophonen.

Les chutes de pierres, les avalanches, les laves torrentielles et autres mouvements de masse similaires génèrent des ondes élastiques dans le sol et dans l'air. Ces signaux peuvent être enregistrés par des sismomètres et des microphones à infrasons et détectés par des algorithmes appropriés. Cela offre des perspectives pour la surveillance et l'alerte des risques naturels gravitaires, car l'activité des processus de chute est souvent accrue avant un effondrement catastrophique. Comme l'installation de sismomètres et d'infrasons est relativement simple et peu coûteuse, cette méthode pourrait jouer un rôle important dans la surveillance des pentes à l'avenir. Nous présentons ici une étude pilote sur l'instabilité de pente "Spitze Stei" dans le canton de Berne. Ce versant se déplace jusqu'à 10 cm par jour et la région a produit des éboulements de centaines de millions de mètres cubes au cours de l'Holocène. Bien qu'il n'y ait actuellement aucun signe de grands effondrements imminents, l'activité de chutes de pierres et de glissements de terrain offre de bonnes conditions pour une surveillance test à l'aide de sismomètres et de microphones à infrasons.

The first step in building an early warning system using seismic signals is to automatically identify events of interest. Here, the first digit distribution of seismic signals generated by debris flows and other surface processes was calculated to validate compliance with Benford's law (BL). A detector model for debris flow events was introduced based on amplitude range and goodness of fit of BL. We show that seismic signals generated by debris flows, landslides, and bedload transport follow the BL. These events release more energy and last longer than rockfalls, which do not follow BL. In the test dataset with 1224 samples, the accuracy of the detector model in identifying debris flow events was 0.75.

Debris flows are among the most dangerous phenomena in mountain environment. Recently the use of seismic and infrasonic recordings has proven to be a powerful tool for studying and monitoring debris flows. However, open questions remain about how signal characteristics refllect flow parameters and fluid dynamic processes. We present a seismo-acoustic analysis of the debris flow activity between 2017 and 2019 at the Illgraben catchment (Switzerland). Seismic and infrasonic amplitudes (maximum root mean square amplitude [RMSA]) and peak frequencies are compared with flow measurements (front velocity, maximum flow depth and density). Front velocity, maximum depth, peak discharge and peak mass flux show a positive correlation with both infrasonic and seismic maximum RMSA, suggesting that seismo-acoustic amplitudes are influenced by these flow parameters. Comparison between seismo-acoustic peak frequencies and flow parameters reveals that, unlike seismic signals, characterized by a constant peak frequency regardless of the magnitude of the flow, infrasound peak frequency decreases with increasing flow velocity, depth and discharge. Our results suggest that infrasound and seismic waves are generated by different source processes, acting at the flow free surface and at the channel bed respectively. We propose that infrasound is likely generated by waves and oscillations that develop at the surface of the flow, which, according to fluid dynamics, are mostly generated wherever the flow encounters significant channel irregularities. Furthermore, the observed positive correlations between seismo-acoustic signal features and flow parameters highlight the potential to use infrasound and seismic measurements for debris-flow monitoring and risk management.

Accumulating evidence suggests that glacier basal motion can produce seismogenic stick-slip events caused by frictional resistance at the ice-bed interface. Frequent and subsequent failure of multiple stick-slip patches may constitute sliding tremor in seismic recordings. We show that during stick-slip tremor at the base of an Alpine glacier, the released seismic moment increases by an order of magnitude. The tremor emerges during melt-intensive days and is associated with water pressure peaks in a nearby subglacial channel. We propose an extended slider-block model that explains the regular occurrence of stick-slip tremor by coupling seismically sliding patches to surrounding aseismically sliding bed regions. The model predicts that the stick-slip patches control the movement of their surrounding smoothly sliding bed regions via elastic coupling and stress transfer. Thus, tremor producing stick-slip patches could affect large-scale ice flow to a degree, which is underestimated in models, with implications on sea level rise projections.