One of the unsolved challenges for deep geothermal projects is how to mitigate unacceptably large, induced earthquakes. Unfortunately, the processes and conditions underpinning induced seismicity are still not sufficiently well understood to make useful predictions of the likely seismic response to geothermal reservoir development and exploitation. This predictability is further reduced by the sparse knowledge of the hydro-mechanical and geological conditions in the deep underground, and our lacking ability to map them well enough with current technologies. The seismic response to deep geothermal operations can, however, be monitored in real-time by seismological methods. These methods are, therefore, the core element of any traffic light system (TLS) for induced seismicity that was proposed in recent years.

In this project SIAMIS-GT (Improvement of SeIsmological Analysis Methods to better support cantonal authorities in questions related to Induced Seismicity in deep GeoThermal projects), we want to improve seismological analysis methods for monitoring induced seismicity by taking advantage of the waveform similarity observed in these earthquake sequences. The goal is to enable the SED to inform authorities, project developers and the population faster and more accurately about induced earthquakes that may occur during geotechnical projects (e.g., geothermal, mining, tunneling, etc.). We plan to apply similarity and repeating earthquake analysis to induced seismicity to improve the understanding of the source mechanics (i.e., distinction of natural versus induced seismicity, improved detection, localization and characterization). With these methods we also hope to be able to resolve aseismic changes in the subsurface that cannot be resolved using classical methods in real-time.

Figure on the left: Illustration of the basic principles of similarity location. Upper left: waveforms of 7 template earthquakes used for template matching of the Diemtigen swarm. Upper right: Epicenter map of 306 relocated earthquakes color-coded by highest similarity to the 7 templates. Red diamonds indicate the location of the templates. Notice that earthquakes cluster closely around the template they are associated to by waveform similarity. Bottom: Relationship between inter-event distance and waveform similarity for each template with respect to the 306 relocated earthquakes. Color indicates the reference template. It can be clearly seen that similarity decreases strongly with inter-event distance and that individual relationships can be derived for every reference template.

In the last few years, major and damaging earthquakes were felt in regions supposedly affected by a low rate of natural seismicity. Such events have become an extremely important topic of discussion in both Europe and North America, since several major events were associated to industrial activities. The main focus of this proposed research is the study of induced seismicity during exploitation of natural underground resources. More specifically, this proposal focuses on one side on a detailed understanding of the deep fault and/or fractures reactivation associated with the fluid injection. On the other side, the proposed research plans to investigate the mitigation of large magnitude induced seismicity, which may pose at risk the affected population and structures, as well as acting as obstacle to the development of new techniques for the exploitation of deep underground resources. First the project aims to understand the physics associated with the induced seismicity caused by anthropogenic activities. Secondly, the project aims to investigate in which conditions fluid-induced seismicity can be used as a tool. For example, the shearing process of fault and fracture, or the creation of new fractures, are needed to increase the permeability, and hence to enhance the fluid circulation at depth, eventually resulting in a more efficient energy production. However, such processes may, at the same time, produce seismicity that can be felt by local population.

• How is the induced seismicity affecting the growth of a geothermal reservoir? When does it act as a potential seal breaching during storage operation?
• Is it possible to use induced seismicity to develop a high permeability system, while constraining the earthquakes maximum magnitude? Or can a gas be stored in a safe (no seismic) way?

Understanding the fundamental processes will help in finding new methods to safely extract energy, whose request nowadays is constantly increasing.

The project will address the above open questions by theoretical studies, laboratory, and applied fieldwork. I plan to use data collected from on-going experiments (laboratory and in-situ) to improve numerical models. Numerical tools to be used include: thermo-hydromechanical modeling (THM), discrete fracture network (DFN), and statistical, as well as hybrid simulators.

The first phase of this project involves the participation in the In-situ Stimulation and Circulation (ISC) experiment at the Grimsel Test Site. Goal of the experiment series is the permeability enhancement of a decameter rock volume intersected with two fault zones. The permeability enhancement process is accompanied by slip along the faults, which is achieved by high-pressure fluid injection (i.e. hydraulic shearing). Fluid injection is performed in several borehole sections distributed over two injection boreholes drilled trough the target fault zones. During the experiment hydraulic processes, mechanical processes and induced micro-seismicity are monitored with an unprecedented multi-sensor monitoring system leading to a unique and high quality dataset.
The goal of the second phase of the project is the detailed processing of recorded micro-seismic data. The objective is to determine a wide range of earthquake source parameters. The adaptation and further development of seismological processing techniques to the decameter scale make this task particularly delicate.

The third phase of the project will bring together micro-seismic data and other measured data during the experiment and shed light into the physical understanding of induced earthquakes.

Finally, forecasting skills of models for induced seismicity are tested on the Grimsel high quality dataset.

Le SED a été sollicité ces dernières années pour la surveillance sismique de plusieurs projets de géothermie profonde en Suisse, et a pu ainsi initier un dialogue constructif avec les exploitants et les autorités cantonales. Sur la base de ces expériences, le SED a élaboré, à destination des autorités octroyant les autorisations et des autorités d’exécution, les premières recommandations concernant la sismicité induite dans les différentes phases d’un projet de géothermie profonde. Ces recommandations ne s’appuient cependant jusqu’ici que sur l’expérience de quelques projets bien documentés. L’expérience pratique pour la mise en œuvre est largement lacunaire jusqu’ici en Suisse. Il est donc difficile d’évaluer si ces recommandations sont applicables dans le contexte local (géologique, politique, culturel) et quelles adaptations doivent être effectuées le cas échéant. Il est également important de se demander dans ce contexte comment les directives et consignes d’autres domaines techniques et administratifs interagissent avec les recommandations pour éviter la sismicité induite, et si elles sont compatibles. D’éventuels conflits peuvent souvent n’être identifiés et résolus que lors de la mise en œuvre du projet et par une étroite collaboration entre les sismologues et les autorités octroyant les autorisations ainsi que les exploitants. Tout aussi importantes sont la mise à jour et la surveillance régulières de toutes les directives et consignes au fur et à mesure de l’acquisition de nouvelles connaissances et de nouvelles expériences. À chaque fois que c’est possible, ceci devrait s’effectuer par consensus avec tous les groupes d’intérêt.

Dans le cadre du projet Geobest-CH, le SED va donc continuer à collecter, dépouiller et interpréter des jeux de données fondamentales de hautes qualité et résolution sur la sismicité induite pour des projets de géothermie profonde. Cela permettra également au SED d’apporter conseils et assistance sismiques aux autorités cantonales et fédérales dans leurs missions d’autorisation et de contrôle lors de toutes les phases d’un projet de géothermie profonde. Le SED veut ainsi contribuer à établir des échelles d’évaluation équivalentes pour le contrôle de l’impact environnemental par-delà les limites cantonales, et permettre à court et moyen terme une unification de cette procédure, et donc une simplification et une accélération notoire de son exécution. Un site web central permettra d’informer le grand public de manière indépendante, actuelle et professionnelle sur le thème de la sismicité induite, afin de contribuer à une discussion objective entre tous les groupes d’intérêt.

Dans le cadre du projet d’exploration GeoBest, le Service Sismologique Suisse (SED) effectue la surveillance du programme géothermique de la ville de Saint-Gall. Pour ceci, le SED, en collaboration avec les ateliers municipaux de Saint-Gall, a installé six nouvelles stations sismologiques dans la région. L’objectif de cette surveillance consiste à détecter d’éventuels petits tremblements de terre (microséismes) aux alentours des forages profonds, et de déterminer s’ils sont provoqués par le projet géothermique ou s’ils sont d’origine naturelle. Par ailleurs, des données de base importantes pour une meilleure compréhension de la géothermie sont collectées dans le cadre du projet, un trésor d’expérience indispensable qui devrait permettre de garantir une sécurité de planification pour les autorités cantonales et les futurs exploitants de projets géothermiques.

Après l’arrêt du projet de géothermie au printemps 2014, le SED mène la surveillance sous forme réduite au minimum jusqu’à septembre 2020. Les ateliers municipaux de Saint-Gall soutiennent cette surveillance dans le cadre du projet UE « Science for Clean Energy (S4CE) ».

The exploitation of underground energy resources as well as the use and expansion of hydropower are, like any other energy technology, not risk free. To address this risk, we develop upon the holistic concept of risk governance and community resilience, advocating a broad picture of risk: In addition to risk analysis and risk management, we also investigate how risk-related decision-making unfolds when a range of actors is involved. This requires coordination and possibly reconciliation between a profusion of roles, perspectives, goals and activities. Developments include: a rigorous common methodology and a consistent modelling approach to hazard, vulnerability, risk, resilience and societal acceptance assessment of energy technologies; a stress test framework and its application to assess the vulnerability and resilience of individual critical energy infrastructures, as well as the first level of interdependencies among these infrastructures; standardized protocols, operational guidelines and/or softwares for monitoring strategies, hazard and risk assessment during all project phases (including real-time procedures), and finally for mitigation and related communication strategies.

In phase II, the main activities of the group are in continuation with phase I but with increased interactions with Swiss P&D projects, including industrial ones and research underground labs. New applications are being developed and tested for induced seismicity risk mitigation. Those are also refined based on P&D project feedback. The main innovations in phase II are the testing of more sophisticated traffic-light systems and the inclusion of seismic risk models in economic models (including energy modelling, cost-benefit analyses and decision making under uncertainty). For the latter, interactions with SCCER CREST via a joint activity will help move forward with some legislative recommendations.

The exploitation of underground energy resources as well as the use and expansion of hydropower, are, like all energy technologies, not risk free. The risks identified in the domain of deep geothermal are induced seismicity and other risks (e.g. borehole blowout, environmental risks). The hazard factors for hydropower are those classically affecting large arch or gravity dams, aggravated by the pronounced topography of the Alpine region and by the rapidly changing climatic conditions in the Alps. How these risks potentially impact our society, is, of course dependent on the vulnerabilities of our buildings, our infrastructure and our communities. Moving towards a safe and more resilient energy sector requires tools for hazard and risk assessment, particularly in the low probability-high consequence event settings. Furthermore, the tools have to be closely integrated with related communication and public engagement strategies, as perceived risks actually do impact energy source design and mitigation strategies. A comprehensive risk governance framework is necessary, but not yet existing. This project aims to develop such a framework by integrating risk assessment and related risk perception models and tools. This should lead to a communication strategy for future projects. We develop a holistic concept of risk governance from a truly multi-disciplinary perspective, advocating a broad picture of risk: not only does it include risk assessment and assessment of ability to recover, but it also looks at how risk perception and risk-related communication can be organized. This work is divided into 6 PhDs (attached to SCCER-SoE T4.1): Induced seismic risk (led by SED); Renewable energy risk management & optimization; Accident risks at dams; Vulnerability of the Swiss built environment; Multi-risks and interdependencies; Assessing and monitoring risk perception.

This is the 3rd subproject of the „ENSI – SED-Erdbebenforschung zu Schweizer Kernanlagen“ project.

Subproject 3 aims at studying the relevancy of induced seismicity for the disposal of nuclear waste at geological depth. We will update/develop physics-based modeling of induced seismicity, and we will apply such models to the case of nuclear waste disposal. Moreover, the approach will be validated by reproducing the observations from the “Fault Slip experiment” at Mont Terri underground laboratory. Finally, numerical modeling will be performed to study the seismicity induced by tunnel excavation, as well as to better understand the possible occurrence of seismicity by temperature changes around disposal site. Collaboration with Subproject 1 and 2 will be crucial to properly estimate the hazard and risk of induced seismicity on geological nuclear waste disposal.

RT-RAMSIS is a joint CTI project between the Swiss Seismological Service SED and GeoEnergie Suisse AG. It builds on research results from previous joint research projects GEOSIM and GEOBEST. We are developing and validating a near real time hazard and risk assessment framework for induced seismicity in geothermal projects. With fluid injection rates and micro-earthquakes recorded in near real-time and prepared as input, the framework uses statistical and hybrid statistical-hydromechanical ensemble models to forecast seismicity in six hour intervals. Subsequent stages then compute probabilistic seismic hazard and risk estimates based on different injection scenarios and thereby help the operator to balance risk and reservoir stimulation efficiency.

Core research activities in this project are in the development of fast, reliable and precise micro-earthquake detection algorithms, in the assessment and weighting of forecast model performance and in the enhancement and validation of statistical and mechanical seismic forecast models.

In this ETHZ-funded project we analyze ambient vibration and earthquake recordings to characterize and investigate the dynamic response of unstable rock slopes. We perform systematic measurements and interpretation of ambient vibrations at known unstable rock slopes, both with single stations and array-configurations. The eigenfrequencies, eigenmodes, directivity, and amplification of ambient vibrations are identified and compared to geotechnical investigations. The interpretation of recordings targets the estimation of the potential landslide volume and is supported by numerical modeling of seismic wave propagation in fractured media. A classification scheme based on the seismic response has been introduced. Each class indicates specific properties of a rock instability. The two main classes found are the volume-controlled sites and the depth-controlled sites. The extensive database will be extended with instabilities on high-alpine permafrost locations.

Moreover, both short-term and long-term monitoring is undertaken to understand the time evolution of the slope structure. Short-term monitoring is performed at the Alpe di Roscioro (Preonzo) site which represents a unique opportunity to monitor a slope close to collapse. A second semi-permanent station will be installed on the large rock instability above Brienz (Grisons). A main goal of the monitoring is to understand the effect of weather and climate on the dynamic behavior of the rock and its stability. Long-term monitoring is based on the analysis of available past recordings from existing seismic networks, especially for stations located at or very close to steep cliffs. In a later stage of the project, the effect of earthquakes on the rock slope stability will be evaluated using numerical modelling. It will be evaluated, if ambient noise measurements can provide a direct proxy for the seismic vulnerability of rock slope instabilities. The expected results have the potential to be applied directly in hazard analysis and risk reduction measures.

The reliable detection of snow avalanches, landslides and rockfall provides the basis for any advanced investigation on triggering mechanisms, risk assessment or the identification of possible precursors. It is well known that such phenomena produce specific seismic signals. Nonetheless, the manual detection on seismic traces is not feasible since it is extremely time consuming and results are influenced by the subjective view of the analyst. Automatic detection is therefore preferred, as unbiased results are obtained in near real-time. In this project, we will take advantage of an automatic classification procedure for continuous seismic signals to detect avalanches, landslides and rockfalls in continuous streams of seismic data. The applied method is independent of previously acquired data and classification schemes, offering the opportunity to detect very rare and highly variable events.

The goal of this project is to improve the detection across the entire Swiss Alps and to better understand possible triggering mechanisms. In addition accurate information on the number and release time of avalanches will provide an important contribution to the avalanche warning service at the SLF. This research is highly relevant since large scale avalanche and landslide monitoring will provide reliable data for further research, the development of near real-time products and improved risk assessment and warning systems.

Scientists from the SED are exploring the potential of Earthquake Early Warning (EEW) in Central America. This is a multi-phase project funded by Swiss Agency for Development and Cooperation (SDC) at the Federal Department of Foreign Affairs (FDFA). In Phase 1 (2016-2017), SED worked with colleagues at INETER (a government agency in Nicaragua responsible for Natural Hazards, who operate the national seismic network), to build and implement a prototype EEW system in Nicaragua. In Phase 2 (2018-2020), this prototype system will be extended to other countries in Central America with the collaboration of local scientists, and we will explore how a public EEW could be implemented.

The region suffers from large tsunamigenic earthquakes generated along the subduction zone, and also moderate crustal earthquakes that have in the recent past produced heavy damage, such as the M6.2 1972 earthquake that devastated Managua. The subduction zone events are often characterized by slow rupture velocity (most recently a M7.7 1992 earthquake - the first slow earthquake ever documented - and a M7.3 in 2014)

Earthquake Early Warning (EEW) is a tool that can rapidly characterize on-going earthquakes, and potentially provide seconds to 10s of seconds notification of impending strong shaking in advance of its occurrence. EEW can play an important role as part of a seismic risk reduction program, which is critical in the Central American region that has such a high seismic hazard. Additionally, making a seismic network capable of operating and maintaining an EEW system requires the network to achieve the highest standards in network performance - including station quality, speed and reliability of data communications, and robustness of the seismic network hub that runs the EEW software. Optimal network performance is also critical for other applications such as tsunami warning, volcano monitoring, and enables downstream scientific studies, eg on local Earth structure.

On 9 June 2016, just weeks after the first software was installed and before efforts to optimise had begun, a shallow M6.3 event occurred on the border with El Salvador that was detected by our system after 29s. Though far from the delay time required for operational EEW, this does demonstrate the promise of the existing infrastructure. With an ideal seismic network (all stations fully operational, recording strong motion and with minimal data delay) the existing network density can provide first EEW alerts within 8-12s for shallow on-shore earthquakes. This corresponds to a blind zone on the order of 20-30km around the epicenter where no advanced warning will be available. Outside the blind zone, for large earthquakes, this type of EEW can provide warning in advance of the strongest shaking for areas that will experience shaking of intensity VI on the Modified Mercalli Scale.  

Transfer of EEW capacity from SED to Central American Institutes is feasible because we all use the same basic software for seismic network monitoring - SeisComP3 - that the SED has used to develop EEW. SED use 2 standard algorithms to provide EEW - the Virtual Seismologist (link methods and software page) and the Finite Fault Detector (link methods and software page). We use the EEW Display (EEWD) (link software page) to deliver desktop alerts to early adopters and testers.

EEW can work in Central America because the seismic networks are already relatively dense and data sharing is well-established and effective - necessary as earthquakes can have impacts beyond a single country.

Currently, EEW alerts in the region are delayed or initially incorrect primarily because the seismic networks are relatively unreliable and not yet optimised for EEW. By adopting a long term strategy that focuses on enabling EEW, including adding additional strong motion stations, this will improve.

The Phase 2 project objectives are to 1) review the performance of, and propose improvements to the seismic network in the region, in an effort to optimise their capacity for EEW; 2) to develop EEW algorithms tailored to the seismicity of Central America, with implementation in standard open source software, and transfer know-how and operation to local monitoring institutions; 3) head towards public EEW by engaging with potential key end-users, civil defence and governments, exploring technical means by which EEW can be provided to public and private users, and communicating with social scientists on what is the most effective EEW message provide to the public, and how to communicate this to the public.

Chile has been struck by a number of very large earthquakes (magnitude 7.5 and greater) and tsunamis in the past. There is a clear need for building a system to provide rapid situational awareness, as well as earthquake and tsunami early warning. Since scientific instruments are expensive, we are exploring the use of dense networks of seismic and geodetic low-cost sensors.

The Earthquake Science Center of the U.S. Geological Survey (USGS) has recently started to deploy a prototype network of dedicated smartphone units along the Chilean coast (Brooks et al., 2016). Each sensor box contains a smartphone with integrated MEMS accelerometer, and an external consumer-quality GPS chip and antenna to determine real-time positioning data. The total cost of each box is on the order of a few hundred dollars, nearly two orders of magnitude lower than scientific-grade installations.

We invert seismic and geodetic real-time data from the smartphone units to obtain finite-fault models of large earthquakes by joint application of the seismic FinDer algorithm developed at the Swiss Seismological Service (SED) (Böse et al., 2012; Böse et al., 2015; Böse et al., 2018) and the geodetic BEFORES algorithm developed by Minson et al. (2014).

The SED collaborates in this project with partners at the USGS, the University of Chile, Chilean National Seismological Center, University of Houston, and GISMatters Inc.

Furthermore, we collaborate with the Centro Seismologico National (CSN) at the University of Chile, Santiago, to rapidly identify large subduction-zone and crustal earthquakes using real-time streams of CSN stations (Carrasco et al., 2017).

California, Oregon, and Washington are currently implementing a public earthquake early warning (EEW) system, called ShakeAlert. The SED collaborates with the ShakeAlert team (US Geological Survey, Caltech, UC Berkeley, University of Washington, and University of Oregon) in the development, implementation, and testing of the Finite-Fault Rupture Detector (FinDer) algorithm (Böse et al., 2012; Böse et al., 2015; Böse et al., 2018).

To detect earthquakes and to generate alerts, ShakeAlert feeds data from the seismic and geodetic networks operated along the US west coast into two algorithms: the point-source EPIC (formerly known as ElarmS and Onsite), and the finite-source FinDer algorithm. Alert messages from EPIC and FinDer are aggregated into a single alert feed.

ShakeAlert started in 2012 as a demonstration system for EEW and provided warnings of imminent strong ground shaking to a selected group of test users. At that stage, also the Virtual Seismologist (VS; Cua and Heaton, 2009), another algorithm implemented and tested by the SED, was part of the system. In 2013, California passed legislation to implement a public warning system. After the US Congress approved the funding in 2014, the transition to a full public system started. A limited public roll-out of ShakeAlert is planned for fall 2018.

SERA (Seismology and Earthquake Engineering Research Infrastructure Alliance in Europe) has the goal to integrate data, products, infrastructures and know-how in seismology and earthquake engineering. The project is funded through the European Union.

In WP 28 (Real-time Earthquake Shaking (JRA 6)) of the SERA project, we work with our colleagues at the University of Naples, GFZ, INCDFP, NOA, EMSC, CNRS, and INGV, to produce rapid shaking forecasts following major earthquakes in Europe or elsewhere:

“In the past decade, real-time seismology has moved from providing post-event information within minutes from earthquake occurrence, to issuing event information during the rupture, as soon as data begin to arrive at the network. Reliability and accuracy of the source parameters are limited by the use of automatic procedures applied to data processing and the availability of the earliest snippets of the wavefield reaching only a fraction of the seismic network. As consequence, early estimates of shaking and possible damage are also accompanied by large uncertainties, impacting the ability to organize a rapid and appropriate response.

This work package applies forward modelling techniques to provide time-evolving prediction maps of the expected ground shaking from regionalized GMPE and tsunamigenic potential of a seismic rupture. The predicted shaking estimates are further constrained by integration of late arriving seismic, accelerometric or GPS data and felt reports as they become available. These goals require the adoption of flexible approaches capable to complement near source data with regional and teleseismic data. Finally, the possible improvement of automatic impact assessment of global earthquakes due to improved shaking estimates will be evaluated.” (from SERA proposal)

We use the 2016-2017 Central Italy earthquake sequence, including the destructive 2016 M6.5 Norcia, M6.0 Amatrice, and M5.9 Visso normal fault earthquakes, for demonstration and testing of our algorithms (e.g. FinDer) and software (e.g. EEWD – Earthquake Early Warning Display).

AlpArray is a European initiative to advance our understanding of orogenesis and its relationship to mantle dynamics, plate reorganizations, surface processes and seismic hazard in the Alps-Apennines-Carpathians-Dinarides orogenic system. The initiative integrates present-day Earth observables with high-resolution geophysical imaging of 3D structure and physical properties of the lithosphere and of the upper mantle, with focus on a high-end seismological array.

AlpArray is a major scientific collaboration with over 40 participant institutions. One of the main actions of the AlpArray initiative is to collect top-quality seismological data from a dense network of temporary broadband seismic stations. This complements the existing permanent broadband stations to ensure homogeneous coverage of the Alpine area, with station spacing on the order of 30km. 24 institutions are currently involved in the AlpArray Seismic Network (AASN), which will eventually install over 250 temporary stations in 12 countries. The AASN officially started on 1 January 2016 and will operate for at least 2 years. A complimentary ocean bottom seismometer (OBS) component is expected in 2017.

The Swiss contribution to the AASN is completed, with 23 temporary stations installed in Switzerland, Italy, Bosnia and Herzogovina, Croatia and Hungary. All national broadband stations also contribute to the AASN.

The Seismology and Geodynamics group (SEG) and the Swiss Seismological Service (SED) at the ETH in Zurich take leading roles in the project. Prof Edi Kissling is the project coordinator. Irene Molinari (SEG), John Clinton (SED) and Gyorgy Hetenyi (former SED, now at the University of Lausanne) lead AlpArray working groups, Irene Molinari also manages the Swiss component of the AASN.

More information is available on the project website.

Um in Zukunft die Gefahren von natürlichen und induzierten Erdbeben besser abschätzen zu können, braucht es ein genaueres Verständnis von Verwerfungszonen in tektonisch aktiven Gebieten wie der Schweiz. Mit Hilfe geophysikalischer Abbildungsverfahren und geologischen Kartierungen wurden in den letzten Jahren zahlreiche Verwerfungen in den schweizerischen Alpen und im nördlichen Alpenvorland identifiziert. Allerdings treten in vielen Fällen Erdbeben abseits dieser kartierten Verwerfungen auf, was die Frage nach den tektonischen Prozessen und Mechanismen aufwirft, die diesen Erdbeben zugrunde liegen.  

Ziel dieses Projektes ist es, durch verbesserte geophysikalische Inversionsverfahren die Strukturen von Verwerfungszonen hochauflösend abzubilden und daraus Erkenntnisse über mechanische Eigenschaften der Bruchsysteme abzuleiten. Dazu werden unter anderem Verfahren der seismischen Tomografie mit hochauflösender Erdbebenlokalisierung kombiniert. Die Anwendung konzentriert sich auf zwei Regionen in der Schweiz: (i) einer äußerst aktiven Erdbebenzone nördlich des Rhônetals im Kanton Wallis, (ii) einer Verwerfungszone nahe St. Gallen, die während Stimulationsmaßnahmen für ein geplantes Geothermiekraftwerk aktiviert wurde.

Durch die Anwendung verbesserter Abbildungsverfahren erwarten wir zum einen neue Erkenntnisse über den Zusammenhang zwischen existierenden Verwerfungen und dem Auftreten von Erdbeben. Zudem sind die beiden Untersuchungsgebiete von hoher gesellschaftlicher Relevanz. Das Wallis ist die Region mit der größten seismischen Gefährdung der Schweiz und ein Großteil der gegenwärtigen Seismizität in diesem Gebiet steht in Verbindung mit der Erdbebenzone nördlich des Rhônetals. Die St. Gallen Verwerfungszone bietet Gelegenheit zur Untersuchung der Erdbebengefährdung im dicht besiedelten Molasse Becken, welches potentieller Standort zukünftiger Geothermieprojekte und atomarer Endlager ist.

This project is carried out under a contract with the National Cooperative for the Disposal of Radioactive Waste (Nagra). It provides an independent monitoring of the earthquake activity in the area of the proposed nuclear waste repositories in northern Switzerland. Detailed microseismic analysis will help to identify active fault zones and provide insights into the underlying seismotectonic processes in the vicinity of proposed sites, which has direct implications on the seismic hazard assessment. The SED provides results of this project in a transparent manner and all data acquired are made available for public access (see SED declaration on transparency [link]).

Within this project, the Swiss Seismological Service (SED) constructed ten new seismic stations in northeastern Switzerland and southern Germany to improve monitoring capabilities for very small earthquakes. Monitoring of weak seismic events in this region is challenging, because the study region is densely populated and sediments of the Molasse basin dominate the surface geology. A novel step-wise optimization approach was developed to ensure an optimum configuration of the new stations. To reduce the seismic background noise, three of the ten new sites were equipped with borehole-sensors, located at depths of 120–150 m below the surface. The new stations are fully operational since December 2013 and will observe the local seismicity in northern Switzerland for a minimum of ten years.

The newly installed stations complement the five stations installed in 2003 under a first agreement with Nagra. With these ‘Nagra’ stations, together with stations of the Swiss National Seismic Network and stations of neighboring networks in Germany, the project aims to monitor earthquakes down to magnitudes of 1.0 and smaller in the study area. We intend to achieve an overall catalog completeness of Mc 1.3 throughout northeastern Switzerland and location errors less than 0.5 km in epicenter and less than 2 km in focal depth within the study region. Accurate earthquake locations are essential for seismotectonic interpretations. Within this project we therefore aim to improve location accuracy, especially for focal depths, of past and future earthquakes in the region.

InSight (Interior exploration using Seismic Investigations, Geodesy and Heat Transport) is a NASA Discovery Program mission that will deploy a single geophysical lander on Mars to study its deep interior. This is the first comprehensive surface-based geophysical investigation of Mars. The overarching mission goals are to illuminate the fundamentals of formation and evolution of terrestrial (Earth-like) planets by investigating the interior structure and processes of Mars, and more specifically to determine the thickness, structure and composition of the crust, mantle and core, and to measure the rate and distribution of seismic activity and the rate of meteorite impacts. The Mission should land in November 2018 and have a lifetime of about two Earth years. A set of 3-component broadband and short period seismometers (collectively known as SEIS) will be deployed beside the lander. In additional, InSight will also deploy a heat flow probe (HP3), a geodetic experiment (RISE), a magnetometer, and meteorological sensors. Seismological investigations of Mars have so far been based on modeling and synthetic data; starting in 2018, waveform data will be returned from Mars and the era of 'Seismology on Mars' will begin.

Building on our expertise and infrastructure for earthquake monitoring and seismic data processing on Earth, the SED will take the lead role in the building a catalogue of seismic events recorded by SEIS (the 'Marsquake Service’). This service will comprise automatic and reviewed event detection and characterization of local and teleseismic events, as well as meteor impacts. The goal of this service is to provide a comprehensive high-quality event catalogue for Mars that is critical to the SEIS project, in particular as input to the development of Martian crustal and deep structure models.

We are adapting advanced single-seismometer analysis techniques developed on the Earth to provide locations for Martian seismicity. Creating the Marsquake Service is a collaboration between the SED and the SEG groups at the ETH Zurich.

SED installed 3 seismological stations in the surroundings of the Mont Terri rock laboratory (St Ursanne, Canton Jura): 2 in the tunnel (stations MTI01 and MTI02) and 1 at the surface (MTI03). The Engineering Seismology group is in charge of the site characterization of these 3 stations. Therefore, a literature review and geophysical experiments are performed in order to propose velocity models able to reproduce the ground motion observed at these stations.

Since 2009, the Swiss Seismological Service is renewing and expanding its strong motion network. During the first phase, a total of 30 new accelerometer stations have been installed between 2009 and 2013, both replacing existing strong motion dial-up stations and installing new stations. During the ongoing second phase, 70 more stations are planned to be installed by 2020, including four borehole installations. The renewal project of the Swiss Strong Motion Network was approved by the Swiss Federal Council in February 2009. The project is monitored and supervised by a steering committee headed by the Swiss Federal Office for the Environment (FOEN).

The goals of the enlargement of the Swiss strong motion network are a better spatial coverage of earthquake-prone regions, a better understanding of site effects and thus the verification and improvement of hazard models.

The epicentral areas of relevant past earthquakes have been instrumented in the first phase, namely Aigle (1584), Glarus (1971), Sarnen (1964), Sion-Sierre (1946), Yverdon (1929), Visp (1855), St. Gallen Rhine Valley (1796/96), Altdorf (1774), Brig (1755), Basel (1356), Churwalden/Vaz (1295/1991), etc. Furthermore, the city areas of Zürich, Geneva, Basel, Bern, Lausanne, St. Gallen, Lucerne, Sion, Solothurn, Locarno, Chur, Sierre are relevant sites for free-field installation.

In the second phase, additional epicentral areas of past earthquakes are targeted: Churwalden (1295), Ardez (1504), Ardon (1524), Arbon (1720), Kreuzlingen (1911), and Moudon (1933), among others. Further urban areas are instrumented as well, e.g. Biel, Fribourg, Neuchâtel, Thun and Winterthur. Another aspect of phase 2 is the densification of the network, especially in earthquake-prone areas such as the Valais.

The site selection is always a trade-off between the scientific objectives and the level of vibration disturbances. Modern stations are sensitive enough to record also small earthquakes, but the signal-to-noise ratio may be too low due to traffic and industries. At all sites, geophysical measurements are performed to characterize the site response. Recorded signals can then be interpreted, and sites classified according to the amplification at the site, which is basic information for improved site-specific seismic hazard studies.

The goal of this project is the historical-critical revision of the Swiss earthquake catalogue for the pre-instrumental and early instrumental period of systematic earthquake observations in Switzerland, 1878–1963, with a special focus on events of intermediate magnitude reaching epicentral intensities of IV–VI (EMS-98).

To ensure correct interpretation of the large amount of heterogeneous scientific and non-scientific earthquake records, it is essential to investigate the historical, scientific and technological context of their production. This is achieved through the approaches of historical sciences. The proposed work relies on a systematic investigation and a historically sound assessment of the reliability and factuality of historically reported events.

By the use of descriptive primary and secondary sources, macroseismic intensity fields will be established for intermediate-size earthquakes with an intensity in the range IV–VI (EMS-98). This will facilitate the assessment of event parameters such as magnitude, location and depth in a calibrated procedure that will be developed in the project. These parameters are verified by the analysis of the historical seismograms of the Swiss Seismological Service’s early instrumental station network.

Earthquake can initiate subaquatic mass transport and sediment deformation structures, the deposits of which are stored in the sedimentary archive. Due to their good preservation potential and improving dating methods, these earthquake-related deposits in the sedimentary archive of lakes can extend the earthquake catalogue back to prehistorical times. The determination of frequency, epicentre and magnitude of these paleoearthquakes can be a key input for improving a probabilistic seismic hazard model, especially in areas such as Switzerland where strong earthquakes have long recurrence rates and might not be recorded in historic times. In order to better interpret the prehistorical earthquake-related sedimentary record, it remains important to understand which historical earthquakes did trigger mass transport deposits and sediment deformation structures, and which earthquakes did not trigger secondary sedimentary effects in the lake basins.

The proposed project is aimed to improve our understanding of which earthquakes, or what kind of shaking, leave traces in the sediments, in order to evaluate the completeness of the sublacustrine paleoseismological record, using Lake Thun as example. The second aim of this project, directly linked to the first one, is to compare the paleoseismological record with the contemporary national seismic hazard model of Switzerland, updated in 2015. Thus, using the intensity ranges determined in the first part of this project and using the paleoseismic database of Switzerland a systematic analysis of frequencies, sources and magnitudes of possible paleoearthquakes will be compared to time series produced by the seismic hazard model of Switzerland.

The third phase of this project is split into 3 subtasks with the main goal to improve regional and local seismic hazard assessment in Switzerland; with particular focus to the sites of potential regions of nuclear repositories. These sub-projects are:

Subproject 1 aims to improve models and develop methods for the prediction of strong ground motion in Switzerland at the surface and at depth. Two main approaches are investigated: ground motion prediction equations (GMPEs) and stochastic simulation models. Both approaches require calibration to the local seismicity and careful consideration of their extrapolation to large magnitude events which have, as yet, not been instrumentally recorded in Switzerland. In this context, we study the effect of source- and site-related parameters, as stress drop and kappa.

Subproject 2 is focused on earthquake scenario modeling for Switzerland and exploration of the physical limits on ground motion. Our modeling combines realistic rupture along irregular fault surfaces and wave propagation in complex heterogeneous media at high frequency, with focus on underground repositories. Moreover we investigate the plastic and non-linear behavior of soft sediments when subject to high-amplitude Mach waves, conducting CPT measurements to calibrate our numerical models.

Subproject 3 focuses on numerical modeling the induced seismicity during tunnel excavation. Simulations will be performed using both thermo-hydro-mechanical coupled model and statistical model. We will adapt existing models for induced earthquakes to the conditions typically met in deep geological repositories. Available geomechanical faulting models will be used during the validation and calibration stage. Finally, the results will be used as input for Subproject 1 and 2.

This group gathers the researchers with competences in site characterization in order to discuss and improve the work performed for different projects involving site characterization, especially for sites with new seismic stations. The available tools for processing, archiving and disseminating are shared within the group. The group is responsible for setting up and filling the SED database for site characterization. A process of review has been established in order to validate the work done before it is made public.

The group reviewed the 30 sites of the SSMNet Renewal phase 1, 6 sites of SSMNet stations in Basel installed in the frame of the Basel Erdbebenvorsorge project, the 10 sites of the NAGRA Network, the installation site of stations in the area of the Mont Terri rock laboratory, 2 sites of SSMNet stations in Liechtenstein and the currently installed sites of the SSMNet Renewal Phase 2.

Within this project, new signal processing techniques for the analysis of ambient vibrations are developed. The focus is on the development of single-station methods and the adaptation of existing techniques to large arrays.

Single-station methods represent an extremely valuable tool. The use of a single station further simplifies the measurement procedure thus enabling a more time and cost effective survey.

Assumptions on the wavefield used when processing small arrays may be no longer valid when analyzing large arrays. For this reason, we investigate adaptations of existing techniques to large arrays.

Target applications of the methods developed within this project include microzonation studies in Switzerland and investigation of the characteristics of the Swiss Molasse Basin.

Recent discoveries include: Non-volcanic tremors (NVT), strong heterogeneity of the relative stress distribution, temporal and along-fault variable aseismic creep, repeating earthquakes, slow slip events. Based on the insight that was gained from these previous studies, we will be able to develop and calibrate an indicative stress-meter for the Earth’s crust. In this project we link for the first time the statistical analysis of the size distribution of earthquakes with complementary observations of fault movement.

This yields an improved understanding of the nature of fault loading cycles. Therefore, the results will provide key understanding to unravelling the predictability of earthquakes and has the potential to radically change the assessment of local seismic hazard for selected well-understood and well-monitored faults.

Quantifying time-varying seismicity rates is fundamentally important to protecting people who live in areas subject to extreme earthquake shaking. One primary difficulty with such assessment is determining how faults interact. Some recent studies have noted the ability of passing earthquake waves to increase the 'triggerability' of a fault in a delayed form of dynamic stressing: after seismic waves pass, faults are more prone to fail in a subsequent earthquake. The deadly Canterbury earthquake sequence has characteristics that suggest it was promoted by such distant, delayed, dynamic triggering. The sequence is also compatible with a model in which low-strain rate areas are efficient at storing and transferring static stresses. This has implications for earthquake clustering and the generation of damaging ground motion. We will apply recently-developed techniques in concert to address three questions: 1) Can we quantify distant and delayed triggering in this sequence? We will address this by correlating increased geodetic crustal velocities in Canterbury following the 2009 M7.8 Dusky Sound earthquake that occurred hundreds of km away. We will apply source scanning and template matching techniques to search Canterbury for microseismicity that was triggered by the M7.8 event. 2) Do earthquakes in low-strain rate areas exhibit more clustering and longer aftershock sequences than their high-strain rate counterparts, and do these earthquakes produce stronger ground motions? We will build a comprehensive model of earthquake generation in low-strain rate areas by using an earthquake simulator to model the evolution of the sequence. 3) Can the simulator model we develop demonstrate skill in seismicity forecast experiments? The model developed in this project could provide a true step change and bring seismology closer to bridging the gap between probabilistic forecasting and deterministic modelling of earthquake hazard.

Our present knowledge about earthquakes does not yet allow us to reliably forecast earthquakes. Therefore, the study of precursors is an essential step in the direction of earthquake forecasting. Precursors can be anomalous seismic patterns or other phenomena such as peculiar animal behavior, or electromagnetic anomalies etc., which indicate the incidence of a large event. We focus in our studies on seismic precursors, namely quiescence, which is expressed through reduced seismic activity, accelerated seismicity (ASR) and short term foreshocks. The mentioned precursors are observed in many selected earthquake sequences in the past. However, there is skepticism if these precursors happen systematically; some studies explain their occurrence rather as a random temporary perturbation of normal seismicity which is accidentally followed by a large earthquake.

We believe that systematic investigations on the occurrence of precursors give essential evidence for or against their existence. We chose to perform these investigations with statistical tools, hence by evaluating location, time and magnitudes of earthquakes from several regional earthquake catalogs of the world, to obtain representative precursor statistics.

We find that small earthquakes, as they occur more frequently, could facilitate the detection of precursory patterns (Mignan, 2014). We study statistical models used to describe earthquake occurrence and the impact of the choice of the lowest magnitude on them (Seif et al, 2016, submitted). Using these models we evaluate if foreshock occurrence differs from normal seismicity. We also want to specify how often foreshock patterns are followed by large events or not (true/false alarm rate). In the future remaining precursory patterns, quiescence and accelerated seismicity, will be investigated in the same way. We hope that the statistical analysis will allow us to better understand the physical processes which lead to the occurrence of precursors.

Statistical seismology is the application of rigorous statistical methods to earthquake science with the goal of improving our knowledge of how the earth works. Within statistical seismology there is a strong emphasis on the analysis of seismicity data in order to improve our scientific understanding of earthquakes and to improve the evaluation and testing of earthquake forecasts, earthquake early warning, and seismic hazards assessments. Given the societal importance of these applications, statistical seismology must be done well. Unfortunately, a lack of educational resources and available software tools make it difficult for students and new practitioners to learn about this discipline. The goal of the Community Online Resource for Statistical Seismicity Analysis (CORSSA) is to promote excellence in statistical seismology by providing the knowledge and resources necessary to understand and implement the best practices.

CORSSA covers a wide variety of themes:

Introductory Material
Statistical Foundations
Understanding Seismicity Catalogs and Their Problems
Models and Techniques for Analyzing Seismicity
Earthquake Predictability and Related Hypothesis Testing

Each of these themes includes a series of articles that are listed in the CORSSA Table of Contents. The series of themes was devised to make it easy for the reader to focus on their personal requirements to get an introduction to statistical seismology, or to learn about the basics of earthquakes, statistics, and/or the intricacies of seismicity catalogs before moving onto applications.

This is the 1st subproject of the „ENSI – SED-Erdbebenforschung zu Schweizer Kernanlagen“ project.

This subproject aims to improve source-scaling and attenuation models and to develop methods for the prediction of strong ground motion in Switzerland both at the surface as well as at depth. Two main approaches are investigated: ground motion prediction equations (GMPEs) and stochastic simulation models. Both approaches require adaptions to the local seismicity and careful consideration of their calibration to Swiss conditions. The Fourier spectral and stochastic models correspond to the current state of research and have some advantages over the empirical attenuation relationships, as it is possible to adjust the model to specific local site conditions. The complete understanding in terms of physical parameterization of such models is crucial in order to decouple different effects, which allow building robust predictive models that scale appropriately to large magnitude events. In this regard, variability in source parameter such as stress drop is crucial. Similarly, variability in site-related attenuation parameter kappa is also need to be well understood. Fourier and duration models from Japanese data are now being developed that will allow the review of the Swiss model for large magnitudes in the different distance ranges and at various rock sites which have, as yet, not been instrumentally recorded in Switzerland. Moreover, we will use recordings of local seismicity in addition to numerical modeling results of related projects to calibrate the predictive models. The long-term goal is to develop an improved stochastic simulation model for Switzerland allowing existing uncertainties to be reduced. In future, such models will also allow an assessment of ground motion caused by induced seismicity due to the activation of existing fractures and/or the generation of new fractures.

This is the 2nd subproject of the „ENSI – SED-Erdbebenforschung zu Schweizer Kernanlagen“ project.

Realistic modeling of earthquake ground motions can be achieved only if the causative fault and the medium where waves propagate are described accurately. To efficiently simulate high-frequency earthquake scenario in Switzerland, we develop a hybrid broadband simulation technique combining state-of-the-art rupture models along irregular fault surfaces and wave propagation in complex heterogeneous media taking into account station-specific scattering parameters. Significant efforts are made to extend the validity of our technique to model ground motion at depth for possible applications at underground repositories. A thorough validation and calibration of sensitive parameters is based on Japanese and Swiss datasets. Along with source and path effects, near-surface site conditions represent an important factor controlling ground motions since soft sediments can significantly amplify the shaking observed during an earthquake. Depending on the level of input ground motion, liquefiable soils have the potential to generate excess water pressure resulting in high-frequency acceleration pulses. Advanced constitutive models of liquefiable soils require the knowledge of so-called dilatancy parameters that describe the potential to generate excess water pressure. These parameters can be determined from field observations by analyzing cone penetration test (CPT) measurements. Accurate modeling of liquefiable soils subject to high-amplitude Mach waves is essential to explore the physical limits of ground motion.