Back to overview

detect-µ - High-resolution Imaging and Analysis of Seismicity Patterns in Swiss Microearthquake Sequences

Applicant Kraft Toni
Number 188615
Funding scheme Project funding (Div. I-III)
Research institution Schweizerischer Erdbebendienst ETH Zürich
Institution of higher education ETH Zurich - ETHZ
Main discipline Geophysics
Start/End 01.11.2019 - 31.10.2022
Approved amount 163'553.00
Show all

All Disciplines (2)

Other disciplines of Earth Sciences

Keywords (7)

Induced Seismicity; Foreshock Occurance; Microseismic Analysis and Monitoring; Earthquake Statistic; Earthquake Source Physics; Earthquake Swarms; Seismicity Patterns

Lay Summary (German)

Erdbeben treten häufig in Sequenzen auf, die unterschiedliche zeitliche und räumliche Erscheinungsmuster haben. Diese Muster werden grob in drei Klassen eingeteilt: Vor-Haupt-Nachbeben-Sequenzen, Haupt-Nachbeben-Sequenzen und Erdbebenschwärme. Es treten aber auch Überlagerungen und Abweichungen von diesen einfachen Klassen auf, die zu sehr einfachen und sehr komplexen Erdbebensequenzen führen können. Aus der Untersuchung dieser Muster kann man viel über die Physik von Erdbeben lernen. Das Studium kleiner Erdbeben kann daher dazu beitragen, grosse Erdbeben besser zu verstehen.
Lay summary

Inhalt und Ziel

Das Hauptziel unseres Projekts ist die systematische Analyse und Klassifizierung der schweizerischen Erdbebensequenzen. Damit wollen wir die Grundlage schaffen und nutzen, die Mechanismen besser zu verstehen, die ihre vielfältigen Erscheinungsmuster erzeugen. Dazu verwenden wir eine kürzlich am Schweizerischen Erdbebendienst (SED) entwickelte Methode, die moderne Verfahren des maschinellen Lernens mit gut etablierten seismologischen und statistischen Auswerteverfahren kombiniert. Mit dieser Methode können wir Kataloge der Sequenzen erstellen, die über den gesamten Untersuchungszeitraum konsistente Detektionsschwellen, Lokalisierungen und Magnituden haben. Im Detail werden wir unter anderem (i) das langfristige Verhalten von Erdbebensequenzen in der Schweiz, (ii) mögliche Vorbeben und Wiederholungsbeben, (iii) die geologischen Bedingungen unter denen verschiedene Erdbebensequenzen auftreten und (iv) deren Reaktion auf externe Einflüsse, wie Regen oder die Gezeiten, untersuchen.


Die Ergebnisse unseres Projektes helfen abzuklären, warum und auf welchen tektonischen Strukturen, Erdbeben in der Schweiz stattfinden. Diese Wissen kann dabei helfen, geologische Prozesse besser zu verstehen (z.B. Alpenbildung) und die Erdbebengefährdung besser abzuschätzen. Unsere Erdbenkataloge können verstehen helfen, wie ein Erdbebenbruch beginnt und dazu beitragen abzuklären, ob dieser Prozess berechenbar ist. Diese Erkenntnisse können auch geotechnischen Anwendung, wie die Geothermie, sicherer machen.

Direct link to Lay Summary Last update: 09.10.2019

Lay Summary (English)

Earthquakes often occur in sequences that have different temporal and spatial patterns. These patterns are roughly divided into three classes: fore-main-aftershock sequences, main-aftershock sequences, and earthquake swarms. However, overlaps and deviations from these simple classes also occur, which can lead to very simple and very complex earthquake sequences. The analysis of these patterns can help us to learn a lot about the physics of earthquakes. The study of small earthquakes can, therefore, contribute to a better understanding of large earthquakes.
Lay summary

Content and aim

The main objective of our project is the systematic analysis and classification of the Swiss earthquake sequences. In this way, we want to create and use the basis for a better understanding of the mechanisms that generate their diverse occurrence patterns. We use a method recently developed at the Swiss Seismological Service (SED) that combines modern machine learning with well-established seismological and statistical analysis techniques. With this method, we can create catalogs of sequences that have consistent detection thresholds, locations, and magnitudes over the entire period of investigation. In detail, we will investigate (i) the long-term behavior of earthquake sequences in Switzerland, (ii) possible foreshocks and repeating earthquakes, (iii) the geological conditions under which different earthquake sequences occur, and (iv) their response to external influences such as rainfall or earth tides.


The results of our project help to clarify why and on which tectonic structures earthquakes occur in Switzerland. This knowledge can help to understand geological processes (e.g., alpine formation) better and to assess earthquake hazard better. Our earthquake catalogs can help to understand how an earthquake rupture starts and to clarify whether this process is deterministic. These findings can also make geotechnical applications, such as geothermal energy, safer.

Direct link to Lay Summary Last update: 09.10.2019

Responsible applicant and co-applicants


Project partner

Associated projects

Number Title Start Funding scheme
169178 A decameter-scale reservoir stimulation experiment - the full hydro-mechanical response of a fault zone to high-pressure water injection 01.11.2016 Project funding (Div. I-III)
154434 SWISS-AlpArray - Assessing Alpine Orogeny in 4D-space-time Frame 01.10.2014 Sinergia
163153 Structure And Mechanics of Seismogenic Fault Zones: Insights from advanced passive and active seismic imaging 01.05.2016 Project funding (Div. I-III)


A general feature of seismicity is that earthquakes tend to cluster in space and time. This results in earthquake sequences that show a rich variety of spatiotemporal occurrence patterns. A primary classification in mainshock-aftershock, foreshock-mainshock-aftershock and earthquake swarm sequences was proposed decades ago. Yet, why earthquake sequences are so different is still a matter of debate. Specifically, foreshock occurrence is still poorly understood. To improve our knowledge in this field, we need to investigate in more detail how different seismotectonic and geological settings, and external forcing influence seismicity patterns, and how earthquakes nucleate in general.Over the past several decades, rupture phenomena have been studied intensively on the centimeter scale in controlled laboratory experiments. Detailed models based on elastic crack theory or rate-and-state friction have been developed from these data sets. These models predict that rupture starts with a nucleation phase, in which rupture first grows slowly and quasi-statically, then accelerates quasi-dynamically until it reaches a critical size, and finally grows on dynamically. These nucleation processes have been observed in many densely instrumented lab experiments, where additionally acoustic emissions (i.e., tiny dynamic instabilities) seem to occur at an acceleration rate in the growing nucleation zone. By contrast, until present, only a limited number of detailed observations of nucleation in real earthquakes exist - mainly for magnitudes above M4. Because of the huge observational gap between the lab and field scale, it is still debated if and how the results obtained at the first can be transferred to the second. This observational gap may also be the reason why foreshock occurrence is predominantly explained by presilp in studies that cover large magnitude ranges, whereas the cascading-up of small dynamic instabilities is favored in the others. This observational gap needs to be eliminated, to advance our understanding of seismicity patterns.Recently, we have developed an analysis method for earthquake sequences that can improve the sensitivity and resolution of existing earthquake catalogs by several orders of magnitude. The method combines a highly sensitive template-matching detector with high-precision earthquake relocation and high-resolution statistical analysis. A semiautomatic workflow enables the efficient analysis of several-decade-long data in a consistent way, with constant high quality. The potential of our method to close the mentioned data gap was recently demonstrated by the analysis of the earthquake sequence of Diemtigen, CH, where we were able to identify and study lab-experiment-like foreshock phenomena for earthquakes with magnitudes below ML3.2.In the proposed project we plan to perform the first systematic high-precision analysis of all earthquake sequences with mainshock magnitude ML>=2.5 that occurred in Switzerland in the last 15 years. We start by improving the existing SED catalog using template-matching. Then perform a high-resolution statistical analysis of all sequences, to study their temporal magnitude-frequency behavior. Finally, we will derive high-precision relative locations, to examine the spatiotemporal behavior of the sequences. For selected sequences active in the project period, we plan to further improve the data quality by installing temporary seismic stations. From this analysis, we will develop a systematic classification of seismicity patterns in Switzerland, and study how they correlate with seismotectonic and geological setting, source depth, and other parameters. We will also investigate the long-term behavior and driving mechanisms of the sequences, as well as the influence of external forcing (e.g., remote triggering; rainfall; earth tides). The data will also allow us to study earthquake nucleation and earthquake physics in general. As a community service, we plan to implement a fully automated analysis scheme for sequences of specific interest and to publish their high-resolution earthquake catalogs via the SED internet data portal.