Clustering earthquake signals and background noises in continuous seismic data with unsupervised deep learning

Seydoux, Léonard; Balestriero, Randall; Poli, Piero; Hoop, Maarten de; Campillo, Míchel; Baraniuk, Richard G.

doi:10.1038/s41467-020-17841-x

Cited by 110 publications

(119 citation statements)

References 40 publications

Supporting

Mentioning

116

Contrasting

Order By: Relevance

“…For what concerns the ground motion amplitude data, the availability of these metadata can be of relevance in combination with the shakemaps to develop new tools for rapid earthquake ground motion estimation. Other applications of the data collection include the adoption of unsupervised ML algorithms to group the waveforms independently of the earthquake location and just on the waveform themselves (e.g., Seydoux et al, 2020). IN-STANCE can also be used, as a dataset with a large number of data, for creating pretrained models when using transfer learning techniques either for seismological or other applications which use time-series data (Otović et al, 2021).…”

Section: Applicationsmentioning

confidence: 99%

INSTANCE – the Italian seismic dataset for machine learning

Michelini

Cianetti

Gaviano

et al. 2021

Earth Syst. Sci. Data

View full text Add to dashboard Cite

Abstract. The Italian earthquake waveform data are collected here in a dataset suited for machine learning analysis (ML) applications. The dataset consists of nearly 1.2 million three-component (3C) waveform traces from about 50 000 earthquakes and more than 130 000 noise 3C waveform traces, for a total of about 43 000 h of data and an average of 21 3C traces provided per event. The earthquake list is based on the Italian Seismic Bulletin (http://terremoti.ingv.it/bsi, last access: 15 February 2020) of the Istituto Nazionale di Geofisica e Vulcanologia between January 2005 and January 2020, and it includes events in the magnitude range between 0.0 and 6.5. The waveform data have been recorded primarily by the Italian National Seismic Network (network code IV) and include both weak- (HH, EH channels) and strong-motion (HN channels) recordings. All the waveform traces have a length of 120 s, are sampled at 100 Hz, and are provided both in counts and ground motion physical units after deconvolution of the instrument transfer functions. The waveform dataset is accompanied by metadata consisting of more than 100 parameters providing comprehensive information on the earthquake source, the recording stations, the trace features, and other derived quantities. This rich set of metadata allows the users to target the data selection for their own purposes. Much of these metadata can be used as labels in ML analysis or for other studies. The dataset, assembled in HDF5 format, is available at http://doi.org/10.13127/instance (Michelini et al., 2021).

show abstract

Section: Applicationsmentioning

confidence: 99%

INSTANCE – the Italian seismic dataset for machine learning

Michelini

Cianetti

Gaviano

et al. 2021

Earth Syst. Sci. Data

View full text Add to dashboard Cite

show abstract

“…Template matching is often used to detect aftershocks (Peng and Zhao, 2009), but also in volcano seismology to detect VT swarms (Passarelli et al, 2018), to characterize pre-eruptive sequences (Lengliné et al, 2016) or Long Period subevent multiplets (Matoza et al, 2015). The unsupervized approach on the other hand is usually more useful for tackling the analysis of continuous data (Seydoux et al, 2020). The first task is to recognize the possible presence of coherent regimes in the Lipari dataset.…”

Section: Machine Learning Approachesmentioning

confidence: 99%

The Seismicity of Lipari, Aeolian Islands (Italy) From One-Month Recording of the LIPARI Array

et al. 2021

View full text Add to dashboard Cite

Seismic activity in volcanic settings could be the signature of processes that include magma dynamics, hydrothermal activity and geodynamics. The main goal of this study is to analyze the seismicity of Lipari Island (Southern Tyrrhenian Sea) to characterize the dynamic processes such as the interaction between pre-existing structures and hydrothermal processes affecting the Aeolian Islands. We deployed a dense seismic array of 48 autonomous 3-component nodes. For the first time, Lipari and its hydrothermal field are investigated by a seismic array recording continuously for about a month in late 2018 with a 0.1–1.5 km station spacing. We investigate the distribution and evolution of the seismicity over the full time of the experiment using self-organized maps and automatic algorithms. We show that the sea wave motion strongly influences the background seismic noise. Using an automatic template matching approach, we detect and locate a seismic swarm offshore the western coast of Lipari. This swarm, made of transient-like signals also recognized by array and polarization analyses in the time and frequency domains, is possibly associated with the activation of a NE-SW fault. We also found the occurrence of hybrid events close to the onshore Lipari hydrothermal system. These events suggest the involvement of hot hydrothermal fluids moving along pre-existing fractures. Seismological analyses of one month of data detect signals related to the regional tectonics, hydrothermal system and sea dynamics in Lipari Island.

show abstract

“…Well-controlled laboratory analogs to faults lack the geologic complexity of the Earth, yet, weak natural background vibrations of a similar sort, that again were thought to be random noise, have been shown to embody information that can be used to predict the onset time of slow slip events in the Cascadia subduction zone 10 . Finally, unsupervised deep learning, in which algorithms are used to discern patterns in data without the benefit of prior labels, applied to seismic waveform data uncovered precursory signals preceding the large and damaging 2017 landslide and tsunami in Greenland 11 .…”

mentioning

confidence: 99%

Machine learning and earthquake forecasting—next steps

2021

View full text Add to dashboard Cite

A new generation of earthquake catalogs developed through supervised machine-learning illuminates earthquake activity with unprecedented detail. Application of unsupervised machine learning to analyze the more complete expression of seismicity in these catalogs may be the fastest route to improving earthquake forecasting.The past 5 years have seen a rapidly accelerating effort in applying machine learning to seismological problems. The serial components of earthquake monitoring workflows include: detection, arrival time measurement, phase association, location, and characterization. All of these tasks have seen rapid progress due to effective implementation of machine-learning approaches. They have proven opportune targets for machine learning in seismology mainly due to the large, labeled data sets, which are often publicly available, and that were constructed through decades of dedicated work by skilled analysts. These are the essential ingredient for building complex supervised models. Progress has been realized in research mode to analyze the details of seismicity well after the earthquakes being studied have occurred, and machinelearning techniques are poised to be implemented in operational mode for real-time monitoring. We will soon have a next generation of earthquake catalogs that contain much more information. How much more? These more complete catalogs typically feature at least a factor of ten more earthquakes (Fig. 1) and provide a higher-resolution picture of seismically active faults.This next generation of earthquake catalogs will not be the single, static objects seismologists are accustomed to working with. For example, less than 2 years after the 2019 Ridgecrest, California earthquake sequence there already exist four next-generation catalogs, each of which were developed with different enhanced detection techniques. Now, and in the future, this will be the norm, and earthquake catalogs will be updated and improved-potentially dramaticallywith time. Second-generation deep learning models 1 that are specifically designed based on earthquake signal characteristics and that mimic the manual processing by analysts, can lead to performance increases beyond those offered by earlier models that adapted neural network architectures from other fields. Those interested in using earthquake catalogs for forecasting can anticipate a shifting landscape with continuing improvements.While these improvements are impressive, the value of the extra information they provide is less clear. What will we learn about earthquake behavior from these deeper catalogs and how might it improve the prospects for the stubbornly difficult problem of earthquake forecasting? Short-term deterministic earthquake prediction remains elusive and is perhaps impossible; however, probabilistic earthquake forecasting is another matter. It remains the subject of focused and sustained attention and it informs earthquake hazard characterization 2 and thus both policy and earthquake risk reduction. A key assumption is that what we learn from...

show abstract

Clustering earthquake signals and background noises in continuous seismic data with unsupervised deep learning

Cited by 110 publications

References 40 publications

INSTANCE – the Italian seismic dataset for machine learning

INSTANCE – the Italian seismic dataset for machine learning

The Seismicity of Lipari, Aeolian Islands (Italy) From One-Month Recording of the LIPARI Array

Machine learning and earthquake forecasting—next steps

Contact Info

Product

Resources

About