journal article Open Access Jan 03, 2024

Earthquake or blast? Classification of local-distance seismic events in Sweden using fully connected neural networks

View at Publisher Save 10.1093/gji/ggae018
Abstract
SUMMARY
Distinguishing between different types of seismic events is a task typically performed manually by expert analysts and can thus be both time and resource expensive. Analysts at the Swedish National Seismic Network (SNSN) use four different event types in the routine analysis: natural (tectonic) earthquakes, blasts (e.g. from mines, quarries and construction) and two different types of mining-induced events associated with large, underground mines. In order to aid manual event classification and to classify automatic event definitions, we have used fully connected neural networks to implement classification models which distinguish between the four event types. For each event, we bandpass filter the waveform data in 20 narrow-frequency bands before dividing each component into four non-overlapping time windows, corresponding to the P phase, P coda, S phase and S coda. In each window, we compute the root-mean-square amplitude and the resulting array of amplitudes is then used as the neural network inputs. We compare results achieved using a station-specific approach, where individual models are trained for each seismic station, to a regional approach where a single model is trained for the whole study area. An extension of the models, which distinguishes spurious phase associations from real seismic events in automatic event definitions, has also been implemented. When applying our models to evaluation data distinguishing between earthquakes and blasts, we achieve an accuracy of about 98 per cent for automatic events and 99 per cent for manually analysed events. In areas located close to large underground mines, where all four event types are observed, the corresponding accuracy is about 90 and 96 per cent, respectively. The accuracy when distinguishing spurious events from real seismic events is about 95 per cent. We find that the majority of erroneous classifications can be traced back to uncertainties in automatic phase picks and location estimates. The models are already in use at the SNSN, both for preliminary type predictions of automatic events and for reviewing manually analysed events.
Topics

No keywords indexed for this article. Browse by subject →

References
35
[1]
Abadi (2015)
[2]
Båth "Short-period rayleigh waves from near-surface events" Phys. Earth planet. Inter. (1975) 10.1016/0031-9201(75)90064-3
[3]
Baumgardt "Regional seismic waveform discriminants and case-based event identification using regional arrays" Bull. seism. Soc. Am. (1990)
[4]
Böðvarsson "The sil data acquisition system—at present and beyond year 2000" Phys. Earth planet. Inter. (1999) 10.1016/s0031-9201(99)00032-1
[5]
Random Forests

Leo Breiman

Machine Learning 2001 10.1023/a:1010933404324
[6]
Burkov (2019)
[7]
SMOTE: Synthetic Minority Over-sampling Technique

N. V. Chawla, K. W. Bowyer, L. O. Hall et al.

Journal of Artificial Intelligence Research 2002 10.1613/jair.953
[8]
Chollet (2015)
[9]
Dysart "Regional seismic event classification at the noress array: seismological measurements and the use of trained neural networks" Bull. seism. Soc. Am. (1990)
[10]
Furumura "Lg wave propagation in the area around japan: observations and simulations" Prog. Earth planet. Sci. (2014) 10.1186/2197-4284-1-10
[11]
Goodfellow (2016)
[12]
Gregersen (2021) 10.1017/9781108779906
[13]
Array programming with NumPy

Charles R. Harris, K. Jarrod Millman, Stéfan J. van der Walt et al.

Nature 2020 10.1038/s41586-020-2649-2
[14]
Hourcade "New cnn-based tool to discriminate anthropogenic from natural low magnitude seismic events" Geophys. J. Int. (2023) 10.1093/gji/ggac441
[15]
Matplotlib: A 2D Graphics Environment

John D. Hunter

Computing in Science & Engineering 2007 10.1109/mcse.2007.55
[16]
Juhlin "Reflection seismic studies over the end-glacial burträsk fault, skellefteå, sweden" Solid Earth (2011) 10.5194/se-2-9-2011
[17]
Kim "Discrimination of earthquakes and explosions in the eastern united states using regional high-frequency data" Geophys. Res. Lett. (1993) 10.1029/93gl01267
[18]
Kintner "Testing a local-distance r g/s g discriminant using observations from the bighorn region, wyoming" Bull. seism. Soc. Am. (2020) 10.1785/0120190188
[19]
Kong "Deep convolutional autoencoders as generic feature extractors in seismological applications" Artif. Intell. Geosci. (2021) 10.1016/j.aiig.2021.12.002
[20]
Combining Deep Learning With Physics Based Features in Explosion‐Earthquake Discrimination

Qingkai Kong, Ruijia Wang, William R. Walter et al.

Geophysical Research Letters 2022 10.1029/2022gl098645
[21]
Kortström "Automatic classification of seismic events within a regional seismograph network" Comput. Geosci. (2016) 10.1016/j.cageo.2015.11.006
[22]
Lindblom "Microearthquakes illuminate the deep structure of the endglacial pärvie fault, Northern Sweden" Geophys. J. Int. (2015) 10.1093/gji/ggv112
[23]
Deep Learning Models Augment Analyst Decisions for Event Discrimination

Lisa Linville, Kristine Pankow, Timothy Draelos

Geophysical Research Letters 2019 10.1029/2018gl081119
[24]
Lund "The Modern Swedish National Seismic Network: two decades of intraplate microseismic observation" Seismol. Res. Lett. (2021) 10.1785/0220200435
[25]
Data Structures for Statistical Computing in Python

Wes McKinney

Proceedings of the Python in Science Conference 2010 10.25080/majora-92bf1922-00a
[26]
Miao "High-accuracy discrimination of blasts and earthquakes using neural networks with multiwindow spectral data" Seismol. Res. Lett. (2020) 10.1785/0220190084
[27]
O’Rourke "Using p/s amplitude ratios for seismic discrimination at local distancesusing p/s amplitude ratios for seismic discrimination at local distances" Bull. seism. Soc. Am. (2016) 10.1785/0120160035
[28]
Pedregosa "Scikit-learn: machine learning in Python" J. Mach. Learn. Res. (2011)
[29]
Press "Two slow surface waves across North America" Bull. seism. Soc. Am. (1952) 10.1785/bssa0420030219
[30]
Sasaki "The truth of the f-measure" Teach Tutor Mater (2007)
[31]
SNSN (1904)
[32]
Srivastava "Dropout: a simple way to prevent neural networks from overfitting" J. Mach. Learn. Res. (2014)
[33]
Tibi "Depth discrimination using rg-to-sg spectral amplitude ratios for seismic events in Utah recorded at local distancesdepth discrimination using rg-to-sg spectral amplitude ratios for seismic events in utah" Bull. seism. Soc. Am. (2018) 10.1785/0120170257
[34]
Tibi "Classification of local seismic events in the utah region: a comparison of amplitude ratio methods with a spectrogram-based machine learning approach" Bull. seism. Soc. Am. (2019) 10.1785/0120190150
[35]
Wessel "The generic mapping tools version 6" Geochem. Geophys. Geosyst. (2019) 10.1029/2019gc008515
Metrics
14
Citations
35
References
Details
Published
Jan 03, 2024
Vol/Issue
236(3)
Pages
1728-1742
License
View
Funding
Swedish Defence Research Agency
Cite This Article
Gunnar Eggertsson, Björn Lund, Michael Roth, et al. (2024). Earthquake or blast? Classification of local-distance seismic events in Sweden using fully connected neural networks. Geophysical Journal International, 236(3), 1728-1742. https://doi.org/10.1093/gji/ggae018
Related

You May Also Like