Automated hypocenter determination of tectonic tremors in the Nankai subduction zone using convolutional neural networks combined with semblance analysis

Abstract Recent advances in deep learning have enhanced our ability to analyze seismic waveforms. Here, we developed and evaluated a convolutional neural network (CNN) model to classify tectonic tremors, earthquakes, and noise in seismic waveform data recorded by a seismic array in the Nankai subduc...

Full description

Saved in:
Bibliographic Details
Main Authors: Amane Sugii, Yoshihiro Hiramatsu, Takahiko Uchide, Kazutoshi Imanishi
Format: Article
Language:English
Published: SpringerOpen 2024-12-01
Series:Earth, Planets and Space
Subjects:
Online Access:https://doi.org/10.1186/s40623-024-02126-0
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846101383653097472
author Amane Sugii
Yoshihiro Hiramatsu
Takahiko Uchide
Kazutoshi Imanishi
author_facet Amane Sugii
Yoshihiro Hiramatsu
Takahiko Uchide
Kazutoshi Imanishi
author_sort Amane Sugii
collection DOAJ
description Abstract Recent advances in deep learning have enhanced our ability to analyze seismic waveforms. Here, we developed and evaluated a convolutional neural network (CNN) model to classify tectonic tremors, earthquakes, and noise in seismic waveform data recorded by a seismic array in the Nankai subduction zone. The trained CNN model achieved high accuracy, with both precision and recall exceeding 97%, and correctly detected 96% of distant earthquakes. The probability of tectonic tremor as a function of the signal-to-noise ratio (SNR) increased steeply from 10 to 90% at an SNR of 4. We highlighted tectonic tremor waveforms using the integrated gradients (IG) method for interpreting CNN models. IG filter averaging over the stations of an array outperforms bandpass filters and other interpretation methods for CNN models in locating tectonic tremors by semblance analysis, providing the largest number of tectonic tremor sources. As reported previously, located sources of tectonic tremor during episodic tremor and slip events migrate along the strike of the subducting plate. The source location error increases significantly at epicentral distances greater than 30 km because of low SNRs. The technique developed in this study equips CNN models with a high ability to distinguish tectonic tremors and earthquakes from noise and to locate tectonic tremors with sources that are not far from seismic stations. Graphical abstract
format Article
id doaj-art-0eb3bdf1e8cf4f81b86eb7b2c81121bc
institution Kabale University
issn 1880-5981
language English
publishDate 2024-12-01
publisher SpringerOpen
record_format Article
series Earth, Planets and Space
spelling doaj-art-0eb3bdf1e8cf4f81b86eb7b2c81121bc2024-12-29T12:12:59ZengSpringerOpenEarth, Planets and Space1880-59812024-12-0176111610.1186/s40623-024-02126-0Automated hypocenter determination of tectonic tremors in the Nankai subduction zone using convolutional neural networks combined with semblance analysisAmane Sugii0Yoshihiro Hiramatsu1Takahiko Uchide2Kazutoshi Imanishi3Graduate School of Natural Science and Technology, Kanazawa UniversityFaculty of Geosciences and Civil Engineering, Institute of Science and Engineering, Kanazawa UniversityGeological Survey of Japan, National Institute of Advanced Industrial Science and Technology (AIST)Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology (AIST)Abstract Recent advances in deep learning have enhanced our ability to analyze seismic waveforms. Here, we developed and evaluated a convolutional neural network (CNN) model to classify tectonic tremors, earthquakes, and noise in seismic waveform data recorded by a seismic array in the Nankai subduction zone. The trained CNN model achieved high accuracy, with both precision and recall exceeding 97%, and correctly detected 96% of distant earthquakes. The probability of tectonic tremor as a function of the signal-to-noise ratio (SNR) increased steeply from 10 to 90% at an SNR of 4. We highlighted tectonic tremor waveforms using the integrated gradients (IG) method for interpreting CNN models. IG filter averaging over the stations of an array outperforms bandpass filters and other interpretation methods for CNN models in locating tectonic tremors by semblance analysis, providing the largest number of tectonic tremor sources. As reported previously, located sources of tectonic tremor during episodic tremor and slip events migrate along the strike of the subducting plate. The source location error increases significantly at epicentral distances greater than 30 km because of low SNRs. The technique developed in this study equips CNN models with a high ability to distinguish tectonic tremors and earthquakes from noise and to locate tectonic tremors with sources that are not far from seismic stations. Graphical abstracthttps://doi.org/10.1186/s40623-024-02126-0Slow earthquakesDenoisingNeural network attributionIntegrated gradients
spellingShingle Amane Sugii
Yoshihiro Hiramatsu
Takahiko Uchide
Kazutoshi Imanishi
Automated hypocenter determination of tectonic tremors in the Nankai subduction zone using convolutional neural networks combined with semblance analysis
Earth, Planets and Space
Slow earthquakes
Denoising
Neural network attribution
Integrated gradients
title Automated hypocenter determination of tectonic tremors in the Nankai subduction zone using convolutional neural networks combined with semblance analysis
title_full Automated hypocenter determination of tectonic tremors in the Nankai subduction zone using convolutional neural networks combined with semblance analysis
title_fullStr Automated hypocenter determination of tectonic tremors in the Nankai subduction zone using convolutional neural networks combined with semblance analysis
title_full_unstemmed Automated hypocenter determination of tectonic tremors in the Nankai subduction zone using convolutional neural networks combined with semblance analysis
title_short Automated hypocenter determination of tectonic tremors in the Nankai subduction zone using convolutional neural networks combined with semblance analysis
title_sort automated hypocenter determination of tectonic tremors in the nankai subduction zone using convolutional neural networks combined with semblance analysis
topic Slow earthquakes
Denoising
Neural network attribution
Integrated gradients
url https://doi.org/10.1186/s40623-024-02126-0
work_keys_str_mv AT amanesugii automatedhypocenterdeterminationoftectonictremorsinthenankaisubductionzoneusingconvolutionalneuralnetworkscombinedwithsemblanceanalysis
AT yoshihirohiramatsu automatedhypocenterdeterminationoftectonictremorsinthenankaisubductionzoneusingconvolutionalneuralnetworkscombinedwithsemblanceanalysis
AT takahikouchide automatedhypocenterdeterminationoftectonictremorsinthenankaisubductionzoneusingconvolutionalneuralnetworkscombinedwithsemblanceanalysis
AT kazutoshiimanishi automatedhypocenterdeterminationoftectonictremorsinthenankaisubductionzoneusingconvolutionalneuralnetworkscombinedwithsemblanceanalysis