Estimating S-wave velocity profiles from horizontal-to-vertical spectral ratios based on deep learning

S-wave velocity (Vs) profile or time averaged Vs to 30 m depth (VS30) is indispensable information to estimate the local site amplification of ground motion from earthquakes. We use a horizontal-to-vertical spectral ratio (H/V) of seismic ambient noise to estimate the Vs profiles or VS30. The measur...

Full description

Saved in:
Bibliographic Details
Main Authors: Koichi Hayashi, Toru Suzuki, Tomio Inazaki, Chisato Konishi, Haruhiko Suzuki, Hisanori Matsuyama
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:Soils and Foundations
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S0038080624001033
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850255828448182272
author Koichi Hayashi
Toru Suzuki
Tomio Inazaki
Chisato Konishi
Haruhiko Suzuki
Hisanori Matsuyama
author_facet Koichi Hayashi
Toru Suzuki
Tomio Inazaki
Chisato Konishi
Haruhiko Suzuki
Hisanori Matsuyama
author_sort Koichi Hayashi
collection DOAJ
description S-wave velocity (Vs) profile or time averaged Vs to 30 m depth (VS30) is indispensable information to estimate the local site amplification of ground motion from earthquakes. We use a horizontal-to-vertical spectral ratio (H/V) of seismic ambient noise to estimate the Vs profiles or VS30. The measurement of H/V is easier, compared to active surface wave methods (MASW) or microtremor array measurements (MAM). The inversion of H/V is non-unique and it is impossible to obtain unique Vs profiles. We apply deep learning to estimate the Vs profile from H/V together with other information including site coordinates, deep bedrock depths, and geomorphological classification. The pairs of H/V spectra (input layer) and Vs profiles (output layer) are used as training data. An input layer consists of an observed H/V spectrum, site coordinates, deep bedrock depths, and geomorphological classification, and an output layer is a velocity profile. We applied the method to the South Kanto Plain, Japan. We measured MASW, MAM and H/V at approximately 2300 sites. The pairs of H/V spectrum together with their coordinates, geomorphological classification etc. and Vs profile obtained from the inversion of dispersion curve and H/V, compose the training data. A trained neural network predicts Vs profiles from the observed H/V spectra with other information. Predicted Vs profiles and their VS30 are reasonably consistent with true Vs profiles and their VS30. The results implied that the deep learning could estimate Vs profile from H/V together with other information.
format Article
id doaj-art-02a6c19ca3cf438d95f370e094555a46
institution OA Journals
issn 2524-1788
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series Soils and Foundations
spelling doaj-art-02a6c19ca3cf438d95f370e094555a462025-08-20T01:56:46ZengElsevierSoils and Foundations2524-17882024-12-0164610152510.1016/j.sandf.2024.101525Estimating S-wave velocity profiles from horizontal-to-vertical spectral ratios based on deep learningKoichi Hayashi0Toru Suzuki1Tomio Inazaki2Chisato Konishi3Haruhiko Suzuki4Hisanori Matsuyama5Kyoto University, Japan; Corresponding author.Mony Exploration Corporation, JapanNon-affiliatedOYO Corporation, JapanOYO Corporation, JapanOYO Corporation, JapanS-wave velocity (Vs) profile or time averaged Vs to 30 m depth (VS30) is indispensable information to estimate the local site amplification of ground motion from earthquakes. We use a horizontal-to-vertical spectral ratio (H/V) of seismic ambient noise to estimate the Vs profiles or VS30. The measurement of H/V is easier, compared to active surface wave methods (MASW) or microtremor array measurements (MAM). The inversion of H/V is non-unique and it is impossible to obtain unique Vs profiles. We apply deep learning to estimate the Vs profile from H/V together with other information including site coordinates, deep bedrock depths, and geomorphological classification. The pairs of H/V spectra (input layer) and Vs profiles (output layer) are used as training data. An input layer consists of an observed H/V spectrum, site coordinates, deep bedrock depths, and geomorphological classification, and an output layer is a velocity profile. We applied the method to the South Kanto Plain, Japan. We measured MASW, MAM and H/V at approximately 2300 sites. The pairs of H/V spectrum together with their coordinates, geomorphological classification etc. and Vs profile obtained from the inversion of dispersion curve and H/V, compose the training data. A trained neural network predicts Vs profiles from the observed H/V spectra with other information. Predicted Vs profiles and their VS30 are reasonably consistent with true Vs profiles and their VS30. The results implied that the deep learning could estimate Vs profile from H/V together with other information.http://www.sciencedirect.com/science/article/pii/S0038080624001033Horizontal-to-vertical spectral ratioS-wave velocityMachine learningInversionMicrotremorSurface wave
spellingShingle Koichi Hayashi
Toru Suzuki
Tomio Inazaki
Chisato Konishi
Haruhiko Suzuki
Hisanori Matsuyama
Estimating S-wave velocity profiles from horizontal-to-vertical spectral ratios based on deep learning
Soils and Foundations
Horizontal-to-vertical spectral ratio
S-wave velocity
Machine learning
Inversion
Microtremor
Surface wave
title Estimating S-wave velocity profiles from horizontal-to-vertical spectral ratios based on deep learning
title_full Estimating S-wave velocity profiles from horizontal-to-vertical spectral ratios based on deep learning
title_fullStr Estimating S-wave velocity profiles from horizontal-to-vertical spectral ratios based on deep learning
title_full_unstemmed Estimating S-wave velocity profiles from horizontal-to-vertical spectral ratios based on deep learning
title_short Estimating S-wave velocity profiles from horizontal-to-vertical spectral ratios based on deep learning
title_sort estimating s wave velocity profiles from horizontal to vertical spectral ratios based on deep learning
topic Horizontal-to-vertical spectral ratio
S-wave velocity
Machine learning
Inversion
Microtremor
Surface wave
url http://www.sciencedirect.com/science/article/pii/S0038080624001033
work_keys_str_mv AT koichihayashi estimatingswavevelocityprofilesfromhorizontaltoverticalspectralratiosbasedondeeplearning
AT torusuzuki estimatingswavevelocityprofilesfromhorizontaltoverticalspectralratiosbasedondeeplearning
AT tomioinazaki estimatingswavevelocityprofilesfromhorizontaltoverticalspectralratiosbasedondeeplearning
AT chisatokonishi estimatingswavevelocityprofilesfromhorizontaltoverticalspectralratiosbasedondeeplearning
AT haruhikosuzuki estimatingswavevelocityprofilesfromhorizontaltoverticalspectralratiosbasedondeeplearning
AT hisanorimatsuyama estimatingswavevelocityprofilesfromhorizontaltoverticalspectralratiosbasedondeeplearning