CaltechAUTHORS
  A Caltech Library Service

Automated Seismic Source Characterisation Using Deep Graph Neural Networks

van den Ende, M. P. A. and Ampuero, J.-P. (2020) Automated Seismic Source Characterisation Using Deep Graph Neural Networks. . (Unpublished) https://resolver.caltech.edu/CaltechAUTHORS:20200630-103659822

[img] PDF - Submitted Version
Creative Commons Attribution.

2114Kb

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20200630-103659822

Abstract

Most seismological analysis methods require knowledge of the geographic location of the stations comprising a seismic network. However, common machine learning tools used in seismology do not account for this spatial information, and so there is an underutilised potential for improving the performance of machine learning models. In this work, we propose a Graph Neural Network (GNN) approach that explicitly incorporates and leverages spatial information for the task of seismic source characterisation (specifically, location and magnitude estimation), based on multi-station waveform recordings. Even using a modestly-sized GNN, we achieve model prediction accuracy that outperforms methods that are agnostic to station locations. Moreover, the proposed method is flexible to the number of seismic stations included in the analysis, and is invariant to the order in which the stations are arranged, which opens up new applications in the automation of seismological tasks and in earthquake early warning systems.


Item Type:Report or Paper (Discussion Paper)
Related URLs:
URLURL TypeDescription
https://doi.org/10.31223/osf.io/nbmztDOIDiscussion Paper
https://doi.org/10.6084/m9.figshare.12231077DOICode
ORCID:
AuthorORCID
van den Ende, M. P. A.0000-0002-0634-7078
Ampuero, J.-P.0000-0002-4827-7987
Additional Information:License: CC-By Attribution 4.0 International. Submitted: May 24, 2020; Last edited: June 29, 2020. We thank the Associate Editor and two anonymous reviewers for their thoughtful comments on the manuscript. MvdE is supported by French government through the UCAJEDI Investments in the Future project managed by the National Research Agency (ANR) with the reference number ANR-15-IDEX-01. The authors acknowledge computational resources provided by the ANR JCJC E-POST project (ANR-14-CE03-0002-01JCJC E-POST). Python codes and the pre-trained model are available from: https://doi.org/10.6084/m9.figshare.12231077. Author asserted no Conflict of Interest.
Group:Seismological Laboratory
Funders:
Funding AgencyGrant Number
Agence Nationale pour la Recherche (ANR)ANR-15-IDEX-01
Agence Nationale pour la Recherche (ANR)ANR-14-CE03-0002-01JCJC E-POST
Subject Keywords:graph neural networks; seismic source
Record Number:CaltechAUTHORS:20200630-103659822
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20200630-103659822
Official Citation:van den Ende, M., & Ampuero, J. (2020, May 25). Automated Seismic Source Characterisation Using Deep Graph Neural Networks. https://doi.org/10.31223/osf.io/nbmzt
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:104161
Collection:CaltechAUTHORS
Deposited By: Tony Diaz
Deposited On:30 Jun 2020 17:45
Last Modified:30 Jun 2020 17:45

Repository Staff Only: item control page