CaltechAUTHORS
  A Caltech Library Service

P-wave arrival picking and first-motion polarity determination with deep learning

Ross, Zachary E. and Meier, Men-Andrin and Hauksson, Egill (2018) P-wave arrival picking and first-motion polarity determination with deep learning. Journal of Geophysical Research. Solid Earth, 123 (6). pp. 5120-5129. ISSN 2169-9313. https://resolver.caltech.edu/CaltechAUTHORS:20180605-092220672

[img] PDF - Published Version
See Usage Policy.

955Kb
[img] PDF - Submitted Version
See Usage Policy.

1198Kb
[img] MS Word (Data S1) - Supplemental Material
See Usage Policy.

732Kb
[img] Plain Text (Data S1) - Supplemental Material
See Usage Policy.

8Mb

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20180605-092220672

Abstract

Determining earthquake hypocenters and focal mechanisms requires precisely measured P wave arrival times and first‐motion polarities. Automated algorithms for estimating these quantities have been less accurate than estimates by human experts, which are problematic for processing large data volumes. Here we train convolutional neural networks to measure both quantities, which learn directly from seismograms without the need for feature extraction. The networks are trained on 18.2 million manually picked seismograms for the Southern California region. Through cross validation on 1.2 million independent seismograms, the differences between the automated and manual picks have a standard deviation of 0.023 s. The polarities determined by the classifier have a precision of 95% when compared with analyst‐determined polarities. We show that the classifier picks more polarities overall than the analysts, without sacrificing quality, resulting in almost double the number of focal mechanisms. The remarkable precision of the trained networks indicates that they can perform as well, or better, than expert seismologists.


Item Type:Article
Related URLs:
URLURL TypeDescription
https://doi.org/10.1029/2017JB015251DOIArticle
https://arxiv.org/abs/1804.08804arXivDiscussion Paper
ORCID:
AuthorORCID
Ross, Zachary E.0000-0002-6343-8400
Meier, Men-Andrin0000-0002-2949-8602
Hauksson, Egill0000-0002-6834-5051
Additional Information:© 2018 American Geophysical Union. Received 17 NOV 2017; Accepted 27 MAY 2018; Accepted article online 4 JUN 2018; Published online 20 JUN 2018. This study was performed using TensorFlow. Data used in this study were collected by the Caltech/USGS Southern California Seismic Network, doi:10.7914/SN/CI, and distributed by the Southern California Earthquake Center, doi:10.7909/C3WD3xH1. The study was supported by the Gordon and Betty Moore Foundation and NSF award EAR‐1550704. There are no real or perceived financial conflicts of interest for any author. Both the trained CNN model, and the entire waveform data set (a single HDF5 file), and the produced focal mechanism catalog are available from the Southern California Earthquake Data Center <http://scedc.caltech.edu/>.
Group:Seismological Laboratory
Funders:
Funding AgencyGrant Number
Gordon and Betty Moore FoundationUNSPECIFIED
NSFEAR-1550704
Subject Keywords:Phase picking; Deep learning; Focal mechanisms; Real‐time seismology; Machine Learning
Issue or Number:6
Record Number:CaltechAUTHORS:20180605-092220672
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20180605-092220672
Official Citation:Ross, Z. E., Meier, M.‐A., & Hauksson, E. (2018). P wave arrival picking and first‐motion polarity determination with deep learning. Journal of Geophysical Research: Solid Earth, 123, 5120–5129. https://doi.org/10.1029/2017JB015251
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:86793
Collection:CaltechAUTHORS
Deposited By: Tony Diaz
Deposited On:05 Jun 2018 17:30
Last Modified:03 Oct 2019 19:48

Repository Staff Only: item control page