Welcome to the new version of CaltechAUTHORS. Login is currently restricted to library staff. If you notice any issues, please email coda@library.caltech.edu
Published May 28, 2018 | Published + Supplemental Material
Journal Article Open

Machine Learning Seismic Wave Discrimination: Application to Earthquake Early Warning

Abstract

Performance of earthquake early warning systems suffers from false alerts caused by local impulsive noise from natural or anthropogenic sources. To mitigate this problem, we train a generative adversarial network (GAN) to learn the characteristics of first‐arrival earthquake P waves, using 300,000 waveforms recorded in southern California and Japan. We apply the GAN critic as an automatic feature extractor and train a Random Forest classifier with about 700,000 earthquake and noise waveforms. We show that the discriminator can recognize 99.2% of the earthquake P waves and 98.4% of the noise signals. This state‐of‐the‐art performance is expected to reduce significantly the number of false triggers from local impulsive noise. Our study demonstrates that GANs can discover a compact and effective representation of seismic waves, which has the potential for wide applications in seismology.

Additional Information

© 2017 American Geophysical Union. Received 9 MAR 2018; Accepted 6 MAY 2018; Accepted article online 11 MAY 2018; Published online 29 MAY 2018. We thank Qingkai Kong and another anonymous reviewer for constructive comments. This research was supported by a Gordon and Betty Moore Foundation grant to Caltech and by the Swiss National Science Foundation. The Japanese waveform data can be downloaded from http://www.kik.bosai.go.jp/ (last accessed October 2017). For southern California we have used waveforms and parametric data from the Caltech/USGS Southern California Seismic Network (doi:10.7914/SN/CI) stored at the Southern California Earthquake Data Center (doi:10.7909/C3WD3xH1). The algorithms were written with Python packages Keras (https://keras.io/) and Scikit‐learn (http://scikit‐learn.org/). The GAN was trained for 2.0 hr, and the Random Forest was trained for 20 min on a PC (NVIDIA GeForce GTX 1050 Ti 4 GB, Intel Core i5‐7300HQ 2.50GHz).

Attached Files

Published - Li_et_al-2018-Geophysical_Research_Letters.pdf

Supplemental Material - grl57440-sup-0001-2018gl077870-si.docx

Files

Li_et_al-2018-Geophysical_Research_Letters.pdf
Files (3.4 MB)
Name Size Download all
md5:58bc6dcb023d45fefb68b574f38901e6
1.4 MB Preview Download
md5:23ed2f09d765aa40c29f5e524f39d9ca
2.1 MB Download

Additional details

Created:
August 21, 2023
Modified:
October 18, 2023