[Withdrawn] Detecting Low Magnitude Seismic Events Using Convolutional Neural Networks

Date: 4/24/2019

Time: 06:00 PM

Room: Grand Ballroom

Currently, many traditional methods are used to detect arrivals in three-component seismic waveform data collected at various distances. Accurately establishing the identity and arrival of these waves in adverse signal-to-noise environments is vital in detecting and locating seismic events. Autocorrelation and template matching techniques are just a few of the various methods that may be used, yet, each have their own limitations.

In this work, we present updated results using convolutional neural networks (CNNs). CNNs have been shown to significantly improve performance at local distances under certain conditions such as induced seismicity. In this work we expand the use of CNNs to more remote distances and lower magnitudes. We explore the advantages and limits of a certain architecture of CNN and update results previously presented.

We describe in detail performance results of our method tuned on a new dataset with expert defined arrival picks. The dataset used is from the Dynamic Network Experiment 2018 (DNE18) and comes from sensors in Utah. We demonstrate the ability to train the CNN on events from the dataset and achieve significantly higher test set performance than standard methods. Furthermore, we validate performance on streaming data, including very low magnitude expert picked arrivals.

Sandia National Laboratories is a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Presenting Author: Robert Forrest

Authors