Machine Learning Seismic Wave Discrimination: Application to Earthquake Early Warning

  title={Machine Learning Seismic Wave Discrimination: Application to Earthquake Early Warning},
  author={Zefeng Li and Men‐Andrin Meier and Egill Hauksson and Zhongwen Zhan and Jennifer Andrews},
  journal={Geophysical Research Letters},
  pages={4773 - 4779}
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… 

Reliable Real‐Time Seismic Signal/Noise Discrimination With Machine Learning

This study develops and compares a series of nonlinear classifiers with variable architecture depths, including fully connected, convolutional and recurrent neural networks, and a model that combines a generative adversarial network with a random forest to suggest that machine learning classifiers can strongly improve the reliability and speed of EEW alerts.

Seismic Event Identification Based on a Generative Adversarial Network and Support Vector Machine

Identifying appropriate seismic events is the primary precondition for conducting meaningful analysis in seismological research. The successful creation of a method to automatically identify

Earthquake Magnitude Estimation Based on Machine Learning: Application to Earthquake Early Warning System

Results from the statistical analysis show that the waveform can be modelled by deep neural network (DNN) models, and recommendation related to estimate the magnitude from seismic raw modelling are better using Classification DNN with larger dataset.

Earthquake detection and location for Earthquake Early Warning Using Deep Learning

This result is a preliminary study of deep learning, which is targeted at the classification of earthquakes p wave and noise signals and its association to estimate early earthquake location using 3 component record channels.

CRED: A Deep Residual Network of Convolutional and Recurrent Units for Earthquake Signal Detection

The Cnn-Rnn Earthquake Detector (CRED) is introduced, a detector based on deep neural networks that holds great promise for lowering the detection threshold while minimizing false positive detection rates.

CREIME—A Convolutional Recurrent Model for Earthquake Identification and Magnitude Estimation

A multi-tasking deep learning model – the C onvolutional R ecurrent model for E arthquake I dentification and M agnitude E stimation (CREIME) that detects the first earthquake signal, from background seismic noise, determines the P-arrival time as well as estimates the magnitude using the raw of P-wave data compared to the previous studies.

Artificial Neural Networks as Emerging Tools for Earthquake Detection

This paper surveys the latest ANN applications to the automatic interpretation of seismic data, with a special focus on earthquake detection, and the estimation of on set times, and gives an insight into the labor of human interpreters, who may face uncertainties in the case of small magnitude earthquakes.

Imbalanced Seismic Event Discrimination Using Supervised Machine Learning

The discrimination between earthquakes and artificial explosions is a significant issue in seismic analysis to efficiently prevent and respond to seismic events. However, the discrimination of

A Novel Approach for Earthquake Early Warning System Design using Deep Learning Techniques

This work focuses on detecting earthquakes by converting seismograph recorded data into corresponding audio signals for better perception and then uses popular Speech Recognition techniques of Filter bank coefficients and Mel Frequency Cepstral Coefficients to extract the features.

Real‐Time Earthquake Early Warning With Deep Learning: Application to the 2016 M 6.0 Central Apennines, Italy Earthquake

A novel deep learning earthquake early warning system that utilizes fully convolutional networks to simultaneously detect earthquakes and estimate their source parameters from continuous seismic waveform streams and evolutionarily improves the solutions by receiving continuous data is developed.



Convolutional neural network for earthquake detection and location

This work leverages the recent advances in artificial intelligence and presents ConvNetQuake, a highly scalable convolutional neural network for earthquake detection and location from a single waveform, and applies it to study the induced seismicity in Oklahoma, USA.

Machine Learning Predicts Laboratory Earthquakes

We apply machine learning to data sets from shear laboratory experiments, with the goal of identifying hidden signals that precede earthquakes. Here we show that by listening to the acoustic signal

The Gutenberg Algorithm: Evolutionary Bayesian Magnitude Estimates for Earthquake Early Warning with a Filter Bank

It is proposed that uncertainties of the earliest warning messages can be reduced substantially if the broadband frequency information of seismic signals is fully exploited and a novel probabilistic algorithm for estimating EEW magnitudes is presented.

Toward earthquake early warning in northern California

[1] Earthquake early warning systems are an approach to earthquake hazard mitigation which takes advantage of the rapid availability of earthquake information to quantify the hazard associated with

A New Trigger Criterion for Improved Real-Time Performance of Onsite Earthquake Early Warning in Southern California

We have implemented and tested an algorithm for onsite earthquake early warning (EEW) in California using the infrastructure of the Southern California Seismic Network (SCSN). The algorithm relies on

Bulletin of the Seismological Society of America

Response spectra are of fundamental importance in earthquake engineering and represent a standard measure in seismic design for the assessment of structural performance. However, unlike Fourier

MyShake: A smartphone seismic network for earthquake early warning and beyond

It is shown that smartphones can record magnitude 5 earthquakes at distances of 10 km or less and develop an on-phone detection capability to separate earthquakes from other everyday shakes, which could be used to deliver rapid microseism maps, study impacts on buildings, and possibly image shallow earth structure and earthquake rupture kinematics.

CISN ShakeAlert: An Earthquake Early Warning Demonstration System for California

The CISN ShakeAlert demonstration system combines estimates and uncertainties determined by three algorithms implemented in parallel to calculate and report at a given time the most probable earthquake magnitude and location, as well as the likelihood of correct alarm.

Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks

This work introduces a class of CNNs called deep convolutional generative adversarial networks (DCGANs), that have certain architectural constraints, and demonstrates that they are a strong candidate for unsupervised learning.

Generative Adversarial Nets

We propose a new framework for estimating generative models via an adversarial process, in which we simultaneously train two models: a generative model G that captures the data distribution, and a