Convolutional neural network for earthquake detection and location

ConvNetQuake is the first neural network for detection and location of earthquakes from seismograms. The recent evolution of induced seismicity in Central United States calls for exhaustive catalogs to improve seismic hazard assessment. Over the last decades, the volume of seismic data has increased exponentially, creating a need for efficient algorithms to reliably detect and locate earthquakes. Today’s most elaborate methods scan through the plethora of continuous seismic records, searching for repeating seismic signals. We leverage the recent advances in artificial intelligence and present ConvNetQuake, a highly scalable convolutional neural network for earthquake detection and location from a single waveform. We apply our technique to study the induced seismicity in Oklahoma, USA. We detect more than 17 times more earthquakes than previously cataloged by the Oklahoma Geological Survey. Our algorithm is orders of magnitude faster than established methods.

[1]  Xiaodong Cui,et al.  Data Augmentation for Deep Neural Network Acoustic Modeling , 2015, IEEE/ACM Transactions on Audio, Speech, and Language Processing.

[2]  A. Ng.  Feature selection, L1 vs. L2 regularization, and rotational invariance , 2004, Twenty-first international conference on Machine learning - ICML '04.

[3]  David B. Harris,et al.  An Autonomous System for Grouping Events in a Developing Aftershock Sequence , 2011 .

[4]  Kristy F. Tiampo,et al.  Surface uplift and time-dependent seismic hazard due to fluid injection in eastern Texas , 2015, Science.

[5]  Gregory C. Beroza,et al.  An Empirical Approach to Subspace Detection , 2014 .

[6]  Heiga Zen,et al.  WaveNet: A Generative Model for Raw Audio , 2016, SSW.

[7]  R. V. Allen,et al.  Automatic phase pickers: Their present use and future prospects , 1982 .

[8]  Jin Wang,et al.  Identification and picking of S phase using an artificial neural network , 1997, Bulletin of the Seismological Society of America.

[9]  Justin Salamon,et al.  Deep Convolutional Neural Networks and Data Augmentation for Environmental Sound Classification , 2016, IEEE Signal Processing Letters.

[10]  D. Harris.  Subspace Detectors: Theory , 2006 .

[11]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[12]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[13]  F. Ringdal,et al.  The detection of low magnitude seismic events using array-based waveform correlation , 2006 .

[14]  David R. Shelly,et al.  A 15 year catalog of more than 1 million low‐frequency earthquakes: Tracking tremor and slip along the deep San Andreas Fault , 2017 .

[15]  Austin A. Holland,et al.  Earthquakes Triggered by Hydraulic Fracturing in South‐Central Oklahoma , 2013 .

[16]  Robert J. Skoumal,et al.  Optimizing multi-station earthquake template matching through re-examination of the Youngstown, Ohio, sequence , 2014 .

[17]  W. Ellsworth.  Injection-Induced Earthquakes , 2013, Science.

[18]  Navdeep Jaitly,et al.  Vocal Tract Length Perturbation (VTLP) improves speech recognition , 2013 .

[19]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[20]  Qingkai Kong,et al.  MyShake: A smartphone seismic network for earthquake early warning and beyond , 2016, Science Advances.

[21]  Richard C. Aster,et al.  Hundreds of Earthquakes per Day: The 2014 Guthrie, Oklahoma, Earthquake Sequence , 2015 .

[22]  G. Abers,et al.  Potentially induced earthquakes in Oklahoma, USA: Links between wastewater injection and the 2011 Mw 5.7 earthquake sequence , 2013 .

[23]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[24]  Mark D. Zoback,et al.  Oklahoma’s recent earthquakes and saltwater disposal , 2015, Science Advances.

[25]  Andrew P. Valentine,et al.  Data space reduction, quality assessment and searching of seismograms: autoencoder networks for waveform data , 2012 .

[26]  Christopher John Young,et al.  A comparison of select trigger algorithms for automated global seismic phase and event detection , 1998, Bulletin of the Seismological Society of America.

[27]  Jocelyn Sietsma,et al.  Creating artificial neural networks that generalize , 1991, Neural Networks.

[28]  Clara E Yoon,et al.  Earthquake detection through computationally efficient similarity search , 2015, Science Advances.

[29]  Ta-Liang Teng,et al.  Artificial neural network-based seismic detector , 1995, Bulletin of the Seismological Society of America.

[30]  R. Herrmann,et al.  Earthquake hypocenters and focal mechanisms in central Oklahoma reveal a complex system of reactivated subsurface strike‐slip faulting , 2015 .

[31]  Geoffrey Zweig,et al.  The microsoft 2016 conversational speech recognition system , 2016, 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[32]  S. Ge,et al.  High-rate injection is associated with the increase in U.S. mid-continent seismicity , 2015, Science.

[33]  J. MacQueen.  Some methods for classification and analysis of multivariate observations , 1967 .