Stuart Russell -- Global seismic monitoring for the Comprehensive Nuclear-Test-Ban Treaty
The interpretation of sensor data from multiple, geographically dispersed sensors
is a ubiquitous challenge in science and engineering. Problems of noise, sensor imperfections, and signal propagation uncertainty, as well as the complexities of data association,
can make the reliable detection of events extremely difficult.
This is particularly true in global monitoring for the
Comprehensive Nuclear-Test-Ban Treaty (CTBT).
In our work, we use real-time data from the UN's International Monitoring System (IMS)
to detect events in the atmosphere, oceans, and underground that might be nuclear explosions.
The IMS is the world's primary global-scale, continuous, real-time system for seismic event
monitoring. Data from over 240 IMS stations (seismic, hydroacoustic, and infrasound) are transmitted via satellite in
real time to the International Data Center (IDC) in Vienna, where
event bulletins are issued daily. Perfect performance
remains well beyond the reach of current technology: the final (SEL3)
bulletin from IDC's automated system, a highly complex and well-tuned
piece of software, misses nearly one third of all seismic events in the
magnitude range of interest, and about half of the reported events are
spurious. A large team of expert analysts post-processes the automatic
bulletins to improve their accuracy to acceptable levels.
Our approach is based on generative Bayesian modelling and inference.
The first-generation model, NET-VISA (NETwork processing via Vertically Integrated Seismic Analysis),
incorporates submodels for event occurrence, signal generation, signal propagation, signal detection, the characteristics of detected signals, and local noise at seismic stations.
A "detected signal" is a blip with an arrival time and amplitude, as estimated by the UN's existing station processing software.
(More than 90% of all such blips are in fact just local station noise.) Given the model and the observed blips, probabilistic inference
produces a hypothesized bulletin of events -- with locations, times, depths, and magnitudes -- that best explains the observations.
NET-VISA inference involves a dynamically constructed graphical model of unbounded size and time-varying structure;
during large events with many aftershocks, the graphical model may contain as many as 500,000 variables.
Compared to the automated SEL3 bulletin, NET-VISA reduces the rate of detection failure by a factor of 2 to 3, maintaining the same false alarm rate.
NET-VISA also detects numerous events that were previously missed by the human analysts.
In November, 2014, the UN announced that NET-VISA would become the new monitoring algorithm for the CTBT.
NETVISA ran continuously at IDC in "development" mode but in frequent use by analysts.
On January 1, 2018, NETVISA became an official part of the verification regime.
The second-generation SIG-VISA model extends the generative model all the way to detailed waveforms, rather than just blips.
Of particular interest is the fact that the SIG-VISA model expresses, via Gaussian processes, the smooth local dependence of waveform structure on the path taken by the signal through the Earth.
This enables SIG-VISA to automatically derive the benefits of seismological techniques such as waveform matching and results in dramatically improved sensitivity.
Recent experiments show a tenfold improvmeent in low-magnitude event detection compared to bulletins produced by expert human analysts.
Publications on NET-VISA
- Stuart Russell, Nimar Arora, Michael Jordan, and Erik Sudderth,
"Vertically Integrated Seismological Analysis I: Modeling."
Eos Transactions of the American Geophysical Union, 90(52), Fall Meeting Supplement, Abstract S33D-08, 2009.
- Nimar Arora, Stuart Russell, and Erik Sudderth,
"Vertically Integrated Seismological Analysis II: Inference."
Eos Transactions of the American Geophysical Union, 90(52), Fall Meeting Supplement, Abstract S31B-1713, 2009.
- Ronan Le Bras, Sheila Vaidya, Jeffrey Schneider, Stuart Russell, and Nimar Arora,
``Status of the Machine Learning Efforts at the International Data Centre of the CTBTO.''
In Proc. Monitoring Research Review (MRR 2010), Orlando, Florida, 2010.
- Nimar S. Arora, Stuart J. Russell, Paul Kidwell, and Erik Sudderth,
``Global seismic monitoring as probabilistic inference.''
Technical Report No. UCB/EECS-2010-108, EECS Department, University of California, Berkeley, 2010.
- Nimar S. Arora, Stuart J. Russell, Paul Kidwell, and Erik Sudderth,
``Global seismic monitoring as probabilistic inference.''
In Advances in Neural Information Processing Systems 23,
MIT Press, 2011.
- Nimar S. Arora, Stuart J. Russell, Paul Kidwell, and Erik Sudderth,
``Global seismic monitoring: A Bayesian approach.''
In Proc. AAAI-11,
San Francisco, 2011.
- Stuart J. Russell, Stephen C. Myers, Nimar S. Arora, David A. Moore, and Erik Sudderth,
``Bayesian Treaty Monitoring: Preliminary Report.''
In Proc. Monitoring Research Review (MRR 2011), Tucson, Arizona, 2011.
- Ronan Le Bras, Stuart Russell, Nimar Arora, and Vera Miljanovic,
``Machine Learning at the CTBTO. Testing and evaluation of the False Events Identification (FEI) and Vertically Integrated Seismic Association (VISA) project.''
In Proc. Monitoring Research Review (MRR 2011), Tucson, Arizona, 2011.
- Nimar Arora, Tony Dear, and Stuart Russell,
``Scalable Probabilistic Inference for Global Seismic Monitoring.''
Eos Transactions of the American Geophysical Union, 92(53), Fall Meeting Supplement, Abstract S43B-2238, 2011.
- Nimar Arora and Stuart Russell,
``A model of seismic coda arrivals to suppress spurious events (abstract).''
In Proc. European Geophysical Union General Assembly, Vienna, 2012.
- Nimar S. Arora, Jeffrey Given, Elena Tomuta, Stuart Russell, and Spilios Spiliopoulos,
``Analyst Evaluation of NET-VISA (Network Processing Vertically Integrated Seismic Analysis) at the CTBTO.''
In Proc. Monitoring Research Review (MRR 2012), Albuquerque, New Mexico, 2012.
- Arora, Nimar, Model-based Bayesian Seismic Monitoring.
PhD thesis, Computer Science Division,
University of California, Berkeley, CA, 2012.
- Nimar S. Arora, Stuart Russell, and Erik Sudderth,
``NET-VISA: Network Processing Vertically Integrated Seismic Analysis.''
In Bulletin of the Seismological Society of America, 103(2A), 709-729, 2013.
Winner of the 2014 Mitchell Prize, jointly awarded by the American Statistical Association and the International Society for Bayesian Analysis to recognize "an outstanding paper that describes how a Bayesian analysis has solved an important applied problem."
- Nimar S. Arora and Stuart Russell,
``Fine-Scale Event Location and Error Analysis in NET-VISA.''
Eos Transactions of the American Geophysical Union, Fall Meeting Supplement, Abstract S31A-2710, 2016.
- R. Le Bras, N. Arora,
N. Kushida, P. Mialle,
I. Bondar, S. Laban, M. Villarroel,
B. Vera, A. Sudakov,
S. Nippress, D. Bowers,
S. Russell, and T. Taylor, ``The Machine-Learning Tool NET-VISA from Cradle to Adulthood - The Next Generation System of the IDC and the SnT Process.''
In CTBT: Science and Technology Conference, Vienna, 2019.
- N. Arora, S. Russell, P. Mialle, R. Le Bras, and P. Nielsen,
``Recent Advances and Status of Generative Modeling for Network Processing at the CTBTO.''
In CTBT: Science and Technology Conference, Vienna, 2019.
- Ronan Le Bras, Nimar Arora, Noriyuki Kushida, Pierrick Mialle, Istvan Bondar, Elena Tomuta, Fekadu Kebede Alamneh, Paulino Feitio, Marcela Villarroel, Beatriz Vera, Alexander Sudakov, Shaban Laban, Stuart Nippress, David Bowers, Stuart Russell, and Tammy Taylor,
``NET-VISA from Cradle to Adulthood. A Machine-Learning Tool for Seismo-Acoustic Automatic Association,''
Pure and Applied Geophysics, DOI 10.1007/s00024-020-02508-x, 2020.
Publications on SIG-VISA
- David A. Moore, Kevin Mayeda, Steve Myers, Min Joon Seo, and Stuart Russell,
``Progress in Signal-Based Bayesian Monitoring.''
In Proc. Monitoring Research Review (MRR 2012), Albuquerque, New Mexico, 2012.
- David Moore and Stuart Russell,
``Fast Gaussian Process Posteriors with Product Trees.''
In Proc. UAI-14,
Quebec City, Canada, 2014.
- David Moore and Stuart Russell, Gaussian Process Random Fields. In Proc. NIPS-15, 2015.
- David A. Moore, Kevin Mayeda, Stephen C. Myers, Stuart J. Russell,
Bayesian Inference for Signal-Based Seismic Monitoring (poster), AGU Fall Meeting, San Francisco, 2015.
- Stuart Russell, Nimar S. Arora, and David Moore,
``Bayesian Monitoring Systems for the CTBT: Historical Development and New Results.''
Eos Transactions of the American Geophysical Union, Fall Meeting Supplement, Abstract S34A-07, 2016.
- David Moore and Stuart Russell,
``Initial Evaluation of Signal-Based Bayesian Monitoring.''
Eos Transactions of the American Geophysical Union, Fall Meeting Supplement, Abstract S31A-2707, 2016.
- David Moore and Stuart Russell,
``Signal-Based Bayesian Seismic Monitoring.''
In Proc. Twentieth International Conference on
Artificial Intelligence and Statistics, Fort Lauderdale, Florida, 2017.
- Moore, David, Signal-Based Bayesian Seismic Monitoring,
PhD thesis, Computer Science Division,
University of California, Berkeley, CA, 2017.