A5062339338- Seismology and Earthquake Studies
- earthquake and tectonic studies
- Earthquake Detection and Analysis
- Landslides and related hazards
- Seismic Waves and Analysis
- Anomaly Detection Techniques and Applications
- Cryospheric studies and observations
- Hydrocarbon exploration and reservoir analysis
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Winter Sports Injuries and Performance
- Methane Hydrates and Related Phenomena
- Fault Detection and Control Systems
- Atmospheric and Environmental Gas Dynamics
- Imbalanced Data Classification Techniques
- Advancements in Semiconductor Devices and Circuit Design
- Geochemistry and Geologic Mapping
- Structural Health Monitoring Techniques
- Geotechnical and Geomechanical Engineering
- Mineral Processing and Grinding
- Time Series Analysis and Forecasting
- Asian Industrial and Economic Development
- Magnetic and Electromagnetic Effects
- Sports Dynamics and Biomechanics
- Hydraulic and Pneumatic Systems
- Software Engineering Research
Los Alamos National Laboratory
2017-2024
Laboratoire de Géologie de l’École Normale Supérieure
2020-2023
Université Paris Sciences et Lettres
2020-2022
Centre National de la Recherche Scientifique
2020-2022
École Normale Supérieure
2020-2022
École Normale Supérieure - PSL
2020-2021
Laboratoire de Physique de l'ENS
2021
University of Cambridge
2017
Forecasting fault failure is a fundamental but elusive goal in earthquake science. Here we show that by listening to the acoustic signal emitted laboratory fault, machine learning can predict time remaining before it fails with great accuracy. These predictions are based solely on instantaneous physical characteristics of acoustical signal, and do not make use its history. Surprisingly, identifies from zone previously thought be low-amplitude noise enables forecasting throughout quake cycle....
Earthquake prediction, the long-sought holy grail of earthquake science, continues to confound Earth scientists. Could we make advances by crowdsourcing, drawing from vast knowledge and creativity machine learning (ML) community? We used Google’s ML competition platform, Kaggle, engage worldwide community with a develop improve data analysis approaches on forecasting problem that uses laboratory data. The competitors were tasked predicting time remaining before next successive quake events,...
Systematic characterization of slip behaviours on active faults is key to unraveling the physics tectonic faulting and interplay between slow fast earthquakes. Interferometric Synthetic Aperture Radar (InSAR), by enabling measurement ground deformation at a global scale every few days, may hold those interactions. However, atmospheric propagation delays often exceed interest despite state-of-the art processing, thus InSAR analysis requires expert interpretation priori knowledge fault...
Curbing methane emissions is among the most effective actions that can be taken to slow down global warming. However, monitoring remains challenging, as detection methods have a limited quantification completeness due trade-offs made between coverage, resolution, and accuracy. Here we show deep learning overcome trade-off in terms of spectral resolution comes with multi-spectral satellite data, resulting tool coverage high temporal spatial resolution. We compare our detections airborne...
Research Article| March 06, 2019 Characterizing Acoustic Signals and Searching for Precursors during the Laboratory Seismic Cycle Using Unsupervised Machine Learning David C. Bolton; Bolton aDepartment of Geosciences, Pennsylvania State University, 201 Old Main, University Park, 16802 U.S.A., dcb31@psu.educhrisjmarone@gmail.com Search other works by this author on: GSW Google Scholar Parisa Shokouhi; Shokouhi bDepartment Engineering Science Mechanics, pxs990@psu.edu, jvr5626@psu.edu Bertrand...
Slow earthquakes may trigger failure on neighboring locked faults that are stressed enough to break, and slow slip patterns evolve before a nearby great earthquake. However, even in the clearest cases such as Cascadia, associated tremor have only been observed intermittent discrete bursts. By training convolutional neural network detect known single seismic station we isolate identify preceding following larger events. The deep can be used for detection of quasi-continuous tremor, providing...
Nearly all aspects of earthquake rupture are controlled by the friction along fault that progressively increases with tectonic forcing, but in general cannot be directly measured. We show can determined at any time, from continuous seismic signal. In a classic laboratory experiment repeating earthquakes, we find signal follows specific pattern respect to friction, allowing us determine fault's position within its failure cycle. Using machine learning, instantaneous statistical...
Abstract Slow slip events result from the spontaneous weakening of subduction megathrust and bear strong resemblance to earthquakes, only slower. This allows us study fundamental aspects nucleation that remain elusive for classic, fast earthquakes. We rely on machine learning algorithms infer slow timing statistics seismic waveforms. find patterns in power follow 14-month cycle Cascadia, arguing favor predictability rupture. Here, we show exponentially increases as slowly slipping portion...
Seismogenic plate boundaries are presumed to behave in a similar manner densely packed granular medium, where fault and blocks systems rapidly rearrange the distribution of forces within themselves, as particles do slowly sheared systems. We use machine learning show that statistical features velocity signals from individual simulated contain information regarding instantaneous global state intermittent frictional stick-slip dynamics. demonstrate combining built more can improve accuracy...
A fundamental problem of optoelectronic simulations is to achieve convergence. We use statistical analysis and machine learning effectively guide the selection next device be examined based upon expected convergence simulation. This active strategy rapidly constructs a model that predicts Poisson-Schrödinger devices simultaneously produces fully converged simulations.
Abstract During the RESOLVE project (“High‐resolution imaging in subsurface geophysics: development of a multi‐instrument platform for interdisciplinary research”), continuous surface displacement and seismic array observations were obtained on Glacier d’Argentière French Alps 35 days May 2018. The data set is used to perform detailed study targeted processes within highly dynamic cryospheric environment. In particular, physical controlling glacial basal motion are poorly understood remain...
Active faults release tectonic stress imposed by plate motion through a spectrum of slip modes, from slow, aseismic slip, to dynamic, seismic events. Slow earthquakes are often associated with tremor, nonimpulsive signals that can easily be buried in noise and go undetected. We present new methodology aimed at improving the detection location tremors hidden within noise. After identifying classic convolutional neural network (CNN), we rely on attribution extract core tremor signatures....
Tectonic faults slip in various manners, ranging from ordinary earthquakes to slow events aseismic fault creep. The frequent occurrence of and their sensitivity stress make them a promising probe the neighboring locked zone where megaquakes take place. This relationship, however, remains poorly understood. We show that Cascadia megathrust is continuously broadcasting tremor-like signal precisely informs displacement rate throughout cycle. posit this provides indirect, real-time access...
Methane is one of the most potent greenhouse gases, and its short atmospheric half-life makes it a prime target to rapidly curb global warming. However, current methane emission monitoring techniques primarily rely on approximate factors or self-reporting, which have been shown often dramatically underestimate emissions. Although initially designed monitor surface properties, satellite multispectral data has recently emerged as powerful method analyze content. spectral resolution instruments...
Over the last two decades, strain and GPS measurements have shown that slow slip on earthquake faults is a widespread phenomenon. Slow also inferred from correlated small amplitude seismic signals known as nonvolcanic tremor low frequency earthquakes (LFEs). has been reproduced in laboratory simulation studies, however fundamental physics of these phenomena their relationship to dynamic rupture remains poorly understood. Here we show that, setting, continuous waves are imprinted with...
The seismogenic plate boundaries are presumed to behave similarly a densely packed granular medium, where fault and blocks systems rapidly rearrange the distribution of forces within themselves, as particles do in slowly sheared systems. We use machine learning show that statistical features velocity signals from individual simulated contain information regarding instantaneous global state intermittent frictional stick-slip dynamics. demonstrate combining built more can improve accuracy...
Low frequency earthquakes (LFEs) originating below the central San Andreas Fault are associated with slow-slip within more ductile portion of crust beneath seismogenic zone. Monitoring efforts over 15 years recorded >1 million LFEs >70 per day. We apply machine learning (ML) to statistical features describing seismic waveforms and estimate LFE daily intensity. Using 4 independent data, ML model produces a 0.68 correlation. The burst-like behavior is reproduced largest misfit occurs during...