Machine-learning-facilitated earthquake and anthropogenic source detections near the Weiyuan Shale Gas Blocks, Sichuan, China

Zhou, Pengcheng; Ellsworth, William L.; Yang, Hongfeng; Tan, Yen Joe; Beroza, Gregory C.; Sheng, M.; Chu, Risheng

doi:10.26464/epp2021053

Cited by 20 publications

(8 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In some studies of exploration geophysics, it has produced results comparable to and in some cases surpassing expert human performance (Bi et al., 2021; Liang et al., 2019; Reading et al., 2015). In seismology, a successful example is the use of deep learning for earthquake detection and phase picking (Jiang et al., 2021; Z. Li, 2021; Mousavi et al., 2020; J. Wang et al., 2019; Wong et al., 2021; Yu & Ma, 2021; L. Zhang et al., 2020; P. C. Zhou et al., 2021; Y. Zhou et al., 2021). However, applications of deep learning methods in seismic structure inversions have been limited thus far, including super resolution images from low resolution by a CycleGAN and crustal thickness estimated from Rayleigh surface wave based on Deep Neural Networks (Cheng et al., 2019; Niu et al., 2020).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning‐BasedH‐κMethod (HkNet) for Estimating Crustal Thickness andVp/VsRatio From Receiver Functions

Wang

Song

2022

JGR Solid Earth

View full text Add to dashboard Cite

A receiver function (RF) is the response of the Earth's structure below a seismometer to an incident teleseismic wave and consists of a series of P-to-S (Ps) or S-to-P (Sp) converted waves generated at structural interfaces (Langston, 1979), mainly those generated by the velocity discontinuities in the crust and upper mantle below the seismometer. Crustal thickness (H) and P-wave and S-wave velocity ratio (κ) are important parameters reflecting the crustal structure and internal material composition and can provide an important basis for regional tectonics and dynamics. Zhu and Kanamori (2000) noted that when the Moho converted phase Ps and crustal multiple waves are considered at the same time, H and κ can be estimated under the assumption of the average crustal P wave velocity. This is currently the most commonly used H-κ method. The H-κ algorithm stacks the amplitudes of RFs at predicted arrival times for Ps and crustal multiples (PpPs and PpSs + PsPs; here referred to as M1 and M2, respectively, for convenience, as in J. Li et al., 2019) for different values of H and κ in a grid search.The assumptions of the H-κ method are very simple, which leads to its limitations. It assumes that the crustal medium is isotropic, the Moho surface is horizontal, and the P-wave velocity of the crust is fixed. However, the above assumptions cannot be satisfied under complex geological settings, so the results of the H-κ method may be inaccurate. The Chinese continent is known to have highly heterogeneous crustal structure and extremely variable Moho depth, as revealed for example, by seismic tomography (

show abstract

Section: Introductionmentioning

confidence: 99%

“…Li, 2021;Mousavi et al, 2020;Wong et al, 2021;Yu & Ma, 2021;L. Zhang et al, 2020;P. C. Zhou et al, 2021;.…”

unclassified

Deep Learning‐BasedH‐κMethod (HkNet) for Estimating Crustal Thickness andVp/VsRatio From Receiver Functions

Wang

Song

2022

JGR Solid Earth

View full text Add to dashboard Cite

show abstract

“…Deep learning seismology has dramatically improved earthquake monitoring performance particularly in earthquake detection and phase picking (Perol et al, 2018;Ross et al, 2018;Mousavi et al, 2020). Earthquake monitoring workflows using deep-learning-based phase pickers have been applied to studying, dense earthquake sequences (Liu et al, 2020;Ross et al, 2020;Tan et al, 2021), induced seismicity (Park et al, 2020;Wang et al, 2020;Chai et al, 2020;Park et al, in press;Zhou et al, 2021), marine seismicity (Gong et al, 2022;Jiang et al, 2022), and magmatic systems (Retailleau, Saurel, Laporte, et al, 2022). These studies have demonstrated that deep learning can: detect up to orders of magnitude more small earthquakes than conventional algorithms, provide a more complete accounting of seismicity, and enable new insight into earthquake behavior.…”

Section: Introductionmentioning

confidence: 99%

QuakeFlow: A Scalable Machine-learning-based Earthquake Monitoring Workflow with Cloud Computing

Zhu¹,

Brian²,

Yang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Earthquake monitoring workflows are designed to detect earthquake signals and to determine source characteristics from continuous waveform data. Recent developments in deep learning seismology have been used to improve tasks within earthquake monitoring workflows that allow the fast and accurate detection of up to orders of magnitude more small events than are present in conventional catalogs.To facilitate the application of machine-learning algorithms to large-volume seismic records at scale, we developed a cloud-based earthquake monitoring workflow, QuakeFlow, that applies multiple processing steps to generate earthquake catalogs from raw seismic data. QuakeFlow uses a deep learning model, PhaseNet, for picking P/S phases and a machine learning model, GaMMA, for phase association with approximate earthquake location and magnitude. Each component in QuakeFlow is containerized, allowing straightforward updates to the pipeline with new deep learning/machine learning models, as well as the ability to add new components, such as earthquake relocation algorithms. We built QuakeFlow in Kubernetes to make it auto-scale for large datasets and to make it easy to deploy on cloud platforms, which enables large-scale parallel processing. We used QuakeFlow to process three years of continuous archived data from Puerto Rico within a few hours, and found more than a factor of ten more events that occurred on much the same structures as previously known seismicity. We applied Quakeflow to monitoring frequent earthquakes in Hawaii and found over an order of magnitude more events than are in the standard catalog, including many events that illuminate the deep structure of the magmatic system. We

show abstract

“…There are several widely used approaches to address this issue. In one approach, small earthquakes that occur proximal to a larger target event are used as empirical Green's functions (EGFs) to remove site and path effects and hence isolate the target event source parameters (Baltay et al, 2011;Mueller, 1985;Frankel & Wennerberg, 1989;Lanza et al, 1999;Hough, 1996Hough, , 1997Mori & Frankel, 1990;Huang et al, 2016;Ruhl et al, 2017;Huang et al, 2017;S. Chu et al, 2019;Folesky et al, 2021).…”

Section: Introductionmentioning

confidence: 99%

Resolving Differences in the Rupture Properties of Prominent Earthquakes in Southern California with Bayesian Source Spectral Analysis

Trugman

2022

Preprint

View full text Add to dashboard Cite

The spectra of earthquake waveforms can provide important insight into rupture processes, but the analysis and interpretation of these spectra is rarely straightforward. Here we develop a Bayesian framework that embraces the inherent data and modeling uncertainties of spectral analysis to infer key source properties. The method uses a spectral ratio approach to correct the observed waveform spectra of nearby earthquakes for path and site attenuation. The objective then is to solve for a joint posterior probability distribution of three source parameters -seismic moment, corner frequency, and high-frequency falloff rate -for each earthquake in the sequence, as well as a measure of rupture directivity for target events with good azimuthal station coverage. While computationally intensive, this technique provides a quantitative understanding of parameter tradeoffs and uncertainties and allows one to impose physical constraints through prior distributions on all source parameters, which guide the inversion when data is limited. We demonstrate the method by analyzing in detail the source properties of 14 different target events of magnitude M5 in southern California that span a wide range of tectonic regimes and fault systems. These prominent earthquakes, while comparable in size, exhibit marked diversity in their source properties and directivity, with clear spatial patterns, depth-dependent trends, and a preference for unilateral directivity. These coherent spatial variations source properties suggest that regional differences in tectonic setting, hypocentral depth or fault zone characteristics may drive variability in rupture processes, with important implications for our understanding of earthquake physics and its relation to hazard.

show abstract

Machine-learning-facilitated earthquake and anthropogenic source detections near the Weiyuan Shale Gas Blocks, Sichuan, China

Cited by 20 publications

References 60 publications

Deep Learning‐BasedH‐κMethod (HkNet) for Estimating Crustal Thickness andVp/VsRatio From Receiver Functions

Deep Learning‐BasedH‐κMethod (HkNet) for Estimating Crustal Thickness andVp/VsRatio From Receiver Functions

QuakeFlow: A Scalable Machine-learning-based Earthquake Monitoring Workflow with Cloud Computing

Resolving Differences in the Rupture Properties of Prominent Earthquakes in Southern California with Bayesian Source Spectral Analysis

Contact Info

Product

Resources

About