LSTM-Autoencoder Network for the Detection of Seismic Electric Signals

Xue, Jiyan; Huang, Qinghua; Wu, Sihong; Nagao, Toshiyasu

doi:10.1109/tgrs.2022.3183389

Cited by 12 publications

(6 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We tested the performances based on different models, including the proposed MagInfoNet, MagNet, RF, and the MINLSTM where LSTM layers (Xue et al., 2022) are used instead of UniMP layers to compare the performance between the GNN and BRNN. Instead of only selecting samples at epicentral distances of less than 1° (Mousavi & Beroza, 2020), we did not make any redundant selective reservation and other filter operations so that the experimental results can reveal the generalization ability and broader scope of application of MagInfoNet.…”

Section: Resultsmentioning

confidence: 99%

MagInfoNet: Magnitude Estimation Using Seismic Information Augmentation and Graph Transformer

Chen

Wang

et al. 2022

Earth and Space Science

View full text Add to dashboard Cite

In this study, we propose a reliable data‐driven tool, MagInfoNet, to enhance the accuracy of magnitude estimation. Its architecture was assembled using the Pre‐Inform and Mag‐Pred modules to replace and update the key functions of traditional seismic analysis workflows. The Pre‐Inform module with the residual network was used for data pretreatment by combining the intrinsic characteristics of seismic signals with the potential features of the arrival and travel times. Meanwhile, using a graph transformer with an improved cyclic graph, the Mag‐Pred module was used to calculate magnitudes by the preprocessed information and the autocorrelation of seismic time series. Training and testing data were randomly selected from the Stanford Earthquake Data Set. The results show that the estimation accuracy, generalization, and robustness of the proposed MagInfoNet are better than those of three machine learning models. Besides, MagInfoNet can perform better for those samples with larger epicentral distances, enhancing the monitoring capacity of existing system for earthquake events in remote areas. Finally, we discuss the interpretability of the explainable MagInfoNet to verify the role of advanced neural network modules.

show abstract

Section: Resultsmentioning

confidence: 99%

MagInfoNet: Magnitude Estimation Using Seismic Information Augmentation and Graph Transformer

Chen

Wang

et al. 2022

Earth and Space Science

View full text Add to dashboard Cite

show abstract

“…These findings have positioned the geoelectric method as one of the most promising geophysical approaches for achieving breakthroughs in short-term earthquake prediction (Zhao et al 2022). Some researchers have tried to conduct earthquake forecasting by analyzing pre-earthquake geoelectric signals (Varotsos & Lazaridou 1991;Uyeda et al 2000Uyeda et al , 2009Du et al 2002;An et al 2013;Xue et al 2022).…”

Section: Introductionmentioning

confidence: 99%

A Numerical Study of Electrokinetically Generated Pre-Earthquake Geoelectric Fields in Layered Porous Media

Zheng,

Ren,

Ren

et al. 2024

Preprint

View full text Add to dashboard Cite

The investigation into the physical mechanisms and behavior of geoelectric fields as potential precursors of earthquakes is a crucial yet challenging task in seismo-electromagnetics. Assuming the electrokinetic effect as a mechanism, we propose a novel numerical modeling algorithm based on the framework of the LAC GRTM to explore the behavior of geoelectric fields in stratified porous media during earthquake preparation. This algorithm incorporates two innovative aspects: (1) employing the Jordan decomposition to address the degeneracy problem encountered in quasi-stationary scenarios, and (2) utilizing digital linear filters for Hankel transforms to conduct the wavenumber integration. The accuracy of the algorithm in computing static displacements in an extremely low porosity half-space model is verified through comparison with analytical solutions. Numerical results from a half-space model demonstrate a strong consistency between the vertical electric field and filtration displacement. Notably, the maximum amplitude of the vertical electric field can reach approximately 2878.8 mV/km, which is detectable by receivers located at considerable distances from the epicenter. In addition, three distinct types of pre-earthquake electric fields are identified: (1) the localized electric field induced by slow P waves, (2) the interface electric responses, and (3) the direct converted electric field from the source. Results from a multi-layer porous model show the significant influence of the water table on the amplitude of the vertical electric field. This can be attributed to the influence of water saturation and permeability on the coefficient that combines the vertical electric field and the vertical filtration displacement.We also analyze the temporal evolution of different fields during earthquake preparation, which demonstrates a robust correlation between the temporal characteristics of pore pressure and that of the vertical electric field. This indicates that the vertical electric field offers an effective means to detect pore pressure evolution during earthquake preparation. Its sensitivity to the strike, rake, and dip angles of potential faults indicates its potential for determining fault geometry related to an impending earthquake. Furthermore, fluctuations in displacements enable a rough estimation of variations in stress and confining pressure. These observations underscore the advantage of integrating geoelectric precursor data with continuous displacement observations to enhance our understanding of earthquake preparation mechanisms.

show abstract

“…Machine learning (ML) applied to geophysical problems is a cutting-edge direction [33]. In recent years, ML has been utilized in many geophysics application areas, including pattern recognition in seismic attributes [34], noise removal [35,36], and inversion tasks [8,[37][38][39][40]. As a representative of the ML, the deep learning (DL) network comes with strong learning and generalization capabilities, enabling excellent performance in electrical resistivity inversion problems [41][42][43].…”

Section: Introductionmentioning

confidence: 99%

Fast Initial Model Design for Electrical Resistivity Inversion by Using Broad Learning Framework

Tao,

Han,

Yang

et al. 2024

Minerals

View full text Add to dashboard Cite

The electrical resistivity method is widely used in near-surface mineral exploration. At present, the deterministic algorithm is commonly employed in three-dimensional (3-D) electrical resistivity inversion to obtain subsurface electrical structures. However, the accuracy and efficiency of deterministic inversion rely on the initial model. In practice, obtaining an initial model that approximates the true subsurface electrical structures remains challenging. To address this issue, we introduce a broad learning (BL) network to determine the initial model and utilize the limited memory quasi-Newton (L-BFGS) algorithm to conduct the 3-D electrical resistivity inversion task. The powerful mapping capability of the BL network enables one to find the model that elucidates the actual observed data. The single-layer BL network makes it efficient and easy to realize, leading to much faster network training compared to that using the deep learning network. Both the synthetic and field experiments suggest that the BL framework could effectively obtain the initial model based on observed data. Furthermore, in comparison to using a homogeneous medium as the initial model, the L-BFGS inversion with the BL framework-designed initial model improves the inversion accuracy of subsurface electrical structures and expedites the convergence speed of the iteration. This study provides an effective approach for fast initial model design in a data-driven manner when the prior information is unavailable. The proposed method can be useful in high-precision imaging of near-surface mineral electrical structures.

show abstract

LSTM-Autoencoder Network for the Detection of Seismic Electric Signals

Cited by 12 publications

References 47 publications

MagInfoNet: Magnitude Estimation Using Seismic Information Augmentation and Graph Transformer

MagInfoNet: Magnitude Estimation Using Seismic Information Augmentation and Graph Transformer

A Numerical Study of Electrokinetically Generated Pre-Earthquake Geoelectric Fields in Layered Porous Media

Fast Initial Model Design for Electrical Resistivity Inversion by Using Broad Learning Framework

Contact Info

Product

Resources

About