Inversion of a velocity model using artificial neural networks

Moya, Aarón; Irikura, Kojiro

doi:10.1016/j.cageo.2009.08.010

Cited by 36 publications

(14 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Seismic tomography approaches based on NNs, with early theory developed by Röth and Tarantola (1994), have also achieved compelling results. Moya and Irikura (2010) apply an NN approach to velocity model inversion. Gupta et al (2018) address the challenges of limited measurements in travel-time tomography using subspace modeling and convolutional NNs.…”

Section: Ground-motion Prediction Using Supervised Learningmentioning

confidence: 99%

Machine Learning in Seismology: Turning Data into Insights

Kong

Trugman

Ross

et al. 2018

Seismological Research Letters

397

197

View full text Add to dashboard Cite

This article provides an overview of current applications of machine learning (ML) in seismology. ML techniques are becoming increasingly widespread in seismology, with applications ranging from identifying unseen signals and patterns to extracting features that might improve our physical understanding. The survey of the applications in seismology presented here serves as a catalyst for further use of ML. Five research areas in seismology are surveyed in which ML classification, regression, clustering algorithms show promise: earthquake detection and phase picking, earthquake early warning (EEW), ground-motion prediction, seismic tomography, and earthquake geodesy. We conclude by discussing the need for a hybrid approach combining data-driven ML with traditional physical modeling.

show abstract

Section: Ground-motion Prediction Using Supervised Learningmentioning

confidence: 99%

Machine Learning in Seismology: Turning Data into Insights

Kong

Trugman

Ross

et al. 2018

Seismological Research Letters

397

197

View full text Add to dashboard Cite

show abstract

“…Neural network-based inversion methods have been applied to various nonlinear tomography problems in the past. Roth and Tarantola [14] first used NNs to estimate subsurface velocity structure from active source seismic waveforms, Moya and Irikura [16] performed velocity inversion with a neural network using waveform data from earthquakes and Araya-Polo et al [17] used semblance gathers as input to a network to invert for velocity structure. Gupta et al [18] used a convolutional network to learn an ensemble of simpler mappings in a low-dimensional space before reconstructing the image by combining the mappings.…”

Section: Introductionmentioning

confidence: 99%

Probabilistic neural network-based 2D travel-time tomography

Earp

Curtis

2020

Neural Comput & Applic

View full text Add to dashboard Cite

Travel-time tomography for the velocity structure of a medium is a highly nonlinear and nonunique inverse problem. Monte Carlo methods are becoming increasingly common choices to provide probabilistic solutions to tomographic problems but those methods are computationally expensive. Neural networks can often be used to solve highly nonlinear problems at a much lower computational cost when multiple inversions are needed from similar data types. We present the first method to perform fully nonlinear, rapid and probabilistic Bayesian inversion of travel-time data for 2D velocity maps using a mixture density network. We compare multiple methods to estimate probability density functions that represent the tomographic solution, using different sets of prior information and different training methodologies. We demonstrate the importance of prior information in such high-dimensional inverse problems due to the curse of dimensionality: unrealistically informative prior probability distributions may result in better estimates of the mean velocity structure; however, the uncertainties represented in the posterior probability density functions then contain less information than is obtained when using a less informative prior. This is illustrated by the emergence of uncertainty loops in posterior standard deviation maps when inverting travel-time data using a less informative prior, which are not observed when using networks trained on prior information that includes (unrealistic) a priori smoothness constraints in the velocity models. We show that after an expensive program of network training, repeated high-dimensional, probabilistic tomography is possible on timescales of the order of a second on a standard desktop computer.

show abstract

“…Thereafter any data set can be mapped to corresponding parameter values under that mapping. NNs have been applied successfully to many geophysical inverse problems, either to find a single deterministic solution that fits the observed data (Röth & Tarantola, 1994;Moya & Irikura, 2010;Araya-Polo et al, 2018;Bianco & Gerstoft, 2018;Kong et al, 2019), or to find a fully probabilistic result representing the posterior pdf (Devilee et al, 1999;Meier et al, 2007aMeier et al, , 2007bShahraeeni & Curtis, 2011;Shahraeeni et al, 2012;de Wit et al, 2013;Käufl et al, 2014Käufl et al, , 2015. The merit of these methods is their efficiency when inverting different data sets: once the NN has been properly trained, the inversion process can be accomplished rapidly (usually in seconds) by feeding each new observed data set into the NN.…”

Section: Introductionmentioning

confidence: 99%