Seismic Tomography with Random Batch Gradient Reconstruction

DOI10.1137/21M1452342MaRDI QIDQ6098518zbMATH OpenFDO

Authors Li H. Chai, Zhongyi Huang, Xu Yang

Publication date 14 June 2023

Published in SIAM Journal on Scientific Computing (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2110.06455

zbMATH Keywords

frozen Gaussian approximationhigh-dimensional optimizationinverse problemsrandom batch methodseismic tomography

Mathematics Subject Classification ID

Asymptotic behavior of solutions to PDEs (35B40) Numerical methods for inverse problems for initial value and initial-boundary value problems involving PDEs (65M32) Numerical methods for partial differential equations, initial value and time-dependent initial-boundary value problems (65M99) Computational methods for problems pertaining to geophysics (86-08) Seismology (including tsunami modeling), earthquakes (86A15) Inverse problems in geophysics (86A22)

Abstract: Seismic tomography solves high-dimensional optimization problems to image subsurface structures of Earth. In this paper, we propose to use random batch methods to construct the gradient used for iterations in seismic tomography. Specifically, we use the frozen Gaussian approximation to compute seismic wave propagation, and then construct stochastic gradients by random batch methods. The method inherits the spirit of stochastic gradient descent methods for solving high-dimensional optimization problems. The proposed idea is general in the sense that it does not rely on the usage of the frozen Gaussian approximation, and one can replace it with any other efficient wave propagation solvers, e.g., Gaussian beam methods and spectral element methods. We prove the convergence of the random batch method in the mean-square sense, and show the numerical performance of the proposed method by two-dimensional and three-dimensional examples of wave-equation-based travel-time inversion and full-waveform inversion, respectively. As a byproduct, we also prove the convergence of the accelerated full-waveform inversion using dynamic mini-batches and spectral element methods.

Recommendations

• Bayesian methods for a particular inverse problem: seismic topography
• A semismooth Newton-CG method for constrained parameter identification in seismic tomography
• Analysis of seismic inversion with optimal transportation and softplus encoding
• 3D Frequency-Domain Seismic Inversion with Controlled Sloppiness

Cites work

• A convergent multiscale Gaussian-beam parametrix for the wave equation
• A frozen Gaussian approximation-based multi-level particle swarm optimization for seismic inversion
• A mathematical justification for the Herman-Kluk propagator
• Convergence of frozen Gaussian approximation for high-frequency wave propagation
• Fast multiscale Gaussian wavepacket transforms and multiscale Gaussian beams for the wave equation
• Frozen Gaussian approximation for 3D seismic tomography
• Frozen Gaussian approximation for high frequency wave propagation
• scientific article; zbMATH DE number 3821182 (Why is no real title available?)
• On the convergence of frozen Gaussian approximation for linear non-strictly hyperbolic systems
• Random batch methods (RBM) for interacting particle systems

Cited in

(8)

• Nonlinear regularization techniques for seismic tomography
• Corrigendum: Seismic tomography is locally ill-posed (2014 Inverse Problems 30 125001)
• Global seismic tomography: the inverse problem and beyond
• Frozen Gaussian approximation for 3D seismic tomography
• scientific article; zbMATH DE number 869707 (Why is no real title available?)
• First arrival traveltime tomography using supervised descent learning technique
• Frozen Gaussian sampling for scalar wave equations
• A guide to stochastic optimisation for large-scale inverse problems

This page was built for publication: Seismic Tomography with Random Batch Gradient Reconstruction

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6098518)