Fourier-DeepONet: Fourier-enhanced deep operator networks for full waveform inversion with improved accuracy, generalizability, and robustness

Zhu, Min; Feng, Shihang; Lin, Youzuo; Lu, Lu

doi:10.1016/j.cma.2023.116300

Fourier-DeepONet: Fourier-enhanced deep operator networks for full waveform inversion with improved accuracy, generalizability, and robustness

Journal Article · Tue Aug 08 00:00:00 EDT 2023 · Computer Methods in Applied Mechanics and Engineering

DOI:https://doi.org/10.1016/j.cma.2023.116300· OSTI ID:2527399

^[1]; Feng, Shihang ^[2]; ^[2]; ^[3]

Yale University, New Haven, CT (United States); Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
Yale University, New Haven, CT (United States)

In this article, full waveform inversion (FWI) infers the subsurface structure information from seismic waveform data by solving a non-convex optimization problem. Data-driven FWI has been increasingly studied with various neural network architectures to improve accuracy and computational efficiency. Nevertheless, the applicability of pre-trained neural networks is severely restricted by potential discrepancies between the source function used in the field survey and the one utilized during training. Here, we develop a Fourier-enhanced deep operator network (Fourier-DeepONet) for FWI with the generalization of seismic sources, including the frequencies and locations of sources. Specifically, we employ the Fourier neural operator as the decoder of DeepONet, and we utilize source parameters as one input of Fourier-DeepONet, facilitating the resolution of FWI with variable sources. To test Fourier-DeepONet, we develop three new and realistic FWI benchmark datasets (FWI-F, FWI-L, and FWI-FL) with varying source frequencies, locations, or both. Our experiments demonstrate that compared with existing data-driven FWI methods, Fourier-DeepONet obtains more accurate predictions of subsurface structures in a wide range of source parameters. Moreover, the proposed Fourier-DeepONet exhibits superior robustness when handling data with Gaussian noise or missing traces and sources with Gaussian noise, paving the way for more reliable and accurate subsurface imaging across diverse real conditions.

View Accepted Manuscript (DOE)

Research Organization:: University of Pennsylvania, Philadelphia, PA (United States)

Sponsoring Organization:: USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR)

Grant/Contract Number:: SC0022953

OSTI ID:: 2527399

Alternate ID(s):: OSTI ID: 2479496

Journal Information:: Computer Methods in Applied Mechanics and Engineering, Journal Name: Computer Methods in Applied Mechanics and Engineering Vol. 416; ISSN 0045-7825

Publisher:: ElsevierCopyright Statement

Country of Publication:: United States

Language:: English

References (30)

U-FNO—An enhanced Fourier neural operator-based deep-learning model for multiphase flow Wen, Gege; Li, Zongyi; Azizzadenesheli, Kamyar Advances in Water Resources, Vol. 163 https://doi.org/10.1016/j.advwatres.2022.104180	journal	May 2022
A comprehensive and fair comparison of two neural operators (with practical extensions) based on FAIR data Lu, Lu; Meng, Xuhui; Cai, Shengze Computer Methods in Applied Mechanics and Engineering, Vol. 393 https://doi.org/10.1016/j.cma.2022.114778	journal	April 2022
Reliable extrapolation of deep neural operators informed by physics or sparse observations Zhu, Min; Zhang, Handi; Jiao, Anran Computer Methods in Applied Mechanics and Engineering, Vol. 412 https://doi.org/10.1016/j.cma.2023.116064	journal	July 2023
DeepM&Mnet: Inferring the electroconvection multiphysics fields based on operator approximation by neural networks Cai, Shengze; Wang, Zhicheng; Lu, Lu Journal of Computational Physics, Vol. 436 https://doi.org/10.1016/j.jcp.2021.110296	journal	July 2021
DeepM&Mnet for hypersonics: Predicting the coupled flow and finite-rate chemistry behind a normal shock using neural-network approximation of operators Mao, Zhiping; Lu, Lu; Marxen, Olaf Journal of Computational Physics, Vol. 447 https://doi.org/10.1016/j.jcp.2021.110698	journal	December 2021
Approximation rates of DeepONets for learning operators arising from advection–diffusion equations Deng, Beichuan; Shin, Yeonjong; Lu, Lu Neural Networks, Vol. 153 https://doi.org/10.1016/j.neunet.2022.06.019	journal	September 2022
Deep Learning for Geophysics: Current and Future Trends Yu, Siwei; Ma, Jianwei Reviews of Geophysics, Vol. 59, Issue 3 https://doi.org/10.1029/2021RG000742	journal	July 2021
Physics-informed machine learning Karniadakis, George Em; Kevrekidis, Ioannis G.; Lu, Lu Nature Reviews Physics, Vol. 3, Issue 6 https://doi.org/10.1038/s42254-021-00314-5	journal	May 2021
Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators Lu, Lu; Jin, Pengzhan; Pang, Guofei Nature Machine Intelligence, Vol. 3, Issue 3 https://doi.org/10.1038/s42256-021-00302-5	journal	March 2021
Generalizing universal function approximators Higgins, Irina Nature Machine Intelligence, Vol. 3, Issue 3 https://doi.org/10.1038/s42256-021-00318-x	journal	March 2021
Operator learning for predicting multiscale bubble growth dynamics Lin, Chensen; Li, Zhen; Lu, Lu The Journal of Chemical Physics, Vol. 154, Issue 10 https://doi.org/10.1063/5.0041203	journal	March 2021
Absorbing boundary conditions for numerical simulation of waves Engquist, B.; Majda, A. Proceedings of the National Academy of Sciences, Vol. 74, Issue 5 https://doi.org/10.1073/pnas.74.5.1765	journal	May 1977
Multifidelity deep neural operators for efficient learning of partial differential equations with application to fast inverse design of nanoscale heat transport Lu, Lu; Pestourie, Raphaël; Johnson, Steven G. Physical Review Research, Vol. 4, Issue 2 https://doi.org/10.1103/PhysRevResearch.4.023210	journal	June 2022
A Physics-Based Neural-Network Way to Perform Seismic Full Waveform Inversion Ren, Yuxiao; Xu, Xinji; Yang, Senlin IEEE Access, Vol. 8 https://doi.org/10.1109/ACCESS.2020.2997921	journal	January 2020
Physics-Guided Data-Driven Seismic Inversion: Recent progress and future opportunities in full-waveform inversion Lin, Youzuo; Theiler, James; Wohlberg, Brendt IEEE Signal Processing Magazine, Vol. 40, Issue 1 https://doi.org/10.1109/MSP.2022.3217658	journal	January 2023
InversionNet: An Efficient and Accurate Data-Driven Full Waveform Inversion Wu, Yue; Lin, Youzuo IEEE Transactions on Computational Imaging, Vol. 6 https://doi.org/10.1109/TCI.2019.2956866	journal	January 2020
Data-Driven Seismic Waveform Inversion: A Study on the Robustness and Generalization Zhang, Zhongping; Lin, Youzuo IEEE Transactions on Geoscience and Remote Sensing, Vol. 58, Issue 10 https://doi.org/10.1109/TGRS.2020.2977635	journal	October 2020
Image Quality Assessment: From Error Visibility to Structural Similarity Wang, Z.; Bovik, A. C.; Sheikh, H. R. IEEE Transactions on Image Processing, Vol. 13, Issue 4 https://doi.org/10.1109/TIP.2003.819861	journal	April 2004
Learning the solution operator of parametric partial differential equations with physics-informed DeepONets Wang, Sifan; Wang, Hanwen; Perdikaris, Paris Science Advances, Vol. 7, Issue 40 https://doi.org/10.1126/sciadv.abi8605	journal	October 2021
MIONet: Learning Multiple-Input Operators via Tensor Product Jin, Pengzhan; Meng, Shuai; Lu, Lu SIAM Journal on Scientific Computing, Vol. 44, Issue 6 https://doi.org/10.1137/22M1477751	journal	November 2022
Source‐independent full‐waveform inversion of seismic data Lee, Ki Ha; Kim, Hee Joon GEOPHYSICS, Vol. 68, Issue 6 https://doi.org/10.1190/1.1635054	journal	November 2003
An overview of full-waveform inversion in exploration geophysics Virieux, J.; Operto, S. GEOPHYSICS, Vol. 74, Issue 6 https://doi.org/10.1190/1.3238367	journal	November 2009
A theory-guided deep-learning formulation and optimization of seismic waveform inversion Sun, Jian; Niu, Zhan; Innanen, Kristopher A. GEOPHYSICS, Vol. 85, Issue 2 https://doi.org/10.1190/geo2019-0138.1	journal	March 2020
Deep-learning seismic full-waveform inversion for realistic structural models Liu, Bin; Yang, Senlin; Ren, Yuxiao GEOPHYSICS, Vol. 86, Issue 1 https://doi.org/10.1190/geo2019-0435.1	journal	January 2021
Physics-guided deep learning for seismic inversion with hybrid training and uncertainty analysis Sun, Jian; Innanen, Kristopher A.; Huang, Chao GEOPHYSICS, Vol. 86, Issue 3 https://doi.org/10.1190/geo2020-0312.1	journal	March 2021
Seismic velocity model building using neural networks: Training data design and learning generalization Alzahrani, Hani; Shragge, Jeffrey GEOPHYSICS, Vol. 87, Issue 2 https://doi.org/10.1190/geo2020-0547.1	journal	March 2022
Integrating deep neural networks with full-waveform inversion: Reparameterization, regularization, and uncertainty quantification Zhu, Weiqiang; Xu, Kailai; Darve, Eric GEOPHYSICS, Vol. 87, Issue 1 https://doi.org/10.1190/geo2020-0933.1	journal	December 2021
Deep-learning tomography Araya-Polo, Mauricio; Jennings, Joseph; Adler, Amir The Leading Edge, Vol. 37, Issue 1 https://doi.org/10.1190/tle37010058.1	journal	January 2018
Deep learning-based artificial bandwidth extension: Training on ultrasparse OBN to enhance towed-streamer FWI Aharchaou, Mehdi; Baumstein, Anatoly The Leading Edge, Vol. 39, Issue 10 https://doi.org/10.1190/tle39100718.1	journal	October 2020
Deep Learning with Adaptive Attention for Seismic Velocity Inversion Li, Fangda; Guo, Zhenwei; Pan, Xinpeng Remote Sensing, Vol. 14, Issue 15 https://doi.org/10.3390/rs14153810	journal	August 2022

Similar Records

Multiscale Data-Driven Seismic Full-Waveform Inversion With Field Data Study

Journal Article · Fri Oct 01 00:00:00 EDT 2021 · IEEE Transactions on Geoscience and Remote Sensing · OSTI ID:1827579

Related Subjects

42 ENGINEERING
Benchmark datasets
Deep learning
Earth Sciences
Fourier-DeepONet
Full waveform inversion
Generalizability
Robustness

Fourier-DeepONet: Fourier-enhanced deep operator networks for full waveform inversion with improved accuracy, generalizability, and robustness

Citation Formats

References (30)

Similar Records

Related Subjects