1887

Abstract

Summary

Seismic waveform modeling is an essential part of full waveform inversion (FWI), which provides high-resolution images of the near surface known by high velocity variations. However, physics-based approaches for solving wave equation, require extremely high computational cost, because they rely on fine grids to obtain accurate results. Physics-informed neural networks (PINNs) learn solutions to a given PDE for a single instance. For any new instances of functional parameters or coefficients, PINNs require the training of a new neural network, and thus suffer from computation and generalization issues. Neural operators like Deep Neural Operator (DeepONet) learns the mappings between infinite-dimensional spaces of functions. The neural operator needs to be trained only once. Obtaining a solution for a new instance of the parameter requires only a forward pass of the network, reducing the major computational issues. To examine the applicability of DeepONet to solve elastic wave equation, we developed a Fourier-based DeepONet, which achieves seismic wave modeling with minor error. Furthermore, it has strong generalization ability to higher resolution and robustness to velocity error.

© EAGE Publications BV
Loading

Article metrics loading...

/content/papers/10.3997/2214-4609.202520060
2025-09-07
2026-02-11

  • From This Site

    /content/papers/10.3997/2214-4609.202520060
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Journal -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. Delouis, B. et al., (2021). “Constraining the point source parameters of the 11 November 2019 Mw 4.9 Le Teil earthquake using multiple relocation approaches, first motion and full waveform inversions.” Comptes Rendus. Géoscience353(S1): 1–24.
     [Google Scholar]
  2. Lehmann, F. et al., (2024). “Synthetic ground motions in heterogeneous geologies: the HEMEW-3D dataset for scientific machine learning.” Earth System Science Data Discussions2024: 1–26.
     [Google Scholar]
  3. LeVeque, R. J. (1998). “Finite difference methods for differential equations.” Draft version for use in AMath585(6): 112.
     [Google Scholar]
  4. Li, Z. et al., (2020). “Fourier neural operator for parametric partial differential equations.” arXivpreprint arXiv:2010.08895.
     [Google Scholar]
  5. Lu, L. et al., (2021). “Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators.” Nature machine intelligence3(3): 218–229.
     [Google Scholar]
  6. Raissi, M. et al., (2019). “Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations.” Journal of Computational physics378: 686–707.
     [Google Scholar]
  7. Süli, E. and D. F.Mayers (2003). An introduction to numerical analysis, Cambridge university press.
     [Google Scholar]
/content/papers/10.3997/2214-4609.202520060
Loading
Seismic Waveform Modelling with Fourier-based DeepONet
Conference Proceedings 2025, 1 (2025); https://doi.org/10.3997/2214-4609.202520060
/content/papers/10.3997/2214-4609.202520060
/content/papers/10.3997/2214-4609.202520060
Loading

Data & Media loading...

Most Cited This Month Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error