1887
Volume 55, Issue 3
  • ISSN: 0812-3985
  • E-ISSN: 1834-7533

Abstract

Preserving low-frequency components in seismic data is challenging due to data acquisition restrictions and the processing of low-cut filtering after surveys. If seismic data lack low frequencies, their resolution deteriorates, leading to inaccurate seismic interpretations. To resolve this problem, we developed a low-frequency reconstruction that employs a modified U-Net neural network. To improve the training of the neural network, we generated training data analogous to unseen data, based on the far-field signature as the wavelet source to retain the spectral characteristics of field data. We addressed potential overfitting by generating a large amount of various synthetic data through wave equation-based modelling using a variety of velocity and density models. After synthesising the seismic data, we implemented a filtering method to produce input data with insufficient low frequencies and label data with sufficient low frequencies. The far-field signature plays an important role in the successful reconstruction of low frequencies due to its intrinsic field data features and greater low-frequency information compared to field data alone. We tested the generalisation of the network using unseen synthetic and field data not used in the training stage, and analyzed the results in the time and frequency domains. Although the input data did not retain frequencies below 10 Hz, the trained network predicted low frequencies that were similar to the desired data. We also produced post-stack sections via simple processing to evaluate low-frequency reconstructions produced by our trained network. The low-frequency reconstruction scheme led to a better understanding of subsurface media.

© 2024 Australian Society of Exploration Geophysicists
Loading

Article metrics loading...

/content/journals/10.1080/08123985.2024.2317129
2024-05-03
2026-01-23

  • From This Site

    /content/journals/10.1080/08123985.2024.2317129
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Journal -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. Aharchaou, M., and A.Baumstein. 2020. Deep learning-based artificial bandwidth extension: Training on ultrasparse OBN to enhance towed-streamer FWI. The Leading Edge39 no. 10: 718–26. doi:10.1190/tle39100718.1.
     https://doi.org/10.1190/tle39100718.1 [Google Scholar]
  2. Brenders, A., J.Dellinger, I.Ahmed, E.Díaz, M.Gherasim, H.Jin, M.Vyas, and J.Naranjo. 2022. The Wolfspar experience with low-frequency seismic source field data: Motivation, processing, and implications. The Leading Edge41 no. 1: 9–18. doi:10.1190/tle41010010.1.
     https://doi.org/10.1190/tle41010010.1 [Google Scholar]
  3. Choi, W., and S.Pyun. 2021. Synthetic training data generation for fault detection based on deep learning. Geophysics and Geophysical Exploration24 no. 3: 89–97. doi:10.7582/GGE.2021.24.3.089.
     https://doi.org/10.7582/GGE.2021.24.3.089 [Google Scholar]
  4. Fang, J., H.Zhou, Y.Elita Li, Q.Zhang, L.Wang, P.Sun, and J.Zhang. 2020. Data-driven low-frequency signal recovery using deep-learning predictions in full-waveform inversion. Geophysics85 no. 6: A37–43. doi:10.1190/geo2020‑0159.1.
     https://doi.org/10.1190/geo2020-0159.1 [Google Scholar]
  5. Gardner, G.F., L.W.Gardner, and A.R.Gregory. 1974. Formation velocity and density—The diagnostic basics for stratigraphic traps. Geophysics39 no. 6: 770–80. doi:10.1190/1.1440465.
     https://doi.org/10.1190/1.1440465 [Google Scholar]
  6. Hanitzsch, C., D.Marion, A.Tura, W.B.Beydoun, D.Jizba, S.Jin, and D.J.Foster. 1998. AVO migration/inversion analysis. In Comparison of seismic inversion methods on a single real data set: Society of Exploration Geophysicists, ed. R. G.Keys and D. J.Foster, 57–74. doi:10.1190/1.9781560802082.ch1.
     https://doi.org/10.1190/1.9781560802082.ch1 [Google Scholar]
  7. Hu*, W.2014. FWI without low frequency data-beat tone inversion. In SEG technical program expanded abstracts 2014, 1116–20. Society of Exploration Geophysicists. doi:10.1190/segam2014‑0978.1.
     https://doi.org/10.1190/segam2014-0978.1 [Google Scholar]
  8. Hu, Y., L.Han, Z.Xu, F.Zhang, and J.Zeng. 2017. Adaptive multi-step full waveform inversion based on waveform mode decomposition. Journal of Applied Geophysics139: 195–210. doi:10.1016/j.jappgeo.2017.02.017.
     https://doi.org/10.1016/j.jappgeo.2017.02.017 [Google Scholar]
  9. Keys, R.G., and D.J.Foster. 1998. Comparison of seismic inversion methods on a single real data set. Society of Exploration Geophysicists. doi:10.1190/1.9781560802082.
     https://doi.org/10.1190/1.9781560802082 [Google Scholar]
  10. Lee, H.Y., S.C.Lim, D.J.Min, B.D.Kwon, and M.Park. 2009. 2D time-domain acoustic-elastic coupled modeling: a cell-based finite-difference method. Geosciences Journal13 no. 4: 407–14. doi:10.1007/s12303‑009‑0037‑x.
     https://doi.org/10.1007/s12303-009-0037-x [Google Scholar]
  11. Lee, H.Y., D.J.Min, B.D.Kown, and H.S.Yoo. 2008. Time-domain elastic wave modeling in anisotropic media using cell-based finite-difference method. Journal of the Korean Society of Mineral and Energy Resources Engineers45 no. 5: 536–45.
     [Google Scholar]
  12. Li, Y.E., and L.Demanet. 2015. Phase and amplitude tracking for seismic event separation. Geophysics80 no. 6: WD59–72. doi:10.1190/geo2015‑0075.1.
     https://doi.org/10.1190/geo2015-0075.1 [Google Scholar]
  13. Li, Y.E., and L.Demanet. 2016. Full-waveform inversion with extrapolated low-frequency data. Geophysics81 no. 6: R339–48. doi:10.4043/26626‑ms.
     https://doi.org/10.4043/26626-ms [Google Scholar]
  14. Madiba, G.B., and G.A.McMechan. 2003. Processing, inversion, and interpretation of a 2D seismic data set from the North Viking Graben, North Sea. Geophysics68 no. 3: 837–48. doi:10.1190/1.1581036.
     https://doi.org/10.1190/1.1581036 [Google Scholar]
  15. Martin, G.S., S.Larsen, and K.Marfurt. 2002. Marmousi-2: An updated model for the investigation of AVO in structurally complex areas. In SEG International exposition and annual meeting, SEG-2002. doi:10.1190/1.1817083.
     https://doi.org/10.1190/1.1817083 [Google Scholar]
  16. Min, D.J., C.Shin, and H.S.Yoo. 2004. Free-surface boundary condition in finite-difference elastic wave modeling. Bulletin of the Seismological Society of America94 no. 1: 237–50. doi:10.3997/2214‑4609‑pdb.5.p264.
     https://doi.org/10.3997/2214-4609-pdb.5.p264 [Google Scholar]
  17. Ovcharenko, O., V.Kazei, M.Kalita, D.Peter, and T.Alkhalifah. 2019. Deep learning for low-frequency extrapolation from multioffset seismic. Geophysics84 no. 6: R989–1001. doi:10.1190/geo2018‑0884.1.
     https://doi.org/10.1190/geo2018-0884.1 [Google Scholar]
  18. Park, B., and W.Ha. 2019. Efficient 3D acoustic wave propagation modeling using a cell-based finite difference method. Geophysics and Geophysical Exploration22 no. 2: 56–61. doi:10.7582/GGE.2019.22.2.056.
     https://doi.org/10.7582/GGE.2019.22.2.056 [Google Scholar]
  19. Ronneberger, O., P.Fischer, and T.Brox. 2015. U-net: Convolutional networks for biomedical image segmentation: Munich, Germany. In Medical image computing and computer-assisted intervention–MICCAI 2015: 18th International conference, proceedings, Part III, Vol.18, 234–41. Springer International Publishing. doi:10.1007/978‑3‑319‑24574‑4_28.
     https://doi.org/10.1007/978-3-319-24574-4_28 [Google Scholar]
  20. Sun, H., and L.Demanet. 2020. Extrapolated full-waveform inversion with deep learning EFWI-CNN. Geophysics85 no. 3: R275–88. doi:10.1190/geo2019‑0195.1.
     https://doi.org/10.1190/geo2019-0195.1 [Google Scholar]
  21. Ten Kroode, F., S.Bergler, C.Corsten, J.W.de Maag, F.Strijbos, and H.Tijhof. 2013. Broadband seismic data—The importance of low frequencies. Geophysics78 no. 2: WA3–14. doi:10.1190/geo2012‑0294.1.
     https://doi.org/10.1190/geo2012-0294.1 [Google Scholar]
  22. Wang, R., and F.Herrmann. 2016. Frequency down extrapolation with TV norm minimization. In SEG Technical Program Expanded Abstracts 2016. Society of Exploration Geophysicists, 1380–84. doi:10.1190/segam2016‑13879674.1.
     https://doi.org/10.1190/segam2016-13879674.1 [Google Scholar]
  23. Wang, M., S.Xu, and H.Zhou. 2020. Self-supervised learning for low frequency extension of seismic data. In SEG Technical Program Expanded Abstracts 2020, Society of Exploration Geophysicists, 1501–5. doi:10.1190/segam2020‑3427086.1.
     https://doi.org/10.1190/segam2020-3427086.1 [Google Scholar]
  24. Wu, X., Z.Geng, Y.Shi, N.Pham, S.Fomel, and G.Caumon. 2020. Building realistic structure models to train convolutional neural networks for seismic structural. Geophysics85 no. 4: WA27–39. doi:10.1190/segam2019‑3214282.1.
     https://doi.org/10.1190/segam2019-3214282.1 [Google Scholar]
  25. Wu, R.S., J.Luo, and B.Wu. 2014. Seismic envelope inversion and modulation signal model. Geophysics79 no. 3: WA13–24. doi:10.1190/geo2013‑0294.1.
     https://doi.org/10.1190/geo2013-0294.1 [Google Scholar]
  26. Zhang, P., L.Han, Z.Xu, F.Zhang, and Y.Wei. 2017. Sparse blind deconvolution based low-frequency seismic data reconstruction for multiscale full waveform inversion. Journal of Applied Geophysics139: 91–108. doi:10.1016/j.jappgeo.2017.02.021.
     https://doi.org/10.1016/j.jappgeo.2017.02.021 [Google Scholar]
/content/journals/10.1080/08123985.2024.2317129
Loading
Low-frequency marine seismic data reconstruction based on the far-field signature using a modified U-Net
Exploration Geophysics 55, 263 (2024); https://doi.org/10.1080/08123985.2024.2317129
/content/journals/10.1080/08123985.2024.2317129
/content/journals/10.1080/08123985.2024.2317129
Loading

Data & Media loading...

  • Article Type: Research Article
Keyword(s): Band-limited data; deep learning; far-field signature; low-frequency reconstruction; modified U-Net

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