Geophysics and Geophysical Exploration (지구물리와물리탐사)

Korean Society of Earth and Exploration Geophysicists (한국지구물리·물리탐사학회)
권호 표지
DOI QR코드

DOI QR Code

A Review of Deep Learning-based Trace Interpolation and Extrapolation Techniques for Reconstructing Missing Near Offset Data

가까운 벌림 빠짐 해결을 위한 딥러닝 기반의 트레이스 내삽 및 외삽 기술에 대한 고찰

  • Jiho Park (Department of Earth Resources and Environmental Engineering, Hanyang University) ;
  • Soon Jee Seol (Department of Earth Resources and Environmental Engineering, Hanyang University) ;
  • Joongmoo Byun (Department of Earth Resources and Environmental Engineering, Hanyang University)
  • 박지호 (한양대학교 자원환경공학과) ;
  • 설순지 (한양대학교 자원환경공학과) ;
  • 변중무 (한양대학교 자원환경공학과)
  • Received : 2023.09.19
  • Accepted : 2023.11.07
  • Published : 2023.11.30
Download PDF
⟨ Previous Next ⟩

Abstract

In marine seismic surveys, the inevitable occurrence of trace gaps in the near offset resulting from geometrical differences between sources and receivers adversely affects subsequent seismic data processing and imaging. The absence of data in the near-offset region hinders accurate seismic imaging. Therefore, reconstructing the missing near-offset information is crucial for mitigating the influence of seismic multiples, particularly in the case of offshore surveys where the impact of multiple reflections is relatively more pronounced. Conventionally, various interpolation methods based on the Radon transform have been proposed to address the issue of the nearoffset data gap. However, these methods have several limitations, leading to the recent emergence of deep-learning (DL)-based approaches as alternatives. In this study, we conducted an in-depth analysis of two representative DL-based studies to scrutinize the challenges that future studies on near-offset interpolation must address. Furthermore, through field data experiments, we precisely analyze the limitations encountered when applying previous DL-based trace interpolation techniques to near-offset situations. Consequently, we suggest that near-offset data gaps must be approached by extrapolation rather than interpolation.

해양 탄성파 탐사 수행 시 송·수신 케이블의 구조적인 거리차에 의해서 필연적으로 발생하는 가까운 벌림(near offset)의 트레이스(trace)빠짐은 뒤따르는 탄성파 자료처리의 결과 및 영상화에 악영향을 끼치게 된다. 특히 가까운 벌림의 자료의 부재는 정확한 탄성파 영상화를 저해하는 다중반사파의 제거에 주요한 인자로 작용하므로 다중반사파의 영향력이 강해지는 천해 및 연안 탐사의 경우 빠짐을 효과적으로 해결해야 한다. 전통적으로 다양한 라돈 변환(Radon transform) 기반의 내삽 방법들이 가까운 벌림 빠짐의 해결책으로 제시되어왔으나 여러 한계점을 보여, 최근 이를 보완하기 위한 딥러닝(deep learning) 기반의 방법들이 제시되고 있다. 이 논문에서는 기존에 제시된 두 가지의 대표적인 딥러닝 기반의 접근법에 대해 면밀히 분석하여 앞으로 가까운 벌림 내삽 연구가 해결해야 하는 문제점들에 대해 깊이 있게 논의한다. 또한 기존의 딥러닝 기반의 트레이스 내삽 기술을 가까운 벌림 상황에 적용할 때 나타나는 한계점을 현장자료 실험을 통해 명확히 분석하여 향후 가까운 벌림 자료 빠짐의 문제는 내삽이 아닌 외삽으로 접근해야 한다는 것을 보여준다.

Keywords

Acknowledgement

이 논문은 2021년도 정부(과학기술정보통신부)의 재원으로 한국연구재단의 지원과(No. 2021R1A2C2014315) 2022년도 정부(교육부, 산업통상자원부)의 재원으로 K-CCUS 추진단의 지원을 받아 수행된 연구입니다(KCCUS20220001, 온실가스 감축 혁신인재양성사업).

References

  1. Dou, T., Li, K., Duan, H., Li, T., Dong, L., and Huang, Z., 2023, MDA GAN: Adversarial-Learning-Based 3-D Seismic Data Interpolation and Reconstruction for Complex Missing, IEEE Transactions on Geoscience and Remote Sensing, 61, 1-14. doi: 10.1109/TGRS.2023.3249476
  2. Dragoset, W., and Jericevic, Z., 1998, Some remarks on surface multiple attenuation, Geophysics, 63(2), 331-789. doi: 10.1190/1.1444377
  3. Hargreaves, N., 2006, Surface multiple attenuation in shallow water and the construction of primaries from multiples, SEG Technical Program Expanded Abstract 2006, 2689-2693. https://doi.org/10.1190/1.2370080
  4. He, T., Wu, B., and Zhu, X., 2021, Seismic Data Consecutively Missing Trace Interpolation Based on Multistage Neural Network Training Process, IEEE Geoscience and Remote Sensing Letters, 19, 1-5. doi: 10.1109/LGRS.2021.3089585
  5. Huang, H., Lin, L., Tong, R., Hu, H., Zhang, Q., Iwamoto, Y., Han, X., Chen, Y., and Wu, J., 2020, UNet3+:A full-scale connected unet for medical image segmentation, arXiv. https://doi.org/10.48550/arXiv.2004.08790
  6. Jia, Y., Yu, S., and Ma, J., 2017, What can machine learning do for seismic data processing? An interpolation application, Geophysics, 82(3), V163-V177. https://doi.org/10.1190/geo2016-0300.1
  7. Kabir, M. M. N., and Verschuur, D. J., 1995, Restoration of missing offsets by parabolic Radon transform, Geophysical Prospecting, 43, 347-368. https://doi.org/10.1111/j.1365-2478.1995.tb00257.x
  8. Kostov, C., de Melo, F. X., Raj, A., Zarkhidze, A., Cooke, A., Miers, G., and Bacon, J., 2015, Multiple attenuation for shallow- water surveys: Notes on old challenges and new opportunities, The Leading Edge, 34(7), 760-768. doi: 10.1190/tle34070760.1
  9. Lopez, G. A., and Verschuur, D. J., 2015, Closed-loop surfacerelated multiple elimination and its application to simultaneous data reconstruction, Geophysics, 80(6), V189-V199. https://doi.org/10.1190/geo2015-0287.1
  10. Ma, J., Wang, X., Gao, R., Zhang, X., Wang, Z., Wei, Y., Ma, J., Zhao, X., Huang, C., Jiang, S., Li, L., Yu, H., and Wang, H., 2017, Jingbian CCS project in China: 2015 update, Energy Procedia, 114, p.5768-5782. https://doi.org/10.1016/j.egypro.2017.03.1715
  11. Madiba, G. B., and McMechan, G. A., 2003, Processing, inversion, and interpretation of a 2D seismic data set from the North Viking Graben, North Sea, Geophysics, 68(3), 837-848. doi: 10.1190/1.1581036
  12. Mandelli, S., Lipari, V., Bestagini, P., and Tubaro, S., 2019, Interpolation and Denoising of Seismic Data using Convolutional Neural Networks, arXiv. https://doi.org/10.48550/arXiv.1901.07927
  13. Pan, S., Chen, K., Chen, J., Qin, Z., Cui, Q., and Li, J., 2020, A partial convolution-based deep-learning network for seismic data regularization1, Computers and Geosciences, 145, 104609. doi: 10.1016/j.cageo.2020.104609
  14. Park, J., Choi, J., Seol, S. J., Byun, J., and Kim, Y., 2021b, A method for adequate selection of training data sets to reconstruct seismic data using a convolutional U-Net, Geophysics, 86(5), V375-V388. https://doi.org/10.1190/geo2019-0708.1
  15. Park, S.-E., Kim, D.-S., Oh, J.-W., Min, D.-J., and Kim, B.-Y., 2021a, Review of the Applicability of 4D Time-Lapse Seismic Monitoring System Based on Frequency-Domain Reverse-Time Migration to Pohang CCS Demonstration, Journal of the Korean Society of Mineral and Energy Resources Engineers, 394-407. (In Korean with English abstract). https://www.jksmer.or.kr/articles/article/KGW2/
  16. Park, J., Yeeh, Z., Seol, S. J., and Byun, J., 2022, Seismic data interpolation using attention-based deep learning, 83rd EAGE Annual Conference 2022, (1), 1-5. https://scholar.google.co.kr/citations?user=2b-At00AAAAJ&hl=ko
  17. Park, M.-H., Lee, C. S., Kim, B.-Y., Kim, J.-H., Kim, K. J., and Shinn, Y. J., 2018. Preliminary results of the pre-injection monitoring survey at an offshore CO2 injection site in the Yeongil Bay, The Journal of Engineering Geology, 28(2), p.247-258. (In Korean with English abstract). https://koreascience.kr/article/JAKO201821142173409.page https://doi.org/10.9720/KSEG.2018.2.247
  18. Qu, S., Verschuur, E., Zhang D., and Chen, Y., 2021, Training deep networks with only synthetic data: Deep-learning-based near-offset reconstruction for (closed-loop) surface-related multiple estimation on shallow-water field data, Geophysics, 86(3), A39-A43. https://doi.org/10.1190/geo2020-0723.1
  19. Trad, D., 2003, Interpolation and multiple attenuation with migration operators, Geophysics, 68(6), 2043-2054. https://doi.org/10.1190/1.1635058
  20. Wang, B., Han, D., and Li, J., 2022, Missing shots and near-offset reconstruction of marine seismic data with towered streamers via self-supervised deep learning, IEEE Transactions on Geoscience and Remote Sensing, 60, 1-9. doi: 10.1109/TGRS.2022.3172145
  21. Wang, B., Zhang, N., Lu, W., and Wang, J., 2019, Deep-learning- based seismic data interpolation: A preliminary result, Geophysics, 84(1), V11-V20. https://doi.org/10.1190/geo2017-0495.1
  22. Wang, Y., Lu, W., and Li, Y., 2023, A self-adaptive antialiasing framework for seismic data interpolation, IEEE Transactions on Geoscience and Remote Sensing, 60, 1-12. doi: 10.1109/tgrs.2023.3272644
  23. Xu, Z., Sopher, D., Juhlin, C., Han, L., and Gong, X., 2018, Radon-domain interferometric interpolation for reconstruction of the near-offset gap in marine seismic data, Journal of Applied Geophysics, 151, 125-141. doi: 10.1016/j.jappgeo.2018.02.012
  24. Yu, J., and Wu, B., 2022, Attention and Hybrid Loss Guided Deep Learning for Consecutively Missing Seismic Data Reconstruction, IEEE Transactions on Geoscience and Remote Sensing, 60, 1-8. doi: 10.1109/TGRS.2021.3068279
  25. Zhang, D., and Verschuur, E., 2019, Integration of closed-loop surface-related multiple estimation and full wavefield migration for shallow water, 81st Annual International Conference and Exhibition, 1-5. https://doi.org/10.3997/2214-4609.201900779