Papers
Topics
Authors
Recent
Search
2000 character limit reached

WISE: full-Waveform variational Inference via Subsurface Extensions

Published 11 Dec 2023 in physics.geo-ph, cs.AI, cs.LG, eess.SP, and stat.AP | (2401.06230v1)

Abstract: We introduce a probabilistic technique for full-waveform inversion, employing variational inference and conditional normalizing flows to quantify uncertainty in migration-velocity models and its impact on imaging. Our approach integrates generative artificial intelligence with physics-informed common-image gathers, reducing reliance on accurate initial velocity models. Considered case studies demonstrate its efficacy producing realizations of migration-velocity models conditioned by the data. These models are used to quantify amplitude and positioning effects during subsequent imaging.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (64)
  1. An overview of full-waveform inversion in exploration geophysics. Geophysics, 74(6):WCC1–WCC26, 2009.
  2. Deep learning for geophysics: Current and future trends. Reviews of Geophysics, 59(3):e2021RG000742, 2021.
  3. An introduction to variational methods for graphical models. Machine learning, 37:183–233, 1999.
  4. Hierarchical text-conditional image generation with clip latents. arXiv preprint arXiv:2204.06125, 1(2):3, 2022.
  5. Felix J Herrmann. President’s page: Digital twins in the era of generative AI. The Leading Edge, 42(11):730–732, 2023.
  6. William W Symes. Migration velocity analysis and waveform inversion. Geophysical prospecting, 56(6):765–790, 2008.
  7. Learning likelihoods with conditional normalizing flows. arXiv preprint arXiv:1912.00042, 2019.
  8. Albert Tarantola. Inversion of seismic reflection data in the acoustic approximation. GEOPHYSICS, 49(8):1259–1266, 08 1984. doi: 10.1190/1.1441754. URL http://dx.doi.org/10.1190/1.1441754.
  9. Bayesian learning via stochastic gradient langevin dynamics. In Proceedings of the 28th international conference on machine learning (ICML-11), pages 681–688, 2011.
  10. Assessing uncertainties in velocity models and images with a fast nonlinear uncertainty quantification method. Geophysics, 83(2):R63–R75, 2018.
  11. Uncertainty quantification in time-lapse seismic imaging: A full-waveform approach. Geophysical Journal International, 222(2):1245–1263, 2020.
  12. Deep bayesian inference for seismic imaging with tasks. GEOPHYSICS, 87(5):S281–S302, 08 2022a. doi: 10.1190/geo2021-0666.1. URL http://dx.doi.org/10.1190/geo2021-0666.1.
  13. Bayesian Data Analysis. Chapman and Hall/CRC, 07 2003. doi: 10.1201/9780429258480. URL http://dx.doi.org/10.1201/9780429258480.
  14. Parameterizing uncertainty by deep invertible networks: An application to reservoir characterization. SEG Technical Program Expanded Abstracts 2020, 09 2020. doi: 10.1190/segam2020-3428150.1. URL http://dx.doi.org/10.1190/segam2020-3428150.1.
  15. Seismic velocity inversion and uncertainty quantification using conditional normalizing flows. In AGU Fall Meeting 2021. AGU, 2021.
  16. Enabling uncertainty quantification for seismic data preprocessing using normalizing flows (nf) – an interpolation example. First International Meeting for Applied Geoscience & Energy Expanded Abstracts, 09 2021. doi: 10.1190/segam2021-3583705.1. URL http://dx.doi.org/10.1190/segam2021-3583705.1.
  17. Learning by example: Fast reliability-aware seismic imaging with normalizing flows. First International Meeting for Applied Geoscience & Energy Expanded Abstracts, 09 2021. doi: 10.1190/segam2021-3581836.1. URL http://dx.doi.org/10.1190/segam2021-3581836.1.
  18. Preconditioned training of normalizing flows for variational inference in inverse problems. arXiv preprint arXiv:2101.03709, 2021.
  19. An introduction to variational inference in geophysical inverse problems. In Advances in Geophysics, volume 62, pages 73–140. Elsevier, 2021.
  20. Wave-equation-based inversion with amortized variational bayesian inference. arXiv preprint arXiv:2203.15881, 2022b.
  21. Adjoint operators enable fast and amortized machine learning based bayesian uncertainty quantification. In Medical Imaging 2023: Image Processing, volume 12464, pages 357–367. SPIE, 2023a.
  22. Reliable amortized variational inference with physics-based latent distribution correction. GEOPHYSICS, 88(3):R297–R322, 04 2023. doi: 10.1190/geo2022-0472.1. URL http://dx.doi.org/10.1190/geo2022-0472.1.
  23. A conditional normalizing flow for accelerated multi-coil mr imaging. arXiv preprint arXiv:2306.01630, 2023.
  24. Amortized normalizing flows for transcranial ultrasound with uncertainty quantification. arXiv preprint arXiv:2303.03478, 2023b.
  25. 3-d bayesian variational full waveform inversion. Geophysical Journal International, 234(1):546–561, 2023.
  26. Refining amortized posterior approximations using gradient-based summary statistics. arXiv preprint arXiv:2305.08733, 2023c.
  27. Bayesian variational time-lapse full-waveform inversion. arXiv preprint arXiv:2308.08805, 2023.
  28. Variational bayesian experimental design for geophysical applications. arXiv preprint arXiv:2307.01039, 2023.
  29. Inference of CO2 flow patterns–a feasibility study. arXiv preprint arXiv:2311.00290, 2023.
  30. Hint: Hierarchical invertible neural transport for density estimation and bayesian inference. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pages 8191–8199, 2021.
  31. Variational inference with normalizing flows. In International conference on machine learning, pages 1530–1538. PMLR, 2015.
  32. Learned multiphysics inversion with differentiable programming and machine learning. The Leading Edge, 42(7):474–486, 2023a.
  33. Bayesflow: Learning complex stochastic models with invertible neural networks. IEEE Transactions on Neural Networks and Learning Systems, 33(4):1452–1466, 04 2022. doi: 10.1109/tnnls.2020.3042395. URL http://dx.doi.org/10.1109/TNNLS.2020.3042395.
  34. Reverse time migration. Geophysics, 48(11):1514–1524, 1983.
  35. Gerard T. Schuster. Least-squares cross-well migration. SEG Technical Program Expanded Abstracts 1993, 01 1993. doi: 10.1190/1.1822308. URL http://dx.doi.org/10.1190/1.1822308.
  36. Least-squares migration of incomplete reflection data. GEOPHYSICS, 64(1):208–221, 01 1999. doi: 10.1190/1.1444517. URL http://dx.doi.org/10.1190/1.1444517.
  37. Generalized massive optimal data compression. Monthly Notices of the Royal Astronomical Society: Letters, 476(1):L60–L64, 2018.
  38. Accelerating extended least-squares migration with weighted conjugate gradient iteration. GEOPHYSICS, 81(4):S165–S179, 07 2016. doi: 10.1190/geo2015-0499.1. URL http://dx.doi.org/10.1190/geo2015-0499.1.
  39. Least-squares reverse time migration: Inversion-based imaging toward true reflectivity. The Leading Edge, 33(9):962–968, 09 2014. doi: 10.1190/tle33090962.1. URL http://dx.doi.org/10.1190/tle33090962.1.
  40. Wave-equation migration velocity analysis. i. theory. Geophysical Prospecting, 52(6):593–606, 2004.
  41. Multisource least-squares extended reverse-time migration with preconditioning guided gradient method. In SEG International Exposition and Annual Meeting, pages SEG–2013. SEG, 2013.
  42. An approximate inverse to the extended born modeling operator. Geophysics, 80(6):R331–R349, 2015.
  43. Inversion velocity analysis in the subsurface-offset domain. Geophysics, 83(2):R189–R200, 2018.
  44. Low-rank representation of omnidirectional subsurface extended image volumes. GEOPHYSICS, 86(3):S165–S183, 03 2021. doi: 10.1190/geo2020-0152.1. URL http://dx.doi.org/10.1190/geo2020-0152.1.
  45. Fons ten Kroode. An omnidirectional seismic image extension. Inverse Problems, 39(3):035003, 01 2023. doi: 10.1088/1361-6420/acb285. URL http://dx.doi.org/10.1088/1361-6420/acb285.
  46. Deep learning for velocity model building with common-image gather volumes. Geophysical Journal International, 228(2):1054–1070, 2022.
  47. Openfwi: Large-scale multi-structural benchmark datasets for full waveform inversion. Advances in Neural Information Processing Systems, 35:6007–6020, 2022.
  48. Building complex synthetic models to evaluate acquisition geometries and velocity inversion technologies. Proceedings, 06 2012. doi: 10.3997/2214-4609.20148575. URL http://dx.doi.org/10.3997/2214-4609.20148575.
  49. Unsupervised learning of full-waveform inversion: Connecting cnn and partial differential equation in a loop. arXiv preprint arXiv:2110.07584, 2021.
  50. Does full waveform inversion benefit from big data? arXiv preprint arXiv:2307.15388, 2023.
  51. Felix J Herrmann. Randomized sampling and sparsity: Getting more information from fewer samples. Geophysics, 75(6):WB173–WB187, 2010.
  52. Devito (v3.1.0): an embedded domain-specific language for finite differences and geophysical exploration. Geoscientific Model Development, 12(3):1165–1187, 2019. doi: 10.5194/gmd-12-1165-2019. URL https://www.geosci-model-dev.net/12/1165/2019/.
  53. Architecture and performance of devito, a system for automated stencil computation. ACM Trans. Math. Softw., 46(1), apr 2020. ISSN 0098-3500. doi: 10.1145/3374916. URL https://doi.org/10.1145/3374916.
  54. A large-scale framework for symbolic implementations of seismic inversion algorithms in julia. GEOPHYSICS, 84(3):F57–F71, 2019. doi: 10.1190/geo2018-0174.1. URL https://doi.org/10.1190/geo2018-0174.1.
  55. slimgroup/judi.jl: v3.2.3, March 2023b. URL https://doi.org/10.5281/zenodo.7785440.
  56. Mathias Louboutin. slimgroup/ImageGather.jl: v0.2.6. Zenodo, 06 2023. doi: 10.5281/ZENODO.7997993. URL https://zenodo.org/record/7997993.
  57. Wave-based inversion at scale on graphical processing units with randomized trace estimation. Geophysical Prospecting, 07 2023. doi: 10.1111/1365-2478.13405. URL http://dx.doi.org/10.1111/1365-2478.13405.
  58. The effect of velocity uncertainty on migrated reflectors: Improvements from relative-depth imaging. Geophysics, 81(1):S21–S29, 2016.
  59. Dave Hale. An efficient method for computing local cross-correlations of multi-dimensional signals. CWP Report, 656:282, 2006.
  60. Least-squares horizons with local slopes and multigrid correlations. Geophysics, 83(4):IM29–IM40, 2018.
  61. Deep-learning-based 3d fault detection for carbon capture and storage. Geophysics, 88(4):IM101–IM112, 2023.
  62. Iterative amortized inference. In International Conference on Machine Learning, pages 3403–3412. PMLR, 2018.
  63. Sergey Fomel. Time-migration velocity analysis by velocity continuation. Geophysics, 68(5):1662–1672, 2003.
  64. Velocity continuation with fourier neural operators for accelerated uncertainty quantification. In SEG International Exposition and Annual Meeting, page D011S092R004. SEG, 2022c.
Citations (7)
View on Semantic Scholar

Summary

No one has generated a summary of this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Sign Up to Summarize

Paper to Video (Beta)

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Sign Up to Generate

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now

Continue Learning

We haven't generated follow-up questions for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now

Collections

Sign up for free to add this paper to one or more collections.

User Circle Single Streamline Icon: https://streamlinehq.com Sign Up