Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
169 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Multiple-Input Fourier Neural Operator (MIFNO) for source-dependent 3D elastodynamics (2404.10115v2)

Published 15 Apr 2024 in cs.LG and physics.geo-ph

Abstract: Numerical simulations are essential tools to evaluate the solution of the wave equation in complex settings, such as three-dimensional (3D) domains with heterogeneous properties. However, their application is limited by high computational costs and existing surrogate models lack the flexibility of numerical solvers. This work introduces the Multiple-Input Fourier Neural Operator (MIFNO) to deal with structured 3D fields representing material properties as well as vectors describing the source characteristics. The MIFNO is applied to the problem of elastic wave propagation in the Earth's crust. It is trained on the HEMEWS-3D database containing 30000 earthquake simulations in different heterogeneous domains with random source positions and orientations. Outputs are time- and space-dependent surface wavefields. The MIFNO predictions are assessed as good to excellent based on Goodness-Of-Fit (GOF) criteria. Wave arrival times and wave fronts' propagation are very accurate since 80% of the predictions have an excellent phase GOF. The fluctuations amplitudes are good for 87% of the predictions. The envelope score is hindered by the small-scale fluctuations that are challenging to capture due to the complex physical phenomena associated with high-frequency features. Nevertheless, the MIFNO can generalize to sources located outside the training domain and it shows good generalization ability to a real complex overthrust geology. When focusing on a region of interest, transfer learning improves the accuracy with limited additional costs, since GOF scores improved by more than 1 GOF unit with only 500 additional specific samples. The MIFNO is the first surrogate model offering the flexibility of an earthquake simulator with varying sources and material properties. Its good accuracy and massive speed-up offer new perspectives to replace numerical simulations in many-query problems.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (43)
  1. Earthquake Early Warning: Advances, Scientific Challenges, and Societal Needs. Annual Review of Earth and Planetary Sciences 47, 361–388. doi:10.1146/annurev-earth-053018-060457.
  2. 3-D Salt and Overthrust Models. SEG/EAGE 3-D Modeling Series No.1.
  3. Probabilistic Seismic Hazard Analysis. Technical Report.
  4. Are Neural Operators Really Neural Operators? Frame Theory Meets Operator Learning. URL: http://arxiv.org/abs/2305.19913, arXiv:2305.19913.
  5. Frequency bias in neural networks for input of non-uniform density, in: Proceedings of the 37th International Conference on Machine Learning, JMLR.org.
  6. Graph neural networks for multivariate time series regression with application to seismic data. International Journal of Data Science and Analytics 16, 317–332. doi:10.1007/s41060-022-00349-6.
  7. Guidance on the Utilization of Earthquake-Induced Ground Motion Simulations in Engineering Practice. Earthquake Spectra 33, 809–835. doi:10.1193/120216eqs219ep.
  8. Message Passing Neural PDE Solvers, in: International Conference on Learning Representations. URL: https://openreview.net/forum?id=vSix3HPYKSU.
  9. Earthquake early warning: Recent advances and perspectives. Earth-Science Reviews 205, 103184. doi:10.1016/j.earscirev.2020.103184.
  10. DeepShake: Shaking Intensity Prediction Using Deep Spatiotemporal RNNs for Earthquake Early Warning. Seismological Research Letters doi:10.1785/0220210141.
  11. 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. Géoscience 353, 1–24. doi:10.5802/crgeos.78.
  12. 2d and 3D elastic wave propagation by a pseudo-spectral domain decomposition method. Journal of Seismology 1, 237–251. doi:10.1023/A:1009758820546.
  13. En-DeepONet: An enrichment approach for enhancing the expressivity of neural operators with applications to seismology. Computer Methods in Applied Mechanics and Engineering 420, 116681. doi:10.1016/j.cma.2023.116681.
  14. GNOT: A general neural operator transformer for operator learning, in: Krause, A., Brunskill, E., Cho, K., Engelhardt, B., Sabato, S., Scarlett, J. (Eds.), Proceedings of the 40th International Conference on Machine Learning, PMLR. pp. 12556–12569. URL: https://proceedings.mlr.press/v202/hao23c.html.
  15. Onsite Early Prediction of PGA Using CNN With Multi-Scale and Multi-Domain P-Waves as Input. Frontiers in Earth Science 9, 626908. doi:10.3389/feart.2021.626908.
  16. Fourier-MIONet: Fourier-enhanced multiple-input neural operators for multiphase modeling of geological carbon sequestration. URL: http://arxiv.org/abs/2303.04778, arXiv:2303.04778.
  17. MIONet: Learning multiple-input operators via tensor product. SIAM Journal on Scientific Computing 44, A3490–A3514. doi:10.1137/22M1477751.
  18. Rapid prediction of earthquake ground shaking intensity using raw waveform data and a convolutional neural network. Geophysical Journal International 222, 1379–1389. doi:10.1093/gji/ggaa233.
  19. Introduction to the spectral element method for three-dimensional seismic wave propagation. Geophysical Journal International 139, 806–822. doi:10.1046/j.1365-246x.1999.00967.x.
  20. Feasibility of Using Fourier Neural Operators for 3D Elastic Seismic Simulations. Technical Report LLNL-TR-854521. Lawrence Livermore National Laboratory. URL: https://www.osti.gov/servlets/purl/2001189.
  21. Time-frequency misfit and goodness-of-fit criteria for quantitative comparison of time signals. Geophysical Journal International 178, 813–825. doi:10.1111/j.1365-246X.2009.04177.x.
  22. Fourier Neural Operator Surrogate Model to Predict 3D Seismic Waves Propagation, in: 5th ECCOMAS Thematic Conference on Uncertainty Quantification in Computational Sciences and Engineering, Athens, Greece. doi:10.7712/120223.10339.20362.
  23. 3D elastic wave propagation with a Factorized Fourier Neural Operator (F-FNO). Computer Methods in Applied Mechanics and Engineering 420, 116718. doi:10.1016/j.cma.2023.116718.
  24. Synthetic ground motions in heterogeneous geologies: The HEMEW-3D dataset for scientific machine learning. Earth System Science Data Discussions 2024, 1–26. doi:10.5194/essd-2023-470.
  25. Maximum likelihood estimation of intrinsic dimension, in: Saul, L., Weiss, Y., Bottou, L. (Eds.), Advances in Neural Information Processing Systems, MIT Press. URL: https://proceedings.neurips.cc/paper_files/paper/2004/file/74934548253bcab8490ebd74afed7031-Paper.pdf.
  26. Fourier Neural Operator for Parametric Partial Differential Equations, in: International Conference on Learning Representations. doi:10.48550/arXiv.2010.08895.
  27. Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators. Nature Machine Intelligence 3, 218–229. doi:10.1038/s42256-021-00302-5.
  28. Key structural parameters affecting earthquake ground motion in 2D and 3D sedimentary structures. Bulletin of Earthquake Engineering 16, 2421–2450. doi:10.1007/s10518-018-0345-5.
  29. Machine Learning in Earthquake Seismology. Annual Review of Earth and Planetary Sciences 51, 105–129. doi:10.1146/annurev-earth-071822-100323.
  30. Imaging and seismic modelling inside volcanoes using machine learning. Scientific Reports 13, 630. doi:10.1038/s41598-023-27738-6.
  31. On the spectral bias of neural networks, in: Chaudhuri, K., Salakhutdinov, R. (Eds.), Proceedings of the 36th International Conference on Machine Learning, PMLR. pp. 5301–5310. URL: https://proceedings.mlr.press/v97/rahaman19a.html.
  32. U-NO: U-shaped Neural Operators. Transactions on Machine Learning Research URL: https://openreview.net/forum?id=j3oQF9coJd.
  33. Comparison of 3D, 2D and 1D numerical approaches to predict long period earthquake ground motion in the Gubbio plain, Central Italy. Bulletin of Earthquake Engineering 9, 2007–2029. doi:10.1007/s10518-011-9289-8.
  34. Regional physics-based simulation of ground motion within the Rhône Valley, France, during the Mw 4.9 2019 Le Teil earthquake. Bulletin of Earthquake Engineering 21, 1747–1774. doi:10.1007/s10518-022-01591-w.
  35. Modeling of random anisotropic elastic media and impact on wave propagation. European Journal of Computational Mechanics 19, 241–253. doi:10.3166/ejcm.19.241-253.
  36. SEM3D: A 3D High-Fidelity Numerical Earthquake Simulator for Broadband (0–10 Hz) Seismic Response Prediction at a Regional Scale. Geosciences 12, 112. doi:10.3390/geosciences12030112.
  37. Factorized fourier neural operators, in: The Eleventh International Conference on Learning Representations. URL: https://openreview.net/forum?id=tmIiMPl4IPa.
  38. Multitechnology characterization of an unusual surface rupturing intraplate earthquake: The M L 5.4 2019 Le Teil event in France. Geophysical Journal International 226, 803–813. doi:10.1093/gji/ggab136.
  39. Prediction of PGA in earthquake early warning using a long short-term memory neural network. Geophysical Journal International , ggad067doi:10.1093/gji/ggad067.
  40. Rapid Seismic Waveform Modeling and Inversion with Neural Operators. IEEE Transactions on Geoscience and Remote Sensing , 1–1doi:10.1109/TGRS.2023.3264210.
  41. Learning to solve the elastic wave equation with Fourier Neural Operators. Geophysics , 1–63doi:10.1190/geo2022-0268.1.
  42. Seismic Aggravation in Shallow Basins in Addition to One-dimensional Site Amplification. Journal of Earthquake Engineering 24, 1477–1499. doi:10.1080/13632469.2018.1472679.
  43. Deep Neural Helmholtz Operators for 3D Elastic Wave Propagation and Inversion. URL: http://arxiv.org/abs/2311.09608, arXiv:2311.09608.
Citations (1)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets