Development of an automated and computationally low-cost method of huge data generation for training deep learning algorithms using direct sequential simulation

Authors

Abstract

Recent successes in inverse problems have led to instant demand for available training datasets. Lack of appropriate datasets, in terms of quantity and quality, has become a challenge in geophysical application of deep learning neural networks. Several methods have been developed to overcome this problem. In this paper, a framework for generating training datasets was introduced using a geostatistical simulation method called direct sequential simulation. The main idea was to extract well logs from different available velocity models and using them as input data for simulation and co-simulation algorithms. Using the secondary image for co-simulation with different correlation coefficients of 0.3, 0.5 and 0.7, various models were generated with different continuities. In this paper, various examples of well-known velocity models were selected to generate more suitable models in terms of geological structures.

Keywords

References

Aghamiry, H. S., Gholami, A., and Operto, S., 2019, Implementing bound constraints and total-variation regularization in extended full-waveform inversion with the alternating direction method of multiplier: Application to large contrast media. Geophysical Journal International, 218(2), 855–872.
Araya-Polo, M., Jennings, J., Adler, A. and Dahlke, T., 2018, Deep-learning tomography. Leading Edge, 37(1), 58–66.
Deutsch, C.V. and Journel, A.G., 1992, GSLIB: Geostatistical software library and user's guide, Oxford Univ Press, New York. 340.
Fang, J., Zhou, H., Elita Li, Y., Zhang, Q., Wang, L., Sun, P. and Zhang, J., 2020, Data-driven low-frequency signal recovery using deep-learning predictions in full-waveform inversion. Geophysics, 85(6), A37–A43.
Herrmann, F. J., A. Siahkoohi, and G. Rizzuti. Learned imaging with constraints and uncertainty quantification. In Neural Information Processing Systems (NeurIPS), 2019, Deep Inverse Workshop.
Johnson, J., Alahi, A. and Fei-Fei, L., 2016, Perceptual losses for real-time style transfer and super-resolution, European conference on computer vision. Springer, pp. 694–711.
Kaur, H., Pham, N., & Fomel, S., 2021, Seismic data interpolation using deep learning with generative adversarial networks. Geophysical Prospecting, 69(2), 307–326.
Laloy, E., Linde, N., Ruffino, C., Hrault, R., Gasso, G. and Jacques, D., 2019, Gradientbased deterministic inversion of geophysical data with generative adversarial networks: Is it feasible? Computers & Geosciences, 133, 104333.
Lecun, Y., Bengio, Y., & Hinton, G., 2015, Deep learning. Nature, 521(7553), 436–444.
Li, S., Liu, B., Ren, Y., Chen, Y., Yang, S., Wang, Y. and Jiang, P. (2020, Deep-Learning Inversion of Seismic Data. IEEE Transactions on Geoscience and Remote Sensing, 58(3), 2135–2149.
Mao, B., Han, L. G., Feng, Q. and Yin, Y. C. (2019). Subsurface velocity inversion from deep learning-based data assimilation. Journal of Applied Geophysics, 167, 172–179.
Ovcharenko, O. V., Kazei, D. P. and Alkhalifah, T., 2019, Style transfer for generation of realistically textured subsurface models: 89th Annual International Meeting, SEG, Expanded Abstracts, doi: 10.1190/segam2019- 3216349.1.
Ren, Y., Xu, X., Yang, S., Nie, L. and Chen, Y., 2020, A Physics-Based Neural-Network Way to Perform Seismic Full Waveform Inversion. IEEE Access, 8, 112266–112277.
Rezaee, H., D. Marcotte, P. Tahmasebi, and Saucier, A., 2015, Multiple-point geostatistical simulation using enriched pattern databases, Stochastic Environ. Res. Risk Assess., 29(3), 893–913
Siahkoohi, A., Rizzuti, G. and Herrmann, F. J., Deep bayesian inference for seismic imaging with tasks. arXiv preprint arXiv:2110.04825, 2021.
Siwei Yu, Jianwei Ma, and Wenlong Wang, 2019, Deep learning for denoisingdeep learning for denoising. Geophysics, 84(6):V333–V350.
Shihang Feng, Youzuo Lin, and Brendt Wohlberg, 2021, Multiscale data-driven seismic full-waveform inversion with field data study. IEEE Transactions on Geoscience and Remote Sensing, pp. 1–14.
Soares, A., 2001, Direct Sequential Simulation and Cosimulation, Mathematical Geology, 33, 911-926.
Tran, T.T., Deutsch, C.V. and Xie, Y., 2001, Direct Geostatistical Simulation With Multiscale Well, Seismic, and Production Data, SPE Annual Technical Conference and Exhibition, Vol. SPE Paper Number 71323. Society of Petroleum Engineers.
Virieux, J., and Operto, S., 2009, An overview of full-waveform inversion in exploration geophysics. Geophysics, 74(6). https://doi.org/10.1190/1.3238367
Wu, Y. and Lin, Y., 2019, InversionNet: An Efficient and Accurate Data-Driven Full Waveform Inversion. IEEE Transactions on Computational Imaging, 6, 419–433.
Yang, F. and Ma, J., 2019, Deep-learning inversion: A next-generation seismic velocity model building method. Geophysics, 84(4), R583–R599.
Volume 8, Issue 3
July 1401
Pages 189-201

Files

Share

How to cite

Statistics