Generative adversarial network as a stochastic subsurface model reconstruction


In geosciences, generative adversarial networks have been successfully applied to generate multiple realizations of rock properties from geological priors described by training images, within probabilistic seismic inversion and history matching methods. Here, the use of generative adversarial networks is proposed not as a model generator but as a model reconstruction technique for subsurface models where we do have access to sparse measurements of the subsurface properties of interest. We use sets of geostatistical realizations as training datasets combined with observed experimental data. These networks are applied to reconstruct nonstationary sedimentary channels and continuous elastic properties, such as P-wave propagation velocity, in the presence and absence of conditioning data. The reconstruction examples shown herein can be considered a post-processing step applied after seismic inversion and performed at those locations where the convergence of the inversion is low, and therefore, the inverted models are associated with high uncertainty. The application examples show the suitability of generative adversarial networks in learning the spatial structure of the data from sets of geostatistical realizations. The generated models reproduce the first- and second-order statistical moments and the spatial covariance matrix of the training dataset.

This is a preview of subscription content, log in to check access.


  1. 1.

    Arjovsky, M., Bottou, L.: Towards principled methods for training generative adversarial networks. (2017). arXiv:1701.04862v1 [stat.ML]

  2. 2.

    Arjovsky, M., Chintala, S., Bottou, L.: Wasserstein generative adversarial networks. Proceedings of the 34th international conference on machine learning. PMLR. 70, 214–223 (2017)

    Google Scholar 

  3. 3.

    Arpat, G., Caers, J.: Conditional simulation with patterns. Math. Geol. 39(2), 177–203 (2007).

    Article  Google Scholar 

  4. 4.

    Azevedo, L., Soares, A.: Geostatistical Methods for Reservoir Geophysics. Springer International Publishing (2017)

  5. 5.

    Barnett, S.A.: Convergence Problems with Generative Adversarial Networks. A dissertation presented for CCD Dissertations on a Mathematical Topic. Mathematical Institute. University of Oxford (2018). arXiv:1806.11382 [cs.LG]

  6. 6.

    Bjorck, J., Gomes, C., Selman, B., Weinerger, K.Q.: Understanding batch normalization. (2018). arXiv:1806.02375[cs.AI]

  7. 7.

    Bosch, M., Mukerji, T., Gonzalez, E.F.: Seismic inversion for reservoir properties combining statistical rock physics and geostatistics: a review. Geophysics. 75(5), 75A165 (2010).

    Article  Google Scholar 

  8. 8.

    Caterini, A.L., Chang, D.E.: A novel representation of neural networks. (2016). arXiv:1610.01549v2 [stat.ML]

  9. 9.

    Caterini, A.L., Chang, D.E.: A geometric framework for convolutional neural networks. (2016). arXiv:1608.04374v2 [stat.ML]

  10. 10.

    Chan, S., Elsheikh, A.H.: Parametrization and generation of geological models with generative adversarial networks. (2017). arXiv: 1708,01810v1 [stat.ML]

  11. 11.

    Chan, S., Elsheikh, A.H.: Parametric generation of conditional geological realizations using generative neural networks. Comput. Geosci. 23, 925–952 (2019)

    Article  Google Scholar 

  12. 12.

    Cox, T.F., Cox, M.A.A.: Multidimensional Scaling. Chapman & Hall (1994)

  13. 13.

    Creswell, A., White, T., Dumoulin, V., Arulkumaran, K., Sengupta, B., Bharath, A.: Generative adversarial networks: an overview. IEEE Signal Process. Mag. 35(1), 53–65 (2018).

    Article  Google Scholar 

  14. 14.

    Daly, C., Caers, J.: Multi-point geostatistics—and introductory overview. First Break. 28(9), 39–47 (2010).

    Article  Google Scholar 

  15. 15.

    Deutsch, C., Journel, A.G.: GSLIB: Geostatistical Software Library and Users’ Guide. Oxford University Press (1998)

  16. 16.

    Doyen P (2007) Seismic Reservoir Characterization: an Earth Modeling Perspective. EAGE

    Google Scholar 

  17. 17.

    Dubrule, O.: Geostatistics for Seismic Data Integration in Earth Models. SEG/EAGE Distinguished Instructor Short Course Number 6, Tulsa (2003)

    Google Scholar 

  18. 18.

    Dupont, E, Zhang, T, Tilke, P, Liang, L, Bailey, W. (2018). Generating realistic geology conditioned on physical measurements with generative adversarial networks. arXiv preprint arXiv:1802.03065

  19. 19.

    Ferreirinha, T., Nunes, R., Azevedo, L., Soares, A., Pratas, F., Tomás, P., Roma, N.: Acceleration of stochastic seismic inversion in OpenCL-based heterogeneous platforms. Comput Geosci. 26–36 (2015).

  20. 20.

    Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., Bengio, Y.: Generative adversarial networks. In: The twenty-eight annual conference on neural information processing systems (NIPS), Montréal, Canada (2014)

    Google Scholar 

  21. 21.

    Grana, D., Della Rossa, E.: Probabilistic petrophysical-properties estimation integrating statistical rock physics with seismic inversion. Geophysics. 75(3), O21–O37 (2010).

    Article  Google Scholar 

  22. 22.

    Gulrajani, I., Ahmed, F., Arjovsky, M. Dumoulin, V., Courville, A. (2017) Improved training of Wasserstein GANs. arXiv:1704.00028 [stat.ML]

  23. 23.

    Laloy, E., Hérault, R., Lee, J., Jacques, D., Linde, N.: Inversion using a new low-dimensional representation of complex binary geological media based on a deep neural network. Adv. Water Resour. 110, 387–405 (2017).

    Article  Google Scholar 

  24. 24.

    Laloy, E., Hérault, R., Lee, J., Jacques, D., Linde, N.: Training-image based geostatistical inversion using a spatial generative adversarial neural network. Water Resour. Res. 54(1), 381–406 (2017).

    Article  Google Scholar 

  25. 25.

    Mariethoz, G., Caers, J.: Multiple-Point Geostatistics: Stochastic Modelling with Training Images, 376 pp. Wiley-Blackwell, Hoboken (2014)

    Google Scholar 

  26. 26.

    Mariethoz, G., Renard, P., Straubhaar, J.: The direct sampling method to perform multiple-point geostatistical simulations. Water Resour. Res. 46(11), 1–14 (2010).

    Article  Google Scholar 

  27. 27.

    Mosser, L., Dubrule, O., Blunt, M.J.: Reconstruction of three-dimensional porous media using generative adversarial neural networks. Phys. Rev. E. 96, 043309 (2017).

    Article  Google Scholar 

  28. 28.

    Mosser L, Kimman W, Dramsch J, Purves S, De la Fuente A, Ganssle G (2018) Rapid seismic domain transfer: seismic velocity inversion and modeling using deep generative neural networks. 80th EAGE conference and exhibition 2018.

  29. 29.

    Mosser, L, Dubrule, O, Blunt, MJ (2018) Conditioning of three-dimensional generative adversarial networks for pore and reservoir-scale models. arXiv preprint arXiv:1802.05622

  30. 30.

    Radford A, Metz L, Chintala A (2016) Unsupervised representation learning with deep convolutional generative adversarial networks. arXiv: 1511.06434v2 [cs.LG]

  31. 31.

    Scheidt, C., Caers, J.: Representing spatial uncertainty using distances and kernels. Math. Geosci. 41, 397–419 (2009)

    Article  Google Scholar 

  32. 32.

    Soares, A.: Direct sequential simulation and cosimulation. Math. Geol. 33(8), 911–926 (2001).

    Article  Google Scholar 

  33. 33.

    Springenberg J T, Dosovitskiy A, Brox T, Riedmiller M (2015) Striving for simplicity: the all convolutional net. arXiv:1412.6806v3 [cs.LG]

  34. 34.

    Strebelle, S.: Conditional simulation of complex geological structures using multiple-point statistics. Math. Geol. 34(1), 1–21 (2002).

    Article  Google Scholar 

  35. 35.

    Tarantola, A.: Inverse Problem Theory. SIAM (2005)

  36. 36.

    Villani, C.: Optimal Transport: Old and New. Springer, Berlin (2009) ISBN 978-3-540-71049-3

    Google Scholar 

  37. 37.

    Yeh, R.A., Chen, C., Lim, T.Y., Schwing, A.G., Hasegawa-Johnson, M., Do, M.N.: Semantic image inpainting with deep generative models. 2017 IEEE conference on computer vision and pattern recognition (CVPR). p. 6882–6890 (2017).

    Google Scholar 

Download references


The authors gratefully acknowledge the support of the CERENA (strategic project FCT-FCT-UIDB/04028/2020). The authors acknowledge Arjovsky, M., Chintala, S., and Bottou, L. for making available the WGAN algorithm ( and Mosser, L., Dubrule, O., and Blunt, M.J. for the geogan conde ( The authors acknowledge the fruitful discussions with the two anonymous reviewers.

Author information



Corresponding author

Correspondence to Leonardo Azevedo.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Azevedo, L., Paneiro, G., Santos, A. et al. Generative adversarial network as a stochastic subsurface model reconstruction. Comput Geosci (2020).

Download citation


  • Generative adversarial network
  • Stochastic modeling
  • Model reconstruction