This study investigates generative adversarial networks (GANs)’ capacity to model multi-facies distributions of meandering systems. Earlier works showed that GANs outperform geostatistical methods in reproducing complex geometry, like the shapes of fluvial channels. However, the reproduction of geological complexity and geological realism remains an issue when modelling fluvial depositional systems. Meandering systems deposit multiple facies and change facies shape following the migration of rivers. Sand accretes at the inner bank of channels, forming the point bar and erodes the plain at the outer bank to create sediments. Channel fills with mud or sand at the bottom after abandonment due to avulsions or meander cut-offs. Those sedimentary processes yield complex geological patterns. This paper proposes further developing a GAN model, Fluvial GAN, to learn complex multi-facies fluvial patterns across depositional variability. We create a set of meandering facies models by a process-based model, FLUMYTM, for training a GAN and assessing how well it can learn fluvial facies distributions representing sedimentary processes. Fluvial GAN has three distinct enhancements: (i) a One-Hot Encoder for better handling of multi-facies distribution, (ii) a Hybrid-discriminator for better learning geological patterns, and (iii) an improved loss function to prevent mode collapse. We compare Fluvial GAN performance with two more standard configurations using qualitative and quantitative geological features assessments. Fluvial GAN vastly reduces the occurrence of a typical unrealistic feature, channels forming isolated loops, which we called ‘closed channel’ in this study. We analyse the diversity of Fluvial GAN generations via a dimensionality reduction algorithm, UMAP, that plots the training dataset and Fluvial GAN generations together in a 2D space. Fluvial GAN provides good coverage of the uncertainty space represented by the training dataset.
This is a preview of subscription content, access via your institution.
Buy single article
Instant access to the full article PDF.
Price excludes VAT (USA)
Tax calculation will be finalised during checkout.
Codes are available at https://github.com/GeoDataScienceUQ/Fluvial_GAN.
Grimaud, J.-L., Ors, F., Lemay, M., Cojan, I., Rivoirard, J.: Preservation and completeness of fluvial meandering deposits influenced by channel motions and overbank sedimentation. J. Geophys. Res. Earth Surface, 2021–006435
Laloy, E., Hérault, R., Jacques, D., Linde, N.: Training-image based geostatistical inversion using a spatial generative adversarial neural network. Water Resour. Res. 54(1), 381–406 (2018)
Chan, S., Elsheikh, A.H.: Parametric generation of conditional geological realizations using generative neural networks. Comput. Geosci. 23(5), 925–952 (2019)
Zhang, T.-F., Tilke, P., Dupont, E., Zhu, L.-C., Liang, L., Bailey, W.: Generating geologically realistic 3D reservoir facies models using deep learning of sedimentary architecture with generative adversarial networks. Pet. Sci. 16(3), 541–549 (2019)
Song, S., Mukerji, T., Hou, J.: Geological facies modeling based on progressive growing of generative adversarial networks (GANs). Comput. Geosci. 25(3), 1251–1273 (2021)
Maharaja, A.: Tigenerator: object-based training image generator. Comput. Geosci. 34(12), 1753–1761 (2008)
Remy, N., Boucher, A., Wu, J.: Applied Geostatistics with SGeMS: A User’s Guide. Cambridge University Press, Cambridge (2009)
Bubnova, A.: On the conditioning of process-based channelized meandering reservoir models on well data. PhD thesis, Université Paris sciences et lettres (2018)
Lopez, S., Cojan, I., Rivoirard, J., Galli, A.: Process-Based Stochastic Modelling: Meandering Channelized Reservoirs. Analogue and Numerical Modelling of Sedimentary Systems: From Understanding to Prediction, pp 139–144. Wiley, Oxford (2009)
Rongier, G.: Lessons learned from simulating fluvial deposits using process-based models and generative adversarial networks. In: 21st Annual Conference of the International Association for Mathematical Geosciences. Beyond Gaussianity: GANs, MPS, Cumulants or Copula approaches? International Association for Mathematical Geosciences (2022)
Bogaart, P.W., Van Balen, R., Kasse, C., Vandenberghe, J.: Process-based modelling of Fluvial system response to rapid climate change—I: model formulation and generic applications. Quat. Sci. Rev. 22 (20), 2077–2095 (2003)
Azevedo, L., Paneiro, G., Santos, A., Soares, A.: Generative adversarial network as a stochastic subsurface model reconstruction. Comput. Geosci. 24(4), 1673–1692 (2020)
Song, S., Mukerji, T., Hou, J.: Bridging the gap between geophysics and geology with generative adversarial networks. IEEE Trans. Geosci. Remote Sens. 60, 1–11 (2021)
Zhang, C., Song, X., Azevedo, L.: U-net generative adversarial network for subsurface facies modeling. Comput. Geosci. 25(1), 553–573 (2021)
Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., Bengio, Y.: Generative adversarial networks. Commun. ACM 63(11), 139–144 (2020)
Hancock, J.T., Khoshgoftaar, T.M.: Survey on categorical data for neural networks. J. Big Data 7(1), 1–41 (2020)
Radford, A., Metz, L., Chintala, S.: Unsupervised representation learning with deep convolutional generative adversarial networks. arXiv:1511.06434 (2015)
Ioffe, S., Szegedy, C.: Batch normalization: Accelerating deep network training by reducing internal covariate shift. In: International Conference on Machine Learning, pp 448–456. PMLR (2015)
LeCun, Y., Bottou, L., Bengio, Y., Haffner, P.: Gradient-based learning applied to document recognition. Proc. IEEE 86(11), 2278–2324 (1998)
Metz, L., Poole, B., Pfau, D., Sohl-Dickstein, J.: Unrolled generative adversarial networks. arXiv:1611.02163 (2016)
Niculescu-Mizil, A., Perlich, C., Swirszcz, G., Sindhwani, V., Liu, Y., Melville, P., Wang, D., Xiao, J., Hu, J., Singh, M., et al.: Winning the KDD cup orange challenge with ensemble selection. In: KDD-Cup 2009 Competition, pp 23–34. PMLR (2009)
Park, T., Liu, M.-Y., Wang, T.-C., Zhu, J.-Y.: Semantic image synthesis with spatially-adaptive normalization. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp 2337–2346 (2019)
Wang, T.-C., Liu, M.-Y., Zhu, J.-Y., Tao, A., Kautz, J., Catanzaro, B.: High-resolution image synthesis and semantic manipulation with conditional GANs. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp 8798–8807 (2018)
Yu, F., Koltun, V.: Multi-scale context aggregation by dilated convolutions. arXiv:1511.07122 (2015)
Wang, P., Chen, P., Yuan, Y., Liu, D., Huang, Z., Hou, X., Cottrell, G.: Understanding convolution for semantic segmentation. In: 2018 IEEE Winter Conference on Applications of Computer Vision (WACV), pp 1451–1460. IEEE (2018)
Dumoulin, V., Visin, F.: A guide to convolution arithmetic for deep learning. arXiv:1603.07285 (2016)
Sun, C., Demyanov, V., Arnold, D.: Comparison of popular generative adversarial network flavours for Fluvial reservoir modelling. In: 82nd EAGE Annual Conference & Exhibition, vol. 2021, pp 1–5. European Association of Geoscientists & Engineers (2021)
Isola, P., Zhu, J.-Y., Zhou, T., Efros, A.A.: Image-to-image translation with conditional adversarial networks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp 1125–1134 (2017)
Miyato, T., Kataoka, T., Koyama, M., Yoshida, Y.: Spectral normalization for generative adversarial networks. arXiv:1802.05957 (2018)
Lim, J.H., Ye, J.C.: Geometric GAN. arXiv:1705.02894 (2017)
Thanh-Tung, H., Tran, T., Venkatesh, S.: Improving generalization and stability of generative adversarial networks. arXiv:1902.03984 (2019)
Heller, P.L., Paola, C.: Downstream changes in alluvial architecture; an exploration of controls on channel-stacking patterns. J. Sediment. Res. 66(2), 297–306 (1996)
Slingerland, R., Smith, N.D.: River avulsions and their deposits. Annu. Rev. Earth Planet. Sci. 32, 257–285 (2004)
Willems, C.J., Nick, H.M., Donselaar, M.E., Weltje, G.J., Bruhn, D.F.: On the connectivity anisotropy in Fluvial hot sedimentary aquifers and its influence on geothermal doublet performance. Geothermics 65, 222–233 (2017)
Loshchilov, I., Hutter, F.: Decoupled weight decay regularization. arXiv:1711.05101(2017)
King, P.: The connectivity and conductivity of overlapping sand bodies. In: North Sea Oil and Gas reservoirs—II, pp 353–362. Springer (1990)
Renard, P., Allard, D.: Connectivity metrics for subsurface flow and transport. Adv. Water Resour. 51, 168–196 (2013)
Pirot, G., Joshi, R., Giraud, J., Lindsay, M.D., Jessell, M.W.: loopui-0.1: indicators to support needs and practices in 3D geological modelling uncertainty quantification. Geosci. Model Dev. 15(12), 4689–4708 (2022)
McInnes, L., Healy, J., Melville, J.: Umap: Uniform manifold approximation and projection for dimension reduction. arXiv:1802.03426 (2018)
McKie, T., Audretsch, P.: Depositional and structural controls on triassic reservoir performance in the Heron Cluster, ETAP, Central North Sea. In: Geological Society, London, Petroleum Geology Conference Series, vol. 6, pp 285–297. Geological Society of London (2005)
Caers, J.: Modeling Uncertainty in the Earth Sciences. Wiley (2011)
Li, C, Wand, M: Precomputed real-time texture synthesis with markovian generative adversarial networks. In: European Conference on Computer Vision, pp 702–716. Springer (2016)
Sun, C, Demyanov, V, Arnold, D.: GAN River-I: A process-based low NTG meandering reservoir model dataset for machine learning studies. Data in Brief 46, 108785 (2023). Elsevier
Caers, J: Modeling Uncertainty in the Earth Sciences. Wiley (2011)
The work contained in this paper contains work conducted during a PhD study undertaken as part of the Natural Environment Research Council (NERC) Centre for Doctoral Training (CDT) in Oil & Gas [grant number NEM00578X/1]. This project is sponsored by Heriot-Watt University via their James Watt Scholarship Scheme whose support is gratefully acknowledged. We would like to thank MINES ParisTech for their software FLUMYTM, Process-based channelized reservoir models. CopyrightⒸMINES PARIS-PSL / ARMINES. Free download from http://cg.ensmp.fr/flumy. We would like to thank Dr Andy Gardiner for his help on sedimentary geology, thank Prof. Ahmed ElSheikh for his suggestions on result evaluation, and thank Dr Zeyun Jiang, Dr Guillaume Rongier and two anonymous reviewers for their reviews that helped improve the paper.
The work contained in this paper contains work conducted during a PhD study undertaken as part of the Natural Environment Research Council (NERC) Centre for Doctoral Training (CDT) in Oil & Gas [grant number NEM00578X/1]. This project is sponsored by Heriot-Watt University via their James Watt Scholarship Scheme.
Conflict of Interests
The authors have no conflicts of interest to declare that are relevant to this article.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Vasily Demyanov and Daniel Arnold contributed equally to this work.
Appendix A: Details of Fluvial GAN model
1.1 A.1 Generator
1.2 A.2 Hybrid-discriminator
1.3 A.3 Training
We use a benchmark GPU, RTX3090, to test our Fluvial GAN model’s training and generation speed. Training 200 epochs take around 23 hours for a single RTX3090 and the pre-trained Fluvial GAN can simulate about 16000 realisations per minute. Figure 18 shows the loss curves during training.
Appendix B: Details of making training dataset
About this article
Cite this article
Sun, C., Demyanov, V. & Arnold, D. Geological realism in Fluvial facies modelling with GAN under variable depositional conditions. Comput Geosci 27, 203–221 (2023). https://doi.org/10.1007/s10596-023-10190-w