{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/towards-a-robust-parameterization-for","title":"Towards a Robust Parameterization for Conditioning Facies Models Using Deep Variational Autoencoders and Ensemble Smoother","arxiv_id":"1812.06900","date":"2018-12-17","proceeding":null,"authors":["Smith W. A. Canchumuni","Alexandre A. Emerick","Marco Aurélio C. Pacheco"],"abstract":"The literature about history matching is vast and despite the impressive\nnumber of methods proposed and the significant progresses reported in the last\ndecade, conditioning reservoir models to dynamic data is still a challenging\ntask. Ensemble-based methods are among the most successful and efficient\ntechniques currently available for history matching. These methods are usually\nable to achieve reasonable data matches, especially if an iterative formulation\nis employed. However, they sometimes fail to preserve the geological realism of\nthe model, which is particularly evident in reservoir with complex facies\ndistributions. This occurs mainly because of the Gaussian assumptions inherent\nin these methods. This fact has encouraged an intense research activity to\ndevelop parameterizations for facies history matching. Despite the large number\nof publications, the development of robust parameterizations for facies remains\nan open problem.\n  Deep learning techniques have been delivering impressive results in a number\nof different areas and the first applications in data assimilation in\ngeoscience have started to appear in literature. The present paper reports the\ncurrent results of our investigations on the use of deep neural networks\ntowards the construction of a continuous parameterization of facies which can\nbe used for data assimilation with ensemble methods. Specifically, we use a\nconvolutional variational autoencoder and the ensemble smoother with multiple\ndata assimilation. We tested the parameterization in three synthetic\nhistory-matching problems with channelized facies. We focus on this type of\nfacies because they are among the most challenging to preserve after the\nassimilation of data. The parameterization showed promising results\noutperforming previous methods and generating well-defined channelized facies.","url_abs":"http://arxiv.org/abs/1812.06900v1","url_pdf":"http://arxiv.org/pdf/1812.06900v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"towards-a-robust-parameterization-for","repo_url":"https://github.com/smith31t/GeoFacies_DL","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}