DecTree: a physics-based geochemical surrogate for surface complexation of uranium on clay
GFZ German Research Centre for Geosciences, Telegrafenberg, 14473 Potsdam, Germany
Related authors
Marco De Lucia, Michael Kühn, Alexander Lindemann, Max Lübke, and Bettina Schnor
Geosci. Model Dev., 14, 7391–7409, https://doi.org/10.5194/gmd-14-7391-2021, https://doi.org/10.5194/gmd-14-7391-2021, 2021
Short summary
Short summary
POET is a parallel reactive transport simulator which implements a mechanism to store and reuse previous results of geochemical simulations through distributed hash tables. POET parallelizes chemistry using a master/worker design with noncontiguous grid partitions to maximize its efficiency and load balance on shared-memory machines and compute clusters.
Morgan Tranter, Maria Wetzel, Marco De Lucia, and Michael Kühn
Adv. Geosci., 56, 57–65, https://doi.org/10.5194/adgeo-56-57-2021, https://doi.org/10.5194/adgeo-56-57-2021, 2021
Short summary
Short summary
Barite formation is an important factor for many use cases of the geological subsurface because it may change the rock.
In this modelling study, the replacement reaction of celestite to barite is investigated.
The steps that were identified to play a role are celestite dissolution followed by two-step precipitation of barite: spontaneous formation of small crystals and their subsequent growth.
Explicitly including the processes improve the usability of the models for quantified prediction.
Marco De Lucia and Michael Kühn
Adv. Geosci., 56, 33–43, https://doi.org/10.5194/adgeo-56-33-2021, https://doi.org/10.5194/adgeo-56-33-2021, 2021
Short summary
Short summary
RedModRphree is an R extension package to leverage the PHREEQC engine for geochemical models, providing convenience functions to efficiently setup computations and program algorithms involving geochemical models. Version 0.3.6 ships with a novel implementation of Pourbaix (potential/pH) diagram computation which considers the full speciation of the solution at each diagram point.
Marco De Lucia and Michael Kühn
Geosci. Model Dev., 14, 4713–4730, https://doi.org/10.5194/gmd-14-4713-2021, https://doi.org/10.5194/gmd-14-4713-2021, 2021
Short summary
Short summary
DecTree evaluates a hierarchical coupling method for reactive transport simulations in which pre-trained surrogate models are used to speed up the geochemical subprocess, and equation-based
full-physicssimulations are called only if the surrogate predictions are implausible. Furthermore, we devise and evaluate a decision tree surrogate approach designed to inject domain knowledge of the surrogate by defining engineered features based on law of mass action or stoichiometric reaction equations.
M. De Lucia, T. Kempka, and M. Kühn
Geosci. Model Dev., 8, 279–294, https://doi.org/10.5194/gmd-8-279-2015, https://doi.org/10.5194/gmd-8-279-2015, 2015
Marco De Lucia, Michael Kühn, Alexander Lindemann, Max Lübke, and Bettina Schnor
Geosci. Model Dev., 14, 7391–7409, https://doi.org/10.5194/gmd-14-7391-2021, https://doi.org/10.5194/gmd-14-7391-2021, 2021
Short summary
Short summary
POET is a parallel reactive transport simulator which implements a mechanism to store and reuse previous results of geochemical simulations through distributed hash tables. POET parallelizes chemistry using a master/worker design with noncontiguous grid partitions to maximize its efficiency and load balance on shared-memory machines and compute clusters.
Morgan Tranter, Maria Wetzel, Marco De Lucia, and Michael Kühn
Adv. Geosci., 56, 57–65, https://doi.org/10.5194/adgeo-56-57-2021, https://doi.org/10.5194/adgeo-56-57-2021, 2021
Short summary
Short summary
Barite formation is an important factor for many use cases of the geological subsurface because it may change the rock.
In this modelling study, the replacement reaction of celestite to barite is investigated.
The steps that were identified to play a role are celestite dissolution followed by two-step precipitation of barite: spontaneous formation of small crystals and their subsequent growth.
Explicitly including the processes improve the usability of the models for quantified prediction.
Marco De Lucia and Michael Kühn
Adv. Geosci., 56, 33–43, https://doi.org/10.5194/adgeo-56-33-2021, https://doi.org/10.5194/adgeo-56-33-2021, 2021
Short summary
Short summary
RedModRphree is an R extension package to leverage the PHREEQC engine for geochemical models, providing convenience functions to efficiently setup computations and program algorithms involving geochemical models. Version 0.3.6 ships with a novel implementation of Pourbaix (potential/pH) diagram computation which considers the full speciation of the solution at each diagram point.
Marco De Lucia and Michael Kühn
Geosci. Model Dev., 14, 4713–4730, https://doi.org/10.5194/gmd-14-4713-2021, https://doi.org/10.5194/gmd-14-4713-2021, 2021
Short summary
Short summary
DecTree evaluates a hierarchical coupling method for reactive transport simulations in which pre-trained surrogate models are used to speed up the geochemical subprocess, and equation-based
full-physicssimulations are called only if the surrogate predictions are implausible. Furthermore, we devise and evaluate a decision tree surrogate approach designed to inject domain knowledge of the surrogate by defining engineered features based on law of mass action or stoichiometric reaction equations.
M. De Lucia, T. Kempka, and M. Kühn
Geosci. Model Dev., 8, 279–294, https://doi.org/10.5194/gmd-8-279-2015, https://doi.org/10.5194/gmd-8-279-2015, 2015
Cited articles
Bradbury, M. H. and Baeyens, B.: A mechanistic description of Ni and Zn sorption on Na-montmorillonite Part II: modelling, J. Contam. Hydrol., 27, 223–248, https://doi.org/10.1016/s0169-7722(97)00007-7, 1997. a
Chatterjee, S.: A New Coefficient of Correlation, J. Am. Stat. Assoc., 116, 2009–2022, https://doi.org/10.1080/01621459.2020.1758115, 2020. a, b, c, d
Claret, F., Dauzeres, A., Jacques, D., Sellin, P., Cochepin, B., De Windt, L., Garibay-Rodriguez, J., Govaerts, J., Leupin, O., Mon Lopez, A., Montenegro, L., Montoya, V., Prasianakis, N. I., Samper, J., and Talandier, J.: Modelling of the long-term evolution and performance of engineered barrier system, EPJ Nuclear Sci. Technol., 8, 41, https://doi.org/10.1051/epjn/2022038, 2022. a, b
De Lucia, M. and Kühn, M.: DecTree v1.0 – chemistry speedup in reactive transport simulations: purely data-driven and physics-based surrogates, Geosci. Model Dev., 14, 4713–4730, https://doi.org/10.5194/gmd-14-4713-2021, 2021. a, b, c
De Lucia, M., Kempka, T., Jatnieks, J., and Kühn, M.: Integrating surrogate models into subsurface simulation framework allows computation of complex reactive transport scenarios, Enrgy. Proced., 125, 580–587, https://doi.org/10.1016/j.egypro.2017.08.200, 2017. a
Demirer, E., Coene, E., Iraola, A., Nardi, A., Abarca, E., Idiart, A., de Paola, G., and Rodríguez-Morillas, N.: Improving the Performance of Reactive Transport Simulations Using Artificial Neural Networks, Transport Porous Med., 149, 271–297, https://doi.org/10.1007/s11242-022-01856-7, 2023. a
Griessenberger, F., Trutschnig, W., and Junker, R. R.: qad: An R‐package to detect asymmetric and directed dependence in bivariate samples, Methods Ecol. Evol., 13, 2138–2149, https://doi.org/10.1111/2041-210x.13951, 2022. a
Guérillot, D. and Bruyelle, J.: Geochemical equilibrium determination using an artificial neural network in compositional reservoir flow simulation, Computat. Geosci., 24, 697–707, https://doi.org/10.1007/s10596-019-09861-4, 2020. a
Hjelle, O.: Approximation of Scattered Data with Multilevel B-splines, Tech. rep., SINTEF, https://www.sintef.no/globalassets/upload/ikt/9011/geometri/mba/mba.pdf (last access: 25 June 2024), 2001. a
Hoeffding, W.: A Non-Parametric Test of Independence, Ann. Math. Stat., 19, 546–557, https://doi.org/10.1214/aoms/1177730150, 1948. a
Hu, G. and Pfingsten, W.: Data-driven machine learning for disposal of high-level nuclear waste: A review, Ann. Nucl. Energy, 180, 109452, https://doi.org/10.1016/j.anucene.2022.109452, 2023. a
Jacques, D., Phung, Q. T., Perko, J., Seetharam, S. C., Maes, N., Liu, S., Yu, L., Rogiers, B., and Laloy, E.: Towards a scientific-based assessment of long-term durability and performance of cementitious materials for radioactive waste conditioning and disposal, J. Nucl. Mater., 557, 153201, https://doi.org/10.1016/j.jnucmat.2021.153201, 2021. a
Jatnieks, J., De Lucia, M., Dransch, D., and Sips, M.: Data-driven Surrogate Model Approach for Improving the Performance of Reactive Transport Simulations, Enrgy. Proced., 97, 447–453, https://doi.org/10.1016/j.egypro.2016.10.047, 2016. a
Junker, R. R., Griessenberger, F., and Trutschnig, W.: Estimating scale-invariant directed dependence of bivariate distributions, Comput. Stat. Data An., 153, 107 058, https://doi.org/10.1016/j.csda.2020.107058, 2021. a, b
Kendall, M. G.: A new measure of rank correlation, Biometrika, 30, 81–93, https://doi.org/10.1093/biomet/30.1-2.81, 1938. a
Kolditz, O., Jacques, D., Claret, F., Bertrand, J., Churakov, S. V., Debayle, C., Diaconu, D., Fuzik, K., Garcia, D., Graebling, N., Grambow, B., Holt, E., Idiart, A., Leira, P., Montoya, V., Niederleithinger, E., Olin, M., Pfingsten, W., Prasianakis, N. I., Rink, K., Samper, J., Szöke, I., Szöke, R., Theodon, L., and Wendling, J.: Digitalisation for nuclear waste management: predisposal and disposal, Environ. Earth Sci., 82, 42, https://doi.org/10.1007/s12665-022-10675-4, 2023. a, b
Laloy, E. and Jacques, D.: Emulation of CPU-demanding reactive transport models: a comparison of Gaussian processes, polynomial chaos expansion, and deep neural networks, Computat. Geosci., 23, 1193–1215, https://doi.org/10.1007/s10596-019-09875-y, 2019. a
Laloy, E. and Jacques, D.: Speeding Up Reactive Transport Simulations in Cement Systems by Surrogate Geochemical Modeling: Deep Neural Networks and k-Nearest Neighbors, Transport Porous Med., 143, 433–462, https://doi.org/10.1007/s11242-022-01779-3, 2022. a
Lee, S., Wolberg, G., and Shin, S.: Scattered data interpolation with multilevel B-splines, IEEE T. Vis. Comput. Gr., 3, 228–244, https://doi.org/10.1109/2945.620490, 1997. a
Marques Fernandes, M., Baeyens, B., Dähn, R., Scheinost, A., and Bradbury, M.: U(VI) sorption on montmorillonite in the absence and presence of carbonate: A macroscopic and microscopic study, Geochim. Cosmochim. Ac., 93, 262–277, https://doi.org/10.1016/j.gca.2012.04.017, 2012. a
Meeussen, J. C. L.: ORCHESTRA: An Object-Oriented Framework for Implementing Chemical Equilibrium Models, Environ. Sci. Technol., 37, 1175–1182, https://doi.org/10.1021/es025597s, 2003. a
Park, J.-S. and Oh, S.-J.: A New Concave Hull Algorithm and Concaveness Measure for n-dimensional Datasets, J. Inf. Sci. Eng., 28, 587–600, https://doi.org/10.6688/JISE.2012.28.3.10, 2012. a
Prasianakis, N., Haller, R., Mahrous, M., Poonoosamy, J., Pfingsten, W., and Churakov, S.: Neural network based process coupling and parameter upscaling in reactive transport simulations, Geochim. Cosmochim. Ac., 291, 126–143, https://doi.org/10.1016/j.gca.2020.07.019, 2020. a
Prasianakis, N., Laloy, E., Jacques, D., Meeussen, J., Miron, G., Kulik, D., Idiart, A., Demirer, E., Coene, E., Cochepin, B., Leconte, M., Savino, M., Samper II, J., De Lucia, M., Churakov, S., Kolditz, O., Yang, C., Samper, J., and Claret, F.: Geochemistry and Machine Learning: methods and benchmarking, Environ. Earth Sci., in review, 2024a. a, b, c, d, e, f
Prasianakis, N., et al.: Geochemistry and Machine Learning: Methods and Benchmarking, Zenodo [data set], https://doi.org/10.5281/zenodo.11274790, 2024b. a
Reshef, D. N., Reshef, Y. A., Finucane, H. K., Grossman, S. R., McVean, G., Turnbaugh, P. J., Lander, E. S., Mitzenmacher, M., and Sabeti, P. C.: Detecting Novel Associations in Large Data Sets, Science, 334, 1518–1524, https://doi.org/10.1126/science.1205438, 2011. a, b
Sochala, P., Chiaberge, C., Claret, F., and Tournassat, C.: Dimension reduction for uncertainty propagation and global sensitivity analyses of a cesium adsorption model, J. Comput. Sci., 75, 102197, https://doi.org/10.1016/j.jocs.2023.102197, 2024. a
Székely, G. J., Rizzo, M. L., and Bakirov, N. K.: Measuring and testing dependence by correlation of distances, Ann. Stat., 35, 2769–2794, https://doi.org/10.1214/009053607000000505, 2007. a, b
Turunen, J. and Lipping, T.: Feasibility of neural network metamodels for emulation and sensitivity analysis of radionuclide transport models, Sci. Rep., 13, 6985, https://doi.org/10.1038/s41598-023-34089-9, 2023. a
Short summary
This paper presents a surrogate modelling approach applied to a geochemical system of uranium sorption on clay, relevant for the safety assessment of nuclear waste repositories. The surrogate uses knowledge about the underlying process to recursively partition the dataset into regions of reduced dimensionality. It achieves high accuracy with a competitive prediction throughput, and is a promising method to speedup computationally demanding coupled reactive transport models.
This paper presents a surrogate modelling approach applied to a geochemical system of uranium...