Abstract
Parameter estimation with uncertainty quantification is essential in groundwater modeling to ensure model quality; however, parameter estimation, especially for non-Gaussian distributed parameters in highly heterogeneous aquifers, is still a great challenge. The ensemble smoother with multiple data assimilation (ES-MDA) is one of the most popular and effective ensemble-based data assimilation algorithms. However, it only works for multi-Gaussian fields, since two-point statistics are used to estimate the co-relation between parameters and state variables. The probability conditioning method (PCM) has the capability to integrate nonlinear flow data into facies simulation, but it has an assumption of homogeneity within each facies. Full characterization of facies and estimates of hydraulic conductivity within each facies are equally important. This work firstly modifies the original PCM, introducing a new probability assignment method, to consider within-facies heterogeneities, and then it is further combined with the ES-MDA to estimate non-Gaussian distributed hydraulic parameters in a groundwater model. The proposed method is evaluated using a two-facies case and a three-facies case in groundwater modeling. Both cases demonstrate that the modified PCM is effective for facies delineation, especially to identify high heterogeneities in each facies, as well as non-Gaussian characteristics with good connectivity within certain facies. The results also show that the performances of data reproduction and model prediction are of high accuracy and low uncertainty, which is attributed to the accurate characterization of the non-Gaussian parameters in the heterogeneous aquifers used.
Résumé
L’évaluation de paramètres avec quantification d’incertitude est essentielle en modélisation des eaux souterraines pour s’assurer de la qualité du modèle. Cependant, l’estimation des paramètres, en particulier pour les paramètres distribués non gaussiens dans les aquifères très hétérogènes, reste un défi majeur. L’ensemble plus fluide avec assimilation de données multiples (ES-MDA est l’un des algorithmes d’assimilation de données basés sur des ensembles les plus populaires et les plus efficaces. Cependant, cela ne fonctionne que pour les champs multi-gaussiens, car les statistiques à deux points sont utilisées pour estimer la co-relation entre les paramètres et les variables d’état. La méthode des probabilités conditionnelles (MPC) a la capacité d’intégrer des données d’écoulement non linéaires dans la simulation de faciès, mais avec une hypothèse d’homogénéité à l’intérieur de chaque faciès. La caractérisation complète des faciès et les estimations de la conductivité hydraulique à l’intérieur de chaque faciès sont également importantes. Ce travail modifie d’abord la MPC originale, en introduisant une nouvelle méthode d’attribution de probabilité, pour prendre en compte les hétérogénéités intra-faciès, puis il est ensuite combiné avec l’ES-MDA pour estimer les paramètres hydrauliques distribués non gaussiens dans un modèle d’eau souterraine. La méthode proposée est évaluée en utilisant un cas à deux faciès et un cas à trois faciès dans la modélisation des eaux souterraines. Les deux cas démontrent que la MPC modifiée est efficace pour la délimitation des faciès, en particulier pour identifier des hétérogénéités élevées dans chaque faciès, ainsi que des caractéristiques non gaussiennes avec une bonne connectivité dans certains faciès. Les résultats montrent également que les performances de reproduction des données et de prédiction des modèles sont d’une grande précision et d’une faible incertitude, ce qui est attribué à la caractérisation précise des paramètres non gaussiens dans les aquifères hétérogènes utilisés.
Resumen
La estimación de parámetros con cuantificación de la incertidumbre es esencial en la modelización de aguas subterráneas para garantizar la calidad del modelo. Sin embargo, la estimación de parámetros, especialmente para los parámetros distribuidos no gaussianos en acuíferos muy heterogéneos, sigue siendo un gran desafío. El ensamble con una asimilación de datos múltiples (ES-MDA) es uno de los algoritmos basados en conjuntos más comunes y eficaces. Sin embargo, sólo funciona para campos multi-Gaussianos, ya que se utilizan estadísticas de dos puntos para estimar la correlación entre los parámetros y las variables de estado. El método de evaluación de la probabilidad (PCM) tiene la capacidad de integrar datos de flujo no lineal en la simulación de facies, pero tiene un supuesto de homogeneidad dentro de cada facie. La caracterización completa de las facies y las estimaciones de la conductividad hidráulica dentro de cada una de ellas son igualmente importantes. Este trabajo modifica en primer lugar el PCM original, introduciendo un nuevo método de asignación de probabilidades, para considerar las heterogeneidades dentro de las facies, y luego se combina con el ES-MDA para estimar los parámetros hidráulicos distribuidos no gaussianos en un modelo de aguas subterráneas. El método propuesto se evalúa utilizando un caso de dos facies y un caso de tres facies en la modelización de aguas subterráneas. Ambos casos demuestran que el PCM modificado es eficaz para la definición de facies, especialmente para identificar altas heterogeneidades en cada una de ellas, así como características no gaussianas con buena conectividad dentro de ciertas facies. Los resultados también muestran que los resultados de la reproducción de datos y la predicción del modelo son de gran exactitud y baja incertidumbre, lo que se atribuye a la caracterización precisa de los parámetros no gaussianos en los acuíferos heterogéneos utilizados.
摘要
不确定性量化参数估计对于地下水模拟至关重要,可以保障模型的质量。但是,参数估计,尤其是高度非均质含水层中非高斯分布参数的估计,仍然是一个巨大的挑战。多源数据同化的集合平滑器(ES-MDA)是当前最流行和最有效的集合数据同化算法之一。但是,该算法仅适用于多元高斯场,它采用两点统计估计参数和状态变量之间的相互关系。概率条件法(PCM)可以将非线性的地下水流数据集成到岩相模拟中,但它假设每个相内均质。完整刻画岩相分布和估计每个相中的水力传导系数同样重要。本研究首先修正了原始的PCM,引入了一种新的概率分配方法,以考虑岩相内的非均质性,然后将其与ES-MDA进一步结合以估计地下水流模型中的非高斯分布水力参数。在地下水模拟中,采用两相和三相算例对所提出的新方法进行了评估。两个算例结果均表明,经修正的PCM可有效地描述岩相分布,特别是识别每个相中的高度非均质性,以及在某些岩相中具有良好连通性的非高斯特征。研究结果还表明,本方法重现观测数据和模型预测的性能具有较高的准确性和较低的不确定性,这归因于准确刻画了非均质含水层的非高斯参数。
Resumo
A estimativa de parâmetros com quantificação de incerteza é essencial na modelagem de águas subterrâneas para garantir a qualidade do modelo. No entanto, a estimativa de parâmetros, especialmente para parâmetros com distribuição não-gaussiana em aquíferos altamente heterogêneos, ainda é um grande desafio. O método ensemble smoother with multiple data assimilation (ES-MDA) é um dos mais populares e eficazes algoritmos de assimilação de dados baseados em conjunto. No entanto, ele só funciona para campos multi-Gaussianos, uma vez que estatísticas de dois pontos são usadas para estimar a correlação entre parâmetros e variáveis de estado. O método de condicionamento de probabilidade (MCP) tem a capacidade de integrar dados de fluxo não linear na simulação de fácies, mas pressupõe homogeneidade dentro de cada fácies. A caracterização completa da fácies e as estimativas da condutividade hidráulica dentro de cada fácies são igualmente importantes. Primeiramente, este trabalho modifica o MCP original, introduzindo um novo método de atribuição de probabilidade, para considerar heterogeneidades dentro da fácies, e então é ainda combinado com o ES-MDA para estimar parâmetros hidráulicos de distribuição não-gaussiana em um modelo de água subterrânea. O método proposto é avaliado usando um caso de duas fácies e um de três fácies na modelagem de águas subterrâneas. Ambos os casos demonstram que o MCP modificado é eficaz para delineamento de fácies, especialmente para identificar altas heterogeneidades em cada fácies, bem como características não gaussianas com boa conectividade dentro de certas fácies. Os resultados também mostram que os desempenhos de reprodução dos dados e do modelo de previsão são de alta exatidão e baixa incerteza, o que é atribuído à caracterização precisa dos parâmetros não gaussianos nos aquíferos heterogêneos utilizados.
Similar content being viewed by others
References
Bear J (1972) Dynamics of fluids in porous materials. Dover, New York
Caers J, Hoffman T (2006) The probability perturbation method: a new look at Bayesian inverse modeling. Math Geol 38:81–100
Canchumuni SA, Emerick AA, Pacheco MA (2017) Integration of ensemble data assimilation and deep learning for history matching facies models. OTC Brasil, Offshore Technology Conference, Rio de Janeiro, 29–31 October 2019
Cao Z, Li L, Chen K (2018) Bridging iterative ensemble smoother and multiple-point geostatistics for better flow and transport modeling. J Hydrol 565:411–421
Carrera J, Alcolea A, Medina A, Hidalgo J, Slooten LJ (2005) Inverse problem in hydrogeology. Hydrogeol J 13(1):206–222
Chang H, Zhang D, Lu Z (2010) History matching of facies distribution with the EnKF and level set parameterization. J Comput Phys 229:8011–8030. https://doi.org/10.1016/j.jcp.2010.07.005
Chen C, Gao G, Honorio J, Gelderblom P, Jaakkola T (2015) Integration of principal-component-analysis and streamline information for the history matching of channelized reserviors. J Pet Technol 1(4):138–141
Chen C, Gao G, Gelderblom P, Jimenez E (2016) Integration of cumulative-distribution-function mapping with principal-component analysis for the history matching of channelized reservoirs. SPE Reserv Eval Eng 19(02):278–293
Chen Y, Oliver DS (2011) Ensemble randomized maximum likelihood method as an iterative ensemble smoother. Math Geosci 44:1–26. https://doi.org/10.1007/s11004-011-9376-z
Chen Y, Zhang D (2006) Data assimilation for transient flow in geologic formations via ensemble Kalman filter. Adv Water Resour 29:1107–1122. https://doi.org/10.1016/j.advwatres.2005.09.007
Chen Y, Oliver DS, Zhang D (2009) Data assimilation for nonlinear problems by ensemble Kalman filter with reparameterization. J Pet Sci Eng 66:1–14. https://doi.org/10.1016/j.petrol.2008.12.002
Dagan G (1985) Stochastic modeling of groundwater flow by unconditional and conditional probabilities: the inverse problem. Water Resour Res 21(1):65–72
De Marsily WG, Delhomme J-P, Delay F, Buoro A (1999) 40 years of inverse problems in hydrogeology. C R Acad Sci Serie Ii Fascicule A 329(2):73-87
Doherty J (2004) PEST: model-independent parameter estimation. User’s manual, 5th edn. Watermark, Brisbane, Australia
Dorn O, Villegas R (2008) History matching of petroleum reservoirs using a level set technique. Inverse Problems. https://doi.org/10.1088/0266-5611/24/3/035015
Emerick AA (2016) Analysis of the performance of ensemble-based assimilation of production and seismic data. J Pet Sci Eng 139:219–239. https://doi.org/10.1016/j.petrol.2016.01.029
Emerick AA, Reynolds AC (2012) History matching time-lapse seismic data using the ensemble Kalman filter with multiple data assimilations. Comput Geosci 16:639–659
Emerick AA, Reynolds AC (2013a) Ensemble smoother with multiple data assimilation. Comput Geosci-Uk 55:3–15. https://doi.org/10.1016/j.cageo.2012.03.011
Emerick AA, Reynolds AC (2013b) Investigation of the sampling performance of ensemble-based methods with a simple reservoir model. Comput Geosci 17:325
Evensen G (2009) Data assimilation: the ensemble Kalman filter. Springer, Berlin
Evensen G (2018) Analysis of iterative ensemble smoothers for solving inverse problems. Comput Geosci. https://doi.org/10.1007/s10596-018-9731-y
Feyen L, Caers J (2005) Multiple-point geostatistics: a powerful tool to improve groundwater flow and transport predictions in multi-modal formations. Geostatistics for Environmental Applications. Springer, Heidelberg, Germany, pp 197–208
Franssen HJH, Alcolea A, Riva M, Bakr M, Wiel NVD, Stauffer F, Guadagnini A (2009) A comparison of seven methods for the inverse modelling of groundwater flow: application to the characterisation of well catchments. Adv Water Resour 32(6):851–872
Gómez-Hernández JJ, Journel AG (1993) Joint sequential simulation of MultiGaussian fields. Geostatistics Troia’92. Springer, Dordrecht, The Netherlands, pp 85–94. https://doi.org/10.1007/978-94-011-1739-5_8
Gómez-Hernández JJ, Hendricks Franssen HJ, Sahuquillo A (2003) Stochastic conditional inverse modeling of subsurface mass transport: a brief review and the self-calibrating method. Stoch Env Res Risk A 17(5):319–328
Gómez-Hernández JJ, Wen XH (1998) To be or not to be multi-Gaussian? A reflection on stochastic hydrogeology. Adv Water Resour 21(1):47–61
Hansen TM, Mosegaard K, Cordua KS (2018) Multiple point statistical simulation using uncertain (soft) conditional data. Comput Geosci 114:1–10
Harbaugh AW, Banta ER, Hill MC, McDonald MG (2000) MODFLOW-2000, the U.S. Geological survey modular ground-water model: user guide to modularization concepts and the ground-water flow process. US Geol Surv Open-File Rep 00-92:121
Jafarpour B, Khodabakhshi M (2011) A probability conditioning method (PCM) for nonlinear flow data integration into multipoint statistical facies simulation. Math Geosci 43:133–164. https://doi.org/10.1007/s11004-011-9316-y
Jafarpour B, McLaughlin DB (2008) History matching with an ensemble Kalman filter and discrete cosine parameterization. Comput Geosci 12:227–244
Journel AG (2002) Combining knowledge from diverse sources: an alternative to traditional data independence hypotheses. Math Geol 34:573–596
Kalman RE (1960) A new approach to linear filtering and prediction problems. Trans ASME J Basic Eng 82(D):35–45
Karahan H, Ayvaz MT (2008) Simultaneous parameter identification of a heterogeneous aquifer system using artificial neural networks. Hydrogeol J 16(5):817–827
Khaninezhad R, Golmohammadi A, Jafarpour B (2019) A pattern-matching method for flow model calibration under training image constraint. Comput Geosci. https://doi.org/10.1016/J.ADVWATRES.2016.04.007
Khodabakhshi M, Jafarpour B (2013) A Bayesian mixture-modeling approach for flow-conditioned multiple-point statistical facies simulation from uncertain training images. Water Resour Res 49(1):328–342
Khodabakhshi M, Jafarpour B (2014) Adaptive conditioning of multiple-point statistical facies simulation to flow data with probability maps. Math Geosci 46(5):573–595
Laloy E, Hérault R, Jacques D, Linde N (2018) Training-image based geostatistical inversion using a spatial generative adversarial neural network. Water Resour Res 54(1):381–406
Lan T, Shi X, Jiang B, Sun Y, Wu J (2018) Joint inversion of physical and geochemical parameters in groundwater models by sequential ensemble-based optimal design. Stoch Env Res Risk A 32:1919–1937. https://doi.org/10.1007/s00477-018-1521-5
Lee SY, Carle SF, Fogg GE (2007) Geologic heterogeneity and a comparison of two geostatistical models: sequential Gaussian and transition probability-based geostatistical simulation. Adv Water Resour 30(9):1914–1932
Li L, Zhou H, Gómez-Hernández JJ, Hendricks Franssen HJ (2012a) Jointly mapping hydraulic conductivity and porosity by assimilating concentration data via ensemble Kalman filter. J Hydrol 428–429:152–169. https://doi.org/10.1016/j.jhydrol.2012.01.037
Li L, Zhou H, Hendricks Franssen HJ, Gómez-Hernández JJ (2012b) Groundwater flow inverse modeling in non-MultiGaussian media: performance assessment of the normal-score ensemble Kalman filter. Hydrol Earth Syst Sci 16:573–590. https://doi.org/10.5194/hess-16-573-2012
Li L, Srinivasan S, Zhou H, Gómez-Hernández JJ (2013) Simultaneous estimation of geologic and reservoir state variables within an ensemble-based multiple-point statistic framework. Math Geosci 46(5):597–623
Li L, Stetler L, Cao Z, Davis A (2018) An iterative normal-score ensemble smoother for dealing with non-Gaussianity in data assimilation. J Hydrol 567:759–766
Liu N, Oliver DS (2005) Ensemble Kalman filter for automatic history matching of geologic facies. J Pet Sci Eng 47:147–161. https://doi.org/10.1016/j.petrol.2005.03.006
Ma W, Jafarpour B (2018a) Pilot points method for conditioning multiple-point statistical facies simulation on flow data. Adv Water Resour 115:219–233
Ma W, Jafarpour B (2018b) An improved probability conditioning method for constraining multiple-point statistical facies simulation on nonlinear flow data. Society of Petroleum Engineers. https://doi.org/10.2118/190077-ms
Ma W, Jafarpour B (2019) Production data integration into complex geologic facies models: exploiting the behavior of multiple-point statistical simulation for effective data conditioning. SPE Reservoir Simulation Conference, Galveston, TX, April 2019
Man J, Zhang J, Li W, Zeng L, Wu L (2016) Sequential ensemble-based optimal design for parameter estimation. Water Resour Res 52:7577–7592. https://doi.org/10.1002/2016wr018736
Mariethoz G, Renard P, Straubhaar J (2010) The direct sampling method to perform multiple-point geostatistical simulations. Water Resour Res 46(11):1–14
Mo S, Zabaras N, Shi X, Wu J (2020) Integration of adversarial autoencoders with residual dense convolutional networks for estimation of non-Gaussian hydraulic conductivities. Water Resour Res 56(2):e2019WR026082
Nan T, Wu J (2011) Groundwater parameter estimation using the ensemble Kalman filter with localization. Hydrogeol J 19(3):547–561
Neuman SP (1973) Calibration of distributed parameter groundwater flow models viewed as a multiple objective decision process under uncertainty. Water Resour Res 9(4):1006–1021
Oliver DS, Chen Y (2008) Improved initial sampling for the ensemble Kalman filter. Comput Geosci 13:13–27. https://doi.org/10.1007/s10596-008-9101-2
Oliver DS, Chen Y (2011) Recent progress on reservoir history matching: a review. Comput Geosci 15(1):185–221
Oliver DS, Cunha LB, Reynolds AC (1997) Markov chain Monte Carlo methods for conditioning a permeability field to pressure data. Math Geol 29(1):61–91
Sarma P, Durlofsky LJ, Aziz K (2008) Kernel principal component analysis for efficient, differentiable parameterization of multipoint geostatistics. Math Geosci 40(1):3–32
Strebelle S (2002) Conditional simulation of complex geological structures using multiple-point statistics. Math Geol 34:1–21
Van Leeuwen PJ, Evensen G (1996) Data assimilation and inverse methods in terms of a probabilistic formulation. Mon Weather Rev 124:2898–2913
Xu T, Gómez-Hernández JJ (2016) Characterization of non-Gaussian conductivities and porosities with hydraulic heads, solute concentrations, and water temperatures. Water Resour Res. https://doi.org/10.1002/2016wr019011
Xu T, Gómez-Hernández JJ (2017) Simultaneous identification of a contaminant source and hydraulic conductivity via the restart normal-score ensemble Kalman filter. Adv Water Resour. https://doi.org/10.1016/j.advwatres.2017.12.011
Xue L, Zhang D (2014) A multimodel data assimilation framework via the ensemble Kalman filter. Water Resour Res 50:4197–4219. https://doi.org/10.1002/2013wr014525
Zhang Y, Green CT, Fogg GE (2013) The impact of medium architecture of alluvial settings on non-Fickian transport. Adv Water Resour 54:78–99
Zheng C (2006) MT3DMS v5.2 supplemental user’s guide: technical report to the U.S. Department of Geological Sciences, University of Alabama. Army Engineer Research and Development Center, Vicksburg, MS, 24 pp
Zhou HY, Gomez-Hernandez JJ, Franssen HJH, Li LP (2011) An approach to handling non-Gaussianity of parameters and state variables in ensemble Kalman filtering. Adv Water Resour 34:844–864. https://doi.org/10.1016/j.advwatres.2011.04.014
Zhou HY, Li LP, Gómez-Hernández JJ (2012a) Characterizing curvilinear features using the localized normal-score ensemble Kalman filter. Abstr Appl Anal. https://doi.org/10.1007/s11004-020-09882-1
Zhou HY, Gómez-Hernández JJ, Li LP (2012b) A pattern-search-based inverse method. Water Resour Res 48(3):1–17
Zhou HY, Gómez-Hernández JJ, Li LP (2014) Inverse methods in hydrogeology: evolution and recent trends. Adv Water Resour 63:22–37
Zinn B, Harvey CF (2003) When good statistical models of aquifer heterogeneity go bad: a comparison of flow, dispersion, and mass transfer in connected and multivariate Gaussian hydraulic conductivity fields. Water Resour Res. https://doi.org/10.1029/2001wr001146
Acknowledgements
The authors thank the two anonymous reviewers and the associate editor Dr. Yong Zhang for their constructive comments, which significantly improved the quality of this paper.
Funding
This work was financially supported by the National Key Research and Development Program of China (No. 2018YFC0406402) and the National Natural Science Foundation of China (No. 41672229 and No. 41730856).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
In case homo and case heter, the flow is assumed to be transient in a 2D confined aquifer with a starting head of 0 m. As shown in 1, the dimension of the aquifer is 600 m × 600 m and the grid size is 10 m both in horizontal x and y directions. The height of the model is 10 m. The upward and downward boundaries are assumed to be impermeable, the head of the left boundary is fixed to be 0 m, and the flux at the right boundary is set as −300 m3/day. Additionally, the porosity and the specific storage are set to 0.3 and 0.0003 m−1 respectively
Meanwhile, there is a line source at the left boundary with a constant concentration of 100 mg/L. It is of interested to note that only advection and dispersion are considered in these two comparing cases. The longitudinal dispersivity and horizontal transverse dispersivity are set to be 10 and 1 m, respectively. The governing equation for aqueous species’ transportation is defined as (Zheng 2006):
where Cn is the aqueous concentration of the nth component [M L−3]; t is the time [T]; D is the diffusion coefficient [L2 T−1]; v = (−K ∇ H)/θ[L2 T−1]; qs is the volumetric flow rate per unit volume of the aquifer [T−1]; θ is the effective porosity; and Cns is the concentration of the source or sink flux of the nth component [M L−3]. The numerical code MT3DMS (Zheng 2006) is used to solve the solute model. The total simulation time is 500 days with 100 time steps.
Rights and permissions
About this article
Cite this article
Lan, T., Shi, X., Chen, Y. et al. Identification of non-Gaussian parameters in heterogeneous aquifers by a modified probability conditioning method through hydraulic-head assimilation. Hydrogeol J 29, 819–839 (2021). https://doi.org/10.1007/s10040-020-02243-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10040-020-02243-6