DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Hydrogeological Model Selection Among Complex Spatial Priors

Abstract

Hydrogeological field studies rely often on a single conceptual representation of the subsurface. This is problematic since the impact of a poorly chosen conceptual model on predictions might be significantly larger than the one caused by parameter uncertainty. Furthermore, conceptual models often need to incorporate geological concepts and patterns in order to provide meaningful uncertainty quantification and predictions. Consequently, several geologically realistic conceptual models should ideally be considered and evaluated in terms of their relative merits. Here, we propose a full Bayesian methodology based on Markov chain Monte Carlo to enable model selection among 2-D conceptual models that are sampled using training images and concepts from multiple-point statistics. More precisely, power posteriors for the different conceptual subsurface models are sampled using sequential geostatistical resampling and Graph Cuts. To demonstrate the methodology, we compare and rank five alternative conceptual geological models that have been proposed in the literature to describe aquifer heterogeneity at the MAcroDispersion Experiment site in Mississippi, USA. We consider a small-scale tracer test for which the spatial distribution of hydraulic conductivity impacts multilevel solute concentration data observed along a 2-D transect. The thermodynamic integration and the stepping-stone sampling methods were used to compute the evidence and associated Bayesmore » factors using the computed power posteriors. We find that both methods are compatible with multiple-point statistics-based inversions and provide a consistent ranking of the competing conceptual models considered.« less

Authors:
ORCiD logo [1]; ORCiD logo [2]; ORCiD logo [1]; ORCiD logo [1]
  1. Univ. of Lausanne, Lausanne (Switzerland). Applied and Environmental Geophysics Group, Inst. of Earth Sciences
  2. British Geological Survey, Environmental Science Centre, Nottingham (United Kingdom)
Publication Date:
Research Org.:
Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)
Sponsoring Org.:
USDOE Office of Science (SC)
OSTI Identifier:
1572041
Grant/Contract Number:  
AC02-05CH11231
Resource Type:
Accepted Manuscript
Journal Name:
Water Resources Research
Additional Journal Information:
Journal Volume: 55; Journal Issue: 8; Journal ID: ISSN 0043-1397
Publisher:
American Geophysical Union (AGU)
Country of Publication:
United States
Language:
English
Subject:
58 GEOSCIENCES

Citation Formats

Brunetti, C., Bianchi, M., Pirot, G., and Linde, N. Hydrogeological Model Selection Among Complex Spatial Priors. United States: N. p., 2019. Web. doi:10.1029/2019wr024840.
Brunetti, C., Bianchi, M., Pirot, G., & Linde, N. Hydrogeological Model Selection Among Complex Spatial Priors. United States. https://doi.org/10.1029/2019wr024840
Brunetti, C., Bianchi, M., Pirot, G., and Linde, N. Fri . "Hydrogeological Model Selection Among Complex Spatial Priors". United States. https://doi.org/10.1029/2019wr024840. https://www.osti.gov/servlets/purl/1572041.
@article{osti_1572041,
title = {Hydrogeological Model Selection Among Complex Spatial Priors},
author = {Brunetti, C. and Bianchi, M. and Pirot, G. and Linde, N.},
abstractNote = {Hydrogeological field studies rely often on a single conceptual representation of the subsurface. This is problematic since the impact of a poorly chosen conceptual model on predictions might be significantly larger than the one caused by parameter uncertainty. Furthermore, conceptual models often need to incorporate geological concepts and patterns in order to provide meaningful uncertainty quantification and predictions. Consequently, several geologically realistic conceptual models should ideally be considered and evaluated in terms of their relative merits. Here, we propose a full Bayesian methodology based on Markov chain Monte Carlo to enable model selection among 2-D conceptual models that are sampled using training images and concepts from multiple-point statistics. More precisely, power posteriors for the different conceptual subsurface models are sampled using sequential geostatistical resampling and Graph Cuts. To demonstrate the methodology, we compare and rank five alternative conceptual geological models that have been proposed in the literature to describe aquifer heterogeneity at the MAcroDispersion Experiment site in Mississippi, USA. We consider a small-scale tracer test for which the spatial distribution of hydraulic conductivity impacts multilevel solute concentration data observed along a 2-D transect. The thermodynamic integration and the stepping-stone sampling methods were used to compute the evidence and associated Bayes factors using the computed power posteriors. We find that both methods are compatible with multiple-point statistics-based inversions and provide a consistent ranking of the competing conceptual models considered.},
doi = {10.1029/2019wr024840},
journal = {Water Resources Research},
number = 8,
volume = 55,
place = {United States},
year = {Fri Jul 05 00:00:00 EDT 2019},
month = {Fri Jul 05 00:00:00 EDT 2019}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Citation Metrics:
Cited by: 12 works
Citation information provided by
Web of Science

Figures / Tables:

Algorithm 1 Algorithm 1: MCMC inversion workflow based on MPS and the extended Metropolis algorithm to enable evidence estimation using power posteriors.

Save / Share:

Works referenced in this record:

Some Tests of Significance, Treated by the Theory of Probability
journal, April 1935

  • Jeffreys, Harold
  • Mathematical Proceedings of the Cambridge Philosophical Society, Vol. 31, Issue 2
  • DOI: 10.1017/S030500410001330X

Bayes Factors
journal, June 1995


An experimental comparison of min-cut/max- flow algorithms for energy minimization in vision
journal, September 2004

  • Boykov, Y.; Kolmogorov, V.
  • IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 26, Issue 9
  • DOI: 10.1109/TPAMI.2004.60

Posterior population expansion for solving inverse problems: POSTERIOR POPULATION EXP.
journal, April 2017

  • Jäggli, C.; Straubhaar, J.; Renard, P.
  • Water Resources Research, Vol. 53, Issue 4
  • DOI: 10.1002/2016WR019550

Nested Sampling
conference, January 2004

  • Skilling, John
  • BAYESIAN INFERENCE AND MAXIMUM ENTROPY METHODS IN SCIENCE AND ENGINEERING: 24th International Workshop on Bayesian Inference and Maximum Entropy Methods in Science and Engineering, AIP Conference Proceedings
  • DOI: 10.1063/1.1835238

Inference from Iterative Simulation Using Multiple Sequences
journal, November 1992


Targeting Bayes factors with direct-path non-equilibrium thermodynamic integration
journal, March 2017

  • Grzegorczyk, Marco; Aderhold, Andrej; Husmeier, Dirk
  • Computational Statistics, Vol. 32, Issue 2
  • DOI: 10.1007/s00180-017-0721-7

Interpretative modelling of a geological cross section from boreholes: sources of uncertainty and their quantification
journal, January 2014


Marginal likelihood estimation via power posteriors
journal, July 2008

  • Friel, N.; Pettitt, A. N.
  • Journal of the Royal Statistical Society: Series B (Statistical Methodology), Vol. 70, Issue 3
  • DOI: 10.1111/j.1467-9868.2007.00650.x

Training‐Image Based Geostatistical Inversion Using a Spatial Generative Adversarial Neural Network
journal, January 2018

  • Laloy, Eric; Hérault, Romain; Jacques, Diederik
  • Water Resources Research, Vol. 54, Issue 1
  • DOI: 10.1002/2017WR022148

Simulating normalizing constants: from importance sampling to bridge sampling to path sampling
journal, May 1998


Geological realism in hydrogeological and geophysical inverse modeling: A review
journal, December 2015


On the efficient sampling of pathways in the transition path ensemble
journal, January 2001

  • Vlugt, Thijs J. H.; Smit, Berend
  • PhysChemComm, Vol. 4, Issue 2
  • DOI: 10.1039/b009865p

Influence of conceptual model uncertainty on contaminant transport forecasting in braided river aquifers
journal, December 2015


Bayesian evidence computation for model selection in non-linear geoacoustic inference problems
journal, December 2010

  • Dettmer, Jan; Dosso, Stan E.; Osler, John C.
  • The Journal of the Acoustical Society of America, Vol. 128, Issue 6
  • DOI: 10.1121/1.3506345

Probabilistic inversion with graph cuts: Application to the Boise Hydrogeophysical Research Site: INVERSION WITH GRAPH CUTS AT THE BHRS
journal, February 2017

  • Pirot, Guillaume; Linde, Niklas; Mariethoz, Grégoire
  • Water Resources Research, Vol. 53, Issue 2
  • DOI: 10.1002/2016WR019347

Analysis of Solute Transport in Flow Fields Influenced by Preferential Flowpaths at the Decimeter Scale
journal, March 2003


Three-dimensional high resolution fluvio-glacial aquifer analog: Part 1: Field study
journal, July 2011


Bayesian model selection for complex geological structures using polynomial chaos proxy
journal, March 2017


Three-dimensional high resolution fluvio-glacial aquifer analog – Part 2: Geostatistical modeling
journal, July 2011


Improving power posterior estimation of statistical evidence
journal, May 2013


Tomogram-based comparison of geostatistical models: Application to the Macrodispersion Experiment (MADE) site
journal, December 2015


Summary statistics from training images as prior information in probabilistic inversion
journal, February 2015

  • Lochbühler, Tobias; Vrugt, Jasper A.; Sadegh, Mojtaba
  • Geophysical Journal International, Vol. 201, Issue 1
  • DOI: 10.1093/gji/ggv008

Quantifying Uncertainty in Subsurface Systems
book, May 2018


The Direct Sampling method to perform multiple-point geostatistical simulations: PERFORMING MULTIPLE-POINTS SIMULATIONS
journal, November 2010

  • Mariethoz, Gregoire; Renard, Philippe; Straubhaar, Julien
  • Water Resources Research, Vol. 46, Issue 11
  • DOI: 10.1029/2008WR007621

The Controlled Thermodynamic Integral for Bayesian Model Evidence Evaluation
journal, April 2016

  • Oates, Chris J.; Papamarkou, Theodore; Girolami, Mark
  • Journal of the American Statistical Association, Vol. 111, Issue 514
  • DOI: 10.1080/01621459.2015.1021006

Conditioning of Multiple-Point Statistics Facies Simulations to Tomographic Images
journal, September 2013

  • Lochbühler, Tobias; Pirot, Guillaume; Straubhaar, Julien
  • Mathematical Geosciences, Vol. 46, Issue 5
  • DOI: 10.1007/s11004-013-9484-z

Geological modeling of submeter scale heterogeneity and its influence on tracer transport in a fluvial aquifer
journal, October 2010

  • Ronayne, Michael J.; Gorelick, Steven M.; Zheng, Chunmiao
  • Water Resources Research, Vol. 46, Issue 10
  • DOI: 10.1029/2010WR009348

Choosing among Partition Models in Bayesian Phylogenetics
journal, August 2010

  • Fan, Y.; Wu, R.; Chen, M. -H.
  • Molecular Biology and Evolution, Vol. 28, Issue 1
  • DOI: 10.1093/molbev/msq224

Calibration of channelized subsurface flow models using nested sampling and soft probabilities
journal, January 2015


Bayesian inverse problem and optimization with iterative spatial resampling: ITERATIVE SPATIAL RESAMPLING
journal, November 2010

  • Mariethoz, Grégoire; Renard, Philippe; Caers, Jef
  • Water Resources Research, Vol. 46, Issue 11
  • DOI: 10.1029/2010WR009274

Monte Carlo sampling of solutions to inverse problems
journal, July 1995

  • Mosegaard, Klaus; Tarantola, Albert
  • Journal of Geophysical Research: Solid Earth, Vol. 100, Issue B7
  • DOI: 10.1029/94JB03097

Inverse problems with non-trivial priors: efficient solution through sequential Gibbs sampling
journal, January 2012

  • Hansen, Thomas Mejer; Cordua, Knud Skou; Mosegaard, Klaus
  • Computational Geosciences, Vol. 16, Issue 3
  • DOI: 10.1007/s10596-011-9271-1

Can uncertainty in geological cross-section interpretations be quantified and predicted?
journal, April 2018

  • Randle, Charles H.; Bond, Clare E.; Lark, R. Murray
  • Geosphere, Vol. 14, Issue 3
  • DOI: 10.1130/GES01510.1

Dealing with spatial heterogeneity
journal, February 2005


Issues in characterizing heterogeneity and connectivity in non-multiGaussian media
journal, January 2008


To be or not to be multi-Gaussian? A reflection on stochastic hydrogeology
journal, February 1998


Model selection on solid ground: Rigorous comparison of nine ways to evaluate Bayesian model evidence
journal, December 2014

  • Schöniger, Anneli; Wöhling, Thomas; Samaniego, Luis
  • Water Resources Research, Vol. 50, Issue 12
  • DOI: 10.1002/2014WR016062

An adaptive multiscale method for density-driven instabilities
journal, July 2012


Modelling guidelines––terminology and guiding principles
journal, January 2004


Improved Nested Sampling and Surrogate-Enabled Comparison With Other Marginal Likelihood Estimators
journal, February 2018

  • Zeng, Xiankui; Ye, Ming; Wu, Jichun
  • Water Resources Research, Vol. 54, Issue 2
  • DOI: 10.1002/2017WR020782

Investigation of Small-Scale Preferential Flow with a Forced-Gradient Tracer Test
journal, August 2010


Image synthesis with graph cuts: a fast model proposal mechanism in probabilistic inversion
journal, December 2015

  • Zahner, Tobias; Lochbühler, Tobias; Mariethoz, Grégoire
  • Geophysical Journal International, Vol. 204, Issue 2
  • DOI: 10.1093/gji/ggv517

Connectivity metrics for subsurface flow and transport
journal, January 2013


A lithofacies approach for modeling non‐ F ickian solute transport in a heterogeneous alluvial aquifer
journal, January 2016

  • Bianchi, Marco; Zheng, Chunmiao
  • Water Resources Research, Vol. 52, Issue 1
  • DOI: 10.1002/2015WR018186

Bayesian Interpolation
journal, May 1992


Computing Bayes Factors Using Thermodynamic Integration
journal, April 2006


A dual-domain mass transfer approach for modeling solute transport in heterogeneous aquifers: Application to the Macrodispersion Experiment (MADE) site
journal, September 2000

  • Feehley, C. Erin; Zheng, Chunmiao; Molz, Fred J.
  • Water Resources Research, Vol. 36, Issue 9
  • DOI: 10.1029/2000WR900148

Spatial connectivity in a highly heterogeneous aquifer: From cores to preferential flow paths: SPATIAL CONNECTIVITY IN A HETEROGENEOUS AQUIFER
journal, May 2011

  • Bianchi, Marco; Zheng, Chunmiao; Wilson, Crystal
  • Water Resources Research, Vol. 47, Issue 5
  • DOI: 10.1029/2009WR008966

Review of strategies for handling geological uncertainty in groundwater flow and transport modeling
journal, February 2012


Patch-based iterative conditional geostatistical simulation using graph cuts: GEOSTATISTICAL SIMULATION USING GRAPH CUTS
journal, August 2016

  • Li, Xue; Mariethoz, Gregoire; Lu, DeTang
  • Water Resources Research, Vol. 52, Issue 8
  • DOI: 10.1002/2015WR018378

On uncertainty quantification in hydrogeology and hydrogeophysics
journal, December 2017


Make the most of your samples: Bayes factor estimators for high-dimensional models of sequence evolution
journal, March 2013

  • Baele, Guy; Lemey, Philippe; Vansteelandt, Stijn
  • BMC Bioinformatics, Vol. 14, Issue 1
  • DOI: 10.1186/1471-2105-14-85

Reversible jump Markov chain Monte Carlo computation and Bayesian model determination
journal, January 1995


Improving Marginal Likelihood Estimation for Bayesian Phylogenetic Model Selection
journal, December 2010


Multiple-point geostatistics for modeling subsurface heterogeneity: A comprehensive review: REVIEW OF MULTIPLE POI
journal, November 2008


Lessons Learned from 25 Years of Research at the MADE Site
journal, September 2010


Estimating Bayes factors via thermodynamic integration and population MCMC
journal, October 2009


Field study of dispersion in a heterogeneous aquifer: 1. Overview and site description
journal, December 1992

  • Boggs, J. Mark; Young, Steven C.; Beard, Lisa M.
  • Water Resources Research, Vol. 28, Issue 12
  • DOI: 10.1029/92WR01756

The Necessity of a Multiple-Point Prior Model
journal, December 2006


Field study of dispersion in a heterogeneous aquifer: 3. Geostatistical analysis of hydraulic conductivity
journal, December 1992

  • Rehfeldt, Kenneth R.; Boggs, J. Mark; Gelhar, Lynn W.
  • Water Resources Research, Vol. 28, Issue 12
  • DOI: 10.1029/92WR01758

Rate-limited mass transfer or macrodispersion: Which dominates plume evolution at the macrodispersion experiment (MADE) site?
journal, March 2000

  • Harvey, Charles; Gorelick, Steven M.
  • Water Resources Research, Vol. 36, Issue 3
  • DOI: 10.1029/1999WR900247

Heterogeneity in Sedimentary Deposits: A Review of Structure-Imitating, Process-Imitating, and Descriptive Approaches
journal, April 1996

  • Koltermann, Christine E.; Gorelick, Steven M.
  • Water Resources Research, Vol. 32, Issue 9
  • DOI: 10.1029/96WR00025

What do you think this is? “Conceptual uncertainty” in geoscience interpretation
journal, January 2007


Measuring the impact of COVID-19 vaccine misinformation on vaccination intent in the UK and USA
journal, February 2021

  • Loomba, Sahil; de Figueiredo, Alexandre; Piatek, Simon J.
  • Nature Human Behaviour, Vol. 5, Issue 3
  • DOI: 10.1038/s41562-021-01056-1

Abdominal multi-organ auto-segmentation using 3D-patch-based deep convolutional neural network
journal, April 2020


The Controlled Thermodynamic Integral for Bayesian Model Evidence Evaluation
text, January 2016


Geostatistics for environmental applications
journal, August 2013


Bayes factors
journal, November 2006


On uncertainty quantification in hydrogeology and hydrogeophysics
text, January 2017


Monte Carlo methods
null, January 1970


The controlled thermodynamic integral for Bayesian model evidence evaluation
text, January 2015


Parallel power posterior analyses for fast computation of marginal likelihoods in phylogenetics
journal, January 2021

  • Höhna, Sebastian; Landis, Michael J.; Huelsenbeck, John P.
  • PeerJ, Vol. 9
  • DOI: 10.7717/peerj.12438

Improving power posterior estimation of statistical evidence
text, January 2012


Summary statistics from training images as prior information in probabilistic inversion
text, January 2017


Training-image based geostatistical inversion using a spatial generative adversarial neural network
text, January 2017


The Controlled Thermodynamic Integral for Bayesian Model Evidence Evaluation
text, January 2016


Training images for the MADE site
dataset, January 2019