Abstract
Minimization of a stochastic cost function is commonly used for approximate sampling in high-dimensional Bayesian inverse problems with Gaussian prior distributions and multimodal posterior distributions. The density of the samples generated by minimization is not the desired target density, unless the observation operator is linear, but the distribution of samples is useful as a proposal density for importance sampling or for Markov chain Monte Carlo methods. In this paper, we focus on applications to sampling from multimodal posterior distributions in high dimensions. We first show that sampling from multimodal distributions is improved by computing all critical points instead of only minimizers of the objective function. For applications to high-dimensional geoscience inverse problems, we demonstrate an efficient approximate weighting that uses a low-rank Gauss-Newton approximation of the determinant of the Jacobian. The method is applied to two toy problems with known posterior distributions and a Darcy flow problem with multiple modes in the posterior.
Article PDF
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Data Availability
No data are used in the manuscript.
Code Availability
Selected Python codes used in the preparation of this manuscript are available at https://bitbucket.org/JanadeWiljes/workspace/projects/RBPS.
References
Abhyankar, S., Brown, J., Constantinescu, E.M., Ghosh, D., Smith, B.F., Zhang, H.: PETSc/TS: A modern scalable ODE/DAE solver library. arXiv:1806.01437 (2018)
Armstrong, M., Galli, A., Beucher, H., Le Lonh, G., Renard, D., Doligez, B., Eschard, R., Geffroy, F.: Plurigaussian Simulations in Geosciences, 2nd edn. Springer, Berlin (2011)
Bardsley, J., Solonen, A., Haario, H., Laine, M.: Randomize-then-optimize: A method for sampling from posterior distributions in nonlinear inverse problems. SIAM J. Sci. Comput. 36(4), A1895–A1910 (2014)
Bardsley, J.M., Cui, T., Marzouk, Y.M., Wang, Z.: Scalable optimization-based sampling on function space. SIAM J. Sci. Comput. 42(2), A1317–A1347 (2020). https://doi.org/10.1137/19m1245220
Bengtsson, T., Bickel, P., Li, B.: Curse-Of-Dimensionality Revisited: Collapse of the Particle Filter in Very Large Scale Systems. In: Probability and Statistics: Essays in Honor of David A. Freedman, pp 316–334. Institute of Mathematical Statistics (2008)
Beskos, A., Crisan, D., Jasra, A.: On the stability of sequential Monte Carlo methods in high dimensions. Ann. Appl. Probab. 24(4), 1396–1445 (2014)
Bjarkason, E.K., O’Sullivan, J.P., Yeh, A., O’Sullivan, M.J.: Inverse modeling of the natural state of geothermal reservoirs using adjoint and direct methods. Geothermics 78, 85–100 (2019). https://doi.org/10.1016/j.geothermics.2018.10.001
Bui-Thanh, T., Ghattas, O., Martin, J., Stadler, G.: A computational framework for infinite-dimensional bayesian inverse problems part i: The linearized case, with application to global seismic inversion. SIAM J. Sci. Comput. 35(6), A2494–A2523 (2013)
Cardiff, M., Barrash, W., Kitanidis, P.K.: A field proof-of-concept of aquifer imaging using 3-D transient hydraulic tomography with modular, temporarily-emplaced equipment. Water Resour. Res. 48(5), W05531 (2012). https://doi.org/10.1029/2011WR011704
Chen, Y., Oliver, D.S.: Ensemble randomized maximum likelihood method as an iterative ensemble smoother. Math. Geosci. 44(1), 1–26 (2012). https://doi.org/10.1007/s11004-011-9376-z
Cui, T., Law, K.J.H., Marzouk, Y.M.: Dimension-independent likelihood-informed MCMC. J. Comput. Phys. 304, 109–137 (2016). https://doi.org/10.1016/j.jcp.2015.10.008
Duncan, A.B., Lelièvre, T., Pavliotis, G.A.: Variance reduction using nonreversible Langevin samplers. J. Stat. Phys. 163(3), 457–491 (2016)
Efendiev, Y., Hou, T., Luo, W.: Preconditioning Markov chain Monte Carlo simulations using coarse-scale models. SIAM J. Sci. Comput. 28(2), 776–803 (2006)
Eydinov, D., Gao, G., Li, G., Reynolds, A.C.: Simultaneous estimation of relative permeability and porosity/permeability fields by history matching production data. J. Can Petrol Technol. 48(12), 13–25 (2009)
Freeze, R.A.: A stochastic-conceptual analysis of one-dimensional groundwater flow in a non-uniform, homogeneous media. Water Resour. Res. 11(5), 725–741 (1975)
Gao, G., Zafari, M., Reynolds, A.C.: Quantifying uncertainty for the PUNQ-s3 problem in a Bayesian setting with RML and enKF. SPE J. 11(4), 506–515 (2006)
Gelman, A.: Inference and Monitoring Convergence. In: Gilks, WR, Richardson, S, Spiegelhalter, DJ (eds.) Markov Chain Monte Carlo in Practice, chap. 8, pp 131–144. Chapman & Hall (1996)
Haario, H., Saksman, E., Tamminen, J.: Adaptive proposal distribution for random walk Metropolis algorithm. Comput. Stat. 14(3), 375–396 (1999)
Haario, H., Saksman, E., Tamminen, J.: An adaptive Metropolis algorithm. Bernoulli 7 (2), 223–242 (2001)
Jardak, M., Talagrand, O.: Ensemble variational assimilation as a probabilistic estimator – part 2: The fully non-linear case. Nonlinear Processes Geophys. 25(3), 589–604 (2018)
Kitanidis, P.K.: Quasi-linear geostatistical theory for inversing. Water Resour. Res. 31(10), 2411–2419 (1995)
Langtangen, H.P., Logg, A.: Solving PDEs in Python: The FEniCS Tutorial I, vol. 1. Springer, Berlin (2016)
Law, K.J.H., Stuart, A.M.: Evaluating data assimilation algorithms. Mon. Weather. Rev. 140 (11), 3757–3782 (2012). https://doi.org/10.1175/MWR-D-11-00257.1
van Leeuwen, P.J., Künsch, H.R., Nerger, L., Potthast, R., Reich, S.: Particle filters for high-dimensional geoscience applications: a review. Q. J. Roy. Meteorol. Soc. 145(723), 2335–2365 (2019). https://doi.org/10.1002/qj.3551
Liu, N., Oliver, D.S.: Evaluation of Monte Carlo methods for assessing uncertainty. SPE J. 8(2), 188–195 (2003). https://doi.org/10.2118/84936-PA
Liu, N., Oliver, D.S.: Automatic history matching of geologic facies. SPE J. 9(4), 188–195 (2004). https://doi.org/10.2118/84594-PA
Logg, A., Mardal, K.A., Wells, G.: Automated solution of differential equations by the finite element method: The FEniCS book, vol. 84. Springer Science & Business Media, Berlin (2012)
MacKay, D.J.C.: Information theory, inference and learning algorithms. Cambridge University Press, Cambridge (2003)
Martin, J., Wilcox, L., Burstedde, C., Ghattas, O.: A stochastic Newton MCMC method for large-scale statistical inverse problems with application to seismic inversion. SIAM J. Sci. Comput. 34 (3), A1460–A1487 (2012)
Martino, L., Elvira, V., Luengo, D., Corander, J.: MCMC-Driven Adaptive Multiple Importance Sampling. In: Interdisciplinary Bayesian Statistics, pp 97–109. Springer, Berlin (2015)
Morzfeld, M., Hodyss, D.: Gaussian approximations in filters and smoothers for data assimilation. Tellus A: Dynamic Meteorology and Oceanography 71(1), 1–27 (2019). https://doi.org/10.1080/16000870.2019.1600344
Oliver, D.S.: Metropolized randomized maximum likelihood for improved sampling from multimodal distributions. SIAM/ASA J. Uncertain. Quantif. 5(1), 259–277 (2017). https://doi.org/10.1137/15M1033320
Oliver, D.S., Chen, Y.: Recent progress on reservoir history matching: a review. Comput. Geosci. 15(1), 185–221 (2011). https://doi.org/10.1007/s10596-010-9194-2
Oliver, D.S., Chen, Y.: Data assimilation in truncated plurigaussian models: impact of the truncation map. Math. Geosci. 50(8), 867–893 (2018). https://doi.org/10.1007/s11004-018-9753-y
Oliver, D.S., Cunha, L.B., Reynolds, A.C.: Markov chain Monte Carlo methods for conditioning a permeability field to pressure data. Math. Geol. 29(1), 61–91 (1997). https://doi.org/10.1007/BF02769620
Oliver, D.S., He, N., Reynolds, A.C.: Conditioning permeability fields to pressure data. In: Proceedings of the European Conference on the Mathematics of Oil Recovery, V, pp. 1–11 (1996)
Roberts, G.O., Rosenthal, J.S.: Examples of adaptive MCMC. J. Comput. Graph. Stat. 18 (2), 349–367 (2009)
Sakov, P., Oliver, D.S., Bertino, L.: An iterative enKF for strongly nonlinear systems. Mon. Weather Rev. 140(6), 1988–2004 (2012)
Tarantola, A.: Inverse problem theory and methods for model parameter estimation society for industrial and applied mathematics (2005)
Tavassoli, Z., Carter, J.N., King, P.R.: An analysis of history matching errors. Comput. Geosci. 9(2), 99–123 (2005)
Villa, U., Petra, N., Ghattas, O.: hIPPYlib: An extensible software framework for large-scale inverse problems. J. Open Source Softw. 3(30), 940 (2018)
Villa, U., Petra, N., Ghattas, O.: HIPPYLib: An extensible software framework for large-scale inverse problems governed by PDEs: Part i: Deterministic inversion and linearized Bayesian inference. ACM Trans. Math. Softw 47(2). https://doi.org/10.1145/3428447 (2021)
Vrugt, J.A., Nualláin, B. O.́, Robinson, B.A., Bouten, W., Dekker, S.C., Sloot, P.M.: Application of parallel computing to stochastic parameter estimation in environmental models. Comput. Geosci. 32 (8), 1139–1155 (2006)
Wang, K., Bui-Thanh, T., Ghattas, O.: A randomized maximum a posteriori method for posterior sampling of high dimensional nonlinear Bayesian inverse problems. SIAM J. Sci. Comput. 40(1), A142–A171 (2018)
White, J.T.: A model-independent iterative ensemble smoother for efficient history-matching and uncertainty quantification in very high dimensions. Environmental Modelling & Software (2018)
Zhang, F., Reynolds, A.C., Oliver, D.S.: The impact of upscaling errors on conditioning a stochastic channel to pressure data. SPE J. 8(1), 13–21 (2003)
Zhang, H., Constantinescu, E.M., Smith, B.F.: PETSc TSAdjoint: A discrete adjoint ODE solver for first-order and second-order sensitivity analysis. arXiv:1912.07696 (2019)
Funding
Open Access funding provided by NORCE Norwegian Research Centre AS. For this work, Yuming Ba was supported by the China Scholarship Council. Dean Oliver was supported by the NORCE Norwegian Research Centre cooperative research project “Assimilating 4D Seismic Data: Big Data Into Big Models” which is funded by industry partners Aker BP, Equinor, Lundin Norway, Repsol, and Total, as well as the Research Council of Norway through the Petromaks2 program. Jana de Wiljes and Sebastian Reich have been partially funded by Deutsche Forschungsgemeinschaft (DFG) - Project-ID 318763901 - SFB1294.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of Interests
The authors declare that there are no conflicts of interest.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Ba, Y., de Wiljes, J., Oliver, D.S. et al. Randomized maximum likelihood based posterior sampling. Comput Geosci 26, 217–239 (2022). https://doi.org/10.1007/s10596-021-10100-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10596-021-10100-y