A Surrogate Modelling Approach Based on Nonlinear Dimension Reduction for Uncertainty Quantification in Groundwater Flow Models

In this paper, we develop a surrogate modelling approach for capturing the output field (e.g. the pressure head) from groundwater flow models involving a stochastic input field (e.g. the hydraulic conductivity). We use a Karhunen–Loève expansion for a log-normally distributed input field and apply manifold learning (local tangent space alignment) to perform Gaussian process Bayesian inference using Hamiltonian Monte Carlo in an abstract feature space, yielding outputs for arbitrary unseen inputs. We also develop a framework for forward uncertainty quantification in such problems, including analytical approximations of the mean of the marginalized distribution (with respect to the inputs). To sample from the distribution, we present Monte Carlo approach. Two examples are presented to demonstrate the accuracy of our approach: a Darcy flow model with contaminant transport in 2-d and a Richards equation model in 3-d.

[1]  Thomas J. Santner,et al.  The Design and Analysis of Computer Experiments , 2003, Springer Series in Statistics.

[2]  Agathe Girard,et al.  Gaussian Processes: Prediction at a Noisy Input and Application to Iterative Multiple-Step Ahead Forecasting of Time-Series , 2003, European Summer School on Multi-AgentControl.

[3]  Emanuele Borgonovo,et al.  Model emulation and moment-independent sensitivity analysis: An application to environmental modelling , 2012, Environ. Model. Softw..

[4]  Xiang Ma,et al.  An adaptive hierarchical sparse grid collocation algorithm for the solution of stochastic differential equations , 2009, J. Comput. Phys..

[5]  Hamed Ketabchi,et al.  Review: Coastal groundwater optimization—advances, challenges, and practical solutions , 2015, Hydrogeology Journal.

[6]  George F. Pinder,et al.  Orthogonal collocation and alternating-direction procedures for unsaturated flow problems , 1987 .

[7]  Fabio Nobile,et al.  A Sparse Grid Stochastic Collocation Method for Partial Differential Equations with Random Input Data , 2008, SIAM J. Numer. Anal..

[8]  J. Gómez-Hernández,et al.  Uncertainty assessment and data worth in groundwater flow and mass transport modeling using a blocking Markov chain Monte Carlo method. , 2009 .

[9]  Claire Welty,et al.  Revisiting the Cape Cod bacteria injection experiment using a stochastic modeling approach. , 2005, Environmental science & technology.

[10]  K. Y. Foo,et al.  An overview of landfill leachate treatment via activated carbon adsorption process. , 2009, Journal of hazardous materials.

[11]  I. Papaioannou,et al.  Numerical methods for the discretization of random fields by means of the Karhunen–Loève expansion , 2014 .

[12]  R. Martin,et al.  One-dimensional solute transport in stratified sands at short travel distances. , 2000, Journal of hazardous materials.

[13]  Harold J. Kushner,et al.  Stochastic processes in information and dynamical systems , 1972 .

[14]  Bithin Datta,et al.  Optimal Management of Coastal Aquifers Using Linked Simulation Optimization Approach , 2005 .

[15]  D. Higdon,et al.  Computer Model Calibration Using High-Dimensional Output , 2008 .

[16]  A.A. Shah,et al.  Manifold learning for the emulation of spatial fields from computational models , 2016, J. Comput. Phys..

[17]  Prasanth B. Nair,et al.  Reduced dimensional Gaussian process emulators of parametrized partial differential equations based on Isomap , 2015, Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[18]  John Doherty,et al.  Predictive uncertainty analysis of a saltwater intrusion model using null‐space Monte Carlo , 2011 .

[19]  D. Xiu.  Efficient collocational approach for parametric uncertainty analysis , 2007 .

[20]  Hongyuan Zha,et al.  Adaptive Manifold Learning , 2004, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[21]  N Mylopoulos,et al.  Stochastic analysis and prioritization of the influence of parameter uncertainty on the predicted pressure profile in heterogeneous, unsaturated soils. , 2006, Journal of hazardous materials.

[22]  Richard C. Peralta,et al.  Optimal design of aquifer cleanup systems under uncertainty using a neural network and a genetic algorithm , 1999 .

[23]  Oskar von Stryk,et al.  A mixed-integer simulation-based optimization approach with surrogate functions in water resources management , 2008 .

[24]  George E. Karniadakis,et al.  A sharp error estimate for the fast Gauss transform , 2006, J. Comput. Phys..

[25]  E. G. Vomvoris,et al.  Stochastic analysis of the concentration variability in a three‐dimensional heterogeneous aquifer , 1990 .

[26]  G. Karatzas.  Developments on Modeling of Groundwater Flow and Contaminant Transport , 2017, Water Resources Management.

[27]  Jianwei Yin,et al.  Incremental Manifold Learning Via Tangent Space Alignment , 2006, ANNPR.

[28]  Henning Prommer,et al.  Modelling the fate of oxidisable organic contaminants in groundwater. In C. T. Miller, M. B. Parlange, and S. M. Hassanizadeh (editors), , 2002 .

[29]  B. Ataie‐Ashtiani,et al.  Mathematical Forms and Numerical Schemes for the Solution of Unsaturated Flow Equations , 2012 .

[30]  A. O'Hagan,et al.  Curve Fitting and Optimal Design for Prediction , 1978 .

[31]  T. Harter,et al.  Parallel simulation of groundwater non-point source pollution using algebraic multigrid preconditioners , 2014, Computational Geosciences.

[32]  Jianping Yin,et al.  Robust local tangent space alignment via iterative weighted PCA , 2011, Neurocomputing.

[33]  张振跃,et al.  Principal Manifolds and Nonlinear Dimensionality Reduction via Tangent Space Alignment , 2004 .

[34]  Dongxiao Zhang,et al.  An efficient, high-order perturbation approach for flow in random porous media via Karhunen-Loève and polynomial expansions , 2004 .

[35]  Hongyu Li,et al.  Supervised Learning on Local Tangent Space , 2005, ISNN.

[36]  C. Currin,et al.  A Bayesian Approach to the Design and Analysis of Computer Experiments , 1988 .

[37]  David Cohn,et al.  Active Learning , 2010, Encyclopedia of Machine Learning.

[38]  Xiang Ma,et al.  Kernel principal component analysis for stochastic input model generation , 2010, J. Comput. Phys..

[39]  George Kourakos,et al.  Pumping optimization of coastal aquifers based on evolutionary algorithms and surrogate modular neural network models , 2009 .

[40]  D. Božić,et al.  Adsorption of heavy metal ions by sawdust of deciduous trees. , 2009, Journal of hazardous materials.

[41]  Bryan A. Tolson,et al.  Review of surrogate modeling in water resources , 2012 .

[42]  Bithin Datta,et al.  Coupled simulation‐optimization model for coastal aquifer management using genetic programming‐based ensemble surrogate models and multiple‐realization optimization , 2011 .

[43]  Qingsheng Zhu,et al.  Adaptive Neighborhood Graph for LTSA Learning Algorithm without FreeParameter , 2011 .

[44]  Dongbin Xiu,et al.  High-Order Collocation Methods for Differential Equations with Random Inputs , 2005, SIAM J. Sci. Comput..

[45]  Jia Wei,et al.  Adaptive neighborhood selection for manifold learning , 2008, 2008 International Conference on Machine Learning and Cybernetics.

[46]  Domenico Baù,et al.  Stochastic management of pump-and-treat strategies using surrogate functions , 2006 .

[47]  J. Simonoff.  Smoothing Methods in Statistics , 1998 .

[48]  Bithin Datta,et al.  Comparative Evaluation of Genetic Programming and Neural Network as Potential Surrogate Models for Coastal Aquifer Management , 2011 .

[49]  Mohammad Rajabi,et al.  Optimal Management of a Freshwater Lens in a Small Island Using Surrogate Models and Evolutionary Algorithms , 2014 .

[50]  A. O'Hagan,et al.  Bayesian emulation of complex multi-output and dynamic computer models , 2010 .

[51]  Fabio Nobile,et al.  A Stochastic Collocation Method for Elliptic Partial Differential Equations with Random Input Data , 2007, SIAM Rev..

[52]  Mohammad Rajabi,et al.  Using Surrogate Models and Evolutionary Algorithms , 2014 .

[53]  Lynn W. Gelhar,et al.  Stochastic subsurface hydrology from theory to applications , 1986 .

[54]  Dongbin Xiu,et al.  The Wiener-Askey Polynomial Chaos for Stochastic Differential Equations , 2002, SIAM J. Sci. Comput..

[55]  C. Axness,et al.  Three‐dimensional stochastic analysis of macrodispersion in aquifers , 1983 .

[56]  P. Møldrup,et al.  Variability of soil potential for biodegradation of petroleum hydrocarbons in a heterogeneous subsurface. , 2010, Journal of hazardous materials.

[57]  B. Mohanty,et al.  A new convergence criterion for the modified Picard iteration method to solve the variably saturated flow equation , 1996 .

[58]  Mohammad Rajabi,et al.  Polynomial chaos expansions for uncertainty propagation and moment independent sensitivity analysis of seawater intrusion simulations , 2015 .

[59]  M. Celia,et al.  A General Mass-Conservative Numerical Solution for the Unsaturated Flow Equation , 1990 .

[60]  Randel Haverkamp,et al.  A Comparison of Numerical Simulation Models For One-Dimensional Infiltration1 , 1977 .

[61]  Bithin Datta,et al.  Stochastic and Robust Multi-Objective Optimal Management of Pumping from Coastal Aquifers Under Parameter Uncertainty , 2014, Water Resources Management.

[62]  Sonja Kuhnt,et al.  Design and analysis of computer experiments , 2010 .

[63]  Linda M. Abriola,et al.  Mass conservative numerical solutions of the head‐based Richards equation , 1994 .

[64]  T. Harter,et al.  A groundwater nonpoint source pollution modeling framework to evaluate long‐term dynamics of pollutant exceedance probabilities in wells and other discharge locations , 2012 .

[65]  R. Ghanem,et al.  Stochastic Finite Elements: A Spectral Approach , 1990 .

[66]  Wolfgang Kinzelbach,et al.  Simulation of reactive processes related to biodegradation in aquifers: 1. Structure of the three-dimensional reactive transport model , 1998 .

[67]  J. Feyen,et al.  Modelling Water Flow and Solute Transport in Heterogeneous Soils: A Review of Recent Approaches☆ , 1998 .