Accelerating uncertainty quantification of groundwater flow modelling using a deep neural network proxy
From MaRDI portal
Publication:2237307
DOI10.1016/J.CMA.2021.113895zbMATH Open1506.76170arXiv2007.00400OpenAlexW3040439833MaRDI QIDQ2237307FDOQ2237307
Authors: Mikkel B. Lykkegaard, Tim Dodwell, David Moxey 
Publication date: 27 October 2021
Published in: Computer Methods in Applied Mechanics and Engineering (Search for Journal in Brave)
Abstract: Quantifying the uncertainty in model parameters and output is a critical component in model-driven decision support systems for groundwater management. This paper presents a novel algorithmic approach which fuses Markov Chain Monte Carlo (MCMC) and Machine Learning methods to accelerate uncertainty quantification for groundwater flow models. We formulate the governing mathematical model as a Bayesian inverse problem, considering model parameters as a random process with an underlying probability distribution. MCMC allows us to sample from this distribution, but it comes with some limitations: it can be prohibitively expensive when dealing with costly likelihood functions, subsequent samples are often highly correlated, and the standard Metropolis-Hastings algorithm suffers from the curse of dimensionality. This paper designs a Metropolis-Hastings proposal which exploits a deep neural network (DNN) approximation of a groundwater flow model, to significantly accelerate MCMC sampling. We modify a delayed acceptance (DA) model hierarchy, whereby proposals are generated by running short subchains using an inexpensive DNN approximation, resulting in a decorrelation of subsequent fine model proposals. Using a simple adaptive error model, we estimate and correct the bias of the DNN approximation with respect to the posterior distribution on-the-fly. The approach is tested on two synthetic examples; a isotropic two-dimensional problem, and an anisotropic three-dimensional problem. The results show that the cost of uncertainty quantification can be reduced by up to 50% compared to single-level MCMC, depending on the precomputation cost and accuracy of the employed DNN.
Full work available at URL: https://arxiv.org/abs/2007.00400
Recommendations
- Data-driven uncertainty quantification for predictive flow and transport modeling using support vector machines
- Scaling Up Bayesian Uncertainty Quantification for Inverse Problems Using Deep Neural Networks
- Uncertainty quantification for porous media flows
- A Hierarchical Multilevel Markov Chain Monte Carlo Algorithm with Applications to Uncertainty Quantification in Subsurface Flow
- Multilevel Markov Chain Monte Carlo
Markov chain Monte Carlouncertainty quantificationsurrogate modelsgroundwater flowdeep neural networks
Artificial neural networks and deep learning (68T07) Probabilistic models, generic numerical methods in probability and statistics (65C20) Flows in porous media; filtration; seepage (76S05)
Cites Work
- Title not available (Why is that?)
- Title not available (Why is that?)
- An adaptive Metropolis algorithm
- A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code
- Monte Carlo sampling methods using Markov chains and their applications
- Equation of state calculations by fast computing machines
- Exploratory designs for computational experiments
- Statistical inverse problems: discretization, model reduction and inverse crimes
- MCMC methods for functions: modifying old algorithms to make them faster
- Dimensionality reduction and polynomial chaos acceleration of Bayesian inference in inverse problems
- Learning about physical parameters: the importance of model discrepancy
- Coarse-gradient Langevin algorithms for dynamic data integration and uncertainty quantification
- Erratum to ``Monte Carlo errors with less errors [Comput. Phys. Comm. 156 (2004) 143-153]
- A Hierarchical Multilevel Markov Chain Monte Carlo Algorithm with Applications to Uncertainty Quantification in Subsurface Flow
- Survey of multifidelity methods in uncertainty propagation, inference, and optimization
- Simulation of normal distributed smooth fields by Karhunen-Loève expansion in combination with kriging
- Random field simulation over curved surfaces: applications to computational structural mechanics
Cited In (11)
- Approximating a finite element model by neural network prediction for facility optimization in groundwater engineering
- Multilevel Delayed Acceptance MCMC
- Efficient uncertainty quantification for dynamic subsurface flow with surrogate by theory-guided neural network
- Use of multifidelity training data and transfer learning for efficient construction of subsurface flow surrogate models
- A deep learning based reduced order modeling for stochastic underground flow problems
- Data-driven uncertainty quantification for predictive flow and transport modeling using support vector machines
- Surrogate optimization of deep neural networks for groundwater predictions
- Dual order-reduced Gaussian process emulators (DORGP) for quantifying high-dimensional uncertain crack growth using limited and noisy data
- Full-field order-reduced Gaussian process emulators for nonlinear probabilistic mechanics
- Propagation of hydropeaking waves in heterogeneous aquifers: effects on flow topology and uncertainty quantification
- Deep capsule encoder–decoder network for surrogate modeling and uncertainty quantification
Uses Software
This page was built for publication: Accelerating uncertainty quantification of groundwater flow modelling using a deep neural network proxy
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2237307)
