Groundwater contamination source estimation based on a refined particle filter associated with a deep residual neural network surrogate

Pan, Zidong; Lu, Wenxi; Bai, Yukun

doi:10.1007/s10040-022-02454-z

Cited by 17 publications

(4 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To handle non‐linear and non‐Gaussian observation/system models, Markov chain Monte Carlo (MCMC) or particle filter (PF) methods can be used as the suitable DA methods to approximate the posterior, even when its exact form is unknown (Moradkhani et al., 2005; Shi et al., 2023; Vrugt, 2016). However, MCMC and PF can become computationally expensive when dealing with complex problems due to the curse of dimensionality, despite recent advances in improving their simulation efficiency (Pan et al., 2022; Pulido & van Leeuwen, 2019; Reuschen et al., 2021; J. Zhang, Vrugt, et al., 2020). Nevertheless, high‐dimensional DA problems can be efficiently implemented if the models are approximately Gaussian.…”

Section: Introductionmentioning

confidence: 99%

A Novel Deep Learning Approach for Data Assimilation of Complex Hydrological Systems

Zhang,

Cao,

Nan

et al. 2024

Water Resources Research

View full text Add to dashboard Cite

In hydrological research, data assimilation (DA) is widely used to fuse the information contained in process‐based models and observational data to reduce simulation uncertainty. However, many popular DA methods are limited by low computational efficiency or their reliance on the Gaussian assumption. To address these limitations, we propose a novel DA method called DA(DL), which leverages the capabilities of DL to model non‐linear relationships and recognize complex patterns. DA(DL) first generates a large volume of training data from the prior ensemble, and then trains a DL model to update the system knowledge (e.g., model parameters in this study) from multiple predictors. For highly non‐linear models, an iterative form of DA(DL) can be implemented. Additionally, strategies of data augmentation and local updating are proposed to enhance DA(DL) for problems involving small ensemble size and the equifinality issue, respectively. In two hydrological DA cases involving Gaussian and non‐Gaussian distributions, DA(DL) shows promising performance compared to two ensemble smoother methods, that is, ES(K) and ES(DL), which respectively apply the Kalman‐ and DL‐based updates. Potential improvements to DA(DL) can be made by designing better DL model architectures, imposing physical constraints to the training of the DL model, and further updating other important variables like model states, forcings and error terms.

show abstract

Section: Introductionmentioning

confidence: 99%

A Novel Deep Learning Approach for Data Assimilation of Complex Hydrological Systems

Zhang,

Cao,

Nan

et al. 2024

Water Resources Research

View full text Add to dashboard Cite

show abstract

“…Ayaz (2021) proposed an ANN to estimate the release history of groundwater pollution source without information about the starting time of the release. Recently, Pan et al (2022) proposed a deep residual neural network as a forward surrogate model combined with an ensemble smoother particle filter in order to estimate the groundwater contamination source together with the aquifer hydraulic conductivity. This work proposes an application of ANN to contaminant source reconstruction with the objective of minimizing the training period and the information required in the inverse procedure.…”

Section: Introductionmentioning

confidence: 99%

Contaminant source identification in groundwater by means of artificial neural network

Secci¹,

Molino²,

Zanini³

2022

Journal of Hydrology

View full text Add to dashboard Cite

“…However, solving the optimization model requires multiple calls to the mathematical simulation model, which will cause a huge computational load. Therefore, a surrogate model strategy is employed to approximate the input-output mapping relationship of the solute transport model (Pan Z D et al, 2022a). Surrogate model techniques fall into three main categories: data-driven, projection, and hierarchicalbased approaches.…”

Section: Introductionmentioning

confidence: 99%

Groundwater contamination source identification based on Sobol Sequences-based Sparrow Search Algorithm with a BiLSTM surrogate model

Pan

2022

Preprint

Self Cite

View full text Add to dashboard Cite

In the traditional linked simulation-optimization method, solving the optimization model requires massive invoking of the groundwater numerical simulation model, which causes a huge computational load. In the present study, a surrogate model of the origin simulation model was developed using a Bidirectional Long and Short-term Memory neural network method (BiLSTM). Compared with the surrogate models built by shallow learning methods (BP neural network) and traditional LSTM methods, the surrogate model built by BiLSTM has higher accuracy and better generalization performance while reducing the computational load. The BiLSTM surrogate model was linked to the optimization model and solved using the Sparrow Search Algorithm based on Sobol sequences (SSAS). SSAS enhances the diversity of the initial population of sparrows by introducing Sobol sequences and introduces nonlinear inertia weights to control the search range and search efficiency. Compared with SSA, SSAS has stronger global search ability and faster search efficiency. And SSAS identifies the contamination source location and release intensity stably and reliably. This study also applied the Cholesky decomposition method to establish a Gaussian field for hydraulic conductivity to evaluate the feasibility of the simulation-optimization method.

show abstract

Groundwater contamination source estimation based on a refined particle filter associated with a deep residual neural network surrogate

Cited by 17 publications

References 46 publications

A Novel Deep Learning Approach for Data Assimilation of Complex Hydrological Systems

A Novel Deep Learning Approach for Data Assimilation of Complex Hydrological Systems

Contaminant source identification in groundwater by means of artificial neural network

Groundwater contamination source identification based on Sobol Sequences-based Sparrow Search Algorithm with a BiLSTM surrogate model

Contact Info

Product

Resources

About