Data-driven physics informed deep learning of solute transport with anomalous diffusion

TL;DR

This paper introduces a deep learning approach using feedforward neural networks to efficiently infer parameters of fractional advection-dispersion equations, improving modeling of anomalous diffusion in porous media.

Contribution

It develops a novel data-driven deep learning algorithm for parameter estimation in fractional PDEs related to solute transport, reducing reliance on extensive experiments.

Findings

Successfully inferred model parameters from synthetic data.

Demonstrated robustness on field data.

Effective in modeling anomalous diffusion phenomena.

Abstract

The fractional advection-dispersion equation (FADE) has attracted increased attention from researchers as it provides an accurate description for challenging phenomenas with long-range time memory and spatial interactions, such as the anomalous diffusion behavior in the solute transport in porous media. Practically, a full characterization of the model parameters, such as the fluid velocity, dispersion coefficient and the order of the fractional derivative, often implies a huge amount of experiments and measurements and thus are hard to be determined. In this paper, we employ the framework of feedforward deep neural networks (DNNs) to develop an efficient data-driven deep learning algorithm for inferring the aforementioned parameters of the FADE, such as the time-dependent space-fractional advection-dispersion equation (sFADE) and the variable-order fractional mobile/immobile equation (VoFMIE), in which the feedforward DNNs are trained to minimize the mean square error loss function formulated by means of the finite difference approximations of sFADE and VoFMIE, respectively. Several numerical experiments, in which we discover the model parameters by the feedforward DNNs for both the synthetic and field data, are presented to demonstrate the effectiveness and robustness of the proposed data-driven deep learning algorithm.