Predicting bacterial transport through saturated porous media using an automated machine learning model

Aug 4, 2026

Upload by Finn

Escherichia coli can migrate from manure-amended soil to groundwater during rainfall or irrigation, so accurate prediction of vertical transport under saturated conditions supports mitigation of microbiological contamination risk. This study compiles 377 datasets from 61 publications and trains six machine-learning models to predict first-order attachment coefficient and spatial removal rate from eight environmental and medium variables. Higher retention scenarios show improved performance, and Gradient Boosting and Extreme Gradient Boosting outperform other algorithms. Key variable importance is dominated by pore water velocity, ionic strength, median grain size, and column length.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Introduction

Background and problem significance

Existing experimental and modeling approaches

Methods

Dataset compilation

Input and target variables

Machine learning model training

Results

Model performance and retention scenarios

Algorithm comparison

Feature importance

Discussion

Environmental risk evaluation value

Generalization to other contaminants

TYPE Original Research PUBLISHED 10 May 2023 DOI 10.3389/fmicb.2023.1152059 OPEN ACCESS EDITED BY Muhammad Shaaban, Bahauddin Zakariya University, Pakistan REVIEWED BY Alex Furman, Technion Israel Institute of Technology, Israel Murray Close, Institute of Environmental Science and Research (ESR), New Zealand *CORRESPONDENCE Xijuan Chen [chenxj@iae.ac.cn](chenxj@iae.ac.cn) RECEIVED 27 January 2023 ACCEPTED 25 April 2023 PUBLISHED 10 May 2023 CITATION Chen F, Zhou B, Yang L, Chen X and Zhuang J (2023) Predicting bacterial transport through saturated porous media using an automated machine learning model. Front. Microbiol. 14:1152059. doi: 10.3389/fmicb.2023.1152059 COPYRIGHT © 2023 Chen, Zhou, Yang, Chen and Zhuang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Predicting bacterial transport through saturated porous media using an automated machine learning model Fengxian Chen 1, Bin Zhou 2, Liqiong Yang 1, Xijuan Chen 1* and Jie Zhuang3 1 Key Laboratory of Pollution Ecology and Environmental Engineering, Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang, Liaoning, China, 2 Faculty of Medicine, University of Augsburg, Augsburg, Germany, 3 Department of Biosystems Engineering and Soil Science, Center for Environmental Biotechnology, The University of Tennessee, Knoxville, TN, United States Escherichia coli, as an indicator of fecal contamination, can move from manureamended soil to groundwater under rainfall or irrigation events. Predicting its vertical transport in the subsurface is essential for the development of engineering solutions to reduce the risk of microbiological contamination. In this study, we collected 377 datasets from 61 published papers addressing E. coli transport through saturated porous media and trained six types of machine learning algorithms to predict bacterial transport. Eight variables, including bacterial concentration, porous medium type, median grain size, ionic strength, pore water velocity, column length, saturated hydraulic conductivity, and organic matter content were used as input variables while the first-order attachment coefficient and spatial removal rate were set as target variables. The eight input variables have low correlations with the target variables, namely, they cannot predict target variables independently. However, using the predictive models, input variables can effectively predict the target variables. For scenarios with higher bacterial retention, such as smaller median grain size, the predictive models showed better performance. Among six types of machine learning algorithms, Gradient Boosting Machine and Extreme Gradient Boosting outperformed other algorithms. In most predictive models, pore water velocity, ionic strength, median grain size, and column length showed higher importance than other input variables. This study provided a valuable tool to evaluate the transport risk of E.coli in the subsurface under saturated water flow conditions. It also proved the feasibility of data-driven methods that could be used for predicting other contaminants’ transport in the environment. KEYWORDS bacterial transport, automated machine learning, first-order attachment coefficient, spatial removal rate, machine learning Highlights • The predictive models showed better performance when bacterial retention was high. • Spatial removal rate is a better target variable than first-order attachment coefficient. • Algorithms based on gradient boosting outperformed other machine learning algorithms. Frontiers in Microb

FAQs

What endpoints does the automated machine learning model predict for E. coli transport?

It predicts the first-order attachment coefficient and the spatial removal rate as target variables.

Which input variables are used to drive the predictions?

Eight variables are used, including bacterial concentration, porous medium type, median grain size, ionic strength, pore water velocity, column length, saturated hydraulic conductivity, and organic matter content.

Which machine learning algorithms performed best and what variables were most important?

Gradient Boosting Machine and Extreme Gradient Boosting outperformed other algorithms. Most models found pore water velocity, ionic strength, median grain size, and column length to have higher importance.

Aug 4, 2026

Upload by Finn