Evolution of ensemble machine learning approaches in water resources management: a review
Top Cited Papers
- 22 May 2025
- journal article
- review article
- Published by Springer Nature in Earth Science Informatics
Abstract
Ensemble Machine Learning (EML) techniques have markedly advanced in hydrological modeling over the past decade, significantly enhancing predictive accuracy for complex tasks such as streamflow prediction, flood forecasting, and groundwater estimation. This study presents a bibliometric analysis of 199 articles and a systematic review of 51 peer-reviewed articles, published from 2017 to December 2024. The bibliometric findings indicate a surge in EML adoption post-2018, driven by increasing demands for precise hydrological models amidst climate change and hydrological variability. The review categorizes EML strategies into Boosting, Bagging, and Stacking methodologies. Boosting methods—particularly Gradient Boosting Machines (GBM), Extreme Gradient Boosting (XGBoost), and LightGBM—were prominently featured in the reviewed studies, often noted for their capability in improved generalization and non-linearity handling. Stacking models, which integrate algorithms like Random Forest (RF), Artificial Neural Networks (ANN), and Long Short-Term Memory (LSTM) networks, were also frequently applied and reported as effective in managing the complexities of hydrological systems. Bagging methods, notably RF, were found effective in stabilizing performance in noisy datasets. Additionally, this study highlights the critical role of data preprocessing techniques such as Singular Spectrum Analysis (SSA), Recursive Feature Elimination (RFE), and Principal Component Analysis (PCA) in optimizing model performance and addressing data heterogeneity challenges. Approximately 60% of the systematic review studies focus on streamflow prediction; among these, hybrid approaches—both combining multiple ML algorithms (ML-ML) and coupling ML with optimization/statistical methods (ML-AO)—were frequently utilized and reported as well-suited for capturing non-linear hydrological dynamics. Despite significant advancements, model interpretability, scalability, and multi-source data integration challenges persist. This paper underscores the need for further research to enhance EML frameworks, mainly through integrating physics-based models and advancements in computational power. The findings provide critical insights into the current state of EML research in hydrology, highlighting essential areas for future exploration amidst growing environmental uncertainties.Keywords
This publication has 133 references indexed in Scilit:
- Urban flood prediction based on PCSWMM and stacking integrated learning modelNatural Hazards, 2024
- Forecasting short- and medium-term streamflow using stacked ensemble models and different meta-learnersStochastic Environmental Research and Risk Assessment, 2024
- Effectiveness of Integrating Ensemble-Based Feature Selection and Novel Gradient Boosted Trees in Runoff Prediction: A Case Study in Vu Gia Thu Bon River Basin, VietnamPure and Applied Geophysics, 2024
- Improved monthly streamflow prediction using integrated multivariate adaptive regression spline with K-means clustering: implementation of reanalyzed remote sensing dataStochastic Environmental Research and Risk Assessment, 2024
- Improved prediction of monthly streamflow in a mountainous region by Metaheuristic-Enhanced deep learning and machine learning models using hydroclimatic dataTheoretical and Applied Climatology, 2023
- Improvement of flood susceptibility mapping by introducing hybrid ensemble learning algorithms and high-resolution satellite imageriesNatural Hazards, 2023
- River flow rate prediction in the Des Moines watershed (Iowa, USA): a machine learning approachStochastic Environmental Research and Risk Assessment, 2022
- Machine learning for hydrologic sciences: An introductory overviewWIREs Water, 2021
- Flood susceptibility modeling in Teesta River basin, Bangladesh using novel ensembles of bagging algorithmsStochastic Environmental Research and Risk Assessment, 2020
- Super ensemble learning for daily streamflow forecasting: large-scale demonstration and comparison with multiple machine learning algorithmsNeural Computing and Applications, 2020