article · Scientific African
Accurate dam spillage forecasting is vital for reducing downstream flood risks and managing water resources, but existing machine learning models frequently miss delayed hydrological responses. To address this, a Lagged-Feedback Random Forest framework was designed to systematically incorporate temporal feedback using a 40-day lag structure. Evaluated on a 15-year daily record from Weija Dam in Ghana, the framework was tested against Long Short-Term Memory networks, standard Random Forests, Artificial Neural Networks, and modern gradient boosting and temporal convolutional models. The lagged model outperformed all competitors, recording an R² of 0.94 and a correlation of 0.97 with minimal systematic error. Models relying solely on current-day data performed poorly regardless of their complexity, demonstrating that access to historical lag data impacted predictive performance far more than model architecture alone.
Predicting dam spillage accurately helps safeguard downstream settlements from severe flooding while supporting sound water resource management. Showing that incorporating historical time delays matters more than using intricate architectures offers clear guidance for engineering practical flood early-warning systems, proving that delayed hydrological dynamics must be captured to make algorithmic predictions operationally useful.
This methodology could enable real-time operational decision-support and early-warning software for dam operators and water resource authorities. The technology sits at an applied and tested stage, having been validated against a 15-year historical dataset at Weija Dam. Moving it towards real-world commercialisation would require embedding the lagged feedback framework into live monitoring infrastructure to guide day-to-day water release schedules.
AI-generated from the published abstract. Always read the original work before citing.
Accurate prediction of dam spillage is essential for mitigating flood risks to downstream communities and for the optimal management of water resources. Existing machine learning models often struggle to capture the delayed hydrological responses that influence reservoir spillage. This limits their operational usefulness for early warning and reservoir management. This study proposes a lag-enhanced Random Forest forecasting framework, termed Lagged-Feedback Random Forest (LFRF), that systematically incorporates temporal feedback from historical data points to improve dam spillage prediction. The framework was benchmarked against three alternative approaches, Long Short-Term Memory (LSTM) networks, standard Random Forest (RF), and Artificial Neural Networks (ANN), as well as four additional contemporary architectures, XGBoost, LightGBM, CatBoost, and a Temporal Convolutional Network (TCN), using the Weija Dam's 15-year daily historical record and a 40-day lag structure. Models were evaluated using RMSE, MAE, and R² (in m³/s, consistent with the observed spillage discharge), alongside the dimensionless Correlation and T-test statistics. LFRF outperformed all other models with an R² of 0.94 and a Correlation of 0.97, with a T-test statistic of 0.078 indicating minimal systematic deviation from observed values; LFRF and LSTM achieved comparable, markedly lower prediction errors than RF and ANN. Standard RF, XGBoost, LightGBM, CatBoost, and ANN were evaluated as naive baselines using only current-day readings, without either the lagged features given to LFRF or the sequence-native architecture used by LSTM and a Temporal Convolutional Network; all five naive models performed similarly poorly regardless of algorithm sophistication. For the Weija Dam dataset and the configurations evaluated, access to lagged temporal information appeared to contribute more strongly to predictive performance than model architecture alone. These findings highlight the value of lagged features in Random Forest ensembles for dam spillage prediction and identify several lag “sweet spots” — 7-8 days, 15-17 days, 25-26 days, and approximately 30-40 days — at which LFRF provided consistently strong predictive performance that could support real-time operational decisions and risk mitigation at the dam.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1016/j.sciaf.2026.e03611
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
