Piezometric Level Forecasting at the Bouhanifia Dam in Algeria Using Machine Learning: GMDH, LSTM, and XGBoost
- Authors: Slimane Benyahia, Bouchrit Rouissat, Nadia Smail, Fatah Touati
- Citation: Acta hydrotechnica, vol. 39, no. 70, pp. 27-46, 2026. https://doi.org/10.15292/acta.hydro.2026.03
- Abstract: Dam safety monitoring relies on the integrated analysis of multiple interdependent parameters, including seepage rates, piezometric levels, reservoir water elevation, pore pressures, structural displacements, and climatic variables. Interpreting these complex datasets using advanced modeling techniques is essential for assessing structural integrity and anticipating potential risks. This study proposes a machine learning-based approach to predicting piezometric levels, a key indicator in the hydro-mechanical behavior analysis of embankment dams. Three advanced models were developed and compared: the Group Method of Data Handling (GMDH), Long Short-Term Memory recurrent neural networks (LSTM), and eXtreme Gradient Boosting (XGBoost). The models were trained and tested on a multi-year time series (2017–2025) comprising reservoir water levels, climatic variables (precipitation and temperature), and measurements from four piezometers installed within the Bouhanifia concrete-faced rockfill dam (CFRD) in Algeria. The results demonstrated high predictive accuracy across all piezometers, supported by robust performance metrics: root mean square errors (RMSE) ranging from 0.066 m to 0.153 m in validation, mean absolute relative errors (MARE) between 0.099% and 0.235%, correlation coefficients (R) exceeding 0.90 for all models and piezometers, Loague and Green coefficients (LG) ranging from 0.471 to 0.709, and variance accounted for (VAF) between 81.1% and 95%. Comparative analysis highlighted the distinct strengths of each model: XGBoost achieved the highest accuracy (RMSE as low as 0.066 m for P01), LSTM demonstrated remarkable stability across all monitoring points with the highest R (0.975 for P03), and GMDH offered a balance of accuracy and interpretability with reduced computational complexity (<10 seconds).
- Keywords: Machine Learning, Piezometric Levels, Dam Safety Monitoring, GMDH, LSTM, XGBoost, Bouhanifia Dam.
- Full text: a39sb.pdf
- References:
- Anastasakis, L., Mort, N. (2001). The development of self-organization techniques in modelling: a review of the group method of data handling (GMDH). Research Report – University of Sheffield Department of Automatic Control and Systems Engineering. Research Report
- Ali, M., H., Abustan, I. (2014). A new novel index for evaluating model performance. JNRD-Journal of Natural Resources and Development, 4, 1-9. https://doi.org/10.5027/jnrd.v4i0.01
- Agence Nationale des Barrages et Transferts (ANBT): Caractéristiques techniques de barrage Bouhanifia. Technical report, Algeria (2025).
- Amiri-Doumari, S., Karimipour, A., Nayebpour, S., N., Hatamiafkoueieh, J. (2023). Integration of group method of data handling (GMDH) algorithm and population-based metaheuristic algorithms for spatial prediction of potential groundwater. Environmental Earth Sciences, 81(20), 485.https://doi.org/10.1007/s12665-023-10931-1
- Bonkoungou, A., A., Zio, S., Sabane, A., Kafando, R., Kabore, A. K., Bissyande, T. F. (2024, June). A comparison of AI methods for Groundwater Level Prediction in Burkina Faso. In IFIP International Conference on Artificial Intelligence Applications and Innovations (pp. 3-16). Cham: Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-63219-8_1
- Cao, W., Wu, X., Li, J., Kang, F. (2025). A review of artificial intelligence in dam engineering. Journal of Infrastructure Intelligence and Resilience, 4(1), 100122.https://doi.org/10.1016/j.iintel.2024.100122
- Chen, T., Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining (pp. 785-794).https://doi.org/10.1145/2939672.2939785
- Fan, J., Ma, X., Wu, L., Zhang, F., Yu, X., Zeng, W. (2019). Light Gradient Boosting Machine: An efficient soft computing model for estimating daily reference evapotranspiration with local and external meteorological data. Agricultural Water Management, 240, 106272.https://doi.org/10.1016/j.agwat.2019.106272
- Güneş Şen, S. (2025). Machine Learning-Based Water Level Forecast in a Dam Reservoir: A Case Study of Karaçomak Dam in the Kızılırmak Basin, Türkiye. Sustainability, 17(18), 8378. https://doi.org/10.3390/su17188378
- Harbi, L., Smail, N., Rouissat, B., Charrak, H. (2024). Assessing single and hybrid AI approaches in conjunction with FEM to enhance seepage prediction in earth dams. Modeling Earth Systems and Environment,10(2), 2421-2433.https://doi.org/10.1007/s40808-023-01903-2
- Hochreiter, S., Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735
- Khorchani, M., Rhayma, N., Pereira, S., Riahi, H. (2025). Exploring the role of artificial intelligence in predicting piezometric levels in homogeneous earth-fill dams. Modeling Earth Systems and Environment, 11(1), 33.https://doi.org/10.1007/s40808-024-02196-9
- Kozelj, D., Fernández, D., A. (2025). Predicting water distribution pipe failures using machine learning and cross-infrastructure data. Acta hydrotechnica, 38(68), 53-64.https://doi.org/10.15292/acta.hydro.2025.05
- Kratzert, F., Klotz, D., Brenner, C., Schulz, K., Herrnegger, M. (2018). Rainfall–runoff modelling using long short-term memory (LSTM) networks. Hydrology and Earth System Sciences, 22(11), 6005-6022.https://doi.org/10.5194/hess-22-6005-2018
- Kumi-Boateng, B., Ziggah, Y., Y. (2020). Feasibility of using Group Method of Data Handling (GMDH) approach for horizontal coordinate transformation. Geodesy and Cartography, 46(2), 55-66. https://doi.org/10.3846/gac.2020.10486
- Lainder, A., D., Wolfinger, R., D. (2022). Forecasting with gradient boosted trees: augmentation, tuning, and cross-validation strategies: Winning solution to the M5 Uncertainty competition. International Journal of Forecasting, 38(4), 1426-1433. https://doi.org/10.1016/j.ijforecast.2021.12.003
- Liao, W., Zhang, Z., Liu, B., Lu, X., Liu, D., Liu, Q., Liu, C. (2025). Intelligent zoning design of concrete-faced rockfill dams using image-parameter fusion enhanced generative adversarial networks. Engineering Structures, 339, 120662. https://doi.org/10.1016/j.engstruct.2025.120662
- Mueller, A., V., Hemond, H., F. (2013). Extended artificial neural networks: incorporation of a priori chemical knowledge enables use of ion selective electrodes for in-situ measurement of ions at environmentally relevant levels. Talanta, 117, 112-118.https://doi.org/10.1016/j.talanta.2013.08.045
- Ni, L., Wang, D., Singh, V., P., Wu, J., Wang, Y., Tao, Y. (2021). Streamflow forecasting using extreme gradient boosting and long short-term memory coupled with wavelet transform. Journal of Hydrology, 595, 126033.https://doi.org/10.1016/j.jhydrol.2020.124901
- Özbey, V., Ergintav, S., Tarı, E. (2024). GNSS time series analysis with machine learning algorithms: a case study for Anatolia. Remote Sensing, 16(17), 3309. https://doi.org/10.3390/rs16173309
- Osman, A., I., A., Ahmed, A., N., Chow, M., F., Huang, Y., F., El-Shafie, A. (2021). Extreme gradient boosting (Xgboost) model to predict the groundwater levels in Selangor Malaysia. Ain Shams Engineering Journal, 12(2), 1545-1556.https://doi.org/10.1016/j.asej.2020.11.011
- Ranković, V., Novaković, A., Grujović, N., Divac, D., Milivojević, N. (2014). Predicting piezometric water level in dams via artificial neural networks. Neural Computing and Applications, 24(5), 1115-1121.https://doi.org/10.1007/s00521-012-1334-2
- Sharghi, E., Nourani, V., Behfar, N., Tayfur, G. (2019). Data pre-post processing methods in AI-based modeling of seepage through earthen dams. Measurement, 147, 106820.https://doi.org/10.1016/j.measurement.2019.07.048
- Shen, C., Asante-Okyere, S., Yevenyo Ziggah, Y., Wang, L., Zhu, X. (2019). Group method of data handling (GMDH) lithology identification based on wavelet analysis and dimensionality reduction as well log data pre-processing techniques. Energies, 12(8), 1509.https://doi.org/10.3390/en12081509
- Waqas, M., Humphries, U., W., Hlaing, P., T., Ahmad, S. (2024). Seasonal WaveNet-LSTM: a deep learning framework for precipitation forecasting with integrated large scale climate drivers. Water, 16(22), 3194.https://doi.org/10.3390/w16223194
- Xie, Z., Chen, L., Li, Y. (2026). Optimization of dam safety monitoring models based on residual compensation: A multidimensional temporal framework integrating TCN-BiLSTM-attention mechanism. Engineering Structures, 348, 121869.https://doi.org/10.1016/j.engstruct.2025.121869
- Yazdi, S., H., Robati, M., Samani, S., Hargalani, F. Z. (2025). Prediction of two groundwater sustainability indicators in semi-arid aquifers using machine learning. Environmental Earth Sciences, 84(11), 294.https://doi.org/10.1007/s12665-025-12253-w
- Zheng, C., Cen, W., Liu, B., Qian, J., Ding, Y., Mo, C. (2025). Hybrid optimization and AI-driven surrogate model for seepage parameters inversion in complex dam foundations. Journal of Hydrology, 134484,https://doi.org/10.1016/j.jhydrol.2025.134484
- Zhou, Y., Zhang, Y., Pang, R., Xu, B. (2021). Seismic fragility analysis of high concrete faced rockfill dams based on plastic failure with support vector machine. Soil Dynamics and Earthquake Engineering, 144, 106587.https://doi.org/10.1016/j.soildyn.2021.106587
- Ziggah, Y., Y., Issaka, Y., Laari, P., B. (2022). Evaluation of different artificial intelligent methods for predicting dam piezometric water level. Modeling Earth Systems and Environment, 8(2), 2715-2731. https://doi.org/10.1007/s40808-021-01263-9