Résumé: This study addresses the critical issue of urban flooding caused by stormwater network overflow, necessitating unified and efficient management measures to handle increasing water volumes and the effects of climate change. The proposed approach aims to improve the precision and efficiency of overflow rate predictions by investigating advanced machine learning algorithms, specifically ensemble methods such as gradient boosting and random forest algorithms. The main contribution lies in introducing the SWN-ML approach, which integrates hydraulic simulations using MIKE + with machine learning to predict average overflow rates for various rainfall durations and return periods. Mike + model was calibrated for the only available observed data of water depth at the outlet point during the storm event of February 4, 2019. The datasets for model calibration used in ML models consisted of many input variables such as peak flow, max depth, length, slope, roughness, and diameter and average overflow rate as output variable. Experimental results show that these methods are effective under a variety of scenarios, with the ensemble methods consistently outperforming classical machine learning models. For example, the models exhibit similar performance metrics with an MSE of 0.023, RMSE of 0.15, and MAE of 0.101 for a 2-h rainfall duration and a 10-year return period. Correlation analysis further confirms the strong correlation between ensemble method predictions and MIKE + simulated models, with values ranging between 0.72 and 0.80, indicating their effectiveness in capturing stormwater network dynamics. These results validate the utility of ensemble learning models in predicting overflow rates in flood-prone urban areas. The study highlights the potential of ensemble learning models in forecasting overflow rates, offering valuable insights for the development of early warning systems and flood mitigation strategies.

Résumé: Water losses (WL) in water distribution networks are a major problem that requires special attention. Reducing WL is one of the most critical problems faced by the managers of water services. The study aims to analyze improvement alternatives to reduce water losses in the water distribution network of the city of Annaba (Eastern Algeria). The methodology applied in this study is based on examining the criteria and sub-criteria influential on the management of WL using the two multiple-criteria decision-making (MCDM) methods. The two MCDM approaches used in this study are the Analytic Hierarchy Process (AHP) and the Fuzzy Analytic Hierarchy Process (Fuzzy-AHP). Six evaluation criteria, twenty-six sub-criteria, and twelve alternatives were chosen to select the best strategies for reducing WL. The three economic (37.71%), technical (26.25%), and operational (14.27%) criteria are essential to take into account during the selection phase of the best alternatives. The results showed that the alternatives, rehabilitation of distribution networks, management and control of pressure and swift repair of water leaks present the best solutions for reducing WL.

Résumé: As a widespread environmental phenomenon, drought exerts significant adverse effects on agricultural productivity and environmental stability. Thorough drought assessment is essential for effective water resource management and monitoring strategies. Traditionally, drought is analyzed temporally and spatially without accounting for trend evolution, though trend studies have gained traction recently. This research addresses this gap by examining trend shifts, including frequency variations of drought classifications and their intensities, using a hybrid Frequency Innovative Trends and Statistics Analysis (FITA) and Innovative Trend Analysis (ITA) approach. The study applies 12-month Standardized Precipitation Index (SPI12), Reconnaissance Drought Index (RDI12), and Streamflow Drought Index (SDI12) across 10 stations in northeastern Algeria 1975–2021. Findings reveal that FITA-ITA enhances drought trend evaluation by offering a detailed microscale analysis of meteorological and hydrological patterns beyond simple increases or decreases. For instance, El Kalla shows a 100% increasing SPI12 trend (ITA slope 1.160), while Khenchela exhibits a 100% decreasing SPI12 trend (ITA slope − 0.846). At Medjaz Amar, SDI12 increases 100% (ITA slope 0.797) despite an 82.61% decreasing RDI12 (ITA slope − 1.294), reflecting streamflow gains amid agricultural droughts. These trends signal heightened risks to groundwater, desertification, and aquatic ecosystems, emphasizing the need for adaptive measures like rainwater harvesting and early warning systems to strengthen resilience in this semi-arid Mediterranean region.

Résumé: This study addresses key hydraulic engineering challenges in turbulent pipe flow - computing flow rate (Q), hydraulic energy slope , and pipe diameter (D) - by introducing the Modified Rough Model Method (MRMM). We propose novel, high-precision explicit equations for D (Eqs. 56 and 60). These achieve maximum relative errors of 0.017 % and 0.0085 %, respectively. We also introduce an innovative friction factor equation (54) with 0.086 % error. Validated across the entire Moody diagram (, and ) using a brute-force approach with over 7 million data points, these non-iterative solutions outperform existing models. A comprehensive set of statistical metrics including Mean Absolute Error (MAE), Root Mean Square Error (RMSE), correlation coefficients (R² and Pearson's R), Bias, Mean Relative Error (MRE), Standard Deviation (SD), Coefficient of Variation (CV), and maximum relative error were employed to assess the accuracy and reliability of the proposed and existing formulas; the results of the Statistical metrics confirm their robustness, establishing a new benchmark for accuracy in pipeline design. This advancement enhances efficiency and reliability in water, oil, and gas transport systems.

Résumé: Knowledge of variation of the water surface elevation in a river is very important for developing hydraulic models able to predict flooding and flood hazard mapping. The roughness and DEMs resolution are factors that can affect the flow characteristics in a river. Manning's roughness plays a significant role because it is not a constant parameter and varies depending on the length of the river. DEM resolution is potentially the crucial element to give better flood estimations. The objective of the present study is to assess the combined effects of Manning roughness and DEMs resolution to predict water surface elevation of the Moudjar River by using HEC-RAS 1D model. In this study, different results were obtained from various combinations of DEM resolution, and Manning roughness coefficients by applying the unsteady HEC-RAS (Hydrologic Engineering Centers- River Analysis System) 1D model on the Moudjar River in Algeria, to predict and quantify the performance regarding actual water surface elevation extent and the generated water surface limits. Data from February 02, 2019, to February 07, 2019, were used to determine predictive accuracy. The simulation results of the stage level of the river flow for different combinations of Manning’s roughness coefficient and DEM’s resolution show the closer values between the observed and simulated stage hydrograph, only in the case of the coarser DEM resolution (50 m) where the results are very distant and unacceptable. Statistical parameters such as Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Nash Sutcliff (NSE) were used to verify the model's accuracy and predictive ability. The results obtained and the methodology applied to the Moudjar River can be used as a useful reference for HEC-RAS modelling and flood predicting in Algerian rivers.

Résumé: The focus of this study is to examine and analysis the impacts of land use land cover (LULC) change for a period between 1985 and 2040 on the peak discharge and runoff volume in a Kebir river catchment using HEC-HMS model and remote sensing-GIS techniques. Therefore, this research started by analyzing changes in LULC by classifying Landsat 5 and Landsat 7 satellite images from 1985 to 2020, whereas the LULC change map of 2040 was obtained by prediction. Data analysis and projection were performed using an integrated Cellular Automata Artificial Neural Network (CA-ANN) methodology within the Modules of Land Use Change Evaluation (MOLUSCE) plugin in QGIS. The accuracy assessment of the classified images was performed by error matrix, where the overall accuracy and Kappa values were found to be 99.27% and 0.98 respectively. The classification results obtained are quite satisfactory, and thus, the classified image is utilized to evaluate changes in LULC throughout the study period. The findings reveal an increase of 0.403%, 0.584%, and 1.020% in agricultural lands, water bodies, and built-up lands respectively. Also, a decrease of 0.722% and 1.285% in Barren lands, and forests, respectively, between 1985 and 2040, was shown. To simulate the changes in the peak discharge and runoff volume, the classified LULC maps of 1985, 2003, 2020, and the predicted LULC map of 2040 are used in the HEC-HMS model during the calibration period from 18/12/1984 to 31/07/1985 and validation period from 01/01/2003 to 31/07/2003. The simulated results show a 1.93% rise in peak discharge during the calibration period and a 2.20% increase during the validation period. In addition, the runoff volume saw a 1.15% increase in the calibration period and a 1.53% increase in the validation period. Further, the performance results of the model were good for both calibration (RMSE = 0.50, NSE = 0.701 to 0.720, KGE = 0.36 to 0.56, and R2 = 0.92) and validation (RMSE = 0.60, NSE = 0.598 to 0.608, KGE = 0.31 to 0.34, and R2 = 0.86 to 0.87).

Résumé: Despite the high importance of coagulation process in drinking water treatment plant (DWTP), challenge remains in effectively linking raw water quality measured at the inlet of the DWTP with coagulant dosage rate. This study proposes an integral modelling framework using hybrid extreme learning machine and Bat metaheuristic algorithm (ELM-Bat) for modelling coagulant dosage rate using water temperature, pH, specific conductance, dissolved oxygen, and water turbidity. The aluminum sulphate (Al2 (SO4)3.18H2O) coagulant is determined using conventional Jar-Test procedure. Results obtained using the hybrid ELM-Bat were compared to those obtained using standalone ELM, outlier robust extreme learning machine (ORELM), online sequential extreme learning machine (OSELM), optimally pruned extreme learning machine (OPELM), and kernel extreme learning machine (KELM). First, the models have been calibrated during the training stage and in a second stage; they are validated using various statistical metrics, i.e., RMSE, MAE, the correlation coefficient (R), and the Nash–Sutcliffe model efficiency (NSE). We found that the hybrid ELM-Bat was significantly more accurate and it has yielded accuracy higher than all other models. During the validation stage, the R and NSE values calculated using the ELM-Bat were ≈0.959 and ≈0.918 exhibiting an improvement rates of approximately (≈15.26% and ≈33.82%), (≈10.35% and ≈21.92%), (≈14.98% and ≈31.89%), (≈7.63% and ≈16.35%), (≈10.99% and ≈23.05%), compared to the values obtained using the ELM, OPELM, OSELM, KELM and ORELM, respectively. Besides, the new ELM-Bat model has shown to have high predictive capabilities, which can be used optimally for calculating the optimal coagulant dosage with high accuracy.

Résumé: In the present study, three machine learning methods were applied for predicting seepage flow through embankment dams, namely (i) support vector regression (SVR), relevance vector machine (RVM), and Gaussian process regression (GPR). The three models were developed using seepage flow (Q: L/mn) and piezometer level (Z:m) measured at several piezometers placed in the corps body of the dam. The proposed models were calibrated and validated using a separate subset. Models evaluation and comparison was successfully achieved using various performances metrics, i.e., coefficient of correlation (R), Nash-Sutcliffe efficiency (NSE), root mean square error (RMSE), and mean absolute error (MAE). Experimental results showed that the proposed models are a good alternative to the in situ measured and contributed significantly in overcoming the case of missing measured seepage flow. The best performances were obtained using the RVM model with R and NSE values of ≈0.909 and ≈0.823, followed by the GPR model with R and NSE values of ≈0.891 and ≈0.767, while the SVR model was ranked as the poorest one exhibiting R and NSE values of ≈0.780 and ≈0.600, respectively. While, a growing number of investigations have focused on testing machine learning in terms of their feasibilities to accurately describe seepage flow, as well as providing important support to our understanding of the factors affecting its fluctuation, the present work was demonstrated that the combination of a wide range of variables can help in simulating seepage flow, and enhance their sensitivity which has help in developing new algorithms.

Résumé: Flood in urban areas is one of the disasters effecting humans. This may occur due to various reasons such as encroachment of water bodies, inadequate carrying capacity of stormwater networks and changes in rainfall patterns.