Water quality prediction plays a critical role in water supply systems (WSSs) management such as optimizing chlorination, identifying anomalies, and tracing pollutants. In contrast to traditional single-objective prediction, multi-node prediction holds increasing importance given the rising monitoring capabilities today. Despite several existing studies on multi-node water quality prediction, how to leverage the spatiotemporal relationships among monitoring point data to enhance prediction accuracy and stability, and how to ensure model robustness under abnormal conditions such as missing data are still problems that have not yet been fully addressed. This study proposes an Attention-based Multivariate Time Series Graph Neural Network (AMTGNN) for multi-node water quality prediction, which dynamically learns the correlations between different locations to jointly model temporal and spatial dependencies for improved prediction accuracy and robustness. The model is tested on a real-world urban-rural WSS. A comparative analysis with other state-of-the-art models is conducted, and the impact of different inputs and the robustness under missing data scenarios are also investigated. Experimental results show that AMTGNN achieves an R^2 of 0.92 when predicting residual chlorine variation 3 h ahead, and maintains a Mean Absolute Percentage Error (MAPE) below 5.5
Groundwater is a vital resource for human consumption and industrial activities. However, anthropogenic activities have become a major source of groundwater contamination, posing a significant environmental concern due to increasing pollution levels. Approximately 90
This article presents the activation of periodate (PI) under simulated sunlight (SSL) for the removal of reactive blue 222 dye (RB-222). Further, the degradation system was evaluated using real industrial effluents, and the utilization of natural sunlight as a light source for the degradation of RB-222 dye was assessed, both of which have been poorly explored in the literature. The optimization study revealed that a removal efficiency of 99.9
The persistent threat of toxic heavy metals like Cr(VI) and Cu(II) in water systems demands advanced, cost-effective solutions beyond conventional adsorbents such as activated carbon, clay minerals, and ion-exchange resins, which often suffer from limited adsorption capacity, slow kinetics, poor stability under acidic or basic conditions, and high regeneration costs. Here, we engineer a robust Ag2MoO4@Kaolinite nanocomposite to overcome these limitations, delivering exceptional adsorption performance and reusability. Advanced characterizations (XRD, FT-IR, SEM, and TEM) confirm the nanocomposite’s optimized structure and active sites. In batch tests, the material achieves > 90
Effective monitoring of surface and groundwater resources is essential for sustainable water management, especially in arid and semi-arid regions where water scarcity poses a critical challenge. Remote sensing technologies have emerged as powerful tools for assessing hydrological conditions with reduced cost and improved spatial and temporal coverage. This study provides important insights into an innovative method for estimating water levels and depths in wells using feeding lakes and remote sensing data, with a focus on the effectiveness of spectral bands and model performance. Unlike previous studies that focused mainly on bathymetric estimation, the proposed approach integrates surface water depth estimation with groundwater fluctuation analysis in nearby wells using a simplified regression-based framework. The wells, from which water depths were estimated, are located in the Kima area of Aswan, Egypt. A linear regression model was successfully developed to estimate pond water depths over different time periods using Sentinel-2 Level 1 C reflectance data and in-situ depth measurements as reference, highlighting the vital role of field data in model calibration and validation. The analysis revealed that bands B2 and B3 showed the strongest correlation with measured depths, with R² values exceeding 0.88, indicating their suitability for bathymetric applications. Model validation showed slightly higher RMSE values in the validation phase (> 0.35 m) compared to calibration, with some estimates exceeding 1–2 m, confirming the model’s reliability. A strong correlation (average cross-correlation coefficient of 0.8) was also observed between pond surface water levels and groundwater levels in nearby wells, reflecting the dynamic interaction between surface and groundwater systems. Applying the methodology to historical data (2015–2023) yielded promising results, demonstrating its potential for long-term monitoring and environmental assessment. These findings highlight the value of remote sensing in hydrological studies, offering a cost-effective, scalable solution for groundwater management and sustainable water resource planning.
Soil erosion is a critical global environmental challenge that threatens food security and undermines sustainability efforts worldwide. The accelerating rate of erosion leads to severe land degradation, making it essential to adopt strategic and localized resource management approaches. Prioritizing sub-watersheds allows for more effective conservation planning, as each watershed exhibits unique hydrological and geomorphological characteristics. This study introduces a comprehensive framework for prioritizing sub-watersheds (SWs) in the Aran Basin using an integrated approach. It combines morphometric analysis with multiple criteria decision-making (MCDM) methods—namely VIKOR (visekriterijumsko kompromisno rangiranje), TOPSIS (technique for order preference by similarity to ideal solution), and ARAS (additive ratio assessment)—alongside a machine learning model, support vector machine (SVM), to ensure a more robust and consensus-driven prioritization. The novelty of this research lies in the innovative integration of machine learning with morphometric and MCDM techniques, resulting in a scientifically rigorous and unified ranking of sub-watersheds. Based on the integrated analysis, SW1, SW8, SW9, and SW12 are identified as high-priority sub-watersheds for soil erosion control, collectively covering approximately 30
Abstract The accurate forecasting of surface soil moisture (SSM) for multiple time steps is vital in the early warning systems of drying and flooding, irrigation scheduling, and hydrological modelling. The current study presents an innovative hybrid structure called ASHA–LSTM–XGB that provides 1, 3, 5, and 7-day forecasting for SSM by combining Long Short-Term Memory networks (LSTM), eXtreme Gradient Boosting (XGBoost), and the Asynchronous Successive Halving Algorithm (ASHA). The proposed model was tested and validated over Izmir Province in Western Anatolia, Türkiye, using satellite mapping of the SMAP Level 4 SSM dataset and 11 ERA5-Land environmental predictors from April 2015 to October 2025. The obtained results clearly indicate that all proposed hybrid approaches showed better performance than the traditional LSTM model for any horizon of the prediction. For the shortest term, BO-LSTM-XGB was found to be the most accurate method (NRMSE = 0.2151, R2 = 0.8720). Meanwhile, the best combination of high prediction accuracy and stability was achieved by ASHA-LSTM-XGB, which demonstrated the best performance in the medium-term period and beyond (NRMSE = 0.2581, R2 = 0.8153; NRMSE = 0.2666, R2 = 0.8027 at 5- and 7-day horizons, respectively). Thus, these results demonstrate the efficiency of integrating sequential learning, non-linear ensembling, and optimal hyperparameter tuning for forecasting problems. Moreover, the results of the ablation study and analysis using SHAP values support the finding that the two most important predictor variables for SSM were the surface temperature and surface radiation variables. The integration of ASHA and LSTM-XGB has resulted in greater stability and generalization of the models. This study is considered the first application of the combined optimization-deep learning-ensemble architecture to forecast multi-step SSM. The proposed method performed better than the conventional LSTM algorithm in all forecasting scenarios and showed robust performance even at a 7-day forecasting lead time, showing promise for applications in short-range soil moisture prediction and environmental monitoring studies.
This study numerically investigates forced convection heat transfer and entropy generation of a water-based ternary hybrid nanofluid (Fe3O4 + CuO + MoS2) inside a two-dimensional square cavity containing three circular hot fins, under the influence of a direction-dependent magnetic field ( γ = 0°–90°). The finite element method (COMSOL Multiphysics 6.2) is employed to solve the governing equations. Four key parameters are systematically examined: Reynolds number (100–1000), Hartmann number (1–100), nanoparticle volume fraction (1 γ = 30° enhances the local Nusselt number by up to 78.6 γ = 0° at Ha = 100, owing to reduced Lorentz damping and the generation of beneficial secondary flows. Moderate Ha ≈ 25 provides a 2.7 ϕ from 1 to 10 γ = 90°) at Ha = 100 reduces peak total entropy generation by 89 γ = 0°, producing a nearly uniform irreversibility distribution, optimal for thermodynamic efficiency. Predictive regression models: Novel multiple linear regression equations for average temperature and entropy generation are derived with high statistical significance (p-value < 10–29) within the investigated parameter ranges, providing rapid design tools without further CFD. These findings establish that γ = 30° with Ha ≈ 25 and low ϕ offers the best heat transfer–entropy trade-off, while γ = 90° is recommended when minimizing irreversibility is the primary objective.
Groundwater plays a crucial role in Southeastern Morocco, where arid climate conditions, recurrent droughts, and increasing agricultural demands make it the primary source of water for domestic and irrigation purposes. In the absence of previous studies on the Bouanane watershed (18,393 km²), this work represents the first scientific exploration aimed at providing reliable data on groundwater quality and paving the way for future research. The study evaluates groundwater suitability for drinking and irrigation through the application of water quality indices, particularly the Drinking Water Quality Index (DWQI) and the Irrigation Water Quality Index (IWQI). Fieldwork was conducted in April 2024 at nine strategically selected sampling stations. Results reveal considerable variability in groundwater quality across the region, shaped by both natural processes and human activities. While some sites meet the requirements for domestic and agricultural use, others show elevated mineralization that poses limitations for irrigation. Traditional indicators such as total hardness (TH), residual sodium carbonate (RSC), electrical conductivity (EC), and magnesium hazard (MH) did not fully capture these spatial variations. In contrast, the DWQI, adapted to regional conditions, effectively highlighted disparities, with values ranging from 64.06 to 242.23 (average 135.2). Based on this index, groundwater quality was classified into three equal categories (good, average, and poor), each representing 33.3
The groundwater distribution in southern Egypt is primarily governed by the region’s inherited rift-related tectonic framework. We integrated aeromagnetic data with available hydrogeochemical and seismic data to investigate structural controls on groundwater accumulation in El-Gallaba Plain along the western margin of the Kom Ombo Basin. The magnetic data were processed using edge-detection filters and source-parameter imaging to estimate the depth of the basement rock. Moreover, three-dimensional geophysical inversion of magnetic data was performed using the iteratively re-weighted least-squares algorithm to map the basin geometry. The results of magnetic data interpretation delineated a NW-SE major trend and minor NE-SW-trending, fault-bounded depressions extending to depths of 2.5–3.0 km, interpreted as sediment-filled or altered-basement fractured basins favorable for groundwater accumulation. Uplifted basement blocks define structural highs that act as hydraulic barriers, compartmentalizing the area’s aquifer system. The interpreted magnetic structures were correlated with previously published land-surface temperature and hydrochemical data, revealing a spatial alignment among deep structural depressions, low-temperature corridors, and zones of fresher, isotopically depleted groundwater. This correlation confirms that reactivated Pan-African and Cenozoic faults control the vertical permeability and groundwater connectivity of the Nubian Sandstone Aquifer. Overall, this study highlights the effectiveness of three-dimensional inversion, along with edge-detection and depth-estimation filters applied to magnetic data to resolve subsurface architecture and interpret groundwater potential in arid, tectonically complex regions.
Sustainable management of water resources in urban areas is a critical issue due to rapid urbanization. Building resilient urban water supply and demand systems requires an appropriate understanding of how anthropogenic activities, climate change, and environmental factors affect water availability and distribution. An integrated, holistic approach is essential to achieve effective urban water resource management. This research reviews the challenges of water scarcity driven by unprecedented urban expansion. The study aims to analyse water issues across the globe and understand their water management system. Systematic technological trends were observed in urban water science, from econometric techniques of the 1980s to geospatial technologies in the 20th century to extensive integration of artificial intelligence in the 21st century. The study aims to highlight the integration of advanced techniques within a Geographic Information System (GIS) framework for urban water management. It provides a comprehensive overview of various spatial techniques and analytical technologies, including Geospatial models, geostatistical models, hydraulic models, and artificial intelligence/machine learning approaches, and discusses how these methodologies are interoperable within a GIS environment to study water. Combining various techniques is a good practice for accurate analysis and prediction-based urban water studies. GIS is a powerful decision-support tool that enables multidisciplinary approaches by integrating diverse spatial and non-spatial datasets on a single digital platform for comprehensive urban water analysis. An integrated GIS framework provides better resource allocation, raises awareness about responsible water use, and promotes water-saving technologies.
Reliable forecasting of water consumption is essential for water resources management because it enables policymakers and utilities to balance supply and demand effectively. This study examines seasonal (three-month horizon) water consumption in Isfahan Province, Iran, using a modeling table of 528 seasonal observations spanning 24 subscriber categories over the period 2016–2021. The predictor set includes subscriber category, year, season, temperature, rainfall, a COVID-19 indicator, death rate, and birth rate. The analysis compares five machine learning algorithms, namely Random Forest (RF), Decision Tree (DT), Gradient Boosting (GB), eXtreme Gradient Boosting (XGBoost), and Categorical Boosting (CatBoost), under a random 80/20 holdout split with 7-fold cross-validation on the training partition. All five models achieve high predictive accuracy, but their relative performance depends on metric choice, computational cost, and sensitivity to low-consumption subscriber categories.
Accurate groundwater level forecasting is crucial for sustainable water resource management and environmental protection. The main objective of this study was to develop hybrid models combining deep learning models with data processing techniques for groundwater level forecasting. Two groundwater level observations in Cangzhou City, Hebei Province, China, were used as case studies. First, the empirical mode decomposition (EMD) and ensemble empirical mode decomposition (EEMD) were used to pre-process the dataset. Then, two dynamic deep learning techniques, namely long short-term memory (LSTM) and bidirectional LSTM (BiLSTM), were used as the base model. The hybrid models combined two deep learning algorithms (LSTM and BiLSTM) with EMD and EEMD, respectively, yielding 4 models: EMD-LSTM, EEMD-LSTM, EMD-BiLSTM, and EEMD-BiLSTM. Five evaluation metrics, including the root mean squared error (RMSE), the mean absolute error (MAE), the Nash-Sutcliffe efficiency (NSE), the Kling-Gupta efficiency (KGE), and the coefficient of determination (R2), were used to evaluate the proposed models. All the proposed models could achieve satisfactory results, with R2, NSE, and KGE values greater than 0.87. The hybrid models performed better than LSTM and BiLSTM, with reductions of up to 75
Wastewater from industries, municipalities, or the agricultural sector contains huge amounts of nitrogenous substances, causing serious risks to the immediate locality and public health. Anammox technology is recognized as a preferred option for treating nitrogenous wastewater due to its sustainability, cost-effectiveness, and eco-friendliness. In addition, this technique is characterized by the absence of organic carbon, lower sludge production, reduced energy consumption and greenhouse gas emissions, and high nitrogen removal efficiency. However, the efficacy of the wastewater treatment process is deterred by the presence of inhibitory pollutants, including microplastics, heavy metals, and antibiotics. Consequently, the coupling of anammox with other biological wastewater treatment processes, including partial nitrification (PN), partial denitrification (PD), simultaneous partial nitrification anammox and denitrification (SNAD), constructed wetlands (CWs), microbial electrolysis cells (MECs), and microbial fuel cells (MFCs), offers an effective approach to sustainable and improved nitrogen removal from wastewater. The addition of biochar, zeolite, graphene, metal ions, and hydrazine in the culture medium promotes anammox bacterial growth and catalytic activity, as well as pollutant removal efficiency. Therefore, this review provides insights into various anammox-coupled biological technologies for the efficient treatment of nitrogenous wastewater and elucidates strategies to optimize anammox performance. Some challenges that impede anammox activity, along with their respective antidotes, are also highlighted.
Biomass based activated carbon as an adsorbent attained significant attention in today’s world for the industrial waste water treatment due to its sustainable nature, high surface area and adsorption ability. In the present study, activated carbon (AC) was developed from of Pinus roxburghii (Chir Pine) dry needle biomass and was magnetically modified through physical activation. The iron oxide nanoparticles were integrated by in-situ technique to impart improved magnetic separability. The prepared samples of magnetic modified activated carbon (MMAC) were systematically characterized for evaluating the surface morphology, presence of surface functional groups, crystalline structure and magnetic properties. The obtained results confirmed regular tubular vacant spaces, which were later on magnetically modified by dispersing Fe₃O₄ nanoparticles in the hollow spaces of AC to obtain strong magnetic responsiveness. The adsorption studies were performed using Malachite Green (MG) industrial dye. The batch adsorption of MG dye from the aqueous solution demonstrated AC’s maximum Langmuir adsorption capacity of Qmax of 416.93 mg/g while the optimized magnetic composite (MMAC3) retained a highly competitive capacity of 369.47 mg/g. AC required complex filtration for recovery and the MMAC3 exhibited high magnetic saturation of 1.22 emu/g using the external magnet from the aqueous solution. Non-linear kinetic analysis and multi-model equilibrium isotherm evaluations (including Langmuir, Freundlich, Temkin, and Dubinin-Radushkevich) were conducted to rigorously define the physical and mass-transfer parameters driving the adsorption capacity. Thermodynamic investigations confirmed that Malachite Green sequestration by both AC and the MMAC 3 composite is a spontaneous, endothermic and entropy-driven physisorption process. The most important finding of the regeneration experiments is the impressive cyclic stability of both the matrices used, with each matrix maintaining more than 80
The photocatalytic degradation of persistent nitro-aromatic pollutants is often limited by poor light utilization and fast electron-hole recombination in conventional TiO2 systems. In this work, TiO2/carbon composites were prepared using activated carbon (AC) and biomass-derived biochars as supports and evaluated for the degradation of p-Nitrophenol (p-NP). Among the different compositions studied, 20 wt
This study explores the adsorption of fluorescein dye (FD) onto bentonite (BT) to enhance its potential in environmental remediation and water treatment. X-ray diffraction (XRD) analysis confirmed montmorillonite as the dominant phase in BT, with significant SiO₂ and quartz peaks. SEM and EDX characterization revealed a regular bulk morphology and the presence of silicon, aluminium, magnesium, iron, and trace elements. Adsorption experiments were conducted to evaluate the effects of adsorbent dosage, solution pH, and temperature. Increasing BT dosage from 20 mg to 400 mg enhanced dye removal efficiency from 21
Surface water is the primary source of drinking water in the Mahanadi River and its distributaries, and it is used for agricultural, industrial as well as domestic purposes. However, increasing urbanization, agricultural activities, and seawater intrusion have raised concerns about groundwater quality. This study evaluates the surface water quality (WQ) in nine selected locations by analysing twelve physicochemical parameters, computing the Weighted Arithmetic (WA) Water Quality Index (WQI), Numerow’s Pollution Index (NPI), Overall Index of Pollution (OIP), Synthetic Pollution Index (SPI), and applying Multivariate techniques namely, Corelation, Cluster Analysis (CA), and Principal Component Analysis (PCA), in order to identify key pollutants. Again, the current work is intended to create and review four machine learning (ML) models, namely, Support Vector Machine (SVM), Random Forest Model (RFM), Gradient Boosting Machine (GBM), and Extreme Gradient Boosting (XGB), for surface water quality prediction and creating spatial water quality maps to direct conservation initiatives in the heavily urbanized and polluted region being monitored. On the basis of physicochemical results, BOD, PO43−, EC, TDS, and F− were major contributors to surface water pollution, particularly in urban and agricultural areas. Findings revealed that WA WQI values ranged from 40.36 to 176.13, indicating that water quality varied from good to unsuitable category. The NPI indicated deteriorated water quality, accounting 77.78
Monitoring the quality of water in rivers is important for environmental sustainability. This study presents a comparison of Decision Tree (DT) and Random Forest (RF) classifiers for multi-class classification of the water quality of the Yamuna River on the basis of temperature, pH, electrical conductivity, BOD, and fecal coliform concentration. To enhance the robustness of the findings, noise of ± 5
Abstract Asphaltene-contaminated sand poses a significant environmental threat and presents substantial challenges as a waste product. The immediate and serious risks it poses to ecosystems and human health demand urgent action. This study presents an innovative carbon-based SiO 2 composite that is synthesized from crude oil-contaminated sand using a carefully controlled thermal treatment process at 600 °C and 1200 °C, with the temperature difference influencing the material’s properties. The synthesized C@SiO 2 was characterized and evaluated for its ability to effectively remove copper ions (Cu²) from water. We thoroughly examined crucial factors that impact the affinity and efficiency of Cu²⁺ sorption, such as contact time, adsorbent dosage, initial concentration, and competitive adsorption behaviors. It is clearly shown that the maximum adsorption achieved 95.06% after 120 min in the presence of a 2 g/L adsorbent for C@SiO 2 -600, while C@SiO 2 -1200 exhibited a lower initial efficiency of 33.02%, reflecting the temperature effect. The Langmuir isotherm best fits the data, indicating monolayer adsorption with a maximum capacity of ~ 30.67 mg/g using sample prepared at 600 °C, which decreases with sample prepared at 1200 °C due to potential thermal degradation. Kinetic studies favor the pseudo-second-order model (R² = 0.9996) for C@SiO 2 -600, suggesting chemisorption, with a reduced fit with C@SiO 2 -1200 . The C@SiO 2 -600 adsorbent demonstrates excellent reusability, maintaining 90% efficiency over four cycles, whereas C@SiO 2 -1200 loses effectiveness by the third cycle, highlighting the temperature difference’s impact on stability. This study demonstrates an innovative and eco-friendly approach to convert asphaltene-contaminated sand into a valuable resource for water treatment, enhancing waste valorization and improving heavy metal remediation, with C@SiO 2 -600 proving more advantageous than C@SiO 2 -1200. To further enhance water purification efficiency, the synthesized C@SiO 2 was incorporated into an acrylic fiber waste matrix to fabricate ultrafiltration (UF) membranes. The resulting composite membranes exhibited improved hydrophilicity, mechanical strength, and antifouling behavior compared to pristine acrylic membranes. The modified membranes (NCM1 &NCM2), containing C@SiO 2 prepared at 600 & 1200 °C, respectively, achieved a pure water flux of approximately 60 LMH and demonstrated excellent removal performance, rejecting up to 98% of humic acid. These results confirm that the integration of upcycled C@SiO 2 into acrylic fiber-based membranes not only enhances pollutant removal efficiency but also contributes to sustainable waste valorization and advanced water treatment solutions.
