Water quality modelling using principal component analysis and artificial neural network

Artificial neural networks to estimate the sorption and desorption of the herbicide linuron in Brazilian soils

Generally, herbicides used in Brazil follow manufacturer's recommendations, which often do not consider soil attributes. Statistical models that include the physicochemical properties of the soil involved in herbicide retention processes could enable greater precision in herbicide dose decision-making. This study evaluated the potential of artificial neural networks (ANNs) to predict the sorption and desorption of the herbicide linuron in Brazilian soils with different attributes. ANNs and multilayer perceptron (MLP) models were built to predict the sorption and desorption of the herbicide linuron. The inputs to the networks were pH, organic matter (OM), clay, cation exchange capacity (CEC), and base saturation (V); the outputs were sorption (Kfs and Qmax) and desorption (Kfd). The performance of the predictive model was assessed by the coefficient of determination (R2), mean absolute relative error (RMSE), mean absolute error (MAE), mean estimation error (MBE), and Pearson's correlation coefficient (r). The best-performing ANNs for predicting Kfs and Kfd comprised the variables pH, OM, CEC, and clay; for predicting Qmax, the ANN comprised the variables pH, OM, and clay. Artificial neural network models have proved to be a valuable tool for predicting the sorption and desorption of the herbicide linuron in soil, helping to minimize environmental impacts by providing accurate estimates and promoting sustainable herbicide use based on soil attributes.

Determination of polystyrene nanospheres and other nanoplastics in water via binding with organic dyes by capillary electrophoresis with laser-induced fluorescence detection

The purpose of this research work was to develop a new method for the quantitative analysis of water samples containing nanoplastics in the presence of microplastics and other colloidal particles. Our approach involved a mixture of fluorescent organic dyes that was added to each water sample for binding with the target nanoplastics. Binding was proven by zeta potential measurements that revealed the point of zero charge shifting from pH 4 for polystyrene nanoparticles, to pH 6.13 after binding with the dye mixture. Centrifugation effectively separated the free dyes from all dye-bound particles in the heterogeneous mixture, thus eliminating any potential interference. Electrokinetic injection of the free dyes in the supernatant allowed efficient separation by capillary electrophoresis (CE), for accurate quantitation individually with laser-induced fluorescence detection. A diode laser was operated at λex of 450 nm to induce fluorescence from the dyes, and an optical interference filter to collect only emission photons with λem of 520 nm. The fluorescence peak intensity decreased for each dye, thereby determining the total binding activity of all plastics and other particles. This new method enables high-throughput screening of water samples for nanoplastics, based on their fast binding with organic dyes in 5 min, rapid analytical separation of dyes by capillary electrophoresis within 10 min, and instantaneous fluorescence intensity measurement of individual dye peaks. Binding percentages as high as 149(±2) %/μg of 9.5-nm polystyrene nanoparticles were attained when using a concentration of 125 μg/mL for each dye. The binding mechanism was mainly attributed to hydrophobic interaction and modified by electrostatic forces. Binding of the four dyes with polystyrene microparticles, casein micelles, and transition metal oxide nanoparticles was verified to demonstrate minimal interference. The method was successfully applied to rapid testing of water samples from various sources, ranging from drinking fountains and household faucets to flowing rivers. The method also applied to a decontamination study wherein a removal of 94 % polystyrene nanospheres (diameter = 80 nm) was achieved by adding only 20 mg of casein powder into 1.6 mL of water containing 36 mg of the nanoplastics initially.

A multi-method approach to assess long-term urbanization impacts on an ecologically sensitive urban wetland in Northeast India

Deepor Beel, a natural wetland fringing the outskirts of the sub-Himalayan city of Guwahati in North-East India, has been under threat of urbanization since the past few decades. With a shrinking perimeter, the wetland - a favorite winter halt of migrating Siberian birds, manages to survive between anthropogenic aggression and ecological existence. This study maps the wetland's aerial shrinkage and environmental health from the 1990s to the 2020s using satellite imagery at five-year intervals. The water quality indicators used are Chlorophyll-a (Chl-a), turbidity, and total suspended solids (TSS) - the optically active parameters commonly used in satellite-image supported monitoring of water bodies. The comparisons indicate that while Chl-a or TSS levels in the wetland appears to have not changed significantly over the years, the expanse of the water-body shows a rapid reduction. Landuse and land cover (LULC) classification reveals maximum built-up area expansion during 2000-2010 at 52.38 %, followed by 21.6 % growth from 2010 to 2020. Additionally, two machine learning (ML) algorithms, artificial neural network (ANN) and random forest (RF), are incorporated to identify predictors from Landsat satellite bands and band ratios that reflect water quality characteristics for the different years. The correlations are validated against field-acquired data for three seasons: pre-monsoon, monsoon and post monsoon of 2021 and pre-monsoon as well as monsoon seasons of 2022. The ML models show encouraging predictions with the Green Normalized Difference Vegetation Index (GNDVI) and Normalized Difference Turbidity Index (NDTI) for evaluation of the Chl-a and TSS parameters. The moderate but increasing Chl-a values indicate the wetland's susceptibility to eutrophication, possibly due to urbanization. Thus, the use of satellite derived data along with machine learning tools and synoptic sampling for water quality assessment and predictions will be beneficial for urban planners and environmental managers for effective wetland management, especially in data poor regions.

The impact of temperature on the growth of Pseudomonas aeruginosa in mineral waters originated from different wells: A predictive approach

This study aimed to evaluate the behavior of Pseudomonas aeruginosa (PSA) in natural mineral water sourced from three different extraction wells and stored at various temperatures (10, 12, 20, 23, and 30 °C) to calculate the kinetic growth parameters of this microorganism through predictive modeling. The physicochemical characterization of waters was also evaluated at the time of collection, and included the analysis of 40 different minerals, and quality parameters such as pH, conductivity, oxidation-reduction potential (ORP), total dissolved solids (TDS), salinity (PSU), and temperature (T). PSA survived in raw mineral water incubated at 12, 20, 23, and 30 °C; however, no growth was observed at 10 °C. Growth curves started with an initial population of ∼ 2.5-3 log CFU/mL, and final PSA populations ranged from 3.5 to 4.9 log CFU/mL. The maximum specific growth rate (μmax) at 30 °C varied among the wells, with Well P-07 showing the highest growth rate (0.2 h-1), followed by Well P-08 (0.195 h-1) and well P-01 (0.133 h-1). At 12 °C, well P-01 exhibited the highest growth rate (μmax = 0.22 h-1), indicating a influence of mineral composition in the growth of PSA. The lag time (λ) also varied, with minimum values of 2.4 ± 0.1 h at 30 °C and maximum values of 41.6 ± 0.2 h at 12 °C. From these primary estimated parameters, it was possible to obtain five robust secondary models to describe the influence of temperature on the maximum growth rates and lag phase of PSA in the well. The estimated PSA growth parameters at 20 and 23 °C were subjected to a hierarchical cluster analysis and correlation plots to verify the influence of the physicochemical composition of the waters on the PSA behavior at each well's specific annual average temperature. This analysis confirmed a positive relationship (p < 0.05) between the presence of minerals (Ca, Fe, Sr, Mn, Na) and ions (SO4-3, Cl-) and the PSA lag phase time. These results underscore the need for tailored water quality management strategies that consider chemical composition and temperature to address specific microbial contamination risks.

Hydrogeochemical investigation and water quality assessment of the Indus River in the semiarid region of Ladakh, India

The decline in water quality, particularly in river water, is a significant concern, especially in semi-arid areas and tourist destinations such as Ladakh. Periodic assessment of water quality could be a crucial step for ensuring its potability and serve as a foundation for formulating effective policies for sustainable water resource management. Consequently, this research aimed to analyze the periodic variations in the water quality of Indus River for domestic and agricultural use, focusing on the impact of geochemical processes within the basin. Various physicochemical parameters namely temperature, pH, electrical conductivity, turbidity, total dissolved solids, total hardness, total alkalinity, K+, Na+, Mg2+, Ca2+, Cl-, SO42-, SiO2, HCO3-, CO32-, and NO3- for sixty-five water samples from seven key locations were assessed during three distinct periods: April-May (early melting period, EMP), July-August (peak melting period, PMP), and October-November (late melting period, LMP), 2023. The ion contents found were in the following order: Ca2+ > Na+ > Mg2+ > K+ and HCO3- > SO42- > Cl- > SiO2 > NO3- > F-, reflecting Ca-HCO3 water types. However, temporal and spatial variation in ion content and hydrochemical facies were observed when the water moved downstream and confluences with the Zanskar River to give calcium‑magnesium-sulphate facies. Water quality indices- Canadian Council of the Ministers of the Environment Water Quality Index (CCMEWQI) and Weighted Arithmetic Water Quality Index (WAWQI) were employed to assess the water quality over these periods, identify long-term trends, evaluate the water quality status, and provide insights into immediate conditions. WAWQI values recorded for EMP (56.89-509.53), PMP (289.82-3419.23), and LMP (55.16-159.14) found were poor to unsuitable for drinking while using CCMEWQI, it was in the marginal range. Additionally, the Wilcox diagram and other important irrigational indices like Percent sodium, Sodium Absorption Ratio, Residual Sodium Carbonate, Magnesium hazard, Permeability Index, Kelly's ratio, Ryznar stability index indicated the suitability of the water for agricultural use in all the periods. Apart from arsenic (54 μg/L), all other heavy metal ions measured were within the permissible limits according to the Heavy Metal Pollution index. Principal Component Analysis identified different principal components contributing to the hydrochemistry of the river water whereas correlational analysis was conducted to understand the correlation among different parameters and their potential source in water. Rapid carbonate mineral reactions and sulphate reduction were found to affect the alkalinity and hardness of the water. This study will serve as a scientific reference and methodological guide for researchers- to understand water chemistry; committee awareness on health and agricultural impacts; and policymakers for decision-making on water use, policy, and conservation.

Dynamics of pollution and trophic status in selected sub-tropical surface water bodies in Haridwar district, India

This study provides a detailed approach to evaluating water quality in the Haridwar district, Uttarakhand, India, by integrating physicochemical and microbiological investigations. It employs multivariate analysis and applies water quality and trophic state indices to evaluate the current state of the water and identify potential sources of contamination. The results from the correlation matrix highlight the dynamic interactions between different water quality parameters. Dissolved oxygen, total alkalinity, total suspended solids, biochemical oxygen demand, ammonium nitrogen, fecal coliform, pH, nitrate, and temperature exhibited varying temporal associations. Furthermore, principal component analysis (PCA) identified pH, temperature, and total alkalinity as key influencers across all seasons, while sewage pollution and agricultural runoff were recurring concerns. The enumeration of NSF WQI (National Sanitation Foundation Water Quality Index) (observed from 33.0 to 41.7) further affirmed consistently poor water quality and provided a quantitative metric for a comprehensive assessment. Trophic state index (ranged from 58.6 to 94.2) analysis indicated hyper-eutrophic conditions driven by nutrient concentrations in specific sites, which was further validated by phytoplankton analysis, which revealed the widespread occurrence of cyanobacteria in nearly all water bodies, signaling severe nutrient enrichment and the risk of potential eutrophication. Despite high nutrient levels causing pollution, the water in these water bodies consistently fell within the safe category for irrigation based on estimated sodium percentage, sodium adsorption ratio, and Wilcox diagram. However, it is not recommended to use these surface water bodies for irrigation due to the presence of high levels of organic pollution and a significant load of pathogenic bacteria.

Water quality assessment of Johor River Basin, Malaysia, using multivariate analysis and spatial interpolation method

In the Johor River Basin, a comprehensive analysis was conducted on 24 water environmental parameters across 33 sampling sites over 3 years, encompassing both dry and wet seasons. A total of 396 water samples were collected and analyzed to calculate the Water Quality Index (WQI). To further assess water quality and pinpoint potential pollution sources, multivariate techniques such as principal component analysis (PCA) and cluster analysis (CA), alongside spatial analysis using inverse distance weighted (IDW) interpolation, were employed. According to the National Water Quality Standard, most of the analyzed physicochemical components fall within Classes II and III, albeit with varying concentrations. However, certain sites exhibited levels of BOD5, TSS, and nutrients such as total nitrogen (TN) and total phosphorus (TP) that exceeded the threshold level of water quality standards, signaling pollution from diverse sources. Notably, all trace elements, with the exception of copper (Cu) and nickel (Ni), remained within the acceptable limits set by WHO guidelines and the National Water Quality Standard. PCA revealed parameter groupings linked to factors such as soil erosion, salinity, wastewater discharge, and fecal contamination, which are key determinants of water quality. The cluster analysis categorized the 33 sampling sites into three distinct clusters, each reflecting the geological setting and varying levels of pollution. The IDW-based spatial distribution indicated significant water quality degradation as the river flows downstream, particularly in regions experiencing rapid agricultural, industrial, and residential development. These activities contribute to the breakdown of organic matter and the release or overflow of wastewater into nearby river systems. This study highlights the effectiveness of integrating data-driven methodologies for surface water quality assessment, offering valuable insights for sustainable watershed management.

Exploring action-law of microplastic abundance variation in river waters at coastal regions of China based on machine learning prediction

Surface waters, particularly the river systems, constitute a vital freshwater resource for human beings and aquatic life on Earth. In economically developed and densely populated coastal regions, river water is facing severe microplastic pollution, posing a threat to public health and ecological safety. Reliable prediction of microplastic abundance (MPA) can significantly reduce the costs associated with microplastic field sampling and analysis. This study employed spatial correlation, geographical detector, principal component analysis and five mainstream machine learning models to analyze 79 datasets of MPAs in seven coastal areas of China and performed correlation, regression and attribution analyses based on 19 terrestrial influencing factors that potentially affect the MPA life cycle processes (generation, aging, and migration). The results showed that the Neural Network (NN) and the Gaussian Process Regression (GPR) models achieved the best prediction performance, with the predicted R2 close to 1. Principal component analysis and Shapley additive explanations concluded that meteorological factors, in particular the annual geotemperature, surface solar radiation, and annual relative humidity, had a key influence on the aging of microplastics. The second key factor in improving the MPA prediction ability was the dynamic description of microplastic migration, which was primarily governed by hydrological factors such as annual precipitation and average terrain slope. Unexpectedly, the effects of land use and level of urbanization were relatively small in describing the generation of microplastics. Only the percentage of built areas was strongly correlated with the MPA levels. Note that the MPA prediction and its contribution factors may vary across different basins. Nevertheless, the findings of this study are applicable to predicting and analyzing the distribution of microplastics in other coastal rivers, and for indicating the main contributing factors, ultimately serving as a basis for guiding microplastic pollution control strategies in different river basins.

Exploration of supramolecular solvent-based microextraction for crystal violet detecting in water samples

This approach highlights the advantages of supramolecular solvents in a new microextraction model. The distinct properties and behavior of this supramolecular solvent provide enhanced extraction capabilities for detecting crystal violet (CV) in water samples. The methodical experimentation was executed to optimize the critical process parameters, providing maximum efficiency of crystal violet extraction at optimal conditions with pH set at 2.7, 186 μL of extraction solvent, extraction time of 3.5 min, and a salt amount of 3.1 % w/v, yielding the best results. Analytical data from extraction experiments under these optimal conditions demonstrated a high extraction percentage. The extraction model exhibited a linear response within the range of 10-800 ng mL-1 of crystal violet, with a detection limit of 2 ng mL-1. This model enables the measurement of CV in water samples with recovery rates exceeding 97 %, offering a straightforward and accessible approach for analysis.

An efficient water quality index forecasting and categorization using optimized Deep Capsule Crystal Edge Graph neural network

The world's freshwater supply, predominantly sourced from rivers, faces significant contamination from various economic activities, confirming that the quality of river water is critical for public health, environmental sustainability, and effective pollution control. This research addresses the urgent need for accurate and reliable water quality monitoring by introducing a novel method for estimating the water quality index (WQI). The proposed approach combines cutting-edge optimization techniques with Deep Capsule Crystal Edge Graph neural networks, marking a significant advancement in the field. The innovation lies in the integration of a Hybrid Crested Porcupine Genghis Khan Shark Optimization Algorithm for precise feature selection, ensuring that the most relevant indicators of water quality (WQ) are utilized. Furthermore, the use of the Greylag Goose Optimization Algorithm to fine-tune the neural network's weight parameters enhances the model's predictive accuracy. This dual optimization framework significantly improves WQI prediction, achieving a remarkable mean squared error (MSE) of 6.7 and an accuracy of 99%. By providing a robust and highly accurate method for WQ assessment, this research offers a powerful tool for environmental authorities to proactively manage river WQ, prevent pollution, and evaluate the success of restoration efforts. PRACTITIONER POINTS: Novel method combines optimization and Deep Capsule Crystal Edge Graph for WQI estimation. Preprocessing includes data cleanup and feature selection using advanced algorithms. Deep Capsule Crystal Edge Graph neural network predicts WQI with high accuracy. Greylag Goose Optimization fine-tunes network parameters for precise forecasts. Proposed method achieves low MSE of 6.7 and high accuracy of 99%.

Unveiling heavy metal pollution dynamics in sediments of river Ulhas, Maharashtra, India: a comprehensive analysis of anthropogenic influence, pollution indices, and health risk assessment

Metals and metalloids tainting sediments is an eminent issue, predominantly in megacities like Mumbai and Navi Mumbai, requiring an exhaustive examination to identify metal levels in river bodies that serve various populations. Thus, utilising pollution indices, multivariate analysis, and health risk assessment studies, we propose a novel investigation to examine the metal content in the Ulhas River sediments, a prominent agricultural and drinking water supply (320 million-litre per day) near Mumbai in Maharashtra, India. The eleven metals and metalloids (As, Cd, Cr, Cu, Fe, Hg, Mn, Ni, Pb, Se, and Zn) were examined monthly from 10 stations totaling 120 sediment specimens from October 2022 to September 2023. Investigations revealed that average values of Cr, Cu, Hg, and Ni exceeded Australian and New Zealand Environment and Conservation Council and Agriculture and Resource Management Council values, while all metals exceeded World surface rock average limits except As. Various pollution indices showed that upstream sites had none to low level contamination, whereas downstream locations had moderate to considerable contamination, suggesting anthropogenic influences. Furthermore, multivariate analysis including correlation, cluster, and principal component analysis identified that sediment pollution was mostly caused by anthropogenic activities. Lastly, health risk assessment indicated Fe was non-carcinogenic to children, whereas Cr and Ni were carcinogenic to children and adults, with children being more susceptible. Thus, from the findings of the study it is clear that, despite low to moderate pollution levels, metals may have significant repercussions, thus requiring long-term planning, frequent monitoring, and metal abatement strategies to mitigate river contamination.

Evaluation of contaminated groundwater for excessive heavy metal presence and its further assessment of the potential risk to public health

Water plays a significant role in human life. However, the contamination of groundwater by heavy metals (HMs) has profound implications for public health. Industrialization, urbanization, and agricultural activities are turning out to be major causes for the increasing concentration of HMs in rapidly industrializing areas like Rohtak district, Haryana, India. The current study aimed at evaluating and predicting the health hazards associated with the radical rise of HMs in the groundwater of Rohtak district. For this purpose, 45 seasonal-based groundwater samples were collected from five blocks in Rohtak district, namely Kalanaur, Meham, Lakhan Majra, Rohtak City, and Sampla, both during pre- and post-monsoon seasons. Besides physicochemical analysis, these groundwater samples were analyzed for the contamination of HMs. The findings revealed that groundwater samples were relatively more contaminated during the post-monsoon period rather than pre-monsoon. The water quality index (WQI), devised to classify water quality into specific classes, depicted the Kalanaur region as "very poor." Another index named the HM pollution index (HPI) denoted the levels of HMs and categorized Kalanaur as most deteriorated, followed by Meham, Lakhan Majra, Sampla, and Rohtak City. Additionally, principal component analysis (PCA) was employed that showed a significant variation in the distribution pattern of HMs, with the major load being attributed to PC1 and PC2 for both seasons. Pearson's correlation analysis indicated a significant association of pH (R2 = 0.917) with HMs (specifically for Cd and Cr). In terms of health risk assessment, carcinogenic human health risk due to Pb and Cr was found to be higher in children than adults. Non-carcinogenic risk, indicative of harmful human health effects, apart from cancer, was calculated in terms of hazard quotient (HQ) and hazard index (HI). Results of the same, designated "children" as a vulnerable category compared with "adults," especially in the Kalanaur, Sampla, and Rohtak City blocks of the study area. The results thus reiterated that Kalanaur is the most contaminated block among the five blocks chosen and should be given urgent attention. The study holds importance as it provides a framework regarding the methodology that should be adapted for the evaluation, management, and protection of groundwater at a regional level, which could further be replicated by environmentalists and hydrogeologists across the world. PRACTITIONER POINTS: Water logging is one of the most common problems in Kalanaur block of Rohtak district, responsible for causing groundwater pollution. Cadmium and lead pollution was prevalent in Rohtak due to electroplating industries, paint industry, automobile sector, and industrial discharge. Bioremediation is one of the suitable techniques that can be used for the treatment of groundwater that involves the use of microorganisms. Efficient use of groundwater resources is necessary for sustainable development.

Assessment of groundwater quality in arid regions utilizing principal component analysis, GIS, and machine learning techniques

Assessing water quality in arid regions is vital due to scarce resources, impacting health and sustainable management.This study examines groundwater quality in Assuit Governorate, Egypt, using Principal Component Analysis, GIS, and Machine Learning Techniques. Data from 217 wells across 12 parameters were analyzed, including TDS, EC, Cl-, Fe++, Ca++, Mg++, Na+, SO4--, Mn++, HCO3-, K+, and pH. The Water Quality Index (WQI) was calculated, and ArcGIS mapped its spatial distribution. Machine learning algorithms, including Ridge Regression, XGBoost, Decision Tree, Random Forest, and K-Nearest Neighbors, were used for predictive analysis. Higher concentrations of Na, K, Ca, Mg, Mn, and Fe were correlated with industrial and densely populated areas. Most samples exhibited excellent or good quality, with a small percentage unsuitable for consumption. Ridge Regression showed the lowest MAPE rates (0.22 % training, 0.26 % in testing). This research highlights the importance of advanced machine learning for sustainable groundwater management in arid regions. Thus, our results could provide valuable assistance to both national and local authorities involved in water management decisions, particularly for water resource managers and decision-makers. This information can aid in the development of regulations aimed at safeguarding and sustainably managing groundwater resources, which are essential for the overall prosperity of the country.

Data-driven evolution of water quality models: An in-depth investigation of innovative outlier detection approaches-A case study of Irish Water Quality Index (IEWQI) model

Recently, there has been a significant advancement in the water quality index (WQI) models utilizing data-driven approaches, especially those integrating machine learning and artificial intelligence (ML/AI) technology. Although, several recent studies have revealed that the data-driven model has produced inconsistent results due to the data outliers, which significantly impact model reliability and accuracy. The present study was carried out to assess the impact of data outliers on a recently developed Irish Water Quality Index (IEWQI) model, which relies on data-driven techniques. To the author's best knowledge, there has been no systematic framework for evaluating the influence of data outliers on such models. For the purposes of assessing the outlier impact of the data outliers on the water quality (WQ) model, this was the first initiative in research to introduce a comprehensive approach that combines machine learning with advanced statistical techniques. The proposed framework was implemented in Cork Harbour, Ireland, to evaluate the IEWQI model's sensitivity to outliers in input indicators to assess the water quality. In order to detect the data outlier, the study utilized two widely used ML techniques, including Isolation Forest (IF) and Kernel Density Estimation (KDE) within the dataset, for predicting WQ with and without these outliers. For validating the ML results, the study used five commonly used statistical measures. The performance metric (R2) indicates that the model performance improved slightly (R2 increased from 0.92 to 0.95) in predicting WQ after removing the data outlier from the input. But the IEWQI scores revealed that there were no statistically significant differences among the actual values, predictions with outliers, and predictions without outliers, with a 95 % confidence interval at p < 0.05. The results of model uncertainty also revealed that the model contributed <1 % uncertainty to the final assessment results for using both datasets (with and without outliers). In addition, all statistical measures indicated that the ML techniques provided reliable results that can be utilized for detecting outliers and their impacts on the IEWQI model. The findings of the research reveal that although the data outliers had no significant impact on the IEWQI model architecture, they had moderate impacts on the rating schemes' of the model. This finding indicated that detecting the data outliers could improve the accuracy of the IEWQI model in rating WQ as well as be helpful in mitigating the model eclipsing problem. In addition, the results of the research provide evidence of how the data outliers influenced the data-driven model in predicting WQ and reliability, particularly since the study confirmed that the IEWQI model's could be effective for accurately rating WQ despite the presence of the data outliers in the input. It could occur due to the spatio-temporal variability inherent in WQ indicators. However, the research assesses the influence of data input outliers on the IEWQI model and underscores important areas for future investigation. These areas include expanding temporal analysis using multi-year data, examining spatial outlier patterns, and evaluating detection methods. Moreover, it is essential to explore the real-world impacts of revised rating categories, involve stakeholders in outlier management, and fine-tune model parameters. Analysing model performance across varying temporal and spatial resolutions and incorporating additional environmental data can significantly enhance the accuracy of WQ assessment. Consequently, this study offers valuable insights to strengthen the IEWQI model's robustness and provides avenues for enhancing its utility in broader WQ assessment applications. Moreover, the study successfully adopted the framework for evaluating how data input outliers affect the data-driven model, such as the IEWQI model. The current study has been carried out in Cork Harbour for only a single year of WQ data. The framework should be tested across various domains for evaluating the response of the IEWQI model's in terms of the spatio-temporal resolution of the domain. Nevertheless, the study recommended that future research should be conducted to adjust or revise the IEWQI model's rating schemes and investigate the practical effects of data outliers on updated rating categories. However, the study provides potential recommendations for enhancing the IEWQI model's adaptability and reveals its effectiveness in expanding its applicability in more general WQ assessment scenarios.

Hybrid WT-CNN-GRU-based model for the estimation of reservoir water quality variables considering spatio-temporal features

Water quality indicators (WQIs), such as chlorophyll-a (Chl-a) and dissolved oxygen (DO), are crucial for understanding and assessing the health of aquatic ecosystems. Precise prediction of these indicators is fundamental for the efficient administration of rivers, lakes, and reservoirs. This research utilized two unique DL algorithms-namely, convolutional neural network (CNNs) and gated recurrent units (GRUs)-alongside their amalgamation, CNN-GRU, to precisely gauge the concentration of these indicators within a reservoir. Moreover, to optimize the outcomes of the developed hybrid model, we considered the impact of a decomposition technique, specifically the wavelet transform (WT). In addition to these efforts, we created two distinct machine learning (ML) algorithms-namely, random forest (RF) and support vector regression (SVR)-to demonstrate the superior performance of deep learning algorithms over individual ML ones. We initially gathered WQIs from diverse locations and varying depths within the reservoir using an AAQ-RINKO device in the study area to achieve this. It is important to highlight that, despite utilizing diverse data-driven models in water quality estimation, a significant gap persists in the existing literature regarding implementing a comprehensive hybrid algorithm. This algorithm integrates the wavelet transform, convolutional neural network (CNN), and gated recurrent unit (GRU) methodologies to estimate WQIs accurately within a spatiotemporal framework. Subsequently, the effectiveness of the models that were developed was assessed utilizing various statistical metrics, encompassing the correlation coefficient (r), root mean square error (RMSE), mean absolute error (MAE), and Nash-Sutcliffe efficiency (NSE) throughout both the training and testing phases. The findings demonstrated that the WT-CNN-GRU model exhibited better performance in comparison with the other algorithms by 13% (SVR), 13% (RF), 9% (CNN), and 8% (GRU) when R-squared and DO were considered as evaluation indices and WQIs, respectively.

Potential risk assessment and occurrence characteristic of heavy metals based on artificial neural network model along the Yangtze River Estuary, China

Pollution from heavy metals in estuaries poses potential risks to the aquatic environment and public health. The complexity of the estuarine water environment limits the accurate understanding of its pollution prediction. Field observations were conducted at seven sampling sites along the Yangtze River Estuary (YRE) during summer, autumn, and winter 2021 to analyze the concentrations of seven heavy metals (As, Cd, Cr, Pb, Cu, Ni, Zn) in water and surface sediments. The order of heavy metal concentrations in water samples from highest to lowest was Zn > As > Cu > Ni > Cr > Pb > Cd, while that in surface sediments samples was Zn > Cr > As > Ni > Pb > Cu > Cd. Human health risk assessment of the heavy metals in water samples indicated a chronic and carcinogenic risk associated with As. The risks of heavy metals in surface sediments were evaluated using the geo-accumulation index (Igeo) and potential ecological risk index (RI). Among the seven heavy metals, As and Cd were highly polluted, with Cd being the main contributor to potential ecological risks. Principal component analysis (PCA) was employed to identify the sources of the different heavy metals, revealing that As originated primarily from anthropogenic emissions, while Cd was primarily from atmospheric deposition. To further analyze the influence of water quality indicators on heavy metal pollution, an artificial neural network (ANN) model was utilized. A modified model was proposed, incorporating biochemical parameters to predict the level of heavy metal pollution, achieving an accuracy of 95.1%. This accuracy was 22.5% higher than that of the traditional model and particularly effective in predicting the maximum 20% of values. Results in this paper highlight the pollution of As and Cd along the YRE, and the proposed model provides valuable information for estimating heavy metal pollution in estuarine water environments, facilitating pollution prevention efforts.

>Water quality prediction of artificial intelligence model: a case of Huaihe River Basin, China

Accurate prediction of water quality contributes to the intelligent management of water resources. Water quality indices have time series characteristics and nonlinearity, but the existing models only focus on the forward time series when long short-term memory (LSTM) is introduced and do not consider the parallel computation on the model. Owing to this, a new neural network called LSTM-multihead attention (LMA) was constructed to predict water quality, using long short-term memory to process time series data and multihead attention for parallel computing and extracting feature information. Additionally, water quality indices have the issues of multiple data types and complex data correlations, as well as missing data and abnormal data problems in water quality data. In order to solve these problems, this study proposes a water quality prediction model called GRA-LMA-based linear interpolation, gray relational analysis and LMA. Two experiments are carried out to verify the predictive performance of the GRA-LMA with the water quality data of the Huaihe River Basin as a case study sample. The first experiment focuses on data processing, including the processing of missing data and abnormal data of water quality data, and the correlation analysis of water quality indices. Linear interpolation is adapted to process the missing data, while a combination of boxplot and histogram is adopted to analyze and eliminate the abnormal data, which is then repaired the abnormal data with linear interpolation. The gray relational analysis is adopted to calculate the correlation between different water quality indices, and water quality indices with high correlation are retained to determine the input variables of the water quality prediction model. The data processing results demonstrate that repairs can be made using linear interpolation without altering the pattern of data change and the model by using the gray relational analysis to reduce the quantity of data it needs as input. In the second experiment, the predictive capacity of GRA-LMA and existing models such as backpropagation neural network (BP), recurrent neural network (RNN), long short-term memory (LSTM), and gate recurrent unit (GRU) was evaluated and compared using different numerical and graphical performance evaluation metrics. Comparative experimental results show that the mean square error of pH, dissolved oxygen, chemical oxygen demand, ammonia nitrogen, electrical conductivity, turbidity, total phosphorus, and total nitrogen of GRA-LMA is reduced to 0.05890, 0.40196, 0.32454, 0.04368, 14.71003, 8.13252, 0.01558, and 0.14345. The results indicate that GRA-LMA has superior adaptability for predicting various water quality indices and can significantly lower the induced prediction error.

Identification of the Contamination Sources by PCBs Using Multivariate Analyses: The Case Study of the Annaba Bay (Algeria) Basin

Persistent Organic Pollutants (POPs), particularly the indicator polychlorinated biphenyls (PCBs), were first quantified in water and sediments of two wadis, Boujemaâ and Seybouse, as well as in the effluents from a fertilizer and phytosanitary production industrial plant (Fertial). Since these contaminated discharges end in Annaba Bay (Algeria) in the Mediterranean Sea, with a significant level of contamination, all the potential sources should be identified. In this work, this task is conducted by a multivariate analysis. Liquid-liquid extraction and gas chromatography/mass spectrometry (GC-MS) methods were applied to quantify seven PCB congeners, usually taken as indicators of contamination. The sum of the PCB concentrations in the sediments ranged from 1 to 6.4 μg/kg dw (dry weight) and up to 0.027 μg/L in waters. Principal component analysis (PCA) and hierarchical cluster analysis (HCA) were used for the multivariate analysis, indicating that the main sources of PCB emissions in the bay are urban/domestic and agricultural/industrial. The outfalls that mostly contribute to the pollution of the gulf are the Boujemaâ wadi, followed by the Seybouse wadi, and finally by the Fertial cluster and more precisely the annex basin of the plant. Although referring to a specific site of local importance, the work aims to present a procedure and a methodological analysis that can be potentially applicable to further case studies all over the world.

Investigation and comparative analysis of ecological risk for heavy metals in sediment and surface water in east coast estuaries of India

The sediments and surface water from 8 stations each from Dhamara and Paradeep estuarine areas were sampled for investigation of heavy metals, Cd, Cu, Pb, Mn, Ni, Zn, Fe, and Cr contamination. The objective of the sediment and surface water characterization is to find the existing spatial and temporal intercorrelation. The sediment accumulation index (I_sed), enrichment index (I_En), ecological risk index (I_EcR) and probability heavy metals (p-HMI) reveal the contamination status with Mn, Ni, Zn, Cr, and Cu showing permissible (0 ≤ I_sed ≤ 1, I_En ˂ 2, I_EcR ≤ 150) to moderate (1 ≤ I_sed ≤ 2, 40 ≤ R_f ≤ 80) contamination. The p-HMI reflects the range from excellent (p-HMI = 14.89-14.54) to fair (p-HMI = 22.31-26.56) in off shore stations of the estuary. The spatial patterns of the heavy metals load index (I_HMc) along the coast lines indicate that the pollution hotspots are progressively divulged to trace metals pollution over time. Heavy metal source analysis coupled with correlation analysis and principal component analysis (PCA) was used as a data reduction technique, which reveals that the heavy metal pollution in marine coastline might originate from redox reactions (FeMn coupling) and anthropogenic sources.