Water Quality Data Sets Research Articles

Semi-Supervised Soft Computing for Ammonia Nitrogen Using a Self-Constructing Fuzzy Neural Network with an Active Learning Mechanism

Ammonia nitrogen (NH3-N) is a key water quality variable that is difficult to measure in the water treatment process. Data-driven soft computing is one of the effective approaches to address this issue. Since the detection cost of NH3-N is very expensive, a large number of NH3-N values are missing in the collected water quality dataset, that is, a large number of unlabeled data are obtained. To enhance the prediction accuracy of NH3-N, a semi-supervised soft computing method using a self-constructing fuzzy neural network with an active learning mechanism (SS-SCFNN-ALM) is proposed in this study. In the SS-SCFNN-ALM, firstly, to reduce the computational complexity of active learning, the kernel k-means clustering algorithm is utilized to cluster the labeled and unlabeled data, respectively. Then, the clusters with larger information values are selected from the unlabeled data using a distance metric criterion. Furthermore, to improve the quality of the selected samples, a Gaussian regression model is adopted to eliminate the redundant samples with large similarity from the selected clusters. Finally, the selected unlabeled samples are manually labeled, that is, the NH3-N values are added into the dataset. To realize the semi-supervised soft computing of the NH3-N concentration, the labeled dataset and the manually labeled samples are combined and sent to the developed SCFNN. The experimental results demonstrate that the test root mean square error (RMSE) and test accuracy of the proposed SS-SCFNN-ALM are 0.0638 and 86.31%, respectively, which are better than the SCFNN (without the active learning mechanism), MM, DFNN, SOFNN-HPS, and other comparison algorithms.

Assessment and prediction of Water Quality Index (WQI) by seasonal key water parameters in a coastal city: application of machine learning models.

The Water Quality Index (WQI) provides comprehensive assessments in river systems; however, its calculation involves numerous water quality parameters, costly in sample collection and laboratory analysis. The study aimed to determine key water parameters and the most reliable models, considering seasonal variations in the water environment, to maximize the precision of WQI prediction by a minimal set of water parameters. Ten statistical or machine learning models were developed to predict the WQI over four seasons using water quality dataset collected in a coastal city adjacent to the Yellow Sea in China, based on which the key water parameters were identified and the variations were assessed by the Seasonal-Trend decomposition procedure based on Loess (STL). Results indicated that model performance generally improved with adding more input variables except Self-Organizing Map (SOM). Tree-based ensemble methods like Extreme Gradient Boosting (XGB) and Random Forest (RF) demonstrated the highest accuracy, particularly in winter. Nutrients (Ammonia Nitrogen (AN) and Total Phosphorus (TP)), Dissolved Oxygen (DO), and turbidity were determined as key water parameters, based on which, the prediction accuracy for Medium and Low grades was perfect while it was over 80% for the Good grade in spring and winter and dropped to around 70% in summer and autumn. Nutrient concentrations were higher at inland stations; however, it worsened at coastal stations, especially in summer. The study underscores the importance of reliable WQI prediction models in water quality assessment, especially when data is limited, which are crucial for managing water resources effectively.

Distributed enhanced multi-objective evolutionary algorithm based on decomposition for cluster analysis in wireless sensor network

Conventional clustering algorithms do not recognize patterns and structures with contradicting objectives in large, distributed datasets. Distributed clustering leverages rapid processing capabilities to allow multiple nodes to work together. This paper proposes a Distributed clustering based on Multiobjective Evolutionary Algorithm by Decomposition (D-MOEA/d) to solve various multiobjective optimization problems in wireless sensor networks (WSNs). In MOEA/d, a multiobjective optimization problem decomposes into several scalar optimization subproblems, each focusing on a distinct objective. Each subproblem is expressed as a clustering problem that uses local data to perform distributed clustering. The proposed method has been extended to achieve improved accuracy in less time by using a smaller feature subset with less redundancy. The Distributed Enhanced MOEA/d (DE-MOEA/d) avoids local optima by achieving diversity in the population using fuzzy-based nearest neighbor selection, sparse population initialization, and evolved mutation operator. This integration improves the accuracy of the clustering process at WSN nodes, ensuring the attainment of well-balanced solutions across multiple optimization criteria in the distributed environment. Average Euclidean and total symmetrical deviations are the two cost functions used to minimize while clustering on the MOEA/d framework. Six real-life WSN datasets are used to assess the performance of the proposed technique: (1) the Delhi air pollution dataset, (2) the Canada weather station dataset, (3) the Thames River water quality dataset, (4) the Narragansett Bay water quality dataset, (5) the Cook Agricultural land dataset and 6) Gordon Soil dataset. The simulation results of both proposed algorithms are compared with Multiobjective distributed particle swarm optimization (DMOPSO) and Distributed K-means (DK-Means). The proposed algorithm DE-MOEA/d performs better in terms of the Silhouette index (SI), Dunn index (DI), Davies–Bouldin index (DBI), and Kruskal–Wallis (KW) statistical test.

High-resolution temperature, salinity and depth data from southeastern Australian estuaries, 2018–2021

Estuaries are the important interface between the land and sea, providing significant environmental, economic, cultural and social values. However, they face unprecedented pressures including eutrophication, harmful algal blooms, habitat loss, and extreme weather due to climate change. Here we present an open access, quality-controlled water quality dataset collected from twelve diverse estuaries spanning 1000 km along the southeastern Australian coastline. Water depth, temperature and salinity data were collected across two years (2018–2021) capturing drought, wildfire and flood periods, using high accuracy Seabird MicroCAT field sensors located within oyster leases. These fully autonomous instruments collected and transmitted data every 10 minutes before downstream quality checking and uploading onto a public website. Simultaneous, high-resolution, longitudinal environmental data collected across multiple estuaries throughout a range of extreme weather events are exceptionally rare in the Southern Hemisphere, yet provide an invaluable resource for the aquaculture industry, researchers and environmental regulators alike.

Lite approaches for long-range multi-step water quality prediction

Forecasting accurate water quality is very important in aquaculture, environment monitoring, and many other applications. Many internal and external factors influence water quality. Therefore, water quality parameters exhibit complex time series characteristics. Consequently, long-range accurate prediction of water quality parameters suffers from poor propagation of information from past timepoints to further future timepoints. Moreover, to synchronise the prediction model with the changes in the time series characteristics, periodic retraining of the prediction model is required and such retraining is to be done on resource-restricted computation devices. In this work, we present a low-cost training approach to improve long-range multi-step water quality prediction. We train a short-range predictor to save training effort. Then, we strive to achieve and/or improve long-range prediction using multi-step iterative ensembling during inference. Experimental results on 9 water quality datasets demonstrate that the proposed method achieves significantly lower error than the existing state-of-the-art approaches. Our approach significantly outperforms the existing approaches in several standard metrics, even in the case of future timepoints at long distances.

The application of transfer machine learning to predict and impute missing sulphate levels in different Acid Mine Drainage treatment plants

An accurately pre-trained stacking ensemble machine learning regressor was used to predict sulphate levels in two other Acid Mine Drainage (AMD) treatment plants using Transfer Learning (TL). The model was trained on the large Central Rand (CR) water quality dataset and was used to predict and impute the sulphate levels in the scanty East Rand (ER) and West Rand (WR) datasets which would not have been sufficient to train ML models from scratch. TL was successfully used to overcome this barrier and rapidly predicted sulphate levels in the East Rand and West Rand plants using the pre-trained model and achieved a high level of accuracy (Mean Squared Error:0.00124, Mean Absolute Error:0.0290 and R2:0.963) for the East Rand plant when comparing the predicted and true sulphate values. No true sulphate values existed for the West Rand plant; however, TL was successful in imputing these missing values and rapidly completed the West Rand dataset by providing the historic sulphate levels. This was possible due to the high degree of similarity between all domains (treatment plants) since they had similar geographic locations, the same treatment process, possessed the same important features and had the same relationships between variables. TL was successful in providing three accurate datasets for AMD sulphate levels, an important accomplishment towards having reliable data for use in design of experiments aimed at recovering valuable resources such as elemental sulphur, gypsum and important metals from AMD.

A Mixed Methods Approach to Understanding the Public Health Impact of a School-Based Citizen Science Program to Reduce Arsenic in Private Well Water.

Exposure to arsenic (As) in well water is a well-documented public health issue for Maine and New Hampshire, as well as for other states in the United States and abroad. Arsenic contamination of well water in these locations is primarily attributed to metasedimentary bedrock that leaches As into groundwater. However, As can also enter groundwater reserves from soils contaminated by the historical use of arsenical pesticides. Approximately half of the households in Maine and New Hampshire rely on private wells, many of which have elevated As. Arsenic exposure has been associated with an increased risk of cancer, cardiovascular disease, reduced infection resistance, and lower intelligence quotient in children. Despite these known health impacts, well water testing and treatment are not universal. We have approached the problem of low well water testing rates in Maine and New Hampshire communities by developing the All About Arsenic (AAA) project, which engages secondary school teachers and students as citizen scientists in collecting well water samples for analysis of As and other toxic metals and supports their outreach efforts to their communities. We assessed this project's public health impact by analyzing student data relative to existing well water quality datasets in both states. In addition, we surveyed private well owners who contributed well water samples to the project to determine the actions taken to mitigate As in well water. Students collected 3,070 drinking water samples for metals testing, and 752 exceeded New Hampshire's As standard of . The AAA data has more than doubled the amount of information available to public health agencies about well water quality in multiple municipalities across both states. Students also collected information about well types and treatment systems. Their data reveal that some homeowners did not know what type of wells they had or whether they had filtration systems. Those with filtration systems were often unaware of the type of system, what the system was filtering for, or whether the system was designed to remove As. Through interviews with pilot survey participants, we learned that some had begun mitigating their exposure to As and other toxic metals in response to test results from the AAA project. A school-based approach to collecting and analyzing private well water samples can successfully reach communities with low testing rates for toxic elements, such as As and other metals. Importantly, information generated through the program can impact household decision-making, and students can influence local and state policymaking by sharing information in their communities. https://doi.org/10.1289/EHP13421.

Predictive Analytics for Water Safety: Data Mining and Supervised Learning in Potability Classification

Water is crucial for survival, especially for consumption, yet its quality is under threat due to human-caused pollution. Contaminated water poses serious health risks, including the transfer of diseases transmitted by water. Therefore, assessing water quality is critical for ensuring its safety for consumption. Data mining and supervised machine learning algorithms can help classify water potability, revealing hidden patterns and correlations between water parameters. This study evaluates the effectiveness of K-Nearest Neighbors (KNN), Naïve Bayes, Support Vector Machine (SVM), and Neural Network methods in categorizing a water quality dataset. The evaluation is aimed at selecting the most accurate procedure, as indicated by the highest accuracy rate. Results show that Neural Network exceeds KNN (81%), Naïve Bayes (63%), and SVM (73%), with a 85% accuracy rate. Keywords : Classification, Data Mining, Supervised Machine Learning, Water Potability

Towards Precise Water Quality Assessment : A Deep Learning Approach with Feature Selection in Smart Monitoring Systems

As water quality concerns intensify, the imperative for accurate monitoring systems grows. This study pioneers a novel approach to precise water quality assessment by integrating deep learning techniques and feature selection in smart monitoring systems. Utilizing k-Nearest Neighbors (KNN), Convolutional Neural Network (CNN), and Inception V3 for classification, along with Random Forest, AdaBoost, and XGBoost for feature selection, the study presents a detailed examination of their performance on water quality datasets. The results showcase notable improvements in both training and testing accuracies for KNN when coupled with Random Forest and varying numbers of estimators. The combination of CNN and AdaBoost exhibits robust performance, underscoring the impact of feature extraction on training and testing accuracies. Inception V3, when integrated with XGBoost, demonstrates nuanced results, emphasizing the significance of feature extraction in enhancing classification outcomes. Specifically, the performance metrics reveal a fusion model using XGBoost and Inception V3 achieving an accuracy of 65.82%, surpassing individual models like Inception V3 (60.05%). Similarly, the combination of AdaBoost and CNN attains a performance of 65.17%, outperforming individual models such as CNN (64.32%). Additionally, the integration of Artificial Neural Networks (ANN) with Random Forest yields a performance of 69.05%, showcasing improvement over standalone ANN (55.79%). The findings underscore the efficacy of deep learning models, particularly when integrated with appropriate feature selection algorithms, in advancing the precision of water quality assessment in smart monitoring systems. This study contributes valuable insights to the field of environmental monitoring, providing a basis for further exploration of synergies between deep learning and feature selection for enhanced accuracy in water quality assessment. The proposed approach holds promise for addressing the critical challenge of precise water quality monitoring in the face of escalating environmental concerns.

Hot and Fresh: Evidence of Climate-Related Suboptimal Water Conditions for Seagrass in a Large Gulf Coast Estuary

Seagrasses have long been a focal point for management efforts aimed at restoring ecosystem health in estuaries worldwide. In Tampa Bay, Florida (USA), seagrass coverage has declined since 2016 by nearly a third (11,518 acres), despite sustained reductions of nitrogen loads supportive of light environments for growth. Changing physical water quality conditions related to climate change may be stressing seagrasses beyond their optimal growth ranges, requiring an assessment to determine if this decline can be linked to climate stress. Three ambient water quality datasets of varying sampling designs and coverage were evaluated to characterize physicochemical environments in Tampa Bay and the potential relationships with seagrass change. Tampa Bay has become hotter and fresher with water temperature increasing by 0.03–0.04 °C per year and salinity decreasing by 0.04–0.06 ppt per year, translating to an increase of 1.3 to 1.7 °C and a decrease of 1.6 to 2.6 ppt over the last 50 years. Additionally, the number of days when temperature was above 30 °C or salinity was below 25 ppt has increased on average across all bay segments by 48 and 37 days, respectively, since 1975. These changes varied spatially and seasonally, with the most dramatic changes observed in the upper bay. Generalized Additive Models provided a weight-of-evidence that recent seagrass declines are somewhat associated with hotter and fresher conditions. Trends in warming and increased precipitation in the region are likely to continue, further creating suboptimal conditions for seagrasses in Tampa Bay. These results should compel resource managers to consider the likelihood that reduced resilience of estuarine resources due to shifting ecological baselines driven by additional climate change drivers will complicate long-standing management paradigms. While conventional management approaches that focus on limiting nutrient loads should be continued, their future effectiveness may be confounded by climate change drivers and warrant additional, complementary interventions and continuous monitoring data to support ecosystem health into the future.

Development of a “big data” groundwater microbial contamination index and spatial comparisons with enteric infection rates in southern Ontario

Big data have become increasingly important for policymakers and scientists but have yet to be employed for the development of spatially specific groundwater contamination indices or protecting human and environmental health. The current study sought to develop a series of indices via analyses of three variables: Non-E. coli coliform (NEC) concentration, E. coli concentration, and the calculated NEC:E. coli concentration ratio. A large microbial water quality dataset comprising 1,104,094 samples collected from 292,638 Ontarian wells between 2010 and 2021 was used. Getis-Ord Gi* (Gi*), Local Moran's I (LMI), and space-time scanning were employed for index development based on identified cluster recurrence. Gi* and LMI identify hot and cold spots, i.e., spatially proximal subregions with similarly high or low contamination magnitudes. Indices were statistically compared with mapped well density and age-adjusted enteric infection rates (i.e., campylobacteriosis, cryptosporidiosis, giardiasis, verotoxigenic E. coli (VTEC) enteritis) at a subregional (N = 298) resolution for evaluation and final index selection. Findings suggest that index development via Gi* represented the most efficacious approach. Developed Gi* indices exhibited no correlation with well density, implying that indices are not biased by rural population density. Gi* indices exhibited positive correlations with mapped infection rates, and were particularly associated with higher bacterial (Campylobacter, VTEC) infection rates among younger sub-populations (p < 0.05). Conversely, no association was found between developed indices and giardiasis rates, an infection not typically associated with private groundwater contamination. Findings suggest that a notable proportion of bacterial infections are associated with groundwater and that the developed Gi* index represents an appropriate spatiotemporal reflection of long-term groundwater quality. Bacterial infection correlations with the NEC:E. coli ratio index (p < 0.001) were markedly different compared to correlations with the E. coli index, implying that the ratio may supplement E. coli monitoring as a groundwater assessment metric capable of elucidating contamination mechanisms. This study may serve as a methodological blueprint for the development of big data-based groundwater contamination indices across the globe.

Spatial Functional Outlier Detection in Multivariate Spatial Functional Data

Multivariate spatial functional data consists of multiple functions of time-dependent attributes observed at each spatial point. This study focuses on detecting spatial outliers in spatial functional data. Firstly, we develop a new method called Mahalanobis Distance Spatial Outlier (MDSO) to detect functional outliers in the data. The method introduces the multivariate functional Mahalanobis semi-distance and multivariate pairwise functional Mahalanobis semi-distance metrics based on the multivariate functional principal components analysis to calculate the dissimilarity between functions at each spatial point. Via simulation, we show that MDSO performs better than the other competing methods. Secondly, MDSO has been extended to detect spatial functional outliers as well. The functional outliers can now be categorized as global or/and local functional outliers. The appropriate number of neighbors and the cut-off point for the degree of isolation are determined via simulation. Finally, we demonstrate the application of the MDSO on a water quality data set obtained from Sungai Klang basin in Malaysia. The results can be used to support the authority in making better decisions on the management of the river basin or other spatial data with time-independent attributes.

Is comprehensive event sampling necessary for constraining process models of water quality? A comparison of high and low frequency phosphorus sampling programs for constraining the HYPE water quality model

Model parameter calibration is an important step in process-based watershed water quality modelling. Calibration is commonly performed using readily available water quality data from long-term monitoring programs that collect samples periodically at a relatively low frequency (i.e. monthly). Higher frequency synoptic and targeted event data is becoming available from more intensive monitoring programs, yet there remains little consensus on whether relatively lower frequency data are sufficient to constrain process-based watershed water quality models. To investigate the effects of using water quality data from differing sampling regimes for calibration, we use an implementation of HYPE (HYdrological Predictions for the Environment), a process-based watershed water quality model, in Southern Ontario, Canada. HYPE was calibrated with two stream water quality datasets, one associated with the long-term Ontario Provincial (Stream) Water Quality Monitoring Network and characterized by routine approximately monthly sampling frequency, and one associated with the Multi-Watershed Nutrient Study and characterized by semi-regular synoptic sampling during baseflow and at increased frequency (∼4 to 8 h) during event flow. Performances varied widely between sites, with validation ranges of daily predictions having Nash-Sutcliffe Efficiencies (NSE) from > −1 to 0.48 for flow and > −1 to 0.98 for total Phosphorus. Medians for daily data during validation ranged from 0.15 to 0.24 for flow and from 0.19 to 0.36 for total phosphorus. We found that model performance and simulated total phosphorus loads were similar for the two calibration datasets, potentially suggesting that the dataset with approximately monthly water quality sampling was adequate to constrain the HYPE model. We attribute this to a combination of similar levels of statistical variability in the calibration datasets and prior knowledge in the HYPE model structure. Furthermore, while model performance was similar when HYPE was calibrated with datasets of differing sampling strategies, there was a large range in model performance within each modelling domain. Results suggest that this range in performance could be due in part to poor representation of cold-weather processes in the HYPE model, as sites with higher mean annual temperature and fewer freezing days had better model performance.

Deep learning-based water quality index classification using stacked ensemble variational mode decomposition

Water is crucial to human survival in general, and determining the WQI (water quality index) is one of the primary aspects. The existing water quality classification models are facing various challenges and gaps that are impeding their effectiveness. These challenges include limited data availability, the intricate nature of water systems, spatial and temporal variability, non-linear relationships, sensor noise, and error, interpretability, and explainability. It is imperative to address these challenges to improve the accuracy and efficacy of the models and to ensure that they continue to serve as reliable tools for monitoring and safeguarding water quality. To solve the issues, this paper proposes a Stacked Ensemble efficient long short-term memory (StackEL) model for an efficient water quality index classification. At first, the raw input data is pre-processed to rescale the input data using data normalization and one-hot encoding. After that, the process known as variational mode decomposition (VMD) is applied to get at the intrinsic mode functions (IMFs). Consequently, feature selection is performed using an extended coati optimization (EX-CoA) algorithm to select the most significant attributes from the feature selection. Here, publicly available datasets, namely the water quality dataset from Kaggle, are used for classification and performed using are used to perform the Stacked Ensemble efficient long short-term memory (StackEL) classification process effectively. To further perfect the proposed prediction model, the Dwarf Mongoose optimization (DMO) method is implemented. Several measures of effectiveness are examined. When compared to other existing models, the suggested model can achieve a high accuracy of 98.85% of the water quality dataset.

Incorporation of water quality index models with machine learning-based techniques for real-time assessment of aquatic ecosystems

Water quality index (WQI) is a well-established tool for assessing the overall quality of fresh inland-waters. However, the effectiveness of real-time assessment of aquatic ecosystems using the WQI is usually impacted by the absence of some water quality parameters in which their accurately in-situ measurements are impossible and face difficulties. Using a rich water quality dataset spanned from 1980 to 2023, we employed four machine learning-based models to estimate the British Colombia WQI (BCWQI) in the Lake Päijänne, Finland, without parameters like chemical oxygen demand (COD) and total phosphorus (TP). Measurement of both COD and TP is time-consuming, needs laboratory equipment and labor costs, and faces sampling-related difficulties. Our results suggest the machine learning-based models successfully estimate the BCWQI in Lake Päijänne when TP and COD are omitted from the dataset. The long-short term memory model is the least sensitive model to exclusion of COD and TP from inputs. This model with the coefficient of determination and root-mean squared error of 0.91 and 0.11, respectively, outperforms the support vector regression, random forest, and neural network models in real-time estimation of the BCWQI in Lake Päijänne. Incorporation of BCWQI with the machine learning-based models could enhance assessment of overall quality of inland-waters with a limited database in a more economical and time-saving way. Our proposed method is an effort to replace the traditional offline water quality assessment tools with a real-time model and improve understanding of decision-makers on the effectiveness of management practices on the changes in lake water quality.

A point-of-use drinking water quality dataset from fieldwork in Detroit, Michigan

Drinking water quality sensor technology has rapidly advanced, facilitating the collection of rich datasets and real-time analytics. However, sensors have not yet been widely applied to monitor drinking water quality in premise plumbing. Richer quality of data in premise plumbing offers an improved understanding of the quality of drinking water present at the point-of-use. In this paper, online drinking water quality sensor nodes were temporarily installed in twenty-four homes in Detroit, Michigan. The water quality sensor nodes took measurements of five drinking water quality parameters every five minutes for four weeks. Additionally, free chlorine and lead were sampled periodically within each home. Together, these data make up a dataset that captures drinking water quality over time in a legacy city with an oversized drinking water system. This dataset offers more frequent measurements amongst more sample homes than are typically available in premise plumbing or at the tap. The data can be used to investigate temporal trends in drinking water quality, including diurnal patterns and anomaly detection. Additionally, this dataset could be utilized to evaluate water quality in comparison with other cities.

Filling up the water quality database to assess the water quality levels and self-cleaning capacities

The one of water resource’s functions is the self-renewable capacity about quality, quantity and energy. To assess the level or fluctuat trend of water resource functions, a complete set of continuous and representative data is needed. The water quality monitoring of environment management is implementing continuous in Ho Chi Minh City. Because of many causes, the water quality data set is asynchronous by years and observation locations. It causes the many difficulties to researches, studies as the assessment of water’s self-cleaning capacity. With the support of two interpolations equations (the Highest order polynomial fitting Curve Function (HopCEF) and Multivariable Regression Correlative Function (MrCEF)) and GIS tool, the water quality and self-cleaning capacity distributions are zoned/mapped with the fully and continuously data set. The research of water quality and self-cleaning capacity is implemented with the filled quality data set period 2012 – 2022 for urban inner canal system in HCMC. Based on the assessment’s results and the mapped distribution of levels, the fluctuations the quality and self-cleaning capacity of inner canal system are made high visualization. And they are a valuable basis for planning and decision making on the solutions of management agencies in HCMC.

Using machine learning models to estimate Escherichia coli concentration in an irrigation pond from water quality and drone-based RGB imagery data

The rapid and efficient quantification of Escherichia coli concentrations is crucial for monitoring water quality. Remote sensing techniques and machine learning algorithms have been used to detect E. coli in water and estimate its concentrations. The application of these approaches, however, is challenged by limited sample availability and unbalanced water quality datasets. In this study, we estimated the E. coli concentration in an irrigation pond in Maryland, USA, during the summer season using demosaiced natural color (red, green, and blue: RGB) imagery in the visible and infrared spectral ranges, and a set of 14 water quality parameters. We did this by deploying four machine learning models – Random Forest (RF), Gradient Boosting Machine (GBM), Extreme Gradient Boosting (XGB), and K-nearest Neighbor (KNN) – under three data utilization scenarios: water quality parameters only, combined water quality and small unmanned aircraft system (sUAS)-based RGB data, and RGB data only. To select the training and test datasets, we applied two data-splitting methods: ordinary and quantile data splitting. These methods provided a constant splitting ratio in each decile of the E. coli concentration distribution. Quantile data splitting resulted in better model performance metrics and smaller differences between the metrics for both the training and testing datasets. When trained with quantile data splitting after hyperparameter optimization, models RF, GBM, and XGB had R2 values above 0.847 for the training dataset and above 0.689 for the test dataset. The combination of water quality and RGB imagery data resulted in a higher R2 value (>0.896) for the test dataset. Shapley additive explanations (SHAP) of the relative importance of variables revealed that the visible blue spectrum intensity and water temperature were the most influential parameters in the RF model. Demosaiced RGB imagery served as a useful predictor of E. coli concentration in the studied irrigation pond

Surface water quality modelling with data scarcity in semi-enclosed coastal regions encompassed distributed islands

Water quality variation in semi-enclosed urban coastal areas with different pollutant sources is a substantial issue. Pollutant entrapment has a significant impact on the lives of the local people. Surface water quality modelling often requires large datasets covering bathymetry, fluid and pollutants boundary conditions, sources and sinks. This is particularly challenging in regions with complex features and poor data availability, such as coastal water bodies featuring a large number of widely distributed islands and estuaries. In this paper, we developed a model of surface water quality and hydrodynamics for Ha Long Bay, in the North of Vietnam and includes 1969 islands, for a one-year period using a water quality dataset obtained for this study. The model utilized extracted bathymetric and geometric data from Admiralty Charts and Admiralty Tide Tables. Water quality coupled with the TELEMAC Model (WAQTEL) Biomass Module was employed to predict pollutant transport in the domain using point and diffused sources while field studies were conducted to collect data for the setting up and calibration of the water quality model. Thirty scattered sampling points were selected, and the water quality parameters were measured during two campaigns. The predicted water level and velocity values matched the local observation data well with a small error (RMSE = 0.19 and 0.16). Both NO3−-N and PO43--P were high near the shoreline and decrease gradually offshore. The maximum concentrations of NO3−-N and PO43--P reached 0.476 mg/L and 0.048 mg/L at the end of 2021 with the RMSE = 0.13 and 0.011, respectively. The levels of NO3−-N and PO43--P and their distributions showed that Ha Long Bay was eutrophic even during the COVID-19 lockdown period.

Physics-guided spatio–temporal neural network for predicting dissolved oxygen concentration in rivers

The prediction of river water quality is key in water resource management. Data-driven machine learning models have been widely used for predicting river water quality. However, these models seldom consider the physical mechanisms of water quality variation, which degrades the accuracy and stability of the prediction results. Hence, we develop a physics-guided spatio–temporal neural network (PGSTNN) model to predict a critical parameter for water quality assessment, i.e. dissolved oxygen. Physical information regarding spatio–temporal interactions in a hydrological network is explicitly considered to construct the architecture of PGSTNN. Two physical rules of dissolved oxygen variation (i.e. Henry’s law and power-scaling law) are established for the loss function of PGSTNN to guarantee the physical consistency of the prediction results. Experiments on the 2020–2021 water quality dataset in Atlanta, USA show that PGSTNN outperforms seven baseline neural network models in terms of prediction accuracy and stability. PGSTNN typically brings at least 10% accuracy (e.g. root mean square error and mean absolute error) improvement over the comparison methods. The proposed PGSTNN may not only improve the emergency response ability of water resource management, but also provide useful ideas for integrating scientific knowledge with machine learning.