literature review for water quality

Literature review: Water quality and public health problems in developing countries

• Article contents
• Figures & tables
• Supplementary Data
• Peer Review
• Reprints and Permissions
• Cite Icon Cite
• Search Site

Eni Muryani; Literature review: Water quality and public health problems in developing countries. AIP Conf. Proc. 23 November 2021; 2363 (1): 050020. https://doi.org/10.1063/5.0061561

Download citation file:

• Ris (Zotero)
• Reference Manager

Water’s essential function as drinking water is a significant daily intake. Contamination by microorganisms (bacteria or viruses) on water sources and drinking water supplies is a common cause in developing countries like Indonesia. This paper will discuss the sources of clean water and drinking water and their problems in developing countries; water quality and its relation to public health problems in these countries; and what efforts that can be make to improve water quality. The method used is a literature review from the latest journals. Water quality is influenced by natural processes and human activities around the water source Among developed countries, public health problems caused by low water quality, such as diarrhea, dysentery, cholera, typhus, skin itching, kidney disease, hypertension, heart disease, cancer, and other diseases the nervous system. Good water quality has a role to play in decreasing the number of disease sufferers or health issues due to drinking and the mortality rate. The efforts made to improve water quality and public health are by improving WASH (water, sanitation, and hygiene) facilities and infrastructure and also WASH education.

Sign in via your Institution

Citing articles via, publish with us - request a quote.

Sign up for alerts

• Online ISSN 1551-7616
• Print ISSN 0094-243X
• For Researchers
• For Librarians
• For Advertisers
• Our Publishing Partners  
• Physics Today
• Conference Proceedings
• Special Topics

pubs.aip.org

• Privacy Policy
• Terms of Use

Connect with AIP Publishing

This feature is available to subscribers only.

Sign In or Create an Account

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 29 March 2024

Reliable water quality prediction and parametric analysis using explainable AI models

• M. K. Nallakaruppan 1 ,
• E. Gangadevi 2 ,
• M. Lawanya Shri 1 ,
• Balamurugan Balusamy 3 ,
• Sweta Bhattacharya 1 &
• Shitharth Selvarajan 4 , 5  

Scientific Reports volume  14 , Article number:  7520 ( 2024 ) Cite this article

258 Accesses

Metrics details

• Electrical and electronic engineering
• Engineering

The consumption of water constitutes the physical health of most of the living species and hence management of its purity and quality is extremely essential as contaminated water has to potential to create adverse health and environmental consequences. This creates the dire necessity to measure, control and monitor the quality of water. The primary contaminant present in water is Total Dissolved Solids (TDS), which is hard to filter out. There are various substances apart from mere solids such as potassium, sodium, chlorides, lead, nitrate, cadmium, arsenic and other pollutants. The proposed work aims to provide the automation of water quality estimation through Artificial Intelligence and uses Explainable Artificial Intelligence (XAI) for the explanation of the most significant parameters contributing towards the potability of water and the estimation of the impurities. XAI has the transparency and justifiability as a white-box model since the Machine Learning (ML) model is black-box and unable to describe the reasoning behind the ML classification. The proposed work uses various ML models such as Logistic Regression, Support Vector Machine (SVM), Gaussian Naive Bayes, Decision Tree (DT) and Random Forest (RF) to classify whether the water is drinkable. The various representations of XAI such as force plot, test patch, summary plot, dependency plot and decision plot generated in SHAPELY explainer explain the significant features, prediction score, feature importance and justification behind the water quality estimation. The RF classifier is selected for the explanation and yields optimum Accuracy and F1-Score of 0.9999, with Precision and Re-call of 0.9997 and 0.998 respectively. Thus, the work is an exploratory analysis of the estimation and management of water quality with indicators associated with their significance. This work is an emerging research at present with a vision of addressing the water quality for the future as well.

Similar content being viewed by others

A proficiency assessment of integrating machine learning (ML) schemes on Lahore water ensemble

Nazish Shahid

Source discrimination of mine water based on the random forest method

Zhenwei Yang, Hang Lv, … Xinyi Wang

Computational assessment of groundwater salinity distribution within coastal multi-aquifers of Bangladesh

Mehdi Jamei, Masoud Karbasi, … Zaher Mundher Yaseen

Introduction

The major part of our earth comprises water and it is extremely important for the survival of all humans and animal species. Water makes up over 326 cubic metres of the planet’s surface, which is almost 71% of its total area out of which 97% is seawater. Only 0.5 percentage of the drinkable water on earth is accessible, while the remaining 2.5 percentage is either trapped in glaciers, polar ice caps, the atmosphere, on soil, is polluted, or lies beneath the earth’s surface far beyond human reach. If the global water supply is 100 L, consequently the amount of drinking water would be only 0.003 L, which is just a teaspoon. Therefore, the management and preservation of drinking water is regarded as a top priority. It is the most critical issue for mankind to address given the extremely limited amount of water that is accessible for use. The quantum of water around the world is represented in Table 1 .

Water is a common and crucial resource shared among all humans, animals, and plants and is a necessity for all species. Each one of these species has its own respective needs for water quality. Total Dissolvable Solids (TDS) of soft water for human consumption range from the best quality stated, which is between 50 mg/dL and 150 mg/dL. Between 150 mg/dL and 300 mg/dL is the next level that can be applied to humans. The plants need water that is between 700mg/dL and 800mg/dL. The animals, especially cattle consume water around the quality of 1000 mg/dL. It is thus evident from all these observations that water quality management is essential to ensure sustainability and a healthy life on Earth. The impact of water quality prediction is crucial at a global level for many reasons. First of all, to get clean and safe water is a basic human necessity and water quality prediction aids to guarantee the availability of potable water for societies worldwide. Water quality is related to public health as polluted water may cause waterborne diseases which could affect millions of humans globally. A sustainable environment is an important aspect of human well-being by preserving ecosystems and biodiversity. The significance of water quality assessment is profound and intricate by various organizations globally. The WHO (World Health Organization) , UNEP (United Nations Environment Programme), EPA (United States Environmental Protection Agency), EEA (European Environment Agency), IWA (International Water Association) and WEF (Water Environment Federation) are fanatical for water quality assessment and addressing the mitigation strategies for water quality challenges. Water quality creates impact on public health globally and resulting in dissemination of waterborne diseases like typhoid, dysentery, cholera, dengue and malaria and cause substantial risks worldwide.

The advancement in computing technologies and artificial intelligence have elevated the standards of water quality assessments 1 . Measurements and estimations about the quality of the water have become easier to calculate and accurate, especially with the development of Industry 4.0 standards and Internet of Things (IoT) sensors. With the integration of IoT sensors, AI solely serves as a supporting tool to automate water quality checks 2 . Classification and Regression models based on machine learning help in determining the water quality. Depending on the outcomes, classification results tend to be binary or multi-classified. Real-time sensor data are collected, given feature labels, and then classified based on the importance of the feature labels. Earlier, these measurements used to be carried out with fuzzy-based decision support systems 3 with subjective decision-making models. AI development has made it possible to classify and analyse quality aspects quantitatively. The accuracy of the water quality assessment has been validated using various performance metrics like accuracy, precision, recall, and f1-score. AI models also support such quantitative analysis, classification of water sources, and prediction of drinkable water as well as identifying the mixing of bouyant pollutants in water sources 4 .

Despite its success in automating tasks and making water quality predictions using diverse models, the AI models lack transparency and are considered black-box where the decisions are derived but the reasoning behind such decisions is not revealed. The present generation validation frameworks for water quality management need justifiability, transparency and explainability, which is possible to be rendered by Explainable AI (XAI) based systems. XAI is a technology that is white-box and answers the uncertainty related to the classification and regression problems of AI. XAI applies a model-agnostic approach, where the machine learning models can be treated independently for interpretation. Additionally, XAI discusses how the model is chosen, how it works, and how it performs categorization. Through the assessment of a problem’s feature weights, XAI also can determine a feature’s relevance. This clarifies how a feature value relates to a certain target class classification. As an example, XAI uses models like Partial Dependency Plots (PDP) 5 , which describes the relationship between the features using lasso functions. This model may identify the linear relationship between two characteristics of water quality data and explain their correlation. In XAI, models like Local Interpretable Model agnostic Explainer(LIME), explain the relationship between a single feature and relevant others in local surrogacy. This infers that, except for the one-row value of the dataset, it is possible to relate a target attribute to the other independent variables. LIME in this regard can be used to explain the target classification for a single row instance about the water quality 6 . In the proposed work, XAI, which employs both local and global surrogates, includes SHAPELY. The model offers a solution that takes into account the importance of each feature in determining the target as well as the dependency between features, the relationship between features, and the explanation of decisions through a variety of plots, including force plots, summary plots, dependency plots, and decision plots. The framework is very adaptable and capable of giving a thorough explanation of the characteristics of the water quality and how they affect the classification of the water quality.

Advantages of the proposed model

Explainable AI plays an important role in improving the interpretability of predictions made by machine learning models. More transparent predictions are generated by these models. In the proposed approach, the authors have employed LIME and SHAP to interpret predictions achieved from machine learning, which identifies inputs as an important metric for selecting the features. By applying the XAI approach, the proposed model provides deep insights into the features and allows informed decision-making in water management processes.

Contributions of the paper

The following points describe the contribution of the proposed work.

The proposed work offers a comprehensive analysis and white-box description of the classification problem for water quality.

The framework incorporates extensive pre-processing of the dataset to ensure it fit to be fed into the XAI model.

Imputation of missing data is carried out to increase the accuracy of the findings.

The proposed work ensures achievement of most significant features, identification of the feature importance, feature dependencies, and feature weights, that enable optimized classification of water quality dataset.

The proposed approach employs both model-based and model-agnostic interpretations, using model-based ML implementations and model-agnostic XAI implementations.

Organization of the paper

Section “ Introduction ” of the paper introduces the problem of the research paper with the description of the unique contributions. Section. Introduction ” also describes the literature review of the related problems on water quality, in related works subsection, with an exhaustive survey of the various applications and case studies pertaining to water quality management using AI and machine learning approaches. Section “ System model and architecture ” describes the methods applied in the proposed work with the implementation of the mathematical model with the algorithm of the proposed work. Section “ Results ” describes the results of various ML and XAI models with relevant tables and graphs. Section “ Discussion ” provides the comparative analysis of the results with a discussion of challenges and solutions of the proposed work. Section “ Conclusion ” concludes the paper with future directions.

Related works

Lu et al. 7 proposed the central environmental protection inspection (CEPI), which was implemented and the causes of transboundary water contamination were investigated. The triple difference technique (DDD) was used to assess how the CEPI affected pollution and the results to determine how significantly water pollution was decreased as well as the significance of CEPI laws for addressing transboundary pollution. Halder et al. 8 , the Turag River’s neighbouring communities are suffering from major health problems as a result of water contamination. For the sustainability of household and aquatic life, the river’s water quality was unsuitable. The study noted that the threshold values for turbidity, total dissolved solids (TDS), chloride (CL-), chemical oxygen demand (COD), carbon dioxide (CO2), and biochemical oxygen demand (BOD) are higher than the standard permissible limits, which may result in health problems like respiratory illnesses, diarrhoea, cholera, dengue, malaria, anaemia, and skin problems. A study evaluating metal pollution management and mitigation tactics on soil and water was presented by Wang et al. 9 . In this study, the remediation of metal contamination from water and soil utilising chemical, physical, and biological approaches was discussed. In this study, the current methods for reducing heavy metal pollution of the soil and water are examined. Elehinafe et al. 10 discussed the importance of water contamination and examined the main cause of water scarcity. The proposed work discussed the effect of hazardous chemicals on the water, including pesticides, heavy metals, and micro-pollutants. This study outlined the numerous technologies that are currently available to eliminate hazardous materials and provide sustainable clean water resources. Mu et al. 11 proposed a solution for the investigation into farmers’ readiness to implement Rural Water Pollution Control (RWPC). This study examines farmers’ viewpoints to improve the quality of life for locals who reside in rural regions and avoid water contamination. To analyse the contributions of contaminants, Wang et al. 12 developed a unique contaminant flux variable model for river water quality assessment. The framework effectively identified the sources of pollution and evaluated the efficacy of projects designed to reduce water pollution. Zadeh et al. 13 proposed WQPs for estimating chemical oxygen demand and biochemical oxygen demand using the MKSVR algorithm. PSO algorithm is used for solving optimization problems. The multiple kernel support vector regression (MKSVR) is compared with SVR and Random Forest Regression and achieves a better accuracy level for BOD prediction. Nagaf et al. 14 presented a framework for assessing the WQI values based on the NSF guidelines. This framework uses four data-driven models such as EPR, M5 MT, GEP and MARS for predicting WQI values in the Karun River. The classification uses 12 water quality parameters and missing values were extracted from the image analysis. Zadeh et al. 15 proposed a model that utilizes gene expression programming, evolutionary polynomial regression, and model trees for predicting WQPs. The biochemical oxygen demand, dissolved oxygen and chemical oxygen demand are used for estimation with nine parameters. The gamma test is used for determining important parameters. Najaf et al. 16 proposed a water quality predicting framework for estimating the water quality index in the Hudson River based on Canadian Council of Ministers of the Environment (CCME) guidelines. The four artificial intelligence techniques M5 MT, Multivariate Adaptive Regression Spline, Evolutionary Polynomial Regression and Gene Expression Programming are used with Landsat 8 OLI-TIRS images. The results proved that the MARS technique achieved the best outcome compared to other models.

Chowdhury et al. 17 emphasized the sources of water contamination which are caused by densely populated industrial areas that are located close to water bodies. The main causes of water contamination are dangerous chemicals and heavy metals. Farmers’ pre-owned pesticides, including different types of carbamate and organophosphorus pesticides, are the main causes of water contamination on agricultural grounds as per the study. Ahivar et al. 18 examined the use of heavy metal pollution indices (HPIs) in soil, water, and sediments. For assessing metal contamination, HPI is considered a crucial instrument. Each method’s pollution index is assessed to interpret the pollution levels. The selection of HPIs based on the parameters and standards for evaluating the quality of the water and soil is offered. Chen et al. 19 presented a study by used various mathematical and statistical approaches to check the quality of water. The factors indicating the water pollution and the seasonal characteristics are evaluated to reduce the river water pollution. The Principal Component Analysis, Cluster Analysis, Network Analysis and Co-Occurrence Analysis were carried out to find the potential source of river water pollution. Fan et al. 20 examined the quality of water using several mathematical and statistical techniques. To lessen river water pollution, the variables implicating contamination and the seasonal traits are assessed. To identify a likely cause of river water pollution, the Principal Component Analysis, Cluster Analysis, Network Analysis, and Co-Occurrence Analysis were performed. Wang et al. 21 formulated the performance indices for explaining the Water-Energy-Pollution nexus (InWEP) effects of scales. The Nexus Pressure Index (NPI) and Nexus Coupling Index (NCI) were used to represent the pollution pressure and the interacted relations. The factors for InWEP were analysed using the Structural Equation Model (SEM) considering four objects namely enterprises, countries, industrial zones and cities. The performance of InWEP was evaluated for the performance metrics - efficiency, structure and location. To evaluate the quality of groundwater surrounding nearby areas in an industrial metropolis, Asomaku 22 evaluated the water pollution indices. Nine samples from three landfills are used in the analysis of the groundwater’s chemical and metal characteristics. The study in Balaram et al. 23 explored many elements that have an impact on water quality, including climate change, industry, aquaculture, mining, and agriculture. For the quantitative and qualitative evaluation of hazardous metals, metal species, isotopes, and other contaminants that are present in water, various ICP-MS techniques are applied. Yuan et al. 24 proposed a water quality monitoring framework using biological sensors for water quality assessment. Borzooei et al. 25 presented a study to estimate the frequency weather events that creates impact on waste water assessment. The Time series data mining approach is used for categorizing the dry and wet weather events. Noori et al. 26 presented a report on decline of groundwater recharge in Iran. The study presents the average amount of ground water recharge is more than the annual runoff 4 utilized WCSPH (A weakly compressible smoothed particle hydrodynamics) model for simulating the near-shore hydrodynamics. The study conducted experimental and numerical evaluation for detecting the causes for mixing the buoyant pollutants in coastal water source. Yeganeh-Bakhtiar 27 presented a framework using MOS (Model Output Statistics) for establishing the statistical relationships among predicator and predicant.

When evaluating water quality using factors like toxicity and pollutants, computer vision and biological sensor systems are utilised in tandem. To retrieve the important data from images taken by a microscope, a microfluidic chip with sensors is utilised. This chip monitors water samples. Figure 1 describes various factors causing water pollution in smart cities including construction activities, atmospheric deposition, natural factors, municipal wastewater, stormwater runoff, incorrect waste disposal, industrial discharges, agricultural runoff, and municipal wastewater. Jeihouni et al. 28 implemented and compared five data mining techniques, including the Ordinary Decision Tree (ODT), Random Forest (RF), Chi-square Automatic Interaction Detector (CHAID), Iterative Dichotomiser 3 (ID3), and Random tree, to identify high-quality water zones. Eight parameters are used in the evaluation process while deriving rules. Compared to the remaining models, the RF performed well, with an accuracy rate of 97.10%. Lee et al. 29 implemented a framework for evaluating the quality of groundwater utilising a Self-Organizing Map (SOM) technique and fuzzy c-means clustering (FCM) was given. The two methods are employed to describe the complex nature of groundwater. SOM employed 91 neurons to categorise 343 groundwater samples, and FCM grouped the water sources into three groups. Agarwal et al. 30 proposed AI based water evaluating technique to predict the water quality index using Particle Swarm Optimization (PSO), Naïve Bayes Classifier (NBC), and Support vector machine (SVM). PSO was used in this regard for optimizing the classifiers wherein the PSO-optimized NBC obtained 92.8% accuracy and PSO-optimized SVM obtained 77.60% accuracy. Table 3 illustrates various existing state-of-art techniques proposed for assessing water quality, its advantages and research gaps.

Figure 1 illustrates the factors causing water pollution. The factors includes Industrial discharges, agricultural runoff, municipal waste water, storm water, improper waste disposal, oil spills and chemical spills, construction wastages, and atmospheric deposition. The factors are very crucial to protect public health and ecosystem , sustainability development, creating public awareness and for pollution prevention.

Factors causing water pollution.

Figure 2 depicts the required physical parameters such as Temperature, Turbidity, Conductivity, Odour and Color represented in percentage, for evaluating the quality of water. Examining the physical parameters is essential for identifying the potential hazards that leads to poor water quality and for preventing ecosystem health.

Physical Parameters.

Figure 3 depicts the necessary chemical parameters, such as pH, Dissolved Oxygen (DO), Total Dissolved Solids (TDS), Nutrients (nitrogen and phosphorus), Total Suspended Solids (TSS), Heavy Metals, and Organic Matter (OM), as well as Chemical Oxygen Demand (COD) and Biochemical Oxygen Demand (BOD) with percentages, that must be measured in order to assess the water’s quality.

Chemical parameters.

Figure 4 presents various supervised learning models for estimating water quality, including Random Forest, Support Vector Machine (SVM), Decision Trees, Neural Networks, and Gradient Boosting Approaches like XGBoost and AdaBoost.

Supervised learning models.

Figure 5 represents various unsupervised learning models such as Principal Component Analysis, Cluster Analysis and Self-Organizing Maps (SOM) for addressing the quality of the water. PCA is a dimensionality reduction approach mainly utilized for analyzing the high dimensional datasets. Cluster analysis techniques are used primarily for grouping water samples based on similarities. SOM technique is principally used for organizing the water quality data.

Unsupervised learning models.

Figure 6 highlights the various Hybrid ML models such as ensemble models with Reinforcement Learning (RL) for addressing the evaluation of quality of water. The various machine learning models can be verified based on the applications, parameters in order to determine the quality of the water, dataset size and its quality based on the assessment of the performance metrics.

Hybrid ML models.

The motivation for the proposed research, along with the research gap analysis with similar existing research works is discussed as per Table 2 . The comparative analysis and research of similar existing works are presented in Table 3 . These two discussions provide a comprehensive understanding of the requirements, that are essentially required in the design of the proposed system and implementation.

Table 3 refers to similar literature review of various models of machine learning such as DT,RF,DCF, SVM, and so on. This table also discusses about various deep learning models such as, Artificial Neural Networks (ANN), Probablistic Neural Network (PNN), Convolution Neural Networks (CNN) and statistical regression models such as Auto-Regression in Moving Average(ARIMA). This table discusses the the research gaps identified and enhanced in the proposed work. These models were mostly numerical evaluations with regression analysis. The proposed model and the system is classifier which deploys XAI framework, to discuss the impact of parameters, that determine the portability of the water with end user perspective. This is towards achieving environmental sustainability on water conservation and harvesting.

Statement of objectives

The proposed work offers a comprehensive analysis and white-box description of the classification problem for water quality . The framework incorporates extensive pre-processing of the dataset to ensure it fits into the XAI model. Imputation of missing data is carried out to increase the accuracy of the findings. The proposed work ensures the achievement of the most significant features, identification of the feature importance, feature dependencies, and feature weights, that enable optimized classification of the water quality dataset. The proposed approach employs both model-based and model-agnostic interpretations, using model-based ML. Donnelly et al. 46 implementations and model-agnostic XAI implementations. The quality of water is greatly challenged by innumerable influencing factors. These factors vary from condition to condition and place to place. For example, Microplastics (MP) are emerging pollutants in the marine environment with potential toxic effects on littoral and coastal ecosystems 47 and as well as identifying the mixing of bouyant pollutants in water sources 4 . The laboratory evaluations show the presence of polyethene (PE) particles in the waves of the ocean with wave steepness Sop of 2–5%. The transportation of which could cause severe water pollution on the seashores 48 .These measurements require quantification and feature analysis when it is evaluated with AI. This is where the XAI plays a vital role in measuring the order and degree of the pollutants causing the quantifiable pollution in the water.

Case studies

Importance of XAI in Water Quality Assessment: The following case studies delineate the advent of the potential impact of XAI, with a groundbreaking revolution in water quality assessment.

Case Study 1: Pollution of Ganges 49 This case study emphasises the Ganga River pollution issue in India, which has an extremely detrimental impact on humans and the entire ecosystem. The Ganga River is polluted by industrial, animal, and human waste. The main source of pollutants is industrial rubber waste, followed by leather and plastic manufacturers who dump their untreated wastewater into the river. The Ganga Action Plan was developed by the Indian government to combat Ganga pollution. This implies the need for the reinforcement of environmental restrictions to improve river quality.

Materials and methods

An effective policy for health protection should thus emphasize providing access to safe drinking water regardless of social and economic diversity. In some places, it is evident from previous studies that investments in access to clean water and sanitation yield economic benefits for any country. It is a significant aspect of eco-friendly health and public safety, as it regulates the appropriateness of water for numerous purposes, such as drinking, agriculture, industry, and recreational purposes. The important key indicators related to water quality are its physical, chemical, and biological characteristics and its sources of pollution. The dependent target class is potability. The other independent features are pH value, hardness, solids (Total Dissolved Solids-TDS), Chloramines, sulfate, conductivity, organic carbon, trihalomethanes, and turbidity. Water’s potability indicates its purity and safety for ingestion. The parameters used and their WHO limits, the hyper-parametric analysis are listed in Table 4 , and the feature description of parameters are listed in Table 5 .

XAI framework facilitates transparent and interpretable explanations of the outcome generated by the ML algorithm-based frameworks. XAI can thus be applied in the present context of water quality assessment to ensure accurate decision-making, thereby, enabling trustworthiness, enhancement of transparency and interpretability of the behaviour of the model.

Hydro-climatic application

XAI framework can be used to solve Hydro-Climatic problems 50 with diverse spatio-temporal scales. XAI is utilized to unveil the nonlinear correlative causes, in which the performance of the model is enhanced. It enables the users to discover new knowledge and further easily understand the rationale behind the decision outcomes.

Groundwater potential predictions

XAI approach can explain the decisions made by ML models for groundwater potential prediction. The user can easily interpret the outcomes and further comprehend the underlying for an outcome in the realm of water quality evaluation for conservation, and sustainability of water management.

Water quality predictions

XAI framework can forecast water quality using metrics and factors with interpretable results. Water quality assessment managers can comprehend the variables and parameters used for outcomes. This forces quality managers to mitigate water quality issues.

Flood hazard risk predictions

Floods can trigger landslides from excessive rainfall. Flooding causes countless casualties and property damage. Disaster warning systems need a flood risk assessment. XAI can forecast rapid water depths and provide timely, interpretable alerts to protect public health and safety.

Environmental impact assessment

XAI approach can be used for assessing the environmental impact on the water pollution incidents, and provide insight for mitigation and management. It enhances transparency and accountability by providing insights into the factors and parameters influencing environmental conditions. The analysis provided by the XAI model helps the stakeholders to identify the most significant factors contributing towards the environmental outcome.

System model and architecture

System model.

Worldwide, numerous water bodies are contaminated by a variety of anthropogenic and natural processes, resulting in a variety of health problems for human life. Thus water quality requires rigorous monitoring and management to prevent pollution. In accordance with WHO guidelines, the polluted water must be treated using the proper water treatment techniques before consumption. The quality of water is contaminated by the incessant addition of toxic chemicals and microbes and also by the relentless addition of local and industrial sewage sludge, trash, and extra hazardous waste that are toxic to humans and society. Many uncertainties are required to be quantified for all machine learning models. The uncertainties such as selecting and gathering the training data, absolute and accurate training data, understanding the machine learning models with performance bounds and drawbacks and finally the uncertainties which are based on the operational data. To minimize the challenges, adhoc steps like studying the model variability and sensitivity analysis are applied. In current years, the validation of water quality has taken active momentum because of ever-increasing water pollutants which spoil water that is dedicated for domestic use and irrigation. Water quality indices (WQIs) are used worldwide very efficiently for the assessment of the quality of both groundwater and other relevant water sources. Machine Learning techniques play a substantial role in identifying the quality of water using explainable AI. Figure 7 depicts the overall architecture of the proposed framework of our study. The dataset used in the study is split into the ratio of 70:30 wherein 70% is used for training and 30% is used for testing. The model is trained using a decision tree, random forest, SVM, logistic regression, and Naive Bayes algorithms. XAI model is implemented in the framework wherein LIME and Shapely are used to provide explainability and interpretability to the results generated by the machine learning model .

Interfacing ML algorithms with XAI.

Decision tree

The decision tree is stated as a recursive partition of the set of all possible instances 27 51 . The goal of a decision tree is to split the data which consequences in maximum information gain 52 . Let L be a sample for learning, L= ( \(v_{1}\) , \(c_{1}\) ), ( \(v_{2}\) , \(c_{2}\) ),( \(v_{i}\) , \(c_{j}\) ). Here, \(v_{1}\) , \(v_{2}\) , \(v_{3}\) , \(v_{i}\) are represented for measurement vectors, and \(c_{1}\) , \(c_{2}\) , \(c_{3}\) , \(c_{j}\) are represented for class labels.The batch conditions are reliant on one of the vector variables denoted as \(s_{i}\) 53 . Let us assume if the \(e_{i}\) of an element fits class label \(c_{i}\) , then \(p_{i}\) is denoted as per the Eq. ( 1 ).

Entropy evaluates the random value from the given samples and the homogeneity of the expected rate of a group of data 54 . To divide the data most optimally, the lowest value of entropy signifies better homogeneity.

L represents the data set evaluated by the entropy, ‘i’ denotes the classes in the set L, and \(e_{i}\) indicates the number of data labels that fit class ’i’ 55 . The least value of entropy is used for choosing the best feature. Information gain enumerates the amount of information provided by a particular characteristic about the target variable to minimize the uncertainty present in the data set. It is calculated by comparing the weighted average of entropy to the original data set after the splitting process. Let us assume that R is the rate for the features ‘f’, \([|{L}^R|]\) denotes the subset of LS so that bf=R 56 . After splitting L on the feature, information gain is given as follows.

The Gini index evaluates the heterogeneity of a selected node in the decision tree. It counts the probability of wrongly identifying data in the node. The Gini index begins from the value 0 to 1, where 0 indicates a pure node and 1 denotes a node that is distributed equally. The Gini index is represented as

Here, \(e_{i}\) represents the quantity of data labels. When the data is divided on class d as L1 and L2 with sizes \(s_{1}\) and \(s_{2}\) , Gini is evaluated as

Due to its comprehensible nature, decision trees can manage both numerical and categorical data with automatic feature selection.

Random forest

Random forest is an ensemble method that groups the results of multiple decision trees to compute predictions with enhanced accuracy. Every decision tree is improved on a random subset of labels from the dataset, to achieve diversity between the trees. When the data in the training label is t, then with replacement ‘n’ data are verified as bootstrap data 57 . This is done to produce the decision tree with training data. When there are ’m’ labels, a \(<<\) m is selected so that ‘a’ values are considered at random from ‘m’. The value ‘a’ is constant when the tree is growing to the highest level. The highest vote is noted as a new instance. (GE*) is the generalization error for the random forest and is denoted as

Here, f(X, Y) is a margin function to count the average number of votes from (X, Y). X denotes the prediction value and Y denotes the classification problem. The margin function is represented as

where ’F’ is for the indicator function. The value for the margin function is indicated as

The average value of a random forest and the mean correlation of the classifiers are combined as generalization errors. The p denotes the mean of the correlation. The generalization error for the upper bound is

Random forest reduces the over-fitting problem compared to a single decision tree. It can effectively manage high-dimensional data.

Support vector machine (SVM)

Let us consider a binary classification problem 1 or −1 to represent the sample variables 58 . When i elements of the sample variable is − 1, it is a positive class. When the i variables of the samples is 1, it is a negative set. Let V_i  = X1, X2,...Xn, Yi, i = 1,2,...n, \(Y\_{i}\in {-1,1}\) , Si indicates i item from the samples. Yi is the i item of the tests performed 59 . To split the samples into two parts, the function f(X) = ZTX+ b is used, where Z is the coefficient vector to normalize the hyperplane. The optimal margin is given as

\(\underbrace{MIN}_{\begin{array}{c} w, b, \\ \varepsilon \end{array}} \left( {\frac{1}{2}}Z^{TZ}+C\sum _{i=1}^{n}\varepsilon _i\right)\)

subject to:

The Lagrangian equation is given as

\(\underbrace{MAX}_{\propto } \left( \sum _{i=1}^{n}{\propto _i-\frac{1}{2}}\sum _{i,j=1}^{n}{\propto _i\propto _jY_iY_jX_iX_j}\right)\)

The Lagrangian equation with the maximum value with \(\propto _i\) a positive multiplier for the equation \(\sum _{i=1}^{n}{\propto _iY_i=0}\) and \(\propto _i\ge 0\) to change the optimal hyperplane 60 is presented. The optimal equation is given as

In the above equation \(\propto _i=0\) of the Lagrangian multiplier is nearest to the margin of the optimal hyperplane denoted as a support vector. This data is linearly separable by the kernel to evaluate the expected result from the instance 61 . The kernel function is denoted as

The generalized linear equation is changed to represent the non-linear dual Lagrangian \(La(\alpha )\) .

\(Lag\left( \propto \right) =\ \sum _{i=1}^{n}{\propto _i-\frac{1}{2}\sum _{i,j=1}^{n}{\propto _i\propto _jY_iY_jK\left( X_i,X_j\right) }}\)

Subject to:

The Lagrangian equation can be used for the separable case as

The SVM algorithm is very effective when the quantity of features is higher than the number of samples 62 .

Logistic regression

Logistic regression is used for binary classification problems to forecast the probability of an occurrence matching to a particular class. If the dependent value is binary, a regression analysis is used. The idea in logistic regression(logreg) is the logarithm ‘logn’ of odds of X, and odds are the ratios of probabilities ‘pb’ of X 63 . The rate of the independent value is termed odds because logistic regression measures the probability of an act that happens over the likelihood of an occurrence that does not happen.

where p is the probability of a positive output and x is the variable. The \(\alpha\) and \(\beta\) , are the logistic regression parameters 64 . The above equation is used for finding the number of occurrences as

\(p=probability(Y=positive\ outcome|X=x,\) a specific value)

For multiple predictors, a logic regression equation can be written as

\(p=probability(Y=positive\ outcome|X_1=x_1,\ldots ,x_k)\)

Here, pb refers to the probability of the positive occurrence of the event, the Y-intercept is \(\alpha\) , the regression coefficient is \(\beta\) , and e is 2.71828. Logistic regression is applied in various domains like finance, healthcare, social sciences, and many more for predicting diseases, credit default, etc.

Naive Bayesian classification

Gaussian Naive Bayes is a probabilistic classification algorithm developed based on Bayes theorem. It refers to the features which represent a normal distribution 65 . It classifies the samples as most likely classified as

If the sample \(Y_{j}\) is a vector, \(x_{j}\) is the \(j^{th}\) value which contains different values of \(y_{j}\) . The attributes used are dependent and it is shown as

Substituting the above equation into Bayes classification, we get

The Gaussian Naive Bayes algorithm is mainly applied for spam filtering, sentiment analysis, and text classification problems where the features must be continuous and follow the Gaussian distribution 66 .

LIME (Local interpretable model-agnostic explanations)

LIME explains the predictions of any kind of classifier by approximating locally along with an interpretable system. It changes the data sample by altering the values of features and monitors the impact of the result. It explains the predictions from every sample 67 . To receive the labels for the current data, alter the samples z ’s into the unique form \(z \in {\mathbb {R}}^d\) . Since the samples x ’ are generated randomly, x samples closer to the unique instance z for weighing are considered. The weight is evaluated as \(\Pi _z(x)\) for measuring the intimacy between the data z to x. The currently weighted data X and the samples formed by f ( x ), are trained as \(g \in G\) , where G is a model. The interpretable model \(\xi (x)\) of the current data g for explaining f ( x ) as

L is the loss function to measure whether g is following the state of f in the nearest neighborhood of z . If the loss function is reduced, the behaviour of g takes the behaviour of f as \(\Pi _z\) . The complexity of the model \(\Omega (g)\) should be low. When \(g(x')\) is considered as a linear function, \(g(x') = \varphi ^T x' + \varphi _0\) , changes the equation into a linear regression task to evaluate \(\varphi\) and \(\varphi _0\) .

SHAP (SHAPELY Additive exPlanations)

SHAP values determine the status of each feature for the prediction of a specific class 68 . The prediction f ( y ), using \(s(y')\) , a model for the binary elements \(x' \in \{0,1\}^M\) with the sets \(\emptyset _i \in {\mathbb {R}}\) , is given as

M refers to the explanation variable.

where f is the model of the SHAP, z refers to the variable, and \(z'\) are the variables chosen. The value \(f_y(x') - f_y(x'\setminus i)\) indicates all the predictions.

In this section two algorithms are discussed: one for the algorithm-based evaluation of water quality 1 and another for the algorithm-based explanation of water quality 2 . These two algorithms provide a holistic analysis and explanation of water quality management.

Algorithm for water quality classification

Algorithm for water quality Explanation

The water quality is assessed in the proposed work based on nine parameters such as pH value, Hardness (Total Dissolved Soils), Sulphate, Chloramines, Trihalomethanes, Conductivity, Organic carbon, and Turbidity. The target class for this dataset is Potability which is binary where 0 indicates that the water is not potable and 1 reflects its potability.

The dataset consisted of high missing values on sulphate and lower missing values on Chloramines and Trihalomethanes. The missing value imputation is hence performed and all the attributes are imputed for the missing values. The target class is converted into a numeric array for the processing of XAI models. This is done with the label encoder application of Python. The dataset is split with a ratio of 80:20 for training and testing.

The correlation analysis is performed on the dataset. The attribute Hardness has a high correlation of 0.34 with the target attribute potability. The next best correlation value is 0.24, which is rendered by the attribute Chloramines, followed by 0.21 produced by the Trihalomethanes attribute. Turbidity is the next better parameter with a correlation value of 0.16. The correlation heat map between the attributes of interest and the target attribute is presented in Fig. 8 .

Correlation analysis for water quality attributes.

The trained dataset is applied with SVM, LR, DT, RF and Gaussian Naive Bayes machine learning models. The SVM did not provide the desired classification and failed to converge for the portable data. The other models generated the results within the desired range and are presented in Table 6 .

The sensitivity and specificity measurements for the Machine learning models are presented in Table 7 . Considering the performance metrics, the results reveal the superiority of the RF model which generates a better outcome in comparison to the other models and thus it has been selected to be fed into the XAI model to provide enhanced interpretability, justifiability and transparency.

The XAI model implementation is performed considering SHAPELY values in the pandas’ application. This application focuses on the value of each feature in determining the target attribute which is potability. The significance of every feature is assessed through the various applications of SHAPELY. The first XAI model generated is the force plot, which provides the minimum and maximum prediction score of the target attribute in a dataset. The blue colored contour shows that a low score is measured and the red color shows a high score. The values at the separation boundary have the highest priority attribute. The force plot is presented in the Figs. 9 and 10 .

Force plot for water quality.

Force plot for potability.

The Global surrogate version of the force plot is presented in Fig. 11 . The blue regions indicate no potability and the red-coloured regions indicate potability. The border areas of the intersection show the attributes which have higher significance for the feature selection. The Sulphate value of 444 at the point of intersection indicates its significance in explaining this test patch for the entire dataset.

Test patch for potability.

The next XAI application of SHAPELY is the summary plot. This plot describes the features in determining binary classification problems. This predicts the scale of low to high for two significant results. The blue contour indicates lower significance towards the prediction and red indicates higher significance. The summary plot is shown in Fig. 12 . The Solids, pH, Sulfate, and Hardness show higher significance in determining the output.

Summary plot for potability.

The dependency plot shows the relationship between two features in the dataset. It provides the output in granular form with a variable-like result rather than simply a graph-like result of a Partial Dependency Plot(PDP). The relationship between the Sulphate and Potability is depicted in Fig. 13 . The mid-range of the dataset provides more granular output, which shows that the Sulphate parameter values are more significant in determining the values of potability in the mid-range of the dataset.

Dependency plot for potability.

The decision plot, which displays how the values of the features affect the goal, is the final model of XAI. This plot is a local surrogate plot, which would only explain a certain data instance, in which what values of the attributes influence the decision to be 1 or 0 as the decision of the model. The decision plot for the potability as 1 is illustrated in Fig. 14 . The potability 0 is illustrated in Fig. 15 .

Decision plot for potability.

The results of the experiment reveal the superiority of the RF model which generates an accuracy of 0.999 followed by DT, generating an accuracy of 0.998. The lowest accuracy is generated by the SVM model of 0.63. The RF is thus chosen for the implementation of the XAI model using SHAPELY. The comparative analysis of the aforementioned various models is depicted in Fig. 16 , considering evaluation metrics accuracy, precision, recall, and f1-score. In the case of all the performance metrics, the RF model outperforms the other models. Figure 17 shows the comparison of the sensitivity and specificity measures. The RF model stands superior in these considerations as well. Thus, the discussion offers a visual representation and justification of the reasoning behind the choice of RF to be included in the XAI framework to offer explainability.

Comparative analysis of machine learning models used.

Comparative analysis of sensitivity and specificity.

Apart from the selection of the RF model, SHAPELY provided five different models to explain the feature importance and relationships. The proposed work presented the force plot, summary plot, test patch, dependency plot, and decision plot. The Final decision plot explained how the classification is carried out using the corresponding values of the independent variables. Thus the black-box classification is explained in the white-box context of XAI. The following section describes the challenges and opportunities of the proposed work with an emphasis on future directions.

The proposed work may be influenced by the following challenges which are described in detail as follows,

Global unity

For the successful implementation of the system, a unanimously accepted implementation is essential. Unfortunately, water quality estimation and related research are limited to consideration of specific datasets acquired for a particular region, wherein the generated results may differ with the changes in geographic location. Thus the generated results can never be considered suitable on a global scale. The parameters that influence the water quality may also vary across the world, and hence the proposed work can never be considered as a universal solution.

Training and re-training

The qualifying attributes that determine the quality of water vary across the globe and hence the proposed model needs to be re-trained 69 when applied to a new environment of study. This would allow the model to unlearn and re-learn new environments. On the contrary, the complexity of the model would also increase. The accuracy and other performance metrics which are measured in the proposed work may drastically decrease as well in a different environment of study. Thus applying this model to versatile environments is complex and would be a challenging task.

Subjective or quantitative

The trade-off from subjective analysis (which was done through fuzzy-based methods in the form of the Analytical Hierarchy Process (AHP) and The Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS)) has improved the performance and ability to classify the models with better accuracy. However, the involvement of a subject matter expert is a missing point in the current research. Despite all the implementation and analysis from an engineering perspective, the involvement of an environmental scientist in any aspect of water research would contribute towards the enhancement of research quality.

Confusing solids

The proposed work identifies Solids as the primary influencing factor that affects potability. In real-world applications, solids can be of any form. For example, in sewage water treatment plants it can be either mud, Fat-Oil-Grease(FOG), or any other substances. Every solid wastage has its way of filtration and impact on water quality, which makes the recordings unstable from time to time. The attributes of research are too complex to handle in real-life scenarios, which acts as an inevitable yet detrimental impact.

Environmental challenges

Water resources are under serious threat due to water scarcity, water contamination, water conflicts and climate changes. Chemical and the municipal wastewater contaminates the water and endangering the life of the aquatic organisms and affect their ability to reproduce. This also makes them an easier prey to their predators. The food cycle and livelihood of the human is also greatly affected by the water contamination. Chemical substances make the water hard to recycle and consume by reducing the regeneration ratios.

Water quality and industrial sustainability

The era of Industry 5.0 focuses on the consumer centric industrial evolution with the idea of environmental sustainability. The futuristic technologies evolve with the improvement of technical viability, with the mission of sustainable development in the environmental aspects. Since the water is an irreplaceable and finite, the demand of the water is increasing with the industrial evolution and the water requirements on manufacturing and production industries would be very much essential as ever. The challenge is enhancement of the water harvesting, recycling and conservation. For all the above said processes quality of the water is the common essential requirement. Thus the quality of the water is more critical in all futuristic technological developments.

Research finding of the proposed work

The following items are presented as the findings are outcomes of the proposed work

The proposed work performs an exploratory analysis with XAI implementation providing an ability to improve the reliability of machine learning models providing explanation and transparency to the classification process.

The proposed work acquires data from a single dataset, where the performance of classification yields optimized results. This result may vary if the model is subjected to a different dataset constituting different features and instances.

The XAI reveals the most significant features contributing towards classification results and also explains the same.

The best fitting machine learning model is chosen for the explanation through an exhaustive analysis and evaluation of all the models considering the essential performance metrics. Thus the results produced by SHAPELY can be considered as the most reliable and acceptable. 

The proposed work also suggests the importance of the subject matter expert, which can extend the usability of the proposed model at the universal level.

The predictions of the proposed work with the support of an explainer, helps end users and consumers to understand the quality of the water they use.

The features related to the classification and explanation, can be further controlled to diminish the levels of chemicals and pollutants in water recycling.

Total dissolvable solids quantification and the feature weights for the same determine the levels of filtration and carbon purification required in the recycling plants.

The proposed work brings insights of pollutants on the seashore and how the explainabilty can support the impurity estimations for such conditions also.

Water quality management impacts almost all aspects of life on earth and clean water is a basic necessity. The proposed work is extremely relevant in this regard wherein an exploratory analysis conducted to analyze and control the factors that deteriorate the quality of the water. The impact of these factors is explained using XAI models. The contribution of the XAI model lies in its ability to explain the role of the underlying parameters towards the classification of water being potable or not, based on their relative importance and unique properties. The XAI model uses SHAPELY considering the probabilistic prediction generated from the Random Forest classifier. This RF model in this regard is chosen as it yields the highest accuracy of 0.999 with sensitivity and specificity of 0.999 and 0.998, which is found to be superior in comparison to the other state-of-the-art models considered in the study. This justifies the reason for the RF to be selected for XAI implementation. The proposed model identifies the parameter “solid” as the most significant in terms of its impact on the potability of water. The proposed model yields optimized and explainable results considering the dataset used in the study. Future work may involve more complex and heterogeneous datasets to generate predictions. In such scenarios, the metric evaluations may differ. The usage of deep learning algorithms could further enhance the examination the solid sediments and generate classification results based on their mass, dimensions, and shape. The use of XAI in such a model would ensure a better explanation of factors relevant to the solid sedimentation in water.

Data availability

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

Zhu, M. et al. A review of the application of machine learning in water quality evaluation. Eco-Environ. Health 1 , 107–116. https://doi.org/10.1016/j.eehl.2022.06.001 (2022).

Article   PubMed   PubMed Central   Google Scholar  

Miller, M., Kisiel, A., Cembrowska-Lech, D., Durlik, I. & Miller, T. Iot in water quality monitoring are we really here?. Sensors 23 , 960. https://doi.org/10.3390/s23020960 (2023).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Akhtar, N. et al. Modification of the water quality index (wqi) process for simple calculation using the multi-criteria decision-making (mcdm) method: A review. Water 13 , 905. https://doi.org/10.3390/w13070905 (2021).

Article   CAS   Google Scholar  

Abolfathi, S. & Pearson, J. Application of smoothed particle hydrodynamics (sph) in nearshore mixing: A comparison to laboratory data. Coastal Eng. Proc. 35 , 1–13 (2017).

Google Scholar  

Hájek, M. et al. A European map of groundwater ph and calcium. Earth Syst. Sci. Data 13 , 1089–1105. https://doi.org/10.5194/essd-13-1089-2021 (2021).

Article   ADS   Google Scholar  

Li, L. et al. Interpretable tree-based ensemble model for predicting beach water quality. Water Res. 211 , 118078. https://doi.org/10.1016/j.watres.2022.118078 (2022).

Article   CAS   PubMed   Google Scholar  

Lu, J. Can the central environmental protection inspection reduce transboundary pollution? Evidence from river water quality data in china. J. Clean. Prod. 332 , 130030 (2022).

Halder, J. N. & Islam, M. N. Water pollution and its impact on the human health. J. Environ. Hum. 2 , 36–46 (2015).

Article   Google Scholar  

Wang, Z. et al. Overview assessment of risk evaluation and treatment technologies for heavy metal pollution of water and soil. J. Clean. Prod. 379 , 134043 (2022).

Elehinafe, F. B., Agboola, O., Vershima, A. D. & Bamigboye, G. O. Insights on the advanced separation processes in water pollution analyses and wastewater treatment: A review. S. Afr. J. Chem. Eng. 48 , 188–200 (2022).

Mu, L., Mou, M., Tang, H. & Gao, S. Exploring preference and willingness for rural water pollution control: A choice experiment approach incorporating extended theory of planned behaviour. J. Environ. Manag. 332 , 117408 (2023).

Wang, Y., Ding, X., Chen, Y., Zeng, W. & Zhao, Y. Pollution source identification and abatement for water quality sections in Huangshui River Basin, China. J. Environ. Manag. 344 , 118326 (2023).

Najafzadeh, M. & Niazmardi, S. A novel multiple-kernel support vector regression algorithm for estimation of water quality parameters. Nat. Resour. Res. 30 , 3761–3775 (2021).

Najafzadeh, M., Homaei, F. & Farhadi, H. Reliability assessment of water quality index based on guidelines of national sanitation foundation in natural streams: Integration of remote sensing and data-driven models. Artif. Intell. Rev. 54 , 4619–4651 (2021).

Najafzadeh, M., Ghaemi, A. & Emamgholizadeh, S. Prediction of water quality parameters using evolutionary computing-based formulations. Int. J. Environ. Sci. Technol. 16 , 6377–6396 (2019).

Najafzadeh, M. & Basirian, S. Evaluation of river water quality index using remote sensing and artificial intelligence models. Remote Sens. 15 , 2359 (2023).

Chowdhury, M. A. Z. et al. Organophosphorus and carbamate pesticide residues detected in water samples collected from paddy and vegetable fields of the Savar and Dhamrai Upazilas in Bangladesh. Int. J. Environ. Res. Public Health 9 , 3318–3329 (2012).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Ahirvar, B. P., Das, P., Srivastava, V. & Kumar, M. Perspectives of heavy metal pollution indices for soil, sediment, and water pollution evaluation: An insight. Total Environ. Res. Themes 6 , 100039 (2023).

Chen, K., Liu, Q.-M., Peng, W.-H., Liu, Y. & Wang, Z.-T. Source apportionment of river water pollution in a typical agricultural city of Anhui province, Eastern China using multivariate statistical techniques with apcs-mlr. Water Sci. Eng. 16 , 165–174 (2023).

Fan, S. et al. Improved multi-criteria decision making method integrating machine learning for patent competitive potential evaluation: A case study in water pollution abatement technology. J. Clean. Prod. 403 , 136896 (2023).

Wang, Z., Wang, C. & Liu, Y. Evaluation for the nexus of industrial water-energy-pollution: Performance indexes, scale effect, and policy implications. Environ. Sci. Policy 144 , 88–98 (2023).

Asomaku, S. O. Quality assessment of groundwater sourced from nearby abandoned landfills from industrial city in Nigeria: Water pollution indices approach. HydroResearch 6 , 130–137 (2023).

Balaram, V., Copia, L., Kumar, U. S., Miller, J. & Chidambaram, S. Pollution of water resources and application of icp-ms techniques for monitoring and management: A comprehensive review. Geosyst. Geoenviron. 2 , 100210 (2023).

Yuan, F., Huang, Y., Chen, X. & Cheng, E. A biological sensor system using computer vision for water quality monitoring. Ieee Access 6 , 61535–61546 (2018).

Borzooei, S. et al. Impact evaluation of wet-weather events on influent flow and loadings of a water resource recovery facility. In New Trends in Urban Drainage Modelling: UDM 2018 11 706–711 (Springer, 2019).

Noori, R. et al. Decline in Iran’s groundwater recharge. Nat. Commun. 14 , 6674 (2023).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Yeganeh-Bakhtiary, A., EyvazOghli, H., Shabakhty, N., Kamranzad, B. & Abolfathi, S. Machine learning as a downscaling approach for prediction of wind characteristics under future climate change scenarios. Complexity 2022 , 8451812 (2022).

Jeihouni, M., Toomanian, A. & Mansourian, A. Decision tree-based data mining and rule induction for identifying high quality groundwater zones to water supply management: a novel hybrid use of data mining and gis. Water Resour. Manag. 34 , 139–154 (2020).

Lee, K.-J. et al. The combined use of self-organizing map technique and fuzzy c-means clustering to evaluate urban groundwater quality in Seoul Metropolitan City, South Korea. J. Hydrol. 569 , 685–697 (2019).

Article   ADS   CAS   Google Scholar  

Agrawal, P. et al. Exploring artificial intelligence techniques for groundwater quality assessment. Water 13 , 1172 (2021).

Wang, Y. et al. Monthly water quality forecasting and uncertainty assessment via bootstrapped wavelet neural networks under missing data for Harbin, China. Environ. Sci. Pollut. Res. 20 , 8909–8923 (2013).

El Bilali, A., Taleb, A. & Brouziyne, Y. Groundwater quality forecasting using machine learning algorithms for irrigation purposes. Agric. Water Manag. 245 , 106625 (2021).

Arabgol, R., Sartaj, M. & Asghari, K. Predicting nitrate concentration and its spatial distribution in groundwater resources using support vector machines (svms) model. Environ. Model. Assess. 21 , 71–82 (2016).

Sajedi-Hosseini, F. et al. A novel machine learning-based approach for the risk assessment of nitrate groundwater contamination. Sci. Total Environ. 644 , 954–962 (2018).

Article   ADS   CAS   PubMed   Google Scholar  

Ransom, K. M., Nolan, B. T., Stackelberg, P., Belitz, K. & Fram, M. S. Machine learning predictions of nitrate in groundwater used for drinking supply in the conterminous united states. Sci. Total Environ. 807 , 151065 (2022).

Yadav, B., Gupta, P. K., Patidar, N. & Himanshu, S. K. Ensemble modelling framework for groundwater level prediction in urban areas of India. Sci. Total Environ. 712 , 135539 (2020).

Tomić, A. Š, Antanasijević, D., Ristić, M., Perić-Grujić, A. & Pocajt, V. A linear and non-linear polynomial neural network modeling of dissolved oxygen content in surface water: Inter-and extrapolation performance with inputs’ significance analysis. Sci. Total Environ. 610 , 1038–1046 (2018).

Zhi, W. et al. From hydrometeorology to river water quality: Can a deep learning model predict dissolved oxygen at the continental scale?. Environ. Sci. Technol. 55 , 2357–2368 (2021).

Srinivas, R., Bhakar, P. & Singh, A. P. Groundwater quality assessment in some selected area of Rajasthan, India using fuzzy multi-criteria decision making tool. Aquat. Procedia 4 , 1023–1030 (2015).

Haghibi, A. H., Nasrolahi, A. H. & Parsaie, A. Water quality prediction using machine learning. J. Water Qual. Res. 53 , 3–13 (2018).

Liu, M. & Lu, J. Support vector machine-an alternative to artificial neuron network for water quality forecasting in an agricultural nonpoint source polluted river?. Environ. Sci. Pollut. Res. 21 , 11036–11053 (2014).

Chen, K. et al. Comparative analysis of surface water quality prediction performance and identification of key water parameters using different machine learning models based on big data. Water Res. 171 , 115454 (2020).

Sagan, V. et al. Monitoring inland water quality using remote sensing: Potential and limitations of spectral indices, bio-optical simulations, machine learning, and cloud computing. Earth-Sci. Rev. 205 , 103187 (2020).

Wu, Y., Zhang, X., Xiao, Y. & Feng, J. Attention neural network for water image classification under iot environment. Appl. Sci. 10 , 909 (2020).

Pu, F., Ding, C., Chao, Z., Yu, Y. & Xu, X. Water-quality classification of inland lakes using landsat8 images by convolutional neural networks. Remote Sens. 11 , 1674 (2019).

Donnelly, J., Daneshkhah, A. & Abolfathi, S. Forecasting global climate drivers using gaussian processes and convolutional autoencoders. Eng. Appl. Artif. Intell. 128 , 107536 (2024).

Abolfathi, S., Cook, S., Yeganeh-Bakhtiary, A., Borzooei, S. & Pearson, J. Microplastics transport and mixing mechanisms in the nearshore region. Coast. Eng. Proc. https://doi.org/10.9753/icce.v36v.papers.63 (2021).

Stride, B., Abolfathi, S., Odara, M. G. N., Bending, G. D. & Pearson, J. Modeling microplastic and solute transport in vegetated flows. Water Resour. Res. 59 , e2023WR034653. https://doi.org/10.1029/2023WR034653 (2023).

Unacademy (2022).

Başağaoğlu, H. et al. A review on interpretable and explainable artificial intelligence in hydroclimatic applications. Water 14 , 1230 (2022).

Habib, M., O’Sullivan, J., Abolfathi, S. & Salauddin, M. Enhanced wave overtopping simulation at vertical breakwaters using machine learning algorithms. PLoS ONE 18 , e0289318 (2023).

Mpia, H., Mburu, L. & Mwendia, S. Applying data mining in graduates’ employability: A systematic literature review. Int. J. Eng. Pedag. 13 , 86–108. https://doi.org/10.3991/ijep.v13i2.33643 (2023).

Raileanu, L. E. & Stoffel, K. Theoretical comparison between the gini index and information gain criteria. Ann. Math. Artif. Intell. 41 , 77–93. https://doi.org/10.1023/b:amai.0000018580.96245.c6 (2004).

Article   MathSciNet   Google Scholar  

Gulati, P., Sharma, A. & Gupta, M. Theoretical study of decision tree algorithms to identify pivotal factors for performance improvement: A review. Int. J. Comput. Appl. 141 , 19–25. https://doi.org/10.5120/ijca2016909926 (2016).

Tangirala, S. Evaluating the impact of GINI index and information gain on classification using decision tree classifier algorithm. Int. J. Adv. Comput. Sci. Appl. 11 , 110277. https://doi.org/10.14569/ijacsa.2020.0110277 (2020).

Xu, P. Review on studies of machine learning algorithms. J. Phys. 1187 , 052103. https://doi.org/10.1088/1742-6596/1187/5/052103 (2019).

Purwanto, A. D., Wikantika, K., Deliar, A. & Darmawan, S. Decision tree and random forest classification algorithms for mangrove forest mapping in Sembilang National Park, Indonesia. Remote Sens. 15 , 16. https://doi.org/10.3390/rs15010016 (2022).

Huang, H. et al. A new fruit fly optimization algorithm enhanced support vector machine for diagnosis of breast cancer based on high-level features. BMC Bioinform. https://doi.org/10.1186/s12859-019-2771-z (2019).

Ji, Y. & Sun, S. Multitask multiclass support vector machines: Model and experiments. Pattern Recogn. 46 , 914–924. https://doi.org/10.1016/j.patcog.2012.08.010 (2013).

Übeyli, E. D. ECG beats classification using multiclass support vector machines with error correcting output codes. Dig. Signal Process. 17 , 675–684. https://doi.org/10.1016/j.dsp.2006.11.009 (2007).

Cortes, C. & Vapnik, V. Support-vector networks. Mach. Learn. 20 , 273–297. https://doi.org/10.1007/bf00994018 (1995).

Ye, F., Lou, X. Y. & Sun, L. F. An improved chaotic fruit fly optimization based on a mutation strategy for simultaneous feature selection and parameter optimization for SVM and its applications. PLoS ONE 12 , e0173516. https://doi.org/10.1371/journal.pone.0173516 (2017).

Peng, C.-Y.J., Lee, K. L. & Ingersoll, G. M. An introduction to logistic regression analysis and reporting. J. Educ. Res. 96 , 3–14. https://doi.org/10.1080/00220670209598786 (2002).

Park, H.-A. An introduction to logistic regression: From basic concepts to interpretation with particular attention to nursing domain. J. Korean Acad. Nurs. 43 , 154. https://doi.org/10.4040/jkan.2013.43.2.154 (2013).

Article   PubMed   Google Scholar  

Chen, H., Hu, S., Hua, R. & Zhao, X. Improved Naive Bayes classification algorithm for traffic risk management. EURASIP J. Adv. Signal Process. https://doi.org/10.1186/s13634-021-00742-6 (2021).

Shen, J. & Fang, H. Human activity recognition using gaussian Naïve Bayes algorithm in smart home. J. Phys. 1631 , 012059. https://doi.org/10.1088/1742-6596/1631/1/012059 (2020).

Gramegna, A. & Giudici, P. SHAP and LIME: An evaluation of discriminative power in credit risk. Front. Artif. Intell. https://doi.org/10.3389/frai.2021.752558 (2021).

Zaremba, L., Zaremba, C. S. & Suchenek, M. Modification of shapley value and its implementation in decision making. Found. Manag. 9 , 257–272. https://doi.org/10.1515/fman-2017-0020 (2017).

Krishnan, S. R. et al. Smart water resource management using artificial intelligence;a review. Sustainability https://doi.org/10.3390/su142013384 (2022).

Download references

Acknowledgements

Acknowledgements should be brief, and should not include thanks to anonymous referees and editors, or effusive comments. Grant or contribution numbers may be acknowledged.

Author information

Authors and affiliations.

School of Computer Science Engineering and Information Systems, Vellore Institute of Technology, Vellore, 632014, India

M. K. Nallakaruppan, M. Lawanya Shri & Sweta Bhattacharya

Department of Computer Science, Loyola College, Chennai, Tamil Nadu, 600034, India

E. Gangadevi

Shiv Nadar University, Delhi-NCR, 201314, India

Balamurugan Balusamy

School of Built Environment, Engineering and Computing, Leeds Beckett University, Leeds, LS13HE, UK

Shitharth Selvarajan

Department of Computer Science, Kebri Dehar University, Kebri Dehar, Ethiopia

You can also search for this author in PubMed   Google Scholar

Contributions

All authors contributed equally in this research work.

Corresponding author

Correspondence to Shitharth Selvarajan .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Nallakaruppan, M.K., Gangadevi, E., Shri, M.L. et al. Reliable water quality prediction and parametric analysis using explainable AI models. Sci Rep 14 , 7520 (2024). https://doi.org/10.1038/s41598-024-56775-y

Download citation

Received : 21 October 2023

Accepted : 11 March 2024

Published : 29 March 2024

DOI : https://doi.org/10.1038/s41598-024-56775-y

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Water Quality
• Explainable AI
• Water Quality Prediction

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

A critical and intensive review on assessment of water quality parameters through geospatial techniques

• Review Article
• Published: 08 June 2021
• Volume 28 , pages 41612–41626, ( 2021 )

Cite this article

• Jaydip Dey 1 , 2 &
• Ritesh Vijay   ORCID: orcid.org/0000-0002-3198-6007 1  

1215 Accesses

9 Citations

7 Altmetric

Explore all metrics

Evaluation of water quality is a priority work nowadays. In order to monitor and map, the water quality for a wide range on different scales (spatial, temporal), the geospatial technique has the potential to minimize the field and laboratory work. The review has emphasized the advance of remote sensing for the effectiveness of spectral analysis, bio-optical estimation, empirical method, and application of machine learning for water quality assessment. The water quality parameters (turbidity, suspended particles, chlorophyll, etc.) and their retrieval techniques are described in a scientific manner. Available satellite, bands, resolution, and spectrum ranges for specific parameters are critically described in this review with challenges in remote sensing for water quality analysis, considering non-optical active parameters. The application of statistical programmes like linear (multiple regression analysis) and non-linear approaches is discussed for better assessment of water quality. Emphasis is given on comparison between different models to increase the accuracy level of remote sensing of water quality assessment. A direction is suggested for future development in the field of estimation of water pollution assessment through geospatial techniques.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

A Critical Review of Remote Sensing Methods for Inland Water Quality Monitoring: Progress, Limitations, and Future Perspectives

M. Raghul & P. Porchelvan

Inland Water Quality Monitoring Using Remote Sensing and GIS Techniques—A Tigris River, Iraq Case Study

Water indices for surface water extraction using geospatial techniques: a brief review

Kapil Kumar Purnam, A. D. Prasad & Padma Ganasala

Data availability

All data, models, and code generated or used during the study appear in the submitted article.

Abayazid HO, El-Adawy A (2019) Assessment of a non-optical water quality property using space-based imagery in Egyptian coastal lake. Journal of Water Resource and Protection 11:713–727. https://doi.org/10.4236/jwarp.2019.116042

Article   CAS   Google Scholar  

Ahn YH, Shanmugam P, Moon JE, Ryu JH (2008) Satellite remote sensing of a low-salinity water plume in the East China Sea. InAnnalesGeophysicae. Copernicus GmbH 26:2019–2035. https://doi.org/10.5194/angeo-26-2019-2008

Article   Google Scholar  

Aiken GR (1985) Humic substances in soil, sediment and water. Geochemistry, isolation and characterization 21:213–214. https://doi.org/10.1002/gj.3350210213

Akbar TA, Hassan Q, Achari G (2010) A remote sensing based framework for predicting water quality of different source waters. The International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences 34:1–4

Google Scholar  

Allan MG, Hicks BJ, Brabyn L (2007) Remote sensing of water quality in the Rotorua lakes. https://hdl.handle.net/10289/3785

Alparslan E, Aydöner C, Tufekci V, Tüfekci H (2007) Water quality assessment at Ömerli Dam using remote sensing techniques. Environ Monit Assess 135(1):391–398. https://doi.org/10.1007/s10661-007-9658-6

Álvarez-Robles JA, Zarazaga-Soria FJ, Latre MA, Béjar R, Muro-Medrano PR (2006) Water quality monitoring based on sediment distribution using satellite imagery. In Proceedings of the 9th AGILE Conference on Geographic Information Science, Visegrad, Hungary (pp. 20-22).approach. Remote Sensing of Environment 240, 111604. 10

Araujo GS, Abessa DM, Soares AM, Loureiro S (2019) Multi-generational exposure to Pb in two monophyletic Daphnia species: individual, functional and population related endpoints. Ecotoxicol Environ Saf 173:77–85. https://doi.org/10.1016/j.ecoenv.2019.02.001

Baban SM (1993) Detecting water quality parameters in the Norfolk Broads, UK, using Landsat imagery. Int J Remote Sens 14(7):1247–1267. https://doi.org/10.1080/01431169308953955

Becker RH, Sultan MI, Boyer GL, Twiss MR, Konopko E (2009) Mapping cyanobacterial blooms in the Great Lakes using MODIS. J Great Lakes Res 35:447–453

Bhatti AM, Rundquist DC, Nasu S, Takagi M (2008) Assessing the potential of remotely sensed data for water quality monitoring of coastal and inland waters

Bonansea M, Rodriguez MC, Pinotti L, Ferrero S (2015) Using multi-temporal Landsat imagery and linear mixed models for assessing water quality parameters in Río Tercero reservoir (Argentina). Remote Sens Environ 158:28–41

Boucher J, Weathers KC, Norouzi H, Steele B (2018) Assessing the effectiveness of Landsat 8 chlorophyll a retrieval algorithms for regional freshwater monitoring. Ecol Appl 28(4):1044–1054. https://doi.org/10.1002/eap.1708

Braga F, Giardino C, Bassani C, Matta E, Candiani G, Strömbeck N, Bresciani M (2013) Assessing water quality in the northern Adriatic Sea from HICO™ data. Remote sensing letters 4(10):1028–1037. https://doi.org/10.1080/2150704X.2013.830203

Brezonik P, Menken KD, Bauer M (2005) Landsat-based remote sensing of lake water quality characteristics, including chlorophyll and colored dissolved organic matter (CDOM). Lake and Reservoir Management 21(4):373–382

Brezonik PL, Olmanson LG, Bauer ME, Kloiber SM (2007) Measuring water clarity and quality in minnesota lakes and rivers: A census-based approach using remote-sensing techniques. Cura Rep 37:3–313

Buiteveld H, Hakvoort JH, Donze M (1994) Optical properties of pure water. In Ocean Optics XII (Vol. 2258, pp. 174-183). International Society for Optics and Photonics. https://doi.org/10.1117/12.190060

Carmichael WW, Boyer GL (2016) Health impacts from cyanobacteria harmful algae blooms: Implications for the North American Great Lakes. Harmful Algae 54:194–212. https://doi.org/10.1016/j.hal.2016.02.002

Chang NB, Imen S, Vannah B (2015) Remote sensing for monitoring surface water quality status and ecosystem state in relation to the nutrient cycle: a 40-year perspective. Crit Rev Environ Sci Technol 45:101–166. https://doi.org/10.1080/10643389.2013.829981

Chapman D(1996) Water quality assessments - A Guide To Use Of Biota, Sediments And Water In Environmental Monitoring. E&FN Spon, an imprint of Chapman & Hall

Chen L (2003) A study of applying genetic programming to reservoir trophic state evaluation using remote sensor data. Int J Remote Sens 24:2265–2275. https://doi.org/10.1080/01431160210154966

Chen Y, Fan C, Teubner K, Dokulil M (2003) Changes of nutrients and phytoplankton chlorophyll-a in a large shallow lake, Taihu, China: an 8-year investigation. Hydrobiologia 506(1-3):273–279. https://doi.org/10.1023/B:HYDR.0000008604.09751.01

Chen L, Tan CH, Kao SJ, Wang TS (2008) Improvement of remote monitoring on water quality in a subtropical reservoir by incorporating grammatical evolution with parallel genetic algorithms into satellite imagery. Water Res 42:296–306. https://doi.org/10.1016/j.watres.2007.07.014

Chipman JW, Olmanson LG, Gitelson AA (2009) Remote sensing methods for lake management: a guide for resource managers and decision-makers . North American Lake Management Society

Chislock MF, Doster E, Zitomer RA, Wilson AE (2013) Eutrophication: causes, consequences, and controls in aquatic ecosystems. Nature Education Knowledge 4 (4):10 https://www.researchgate.net/publication/285683019

Claverie M, Ju J, Masek JG, Dungan JL, Vermote EF, Roger JC, Skakun SV, Justice C (2018) The harmonized Landsat and Sentinel-2 surface reflectance data set. Remote Sens Environ 219:145–161. https://doi.org/10.1016/j.rse.2018.09.002

Craig SE, Jones CT, Li WK, Lazin G, Horne E, Caverhill C, Cullen JJ (2012) Deriving optical metrics of coastal phytoplankton biomass from ocean colour. Remote Sens Environ 119:72–83. https://doi.org/10.1016/j.rse.2011.12.007

Crittenden JC, Hand DW, Howe KJ, Rhodes Trussell R, Tchobanoglous G(2012) Water treatment principles and design, thirded. John Wiley & Sons

D'Alimonte D, Zibordi G, Berthon JF (2004) Determination of CDOM and NPPM absorption coefficient spectra from coastal water remote sensing reflectance. IEEE Trans Geosci Remote Sens 42 (8):1770–1777. https://doi.org/10.1109/TGRS.2004.831444

Davis TW, Berry DL, Boyer GL, Gobler CJ (2009) The effects of temperature and nutrients on the growth and dynamics of toxic and non-toxic strains of Microcystis during cyanobacteria blooms. Harmful Algae 8:715–725. https://doi.org/10.1016/j.hal.2009.02.004

Dekker AG, Peters SW (1993) The use of the thematic mapper for the analysis of eutrophic lakes: a case study in the Netherlands. Int J Remote Sens 14 (5):799–821. https://doi.org/10.1080/01431169308904379

Dekker AG, Zamurovic-Nenad Z, Hoogenboom HJ, Peters SWM (1996) Remote sensing, ecological water quality modelling and in-situ measurements: a case study in shallow lakes. Hydrol Sci J 41(4):531–547. https://doi.org/10.1080/02626669609491524

Dierberg FE, Carriker NE (1994) Field testing two instruments for remotely sensing water quality in the Tennessee Valley. Environ Sci Technol 28:16–25. https://doi.org/10.1021/es00050a004

Dörnhöfer K, Oppelt N (2016) Remote sensing for lake research and monitoring–Recent advances. Ecol Indic 64:105–122. https://doi.org/10.1016/j.ecolind.2015.12.009

D'Sa EJ (2008) Colored dissolved organic matter in coastal waters influenced by the Atchafalaya River, USA: effects of an algal bloom. J Appl Remote Sens 2 (1):023502. https://doi.org/10.1117/1.2838253

D'Sa EJ, Miller RL (2003) Bio-optical properties in waters influenced by the Mississippi River during low flow conditions. Remote Sens Environ 84 (4):538–549. https://doi.org/10.1016/S0034-4257(02)00163-3

Elhag M, Gitas I, Othman A, Bahrawi J, Gikas P (2019) Assessment of water quality parameters using temporal remote sensing spectral reflectance in arid environments, Saudi Arabia. Water 11 (3):556. https://doi.org/10.3390/w11030556

Fiorani L, Fantoni R, Lazzara L, Nardello I, Okladnikov I, Palucci A (2006) Lidar calibration of satellite sensed CDOM in the southern ocean. EARSeLeProc 5 (1):89–99

Flink P, Lindell LT, Östlund C (2001) Statistical analysis of hyperspectral data from two Swedish lakes. Sci Total Environ 68:155–169. https://doi.org/10.1016/S0048-9697(00)00686-0

Gholizadeh MH, Melesse AM, Reddi L (2016) A comprehensive review on water quality parameters estimation using remote sensing techniques. Sensors 16 (8):1298. https://doi.org/10.3390/s16081298

Giardino C, Brando VE, Dekker AG, Strömbeck N, Candiani G (2007) Assessment of water quality in Lake Garda (Italy) using Hyperion. Remote Sens Environ 109 (2):183–195. https://doi.org/10.1016/j.rse.2006.12.017

Giardino C, Bresciani M, Stroppiana D, Oggioni A, Morabito (2013) Optical remote sensing of lakes: an overview on Lake Maggiore. J Limnol 73. https://doi.org/10.4081/jlimnol.2014.817

Gitelson A, Garbuzov G, Szilagyi F, Mittenzwey KH, Karnieli A, Kaiser A (1993) Quantitative remote sensing methods for real-time monitoring of inland waters quality. Int J Remote Sens 14:1269–1295. https://doi.org/10.1080/01431169308953956

Gómez JA, Alonso CA, García AA (2011) Remote sensing as a tool for monitoring water quality parameters for Mediterranean Lakes of European Union water framework directive (WFD) and as a system of surveillance of cyanobacterial harmful algae blooms (SCyanoHABs). Environ Monit Assess 181(1):317–334. https://doi.org/10.1007/s10661-010-1831-7

Gray JR (2000) Comparability of suspended-sediment concentration and total suspended solids data (No. 4191). US Department of the interior, US Geological Survey

Hansen CH, Williams GP, Adjei Z, Barlow A, Nelson EJ, Miller AW (2015) Reservoir water quality monitoring using remote sensing with seasonal models: case study of five central-Utah reservoirs. Lake and Reservoir Management 31(3):225–240. https://doi.org/10.1080/10402381.2015.1065937

Harrington JA, Repic RL (1995) Hyperspectral and video remote sensing of oklahoma lakes. In Papers and Proceedings of Applied Geography Conferences (Vol. 18, pp. 79-86).APPLIED GEOGRAPHY CONFERENCES.

Helms JR, Stubbins A, Ritchie JD, Minor EC, Kieber DJ, Mopper K (2008) Absorption spectral slopes and slope ratios as indicators of molecular weight, source, and photobleaching of chromophoric dissolved organic matter. Limnol.Oceanogr. 53:955–969. https://doi.org/10.4319/lo.2008.53.3.0955

Hicks BJ, Stichbury GA, Brabyn LK, Allan MG, Ashraf S (2013) Hindcasting water clarity from Landsat satellite images of unmonitored shallow lakes in the Waikato region, New Zealand. Environ Monit Assess 185:7245–7261. https://doi.org/10.1007/s10661-013-3098-2

Huang JJ, Guo H, Chen B, Guo X, Singh VP (2020) Retrieval of non-optically active parameters for small scale urban waterbodies by a machine learning-based strategy. doi: https://doi.org/10.20944/preprints202004.0111.v1

Hunter PD, Tyler AN, Carvalho L, Codd GA, Maberly SC (2010) Hyperspectral remote sensing of cyanobacterial pigments as indicators for cell populations and toxins in eutrophic lakes. Remote Sens Environ 114(11):2705–2718. https://doi.org/10.1016/j.rse.2010.06.006

IOCCG (2006a) Remote sensing of inherent optical properties: fundamentals, tests of algorithms, and applications. IOCCG Report number 5. Dartmouth, NS: IOCCG

IOCCG (2006b) In: Lee, Z. (Ed.), Remote sensing of inherent optical properties: fundamentals, tests of algorithms, and applications. Reports of the International Ocean- Colour Coordinating Group, No. 5, Dartmouth, Canada IOCCG

IOCCG (2018) Earth observations in support of global water quality monitoring. In: Greb S, Dekker A, Binding C (eds) IOCCG Report Series. Canada, International Ocean Colour Coordinating Group, Dartmouth

Jordan YC, Ghulam A, Chu ML (2014) Assessing the impacts of future urban development patterns and climate changes on total suspended sediment loading in surface waters using geoinformatics. J Environ Inf 24:65–79. https://doi.org/10.3808/JEI.201400283

Kallio K, Kutser T, Hannonen T, Koponen S, Pulliainen J, Vepsäläinen J, Pyhlaähti T (2001) Retrieval of water quality from airborne imaging spectrometry of various lake types in different seasons. Sci Total Environ 268:59–77. https://doi.org/10.1016/S0048-9697(00)00685-9

Keiner LE, Brown CW (1999) Estimating oceanic chlorophyll concentrations with neural networks. Int J Remote Sens 20:189–194.10.1080/014311699213695

Khorram S, Cheshire H, Geraci AL, ROSA GL (1991) Water quality mapping of Augusta Bay, Italy from Landsat-TM data. Int J Remote Sens 12(4):803–808. https://doi.org/10.1080/01431169108929696

Kloiber SM, Brezonik PL, Bauer ME (2002) Application of Landsat imagery to regional-scale assessments of lake clarity. Water Res 36(17):4330–4340. https://doi.org/10.1016/S0043-1354(02)00146-X

Kneizys FX, Shettle EP, Abreu LW, Chetwynd JH, Anderson GP(1988) Users guide to LOWTRAN 7 (No. AFGL-TR-88-0177). AFB MA, AIR FORCE GEOPHYSICS LAB HANSCOM

Kritzberg ES, Langenheder S, Lindström ES (2006) Influence of dissolved organic matter source on lake bacterioplankton structure and function–implications for seasonal dynamics of community composition. FEMS Microbiol Ecol 56 (3):406–417. https://doi.org/10.1111/j.1574-6941.2006.00084.x

Kudela RM, Palacios SL, Austerberry DC, Accorsi EK, Guild LS, Torres-Perez J (2015) Application of hyperspectral remote sensing to cyanobacterial blooms in inland waters. Remote Sens Environ 167:196–205. https://doi.org/10.1016/j.rse.2015.01.025

Kutser T (2009) Passive optical remote sensing of cyanobacteria and other intense phytoplankton blooms in coastal and inland waters. Int J Remote Sens 30:4401–4425. https://doi.org/10.1080/01431160802562305

Kutser T, Pierson DC, KallioKY RA, Sobek S (2005) Mapping lake CDOM by satellite remote sensing. Remote Sens Environ 94:535–540. https://doi.org/10.1016/j.rse.2004.11.009

Kutser T, Paavel B, Verpoorter C, Kauer T, Vahtmäe E (2012) Remote sensing of water quality in optically complex lakes. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences 39:B8

Lacaux JP, Tourre YM, Vignolles C, Ndione JA, Lafaye M (2007) Classification of ponds from high-spatial resolution remote sensing: application to Rift Valley Fever epidemics in Senegal. Remote Sens Environ 106:66–74. https://doi.org/10.1016/j.rse.2006.07.012

Le C, Hu C, Cannizzaro J, English D, Muller-Karger F, Lee Z (2013) Evaluation of chlorophyll-a remote sensing algorithms for an optically complex estuary. Remote Sens Environ 129:75–89. https://doi.org/10.1016/j.rse.2012.11.001

Lim J, Choi M (2015) Assessment of water quality based on Landsat 8 operational land imager associated with human activities in Korea. Environ Monit Assess 187:384. https://doi.org/10.1007/s10661-015-4616-1

Lindell LT, Steinvall O, Jonsson M, Claesson T (1985) Mapping of coastal-water turbidity using Landsat imagery. Int J Remote Sens 6(5):629–642. https://doi.org/10.1080/01431168508948486

Ma L, Liu Y, Zhang X, Ye Y, Yin G, Johnson BA (2019) Deep learning in remote sensing applications: a meta-analysis and review. ISPRS J Photogramm Remote Sens 152:166–177. https://doi.org/10.1016/j.isprsjprs.2019.04.015

Mahato LL, Pathak AK, Kapoor D, Patel N, Murthy M (2014) Surface water monitoring and evaluation of indravati reservoir using the application of principal component analysis using satellite remote sensing technology. In: Proceedings of Map Asia 2004. Beijing, China, pp 26–29

Maier PM, Keller S (2018) Machine learning regression on hyperspectral data to estimate multiple water parameters. In2018 9th Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS) (pp. 1-5). IEEE.DOI: https://doi.org/10.1109/WHISPERS.2018.8747010

Maritorena S, Guillocheau N (1996) Optical properties of water and spectral light absorption by living and non-living particles and by yellow substances in coral reef waters of French Polynesia. Mar Ecol Prog Ser 131:245–255. https://doi.org/10.3354/meps13124

Matthews MW, Odermatt D(2015) Improved algorithm for routine monitor-ing of cyanobacteria and eutrophication in inland and near-coastal waters. Remote Sens. V. Sagan, et al. Earth-Science Reviews 205 (2020) 103187 Environ. 156. https://doi.org/10.1016/j.rse.2014.10.010

Matthews MW, Bernard S, Robertson L (2012) An algorithm for detecting trophic status (chlorophyll-a), cyanobacterial-dominance, surface scums and floating vegetation in inland and coastal waters. Remote Sens Environ 124:637–652. https://doi.org/10.1016/j.rse.2012.05.032

McQueen DJ, Post JR, Mills EL (1986) Trophic relationships in freshwater pelagic ecosystems. Can J Fish Aquat Sci 43(8):1571–1581. https://doi.org/10.1139/f86-195

Menken KD, Brezonik PL, Bauer ME (2006) Influence of chlorophyll and colored dissolved organic matter (CDOM) on lake reflectance spectra: Implications for measuring lake properties by remote sensing. Lake and Reservoir Management 22(3):179–190. https://doi.org/10.1080/07438140609353895

Miao S, Liu C, Qian B, Miao Q (2020) Remote sensing-based water quality assessment for urban rivers: a study in linyi development area. Environ Sci Pollut Res 27(28):34586–34595. https://doi.org/10.1007/s11356-018-4038-z

Mishra S, Mishra DR (2012) Normalized difference chlorophyll index: a novel model for remote estimation of chlorophyll-a concentration in turbid productive waters. Remote Sens Environ 117:394–406. https://doi.org/10.1016/j.rse.2011.10.016

Mobley CD(1998) Hydrolight 4.0 users guide

Mohd Hasmadi I, Norsaliza U (2010) Analysis of SPOT-5 data for mapping turbidity level of river klang. Peninsular Malaysia Appl Remote Sens J 1:14–18

Moore C, Barnard A, Fietzek P, Lewis MR, Sosik HM, White S, Zielinski O (2009) Optical tools for ocean monitoring and research. Ocean Sci 5:661–684. https://doi.org/10.5194/os-5-661-2009

Muller-Karger FE (1992) Remote sensing of marine pollution: a challenge for the 1990s. Mar Pollut Bull 25(1-4):54–60. https://doi.org/10.1016/0025-326X(92)90186-A

Müller-Navarra DC, Brett MT, Park S, Chandra S, Ballantyne AP, Zorita E, Goldman CR (2004) Unsaturated fatty acid content in seston and tropho-dynamic coupling in lakes. Nature 427(6969):69–72. https://doi.org/10.1038/nature02210

Murphy KP (2012) Machine learning : a probabilistic perspective. MIT Press, Cambridge, Mass

Murray C, Markager S, Stedmon CA, Juul-Pedersen T, Sejr MK, Bruhn A (2015) The influence of glacial melt water on bio-optical properties in two contrasting Greenlandic fjords. EstuarCoast Shelf Sci 163:72–83. https://doi.org/10.1016/j.ecss.2015.05.041

Myint SW, Walker ND (2002) Quantification of surface suspended sediments along a river dominated coast with NOAA AVHRR and SeaWiFS measurements: Louisiana, USA. Int J Remote Sens 23(16):3229–3249. https://doi.org/10.1080/01431160110104700

Neville RA, Gower JF (1977) Passive remote sensing of phytoplankton via chlorophyll α fluorescence. J Geophys Res 82:3487–3493. https://doi.org/10.1029/JC082i024p03487

Odermatt D, Kiselev V, Heege T, Kneubühler M, Itten KI(2008) Adjacency effect considerations and air/water constituent retrieval for Lake Constance. In: aProceedings of the 2nd MERIS/(A) ATSR user workshop. Frascati, Italy. Vol. 1.of total phosphorus (TP) in three central indiana water supply reservoirs. Water Air

Ogashawara I, Mishra DR, Mishra S, Curtarelli MP, Stech JL (2013) A performance review of reflectance based algorithms for predicting phycocyanin concentrations in inland waters. Remote Sens 5:4774–4798. https://doi.org/10.3390/rs5104774

Olmanson LG, Bauer ME, Brezonik PL (2008) A 20-year Landsat water clarity census of Minnesota’s 10,000 lakes. Remote Sens Environ 112:4086–4097

Olmanson LG, Brezonik PL, Bauer ME (2011) Evaluation of medium to low resolution satellite imagery for regional lake water quality assessments. Water Resour Res 47. https://doi.org/10.1016/j.rse.2007.12.013

Olmanson LG, Brezonik PL, Bauer ME (2013) Airborne hyperspectral remote sensing to assess spatial distribution of water quality characteristics in large rivers: the Mississippi river and its tributaries in Minnesota. Remote Sens Environ 130:254–265. https://doi.org/10.1016/j.rse.2012.11.023

Osinska-Skotak K, Kruk M, Mróz M (2007) The spatial diversification of lake water quality parameters in Mazurian lakes in summertime. Millpress, Rotterdam, The Netherlands

Paerl HW, Otten TG (2013) Harmful cyanobacterial blooms: causes, consequences, and controls. Microb.Ecol 65:995–1010. https://doi.org/10.1007/s00248-012-0159-y

Pahlevan N, Smith B, Schalles J, Binding C, Cao Z, Ma R, Alikas K, Kangro K, Gurlin D, Hà N, Matsushita B (2020) Seamless retrievals of chlorophyll-a from Sentinel-2 (MSI) and Sentinel-3 (OLCI) in inland and coastal waters: a machine-learning approach. Remote Sens Environ 240:111604. https://doi.org/10.1016/j.rse.2019.111604

Panda SS, Garg V, Chaubey I (2004) Artificial neural networks application in lake water quality estimation using satellite imagery. J Environ Inf 4:65–74

Peña-Martínez R, Ruiz-Verdú A, Domínguez-Gómez JA (2004). Mapping of photosynthetic pigments in Spanish inland waters using MERIS imagery. In Proceedings of the 2004 Envisat& ERS Symposium, Salzburg, Austria (pp. 6-10)

Pereira LS, Andes LC, Cox AL, Ghulam A (2018) Measuring suspended-sediment concentration and turbidity in the Middle Mississippi and Lower Missouri rivers using Landsat data. J Am Water Resour Assoc 54:440–450. https://doi.org/10.1111/1752-1688.12616

Pérez GL, Galí M, Royer SJ, Sarmento H, Gasol JM, Marrasé C, SimóR (2016) Bio-optical characterization of offshore NW Mediterranean waters: CDOM contribution to the absorption budget and diffuse attenuation of downwelling irradiance. Deep-Sea Res I Oceanogr Res Pap 114:111–127. https://doi.org/10.1016/j.dsr.2016.05.011

Peterson KT, Sagan V, Sidike P, Cox A, Martinez M (2018) Suspended sediment concentration estimation from landsat imagery along the lower missouri and middle mississippi rivers using an extreme learning machine. Remote Sens 10:1503. https://doi.org/10.3390/rs10101503

Peterson KT, Sagan V, Sidike P, Hasenmueller EA, Sloan JJ, Knouft JH (2019) Machine learning based ensemble prediction of water quality variables using featurelevel 1 and decision-level fusion with proximal remote sensing. PhotogrammEng Remote Sens 85(4):269–280. https://doi.org/10.14358/PERS.85.4.269

Peterson KT, Sagan V, Sloan J (2020) Deep learning-based water quality estimation and anomaly detection using Landsat-8/Sentinel-2 virtual constellation and cloud computing. GIScience& Remote Sensing 57:510–525. https://doi.org/10.1080/15481603.2020.1738061

Preisendorfer RW (1961) Application of radiative transfer theory to light measurements in the sea. Union GeodGeophys Inst Monogr 10:11–30

Qi L, Hu CM, Duan HT, Barnes BB, Ma RH (2014) An EOF-based algorithm to estimate chlorophyll a concentrations in Taihu Lake from MODIS land-band measurements: implications for near real-time applications and forecasting models. Remote Sens 6:10694–10715. https://doi.org/10.3390/rs61110694

Rajalahti T, Arneberg R, Berven FS, Myhr KM, Ulvik RJ, Kvalheim OM (2009) Biomarker discovery in mass spectral profiles by means of selectivity ratio plot. ChemomIntell Lab Syst 95:35–48. https://doi.org/10.1016/j.chemolab.2008.08.004

Rastogi RP, Madamwar D, Incharoensakdi A (2015) Bloom dynamics of cyanobacteria and their toxins: environmental health impacts and mitigation strategies. Front Microbiol 6:1254. https://doi.org/10.3389/fmicb.2015.01254

Ritchie JC, Schiebe FR, McHenry R (1976) Remote sensing of suspended sediment in surface waters. PhotogrammEng Remote Sens 69:695–714

Ritchie JC, Zimba PV, Everitt JH (2003) Remote sensing techniques to assess water quality. Photogramm Eng Remote Sens 69(6):695–704. https://doi.org/10.14358/PERS.69.6.695

Rundquist DC, Han L, Schalles JF, Peake JS (1996) Remote measurement of algal chlorophyll in surface waters: the case for the first derivative of reflectance near 690 nm. Photogramm Eng Remote Sens 62(2):195–200

Sagan V, Peterson KT, Maimaitijiang M, Sidike P, Sloan J, Greeling BA, Maalouf S (2020) Adams C (2020) Monitoring inland water quality using remote sensing: potential and limitations of spectral indices, bio-optical simulations, machine learning, and cloud computing. Earth Sci Rev 22:103187. https://doi.org/10.1016/j.earscirev.2020.103187

Santos IR, Costa RC, Freitas U, Fillmann G (2008) Influence of effluents from a wastewater treatment plant on nutrient distribution in a coastal creek from southern Brazil. Braz Arch Biol Technol 51(1):153–162. https://doi.org/10.1590/S1516-89132008000100019

Satapathy DR, Vijay R, Kamble SR, Sohony RA (2010) Remote sensing of turbidity and phosphate in creeks and coast of Mumbai: an effect of organic matter. Trans GIS 146:811–832. https://doi.org/10.1111/j.1467-9671.2010.01234.x

Savitzky A, Golay MJ (1964) Smoothing and differentiation of data by simplified least squares procedures. Anal Chem 36:1627–1639. https://doi.org/10.1021/ac60214a047

Sawaya KE, Olmanson LG, Heinert NJ, Brezonik PL, Bauer ME (2003) Extending satellite remote sensing to local scales: land and water resource monitoring using high-resolution imagery. Remote Sens Environ 88(1-2):144–156. https://doi.org/10.1016/j.rse.2003.04.006

Schroeder T, Brando VE, Cherukuru N, Clementson L, Blondeau-Patissier D, Dekker AG, Schaale M, Fischer J (2008) Remote sensing of apparent and inherent optical properties of Tasmanian coastal waters: application to MODIS data. InProceedings of the XIX Ocean Optics Conference, Barga, Italy. (pp. 6-10).: https://www.researchgate.net/publication/260106801

Shafique NA, Fulk F, Autrey BC, Flotemersch J (2003) Hyperspectral remote sensing of water quality parameters for large rivers in the Ohio River basin. In First interagency conference on research in the watershed, Benson, AZ (pp. 216-221).

Shirke S, Pinto SM, Kushwaha VK, Mardikar T, Vijay R (2016) Object-based image analysis for the impact of sewage pollution in Malad Creek, Mumbai, India. Environ Monit Assess 188(2):95. https://doi.org/10.1007/s10661-015-4981-9

Sidike P, Sagan V, Maimaitijiang M, Maimaitiyiming M, Shakoor N, Burken J, Mockler T, Fritschi FB (2019) dPEN: Deep progressively expanded network for mapping heterogeneous agricultural landscape using WorldView-3 satellite imagery. Remote Sens Environ 221:756–772. https://doi.org/10.1016/j.rse.2018.11.031

Somvanshi S, Kunwar P, Singh NB, Shukla SP, Pathak V (2012) Integrated remote sensing and GIS approach for water quality analysis of Gomti river, Uttar Pradesh. Int J Environ Sci 3(1):62–74

CAS   Google Scholar  

Song Y, Song XD, Jiang H, Guo ZB, Guo QH (2010) Quantitative remote sensing retrieval for algae in inland waters. Spectrosc Spectr Anal 30(4):1075–1079. https://doi.org/10.3964/j.issn.1000-0593(2010)04-1075-05

Song K, Lu D, Li L, Li S, Wang Z, Du J (2012b) Remote sensing of chlorophyll-a concentration for drinking water source using genetic algorithms (GA): partial least square (PLS) modeling. Ecological Informatics 10:25–36. https://doi.org/10.1016/j.ecoinf.2011.08.006

Song K, Li L, Li S, Tedesco L, Hall B, Li L (2012c) Hyperspectral remote sensing of total phosphorus (TP) in three central Indiana water supply reservoirs. Water Air Soil Pollut 223(4):1481–1502. https://doi.org/10.1007/s11270-011-0959-6

Sudheer KP, Chaubey I, Garg V (2006) Lake water quality assessment from landsat thematic mapper data using neural network: an approach to optimal band combination selection1. JAWRA Journal of the American Water Resources Association 42(6):1683–1695. https://doi.org/10.1111/j.1752-1688.2006.tb06029.x

Swanson H, Zurawell R (2006) Steele Lake water quality monitoring report. Monitoring and Evaluation Branch, Environmental Assurance Division, Alberta Environment, Edmonton, AB, Canada

Tehrani NC, D'Sa EJ, Osburn CL, Bianchi TS, Schaeffer BA (2013) Chromophoric dissolved organic matter and dissolved organic carbon from sea-viewing wide field-of-view sensor (SeaWiFS), Moderate Resolution Imaging Spectroradiometer (MODIS) and MERIS Sensors: case study for the Northern Gulf of Mexico. Remote Sens 5(3):1439–1464. https://doi.org/10.3390/rs5031439

Thurman EM(1985) Organic Geochemistry of Natural Waters.

Tiwari SP, Shanmugam P (2011) An optical model for the remote sensing of coloured dissolved organic matter in coastal/ocean waters. Estuar Coast Shelf Sci 93(4):396–402. https://doi.org/10.1016/j.ecss.2011.05.010

Turner D (2010) Remote sensing of chlorophyll a concentrations to support the Deschutes basin lake and reservoirs TMDLs. Department of Environmental Quality, Portland, OR, USA

Twardowski MS, Boss E, Sullivan JM, Donaghay PL (2004) Modeling the spectral shape of absorption by chromophoric dissolved organic matter. Mar Chem 89(1-4):69–88. https://doi.org/10.1016/j.marchem.2004.02.008

Tyler AN, Svab E, Preston T, Présing M, Kovács WA (2006) Remote sensing of the water quality of shallow lakes: a mixture modelling approach to quantifying phytoplankton in water characterized by high-suspended sediment. Int J Remote Sens 27:1521–1537. https://doi.org/10.1080/01431160500419311

Van der Meer FD, De Jong SM, editors (2011) Imaging spectrometry: basic principles and prospective applications. Springer Science & Business Media.

Vijay R, Dey J, Sakhre S, Kumar R (2020) Impact of urbanization on creeks of Mumbai, India: a geospatial assessment approach. J Coast Conserv 24(1):1–16, 04. https://doi.org/10.1007/s11852-019-0072-y

Vijay R, Kushwaha VK, Pandey N, Nandy T, Wate SR (2015) Extent of sewage pollution in coastal environment of Mumbai, India: an object-based image analysis. Water and Environment Journal 29(3):365–374. https://doi.org/10.1111/wej.12115

Vijay R, Pinto SM, KushwahaVK, Pal S, Nandy T (2016) A multi-temporal analysis for change assessment and estimation of algal bloom in Sambhar Lake, Rajasthan, India. Environ Monit Assess188(9). https://doi.org/10.1007/s10661-016-5509-7

Vincent RK, Qin XM, McKay RM, Miner J, Czajkowski K, Savino J, Bridgeman T (2004) Phycocyanin detection from LANDSAT TM data for mapping cyanobacterial blooms in Lake Erie. Remote Sens Environ 89:381–392. https://doi.org/10.1016/j.rse.2003.10.014

Visser PM, Ibelings BW, Mur LR, Walsby AE (2005) The ecophysiology of the harmful cyanobacterium Microcystis: features explaining its success and measures for its control. In: Huisman J, Matthijs HCP, Visser PM (eds) Harmful cyanobacteria. Springer-Verlag, Berlin, pp 109–142. https://doi.org/10.1007/1-4020-3022-3_6

Chapter   Google Scholar  

Vollenweider RA (1976) Advances in defining critical loading levels for phosphorus in lake eutrophication. Memoriedell'IstitutoItaliano di Idrobiologia, Dott. Marco de MarchiVerbaniaPallanza.

Wang F, Han L, Kung HT, Van Arsdale RB (2006) Applications of Landsat-5 TM imagery in assessing and mapping water quality in Reelfoot Lake, Tennessee. Int J Remote Sens 27(23):5269–5283. https://doi.org/10.1080/01431160500191704

Wang X, Ma L, Wang X (2010) Apply semi-supervised support vector regression for remote sensing water quality retrieving. In2010 IEEE International Geoscience and Remote Sensing Symposium (pp. 2757-2760). IEEE.DOI: https://doi.org/10.1109/IGARSS.2010.5653832

Wang MH, Nim CJ, Son S, Shi W (2012) Characterization of turbidity in Florida’s Lake Okeechobee and Caloosahatchee and St. Lucie Estuaries using MODIS-Aqua measurements. Water Res 46:5410–5422. https://doi.org/10.1016/j.watres.2012.07.024

Wang MH, Son SH, Zhang YL, Shi W (2013) Remote sensing of water optical property for China's Inland Lake Taihu using the SWIR atmospheric correction with 1640 and 2130 nm bands. Ieee Journal of Selected Topics in Applied Earth Observations and Remote Sensing 6:2505–2516. https://doi.org/10.1109/JSTARS.2013.2243820

Wang X, Zhang F, Kung H, Ghulam A, Trumbo A, Yang J, Ren Y, JingY (2017) Evaluation and estimation of surface water quality in an arid region based on EEMPARAFAC and 3D fluorescence spectral index: a case study of the Ebinur Lake Watershed, China. Catena 155:62–74. https://doi.org/10.1016/j.catena.2017.03.006

Wass PD, Marks SD, Finch JW, Leeks GJ, Ingram JK (1997) Monitoring and preliminary interpretation of in-river turbidity and remote sensed imagery for suspended sediment transport studies in the Humber catchment. Sci Total Environ 194:263–283. https://doi.org/10.1016/S0048-9697(96)05370-3

Wu M, Zhang W, Wang X, Luo D (2009) Application of MODIS satellite data in monitoring water quality parameters of Chaohu Lake in China. Environ Monit Assess 148(1-4):255–264. https://doi.org/10.1007/s10661-008-0156-2

Wu C, Wu J, Qi J, Zhang L, Huang H, Lou L, Chen Y (2010) Empirical estimation of total phosphorus concentration in the mainstream of the Qiantang River in China using Landsat TM data. Int J Remote Sens 31(9):2309–2324. https://doi.org/10.1080/01431160902973873

Wynne TT, Stumpf RP, Tomlinson MC, Warner RA, Tester PA, Dyble J, Fahnenstiel GL (2008) Relating spectral shape to cyanobacterial blooms in the Laurentian Great Lakes. Int J Remote Sens 29:3665–3672

Zeinelabdeina KE, Albielyb AI (2008) Ratio image processing techniques: a prospecting tool for mineral deposits, Red Sea Hills. NE Sudan Int Arch Photogramm Remote Sens Spat Inf Sci 37:1981–1984

Zhang YZ, Pulliainen J, Koponen S, Hallikainen M (2002) Water quality studies of combined optical, thermal infrared, and microwave remote sensing. Microw Opt Technol Lett 34:281–285. https://doi.org/10.1002/mop.10438

Zhang YC, Ma RH, Duan HT, Loiselle SA, Xu JD, Ma MX (2014) A novel algorithm to estimate algal bloom coverage to subpixel resolution in Lake Taihu. Ieee Journal of Selected Topics in Applied Earth Observations and Remote Sensing 7:3060–3068. https://doi.org/10.1109/JSTARS.2014.2327076

Zhu W, Yu Q, Tian YQ, Chen RF, Gardner GB (2011) Estimation of chromophoric dissolved organic matter in the Mississippi and Atchafalaya river plume regions using above-surface hyperspectral remote sensing. J Geophys Res Oceans 116(C2). https://doi.org/10.1029/2010JC006523

Zhu W, Huang L, Sun N, Chen J, Pang S (2020) Landsat 8-observed water quality and its coupled environmental factors for urban scenery lakes: a case study of West Lake. Water Environ Res 92(2):255–265. https://doi.org/10.1002/wer.1240

Download references

Acknowledgements

Authors are thankful to the Director of CSIR-National Environmental Engineering Research Institute (NEERI), Nagpur, for providing the necessary infrastructure and support to carry out this study.

Author information

Authors and affiliations.

Wastewater Technology Division, CSIR-National Environmental Engineering Research Institute (NEERI), Nagpur, Maharashta, 440020, India

Jaydip Dey & Ritesh Vijay

Academy of Scientific and Innovative Research (AcSIR), Ghaziabad, Uttar Pradesh, 201002, India

You can also search for this author in PubMed   Google Scholar

Contributions

Jaydip Dey: Intensive review, sensors and analysis of data, limitations, and future research

Ritesh Vijay: Conceptualization, review, quality checking of data, and final editing

Corresponding author

Correspondence to Ritesh Vijay .

Ethics declarations

Ethics approval and consent to participate.

Not applicable.

Consent for publication

Competing interests.

The authors declare no competing interests.

Additional information

Responsible Editor: Xianliang Yi

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Dey, J., Vijay, R. A critical and intensive review on assessment of water quality parameters through geospatial techniques. Environ Sci Pollut Res 28 , 41612–41626 (2021). https://doi.org/10.1007/s11356-021-14726-4

Download citation

Received : 11 January 2021

Accepted : 01 June 2021

Published : 08 June 2021

Issue Date : August 2021

DOI : https://doi.org/10.1007/s11356-021-14726-4

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Water quality
• Optically active constituents
• Inherent and apparent properties
• Empirical methods
• Bio-optical methods
• Remote sensing
• Find a journal
• Publish with us
• Track your research

• Open access
• Published: 10 May 2019

Factors influencing perceptions of private water quality in North America: a systematic review

• Abraham Munene   ORCID: orcid.org/0000-0001-9546-3574 1 &
• David C. Hall 1  

Systematic Reviews volume  8 , Article number:  111 ( 2019 ) Cite this article

4004 Accesses

20 Citations

Metrics details

An estimated four million and 43 million people in Canada and the USA use private water supplies. Private water supplies are vulnerable to waterborne disease outbreaks. Private water supplies in Canada and the USA are often unregulated and private water management is often a choice left to the owner. Perceptions of water quality become important in influencing the adoption of private water stewardship practices, therefore safeguarding public health.

We conducted a systematic literature review to understand factors that shape perceptions of water quality among private water users. We searched six computer databases (Web of science, Medline, Scopus, EBSCO, PubMed and Agricola). The search was limited to primary peer-reviewed publications, grey literature and excluded conference proceedings, review articles, and non-peer review articles. We restricted the search to papers published in English and to articles which published data on surveys of private water users within Canada and the USA. The search was also restricted to publications from 1986 to 2017. The literature search generated 36,478 records. Two hundred and four full text were reviewed.

Fifty-two articles were included in the final review. Several factors were found to influence perceptions of water quality including organoleptic preferences, chemical and microbiological contaminants, perceived risks, water well infrastructure, past experience with water quality, external information, demographics, in addition to the values, attitudes, and beliefs held by well owners.

Conclusions

Understanding the factors that shape perceptions of water quality among private water users is an important step in developing private water management policies to increase compliance towards water testing and treatment in Canada and the USA. As many jurisdictions in Canada and the USA do not have mandatory private water testing or treatment guidelines, delineating these factors is an important step in informing future research and guiding policy on the public health of private water systems.

Peer Review reports

Introduction

An estimated four million and 43 million people in Canada and the USA use private water supplies [ 1 , 2 ]. In the absence of municipal water distribution systems in rural populations, private water supplies are an alternative source of domestic water in developed countries. Private water supplies are vulnerable to waterborne disease outbreaks [ 3 , 4 ]. Several chemical and microbiological contaminants can contaminate private water supplies. Nutrients (e.g. nitrates), pathogens, pharmaceuticals, hormones, heavy metals, nanomaterials and personal care products are some contaminants that have been identified in well water [ 5 , 6 , 7 , 8 , 9 , 10 ]. These contaminants are associated with illnesses including gastrointestinal illnesses, liver and kidney problems, endocrine disruption, cancer, reproductive issues and neurological disorders [ 11 ].

Private water supplies in Canada and the USA are often unregulated. Management of private water supplies (e.g. water wells, cisterns or boreholes) is the responsibility of the owner. As a guide to drinking water quality standards, the Guidelines for Canadian Drinking Water Quality set out national drinking water standards in Canada. Similarly, the Safe Water Drinking Act is used to set out national and enforceable drinking water guidelines in the USA. However, the legislation excludes private water sources that serve less than 25 people. Private water systems are defined by water systems that serve 25 people or less for at least 60 days within a year and have up to 15 service connections. Water wells make up the majority of private water systems with cisterns and residential wells also considered as private water systems [ 12 ]. Approximately four million and over 13 million people are estimated to rely on unregulated private water wells in Canada and the USA [ 13 , 14 ]. Although both Canada and the USA have national guidelines for the minimum standards of drinking water quality, there may be jurisdictional differences in the contaminants that are assessed [ 15 ]. Furthermore, individual provinces or states may have their own regulations on the construction of new wells, how many service connections can be served by a private water supply, and water testing recommendations, with some provinces or states requiring mandatory testing of wells upon the acquisition of new properties [ 12 , 16 ]. Unlike municipal water supplies which may be regularly monitored and treated, few regulations cater to testing and treatment of private water supplies in Canada and the USA [ 1 , 16 , 17 , 18 ].

In the absence of regulations on the management of private water supplies, compliance to private water testing and treatment becomes an essential mitigation strategy in protecting the health of private water users from diseases that could be contracted from consuming contaminated water. Recent studies indicate that compliance towards private water testing and treatment recommendations in various jurisdictions is low [ 2 , 19 ]. Roche et al. (2013) found that nearly 80% of respondents in their survey tested water quality at frequencies below the current provincial recommendations. Perceptions of water quality may influence the adoption and implementation of private water management practices [ 20 ]. The choice of when to test water quality, what to test for, and what treatment devices to use on private water systems are decisions that are based on both perception and knowledge of risks to private water contamination.

Perceptions have been broadly defined as a human being’s primary cognitive contact with the environment or simply the way in which we understand the world around us using our senses [ 21 ]. However, this narrow definition based on the sensory appraisal we make to understand our environment is myopic and does not capture the complexity of factors involved in shaping perceptions. Perception also has subjective components that are associated with learning and past experiences that are mediated by attention, memory, and the ability to retrieve information from memory [ 22 ].

Consequently, this raises the question; what factors are important in shaping the perceptions private water users have of their water quality? Little is known about the factors that influence perceptions of water quality among private water users. We conducted a systematic review of studies on people reliant on private water systems for domestic use in both Canada and the USA to determine the factors that influence the perceptions of water quality within these two countries. Describing and understanding the factors that shape perceptions of water quality among private water users is an important step in developing well water management policies to increase compliance towards private water stewardship practices such as water testing and treatment in Canada and the USA. To guide the scope of our systematic review, we wanted to answer the main question. What factors predominantly drive perceptions of private water quality in Canada and the USA?

PICO framework

A PICO framework [ 23 ] was used to help guide the questions of the review. As most studies included were observational, assessment for control groups was not feasible as there would be no adequate comparison for perceptions held by private water users to a similar group (Table  1 ).

Search strategy

Literature searches were made on both health and environmental databases. A search strategy was developed in consultation with a research librarian and the review team. Our review was informed by methods for conducting systematic reviews in agri-food research [ 24 ]. We searched six computer databases (Web of science, Medline, Scopus, EBSCO, PubMed, and Agricola). The search was conducted between January and December 2017. The search was limited to primary peer-reviewed publications and grey literature. The search excluded conference proceedings, review articles, and non-peer review articles. We restricted the search to articles published in English and to articles which published data on private water users within Canada and the USA. The search was restricted to publications within the last 31 years (01/01/1986–31/12/2017). This time frame was used to capture recent amendments in regulations within the Safe Water Drinking Act and the Guidelines for Canadian Drinking Water Quality which may influence what substances are considered as drinking water contaminants and at what maximum acceptable concentration (MAC). A combination of search phrases was used for each database but consisted of major search domains with associated synonyms required to capture relevant articles. Keywords searched were private water , domestic water , household water , well water , drinking water , perceptions , knowledge , belief , attitude , information , awareness , testing , treatment , survey, and rural . Reference lists for relevant primary articles and review articles were screened. Articles fitting the inclusion criteria, that is, articles that were published in English, articles that conducted surveys on human participants relying on private water sources through questionnaires or interviews, articles that surveyed participants in Canada or the USA, articles that were primary research and articles that had the outcomes of private water testing, treatment or investigate alternative water use in the context of private water users were added to the final list. All study approaches were considered including quantitative, qualitative, and mixed methods. As the focus was on perceptions of water quality among private water users, studies that directly surveyed private water owners were included in the final literature search (Table  2 with key terms used and the number of papers generated for each phrase search is provided. See Additional file  1 : Table S1).

Data extraction

Each paper included in the final review was read independently by the two authors and then assessed for relevancy in the review. The lead author extracted the following information: the main purpose of the study, the study population, study approach, methods of data collection, whether theoretical frameworks were used in the study, notes on the context of use of the private water systems, whether a formal intervention was present and results. The author also constructed a table identifying the study type, demographics, the intervention being evaluated and results of relevance to the present study (Additional file  2 : Table S2). The second author independently verified data extraction and tabulation for the included articles. Each article included in the final list was independently rated by both authors for relevance to the review. Both authors met regularly over a period of 4 months to discuss the findings. In instances of disagreement, articles were reassessed independently, and consensus was reached following deliberation and discussion by the authors. A PRISMA flow diagram was used to narrow our selection of articles [ 25 ]. Articles were preliminarily screened by (1) reviewing the article titles generated by the keywords search, (2) reviewing article abstracts, (3) reviewing the full articles, and (4) sorting on relevance for the review.

Quality of study and risk of bias

As a measure of the quality of study, articles were evaluated by whether they were published in a peer reviewed journal (as the assumption is that articles published in reviewed journals have been adequately scrutinised by reviewers before publication) or were technical reports. Reviewers also ranked the quality of the study relative to the review’s objectives on a scale. A risk of bias assessment from each study was conducted using the Strobe checklist assessment for risk of bias. Studies were ranked on a scale of 1 (high quality) to 4 (low quality) for their relevance to the review and based on the strobe checklist.

The database search included 36,478 articles using the keyword search. Web of Science ( n  = 4160), Medline OVID ( n  = 286), Scopus ( n  = 3875), PubMed ( n  = 4072), Agricola ( n  = 5506), EBSCO ( n  = 18,579). Ultimately, 204 papers were examined intensively of which 152 articles were excluded for not meeting the relevance criteria for this study (Fig.  1 ). Fifty-two studies were included in the final review. Of the 52 studies identified, 44 exclusively focused on surveys delivered to private water supply owners while ten studies surveyed both residents with private and municipal supplies. Most of the articles ( n  = 35) were from the USA while 17 articles reported on private water users in Canada. All studies were observational. Most of the studies used a cross-sectional design ( n  = 49) with the rest reporting on case control studies. Studies were also classified as quantitative ( n  = 48), mixed methods ( n  = 3) or qualitative ( n  = 3). Survey administration methods varied. Questionnaire mail deliveries were used in 35 out of 52 studies and telephone surveys were used in 11 out of 52 studies. Other methods used to elicit participation included face to face interviews (3 out of 52) and focus groups (6 out of 52).

PRISMA flow diagram for study selection

This systematic review included 52 journal articles with data collected on over 35,000 well water owners across Canada ( n  = 14,793) and the USA ( n  = 22,420). Perceptions of well water quality across Canada and the USA were found to be influenced by several factors. The main factors identified through this review were organoleptic properties of water, knowledge of chemical and microbiological contaminants, perceived risk, demographic factors, past experience with water quality, external information, values, attitudes, and beliefs about water, and water infrastructure.

Organoleptic properties of water

Private water owners primarily relied on the sensory properties of drinking water sourced from their wells when it came to decisions regarding well management options. Decisions on when to test water quality or the choice to consume water were often instigated by changes in either the taste, look or smell of the water [ 1 , 2 , 17 , 18 , 19 ]. Satisfaction with the organoleptic properties of water was not necessarily equated to concern over drinking water sourced from the wells. For example, although most respondents rated the organoleptic properties of their water from ‘good’ to ‘very good’, nearly 80% of respondents to the survey indicated being concerned about their water quality [ 1 ]. In contrast, organoleptic properties of drinking water were congruent with the perceptions of the safety of water for consumption. About 67% of participants who had issues with the organoleptic properties did not consider their water as safe to consume [ 26 ]. Sensory cues derived from the organoleptic properties of water were not only limited to water consumed but also to other water uses. For example, some people reported on the hardness of their well water as it tended to discolour their appliances or plumbing systems [ 17 ]. Due to psychological factors, people expect sensorial information on the taste odour, and colour of water to be congruent [ 20 ]. However, what is not clear from the studies is what sense dominated when well water owners indicated a change in their well water quality. Evidence on how well water owners perceive the taste, smell, and odour of water sourced from their wells relative to alternative water sources such as bottled water or municipal tap water was also evaluated in some studies. Well water owners were unwilling to change to municipal water supplies due to their personal preference for the taste of their well water and fear of ‘chemicals’ in city water [ 27 ]. Similarly, Jones et al. (2005) found that well water owners preferred their well water over bottled water due to preferences in taste and scepticism to where the bottled water came from.

Chemical and microbiological contaminants

Due to the soluble properties of water, several chemical and microbiological substances can be found in private water sources. Some chemical and microbiological substances can pose a health risk to individuals consuming well water. Of the 52 articles included, 13 out of 52 exclusively focused on assessing exposure to naturally occurring arsenic. Nitrate exposure was exclusively evaluated in 5 out of 52 articles. Radon exposure was exclusively evaluated in 1 out of 52 articles while 2 out of 52 articles evaluated the exposure of Escherichia coli and total coliforms on well water. Thirty-four articles were non-specific towards the chemical or microbiological contaminants (e.g. general microbiological and chemical contamination or a combination of both). For studies exclusively focusing on exposure to one contaminant, some studies were clear about the thresholds for the MAC for contaminants. The MAC is the specific level of a contaminant that is allowed in water for a specific purpose (e.g., human consumption). The contaminants assessed for and the MAC for specific contaminants may vary regionally [ 15 ]. However, the MAC’s of several contaminants in Canada and the USA are similar and reflect standards set out by the US Environmental Protection Agency. Studies examining naturally occurring arsenic as an exposure often quoted 10 μg/l as the MAC [ 2 , 28 , 29 ]. For nitrate as an exposure, the MAC used was 10 mg/L in six studies [ 30 , 31 , 32 ]. Lead exposure was assessed in one of the studies [ 33 ] with some studies quoting MAC’s for each contaminant assessed [ 34 ]. However, in some of the studies that assessed multiple exposure to contaminants or that were non-specific to a contaminant, MAC’s were not used [ 26 , 35 , 36 , 37 ].

Some studies also included a water testing component to evaluate the prevalence of contaminants of interest in their samples (Table  3 ).

Knowledge of the level of contaminants within well water was an important factor when well owners had to decide on treatment systems to use in their wells. For example, half of respondents indicated they would begin treating or finding other water sources before the concentration reached the MAC 10 mg/l of nitrates in their well water. Interestingly, a similar proportion of participants indicated that they would wait until the concentration of nitrates in their water was > 10 mg/l or higher [ 32 ]. However, the authors noted that stated intentions differed from the actual responses with only 21.9% opting to use a treatment system and about 25% opting to switch to bottled water and drilling a new well upon learning of exceedances. Flanagan et al. (2015) found that about 43% of well water owners installed water treatments with a further 30% seeking alternative water sources after being informed of exceedances in the MAC of arsenic in their well water. Therefore, even though some well owners knew their water wells exceeded the MAC for nitrates, their decision to adopt treatment or use alternative water sources may have been influenced by the perceived risk of the contaminant towards their health. These findings demonstrate the complexity in how the appraisal of the risks posed by contaminants may be highly subjective to individuals.

Perceived risk

Individuals respond to hazards they perceive within their environment. Risk perception is defined as the subjective judgement that an individual makes about the characteristics and severity of a risk [ 38 ]. In order for an individual to make a decision on whether or not to use treatments or seek alternative drinking water sources, they must first identify the hazard (e.g. nitrates), evaluate the risk of contamination based on potential risk factors in their environment and their exposure to hazards (e.g. test for contamination in an area with extensive manure or fertiliser spread and understand how likely they are to be exposed to nitrate contamination) and finally they must understand the consequences of the hazard and their ability to control those consequences (e.g. know about a health risk such as methemoglobinemia and make a judgement on the severity of the methemoglobinemia towards their own health). Given there are several contaminants that may be considered hazards to well water, the process of assessing the risk of general well water contamination without a specific preidentified hazard may be problematic for well owners therefore making the decision of treatment options, whether to switch to an alternative or what and when to test for water quality more difficult [ 1 ]. Individuals were less likely to drink well water if they thought there were health risks associated with consuming water with arsenic [ 39 ]. Similarly, well water owners were less likely to drink well water if they perceived a risk in drinking well water regardless of aesthetic concerns [ 40 , 41 ]. Perceived risk factors within the environment could also influence what people think of their well water quality. Participants reported proximity to livestock, proximity to septic systems, proximity to oil and gas activities, proximity to mining areas, proximity to nuclear power plants, flooding, severe runoff events, and drought as environmental risks that caused concern and motivated well owners to test their water [ 27 , 34 , 42 , 43 , 44 , 45 , 46 ]. However, the perceptions of water quality in response to environmental risk factors were indirectly mediated by actual changes in the aesthetic properties of water as some participants noted.

Demographic factors

Demographics can influence the choices well water owners make of drinking water options. Factors such a participant’s education, income, number of years within a residence, and place of residence have been noted as important factors that influenced perceptions of water quality and the willingness to use water treatment [ 26 , 29 , 30 , 47 , 48 ]. Low education and income were more likely to result in the lack of use of well water treatment devices [ 26 , 30 ]. Low education and income may also be socioeconomic factors that predispose well water owners to certain risk factors. Garcia et al. (2016) noted that residents living in underprivileged communities within New Mexico had unreliable drinking water systems, poor sanitation, and a lack of access to water testing and treatment. Despite the risk of arsenic being randomly distributed within socioeconomic groups, individuals with lower income and lower education were less likely to adopt protective behaviours such as well testing and treatment for their water wells [ 49 ]. Furthermore, psychological factors influencing testing and treatment were more prevalent among those with higher income and education. Similarly, higher education and income were positively associated with the decision to test well water quality and use water treatment devices [ 50 , 51 ]. Education and income were not always associated with positive outcomes on treatment and testing. No significant association was found between education and stewardship behaviours conducted by well water owners [ 52 ]. Similarly, no significant association was found between education and income and the use of well water treatments [ 34 , 46 ]. In contrast, Shaw et al. (2005) found a negative association between income, education, and the decision to use well water treatments. The number of years an individual had lived at a residence and the length of time they had used their well water also seemed to play an important role in predicting water testing and treatment behaviour. This is because well owners may get habituated to their drinking water source. Shaw et al. (2005) found that the longer an individual had lived in the household, the less likely they were to engage in well water testing behaviour. Similarly, the longer an individual had lived within the household, the less likely they were to conduct a water quality test within the last 5 years and the less likely they were willing to submit a water quality test [ 18 ]. However, some studies failed to find a significant association between the number of years lived within the home and water treatment practices [ 51 ].

Age and gender have also been explored as demographic variables that can influence perceptions of private water quality. Evidence to show associations between age and gender on perceptions of well water quality has been sparse. Age and gender did not predict well water testing behaviour among well owners [ 50 ]. With respect to gender, a significant association has been found between women and the use of well water treatment systems. This is because the presence of children within a household may be identified as a reason for concern among parents and a reason for well water owners to choose alternative drinking water sources [ 44 , 45 , 50 , 53 , 54 ].

Past experience

The role of past experience with water quality issues is important. Past negative experiences with well water quality were found to predict well water testing behaviour [ 55 ]. These experiences were either on the individual well or within the well owners’ community. Learning of water contamination among neighbours and experiencing unexplained gastrointestinal illness were noted as motivators for individuals to conduct well water testing [ 18 ]. Despite past negative experiences being noted to influence perception of drinking water quality, determining the validity of reported past negative experiences may be subject to recall bias among surveyed participants. Furthermore, participants may not always attribute personal health problems, such as gastrointestinal illness, to drinking water from their wells. As gastrointestinal illnesses may be underreported and deemed controllable, it may be difficult to get an accurate representation of how past negative experiences with gastrointestinal illness influence perceptions of well water quality [ 54 , 56 , 57 ]. For well water contaminants which do not present direct clinical symptoms and may have severe health consequences due to chronic exposure (e.g., arsenic), the role of past experience associated with negative health outcomes on perception of well water quality is difficult to determine. However, past negative experience with contamination indicated by well water testing may change the perspective of well water owners with regards to the safety of their drinking water [ 32 ].

Previous positive experience with water quality testing may also influence the likelihood of well owners testing water quality in the future. For example, well owners reported being more confident in their well water supplies and therefore less likely to test their well water quality if the result of the water quality test they had conducted in the past showed no evidence of contamination [ 1 ]. Recurrent problems with well water quality as indicated by water quality test may also cause individuals to worry more about their well water quality and therefore conduct frequent testing. For example, well owners who were identified as being high risk for arsenic contamination through water testing and who knew they were at a higher risk of arsenic contamination were more likely to conduct well water testing than individuals who were identified as low risk for arsenic contamination [ 49 ]. Similarly, well owners who had engaged in previous water testing and were aware of water quality issues were more likely to conduct routine testing [ 58 ].

External information

The impact of external information on changing perceptions towards well water quality to promote testing or treatment has been explored. External information sources may be in the from media campaigns, educational awareness programs or from prompts given by members of the society to encourage a behaviour. The format of the information presented may by varied including pamphlets and flyers distributed by public and private water public health agencies, news items, advertisements or advisories distributed through print media, social media, television or radio, information workshops, information solicited directly from water public health agencies (e.g. through phone calls) or information gathered from social informants (e.g. neighbours and friends) [ 2 , 29 , 40 , 58 , 59 ]. Participants’ responses to educational material may be varied. Nearly 43% of participants installed water treatment systems in response to elevated arsenic levels while nearly 31% switched to alternative drinking water sources [ 18 ]. Similarly, well owners were more likely to report higher arsenic testing rates in towns that had received educational intervention programs when compared to towns that did not receive programs [ 60 ]. In response to media reports on the risk of cancer associated with arsenic exposure, only 18% of participants used mitigation strategies that were useful against arsenic despite 66% having arsenic concentrations above the MAC [ 29 , 39 , 49 ]. Chappells et al. (2015) found that nearly 25% of participants reported making some change to their well water management practice in response to information received from either private testing laboratories or government departments. Well owners were more likely to engage in well testing programs after the dissemination of well management information through a well stewardship program [ 55 ]. Information on well water quality in the form of testing results can also be used to change participants’ perceptions of the safety of their drinking water [ 31 ]. Interestingly, not all information campaigns may increase water well stewardship. Nearly 28% of participants did not take any well stewardship action despite being aware of elevated arsenic concentrations within their well water [ 18 ]. Therefore, exposure to media or other forms of external information may not be sufficient to modify well stewardship behaviour [ 58 ].

Values, attitudes, and beliefs

Values, attitudes, and beliefs towards health or environmental protection may also influence well owners' willingness to adopt well stewardship practices. Well owners’ decisions to conduct stewardship practices were more influenced by whether they were satisfied with their water quality and with their knowledge and beliefs of water quality [ 46 ]. Satisficing was where well owners took on a simple belief about their water well and did not develop a strong enough knowledge base to accurately make judgements of their water quality. Furthermore, most individuals in their survey believed that it was best to not to do anything with the water well unless they had issues with it. Participants also held a wide variety of beliefs when it came to their water wells and these beliefs were not necessarily associated with negative health consequences. The role of imperfect and incomplete knowledge (e.g. wrong beliefs about aquifers and the origin of water in water wells) in the decisions of whether to adopt well stewardship practices was identified as a possible barrier [ 46 ].

Well water infrastructure

Available infrastructure, both physical and services available for well water quality maintenance, may also influence stewardship practices. The availability of free well water testing services has often been used to encourage water quality testing among well water owners [ 1 , 18 , 19 , 52 ]. Despite testing services being offered for free in several jurisdictions in Canada and the USA, compliance towards well water testing recommendations is usually low [ 19 , 60 ]. Several barriers have been identified that inhibit well water owners from conducting regular testing. Individual well owners may face multiple barriers when deciding to go through with water testing [ 46 , 52 ] (Table  4 ). To increase compliance towards well water testing, several studies have solicited participants’ suggestions on how to increase routine well water testing. Making pick up and drop off of water sampling kits more accessible, increasing reminders to participants to conduct water quality tests, increasing educational awareness forums, providing incentives, enforcing penalties or making well water testing mandatory through legislation have all been stated as possible measures to increase compliance towards well water testing. The availability and accessibility of infrastructure for well water treatment may also influence habits towards well water protection. However, few studies have explored the reasons behind well water owner’s choice of well water treatments

This systematic review included 52 journal articles with data collected from well water owners in Canada and the USA. Perceptions of well water quality across Canada and the USA were found to be influenced by several factors. Main factors identified through this review were organoleptic properties of water, knowledge of chemical and microbiological contaminants, perceived risk, demographic factors, past experience with water quality, external information, values, attitudes, and beliefs about water, and water infrastructure. The reliance on the organoleptic properties of water to make judgements on the safety of drinking water by private water users is profound and has been identified as a key factor in other reviews [ 20 ]. To the best of our knowledge, only two previous literature reviews [ 20 , 61 ] had attempted to provide a review on factors influencing perceptions of water quality.

Well water management practices are discussed in the context of testing and/or treatments. Well water testing practices often tend to be the focus for researchers and intervention strategies [ 1 , 19 , 52 , 60 , 62 ]. Widespread adoption of well testing and compliance towards recommendations set for testing tend to be problematic for well water owners to achieve. Interventions focusing on modifying well water testing behaviour based on incentives, legislation, education or community outreach activities have had moderate success on increasing compliance towards well water testing [ 2 , 52 , 60 ].

Interventions based on getting well water owners to adopt well water treatment are contingent on well owners understanding contaminants and the potential health risks they may pose. However due to the variety of possible contaminants found within well water, it may be very difficult to prescribe treatment devices, unless a contaminant is identified through testing, as one device may not be effective at removing all contaminants. The use of multiple well water treatment devices may offer more protection against several contaminants; however, water testing will still need to verify well water quality and identify possible risks to a well. Therefore, educating private water users on options available for them with respect to water treatment may enable private water users make more informed decisions based on the identified risks to their private water sources. The need for more information on water treatment has been identified in previous surveys [ 1 , 19 , 59 ]. Information from this study will be useful in informing private water users, researchers, and educators on some of the present gaps in the literature and research areas that need to be expanded on.

Gaps identified

Despite the focus on well water testing, very few studies have tried to discriminate which health risks are perceived to be associated with drinking water contamination and more specifically towards individual contaminants [ 63 ]. More studies are required to address this gap in knowledge between the perception of well water quality and the potential health consequences well water owners attribute to well water contamination.

Maintenance of well water stewardship behaviour such as testing, post intervention, is also an issue that has yet to be adequately addressed. Despite the role research may play in active surveillance of well water and instigating well owners to conduct water testing during the duration of the research program, there is very little evidence that behaviour such as water testing is continued after the research programs or other intervention programs end. Future research should look into assessing if well water testing behaviour is maintained among well owners and this could be done by broadening the methods to include cohort studies and not only cross-sectional designs. Broadening active surveillance periods using research may also help in determining the period prevalence of well contamination over time and address reliability issues associated with surveys by following up on well owners’ behaviour, in addition to determining the maintenance of well water stewardship practices.

Despite the amount of research that has been conducted on well water testing behaviour, compliance towards well water testing recommendations is still considered low in many jurisdictions. Changes in technology over the last 30 years and increased internet connectivity in Canada and the USA may provide well water owners with more access to information regarding their water wells. However, a potential problem that arises is what information sources should well owners trust given that current policies in well stewardship are only recommendations. More studies need to be conducted on the quality of information provided for by interventions such as educational programs or online information. Assessing the quality of information and how it is understood by well water owners may influence the adoption of well stewardship behaviours and may be important in dealing with satisficing and complacency among well owners. Furthermore, more research needs to be conducted on sources of information private water owners have access to and the uptake of information based on its trustworthiness [ 46 , 47 ].

The adoption of qualitative and mixed method designs to further study perceptions of private water quality over the last decade and the shift away from quantitative studies has helped in developing a richer understanding of the issues faced by well water owners with respect to water quality. Qualitative and mixed methods research may be more beneficial in capturing the unique personal experiences and knowledge private water owners have of their water quality. Furthermore, incorporating the voice of private water owners in research may be an important step in developing well management policy and practices that will directly tap into the needs of private water users.

Despite having identified factors that influence well owners' perceptions of well water quality, it is important to note the paucity of research on how combinations of these factors influence well stewardship behaviour. There is very little evidence to suggest that perceptions of well water quality and well stewardship practices (i.e. testing and treatment) are driven by a single factor and are more likely to be influenced by a combination of several factors. While research to date has done an adequate job of identifying factors that influence perceptions of well water quality and predictors of well water stewardship, there is a knowledge gap in how these factors interact with each other to produce the desired outcome (e.g. well testing) in well owners. For example, although external information (e.g. educational forums) may help encourage well testing, if well owners conduct a well test and have a negative test result due to the educational program, how does the past experience of having a negative well test result influence both their appraisal of susceptibility to well water contamination and their willingness to test their water in the future. More research is required on how factors that influence perceptions of water quality may act synergistically or antagonistically to influence well stewardship behaviour.

This review summarises research that has been conducted on well water owners’ perceptions of water quality over the last 30 years while identifying questions and areas that need further development in research. Policies and recommendations for well water testing, treatment, and other management practices are highly contextual to the regions; however, this study summarises the most pertinent factors driving perceptions of private water based on research that has been conducted.

Limitations

Publication bias may have been present due to the selection of articles from peer reviewed journals. Furthermore, because we only selected articles published within the last 30 years, there may have been a time lag bias with the selection of articles [ 64 ]. Despite the search for articles and selection of articles relevant for the review being restricted to the language spoken by the authors, no systematic bias has been found in reviews published in English [ 65 ].

Although their may be relationships between education and income to private water stewardship behaviours, it was difficult to operationalise or standardise income and education variables. This was because of differences in education standards and currency between Canada and the USA, income levels within different jurisdictions, and changes to income and education levels over a 30-year period. Furthermore, it was difficult to operationalise variables such as income and education levels because of differences in the what researchers choose to operationalise as ‘low education’ and ‘low income’ within their studies.

Given that perceptions of water quality among private water users are influenced by several factors, researchers, educators and policy makers should appreciate the heterogeneity and interplay of these factors when planning private water management programs or developing policies. Education and communication strategies that focus more on individual well owners and their needs, based on risks identified around their well, need to be adopted as opposed to blanket policies or programs. The use of questionnaire surveys and qualitative research to identify the needs of individual well owners may help. This is especially pertinent because of the different interacting, and sometimes confounding, factors that may motivate private water users to comply with water testing and treatment recommendations.

Jones AQ, Dewey CE, Dore K, Majowicz SE, McEwen SA, David WT, Eric M, Carr DJ, Henson SJ. Public perceptions of drinking water: a postal survey of residents with private water supplies. BMC Public Health. 2006;6:94.

Article   Google Scholar  

Flanagan SV, Marvinney RG, Zheng Y. Influences on domestic well water testing behavior in a Central Maine area with frequent groundwater arsenic occurrence. Sci Total Environ. 2015;505:1274–81.

Article   CAS   Google Scholar  

Said B, Wright F, Nichols GL, Reacher M, Rutter M. Outbreaks of infectious disease associated with private drinking water supplies in England and Wales 1970–2000. Epidemiol Infect. 2003;130:469–79.

Schuster CJ, Ellis AG, Robertson WJ, Charron DE, Aramini JJ, Marshall BJ, Medeiros DT. Infectious disease outbreaks related to drinking water in Canada, 1974-2001. Can J Public Health. 2005;96:254–8.

PubMed   Google Scholar  

Hooda PS, Edwards AC, Anderson HA, Miller A. A review of water quality concerns in livestock farming areas. Sci Total Environ. 2000;250:143–67.

Campagnolo ER, Johnson KR, Karpati A, Rubin CS, Kolpin DW, Meyer MT, Esteban JE, Currier RW, Smith K, Thu KM, McGeehin M. Antimicrobial residues in animal waste and water resources proximal to large-scale swine and poultry feeding operations. Sci Total Environ. 2002;299:89–95.

Kolodziej EP, Harter T, Sedlak DL. Dairy wastewater, aquaculture, and spawning fish as sources of steroid hormones in the aquatic environment. Environ Sci Technol. 2004;38:6377–84.

Johnson AC, Williams RJ, Matthiessen P. The potential steroid hormone contribution of farm animals to freshwaters, the United Kingdom as a case study. Sci Total Environ. 2006;362:166–78.

Burkholder J, Libra B, Weyer P, Heathcote S, Kolpin D, Thorne PS, Wichman M. Impacts of waste from concentrated animal feeding operations on water quality. Environ Health Persp. 2007;115:308–12.

Coleman BL, Louie M, Salvadori MI, McEwen SA, Neumann N, Sibley K, Irwin RJ, Jamieson FB, Daignault D, Majury A, et al. Contamination of Canadian private drinking water sources with antimicrobial resistant Escherichia coli. Water Res. 2013;47:3026–36.

Villanueva CM, Kogevinas M, Cordier S, Templeton MR, Vermeulen R, Nuckols JR, Nieuwenhuijsen MJ, Levallois P. Assessing exposure and health consequences of chemicals in drinking water: current state of knowledge and research needs. Environ Health Persp. 2014;122:213–21.

Centre for Disease Control (CDC). Private Groundwater wells 2019. Available from: https://www.cdc.gov/healthywater/drinking/private/wells/index.html .

Charrois JWA. Private drinking water supplies: challenges for public health. Can Med Assoc J. 2010;182(10):1061–4.

US Environmental Protection Agency (EPA). Private Drinking Water Wells 2019. Available from: https://www.epa.gov/privatewells .

Dunn G, Bakker K, Harris L. Drinking water quality guidelines across Canadian provinces and territories: jurisdictional variation in the context of decentralized water governance. Int J Environ Res Public Health. 2014;11:4634–51.

Fox MA, Nachman KE, Anderson B, Lam J, Resnick B. Meeting the public health challenge of protecting private wells. Proceedings and recommendations from an expert panel workshop. Sci Total Environ. 2016;554:113–8.

Jones AQ, Dewey CE, Dore K, Majowicz SE, McEwen SA, Waltner-Toews D, Henson SJ, Mathews E. Public perception of drinking water from private water supplies: focus group analyses. BMC Public Health. 2005;5:129.

Flanagan SV, Marvinney RG, Johnston RA, Yang Q, Zheng Y. Dissemination of well water arsenic results to homeowners in Central Maine: influences on mitigation behavior and continued risks for exposure. Sci Total Environ. 2015;505:1282–90.

Roche SM, Jones-Bitton A, Majowicz SE, Pintar KDM, Allison D. Investigating public perceptions and knowledge translation priorities to improve water safety for residents with private water supplies: a cross-sectional study in Newfoundland and Labrador. BMC Public Health. 2013;13:1225.

de França Doria M. Factors influencing public perception of drinking water quality. Water Policy. 2010;12:1–19.

Deonna JA. Emotion, perception and perspective. Dialectica. 2006;60:29–46.

Hochberg J. Perception: toward the recovery of a definition. Psychol Rev. 1956;63:400–5.

Schardt C, Adams MB, Owens T, Keitz S, Fontelo P. Utilization of the PICO framework to improve searching PubMed for clinical questions. BMC Med Inform Decis. 2007;7:16.

Sargeant JM, Rajic A, Read S, Ohlsson A. The process of systematic review and its application in agri-food public-health. Prev Vet Med. 2006;75:141–51.

Moher D, Liberati A, Tetzlaff J, Altman DG, The PG. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med. 2009;6:e1000097.

Garcia LB, Sobin C, Tomaka J, Santiago I, Palacios R, Walker WS. A comparison of water-related perceptions and practices among West Texas and South New Mexico Colonia residents using hauled-stored and private well water. J Environ Health. 2016;79:14–20.

Murti M, Yard E, Kramer R, Haselow D, Mettler M, McElvany R, Martin C. Impact of the 2012 extreme drought conditions on private well owners in the United States, a qualitative analysis. BMC Public Health. 2016;16:430.

Colt JS, Baris D, Clark SF, Ayotte JD, Ward M, Nuckols JR, Cantor KP, Silverman DT, Karagas M. Sampling private wells at past homes to estimate arsenic exposure: a methodologic study in New England. J Expo Anal Env Epid. 2002;12:329–34.

Shaw WD, Walker M, Benson M. Treating and drinking well water in the presence of health risks from arsenic contamination. Results from a US hot spot. Risk Anal. 2005;25:1531–43.

Schwartz JJ, Waterman AB, Lemley AT, Wagenet LP, Landre P, Allee DJ. Homeowner perceptions and management of private water supplies and wastewater treatment systems. J Soil Water Conserv. 1998;53:315–9.

Google Scholar  

Poe GL, van HM E, VandenBerg TP, Bishop RC. Do participants in well water testing programs update their exposure and health risk perceptions? J Soil Water Conserv. 1998;53:320.

Lewandowski AM, Montgomery BR, Rosen CJ, Moncrief JF. Groundwater nitrate contamination costs: a survey of private well owners. J Soil Water Conserv. 2008;63:153–61.

Pieper KJ, Krometis LAH, Gallagher DL, Benham BL, Edwards M. Incidence of waterborne lead in private drinking water systems in Virginia. J Water Health. 2015;13:897–908.

Postma J, Butterfield PW, Odom-Maryon T, Hill W, Butterfield PG. Rural children’s exposure to well water contaminants: implications in light of the American Academy of Pediatrics’ recent policy statement. J Am Acad Nurse Prac. 2011;23:258–65.

Mechenich C, Shaw BH. Chemical use practices and opinions about groundwater contamination in two unsewered subdivisions. J Environ Health. 1994;56:17.

CAS   Google Scholar  

Strauss B, King W, Ley A, Hoey JR. A prospective study of rural drinking water quality and acute gastrointestinal illness. BMC Public Health. 2001;1:8.

Schade CP, Wright N, Gupta R, Latif DA, Jha A, Robinson J. Self-reported household impacts of large-scale chemical contamination of the public water supply, Charleston, West Virginia, USA. PLOS One. 2015;10:e0126744.

Slovic P, Fischhoff B, Lichtenstein S. Why study risk perception? Risk Anal. 1982;2:83–93.

Walker M, Shaw WD, Benson M. Arsenic consumption and health risk perceptions in a rural western US area. J Am Water Resour As. 2006;42:1363–70.

McLeod L, Bharadwaj L, Waldner C. Risk factors associated with the choice to drink bottled water and tap water in rural Saskatchewan. Int J Env Res Pub He. 2014;11:1626–46.

McLeod L, Bharadwaj L, Waldner C. Risk factors associated with perceptions of drinking water quality in rural Saskatchewan. Canadian Water Resources Journal / Revue canadienne des ressources hydriques. 2015;11(2):1–11.

Levallois P, Theriault M, Rouffignat J, Tessier S, Landry R, Ayotte P, Girard M, Gingras S, Gauvin D, Chiasson C. Groundwater contamination by nitrates associated with intensive potato culture in Quebec. Sci Total Environ. 1998;217:91–101.

Acharya M, Kalischuk RG, Klein KK, Bjornlund H. Farmstead drinking water sources, concerns and safety practices of livestock farm families in southern Alberta, Canada. Water Pollution IX. 2008;111:627–36.

McSpirit S, Reid C. Residents’ perceptions of tap water and decisions to purchase bottled water: a survey analysis from the Appalachian, big Sandy coal mining region of West Virginia. Soc Nat Resour. 2011;24:511–20.

Merkel L, Bicking C, Sekhar D. Parents’ perceptions of water safety and quality. J Commun Health. 2012;37:195–201.

Summers R. Alberta well water survey. A report prepared for Alberta environment. Edmonton: University of Alberta; 2010.

Chappells H, Campbell N, Drage J, Fernandez CV, Parker L, Dummer TJB. Understanding the translation of scientific knowledge about arsenic risk exposure among private well water users in Nova Scotia. Sci Total Environ. 2015;505:1259–73.

Mahler R, Barber M, Shafii B. Drinking water issues and concerns of urban residents of the Pacific Northwest, USA Urban Water 2; 2014. p. 139.

Flanagan SV, Spayd SE, Procopio NA, Marvinney RG, Smith AE, Chillrud SN, Braman S, Zheng Y. Arsenic in private well water part 3 of 3: socioeconomic vulnerability to exposure in Maine and New Jersey. Sci Total Environ. 2016;562:1019–30.

Straub CL, Leahy JE. Application of a modified health belief model to the pro-environmental behavior of private well water testing. J Am Water ResouR As. 2014;50:1515–26.

Lothrop N, Wilkinson ST, Verhougstraete M, Sugeng A, Loh MM, Klimecki W, Beamer PI. Home water treatment habits and effectiveness in a rural Arizona community. Water. 2015;7:1217–31.

Kreutzwiser R, de Loe R, Imgrund K, Conboy MJ, Simpson H, Plummer R. Understanding stewardship behaviour: factors facilitating and constraining private water well stewardship. J Environ Manag. 2011;92:1104–14.

Schubert C, Knobeloch L, Kanarek MS, Anderson HA. Public response to elevated nitrate in drinking water wells in Wisconsin. Arch Environ Health. 1999;54:242–7.

Feinman SJ, Ryan PB, Toth B, Honey WA, Gargano JW. Primary drinking water source and acute gastrointestinal illness: New Mexico, 2007. Water Qual Expos Hea. 2015;7:285–94.

Imgrund K, Kreutzwiser R, de Loe R. Influences on the water testing behaviors of private well owners. J Water Health. 2011;9:241–52.

Jones AQ, Majowicz SE, Edge VL, Thomas MK, MacDougall L, Fyfe M, Atashband S, Kovacs SJ. Drinking water consumption patterns in British Columbia: an investigation of associations with demographic factors and acute gastrointestinal illness. Sci Total Environ. 2007;388:54–65.

Johnson BB. Public views on drinking water standards as risk indicators. Risk Anal. 2008;28:1515–30.

Renaud J, Gagnon F, Michaud C, Boivin S. Evaluation of the effectiveness of arsenic screening promotion in private wells: a quasi-experimental study. Health Promot Int. 2011;26:465–75.

Ridpath A, Taylor E, Greenstreet C, Martens M, Wicke H, Martin C. Description of calls from private well owners to a national well water hotline, 2013. Sci Total Environ. 2016;544:601–5.

Flanagan SV, Spayd SE, Procopio NA, Chillrud SN, Ross J, Braman S, Zheng Y. Arsenic in private well water part 2 of 3: who benefits the most from traditional testing promotion? Sci Total Environ. 2016;562:1010–8.

Morris L, Wilson S, Kelly W. Methods of conducting effective outreach to private well owners—a literature review and model approach. J Water Health. 2016;14:167–82.

Flanagan SV, Spayd SE, Procopio NA, Chillrud SN, Braman S, Zheng Y. Arsenic in private well water part 1 of 3: impact of the New Jersey private well testing act on household testing and mitigation behavior. Sci Total Environ. 2016;562:999–1009.

Levallois P, Grondin J, Gingras S. Evaluation of consumer attitudes on taste and tap water alternatives in Québec. Water Sci Technol. 1999;40(6):135–9.

Egger M, Dickersin K, Smith GD. Problems and limitations in conducting systematic reviews. In Systematic Reviews in Health Care; 2008.

Morrison A, Polisena J, Husereau D, Moulton K, Clark M, Fiander M, Mierzwinski-Urban M, Clifford T, Hutton B, Rabb D. The effect of English-language restriction on systematic review-based meta-analyses: a systematic review of empirical studies. Int J Technol Assess. 2012;28:138–44.

Laflamme DM, VanDerslice JA. Using the behavioral risk factor surveillance system (BRFSS) for exposure tracking: experiences from Washington state. Environ Health Persp. 2004;112:1428–33.

Severtson DJ, Baumann LC, Brown RL. Applying a health behavior theory to explore the influence of information and experience on arsenic risk representations, policy beliefs, and protective behavior. Risk Anal. 2006;26:353–68.

Severtson DJ, Baumann LC, Brown RL. Applying the common-sense model to measure representations of arsenic contaminated well water. J Health Commun. 2008;13:538–54.

Slotnick MJ, Meliker JR, Nriagu JO. Effects of time and point-of-use devices on arsenic levels in southeastern Michigan drinking water, USA. Sci Total Environ. 2006;369:42–50.

Kite-Powell AC, Harding AK. Nitrate contamination in Oregon well water: geologic variability and the public's perception. J American Water Resour As. 2006;42:975–87.

Tabbot PN, Robson MG. The New Jersey residential well-testing program—a case study: Randolph township. J Environ Health. 2006;69:15–9.

CAS   PubMed   Google Scholar  

Hexemer AM, Pintar K, Bird TM, Zentner SE, Garcia HP, Pollari F. An investigation of bacteriological and chemical water quality and the barriers to private well water sampling in a southwestern Ontario community. J Water Health. 2008;6:521–5.

Swistock BR, Clemens S, Sharpe WE, Rummel S. Water quality and management of private drinking water wells in Pennsylvania. J Environ Health. 2013;75:60–6.

Paul MP, Rigrod P, Wingate S, Borsuk ME. A community-driven intervention in Tuftonboro, New Hampshire, succeeds in altering water testing behavior. J Environ Health. 2015;78:30–9.

CAS   PubMed   PubMed Central   Google Scholar  

Pintar KDM, Waltner-Toews D, Charron D, Pollari F, Fazil A, McEwen SA, Nesbitt A, Majowicz S. Water consumption habits of a south-western Ontario community. J Water Health. 2009;7:276–92.

Yu ZM, TJB D, Adams A, Murimboh JD, Parker L. Relationship between drinking water and toenail arsenic concentrations among a cohort of Nova Scotians. J Expo Sci Env Epid. 2013;24:135.

Malecki KMC, Schultz AA, Severtson DJ, Anderson HA, VanDerslice JA. Private-well stewardship among a general population-based sample of private well-owners. Sci Total Environ. 2017;601-602:1533–43.

Download references

Acknowledgements

We would like to thank the Diane Lorenzetti for her advise in formulating the search and selection strategy as well as Dr. Jocelyn Lockyer and Dr. Sylvia Checkley for their comments on developing the systematic review and their review of the manuscript.

This research was supported Alberta Innovates Energy and Environment Solutions (AIEES) grant number 2074 to Dr. David Hall and was the part of the literature review to inform the study investigating perceptions of well water quality in Alberta associated with livestock.

Availability of data and materials

All data generated or analysed during this study are included in this published article and its supplementary information files.

Author information

Authors and affiliations.

Department of Ecosystem and Public Health, Faculty of Veterinary Medicine, University of Calgary, 3280 Hospital Drive NW, Calgary, AB, T2N 4Z6, Canada

Abraham Munene & David C. Hall

You can also search for this author in PubMed   Google Scholar

Contributions

AM and DCH conceptualised and designed the study. AM conducted the literature search. AM and DCH analysed the data. AM drafted the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Abraham Munene .

Ethics declarations

Ethics approval and consent to participate.

The study was conducted under the University of Calgary study number 1025400-3-1 and approved by the Conjoint Faculties Ethics Research Board and by the Research Ethics Board (REB13-0473).

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Additional files

Additional file 1:.

Table S1. Search terms used and papers generated on each database with search terms. (DOCX 24 kb)

Additional file 2:

Table S2. Data abstraction for articles included in the systematic review. (XLSX 96 kb)

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License ( http://creativecommons.org/licenses/by/4.0/ ), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated.

Reprints and permissions

About this article

Cite this article.

Munene, A., Hall, D.C. Factors influencing perceptions of private water quality in North America: a systematic review. Syst Rev 8 , 111 (2019). https://doi.org/10.1186/s13643-019-1013-9

Download citation

Received : 23 May 2018

Accepted : 01 April 2019

Published : 10 May 2019

DOI : https://doi.org/10.1186/s13643-019-1013-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Private water
• Water wells
• Perceptions
• Well stewardship
• Public health

Systematic Reviews

ISSN: 2046-4053

• Submission enquiries: Access here and click Contact Us
• General enquiries: [email protected]

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Int J Environ Res Public Health

Drinking Water Quality and Human Health: An Editorial

Patrick levallois.

1 Direction de la santé environnementale et de la toxicologie, Institut national de la santé publique du Québec, QC G1V 5B3, Canada

2 Département de médecine sociale et préventive, Faculté de médecine, Université Laval, Québec, QC G1V 0A6, Canada

Cristina M. Villanueva

3 ISGlobal, 08003 Barcelona, Spain; [email protected]

4 Universitat Pompeu Fabra (UPF), 08002 Barcelona, Spain

5 Consortium for Biomedical Research in Epidemiology and Public Health (CIBERESP), Carlos III Institute of Health, 28029 Madrid, Spain

6 IMIM (Hospital del Mar Medical Research Institute), 08003 Barcelona, Spain

Drinking water quality is paramount for public health. Despite improvements in recent decades, access to good quality drinking water remains a critical issue. The World Health Organization estimates that almost 10% of the population in the world do not have access to improved drinking water sources [ 1 ], and one of the United Nations Sustainable Development Goals is to ensure universal access to water and sanitation by 2030 [ 2 ]. Among other diseases, waterborne infections cause diarrhea, which kills nearly one million people every year. Most are children under the age of five [ 1 ]. At the same time, chemical pollution is an ongoing concern, particularly in industrialized countries and increasingly in low and medium income countries (LMICs). Exposure to chemicals in drinking water may lead to a range of chronic diseases (e.g., cancer and cardiovascular disease), adverse reproductive outcomes and effects on children’s health (e.g., neurodevelopment), among other health effects [ 3 ].

Although drinking water quality is regulated and monitored in many countries, increasing knowledge leads to the need for reviewing standards and guidelines on a nearly permanent basis, both for regulated and newly identified contaminants. Drinking water standards are mostly based on animal toxicity data, and more robust epidemiologic studies with an accurate exposure assessment are rare. The current risk assessment paradigm dealing mostly with one-by-one chemicals dismisses potential synergisms or interactions from exposures to mixtures of contaminants, particularly at the low-exposure range. Thus, evidence is needed on exposure and health effects of mixtures of contaminants in drinking water [ 4 ].

In a special issue on “Drinking Water Quality and Human Health” IJERPH [ 5 ], 20 papers were recently published on different topics related to drinking water. Eight papers were on microbiological contamination, 11 papers on chemical contamination, and one on radioactivity. Five of the eight papers were on microbiology and the one on radioactivity concerned developing countries, but none on chemical quality. In fact, all the papers on chemical contamination were from industrialized countries, illustrating that microbial quality is still the priority in LMICs. However, chemical pollution from a diversity of sources may also affect these settings and research will be necessary in the future.

Concerning microbiological contamination, one paper deals with the quality of well water in Maryland, USA [ 6 ], and it confirms the frequent contamination by fecal indicators and recommends continuous monitoring of such unregulated water. Another paper did a review of Vibrio pathogens, which are an ongoing concern in rural sub-Saharan Africa [ 7 ]. Two papers focus on the importance of global primary prevention. One investigated the effectiveness of Water Safety Plans (WSP) implemented in 12 countries of the Asia-Pacific region [ 8 ]. The other evaluated the lack of intervention to improve Water, Sanitation and Hygiene (WASH) in Nigerian communities and its effect on the frequency of common childhood diseases (mainly diarrhea) in children [ 9 ]. The efficacies of two types of intervention were also presented. One was a cost-effective household treatment in a village in South Africa [ 10 ], the other a community intervention in mid-western Nepal [ 11 ]. Finally, two epidemiological studies were conducted in industrialized countries. A time-series study evaluated the association between general indicators of drinking water quality (mainly turbidity) and the occurrence of gastroenteritis in 17 urban sites in the USA and Europe. [ 12 ] The other evaluated the performance of an algorithm to predict the occurrence of waterborne disease outbreaks in France [ 13 ].

On the eleven papers on chemical contamination, three focused on the descriptive characteristics of the contamination: one on nitrite seasonality in Finland [ 14 ], the second on geogenic cation (Na, K, Mg, and Ca) stability in Denmark [ 15 ] and the third on historical variation of THM concentrations in french water networks [ 16 ]. Another paper focused on fluoride exposure assessments using biomonitoring data in the Canadian population [ 17 ]. The other papers targeted the health effects associated with drinking water contamination. An extensive up-to-date review was provided regarding the health effects of nitrate [ 18 ]. A more limited review was on heterogeneity in studies on cancer and disinfection by-products [ 19 ]. A thorough epidemiological study on adverse birth outcomes and atrazine exposure in Ohio found a small link with lower birth weight [ 20 ]. Another more geographical study, found a link between some characteristics of drinking water in Taiwan and chronic kidney diseases [ 21 ]. Finally, the other papers discuss the methods of deriving drinking water standards. One focuses on manganese in Quebec, Canada [ 22 ], another on the screening values for pharmaceuticals in drinking water, in Minnesota, USA [ 23 ]. The latter developed the methodology used in Minnesota to derive guidelines—taking the enhanced exposure of young babies to water chemicals into particular consideration [ 24 ]. Finally, the paper on radioactivity presented a description of Polonium 210 water contamination in Malaysia [ 25 ].

In conclusion, despite several constraints (e.g., time schedule, fees, etc.), co-editors were satisfied to gather 20 papers by worldwide teams on such important topics. Our small experience demonstrates the variety and importance of microbiological and chemical contamination of drinking water and their possible health effects.

Acknowledgments

Authors want to acknowledge the important work of the IJERPH staff and of numbers of anonymous reviewers.

Author Contributions

P.L. wrote a first draft of the editorial and approved the final version. C.M.V. did a critical review and added important complementary information to finalize this editorial.

This editorial work received no special funding.

Conflicts of Interest

The authors declare no conflict of interest.