Assessing water quality in North-East Algeria: a comprehensive study using water quality index (WQI) and PCA

Water quality is vital for human health, ecosystems, and well-being due to various reasons. Clean drinking water is crucial for survival and preventing waterborne diseases, especially in vulnerable groups (Kumar et al. 2022). Agriculture relies on water quality to ensure safe and healthy crops, while polluted water can harm aquatic ecosystems (Syafrudin et al. 2021). Impaired water quality also impacts tourism, fisheries, and industries, leading to economic losses (Russ et al. 2022). Treating poor water quality is expensive and requires advanced technologies (Palansooriya et al. 2020). Preserving water quality is essential for environmental conservation and protecting endangered species (Valera et al. 2019). Moreover, water quality affects recreational activities, drinking water scarcity, and sustainable development goals (Chen et al. 2020). To ensure a healthier future, collective efforts are necessary to prevent pollution, promote responsible water use, and implement effective water management practices. Water quality assessment involves various methods tailored to measure specific parameters and contaminants present in the water (Altenburger et al. 2019). These methods include analyzing physical characteristics like temperature, turbidity, color, and odor using handheld meters or observation (Bwadi et al. 2021). Chemical analysis determines concentrations of substances like pH, dissolved oxygen, nutrients, heavy metals, and pesticides through spectrophotometry and colorimetric tests. Biological assessments study organisms like macroinvertebrates and algae to gauge overall aquatic ecosystem health (Kotalik et al. 2021). Microbiological tests detect harmful microorganisms, while bioassays expose living organisms to assess water toxicity (do Nascimento et al. 2023). Remote sensing uses satellite imagery for larger water bodies, and real-time monitoring with automated sensors provides immediate data on water quality (Zhuang et al. 2022). Biosensors, isotope analysis, and traditional sampling with laboratory analysis also contribute to comprehensive water quality understanding, facilitating better water resource management and protection decisions (Bieroza et al. 2023).

A water quality index (WQI) is a numerical rating used to assess and communicate the overall quality of water in a specific area. It considers multiple water quality parameters, such as physical, chemical, and biological characteristics, to calculate a single value representing water quality. The index typically ranges from 0 to 100, with higher values indicating better water quality (Uddin et al. 2021). It simplifies complex data for easy understanding by the public and decision-makers, enables comparisons over time and across locations, communicates potential risks to human health and the environment, and aids policymakers in prioritizing actions to improve water quality. However, the WQI's simplicity may not capture all water quality nuances, and it should be used in conjunction with comprehensive monitoring and analysis while considering local conditions and assessment objectives (Giri 2021).

Scientific research on the WQI is extensive and widespread (Wong et al. 2020). Researchers and environmental agencies worldwide have worked to develop, refine, and validate various WQI models (Uddin et al. 2021). They have explored its effectiveness in assessing water quality and its applicability in different regions and water bodies. Studies have focused on validating and calibrating WQI models, comparing multiple approaches, and evaluating its role in decision-making and resource allocation (Akdogan & Güven 2023). Long-term studies using WQI have helped identify water quality trends over time and understand the impacts of environmental policies and climate change (Vatanpour et al. 2020). Researchers have also investigated the connection between WQI values and ecosystem health and human health risks. As scientific knowledge advances, the WQI continues to be a valuable tool for water quality management and is expected to gain more significance in addressing environmental challenges and promoting sustainable water resource management.

In the last two decades, numerous studies have demonstrated the deterioration of groundwater quality in various countries, as highlighted by researchers such as Jeong (2001), Moon et al. (2004), Adimalla (2019), and Gaikwad et al. (2020a, 2020b); Asmamaw & Debie (2023). Developing nations encounter challenges in accessing safe drinking water, contributing to multiple public health issues, as noted by Abedin et al. (2019). Additionally, unattended agricultural activities and domestic effluents exacerbate the degradation of water quality, as emphasized by Khan et al. (2021). Effective water quality management is vital for comprehensive integrated water resource management (Xiang et al. 2021). Unfortunately, declining rainfall, irregular seasonal patterns, desertification, and environmental degradation worsen the water problem (Singh et al. 2021). Therefore, regular and continuous monitoring of water quality is urgently required to address and preserve it (Yan et al. 2022). Contaminated water poses serious threats to public health and ecosystems, causing waterborne diseases and environmental degradation (Panaskar et al. 2016; Mukate et al. 2018; Wagh et al. 2018; Nabi et al. 2019). As environmental pollutants continue to emerge, improving water quality has become a critical concern for water resource management experts to protect human health and aquatic life (Ngqwala & Muchesa 2020).

Algeria, as many other countries, has not been spared from issues related to surface and groundwater, as well as their quality. Several studies have been conducted in arid or semi-arid regions to address these water challenges as highlighted by Kouadri et al. (2021) and Bouderbala (2021).

In North-East of Algeria, especially Annaba Province, which is currently facing significant challenges due to water shortages and intermittent water distribution (Berredjem et al. 2023). The water demand was estimated at 98 million cubic metres (MCM) by the year 2021, with the potential to reach 148 MCM by the year 2070 (Berredjem et al. 2023). In the face of this water shortage, some researchers have directed their focus toward investigating the deterioration of water quality as highlighted by Attoui et al. (2016) and Hafsi & Boutaghane (2022). The research aimed to assess water quality in North-East Algeria using a temporal monitoring system extends from January to December 2021 for physico-chemical parameters in raw water obtained from the Cheffia Dam, Oued El Aneb, and Treat boreholes, where the water of these three sources is not treated, as the WQI and PCA are being used in order to ensure that the drinking water supplied to the vast majority of the population through the distribution network met the necessary potability criteria to protect public health, or the transition to water treatment when the WQI indicates poor water quality to unsuitable for consumption, as well as assisting in identifying appropriate treatment systems.

One of Annaba's prominent features is its dense hydrographic network, highlighted by Lake Fetzara, an expansive freshwater body spanning 18,670 ha, and the Seybouse River, stretching over 127.5 km².

The available water resources in the city of Annaba are estimated at approximately 92.6 MCM/year. These resources are distributed among various sources, including surface waters from the Cheffia Dam (44 MCM/year) and the Mexa Dam (21 MCM/year), as well as hillside reservoirs (7 MCM/year). Groundwater also plays a significant role, sourced from wells, springs (0.6 MCM/year), and boreholes (20 MCM/year).

In total, 36 samples across the study area were collected. The experimental setup for studying the water sample contents involved the use of 1.5-L mineral water bottles (Figure 2(d)). These bottles were thoroughly rinsed with distilled water to eliminate any potential residues. Each bottle was meticulously labeled to ensure precise traceability of the collected samples. Water samples were collected using sterile gloves, ensuring the preservation of sample integrity and preventing any external contamination. After collection, the samples were stored in a cooler at a constant temperature of 4 °C, following the procedures recommended by Rodier et al. (1996). This storage method ensures the stability of the chemical properties of the samples until their subsequent analysis. Sampling was carried out systematically, with a monthly frequency throughout the year 2021, specifically in the morning around 10 am, thus covering all seasons and weather conditions of the year. This methodical approach ensures the representativity of the collected samples, providing a solid foundation for the in-depth analysis of water content over time.

Water analysis is a critical aspect of environmental monitoring and quality control (QC), involving the determination of various chemical and physical parameters. The laboratory analyses encompassed 36 samples for 16 parameters, totaling 576 analyses. Three main analytical methods, namely electrochemical, volumetric, and spectrophotometric, are employed for comprehensive water analysis.

The electrochemical method relies on redox reactions at an electrode in a solution, utilizing voltammetry for measuring electrical parameters and analyzing chemical compounds. This method is implemented through a multiparameter instrument like the HACH HQ4100 for parameters such as pH, temperature, conductivity, salinity, and turbidity.

The volumetric method, as described by Genin et al. (2007), involves precise measurement of reactive solution volumes to determine substance concentration. In water analysis, it is applied using the Paillaise spectrophotometer or HACH DR3900 for measuring calcium and magnesium concentrations, as well as total hardness (TH).

Spectrophotometric analysis, based on the absorption or transmission of light by substances in solution, plays a crucial role in determining chemical concentrations. For potassium and sodium analysis, a photometric method is employed, measuring light absorption by these ions using a photometer. The variation in absorption is directly proportional to the concentrations of potassium and sodium in the solution. Moreover, Table 2 provides information on various water parameters, their respective units, and the corresponding analysis methods and instruments used.

Calibrating and adjusting measurement instruments play a crucial role in ensuring the reliability and accuracy of analytical data in laboratory settings. In this study, we conducted a calibration and adjustment process at the laboratory level for four instruments from the HACH range: the HACH HQ4100 Multiparameter, HACH 54650 Conductometer, HACH 2100 Turbidimeter, and HACH DR3900 Spectrophotometer. For each of these instruments, we utilized standard calibration solutions of known concentrations to fine-tune the measured values according to the manufacturer's instructions. Emphasizing preventive maintenance of components such as electrodes and measurement cells is essential to ensure accurate results. Meticulously documenting all performed calibrations is crucial for maintaining measurement traceability and ensuring compliance with established quality standards and laboratory best practices. This process contributes to enhancing the credibility of analyses conducted in scientific environments.

The term ‘Water Quality Index (WQI)’ refers to a scoring method aimed at assessing the overall impact of various water quality parameters on the suitability of water for human consumption (Mitra et al. 2006). The process of calculating the WQI involves three distinct steps:

In the initial stage, the nine parameters [pH, electrical conductivity (CE), calcium (Ca), magnesium (Mg), sodium (Na), potassium (K), chloride (Cl), sulfate (SO₄), and nitrate (NO₃)] serve as key indicators of water quality. These parameters primarily correspond to the different minerals present in water. It is important to note that water with elevated levels of solids might potentially lead to constipation or exert laxative effects (Hajji et al. 2018). Recognizing the pivotal role of parameters such as chlorides, sulfates, and nitrates in water quality assessment, the highest weight of 5 has been attributed to them (Vasanthavigar et al. 2010; Srinivasamoorthy et al. 2014). Meanwhile, other parameters like potassium, sodium, magnesium, and calcium have been assigned weightings ranging from 2 to 4, based on their respective significance in determining water quality. Refer to Table 3 for detailed weight assignments.

The sub-index for each parameter is SI_i, and the number of parameters is n. The calculated WQI has been divided into five categories shown in Table 4.

Another commonly used approach involves statistical methods such as PCA. It is a statistical method utilized for dimensional reduction of data while retaining the highest quantity and quality of the input data. It aggregates the information being dispersed in many dimensions into a smaller dimension which are independent of each other. PCA is a commonly used method for selecting independent variables and discarding highly correlated or redundant variables. The PCA involves five main steps as follows:

• 1.
  The original data matrix is listed as given by:

  

• 2.
  The originally measured data are standardized to mitigate the impact of dimension with z score standardization:

  

• 3.
  Correlation coefficient matrix is determined based on the following equation:

  
• 4.
  The eigenvalues and eigenvectors are then calculated as follows:

  

This analysis condenses the original variables into a more compact set of novel variables, leading to a minor loss of information from the original parameters.

The new personal computers undergo a process known as varimax rotation, resulting in the creation of varimax factors. This technique aims to reduce the dimensionality of the original data and identify key variables for straightforward interpretation (Ul-Saufie et al. 2010). By examining component loadings, significant variables are categorized, enabling source identification. Notably, a factor loading of 0.75 is indicative of a robust and significant factor, while loadings falling within the range of 0.75–0.5 range are considered moderate, and those within 0.5–0.3 are deemed weak.

The significance of each component is determined by eigenvalues, indicating the extent of variance explained by individual components within the sample. Components with eigenvalues of at least 1 are retained to interpret the variation in surface water. In cases where multiple variables are chosen for a PC, a multivariate correlation analysis is employed to decide which variable to include. Highly correlated variables are regarded as duplicative, and only the variable with the highest loading is retained. Conversely, if highly loaded variables display no correlation, each one is deemed important and consequently incorporated.

The samples were collected from three different sites (Cheffia Dam, Oued El Aneb borehole, and Treat borehole) during 12 months of 2021. The maximum/minimum and analytic results for each parameter are summarized in Table 5. The values are compared against recommended standards such as Algerian potability standards (AS) and WHO guidelines.

The results of the monthly variation of the WQI during the year 2021 for the three study sites reveal a range of fluctuations. For the dam, the variation spans from 37.3 to 86.3, while for the two wells, it extends from 45.5 to 91.5. These values unequivocally demonstrate that the waters in the study area belong to both the categories of good water and excellent water.

The utilization of principal component analysis (PCA) in this study offers a comprehensive understanding of the intricate relationships among physico-chemical variables in the investigated study sites (Meybeck et al. 1996). The PCA employed to analyze 16 measured parameters, allowing for the exploration of both the interrelationships among variables and the spatial distribution of the study sites.

Figure 5 depict the correlation circle of parameters, illustrating the use of the horizontal axis F1 (First component) and the vertical axis F2 (second component), accompanied by theprojection of samples onto the plane defined by the first two factorial axes, namely F1 and F2.

Physico-chemical analysis of three distinct water sources, namely Cheffia Dam (S1), Oued El Aneb Well (S2), and Treat Well (S3), provides valuable insights into their characteristics and potential health implications. The study covers various parameters, and the main findings are interpreted and explained as follows.

Firstly, pH levels reveal a range from 7.17 to 8.51, indicating slightly alkaline conditions. Importantly, these values fall within the recommended range for drinking water (6.5–9.5) and comply with both Algerian drinking water standards and WHO standards. The pH range of 7.2–7.8, considered ideal for maintaining good health according to previous research (WHO 2011; Saleem et al. 2022), is within the observed values. Overall, pH analysis suggests that the water sources are suitable for consumption, in accordance with established standards and guidelines.

Regarding water temperature, recorded values range from 16.3 to 21.3 °C for the three sources. These temperatures fall within a reasonable range and do not exceed 25 °C, which is a positive sign for mitigating the risk of pollution. The conclusions align with previous research findings by Nguefack et al. (2018) highlighting the importance of temperature in assessing water source vulnerability. The observed temperature range supports the overall health of the water sources.

Conductivity and salinity values show considerable variation, indicating significant mineralization. Some measurements exceed Algerian drinking water standards, raising concerns about water suitability for consumption. However, salinity values fall within a range that, with proper management, may not pose significant issues. It is emphasized that considering the local context and balancing mineral content against health guidelines are crucial. Close monitoring and appropriate measures are needed to ensure water remains safe to consume.

Turbidity values range from 0.33 to 54 Nephelometric Turbidity Unit (NTU), with Cheffia Dam water exceeding drinking water standards and WHO recommendations. High turbidity suggests potential issues with sediment and particles, requiring enhanced treatment or protective measures. This observation highlights the need for targeted interventions to address specific challenges associated with Cheffia Dam water.

Nitrate concentrations and other chemical parameters are below recommended guidelines, including nitrite, sodium, sulfate, phosphate, and TH. However, potassium levels exceed recommendations, posing a potential problem. Additionally, calcium content in Cheffia Dam water exceeds AS recommendations, while magnesium content varies but remains within acceptable limits. These results underscore the importance of monitoring specific chemical parameters to ensure comprehensive water quality.

Finally, chloride concentrations vary, with Oued El Aneb Well having the highest level. Iron content is within acceptable limits for S2 and S3 but exceeds AS standards in Cheffia Dam water. It is essential to address elevated chloride and iron levels in Cheffia Dam water to comply with established standards.

Overall, the comprehensive analysis provides a nuanced understanding of water quality in these sources, highlighting both strengths and areas requiring attention to maintain safe drinking water.

Firstly, the data highlight notable temporal variations in water quality across locations over the months. For example, Cheffia Dam (S1) shows substantial fluctuations in WQI values, ranging from a minimum of 37.29 in February to a maximum of 86.30 in January. This variation signifies a dynamic change in water quality over a year.

Secondly, there is a discernible spatial disparity among the locations. Oued El Aneb Well (S2) consistently shows lower WQI values compared to other locations, particularly evident in February and March. This disparity suggests potential differences in environmental factors or pollution sources affecting water quality at these sites.

Additionally, the data suggest the presence of seasonal trends influencing water quality. Generally, there is a trend where winter months (January, February) exhibit higher WQI values compared to summer months (July, August). This observation could be attributed to seasonal variations in precipitation, temperature, or changes in human activities impacting water quality.

Overall, the majority of months indicate ‘Good’ water quality, suggesting that water is generally acceptable for various uses. However, there are cases of ‘Excellent’ water quality, indicating even better conditions in terms of natural conditions, climate, and human activities, while a few months display lower WQI values, indicating potential concerns.

Cheffia Dam consistently maintains higher WQI values compared to other locations, indicating relatively better water quality. Similarly, Treat Well (S3) generally maintains good water quality, with some months displaying an ‘Excellent’ classification.

In contrast, Oued El Aneb Well (S2) exhibits more fluctuations in WQI values, potentially indicating greater vulnerability or sensitivity to environmental changes or pollution sources.

Comparing WQI results with national and international standards, it can be noted that although some parameters, such as conductivity, exceed recommended limits, the majority of water quality parameters at the three sites meet Algerian and WHO standards, indicating good to excellent water quality. This is consistent with WQI values and water quality throughout. However, it is crucial to continue monitoring and addressing any parameters that may deviate from standards over time.

The figure provides a comprehensive overview of water quality dynamics across different locations and months. Interpretation of these results could guide further research and monitoring efforts to understand underlying factors influencing these variations. It is crucial to address these factors to maintain or improve water quality in the studied areas, potentially requiring targeted interventions or environmental management strategies.

The findings, summarized through projection onto a two-dimensional plane capturing 46.18% of the total variance, reveal two notable associations. Axis 1 and Axis 2 contribute 30.13 and 16.05% of the total variance, respectively. The first association, characterized by negative correlations, includes parameters such as conductivity, salinity, Mg, Na, NO₂, temperature (T°C), and TH, explaining 30.6% of the total variance. The second association, marked by positive correlations, encompasses pH, turbidity, K, Ca, NO₃, PO₄, Iron, and SO₄, representing 27.14% of the total variance.

The PCA investigation further unveils significant interrelationships, highlighting negative connections between conductivity and salinity, as well as between iron and sulfate. Importantly, each study site exhibits distinctive characteristics, leading to the identification of unique site/variable patterns. Notably, fluctuations in the physical water parameters of specific sites, such as Cheffia Dam, Oued El Aneb, and Treat boreholes, are intricately linked to chemical properties and influenced by spatio-temporal variations.

The analysis of Plans 1–2 concerning individuals reveals three distinct groups. The first group encompasses individuals S1.1–S1.12, located in the positive domain of the factorial axis, representing Cheffia Dam analyses. These values correspond to surface waters and exhibit lower mineralization compared to sites 2 and 3, which are situated in the negative domain. The second group comprises individuals (S2.1, S2.2, S2.3, S3.1, S3.2, S3.3); these individuals represent the first 3 months of both drillings and are characterized by low nutrient levels for both drills. The third group, consisting of the remaining individuals, demonstrates a progressive increase in all parameters.

The application of PCA in this study provides valuable insights into the complex relationships and spatial differentiation within the study sites. The positive and negative associations revealed by the analysis offer a nuanced understanding of how various physico-chemical parameters interact, contributing to the overall variability observed in the water characteristics of the studied locations. This study not only contributes to the understanding of local water quality but also suggests potential avenues for future research, such as exploring the environmental factors influencing the observed patterns and their implications for ecosystem health and management.

Although chemical analyses and WQI are crucial tools for identifying water quality, it is imperative to recognize and understand some limitations in order to achieve a more comprehensive and precise assessment of water quality in the future. These limitations include challenges associated with using a single index, which may simplify the complexity of water quality by overlooking specific nuances for each parameter and the potential subjectivity in establishing weights for different parameters in WQI calculation. The temporal and spatial variability of water conditions can lead to fluctuating and unrepresentative results of the actual water quality. Additionally, the analytical limitations of chemical methods may result in estimation errors, especially in cases of very low or high concentrations. The evolution of water quality standards, the influence of climatic factors, and the neglect of biological aspects are dimensions that require particular attention for a more holistic assessment of water quality in the future.

This comprehensive study underscores the critical importance of maintaining high water quality for the well-being of Annaba's population and the sustainability of its ecosystems. Through an extensive evaluation of water quality from January to December 2021, focusing on the Cheffia dam, Oued El Aneb, and Treat boreholes, valuable insights have been gained using the WQI and various physico-chemical parameters.

The findings indicate excellent water quality in autumn and generally good water quality throughout the year, therefore suitable for consumption, with distinct characteristics observed for each site through principal component analysis PCA. The analysis of water parameters indicates an adherence to standards; however, challenges such as elevated turbidity in the Cheffia dam and fluctuations in certain chemical parameters necessitate targeted interventions for maintaining safe drinking water.

While this study provides valuable insights, it is crucial to acknowledge limitations, such as the use of a single index and potential subjectivity in assigning weights to parameters.

It is essential for practitioners to strengthen monitoring efforts and implement adaptive management strategies to address potential concerns, considering evolving standards, climatic factors, and biological aspects. Administrators should develop adaptive public policies that target specific interventions for each water source, addressing challenges related to turbidity, mineralization, and chemical compounds.

Engineers can contribute by conducting detailed studies on turbidity sources in the Cheffia Dam to guide treatment strategies and exploring innovative approaches to reduce mineralization and address specific chemical parameters using advanced technologies. Longitudinal studies on the effectiveness of management measures implemented following these findings will provide crucial insights for ongoing and sustainable improvement of water quality in Annaba.

This research sets a significant milestone for future water resource management decisions in Annaba, offering valuable insights into temporal and spatial variations and guiding sustainable water management practices. Future research should focus on in-depth seasonal variations, specific turbidity sources, adaptive water resource management, and innovative technologies, aiming for the long-term enhancement of water quality and the preservation of ecosystems.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.