1 Introduction

Water is the vital natural resource with social and economic values for human beings (Kumar 2018). Without water, existence of man would be threatened (Zhang 2017). The most important drinking sources in the world are surface water and groundwater (Paun et al. 2016).

Currently, more than 1.1 billion people do not have access to clean drinking water and it is estimated that nearly two-thirds of all nations will experience water stress by the year 2025 (Kumar 2018).

With the extensive social and economic growth, such as human factors, climate and hydrology may lead to accumulation of pollutants in the surface water that may result in gradual change of the water source quality (Shan 2011).

The optimal quantity and acceptable quality of water is one of the essential needs to survive as mentioned earlier, but the maintenance of an acceptable quality of water is a challenge in the sector of water resources management (Mukate et al. 2019). Accordingly, the water quality of water bodies can be tested through changes in physical, chemical and biological characteristics related to anthropogenic or natural phenomena (Britto et al. 2018).

Therefore, water quality of any specific water body can be tested using physical, chemical and biological parameters also called variables, by collecting samples and obtaining data at specific locations (Britto et al. 2018; Tyagi et al. 2013).

To that end, the suitability of water sources for human consumption has been described in terms of Water Quality Index (WQI), which is one of the most effective ways to describe the quality of water, by reducing the bulk of information into a single value ranging between 0 and 100 (Tyagi et al. 2013).

Hence, the objective of the study is to review the WQI concept by listing some of the important water quality indices used worldwide for water quality assessment, listing the advantages and disadvantages of the selected indices and finally reviewing some water quality studies worldwide.

2 Water quality index

2.1 History of water quality concept

In the last decade of the twentieth century, many organizations involved in water control, used the water quality indices for water quality assessment (Paun et al. 2016). In the 1960’s, the water quality indices was introduced to assess the water quality in rivers (Hamlat et al. 2017).

Horton (1965), initially developed a system for rating water quality through index numbers, offering a tool for water pollution abatement, since the terms “water quality” and “pollution” are related. The first step to develop an index is to select a list of 10 variables for the index’s construction, which are: sewage treatment, dissolved oxygen (DO), pH, coliforms, electroconductivity (EC), carbon chloroform extract (CCE), alkalinity, chloride, temperature and obvious pollution. The next step is to assign a scale value between zero and 100 for each variable depending on the quality or concentration. The last step, is to designate to each variable is a relative weighting factor to show their importance and influence on the quality index (the higher the assigned weight, the more impact it has on the water quality index, consequently it is more important) (Horton 1965).

Later on, Brown et al. (1970) established a new water quality index (WQI) with nine variables: DO, coliforms, pH, temperature, biochemical oxygen demand (BOD), total phosphate, nitrate concentrations, turbidity and solid content based on a basic arithmetic weighting using arithmetic mean to calculate the rating of each variable. These rates are then converted not temporary weights. Finally, each temporary weight is divided by the sum of all the temporary weights in order to get the final weight of each variable (Kachroud et al. 2019a; Shah and Joshi 2017). In 1973, Brown et al., considered that a geometric aggregation (a way to aggregate variables, and being more sensitive when a variable exceeds the norm) is better than an arithmetic one. The National Sanitation Foundation (NSF) supported this effort (Kachroud et al. 2019a; Shah and Joshi 2017).

Steinhart et al. (1982) developed a novel environmental quality index (EQI) for the Great Lakes ecosystem in North America. Nine variables were selected for this index: biological, physical, chemical and toxic. These variables were: specific conductance or electroconductivity, chloride, total phosphorus, fecal Coliforms, chlorophyll a, suspended solids, obvious pollution (aesthetic state), toxic inorganic contaminants, and toxic organic contaminants. Raw data were converted to subindex and each subindex was multiplied by a weighting factor (a value of 0.1 for chemical, physical and biological factors but 0.15 for toxic substances). The final score ranged between 0 (poor quality) and 100 (best quality) (Lumb et al. 2011a; Tirkey et al. 2015).

Dinius (1987), developed a WQI based on multiplicative aggregation having a scale expressed with values as percentage, where 100% expressed a perfect water quality (Shah and Joshi 2017).

In the mid 90’s, a new WQI was introduced to Canada by the province of British Columbia, and used as an increasing index to evaluate water quality (Lumb et al. 2011b; Shah and Joshi 2017). A while after, the Water Quality Guidelines Task Group of the Canadian Council of Ministers of the Environment (CCME) modified the original British Columbia Water Quality Index (BCWQI) and endorsed it as the CCME WQI in 2001(Bharti and Katyal 2011; Lumb et al. 2011b).

In 1996, the Watershed Enhancement Program (WEPWQI) was established in Dayton Ohio, including water quality variables, flow measurements and water clarity or turbidity. Taking into consideration pesticide and Polycyclic Aromatic Hydrocarbon (PAH) contamination, is what distinguished this index from the NSFWQI (Kachroud et al. 2019a, b).

Liou et al. (2003) established a WQI in Taiwan on the Keya River. The index employed thirteen variables: Fecal coliforms, DO, ammonia nitrogen, BOD, suspended solids, turbidity, temperature, pH, toxicity, cadmium (Cd), lead (Pb), copper (Cu) and zinc (Zn). These variables were downsized to nine based on environmental and health significance: Fecal coliforms, DO, ammonia nitrogen, BOD, suspended solids, turbidity, temperature, pH and toxicity. Each variable was converted into an actual value ranging on a scale from 0 to 100 (worst to highest). This index is based on the geometric means (an aggregation function that could eliminate the ambiguous caused from smaller weightings) of the standardized values (Akhtar et al. 2021; Liou et al. 2004; Uddin et al. 2021).

Said et al. (2004) implemented a new WQI using the logarithmic aggregation applied in streams waterbodies in Florida (USA), based on only 5 variables: DO, total phosphate, turbidity, fecal coliforms and specific conductance. The main idea was to decrease the number of variables and change the aggregation method using the logarithmic aggregation (this function does not require any sub-indices and any standardization of the variables). This index ranged from 0 to 3, the latter being the ideal value (Akhtar et al. 2021; Kachroud et al. 2019a, b; Said et al. 2004; Uddin et al. 2021).

The Malaysian WQI (MWQI) was carried out in 2007, including six variables: DO, BOD, Chemical Oxygen Demand (COD), Ammonia Nitrogen, suspended solids and pH. For each variable, a curve was established to transform the actual value of the variable into a non-dimensional sub-index value.

The next step is to determine the weighting of the variables by considering the experts panel opinions. The final score is determined using the additive aggregation formula (where sub-indices values and their weightings are summed), extending from 0 (polluted) to 100 (clean) (Uddin et al. 2021).

The Hanh and Almeida indices were established respectively in 2010 on surface water in Vietnam and 2012 on the Potrero de los Funes in Argentina, based on 8 (color, suspended solids, DO, BOD, COD, chloride, total coliforms and orthophosphate) and 10 (color, pH, COD, fecal coliforms, total coliforms, total phosphate, nitrates, detergent, enterococci and Escherichia coli.) water quality variables. Both indices were based on rating curve- based sum-indexing system (Uddin et al. 2021).

The most recent developed WQI model in the literature was carried out in 2017. This index tried to reduce uncertainty present in other water quality indices. The West Java Water Quality Index (WJWQI) applied in the Java Sea in Indonesia was based on thirteen crucial water quality variables: temperature, suspended solids, COD, DO, nitrite, total phosphate, detergent, phenol, chloride, Zn, Pb, mercury (Hg) and fecal coliforms. Using two screening steps (based on statistical assessment), parameter (variable) redundancy was determined to only 9: temperature, suspended solids, COD, DO, nitrite, total phosphate, detergent, phenol and chloride. Sub-indices were obtained for those nine variables and weights were allocated based on expert opinions, using the same multiplicative aggregation as the NSFWQI. The WJWQI suggested 5 quality classes ranging from poor (5–25) to excellent (90–100) (Uddin et al. 2021).

2.2 Phases of WQI development

Mainly, WQI concept is based on many factors as displayed in Fig. 1 and described in the following steps:

Fig. 1
figure 1

Phases of WQI development

Full size image
  1. 1.

    Parameter selection for measurement of water quality (Shah and Joshi 2017):


    The selection is carried out based on the management objectives and the environmental characteristics of the research area (Yan et al. 2015). Many variables are recommended, since they have a considerable impact on water quality and derive from 5 classes namely, oxygen level, eutrophication, health aspects, physical characteristics and dissolved substances (Tyagi et al. 2013).

  2. 2.

    Transformation of the raw data parameter into a common scale (Paun et al. 2016):


    Different statistical approach can be used for transformation, all parameters are transformed from raw data that have different dimensions and units (ppm, saturation, percentage etc.) into a common scale, a non-dimensional scale and sub-indices are generated (Poonam et al. 2013; Tirkey et al. 2015).

  3. 3.

    Providing weights to the parameters (Tripathi and Singal 2019):


    Weights are assigned to each parameter according to their importance and their impact on water quality, expert opinion is needed to assign weights (Tirkey et al. 2015). Weightage depends on the permissible limits assigned by International and National agencies in water drinking (Shah and Joshi 2017).

  4. 4.

    Aggregation of sub-index values to obtain the final WQI:


    WQI is the sum of rating and weightage of all the parameters (Tripathi and Singal 2019).

It is important to note that in some indices, statistical approaches are commonly used such as factor analysis (FA), principal component analysis (PCA), discriminant analysis (DA) and cluster analysis (CA). Using these statistical approaches improves accuracy of the index and reduce subjective assumptions (Tirkey et al. 2015).

2.3 Evolution of WQI research

2.3.1 Per year

According to Scopus (2022), the yearly evolution of WQI's research is illustrated in Fig. 2 (from 1978 till 2022).

Fig. 2
figure 2

Evolution of WQI research per year (Scopus 2022)

Full size image

Overall, it is clear that the number of research has grown over time, especially in the most recent years. The number of studies remained shy between 1975 and 1988 (ranging from 1 to 13 research). In 1998, the number improved to 46 studies and increased gradually to 466 publications in 2011.The WQI's studies have grown significantly over the past decade, demonstrating that the WQI has become a significant research topic with the goal of reaching its maximum in 2022 (1316 studies) (Scopus, 2022).

2.3.2 Per country

In Fig. 3, the development of WQI research is depicted visually per country from 1975 to 2022.

Fig. 3
figure 3

Evolution of WQI research per country (Scopus 2022)

Full size image

According to Scopus (2022), the top three countries were China, India and the United States, with 2356, 1678 and 1241 studies, respectively. Iran, Brazil, and Italy occupy the fourth, fifth, and sixth spots, respectively (409, 375 and 336 study). Malaysia and Spain have approximately the same number of studies, respectively 321 and 320 study. The studies in the remaining countries decrease gradually from 303 document in Spain to 210 documents in Turkey. This demonstrates that developing nations, like India, place a high value on the development of water quality protection even though they lack strong economic power, cutting-edge technology, and a top-notch scientific research team. This is because water quality is crucial to the long-term social and economic development of those nations (Zhang 2019).

2.4 Different methods for WQI determination

Water quality indices are tools to determine water quality. Those indices demand basic concepts and knowledge about water issues (Singh et al. 2013). There are many water quality indices such as the: National Sanitation Foundation Water Quality Index (NSFWQI), Canadian Council of Ministers of Environment Water Quality Index (CCMEWQI), Oregon Water Quality Index (OWQI), and Weight Arithmetic Water Quality Index (WAWQI) (Paun et al. 2016).

These water quality indices are applied in particular areas, based on many parameters compared to specific regional standards. Moreover, they are used to illustrate annual cycles, spatio-temporal variations and trends in water quality (Paun et al. 2016). That is to say that, these indices reflect the rank of water quality in lakes, streams, rivers, and reservoirs (Kizar 2018).

Accordingly, in this section a general review of available worldwide used indices is presented.

2.4.1 National sanitation foundation (NSFWQI)

The NSFWQI was developed in 1970 by the National Sanitation Foundation (NSF) of the United States (Hamlat et al. 2017; Samadi et al. 2015). This WQI has been widely field tested and is used to calculate and evaluate the WQI of many water bodies (Hamlat et al. 2017). However, this index belongs to the public indices group. It represents a general water quality and does not take into account the water’s use capacities, furthermore, it ignores all types of water consumption in the evaluation process (Bharti and Katyal 2011; Ewaid 2017).

The NSFWQI has been widely applied and accepted in Asian, African and European countries (Singh et al. 2013), and is based on the analysis of nine variables or parameters, such as, BOD, DO, Nitrate (NO3), Total Phosphate (PO4), Temperature, Turbidity, Total Solids(TS), pH, and Fecal Coliforms (FC).

Some of the index parameters have different importance, therefore, a weighted mean for each parameter is assigned, based on expert opinion which have grounded their opinions on the environmental significance, the recommended principles and uses of water body and the sum of these weights is equal to 1 (Table 1) (Ewaid 2017; Uddin et al. 2021).

Table 1 Weight scores of the nine NSF-WQI parameters
Full size table

Due to environmental issues, the NSFWQI has changed overtime. The TS parameter was substituted by the Total Dissolved Solids (TDS) or Total Suspended Solids (TSS), the Total Phosphate by orthophosphate, and the FC by E. coli (Oliveira et al. 2019).

The mathematical expression of the NSFWQI is given by the following Eq. (1) (Tyagi et al. 2013):

$${\text{NSFWQI}} = \mathop \sum \limits_{{{\text{i}} = 1}}^{{\text{n}}} {\text{QiWi}}$$
(1)

where, Qi is the sub-index for ith water quality parameter. Wi is the weight associated with ith water quality parameter.n is the number of water quality parameters.

This method ranges from 0 to 100, where 100 represents perfect water quality conditions, while zero indicates water that is not suitable for the use and needs further treatment (Samadi et al. 2015).

The ratings are defined in the following Table 2.

Table 2 Colors and definition used in the classification of pollution using NSFWQI (Roozbahani and Boldaji 2013)
Full size table

In 1972, the Dinius index (DWQI) happened to be the second modified version of the NSF (USA). Expended in 1987 using the Delphi method, the DWQI included twelve parameters (with their assigned weights): Temperature (0.077), color (0.063), pH (0.077), DO (0.109), BOD (0.097), EC (0.079), alkalinity (0.063), chloride (0.074), coliform count (0.090), E. coli (0.116). total hardness (0.065) and nitrate (0.090). Without any conversion process, the DWQI used the measured variable concentrations directly as the sub-index values (Kachroud et al. 2019b; Uddin et al. 2021).

Sukmawati and Rusni assessed in 2018 the water quality in Beratan lake (Bali), choosing five representative stations for water sampling representing each side of the lake, using the NSFWQI. NSFWQI’s nine parameters mentioned above were measured in each station. The findings indicated that the NSFWQI for the Beratan lake was seventy-eight suggesting a good water quality. Despite this, both pH and FC were below the required score (Sukmawati and Rusni 2019).

The NSFWQI indicated a good water quality while having an inadequate value for fecal coliforms and pH. For that reason, WQIs must be adapted and developed so that any minor change in the value of any parameter affects the total value of the water quality index.

A study conducted by Zhan et al. (2021), concerning the monitoring of water quality and examining WQI trends of raw water in Macao (China) was established from 2002 to 2019 adopting the NSFWQI. NSFWQI's initial model included nine parameters (DO, FC, pH, BOD, temperature, total phosphates, and nitrates), each parameter was given a weight and the parameters used had a significant impact on the WQI calculation outcomes. Two sets of possible parameters were investigated in this study in order to determine the impact of various parameters. The first option was to keep the original 9-parameter model, however, in the second scenario, up to twenty-one parameters were chosen, selected by Principal Component Analysis (PCA).

The latter statistical method was used to learn more about the primary elements that contributed to water quality variations, and to calculate the impact of each attribute on the quality of raw water. Based on the PCA results, the 21-parameter model was chosen. The results showed that the quality of raw water in Macao has been relatively stable in the period of interest and appeared an upward trend overall. Furthermore, the outcome of environmental elements, such as natural events, the region's hydrology and meteorology, can have a significant impact on water quality. On the other hand, Macao's raw water quality met China's Class III water quality requirements and the raw water pollution was relatively low. Consequently, human activities didn’t have a significant impact on water quality due to effective treatment and protection measures (Zhan et al. 2021).

Tampo et al. (2022) undertook a recent study in Adjougba (Togo), in the valley of Zio River. Water samples were collected from the surface water (SW), ground water (GW) and treated wastewater (TWW), intending to compare the water quality of these resources for irrigation and domestic use.

Hence, WQIs, water suitability indicators for irrigation purposes (WSI-IPs) and raw water quality parameters were compared using statistical analysis (factor analysis and Spearman’s correlation).

Moreover, the results proposed that he water resources are suitable for irrigation and domestic use: TWW suitable for irrigation use, GW suitable for domestic use and SW suitable for irrigation use.

The NSFWQI and overall index of pollution (OPI) parameters were tested, and the results demonstrated that the sodium absorption ratio, EC, residual sodium carbonate, Chloride and FC are the most effective parameters for determining if water is suitable for irrigation.

On the other hand, EC, DO, pH, turbidity, COD, hardness, FC, nitrates, national sanitation foundation's water quality index (NSFWQI), and overall index of pollution (OPI) are the most reliable in the detection of water suitability for domestic use (Tampo et al. 2022).

Following these studies, it is worth examining the NSFWQI. This index can be used with other WQI models in studies on rivers, lakes etc., since one index can show different results than another index, in view of the fact that some indices might be affected by other variations such as seasonal variation.

Additionally, the NSFWQI should be developed and adapted to each river, so that any change in any value will affect the entire water quality. It is unhelpful to have a good water quality yet a low score of a parameter that can affect human health (case of FC).

2.4.2 Canadian council of ministers of the environment water quality index (CCMEWQI)

The Canadian Water Quality Index adopted the conceptual model of the British Colombia Water Quality Index (BCWQI), based on relative sub-indices (Kizar 2018).

The CCMEWQI provides a water quality assessment for the suitability of water bodies, to support aquatic life in specific monitoring sites in Canada (Paun et al. 2016). In addition, this index gives information about the water quality for both management and the public. It can furthermore be applied in many water agencies in various countries with slight modification (Tyagi et al. 2013).

The CCMEWQI method simplifies the complex and technical data. It tests the multi-variable water quality data and compares the data to benchmarks determined by the user (Tirkey et al. 2015). The sampling protocol requires at least four parameters sampled at least four times but does not indicate which ones should be used; the user must decide ( Uddin et al. 2021). Yet, the parameters may vary from one station to another (Tyagi et al. 2013).

After the water body, the objective and the period of time have been defined the three factors of the CWQI are calculated (Baghapour et al. 2013; Canadian Council of Ministers of the Environment 1999):

  1. (1)

    The scope (F1) represents the percentage of variables that failed to meet the objective (above or below the acceptable range of the selected parameter) at least once (failed variables), relative to the total number of variables.

    $${\text{F1}} = \left( {\frac{{{\text{Number }}\;{\text{of }}\;{\text{failed }}\;{\text{variables}}}}{{{\text{Total}}\;{\text{ number}}\;{\text{ of }}\;{\text{variables}}}}} \right) \times 100$$
    (2)
  2. (2)

    The frequency (F2) represents the percentage of tests which do not meet the objectives (above or below the acceptable range of the selected parameter) (failed tests).

    $${\text{F2}} = \left( {\frac{{{\text{Number }}\;{\text{of }}\;{\text{failed }}\;{\text{tests}}}}{{{\text{Total}}\;{\text{ number}}\;{\text{ of }}\;{\text{tests}}}}} \right) \times 100$$
    (3)
  3. (3)

    The amplitude represents the amount by which failed tests values did not meet their objectives (above or below the acceptable range of the selected parameter). It is calculated in three steps.

    1. a.

      The excursion is termed each time the number of an individual parameter is further than (when the objective is a minimum, less than) the objective and is calculated by two Eqs. (4,5) referring to two cases. In case the test value must not exceed the objective:

      $${\text{excursion}}_{{\text{i}}} = \left( {\frac{{{\text{Failed }}\;{\text{test }}\;{\text{value}}_{{\text{i}}} }}{{{\text{Objective}}_{{\text{i}}} }}} \right) - 1$$
      (4)

      For the cases in which the test value must not fall below the objective:

      $${\text{ excursion}}_{{\text{i}}} = \left( {\frac{{{\text{Objective}}_{{\text{i}}} }}{{{\text{Failed}}\;{\text{ test }}\;{\text{value}}_{{\text{i}}} { }}}} \right) - 1$$
      (5)
    2. b.

      The normalized sum of excursions, or nse, is calculated by summing the excursions of individual tests from their objectives and diving by the total number of tests (both meetings and not meeting their objectives):

      $${\text{nse}} = \frac{{\mathop \sum \nolimits_{{{\text{i}} = 1}}^{{\text{n}}} {\text{excursion}}_{{\text{i}}} }}{{\text{number of tests}}}$$
      (6)
    3. c.

      F3 is then calculated an asymptotic function that scales the normalized sum of the excursions from objectives (nse) to yield a range between 0 and 100:

      $${\text{F3}}\, = \,\frac{{{\text{nse}}}}{{0.01{\text{nse}} + 0.01}}$$
      (7)

Finally, the CMEWQI can be obtained from the following equation, where the index changes in direct proportion to changes in all three factors.

$${\text{CCMEWQI}} = 100 - \left( {\frac{{\sqrt {{\text{F}}1^{2} + {\text{F}}2^{2} + {\text{F}}3^{2} } }}{1.732}} \right)$$
(8)

where 1.732 is a scaling factor and normalizes the resultant values to a range between 0 and 100, where 0 refers to the worst quality and one hundred represents the best water quality.

Once the CCME WQI value has been determined, water quality in ranked as shown in Table 3

Table 3 Water quality categorizations according to CCMEWQI (Kizar 2018; Canadian Council of Ministers of the Environment 1999)
Full size table

Ramírez-Morales et al. (2021) investigated in their study the measuring of pesticides and water quality indices in three agriculturally impacted micro catchments in Costa Rica between 2012 and 2014. Surface water and sediment samples were obtained during the monitoring experiment.

The specifications of the water included: Pesticides, temperature, DO, oxygen saturation, BOD, TP, NO3, sulfate, ammonium, COD, conductivity, pH and TSS.

Sediment parameters included forty-two pesticides with different families including carbamate, triazine, organophosphate, phthalimide, pyrethroid, uracil, benzimidazole, substituted urea, organochlorine, imidazole, oxadiazole, diphenyl ether and bridged diphenyl.

WQIs are effective tools since they combine information from several variables into a broad picture of the water body's state. Two WQIs were calculated using the physicochemical parameters: The Canadian Council of Ministers of the Environment (CCME) WQI and the National Sanitation Foundation (NSF) WQI.

These were chosen since they are both extensively used and use different criteria to determine water quality: The NSF WQI has fixed parameters, weights, and threshold values, whereas the CCME has parameters and threshold values that are customizable.

The assessment of water quality using physico-chemical characteristics and the WQI revealed that the CCME WQI and the NSF WQI have distinct criteria. CCME WQI categorized sampling point as marginal/bad quality, while most sampling locations were categorized as good quality in the NSF WQI. Seemingly, the water quality classifications appeared to be affected by seasonal variations: during the wet season, the majority of the CCME WQI values deteriorated, implying that precipitation and runoff introduced debris into the riverbed. Thus, it’s crucial to compare WQIs because they use various factors, criteria, and threshold values, which might lead to different outcomes (Ramírez-Morales et al. 2021).

Yotova et al. (2021) directed an analysis on the Mesta River located between Greece and Bulgaria. The Bulgarian section of the Mesta River basin, which is under the supervision of the West-Aegean Region Basin Directorate, was being researched. The goal was to evaluate the surface water quality of ten points of the river using a novel approach that combines composite WQI developed by the CCME and Self organizing map (SOM) on the required monitoring data that include: DO, pH, EC, ammonium, nitrite, nitrate, total phosphate, BOD and TSS.

The use of WQI factors in SOM calculations allows for the identification of specific WQI profiles for various object groups and identifying groupings of river basin which have similar sampling conditions. The use of both could reveal and estimate the origin and magnitude of anthropogenic pressure. In addition, it might be determined that untreated residential wastewaters are to blame for deviations from high quality requirements in the Mesta River catchment.

Interestingly, this study reveals that WQI appear more accurate and specific when combined with a statistical test such as the SOM (Yotova et al. 2021).

2.4.3 Oregon water quality index (OWQI)

The Oregon Water Quality Index is a single number that creates a score to evaluate the water quality of Oregon’s stream and apply this method in other geographical region (Hamlat et al. 2017; Singh et al. 2013). The OWQI was widely accepted and applied in Oregon (USA) and Idaho (USA) (Sutadian et al. 2016).

Additionally, the OWQI is a variant of the NSFWQI, and is used to assess water quality for swimming and fishing, it is also used to manage major streams (Lumb et al. 2011b). Since the introduction of the OWQI in 1970, the science of water quality has improved noticeably, and since 1978, index developers have benefited from increasing understanding of stream functionality (Bharti and Katyal 2011). The Oregon index belongs to the specific consumption indices group. It is a water classification based on the kind of consumption and application such as drinking, industrial, etc. (Shah and Joshi 2017).

The original OWQI dropped off in 1983, due to excessive resources required for calculating and reporting results. However, improvement in software and computer hardware availability, in addition to the desire for an accessible water quality information, renewed interest in the index (Cude 2001).

Simplicity, availability of required quality parameters, and the determination of sub-indexes by curve or analytical relations are some advantages of this approach (Darvishi et al. 2016a). The process combines eight variables including temperature, dissolved oxygen (percent saturation and concentration), biochemical oxygen demand (BOD), pH, total solids, ammonia and nitrate nitrogen, total phosphorous and bacteria (Brown 2019). Equal weight parameters were used for this index and has the same effect on the final factor (Darvishi et al. 2016a; Sutadian et al. 2016).

The Oregon index is calculated by the following Eq. 9 (Darvishi et al. 2016a):

$${\text{OWQI}} = \frac{{\text{n}}}{{\sqrt {\mathop \sum \nolimits_{{{\text{i}} = 1}}^{{\text{n}}} \frac{1}{{{\text{SI}}^{2} }}} }}$$
(9)

where,n is the number of parameters (n = 8) SIi is the value of parameter i.

Furthermore, the OWQI scores range from 10 for the worse case to 100 as the ideal water quality illustrated in the following Table 4 (Brown 2019).

Table 4 Average values of river water index according to OWQI index (Darvishi et al. 2016a)
Full size table

Kareem et al. (2021) using three water quality indices, attempted to analyze the Euphrates River (Iraq) water quality for irrigation purposes in three different stations: WAWQI, CCMEWQI AND OWQI.

For fifteen parameters, the annual average value was calculated, which included: pH, BOD, Turbidity, orthophosphate, Total Hardness, Sulphate, Nitrate, Alkalinity, Potassium Sodium, Magnesium, Chloride, DO, Calcium and TDS.

The OWQI showed that the river is “very poor”, and since the sub-index of the OWQI does not rely on standard-parameter compliance, there are no differences between the two inclusion and exclusion scenarios, which is not the case in both WAWQI and CCMEWQI (Kareem et al. 2021).

Similarly, the OWQI showed a very bad quality category, and it is unfit for human consumption, compared to the NSFWQI and Wilcox indices who both showed a better quality of water in Darvishi et al., study conducted on the Talar River (Iran) (Darvishi et al. 2016b).

2.4.4 Weighted arithmetic water quality index (WAWQI)

The weighted arithmetic index is used to calculate the treated water quality index, in other terms, this method classifies the water quality according to the degree of purity by using the most commonly measured water quality variables (Kizar 2018; Paun et al. 2016).This procedure has been widely used by scientists (Singh et al. 2013).

Three steps are essential in order to calculate the WAWQI:

  1. (1)

    Further quality rating or sub-index was calculated using the following equation (Jena et al. 2013):

    $${\text{Qn}} = { 1}00 \times \frac{{\left[ {{\text{Vn}} - {\text{Vo}}} \right]}}{{\left[ {{\text{Sn}} - {\text{Vo}}} \right]}}$$
    (10)

    where,

    Qn is the quality rating for the nth water quality parameter.

    Vn is the observed value of the nth parameter at a given sampling station.

    Vo is the ideal value of the nth parameter in a pure water.

    Sn is the standard permissible value of the nth parameter.

    The quality rating or sub index corresponding to nth parameter is a number reflecting the relative value of this parameter in polluted water with respect to its permissible standard value (Yogendra & Puttaiah 2008).

  2. (2)

    The unit weight was calculated by a value inversely proportional to the recommended standard values (Sn) of the corresponding parameters (Jena et al. 2013):

    $${\text{Wn}} = \frac{{\text{K}}}{{{\text{Sn}}}}$$
    (11)

    where,

    Wn is the unit weight for the nth parameter.

    K is the constant of proportionality.

    Sn is the standard value of the nth parameter.

  3. (3)

    The overall WQI is the aggregation of the quality rating (Qn) and the unit weight (Wn) linearly (Jena et al. 2013):

    $${\text{WQI}} = \frac{{\sum {\text{Qn Wn}}}}{{\sum {\text{Wn}}}}$$
    (12)

After calculating the WQI, the measurement scale classifies the water quality from “unsuitable water” to “excellent water quality” as given in the following Table 5.

Table 5 WAWQI and status of water quality (Yogendra and Puttaiah 2008)
Full size table

Sarwar et al. (2020) carried out a study in Chaugachcha and Manirampur Upazila of Jashore District (Bangladesh). The goal of this study was to determine the quality of groundwater and its appropriateness for drinking, using the WAWQI including nine parameters: turbidity, EC, pH, TDS, nitrate, ammonium, sodium, potassium and iron. Many samplings point was taken from Chaugachcha and Manirampur, and WQI differences were indicated (ranging from very poor to excellent). These variations in WQI were very certainly attributable to variances in geographical location. Another possibility could be variations in the parent materials from which the soil was created, which should be confirmed using experimental data. It is worth mentioning that every selected parameter was taken into consideration during calculation. Similarly, the water quality differed in Manirampur due to the elements contained in the water samples that had a big impact on the water quality (Sarwar et al. 2020).

In 2021, García-Ávila et al. undertook a comparative study between the CCMEWQI and WAWQI for the purpose of determining the water quality in the city of Azogues (Ecuador). Twelve parameters were analyzed: pH, turbidity, color, total dissolved solids, electrical conductivity, total hardness, alkalinity, nitrates, phosphates, sulfates, chlorides and residual chlorine over 6 months. The average WAWQI value was calculated suggesting that 16.67% of the distribution system was of 'excellent' quality and 83.33% was of 'good' quality, while the CCMEWQI indicated that 100% of the system was of ‘excellent’ quality.

This difference designated that the parameters having a low maximum allowable concentration have an impact on WAWQI and that WAWQI is a valuable tool to determine the quality of drinking water and have a better understanding of it (García-Ávila et al. 2022a, b).

2.4.5 Additional water quality indices

The earliest WQI was based on a mathematical function that sums up all sub-indices, as detailed in the 2.1. History of water quality concept section (Aljanabi et al. 2021). The Dinius index (1972), the OWQI (1980), and the West Java index (2017) were later modified from the Horton index, which served as a paradigm for later WQI development (Banda and Kumarasamy 2020).

Based on eleven physical, chemical, organic, and microbiological factors, the Scottish Research Development Department (SRDDWQI) created in 1976 was based on the NSFWQI and Delphi methods used in Iran, Romania, and Portugal. Modified into the Bascaron index (1979) in Spain, which was based on 26 parameters that were unevenly weighted with a subjective representation that allowed an overestimation of the contamination level. The House index (1989) in the UK valued the parameters directly as sub-indices. The altered version was adopted as Croatia's Dalmatian index in 1999.

The Ross WQI (1977) was created in the USA using only 4 parameters and did not develop into any further indices.

In 1982, the Dalmatian and House WQI were used to create the Environmental Quality Index, which is detailed in Sect. 2.1. This index continues to be difficult to understand and less powerful than other indices (Lumb et al. 2011a; Uddin et al. 2021).

The Smith index (1990), is based on 7 factors and the Delphi technique in New Zealand, attempts to eliminate eclipsing difficulties and does not apply any weighting, raising concerns about the index's accuracy (Aljanabi et al. 2021; Banda and Kumarasamy 2020; Uddin et al. 2021).

The Dojildo index (1994) was based on 26 flexible, unweighted parameters and does not represent the water's total quality.

With the absence of essential parameters, the eclipse problem is a type of fixed-parameter selection. The Liou index (2004) was established in Taiwan to evaluate the Keya River based on 6 water characteristics that were immediately used into sub-index values. Additionally, because of the aggregation function, uncertainty is unrelated to the lowest sub-index ranking (Banda and Kumarasamy 2020; Uddin et al. 2021).

Said index (2004) assessed water quality using only 4 parameters, which is thought to be a deficient number for accuracy and a comprehensive picture of the water quality. Furthermore, a fixed parameter system prevents the addition of any new parameters.

Later, the Hanh index (2010), which used hybrid aggregation methods and gave an ambiguous final result, was developed from the Said index.

In addition to eliminating hazardous and biological indicators, the Malaysia River WQI (MRWQI developed in the 2.1 section) (2007) was an unfair and closed system that was relied on an expert's judgment, which is seen as being subjective and may produce ambiguous findings (Banda and Kumarasamy 2020; Uddin et al. 2021).

Table illustrated the main data of the studies published during 2020–2022 on water quality assessments and their major findings:

2.5 Advantages and disadvantages of the selected water quality indices

A comparison of the selected indices is done by listing the advantages and disadvantages of every index listed in the Table 7 below.

2.6 New attempts of WQI studies

Many studies were conducted to test the water quality of rivers, dams, groundwater, etc. using multiple water quality indices throughout the years. Various studies have been portrayed here in.

Massoud (2012) observed during a 5-year monitoring period, in order to classify the spatial and temporal variability and classify the water quality along a recreational section of the Damour river using a weighted WQI from nine physicochemical parameters measured during dry season. The WWQI scale ranged between “very bad” if the WQI falls in the range 0–25, to “excellent” if it falls in the range 91–100. The results revealed that the water quality of the Damour river if generally affected by the activities taking place along the watershed. The best quality was found in the upper sites and the worst at the estuary, due to recreational activities. If the Damour river is to be utilized it will require treatment prior any utilization (Massoud 2012).

Rubio-Arias et al. (2012) conducted a study in the Luis L. Leon dam located in Mexico. Monthly samples were collected at 10 random points of the dam at different depths, a total of 220 samples were collected and analyzed. Eleven parameters were considered for the WQI calculation, and WQI was calculated using the Weighted WQI equation and could be classified according to the following ranges: < 2.3 poor; from 2.3 to 2.8 good; and > 2.8 excellent. Rubio-Arias et al., remarked that the water could be categorized as good during the entire year. Nonetheless, some water points could be classified as poor due to some anthropogenic activities such as intensive farming, agricultural practices, dynamic urban growth, etc. This study confirms that water quality declined after the rainy season (Rubio-Arias et al. 2012).

In the same way, Haydar et al. (2014) evaluated the physical, chemical and microbiological characteristics of water in the upper and lower Litani basin, as well as in the lake of Qaraaoun. The samples were collected during the seasons of 2011–2012 from the determined sites and analyzed by PCA and the statistical computations of the physico-chemical parameters to extract correlation between variables. Thus, the statistical computations of the physico-chemical parameters showed a correlation between some parameters such as TDS, EC, Ammonium, Nitrate, Potassium and Phosphate. Different seasons revealed the presence of either mineral or anthropogenic or both sources of pollution caused by human interference from municipal wastewater and agricultural purposes discharged into the river. In addition, temporal effects were associated with seasonal variations of river flow, which caused the dilution if pollutants and, hence, variations in water quality (Haydar et al. 2014).

Another study conducted by Chaurasia et al., (2018), proposed a groundwater quality assessment in India using the WAWQI. Twenty-two parameters were taken into consideration for this assessment, however, only eight important parameters were chosen to calculate the WQI. The rating of water quality shows that the ground water in 20% of the study area is not suitable for drinking purpose and pollution load is comparatively high during rainy and summer seasons. Additionally, the study suggests that priority should be given to water quality monitoring and its management to protect the groundwater resource from contamination as well as provide technology to make the groundwater fit for domestic and drinking (Chaurasia et al. 2018).

Daou et al. (2018) evaluated the water quality of four major Lebanese rivers located in the four corners of Lebanon: Damour, Ibrahim, Kadisha and Orontes during the four seasons of the year 2010–2011. The assessment was done through the monitoring of a wide range of physical, chemical and microbiological parameters, these parameters were screened using PCA. PCA was able to discriminate each of the four rivers according to a different trophic state. The Ibrahim River polluted by mineral discharge from marble industries in its surroundings, as well as anthropogenic pollutants, and the Kadisha river polluted by anthropogenic wastes seemed to have the worst water quality. This large-scale evaluation of these four Lebanese rivers can serve as a water mass reference model (Daou et al. 2018).

Moreover, some studies compared many WQI methods. Kizar (2018), carried out a study on Shatt Al-Kufa in Iraq, nine locations and twelve parameters were selected. The water quality was calculated using two methods, the WAWQI and CWQI. The results revealed the same ranking of the river for both methods, in both methods the index decreased in winter and improved in other seasons (Kizar 2018).

On the other hand, Zotou et al. (2018), undertook a research on the Polyphytos Reservoir in Greece, taking into consideration thirteen water parameters and applying 5 WQIs: Prati’s Index of Pollution (developed in 1971, based on thirteen parameter and mathematical functions to convert the pollution concentration into new units. The results of PI classified water quality into medium classes (Gupta and Gupta 2021). Bhargava’s WQI (established in 1983, the BWQI categorize the parameters according to their type: bacterial indicators, heavy metals and toxins, physical parameters and organic and inorganic substances. The BWQI tends to classify the water quality into higher quality classes, which is the case in the mentioned study (Gupta and Gupta 2021). Oregon WQI, Dinius second index, Weighted Arithmetic WQI, in addition to the NSF and CCMEWQI. The results showed that Bhargava and NSF indices tend to classify the reservoir into superior quality classes, Prati’s and Dinius indices fall mainly into the middle classes of the quality ranking, while CCME and Oregon could be considered as “stricter” since they give results which range steadily between the lower quality classes (Zotou et al. 2018).

In their study, Ugochukwu et al. (2019) investigated the effects of acid mine drainage, waste discharge into the Ekulu River in Nigeria and other anthropogenic activities on the water quality of the river. The study was performed between two seasons, the rainy and dry season. Samples were collected in both seasons, furthermore, the physic-chemistry parameters and the heavy metals were analyzed. WQI procedure was estimated by assigning weights and relative weights to the parameters, ranking from “excellent water” (< 50) to “unsuitable for drinking” (> 300). The results showed the presence of heavy metals such as lead and cadmium deriving from acid mine drainage. In addition, the water quality index for all the locations in both seasons showed that the water ranked from “very poor” to “unsuitable for drinking”, therefore the water should be treated before any consumption, and that enough information to guide new implementations for river protection and public health was provided (Ugochukwu et al. 2019).

The latest study in Lebanon related to WQI was carried out by El Najjar et al. (2019), the purpose of the study was to evaluate the water quality of the Ibrahim River, one of the main Lebanese rivers. The samples were collected during fifteen months, and a total of twenty-eight physico-chemical and microbiological parameters were tested. The parameters were reduced to nine using the Principal Component Analysis (PCA) and Pearson Correlation. The Ibrahim WQI (IWQI) was finally calculated using these nine parameters and ranged between 0 and 25 referring to a “very bad” water quality, and between 91 and 100 referring to an “excellent” water quality. The IWQI showed a seasonal variation, with a medium quality during low -water periods and a good one during high-water periods (El Najjar et al. 2019).

3 Conclusion

WQI is a simple tool that gives a single value to water quality taking into consideration a specific number of physical, chemical, and biological parameters also called variables in order to represent water quality in an easy and understandable way. Water quality indices are used to assess water quality of different water bodies, and different sources. Each index is used according to the purpose of the assessment. The study reviewed the most important indices used in water quality, their mathematical forms and composition along with their advantages and disadvantages. These indices utilize parameters and are carried out by experts and government agencies globally. Nevertheless, there is no index so far that can be universally applied by water agencies, users and administrators from different countries, despite the efforts of researchers around the world (Paun et al. 2016). The study also reviewed some attempts on different water bodies utilizing different water quality indices, and the main studies performed in Lebanon on Lebanese rivers in order to determine the quality of the rivers (Table 6).

Table 6 Various research projects carried out on WQIs
Full size table

As mentioned in the article (Table 7); WQIs may undergo some limitations. Some indices could be biased, others are not specific, and they may not get affected by the value of an important parameter. Therefore, there is no interaction between the parameters.

Table 7 Advantages and disadvantages of the selected water quality indices
Full size table

Moreover, many studies exhibited a combination between WQIs and statistical techniques and analysis (such as the PCA, Pearson’s correlation etc.). with a view to obtain the relation between the parameters and which parameter might affect the water quality.

In other research, authors compared many WQIs to check the difference of water quality according to each index. Each index can provide different values depending on the sensitivity of the parameter. For that reason, WQIs should be connected to scientific advancements to develop and elaborate the index in many ways (example: ecologically). Therefore, an advanced WQI should be developed including first statistical techniques, such as Pearson correlation and multivariate statistical approach mainly Principal Component Analysis (PCA) and Cluster Analysis (CA), in order to determine secondly the interactions and correlations between the parameters such as TDS and EC, TDS and total alkalinity, total alkalinity and chloride, temperature and bacteriological parameters, consequently, a single parameter could be selected as representative of others. Finally, scientific and technological advancement for future studies such as GIS techniques, fuzzy logic technology to assess and enhance the water quality indices and cellphone-based sensors for water quality monitoring should be used.