**1.2 Water quality**

Water quality is represented by a set of physical, chemical, biological and bacteriological characteristics. These characteristics are also called parameters or indicators. Physical, chemical, biological and bacteriological parameters allow water to be classified in some categories, leading to its utilization for a specific use.

Water quality requirements depend on the purposes for which the water will be used. Thus, drinking water must not contain chemicals or micro-organisms which can affect the human health. Water used in agriculture must not contain large amounts of sodium ions, high concentrations of nitrates or high concentrations of other contaminants. Requirements for water use in industry are less rigorous than drinking water [3].

Water quality also depends on the type of water source and changes with geological, meteorological and land use conditions. The World Health Organization (WHO) has established regulations and standards for water safety in support of public health [3]. The European Union has, also, established a legal framework for water protection [4]. Water quality criteria in all countries have been established in accordance with the WHO guidelines [3]. In European countries, the framework directives of the European Union are closely followed [4].

The EU Framework Directive requires that operational monitoring should be specific and based on monitoring relevant biological, hydro-morphological and physic-chemical parameters. These world environmental monitoring systems provide for water quality measurements in three categories of parameters:


Water quality monitoring indicators established by the European Union rules are grouped according to their common characteristics as it follows [4, 5]:

1. Indicators that give information about oxygen condition: Dissolved Oxygen, Biochemical Oxygen Consumption (CBO5), Chemical Oxygen Consumption with chromium, Chemical Oxygen Consumption with manganese;

*Water Quality Parameters and Monitoring Soft Surface Water Quality Using Statistical… DOI: http://dx.doi.org/10.5772/intechopen.97372*


To understand the overall health of an ecosystem and the condition of water, a number of water quality parameters or indicators must be analyzed and monitored. In 1998, Sene and Farquharson [6] stated that monitoring of the surface water quality is necessary to assess spatial and temporal regional variations. The process of monitoring the quality of ambient water has led to the development of water standards and the periodic assessments of the environment.

The monitoring program and the parameters to be measured for the study of water quality should be chosen specifically for each locality and each type of water. Although many parameters of water are important for human health or the health of an ecosystem, the analysis of all parameters is not feasible. The standards recommend the analysis of specific parameters for both drinking water and nondrinking water [6, 7].

Chemical and physical parameters are important in the rapid determination of water quality while biological parameters provide a detailed and complex analysis of the environment [8].

#### *1.2.1 General physico-chemical parameters*

*Temperature* is an important parameter that influences the chemical properties of water. Temperature affects the density and stratification of water, the density and viscosity of transported sediments, solubility of dissolved gases, vapor pressure [9].

*pH* is determined only at the place where the sample was collected, directly from the water source while also determining the air temperature. Due to the presence of carbon dioxide, bicarbonates, and carbonates, the pH of the water varies very little from the neutral pH. The pH of natural water is usually in the range 6.5–8. The pH of the wastewater can be alkaline if pH is higher than 7 or acidic if pH is lower than 7 [10].

*The conductivity* of water is given by the presence of ions in the solution, ions that have the property of transmitting electric current. The higher the ionic concentration of the solution is, the higher the conductivity gets. The conductivity value depends on the amount of substances dissolved in the water. As a rule, high turbidity value also implies a high conductivity [10].

*Turbidity* expresses the amount of light reflected or absorbed by particles suspended in a water sample and is a measure of its relative clarity. Turbidity is due to solid particles in the form of suspensions or in a colloidal state [10].

#### *1.2.2 Chemical parameters*

#### *1.2.2.1 Oxygen regime indicators*

*Dissolved Oxygen* is an indicator of water quality whose values are dependent on the type of the water. The amount of oxygen dissolved in water depends on water temperature, air pressure and the quantity of acidic substances and microorganisms.

Oxygen is necessary for aquatic life. A series of aerobic chemical processes take place through dissolved oxygen: the oxidation processes of organic matter, oxidation of mineral substances, and bio-chemical decomposition of the dead bodies in water [9]. With the decrease of oxygen, the self-purification capacity of natural water is reduced, favoring the persistence of pollution with its undesirable consequences. Other indicators of oxygen regime are Biochemical Oxygen Consumption (CBO5) and Chemical Oxygen Consumption (CCO). *Biochemical Oxygen Consumption* (CBO5) is the amount of oxygen consumed by microorganisms, during a 5 day period, for the biochemical decomposition of organic substances contained in water, at a temperature of 20°C. *Chemical Oxygen Consumption with chromium* (CCOCr) is an integral index of the existence of difficult degradable organic substances. *Chemical Oxygen Consumption with manganese* (CCOMn) is a comprehensive index of the existence of easily degradable organic substances [11].

## *1.2.2.2 Biogenic indicators*

*Nitrites* in water represent the incomplete oxidation of organic nitrogen. Their presence in the water indicates an old pollution, because the transformation of organic substances containing nitrogen under the action of microorganisms first convert into ammonia then ammonia converts into nitrites. Therefore, the concentration of nitrites in the water may indicate an old pollution because all these transformations take time. Under normal oxygenation of natural water, nitrogen appears in the form of nitrates. The chemical forms of nitrite and ammonium are present when water pollution occurs and are toxic to living organisms [10].

*Nitrates* represent the final stage of oxidation of organic nitrogen. If ammonia, nitrites and nitrates are present simultaneously in the water, this indicates a continuous pollution. The simultaneous presence of ammonia and nitrates in the water indicates an intermittent pollution [11].

The phosphate content in natural water is relatively low. High amounts of phosphorus in water can come from excessive use of nitrogen and phosphorus fertilizers. Higher concentrations of phosphorus in surface water can result in eutrophication.

#### *1.2.2.3 Salinity indicators*

*Salinity* is the content of mineral salts in water, mainly metal salts such as sodium, magnesium and calcium. The salts present in natural water are formed by the following cations Ca2+, Mg2+, Na+ , K<sup>+</sup> and anions HCO3 , SO4 <sup>2</sup>, Cl.

*The chlorides* in the water come either from natural soil layers, pollution or animal origin. The amount of chlorides that are released from the soil is relatively constant and varies slightly over time. A significant increase in chloride content is usually an index of organic pollution [11].

*Hardness* is an indirect indicator of the degree of mineralization of water.

#### *1.2.2.4 Heavy metals*

Heavy metals are those metals that have a high density (i.e. 5 g/cm3 ) [10]. In low concentrations, heavy metal ions are essential for the development of metabolic processes in plants and animals. These metals (e.g., cadmium, chromium, cobalt, lead, nickel, mercury, selenium) can come from natural or anthropogenic processes. If certain concentrations are exceeded, then they become toxic substances for the living organisms.

*Water Quality Parameters and Monitoring Soft Surface Water Quality Using Statistical… DOI: http://dx.doi.org/10.5772/intechopen.97372*

#### *1.2.3 Biological and bacteriological indicators*

Water quality and its changes due to various forms of pollution may influence the composition of aquatic biocenoses. Biological analysis consists of an inventory of phytoplankton, zooplankton, benthic organisms or periphyton from water samples.

The microbial flora found in the water can be classified into two categories: water-specific microbial flora and microbial impurity flora. Water-specific microbial flora consists of microorganisms that commonly inhabit water and soil: cocci bacilli, different fungi and bacterial species which play a role in the natural degradation processes of organic substances. Microbial impurity flora consists of species of microorganisms of human or animal origin. This category can include pathogenic saprophytes. These microbes are generally accompanied by high concentrations of organic matter which provide their nutritional support [11].

In bacteriological analysis of water, the total number of germs and the determination of the bacillus coli have been adopted as bacteriological indicators.

#### **2. Statistical analyses for assessing the surface water quality parameters**

Water quality is determined by the biological, chemical and physical parameters of the water. Most often, it is not enough to measure these water quality indicators. In order to draw some solid conclusions, it is necessary to apply adequate statistical method to the measurements. These statistical methods can provide useful information that can lead to actionable advice regarding water management. There are a large number of statistical methods for examining water quality.

The main differences between these methods are the statistical techniques used and the significance of the values determined for each parameter. Statistical indices developed using water quality parameters can be linear, non-linear, segmented linear or segmented non-linear [12]. In order to have a global vision of the changes of the water quality in space and in time, various indices have been developed [13].

The water quality index (WQI) is represented by a number that expresses the general water quality in a particular location, over time, based on several water quality parameters. The aim of this index is to transform a large number of complex water quality measurements into information that is easy for water managers and the public to understand and to use. Are a multitude of methods for calculating water quality indices (WQI). In the following, we present the weighted average method. This method was proposed by Horton in 1965 and developed by Brown et al. in the year 1970 [14].

For the calculation of the WQI, the following expression was used [15–19]:

$$WQI = \frac{\sum q\_n W\_n}{\sum W\_n} \tag{1}$$

where:

*n* is the number of the water quality parameters. *Wn* is the Unit Weight:

$$\mathcal{W}\_n = \frac{K}{\mathcal{S}\_n} \tag{2}$$

*qn* is the Quality rating:

*Promising Techniques for Wastewater Treatment and Water Quality Assessment*

$$\mathbf{q\_n} = \frac{\mathbf{V\_n} - \mathbf{V\_{id}}}{\mathbf{S\_n} - \mathbf{V\_{id}}} \times \mathbf{100} \tag{3}$$

where:

*Vn* represents the measured value for the nth parameter of the water corresponding to a given sample;

*Vid* is the ideal value for the nth parameter corresponding to pure water; Vid values are zero for most parameters except for pH and dissolved oxygen. Vid for the pH is 7 and for dissolved oxygen (DO) is 14.6 mg/l [20].

*Sn* represents the standard value allowed for the nth parameter;

*K* it is a constant of proportionality calculated with the formula:

$$K = \frac{1}{\sum \frac{1}{\mathbb{S}\_u}}.\tag{4}$$

Water quality is Excellent if the WQI index score is between 0 and 25; Good for values of 26–50; Poor for WQI = 51–75; and, Very Poor for values between 76 and 100. If the value of the WQI index exceeds the value of 100, then the water is unsuitable for drinking and cannot be transformed into drinking water by any process [19, 21].

To study the relationship between two parameters of water samples, several correlation coefficients can be used. The statistics used most often are Pearson and Spearman coefficients. Linear correlation can be determined using the Pearson correlation coefficient while non-linear correlation can be determined using the Spearman coefficient. The Pearson correlation coefficient is a statistical technique that measures and describes the degree of linear association between two normally distributed continuous quantitative variables [21]. Let *x* and *y* be two variables, in our case two indicators of water quality. The Pearson coefficient, r, is calculated using the expression:

$$r = \frac{\mathbb{S}\_{\text{xy}}}{\mathbb{S}\_{\text{x}} \bullet \mathbb{S}\_{\text{y}}}; r = \frac{\sum (\mathbb{x} - \overline{\mathbb{x}})(\mathbb{y} - \overline{\mathbb{y}})}{\sqrt{\sum \left(\mathbb{x} - \overline{\mathbb{x}}\right)^{2} \left(\mathbb{y} - \overline{\mathbb{y}}\right)^{2}}} \tag{5}$$

where *Sxy* represents the covariance, *Sx* and *Sy* are the standard deviations of the two variables x and y; *x* and *y* and are the mean values of the two variables x and y [21]. The Pearson coefficient takes values between �1 and + 1. The value of the coefficient indicates the strength of the relationship between parameters while the sign of the coefficient indicates the direction of the linear association. If the sign is positive, the two variables are directly correlated and, if the sign is negative, the two variables are inversely correlated. The closer of the Pearson correlation coefficient is to the value of 1, the stronger the "intensity" of the linear relationship between the two variables [21]. The variables x and y are independent if r has the value 0 (**Figure 1**).

The minimum value of the Pearson coefficient (r = 0) is not an indicator of independence of the two characteristics (variables), but only of their noncorrelation. The coefficient of determination (r<sup>2</sup> ) is the square of the Pearson coefficient. The coefficient of determination indicates the percentage of the total variation of the dependent variable (y) which is explained by the independent variable (x).

Spearman method is a non-parametric method used when the relationship between two variables is not linear (monotonic correlation) [23–25]. The Spearman coefficient addresses some limitations of the Pearson coefficient. It is denoted either *Water Quality Parameters and Monitoring Soft Surface Water Quality Using Statistical… DOI: http://dx.doi.org/10.5772/intechopen.97372*

**Figure 1.** *Examples of the geometric significance of the Pearson coefficient.*

with ρ or with rS and represents an alternative to the Pearson coefficient. To calculate the coefficient, the data must have an order or rank. Coefficient can be calculated using the formula:

$$r\_S = 1 - \frac{\mathfrak{G}\sum\_{i}^{n} d\_i^2}{n^3 - n} \tag{6}$$

where:

*n* is the number of pairs of values ordered in ascending order,

*di* is the difference between the orders of each pair(*i*) of values or the rank of the value:

$$d\_i = r\_{\mathbf{x}\_i} - r\_{\mathbf{y}\_i} \tag{7}$$

where:

*rxi* is the rank of the value of *xi* in the ascending ordered system and *ryi* is the rank of the value *yi* in the ascending ordered system [21–25].

Eq. (6) is usually used when all n ranks are distinct integers or do not have tied ranks. When there are tied ranks, Eq. (6) is replaced by the following form:

$$r\_S = \frac{S\_{xy}}{S\_x \bullet S\_y} = \frac{\frac{1}{n} \sum\_{i}^{n} (r\_{\chi\_i} - \overline{r\_{\chi}}) \left(r\_{\overline{\chi}\_i} - \overline{r\_{\overline{\chi}}}\right)}{\sqrt{\frac{1}{n} \sum\_{i}^{n} (r\_{\chi\_i} - \overline{r\_{\chi}})^2 \frac{1}{n} \sum\_{i}^{n} \left(r\_{\overline{\chi}\_i} - \overline{r\_{\overline{\chi}}}\right)^2}} \tag{8}$$

where *rx* and *ry* are the mean ranks of value *x* and value *y* [23].

Spearman coefficient values are in the range [�1, 1]. The interpretation of these values is similar with that of the Pearson coefficient [21].

For a correct interpretation, the correlation coefficient must be accompanied by a significance test. The correlation coefficient has statistical significance if the value level of confidence factor p < 0.05. This significance coefficient p means the probability of making erroneous statements. If p < 0.05, we could reject the null hypothesis H0 and the computed results has certain statistical significance [24]. If the p result of the test is less than the significance threshold α (α = 0.05), hypothesis H1 is accepted: there is monotonic correlation. If p is greater than 0.05, then the H0 hypothesis is valid, which considers that there is no monotonic correlation [25].
