The hydrochemical data provides an overview of the physical-chemical parameters measured in the groundwater and surface water samples. The pH values ranged from 7.2 to 8.1 in the Pleistocene aquifer, from 7.1 to 7.6 in the Miocene aquifer and from 7.1 to 8.3 in surface water reflected that the groundwater and surface water samples are somewhat neutral to slightly alkaline. The electric conductivity(EC) values represent water's dissolved salt content and higher values typically reflect higher concentrations of ions in the water (Prasanth et al. 2012). The EC values ranged from 8520 to 19410 μS/cm in the Pleistocene groundwater, from 4290 to 7920 μS/cm in the Miocene groundwater and from 3970- 64410 μS/cm in the surface water. The EC value differences are attributable to the composition of the aquifer rocks. The total dissolved solids (TDS) values ranged from 5041 to 11741 mg/l in the Pleistocene groundwater, from 2360 to 4742 mg/l in the Miocene groundwater and from 2455 to 41958 mg/l in the surface water. Major ions are observed in Na+ and Cl− correspondingly as prominent cation and anion species in both the groundwater and surface water sample concentrations. The concentration of Na+ varies from 1450 to 3500 mg/L in the Pleistocene groundwater, from 580 to 1380 mg/L in the Miocene groundwater, and from 740 to13400 mg/L in the surface water. The ionic concentration of Cl− is highest among all the ions and the concentration of Cl− in pleistocene groundwater is between 2211 and 6076 mg/L, in Miocene groundwater is between 951 to 2262 mg/L, and in surface water from 1054 to 20567 mg/L. Both the surface and the groundwater display significantly higher Na+ and Cl− concentrations, indicating that seawater most likely affects the water quality in the area under investigation and this indicates the mixing of groundwater with the matrix of marine aquifers. The concentration of K+ shows the least variation with the range of 52 to 100 mg/L in the Pleistocene groundwater, from 32 to 55 mg/l in the Miocene groundwater and 28 to 600 mg/L in the surface water. In the Pleistocene groundwater, Ca2+ and Mg2+ concentrations range between 131 - 767 mg/L and 152 - 270 mg/L, respectively, also in the Miocene groundwater Ca2+ and Mg2+ concentrations range between 129 - 306 mg/L and 113-164 mg/L. Similarly, the concentrations of SO42− and alkalinity vary from 344 to 2650 mg/L and 50 to 205 mg/L, in the Pleistocene groundwater, while in the Miocene groundwater ranges from 140 to1340mg/L and 135 to 275 mg/ L, in each case. The dissolving of marine sediments is the result of the high salinity of groundwater. Table (3), Figure (6).
A multi-rectangular hydrochemical facies evolution diagram (HFE) can be employed to determine the dynamics of seawater intrusion, considering the percentages of major ions, showing the intruding and freshening phases. Figure (7) shows that the majority of samples are appropriate for a phase of intrusion. The Na-Cl facies signifies the state of aquifer. The samples shown in HFE-D confirms the hypotheses regarding salinization with the exception of one sample (G1), is scattered in the field of freshening. The methodology of classification proposed in the present study takes into account that of Gimenez FE (2010).
5.2. Statistical analysis
5.2.1. Cluster analysis
Cluster analysis (CA) is a statistical analysis based on their similarity used in grouping samples in clusters and it is one of the most frequently used for evaluating the surface water effect on groundwater quality. The dendogram of two aquifers and surface water obtained by agglomerative hierarchical clustering are shown in Figures (8and19).Two dendrograms were produced which are certified to illustrate significant pollution. The diagram of CA calculated by using all detected parameters in water depend on chemical characteristics such as pH, EC, Ca, Mg, CO3, HCO3, Na, K, SO4, Cl, BOD, TOC, COD, NO3,NO2, NH4. The dendogram in Pleistocene was grouped by HCA into two groups Figure (8). Cluster 1 (C-1) contains 22samples (18 groundwater and 4 surface water); this cluster shows pollution and fecal contamination. Appearance of contaminated water at a sewage treatment plant in the study area, the remaining 2 samples, including two surface water is collective in cluster 2 (C-2). The variation of water quality parameters in cluster 1 of samples is presented in Figure (8) which indicated that the influence of sewage treatment plant on groundwater. In the Miocene aquifer, the dendogram also was classify into two clusters Figure (9). (C-1) includes 11 samples (7 groundwater and 4 surface water), (C-2) corresponding to surface water. The C-1 samples had substantially higher levels of water quality parameters, the predominance of the average cations being Na+ > Mg2+ > Ca2+ > K+ for both samples' clusters. The mean anion levels are in the category of Cl− > SO4 2− > HCO3.
5.2.2. Factorial analysis
The factor analysis was executed on 7 variables such as boron(B), strontium(Sr), biological oxygen demand(BOD), phosphorous(P), chromium(Cr), nickel(Ni) and total organic carbon(TOC) for the Pleistocene and Miocene groundwater, so that is identified the influence of surface water and treatment plant on groundwater. An eign value measures the relevance of the factor: the factors with maximum eign values are the most significant. Eigenvalues of 1.0 or larger are reflected significant values (Shrestha and Kazama 2007). Thus the classification for factor loads is 'strong,' 'moderate' and 'weak' which correspond to >0.75, 0.75-0.50 and 0.50-0.30 absolute loading values, respectively (Liu et al. 2003). Table (4) provides variable loads and explained variation and high loading values. The four factors of Miocene groundwater include more than 81.8 % of the total cumulative variance, and the three factors of Pleistocene groundwater include more than 71 % of the total cumulative variance respecting water quality data sets. The most significant parameters with strong positive factors in Miocene groundwater are B, Sr, BOD, Ni, and TOC and in Pleistocene groundwater are B, Sr, BOD, Ni, and TOC. Strontium and Boron with positive strong loading value have contributed as the most important parameters to changes in water quality in Miocene and Pleistocene ground water. This means that a significant amount of inorganic nutrients is caused by agriculture. The biochemical demands for oxygen, nickel and total organic carbon with strong factor loads are key parameters in the variability of water quality and it explains the fact that waste water and industrial wastewater enters the study area and causes significant pollution caused by the treatment plant. (Pejman et al. 2009).
5.3. Pollution indices
5.3.1. Heavy metal pollution index:
The mean concentrations were calculated to guess the heavy metal pollution index (HPI). Calculated index values and unit weightage values were listed in Table (1). In the current study, metals for example V, Cu, Mo, Cr, B, Fe, Cd, Mn, Ni, Pb, Al, and Zn were measured. The HPI for the study area is intended by integrating the mean concentration values of confirmed heavy metals. The particulars of the calculation are existing in Table (8). HPI is categorized into five classes; excellent (0–25), good (26–50), poor (51–75), very poor (76–100), and unsuitable (100). 72.2% of the Pleistocene aquifer samples is considered unsuitable for drinking purposes, 16.7% poor, 5.5% very poor, and the remaining samples5.5% is considered good, on the other hand, 71.5% of the Miocene aquifer samples is considered unsuitable for drinking purposes, and the remaining 28.5% is considered good. The results were assessed that in both aquifers (Pleistocene, Miocene), the heavy metal pollution index exceeds 100 in the majority of the samples. Wells are shown to be contaminated by heavy metals. It was estimated that the region of the research would be affected by heavy metal leakage from the water treatment plant, as shown in Figure (10). The water treatment plant has not treated the inorganic matters especially the heavy metals.
5.3.2. Nitrate pollution index:
Nitrate levels were ranged from 1 to 13.2 mg/L with an average of 4.03 mg/L in the study area. Nitrate was organized into three groups, 1) low (<20 mg/L), 2) medium (≥ 20 mg/L to <50 mg/L), and 3) high (≥50 mg/L). The concentration of nitrate in all the samples in the studied area is less than 20 mg/L. Five classifications of water have been determined according to NPI values: clean, light pollution, moderate pollution, significant pollution, and highly significant pollution, with NPI values of <0, 0-1, 1-2, 2-3, and >3, respectively. NPI is smaller than zero for all groundwater samples in the class clean Table (8).
5.3.3. Pollution index of groundwater:
The relative contribution of the pollutants from each ground water sample was evaluated in a Pollution Index (PIG) assessment. The chemical water quality (Ow) of pH and NO3 of less than 0.1 Table (6) shows a low impact on groundwater contamination in the current study. Depends on the data in Table (6), Pb and Fe had the greatest effect on sample water quality. This is evident in the values of Ow and PIG achieved. In this study the final PIG values were around 1.4 and 6.0. In five categories, the level of drinking water pollution is divided: PIG < 1.0 indicates insignificant pollution; 1.0–1.5 refers to the low pollution, 1.5-2.0 is moderate pollution; 2.0–2.5 signifies high pollution; and PIG > 2.5 shows very high pollution Table (7) (Rao 2012; Rao et al. 2018; Rao and Maya 2019). Based on this classification, 94.4% of Pleistocene water samples were found to be very high polluted, while the remaining samples, 5.6%, were highly polluted and unsuited for drinking. Nevertheless, Miocene aquifers include a very low pollution level of 14.3%, moderate pollution of 14.3%, high pollution of 28.6% and very high pollution of 42.8%. The highest anthropogenic input is probably found in the samples identified as unfit to drink and is observed from both the north and south regions of the study area Figure (11).
5.3.4. Ecological risk index:
For each heavy metals and water sample, the RI (potential ecological risk) was initially identified during the ERI evaluation. Table (7). According to (Bhutiani et al. 2017; Adimalla and Wang 2018; Taiwo et al. 2019), RI is divided into five, to reflect its impact on sample quality of heavy metal: RI < 40 (low potential risk), 40 ≤ RI < 80 (moderate potential risk), 80 ≤ RI < 160 (considerable potential risk), 160 ≤ RI < 320 (high potential risk), and ≥ 320 (very high potential risk). In the current study, depend on this classification; Cd poses moderate potential risks to sample G14, considerable potential risk index to samples G11, G13, G15, and G29, high potential risk index to samples G9, G16, G19 and G7, very high potential to sample G27. However, the samples G6, G13, G19, G1, G26, G27 and G29 have been found to be moderately at risk by Pb Table (8).
The final ERI values achieved from this analysis ranged from 10.9 to 388.3 Table (8). The ground water can be divided into four categories based on the ERI values: ERI < 150 (low risk), 150 < ERI (moderate ecological risk), 300 < ERI < 600 (considerable risk), and 276 > 600 (very high risk) (Adimalla and Wang 2018; Taiwo et al. 2019). According to this classification scheme, 66.7% of the Pleistocene aquifer samples have low ecological risk, 22.2% have moderate risk, and 11% are considerable risks. However, 42.8% of the Miocene aquifer has low ecological risk, 22.8% have moderate risk and 14.4% are a considerable risk Figure (12).