Delineation of Salinization and Recharge Sources Affecting Groundwater Quality Using Chemical and Isotopic Indices in the Northwest Coast, Egypt

: Salinization of coastal aquifers is a serious issue affected by climate change and enhanced by overexploitation of groundwater resources. This research aims to explore the hydrogeochemical processes that cause salinization of groundwater in coastal aquifers, such as the area located between Barrani and Baqbaq, on the northwestern coast of Egypt. Various techniques were applied, including Gibbs plots and hydrochemical facies diagrams (HFE-D), ion ratios and stable isotope bivariate plots, statistical analyses, a groundwater quality index for seawater intrusion (GQI SWI ), and a seawater mixing index (SMI). Based on the total dissolved solids (TDS), groundwater can be classiﬁed into four groups: slightly saline (9%), moderately saline (45%), highly saline (43%), and salty water (3%). The geochemical properties were further catergorized on the basis of other parameters and ion ratios, such as Ca excess , Na deﬁcit , Na/Cl, Cl/HCO 3 , and Br/Cl, which suggest the inﬂuence of cation exchange, seawater, and marine sediment dissolution. Additionally, stable isotopes indicated two groups. One of these has relatively high salinity and low isotopic content and is affected by the leaching and dissolution of marine deposits. The other group is enriched in δ 18 O and δ D content, with much higher salinity due to mixing with seawater and evaporation. The GQI SWI categorizes groundwater as saline and mixed (55 and 41%, respectively), followed by saltwater (4%), whereas the SMI calculations indicate that about 10% of the groundwater samples are impacted by seawater. Finally, the areal distribution of GQI SWI and SMI identiﬁed some patches along the coastline as well as other inland places located about 12.5 km away from the sea that have undergone saltwater intrusion. In conclusion, overexploitation of groundwater should be avoided because the amount of annual rainfall is very limited.


Introduction
Water demand has increased, especially in coastal areas, with the rapid growth of residential, agricultural, industrial, and tourism activities. Coastal aquifers are considered one of the main sources of freshwater supplies in many countries worldwide, especially in the Mediterranean [1]. The geological heterogeneity as well as spatial and temporal variability in flow patterns play a critical role in governing the distribution of fresh and saline waters in the coastal aquifers [2]. Salinization of groundwater takes place in many coastal aquifers [3][4][5]. It has been found through previous studies that sources and causes of groundwater salinization in coastal aquifers are due to several processes. Among these are water-rock interactions, cation exchange, redox reactions, carbonate and evaporate mineral dissolution, wastewater disposal, and intrusion of fossil seawater and modern seawater induced by groundwater overexploitation [6][7][8][9]. All of these are linked to climate change's effects on precipitation and patterns of recharge as well as to evapotranspiration, which in turn puts pressure on the coastal aquifer systems. This likely speeds up the quantitative and qualitative degradation of groundwater. Therefore, in order to support

Geomorphological and Geological Aspects
Geomorphologically, the landforms within the study area reflect the influences of endo-genetic factors (e.g., faulting, folding, lithologic features, etc.) and exo-genetic factors (e.g., climatic conditions, weathering, deposition, erosion, etc.) [46,47]. These factors result in the emergence of various landforms, such as tablelands, ridges, depressions, and dunes as well as drainage lines, all of which are affected by the distribution of surface runoff in addition to the accumulation and storage of groundwater. Three main geomorphologic units were identified: the southern tableland, the piedmont plain, and the coastal plain [27] (Figure 2). The southern tableland extends southwards to the Qattara Depression, with a maximum elevation of 250 m above sea level. The northern bounding slopes of this tableland are usually dissected by the drainage lines that discharge to the coastal plain. The piedmont plain, with low land and hills, is a transitional zone between the tableland to the south and the coastal plain to the north. Its elevation ranges from 30 to 90 m above sea level, while it varies between 2 and 25 kilometers wide. The inland depressions can be seen within this plain in between the ridges [48]. The coastal plain occupies a

Geomorphological and Geological Aspects
Geomorphologically, the landforms within the study area reflect the influences of endo-genetic factors (e.g., faulting, folding, lithologic features, etc.) and exo-genetic factors (e.g., climatic conditions, weathering, deposition, erosion, etc.) [46,47]. These factors result in the emergence of various landforms, such as tablelands, ridges, depressions, and dunes as well as drainage lines, all of which are affected by the distribution of surface runoff in addition to the accumulation and storage of groundwater. Three main geomorphologic units were identified: the southern tableland, the piedmont plain, and the coastal plain [27] ( Figure 2). The southern tableland extends southwards to the Qattara Depression, with a maximum elevation of 250 m above sea level. The northern bounding slopes of this tableland are usually dissected by the drainage lines that discharge to the coastal plain. The piedmont plain, with low land and hills, is a transitional zone between the tableland to the south and the coastal plain to the north. Its elevation ranges from 30 to 90 m above sea level, while it varies between 2 and 25 kilometers wide. The inland depressions can be seen within this plain in between the ridges [48]. The coastal plain occupies a narrow strip of land that extends parallel to the Mediterranean Sea, with elevations ranging from 0 to 50 meters above sea level. The existence of alternating low-lying ridges separated by narrow depressions along the coast reflects the influence of lithologic and structural conditions as well as the fluctuation of sea level [23]. narrow strip of land that extends parallel to the Mediterranean Sea, with elevations ranging from 0 to 50 meters above sea level. The existence of alternating low-lying ridges separated by narrow depressions along the coast reflects the influence of lithologic and structural conditions as well as the fluctuation of sea level [23]. Geologically, the study area, which is a part of the northwestern Mediterranean coast, is mainly underlain by sedimentary rocks ranging in age from Tertiary to Quaternary [23]. The Miocene deposits are represented by Moghra and Marmarica Formations. The Moghra Formation is composed of argillaceous limestone intercalated with sand and shale related to fluviatile to fluviomarine deposits, while the overlying shallow marine rocks belong to the Marmarica Formation. It is formed of fracture white limestone and a lower grey calcarenite interbedded with clay lenses. The overlaying Quaternary deposits are exposed in the study area. They are formed by a thin cover of drift sands and loamy deposits covering mainly low-lying areas and the floors of narrow valleys dissecting the tableland ( Figure 3). The study area's climatic conditions are typically arid, with a long, hot, dry summer, a mild winter with little rainfall, high evaporation, and moderate to high relative humidity, with a mean annual rainfall of 155 mm [49]. Hydrologically, the Quaternary and Miocene aquifers are considered as the two main water-bearing formations in the area under consideration. The Quaternary carbonate aquifer comprises wadi fill and oolitic limestone deposits, with a thickness ranging from 10 m to about 40 m [50,51]. It is mainly recharged from the infiltration of the precipitation Geologically, the study area, which is a part of the northwestern Mediterranean coast, is mainly underlain by sedimentary rocks ranging in age from Tertiary to Quaternary [23]. The Miocene deposits are represented by Moghra and Marmarica Formations. The Moghra Formation is composed of argillaceous limestone intercalated with sand and shale related to fluviatile to fluviomarine deposits, while the overlying shallow marine rocks belong to the Marmarica Formation. It is formed of fracture white limestone and a lower grey calcarenite interbedded with clay lenses. The overlaying Quaternary deposits are exposed in the study area. They are formed by a thin cover of drift sands and loamy deposits covering mainly low-lying areas and the floors of narrow valleys dissecting the tableland (Figure 3). The study area's climatic conditions are typically arid, with a long, hot, dry summer, a mild winter with little rainfall, high evaporation, and moderate to high relative humidity, with a mean annual rainfall of 155 mm [49].
narrow strip of land that extends parallel to the Mediterranean Sea, with elevations ranging from 0 to 50 meters above sea level. The existence of alternating low-lying ridges separated by narrow depressions along the coast reflects the influence of lithologic and structural conditions as well as the fluctuation of sea level [23]. Geologically, the study area, which is a part of the northwestern Mediterranean coast, is mainly underlain by sedimentary rocks ranging in age from Tertiary to Quaternary [23]. The Miocene deposits are represented by Moghra and Marmarica Formations. The Moghra Formation is composed of argillaceous limestone intercalated with sand and shale related to fluviatile to fluviomarine deposits, while the overlying shallow marine rocks belong to the Marmarica Formation. It is formed of fracture white limestone and a lower grey calcarenite interbedded with clay lenses. The overlaying Quaternary deposits are exposed in the study area. They are formed by a thin cover of drift sands and loamy deposits covering mainly low-lying areas and the floors of narrow valleys dissecting the tableland ( Figure 3). The study area's climatic conditions are typically arid, with a long, hot, dry summer, a mild winter with little rainfall, high evaporation, and moderate to high relative humidity, with a mean annual rainfall of 155 mm [49]. Hydrologically, the Quaternary and Miocene aquifers are considered as the two main water-bearing formations in the area under consideration. The Quaternary carbonate aquifer comprises wadi fill and oolitic limestone deposits, with a thickness ranging from 10 m to about 40 m [50,51]. It is mainly recharged from the infiltration of the precipitation Hydrologically, the Quaternary and Miocene aquifers are considered as the two main water-bearing formations in the area under consideration. The Quaternary carbonate aquifer comprises wadi fill and oolitic limestone deposits, with a thickness ranging from 10 m to about 40 m [50,51]. It is mainly recharged from the infiltration of the precipitation over its outcropping rocks [50]. The Quaternary aquifer is directly connected to the Mediterranean Sea, which has a significant impact on groundwater salinity [51]. Groundwater, in this aquifer, exists under unconfined conditions and is represented by six groundwater samples. The depth to water ranges from 3.15 m (well 6) to 9.30 m (well 2), while the water level ranges between 2 m below sea level (well 1) and +4.7 m (well 3). The Miocene limestone aquifer of middle Miocene age (Marmarica Formation) is the main water-bearing formation in the study area. It is mainly composed of successions of chalky, marly, argillaceous, and dolomitic limestones interbedded with clay lenses. The groundwater in this aquifer exists under two conditions: perched groundwater, where the water level is above mean sea level, and main water table, where the water is free or semi-confined. The perched water table is recharged indirectly by the overlying fractured rocks, while the fractured free to semi-confined aquifer is indirectly recharged by natural groundwater movement from south to north [51]. The degree of recharge depends on the nature of the fault from which this aquifer derives its water. The depth to water ranges from 16 m (well 11) to 72 m (well 41), whereas the water level ranges between 5.3 m below sea level (well 12) and +8 m (well 33). Despite the fact that there is a general relationship between water level and salinity because salinity frequently rises with the direction of water flow, Figure 4 shows that the relationship between the two variables is not clearly defined in each of the Quaternary and Miocene aquifers. This characteristic could mean that other hydrological factors, rather than the absolute water level, have a significant impact on the salinity of the water, as is the case in some localities within the study area, which showed perched aquifer conditions. In general, the figure shows a general direction of groundwater flow from southeast to northwest.
over its outcropping rocks [50]. The Quaternary aquifer is directly connected to the Mediterranean Sea, which has a significant impact on groundwater salinity [51]. Groundwater, in this aquifer, exists under unconfined conditions and is represented by six groundwater samples. The depth to water ranges from 3.15 m (well 6) to 9.30 m (well 2), while the water level ranges between 2 m below sea level (well 1) and +4.7 m (well 3). The Miocene limestone aquifer of middle Miocene age (Marmarica Formation) is the main water-bearing formation in the study area. It is mainly composed of successions of chalky, marly, argillaceous, and dolomitic limestones interbedded with clay lenses. The groundwater in this aquifer exists under two conditions: perched groundwater, where the water level is above mean sea level, and main water table, where the water is free or semi-confined. The perched water table is recharged indirectly by the overlying fractured rocks, while the fractured free to semi-confined aquifer is indirectly recharged by natural groundwater movement from south to north [51]. The degree of recharge depends on the nature of the fault from which this aquifer derives its water. The depth to water ranges from 16 m (well 11) to 72 m (well 41), whereas the water level ranges between 5.3 m below sea level (well 12) and +8 m (well 33). Despite the fact that there is a general relationship between water level and salinity because salinity frequently rises with the direction of water flow, Figure  4 shows that the relationship between the two variables is not clearly defined in each of the Quaternary and Miocene aquifers. This characteristic could mean that other hydrological factors, rather than the absolute water level, have a significant impact on the salinity of the water, as is the case in some localities within the study area, which showed perched aquifer conditions. In general, the figure shows a general direction of groundwater flow from southeast to northwest. According to Morad et al. [51], the maximum discharge from the Quaternary and Miocene aquifers in the Barrani area is 52,200 m 3 /day and 135,000 m 3 /day, respectively. A hydrogeological cross-section extending from the southeast to the northwest towards the city of Baqbaq, with a distance of about 50 km ( Figure 5), indicates that groundwater recharge is possible, with precipitation on the tablelands to the south and flowing to the Mediterranean Sea to the north. Because the karst aquifer's rocks are mostly fractured, According to Morad et al. [51], the maximum discharge from the Quaternary and Miocene aquifers in the Barrani area is 52,200 m 3 /day and 135,000 m 3 /day, respectively. A hydrogeological cross-section extending from the southeast to the northwest towards the city of Baqbaq, with a distance of about 50 km ( Figure 5), indicates that groundwater recharge is possible, with precipitation on the tablelands to the south and flowing to the Mediterranean Sea to the north. Because the karst aquifer's rocks are mostly fractured, recharge is possible and increasing along the drainage lines, where these drainage lines act as water collectors [52]. recharge is possible and increasing along the drainage lines, where these drainage lines act as water collectors [52].

Methodology
Fifty-eight groundwater samples were collected, throughout 2021, from all the available water points. These water points represent shallow hand dug and deep wells. The Quaternary aquifer was represented mostly by hand dug wells (6 samples), while the Miocene aquifer was primarily represented by drilled wells (52 samples). All the studied groundwater wells are private, with the exception of three governmental wells (12,23,44), which have desalination plants installed. Additionally, two water samples representing the local rainfall and the Mediterranean Sea were collected during the winter of 2020 to be used as reference waters. The rainwater sample (59) was collected from the rain gauge station located in the support center of the Desert Research Center in the Sidi Barani area, while the seawater sample (60) was collected from the Mediterranean Sea. The field activities included depth to water measurements and geographic locations of the wells, in addition to measuring the hydrogen ion activity (pH) and electrical conductivity (EC). During the field trip, only one water sample was collected from each well in order to perform the necessary chemical analyses. Another set of water samples was collected from selected wells for the purpose of conducting stable isotope analyses. The collected water samples were analyzed for major and minor ion concentrations (such as Ca, Mg, Na, CO3, HCO3, SO4, Cl, SiO2, and Br), in addition to environmental stable isotopes including δ 18 O and δ 2 H. The concentrations of both CO3 and HCO3 were measured by the titration method using 0.01 normal H2SO4. The concentrations of the other major and minor ions were detected using ion chromatography (Dionex, ICS-1100, Thermo Fisher Scientific Inc., Waltham, MA, USA). All of the aforementioned analyses were conducted according to the methods adopted by Rainwater and Thatcher, Fishman and Friedman, and the American Society for Testing and Materials [53][54][55] at the Laboratory of the Desert Research Center, Egypt. The results of all of the analyzed water samples were within the acceptable error limit (±5).
The concentration of stable isotopes for some selected water samples were estimated according to the method described by Coplen et al. and Coplen [56,57] at the Center for Stable Isotopes at the University of New Mexico, USA. The results were reported in delta

Methodology
Fifty-eight groundwater samples were collected, throughout 2021, from all the available water points. These water points represent shallow hand dug and deep wells. The Quaternary aquifer was represented mostly by hand dug wells (6 samples), while the Miocene aquifer was primarily represented by drilled wells (52 samples). All the studied groundwater wells are private, with the exception of three governmental wells (12,23,44), which have desalination plants installed. Additionally, two water samples representing the local rainfall and the Mediterranean Sea were collected during the winter of 2020 to be used as reference waters. The rainwater sample (59) was collected from the rain gauge station located in the support center of the Desert Research Center in the Sidi Barani area, while the seawater sample (60) was collected from the Mediterranean Sea. The field activities included depth to water measurements and geographic locations of the wells, in addition to measuring the hydrogen ion activity (pH) and electrical conductivity (EC). During the field trip, only one water sample was collected from each well in order to perform the necessary chemical analyses. Another set of water samples was collected from selected wells for the purpose of conducting stable isotope analyses. The collected water samples were analyzed for major and minor ion concentrations (such as Ca, Mg, Na, CO 3 , HCO 3 , SO 4 , Cl, SiO 2 , and Br), in addition to environmental stable isotopes including δ 18 O and δ 2 H. The concentrations of both CO 3 and HCO 3 were measured by the titration method using 0.01 normal H 2 SO 4 . The concentrations of the other major and minor ions were detected using ion chromatography (Dionex, ICS-1100, Thermo Fisher Scientific Inc., Waltham, MA, USA). All of the aforementioned analyses were conducted according to the methods adopted by Rainwater and Thatcher, Fishman and Friedman, and the American Society for Testing and Materials [53][54][55] at the Laboratory of the Desert Research Center, Egypt. The results of all of the analyzed water samples were within the acceptable error limit (±5).
The concentration of stable isotopes for some selected water samples were estimated according to the method described by Coplen et al. and Coplen [56,57] at the Center for Stable Isotopes at the University of New Mexico, USA. The results were reported in delta per mille (‰) notation for O-18 and deuterium ( 2 H) to identify the recharge sources as well as the factors affecting groundwater quality change. The assessment of the different hydrogeochemical and mixing processes were studied using multiple techniques, including the groundwater quality index for seawater intrusion (GQI SWI ), hydrochemical

Gibbs Diagram
This diagram, a simple plot of the total dissolved solids (TDS) versus the weight ratio of Na/(Na + Ca) or Cl/(Cl + HCO 3 ), is used to identify the relationships between the chemical water composition and the types of the rocks in which the water circulates [58]. Three distinct fields, including precipitation dominance, evaporation dominance, and rock weathering dominance, constitute the segments in the Gibbs diagram.

Hydrochemical Facies Evolution Diagram (HFE-D)
This diagram was suggested by Gimenez Forcada [59] as an alternative to other hydrochemical diagrams such as the piper diagram, to clearly identify the hydrogeochemical changes in groundwater during recharge and saltwater intrusion processes. The importance of this diagram is that it helps in the identification of the salinization-freshening phases in coastal aquifers. In addition, it contributes to a clearer identification of possible groups of samples and their evolution trends in the aquifer [60,61]. In this diagram, four main facies are defined, namely NaCl, CaCl 2 , NaHCO 3 , and CaHCO 3 . Each group is divided in turn into four other subfacies, resulting in sixteen subdivisions. These subdivisions are the following: 1. Na-HCO 3 [62]. The facies in this diagram are determined by the percentage of Ca and Na cations, as well as by HCO 3 (or SO 4 ) and Cl − anions, in relation to the sum of cations and anions. The term Mix is used for the facies names to indicate that the percentage of the cations or anions is less than 50% but greater than the percentage of any of the other cations and anions considered. For example, if a sample is plotted in subdivision 6, it belongs to MixNa-MixHCO 3 /MixSO 4 , which means none of the Na, HCO 3 , and SO 4 ions reached 50%, yet Na% remains greater than Ca and Mg, and similarly, HCO 3 and SO 4 will be greater than Cl. In addition to the mixing processes, the diagram also explains both the direct and reverse exchange reactions.

Correlation Coefficients (Ion-Ion Relationships)
Changes in the chemical composition of groundwater are affected by many factors, such as rock-water interaction, mixing, and ion-exchange processes. Some parameters and ion ratios calculated in meq/L (such as Ca excess , Na deficit , Na/Cl, Mg/Ca, SO 4 /Cl, and Br/Cl) are addressed. They were correlated with each other as well as with the TDS in bivariate diagrams in order to distinguish between the mixing mechanisms of fresh water and saline water from other chemical reactions, such as rock weathering and ion exchange.

Principal Component Analysis (PCA)
Principal component analysis (PCA) is a data transformation technique that attempts to visualize a simple underlying structure that is assumed to exist within a multivariate data set [63]. It quantifies the relationship between different variables by computing the correlation matrix for the entire data set. As a result, the data set is easily summarized without losing much information [64]. When using this technique, data sets are first standardized, and then the correlation matrix is created. The eigenvalues and factor loadings for the correlation matrix are determined and a scree plot is drawn. The extraction factors are based on the variances and co-variances of the variables. The eigenvalues and eigenvectors are evaluated, which represent the amount of variance explained by each factor. An eigenvalue greater than 1 is set as a criterion to extract factors [65]. Finally, by the process of rotation, the loading of each variable on one of the extracted factors is maximized, and the loadings of all the other factors are minimized. These factor loadings are useful in grouping the water quality parameters and providing information for interpreting the data. This study considered pH, EC, Na, K Ca, Mg, Cl, CO 3 , HCO 3 , and SO 4 as water quality parameters. Statistical Package for the Social Sciences (SPSS) version 16 was used for the statistical analysis.

Groundwater Quality Index for Seawater Intrusion (GQI SWI )
The GQI SWI index was developed by Tomaszkiewicz et al. [66] as another tool, calculated according to Piper [67], to interpret groundwater quality change (in terms of seawater salinization). Equation (1) is developed from Equations (2)-(4): where GQI Piper (mix) is the freshwater seawater mixing index of the piper diagram; GQI fsea is seawater fraction index.
The seawater fraction (f sea ) is calculated according to the following equation: Chloride ion is used as a conservative parameter, where m Cl sea − m Cl f resh is equal to about 35‰ (grams of salt per kilogram) seawater, based on the concentration of Cl − in mmol/L [66,68]. GQISW is a numerical indicator used to interpret data from the Piper diagram as well as seawater intrusion. It ranges from 0 to 100, with 100 representing freshwater and 0 representing seawater.

Seawater Mixing Index (SMI)
The seawater mixing index (SMI) is applied to calculate the degree of seawater mixing [69]. Major ion compositions, including (Cl, SO 4 , Na and Mg), are used to calculate this index according to the following equation: The constants a, b, c, and d are estimated based on the relative concentration proportion of Na + , Cl − , Mg 2+ , and SO 4 2− in seawater and recorded as 0.31, 0.04, 0.57, and 0.08, respectively. C is the measured ion concentration in mg/L. T represents the values of threshold of the selected ions, which can be estimated from the cumulative probability curves by determining the inflection points for each ion. The water is said to be impacted by seawater mixing if the SMI is more than 1 [70].

Results and Discussion
Groundwater characterization mainly depends on its chemical constituents. The concentrations of these constituents are mostly a result of the interaction between groundwater and its host rocks. In this part, different aspects are addressed, including hydrochemical characterization, salinization process identification, and mixing ratio quantification.

Hydrochemistry and Its Prevailing Factors
The physical, chemical, and isotopic characteristics of groundwater samples from the shallow Quaternary (samples 1 to 6) and deeper Miocene (samples 7 to 58) wells are presented in Table 1 and Figure 6. The total dissolved solids (TDS) displayed high variance, ranging from 1113 mg/L to 51,975 mg/L (Avg. 10,290 mg/L). The high values and extreme outliers in TDS values are probably due to salt water. The major cations followed the order Na (260-15,000 mg/L, Avg., 2738 mg/L) > Mg (46-2292 mg/L, Avg., 500 mg/L), or Ca (47-1887 mg/L, Avg. 423 mg/L), while the major anions are of the order Cl (406-29,200 mg/L, Avg., 4786 mg/L) > SO 4 (200-4300 mg/L, Avg., 1778 mg/L) > HCO 3 (49-460 mg/L, Avg., 174), indicating that groundwater is dominated by two hydrochemical types: Na-Cl and Na-SO 4 . Groundwater samples collected from the Quaternary and Miocene aquifers were classified into four groups according to USGS [71]. Group A, which is classified as slightly saline water (5 samples, and its percentage is 9%), ranged from 1000 to less than 3000 mg/L. Group B (26 samples, about 45%), with salinity ranging from 3000 to 10,000 mg/L, is classified as moderately saline water. Group C (25 samples of about 43%), whose salinity ranged from 10,000 to 35,000 mg/L, is classified as highly saline water. Finally, group D, with a very high salinity (more than 35,000 mg/L), is represented by two samples of about 3%. The majority of the studied groundwater samples, which mainly belong to the Miocene aquifer, are represented by Group B and C, while almost all the Quaternary samples belong to Group D. The hydrogen ion activity (pH) values reveal that groundwaters are neutral to slightly alkaline, with a mean pH value of 7.18. The water type shown in Table 1 is mainly characterized as a Cl-Na type, except two samples are of Cl-SO 4 -Na type, indicating advanced evolutionary stages of groundwater.
Groundwater characterization mainly depends on its chemical constituents. The concentrations of these constituents are mostly a result of the interaction between groundwater and its host rocks. In this part, different aspects are addressed, including hydrochemical characterization, salinization process identification, and mixing ratio quantification.

Hydrochemistry and Its Prevailing Factors
The physical, chemical, and isotopic characteristics of groundwater samples from the shallow Quaternary (samples 1 to 6) and deeper Miocene (samples 7 to 58) wells are presented in Table 1 and Figure 6. The total dissolved solids (TDS) displayed high variance, ranging from 1113 mg/L to 51,975 mg/L (Avg. 10,290 mg/L). The high values and extreme outliers in TDS values are probably due to salt water. The major cations followed the order Na (260-15,000 mg/L, Avg., 2738 mg/L) > Mg (46-2292 mg/L, Avg., 500 mg/L), or Ca (47-1887 mg/L, Avg. 423 mg/L), while the major anions are of the order Cl (406-29,200 mg/L, Avg., 4786 mg/L) > SO4 (200-4300 mg/L, Avg., 1778 mg/L) > HCO3 (49-460 mg/L, Avg., 174), indicating that groundwater is dominated by two hydrochemical types: Na-Cl and Na-SO4. Groundwater samples collected from the Quaternary and Miocene aquifers were classified into four groups according to USGS [71]. Group A, which is classified as slightly saline water (5 samples, and its percentage is 9%), ranged from 1000 to less than 3000 mg/L. Group B (26 samples, about 45%), with salinity ranging from 3000 to 10,000 mg/L, is classified as moderately saline water. Group C (25 samples of about 43%), whose salinity ranged from 10,000 to 35,000 mg/L, is classified as highly saline water. Finally, group D, with a very high salinity (more than 35,000 mg/L), is represented by two samples of about 3%. The majority of the studied groundwater samples, which mainly belong to the Miocene aquifer, are represented by Group B and C, while almost all the Quaternary samples belong to Group D. The hydrogen ion activity (pH) values reveal that groundwaters are neutral to slightly alkaline, with a mean pH value of 7.18. The water type shown in Table 1 is mainly characterized as a Cl-Na type, except two samples are of Cl-SO4-Na type, indicating advanced evolutionary stages of groundwater.

Groundwater Salinization and Mixing Processes
Different geochemical mechanisms, including rock-water interaction, evaporation, and precipitation, were identified when water samples were plotted on Gibbs diagrams (Figure 7). It is clearly shown on these diagrams that the water chemistry of almost all the Quaternary and Miocene groundwater samples is mainly controlled by evaporationprecipitation processes and seawater intrusion, plotted in the upper right corner of the diagram. It is also apparent that most of the samples with Na/(Na + Ca) or Cl/(Cl + HCO 3 ) ratios greater than 0.6 and TDS values greater than 5000 mg/L were mainly affected by saline water mixing or evaporation. This plot showed also that few samples were located in the rock dominance region, indicating that rock-water interaction has little impact on groundwater chemistry.

Groundwater Salinization and Mixing Processes
Different geochemical mechanisms, including rock-water interaction, evaporation, and precipitation, were identified when water samples were plotted on Gibbs diagrams (Figure 7). It is clearly shown on these diagrams that the water chemistry of almost all the Quaternary and Miocene groundwater samples is mainly controlled by evaporation-precipitation processes and seawater intrusion, plotted in the upper right corner of the diagram. It is also apparent that most of the samples with Na/(Na + Ca) or Cl/(Cl + HCO3) ratios greater than 0.6 and TDS values greater than 5000 mg/L were mainly affected by saline water mixing or evaporation. This plot showed also that few samples were located in the rock dominance region, indicating that rock-water interaction has little impact on groundwater chemistry. Groundwater samples were plotted in the Hydrochemical Facies Evolution Diagram (HFE-D); seawater and freshwater were connected through a mixing line (Figure 8). It is shown that the dominant water facies type is Na-Cl, where almost all samples were plotted in subdivision 4, and only two samples are of Mix/Na-Cl, located in subdivision 8. This means that Na ions exceed 50% of the total percent of cations as is also the case with respect to Cl ions. Samples can be divided into two groups. The first group is located below and to the right of the seawater/freshwater mixing line, indicating seawater/saltwater intrusion. The other group is situated down to the left of the line. While it still has the same facies, it is somewhat affected by recharge from fresh water. This diagram also indicates that a group of samples were subjected to direct cation exchange, with Na ions being released to the solution, while other samples were affected by an inverse cation exchange reaction. Delineation of the salinization processes in the study area will be more clearly elucidated in the following sections. Groundwater samples were plotted in the Hydrochemical Facies Evolution Diagram (HFE-D); seawater and freshwater were connected through a mixing line (Figure 8). It is shown that the dominant water facies type is Na-Cl, where almost all samples were plotted in subdivision 4, and only two samples are of Mix/Na-Cl, located in subdivision 8. This means that Na ions exceed 50% of the total percent of cations as is also the case with respect to Cl ions. Samples can be divided into two groups. The first group is located below and to the right of the seawater/freshwater mixing line, indicating seawater/saltwater intrusion. The other group is situated down to the left of the line. While it still has the same facies, it is somewhat affected by recharge from fresh water. This diagram also indicates that a group of samples were subjected to direct cation exchange, with Na ions being released to the solution, while other samples were affected by an inverse cation exchange reaction. Delineation of the salinization processes in the study area will be more clearly elucidated in the following sections. Sustainability 2022, 14, x FOR PEER REVIEW 13 of 24

Ion Ratios and Correlation Coefficients
Ionic chemical ratios are useful for the identification of different hydrochemical processes affecting water quality as well as the impact of seawater intrusion on groundwater chemistry [72,73]. The ion ratios selected are summarized in Table 2. The Na/Cl ratio is less than unity in the majority of the Miocene samples and in half of the Quaternary samples, while the rest of the samples showed values greater than unity. The relationship between Cl and Na/Cl ( Figure 9A) showed that many samples plot on the 1:1 line, indicating the role of halite dissolution as a major process contributing to groundwater salinization [68]. A few samples showed slight increases in sodium relative to chloride, reflecting the effect of recharge from rainwater and/or ion exchange. The majority of samples plot slightly below that line, with a higher chloride content compared to sodium, indicating the probability of seawater intrusion.

Ion Ratios and Correlation Coefficients
Ionic chemical ratios are useful for the identification of different hydrochemical processes affecting water quality as well as the impact of seawater intrusion on groundwater chemistry [72,73]. The ion ratios selected are summarized in Table 2. The Na/Cl ratio is less than unity in the majority of the Miocene samples and in half of the Quaternary samples, while the rest of the samples showed values greater than unity. The relationship between Cl and Na/Cl ( Figure 9A) showed that many samples plot on the 1:1 line, indicating the role of halite dissolution as a major process contributing to groundwater salinization [68]. A few samples showed slight increases in sodium relative to chloride, reflecting the effect of recharge from rainwater and/or ion exchange. The majority of samples plot slightly below that line, with a higher chloride content compared to sodium, indicating the probability of seawater intrusion. Sustainability 2022, 14, x FOR PEER REVIEW 14 of 24 The Mg/Ca ratio is used to interpret the impact of the aquifer matrix dominated by carbonate rocks as well as the effect of seawater on groundwater; it tends to be 1 in freshwater and approaches 5 in seawater [74,75]. In Figure 9B, there are three groups. The first is a group with slightly lower salinity (less than or equal 5000 mg/L) and a higher value of Mg/Ca, reflecting cation exchange and recharge from freshwater. The second group, which represents the majority of samples, has higher values of TDS as well as Mg/Ca. This group is mainly affected by marine rock-water interaction and seawater intrusion processes. The third group, with lower Mg/Ca and relatively higher Cl content, indicates a long residence time in carbonate rocks and recharge from fresh water. The SO4/Cl ratio is considered as evidence of the influence of marine deposits as well as seawater on groundwater, with values below unity in almost all samples ( Figure 9C). As Br and Cl are relatively conservative in hydrological systems, the Br/Cl ratio is used to distinguish whether groundwater salinity is of marine or non-marine origins [76]. As shown in both Table 2 and Figure 9D, the Br/Cl ratios of most of the studied groundwater samples are within the The Mg/Ca ratio is used to interpret the impact of the aquifer matrix dominated by carbonate rocks as well as the effect of seawater on groundwater; it tends to be 1 in freshwater and approaches 5 in seawater [74,75]. In Figure 9B, there are three groups. The first is a group with slightly lower salinity (less than or equal 5000 mg/L) and a higher value of Mg/Ca, reflecting cation exchange and recharge from freshwater. The second group, which represents the majority of samples, has higher values of TDS as well as Mg/Ca. This group is mainly affected by marine rock-water interaction and seawater intrusion processes. The third group, with lower Mg/Ca and relatively higher Cl content, indicates a long residence time in carbonate rocks and recharge from fresh water. The SO 4 /Cl ratio is considered as evidence of the influence of marine deposits as well as seawater on groundwater, with values below unity in almost all samples ( Figure 9C). As Br and Cl are relatively conservative in hydrological systems, the Br/Cl ratio is used to distinguish whether groundwater salinity is of marine or non-marine origins [76]. As shown in both Table 2 and Figure 9D, the Br/Cl ratios of most of the studied groundwater samples are within the normal seawater range (0.0040-0.0070) as salinity increases, with a mean value of about 0.006, indicating that the salinization of these groundwaters resulted from dissolution of marine sediments and mixing with seawater.
The slight decrease of the Br/Cl ratio with higher salinity is likely due to interaction with rocks rich in evaporites. On the other hand, samples with relatively low salinity but with a higher Br/Cl ratio suggest the impact of anthropogenic activities on groundwaters [77]. This could be explained as a result of the impact of the on-site wastewater disposal on groundwater, as some residents resort to using abandoned wells for sewage in the area under consideration. The relationship between Na deficit and Ca excess in basinal fluids is presented according to Davisson et al. and Davisson and Criss [78,79] using two parameters: Ca excess = 2[Ca meas − (Ca/Cl) SW × Cl meas ]/40.08 Na deficit = [(Na/Cl) SW × Cl meas − Na meas ]/22.99 A straight Na deficit Ca excess regression line with a 1:1 slope, known as the Basinal Fluid Line (BFL; [79]), is plotted for the majority of basinal brines in the world, showing waterrock interaction with a net cation exchange ratio of 1 Ca for 2 Na ( Figure 9E). This diagram shows the various chemical processes that the groundwater has undergone. Such processes include (i) halite dissolution with a trend that extends from the origin (0,0 point) towards the negative quadrant of the plot; (ii) cation exchange for the samples plotted on the 1:1 line; and (iii) seawater evaporation after calcite precipitation following a nearly horizontal trend, with a slope of 0.123. Further evaporation with a decrease in Na concentration will result in the precipitation of halite, as indicated by the samples plotted slightly above the seawater sample.

Saturation Indices
PHREEQC version 3 [80] calculated saturation indices by comparing the chemical activities of the mineral's dissolved ions (ion activity product, IAP) with their solubility product (Ksp). The saturation indices (SI) for selected minerals, including anhydrite, calcite, dolomite, gypsum, and halite, were calculated to confirm the precipitation and dissolution of these minerals. They have negative values when the minerals tend to dissolve, positive values when the minerals tend to precipitate, and zero values when the water and minerals are in chemical equilibrium. The minerals chosen were based on the major ions encountered in the study area's groundwater. Figure 10 depicts a bivariate diagram of SI values for anhydrite, calcite, dolomite, gypsum, and halite against the chloride concentration of the studied groundwater samples. It shows a positive correlation between Cl concentration and SI values. The plot shows that groundwater samples are undersaturated with respect to calcite and dolomite, with 78% and 40%, respectively. All samples exhibited undersaturation in relation to anhydrite, gypsum, and halite. The positive values of the calculated SI with respect to calcite and dolomite were about 22% and 60%, respectively, and the negative values of SI for all groundwater samples with respect to anhydrite and gypsum suggest that these salts have a major role in the enrichment of groundwater with calcium and sulphate.
This figure also shows that there are three groups; group I, with intermediate Cl concentration; group II, with a higher Cl concentration; group III, with much higher Cl concentration, which reached or exceeded the seawater values (as in samples 1 and 11). Group I is plotted near the midway between the recharge water (rainwater) and the seawater, indicating the leaching and dissolution of marine sediments. gp II is closely plotted next to seawater, reflecting mixing with seawater as a result of groundwater overexploitation and scarcity of recharge from precipitation. Finally, gp III includes seawater and samples subjected to the dissolution of salt marches rich in evaporates. This figure also shows that there are three groups; group I, with intermediate Cl concentration; group II, with a higher Cl concentration; group III, with much higher Cl concentration, which reached or exceeded the seawater values (as in samples 1 and 11). Group I is plotted near the midway between the recharge water (rainwater) and the seawater, indicating the leaching and dissolution of marine sediments. gp II is closely plotted next to seawater, reflecting mixing with seawater as a result of groundwater overexploitation and scarcity of recharge from precipitation. Finally, gp III includes seawater and samples subjected to the dissolution of salt marches rich in evaporates.

Stable Isotopes
Environmental stable isotopes such as Oxygen-18 and deuterium are good indicators for tracing the recharge sources and the origin of groundwaters, as they do not participate in geochemical reactions except in geothermal systems. Isotope exchange usually takes place within geothermal water systems only at elevated temperatures (>250 °C) [81]. The stable isotopic compositions of the collected water samples are listed in Table 1

Stable Isotopes
Environmental stable isotopes such as Oxygen-18 and deuterium are good indicators for tracing the recharge sources and the origin of groundwaters, as they do not participate in geochemical reactions except in geothermal systems. Isotope exchange usually takes place within geothermal water systems only at elevated temperatures (>250 • C) [81]. The stable isotopic compositions of the collected water samples are listed in Table 1 The distinctive trend expressing the mixing of groundwater with more evaporated water (seawater) is represented by the mixing line (ML) (Figure 11a). This line is described by the equation δD = 5.45 δ 18 O − 0.46‰, which is similar to the equation proposed by Olive [84]. It is shown that almost all groundwater samples were situated on that line and can The distinctive trend expressing the mixing of groundwater with more evaporated water (seawater) is represented by the mixing line (ML) (Figure 11a). This line is described by the equation δD = 5.45 δ 18 O − 0.46‰, which is similar to the equation proposed by Olive [84]. It is shown that almost all groundwater samples were situated on that line and can be divided into two groups. First is the group with depleted isotope values, indicating that it is of meteoric origin. The other group, with more enriched in δ 18 O and δD values, is assumed to have been exposed to mixing with seawater and/or the fractionation during evaporation from the shallow groundwater aquifers. The purpose of using the relationship between TDS and δ 18 O is to distinguish between samples that were exposed to evaporation from those that were subjected to leaching and dissolution. It is noted from Figure 11b that the majority of groundwater samples, characterized by relatively high salinity, were mainly subjected to leaching and dissolution of marine deposits. The remaining group of samples, of much higher salinity, was affected by dissolution of evaporites (such as halite, anhydrite, and/or gypsum).

Principle Component Analysis (PCA)
Principle component analysis was applied to identify relationships between different variables, including pH, TDS, Ca, Mg, Na, K, CO 3 , HCO 3 , SO 4 , Cl, δ 18 O, δD, SMI, and GQI SWI , and to extract different factors affecting groundwater quality (Tables 3 and 4). The correlations established between the selected parameters (Table 3) show that there is a strong correlation between TDS and the concentrations of the variables Ca, Mg, Na, SO 4 , Cl, Br, δ 18 O, δD, and SMI (r 2 ≥ 0.7), indicating that such variables are enriched due to mineralization processes. The high correlation between TDS and both δ 18 O and δD is evidence of the exposure of groundwater to evapoconcentration processes. Although the aquifer matrix is dominated by carbonate rocks, bicarbonates are not correlated with calcium, indicating a source other than calcite dissolution. Meanwhile, calcium is well correlated with SO 4 (r 2 ≥ 0.53), suggesting gypsum dissolution. The factor loadings of the correlation matrix were defined based on eigenvalues that exceeded unity [65]. Three factors, with a total cumulative variance of about 80%, were assigned to explain changes made to groundwater (Table 4). Such factors are presented in two biplots to clearly illustrate their correlation with the different variables ( Figure 12). The first factor f 1 (Figure 12a), with about 60% of the total variance, has high positive loading with the variables TDS, Ca, Mg, Na, Cl, and SO 4 , δ 18 O, δD, and SMI and also has strong negative loading with GQI SWI . This strong correlation can be interpreted as a result of different mechanisms. Among these are (1) rock-water interaction, (2) cation exchange, and (3) seawater intrusion. The second factor f 2 (Figure 12a) accounts for 11% of the total variance, which has a moderate positive loading with SO 4 , δ 18 O, and δD as well as high positive loading with K. This suggests evaporation and dissolution of evaporites occurred in the shallower aquifers. The third factor f 3 (Figure 12b), with about 9% of the total variance, is positively correlated with HCO 3 and SiO 2 . This factor represents the freshening and the impact of the weathered silicate minerals on groundwater.

Groundwater Quality Index (GQISWI) and Seawater Mixing Index (SMI)
The GQISWI varies from 0 to 100, and the values above 75 can be considered as freshwater, while when below 50, groundwater is classified as saline or seawater. If groundwater has GQISWI values ranging between 50 and 75, it will be deemed as mixed water. The calculated GQISWI values presented in Table 5 reveal that the study area is dominated by saline and mixed groundwater (55 and 41%, respectively), followed by saltwater (4%).

Groundwater Quality Index (GQI SWI ) and Seawater Mixing Index (SMI)
The GQI SWI varies from 0 to 100, and the values above 75 can be considered as freshwater, while when below 50, groundwater is classified as saline or seawater. If groundwater has GQI SWI values ranging between 50 and 75, it will be deemed as mixed water. The calculated GQI SWI values presented in Table 5 reveal that the study area is dominated by saline and mixed groundwater (55 and 41%, respectively), followed by saltwater (4%). The SMI was calculated when the cumulative probability percentages of Cl, SO 4 , Na, and Mg ions were plotted against their concentration in mg/L ( Figure 13). The inflection points of each plot characterize the threshold values (T) of each ion that were substituted in Equation (5) to calculate the SMI values for each groundwater sample. If the SMI value was more than 1, the water may have been influenced by seawater mixing [69,70]. The results of the calculated SMI show that about 90% (52 samples) of the groundwater samples are of meteoric origin, and the remaining 10% (6 samples) are related to saline groundwater. This hypothesis is compatible with the results obtained from the ion ratios and stable isotopes analyses. It should be noted that the variation in SMI and GQI SWI calculated values is due to how each method characterizes groundwater. In other words, the SMI considers seawater as a pollutant, and its percentage in groundwater is calculated, so that if the index value exceeds unity, this indicates seawater intrusion. The GQI index, on the other hand, classifies groundwater based on its origin and takes into account the changes that occurred as a result of various geochemical processes, including intrusion of sea water. The spatial distribution of the groundwater quality index (GQISWI) as well as the seawater mixing index (SMI) were used as complementary tools to delineate the salinization zones in accordance with the extent to which groundwater was affected either by freshwater mixing with seawater or rock-water interaction. According to the spatial distribution maps for GQISWI and SMI ( Figure 14A,B), it was revealed that almost all the ground- The spatial distribution of the groundwater quality index (GQI SWI ) as well as the seawater mixing index (SMI) were used as complementary tools to delineate the salinization zones in accordance with the extent to which groundwater was affected either by freshwater mixing with seawater or rock-water interaction. According to the spatial distribution maps for GQI SWI and SMI ( Figure 14A,B), it was revealed that almost all the groundwater in the study area is mainly of meteoric origin, which in turn reaches a progressive evolutionary stage due to the interaction with marine sediments and mixing with salt water. Patches that have undergone saltwater intrusion were identified in some locations along the shoreline of the Mediterranean Sea. However, saline water was also identified in some other inland locations west of the study area, about 12.5 km away from the sea. Figure 13. Threshold values (T) of (A) Cl, (B) SO4, (C) Na, and (D) Mg in the studied groundwater samples. The intersection of the background concentration (highlighted in blue) and the trend line representing the anomalous concentrations (highlighted in orange) indicates the regional threshold of relevant element.
The spatial distribution of the groundwater quality index (GQISWI) as well as the seawater mixing index (SMI) were used as complementary tools to delineate the salinization zones in accordance with the extent to which groundwater was affected either by freshwater mixing with seawater or rock-water interaction. According to the spatial distribution maps for GQISWI and SMI ( Figure 14A,B), it was revealed that almost all the groundwater in the study area is mainly of meteoric origin, which in turn reaches a progressive evolutionary stage due to the interaction with marine sediments and mixing with salt water. Patches that have undergone saltwater intrusion were identified in some locations along the shoreline of the Mediterranean Sea. However, saline water was also identified in some other inland locations west of the study area, about 12.5 km away from the sea.  Figures (A,B) indicated the areas, which are thought to be affected by seawater intrusion based on GQI SWI , and SMI calculations, respectively.

Conclusions
This paper highlighted the different processes and mechanisms impacting groundwater salinization using multi-hydrochemical parameters and stable isotopes. The groundwater in the study area is classified as brackish, saline, and highly saline water. Different methods, including ion ratio relationships, isotope bivariate plots, evolution diagrams, and statistical analyses, were applied to recognize the main geochemical processes controlling the salinization processes. According to Gibbs classifications, groundwaters are dominated by evaporation followed by rock-water interaction. The HFE-D indicated mixing with saline water and the occurrence of both reverse and direct cation exchange. The correlation of the ion ratios Na/Cl, Mg/Ca, SO 4 /Cl, and Br/Cl with TDS indicates the effect of marine sediments and seawater as well as the anthropogenic activities on some of the studied groundwater samples. Furthermore, the Na deficit and Ca excess bi-plot revealed different processes, including halite dissolution, cation exchange, and seawater evaporation after calcite precipitation. The δ 18 O vs. δD and TDS vs. δ 18 O relationships revealed that some samples were exposed to mixing with seawater and/or fractionation during the evaporation from shallow groundwater aquifers. The factors extracted by PCA suggest the aforementioned processes, in addition to the impact of freshening and silicate weathering on groundwater. The mixing with seawater was confirmed in some locations along the coastline and extended to about 12.5 km away from shore in the Bqabaq area. Additionally, statistical analyses as well as hydrogen and oxygen-18 stable isotope relationships identified the effect of recharge by rainwater in few locations. The findings from this study can be used to enhance our understanding of the different hydrogeochemical processes that led to the deterioration in the groundwater quality. Moreover, it sheds light on the locations that have experienced groundwater quality deterioration as well as other areas that are not or are little affected by seawater intrusion. This will enable the decision makers to take the necessary measures to use groundwater in those areas in a way that ensures its sustainability and does not lead to any further deterioration.