Evaluation of groundwater quality in central Saudi Arabia using hydrogeochemical characteristics and pollution indices

The groundwater quality and heavy metal (HM) contamination were evaluated in palm farms, central Saudi Arabia, using pollution indices, irrigation quality parameters, and multivariate statistical analyses. Thirty groundwater samples were collected in October 2020 for major anions, cations, and HMs analyses and interpretation. The results showed that the average concentrations of total dissolved solids (TDS), Ca+, Na+, K+, Cl−, SO42−, and F− were greater than the permissible limits of the WHO standards for drinking water. The groundwater facies types were Ca–Na–SO4–Cl (23 samples), Ca–Cl–SO4, (4 samples), and Ca–SO4–Cl type (3 samples). The groundwater quality index indicated that 15 groundwater samples were of good quality and 15 were of poor quality, whereas the metal index and heavy metal pollution index indicated that all samples were categorized as slightly affected and with low pollution, respectively. The variation is attributed to the increasing average concentrations of some ions and decreasing HMs. The dissolution/precipitation of silicates, gypsum, and carbonates and soil leaching were the natural factors affecting groundwater chemistry, whereas higher PO43−, NO3−, F−, Pb, and Zn values in some samples may be attributed to human activities from the extensive use of fertilizers and pesticides on the investigated farms.


Introduction
Groundwater is a critical resource for domestic and agricultural needs, especially in arid countries, where the geochemical processes resulting from aquifer-water interaction often control its quality in such minimal rainfall recharge (Alharbi 2018;Alshehri et al. 2021). The accumulation of heavy metals (HMs) in water resources is potentially harmful to human health because of their accumulative characteristics, toxicity, and contamination of food sources. Therefore, the pollution of water resources with HMs has become a global problem (Hardaway et al. 2004;Sarkar et al. 2011;Alshahri and El-Taher 2018). Hydrochemical characteristics including cations, anions, heavy metals, nitrates, chlorides, and organics of the groundwater to determine water quality indices should be used to help sustain the groundwater (El-Sayed et al. 2012;Ledesma-Ruiz et al. 2015;Alghamdi et al. 2020). The chemical characteristics of groundwater, such as pH, dissolved salts and gases, metals, and organics, are controlled by the exchange of cations within the geological aquifer, dissolution of minerals, and evaporation and redox reactions (Matthess 1982;Kumar et al. 2006).
The Kingdom of Saudi Arabia (KSA) is located in an arid environment and has limited renewable water resources (Al-Omran et al. 2016). The KSA depends mainly on groundwater and seawater desalination to cover its need for drinking and irrigation as KSA does not have rivers or freshwater lakes. The recent years have witnessed great focus directed towards expanding agricultural projects based on developing groundwater resources. This stress on groundwater is the main cause of its deterioration (Al-Omran et al. 2016). There is a potential risk of water scarcity in the future due to a rapidly changing climate and anthropogenic activities (Fallatah 2020). The kingdom depends largely on the desalination of seawater and groundwater for different purposes such as drinking, irrigation, and industry (Saud and Abdullah 2009;Alghamdi et al. 2020). The thick Mesozoic and Cenozoic sedimentary rocks function as prolific aquifer for groundwater in central Saudi Arabia (Alharbi and Zaidi 2018). Overdependence on groundwater resources, especially in arid countries, often leads to decreased groundwater levels and deteriorates its quality (Alharbi 2018). The rapid population growth in and extension of agricultural activities around Saudi Arabia have increased the need for freshwater resources (Khanfar 2008;Al-Hammad and Abd El-Salam 2016). Saudi Arabia is the second-largest producer of date palm (Suleman 2014;Salama et al. 2019). The Al Uyaynah-Al Jubailah region ( Fig. 1) is located 40-55 km northwest of Riyadh, inside the narrow, dry riverbed of Wadi Hanifa, which continues southwards through Dir'iyyah and Riyadh.
The Al Uyaynah-Al Jubailah region has many agricultural farms producing several types of dates, leafy green plants (e.g., lettuce, arugula, and radish), vegetables (e.g., pepper, eggplant, tomato, watermelon, onion, and potato), and fruits Fig. 1 Hydrogeological map of the study area including groundwater types, confined groundwater boundary, and sampling numbers (e.g., citrus, orange, pomegranate, lemon, and grapes). As groundwater is an important source of HMs for these plants and can reach humans directly or indirectly, this study evaluates the groundwater quality and HM contamination in some palm farms in central Saudi Arabia in the Al Uyaynah-Al Jubailah region and documents the possible sources of contamination using hydrogeochemical characteristics, pollution indices, and multivariate analyses.

Methods and materials
The present study was conducted in the Al Uyaynah-Al Jubailah region, central Saudi Arabia, at 24°53′14.4″ to 24°54′59.4″ N and 46°20′41.5″ to 46°25′47.2″ E (Fig. 1). Geologically, the carbonates of the Upper Jurassic Hanifa, Jubaila, and Arab formations and Quaternary sediments dominate the study area (Al Husseini and Mathews 2006;Hussein et al. 2012;Youssef and El Sorogy 2015;Tawfik et al. 2016). The Hanifa Formation consists of a lower muddy carbonate unit and an upper stromatoporoid and lagoonal carbonate unit and disconformably overlies the Tuwaiq Mountain Formation (Al Husseini 2009;El-Sorogy and Al-Kahtany 2015;El-Sorogy et al. 2016;Al-Dabbagh and El-Sorogy 2016). The Jubaila Formation consists of moderately deep marine carbonates overlain by a shallow marine stromatoporoid-associated assemblage and disconformably lies on the Hanifa Formation (Powers et al. 1966;Al Husseini 2009;Hughes et al. 2009;El-Asmar et al. 2015;Khalifa et al. 2021). A total of 30 groundwater samples were collected in October 2020 from 25-to 100-m-deep wells in Al Uyaynah-Al Jubailah farms, central Saudi Arabia (Fig. 1). Twenty samples were collected from wells drilled in Quaternary sediments, 9 samples in Jubaila limestone, and one sample in the Hanifa Formation (Fig. 2). The elevation of the water table varied from 700 m above mean sea level (AMSL) at well number 2, in the west part of the study area, to about 680 m AMSL at well number 30. Figure 2 shows the water table distribution, the direction of groundwater flow, and the spatial distribution of the total dissolved solids (TDS) in the study area. The groundwater flow followed the topography of the study area, and the TDS values are increased with the direction of the groundwater flow. The hydrogeochemical parameters (pH, electrical conductivity [EC], and TDS), ions (SiO 2 , Cl − , NO 3 − , F − , PO 4 3− , and SO 4 2− and HCO 3 − , Mg 2+ , Ca 2+ , K + , and Na + ), and HMs (Hg, Al, Sb, Cu, Cr, B, Pb, Ni, Se, Cd, As, Zn) were analyzed in the laboratories of King Saud University. The hydrogen ion concentration (pH) and EC were measured using a portable EC/pH meter (Hanna,. The ions of magnesium and calcium were determined using the titration method with ethylenediaminetetraacetic acid. Potassium and sodium were determined by a flame photometer (Corning 400). Bicarbonate was determined using acid titration. Nitrate and boron were established using methods utilizing phenoldisulfonic acid and azomethine-H, respectively. Chloride was determined using silver nitrate titration. Sulfate was estimated using a turbidity procedure. Fluoride was determined by using a fluoride selective electrode. Heavy metals were determined using an inductively coupled plasma-mass spectrometer (ICP-MS). The Piper plot is prepared to determine the groundwater facies. The groundwater quality index (GWQI), heavy metal pollution index (HPI), and metal index (MI) are used as pollution indices to document water quality. The following are the procedures and classification of these indices.

Groundwater quality index (GWQI)
Each of the 11 parameters has been assigned a weight (wi) according to its relative importance in the overall quality of drinking water as shown in Table 1. The relative weight (Wi) is computed from the following equation: Wi ¼ wi=∑wi where Wi is the relative weight and wi is the weight of each parameter.
The quality rating scale (qi) for each parameter is calculated by dividing the parameter concentration in each water sample by its respective standard (WHO 2011) multiplied by 100: where qi is the quality rating, Ci is the concentration of each chemical parameter in each water sample in mg/L, and Si is the WHO (2011) standard for each chemical parameter. Finally, the Wi and qi are used to calculate the SIi for each chemical parameter and then the GWQI is calculated from the following equation:

Heavy metal pollution index (HPI)
The HPI index is calculated as follows (Mohan et al. 1996): where Wi is the relative weight of each parameter and MAC is the maximum allowable concentration in drinking water.
An individual quality rating (Qi) is computed for each parameter using the following equation: where Mi is the monitored value of the heavy metal in the water sample, Ii is the ideal value of the parameter, and Si is the standard value of the parameter. The overall index is computed using the following equation:

Metal index (MI)
This index can be expressed by the following equation: where MI is the metal index, C is the concentration of each element in the solution, and MAC is the maximum allowed The status of groundwater for irrigation is measured using the sodium adsorption ratio (SAR), sodium percentage (Na%), Kelly's ratio (KR), and magnesium ratio (MR) (Richards 1954;Doneen 1964;Kelley 1963;Raghunath 1987). The following are the equations of calculation of this parameter and their classification.
Kelly's ratio (KR) All the values are expressed in meq/L. The ratio classifies the groundwater quality into two groups. KR < 1 (safe) and KR > 1 (unsafe). It is proposed by Raghunath (1987) and classifies the groundwater quality into two groups: MR ˂ 50% (suitable) and MR > 50% (unsuitable).

Magnesium ratio (MR)
Supplementary Table 1 shows the coordinates of groundwater boreholes (samples), hydrogeochemical parameters, major anions, major cations, and HMs. Statistical analyses were conducted using SPSS software. Principal component analysis (PCA), Pearson's correlation coefficients, and hierarchical cluster analysis (HCA; Q and R modes) were used to identify the possible sources of HMs in the groundwater samples investigated Alshehri et al. 2021).

Results and discussion
Hydrogeochemical characteristics Table 1 shows the hydrogeochemical dataset. The groundwater pH ranges from 6.5 (sample 25) to 7.8 (sample 21), with an average of 7.3, implying neutral to weak acidic waters, and falls within the standards prescribed for drinking water (WHO 2014). In assessing the groundwater quality for irrigation, TDS is a critical parameter (Salifu et al. 2017) and varies from 1088 mg/L in sample 3 to 3815 mg/ L in sample 25, with an average of 2334 mg/L, indicating values greater than the acceptable limits (1000 mg/L) of the World Health Organization (WHO 2011). Regarding irrigation criteria (Ayers and Westcott 1985;UCCC 1974), 8 samples were categorized as slight to moderate (TDS = 450-2000 mg/L) and 20 samples were severe (TDS > 2000 mg/L). In drinking water, TDS higher than 500 mg/L could cause gastrointestinal infections in consumers (Dar et al. 2011;Gnanachandrasamy et al. 2018).
Ca 2+  , and F − were greater than the permissible limits of WHO standards for drinking water ( Table 1). The higher values of these cations and anions could be attributed to the ion exchange reactions or silicate weathering (Li et al. 2016).
The saturation indices of halite, dolomite, anhydrite, gypsum, calcite, and aragonite were negative (Fig. 3), indicating undersaturated conditions regarding the capacity of groundwater to dissolve more minerals (Yidana et al. 2010). Some gypsum and calcite samples were almost close to the saturation phase because of the prolonged interaction of groundwater with carbonate aquifers and gypsiferous layers (Deutsch and Siegel 1997). All groundwater samples fell within the Cadominant type on the cationic triangle, whereas 90% fell within the SO 4 -dominant type, 4% within the Cldominant type, and 6% within the nondominant type on the anionic triangle (Fig. 4). SO 4 2− , Ca 2+ , Cl − , and Na + were the most dominant ions, and based on their dominance, the groundwater facies were classified into three types (Figs. 1 and 4): Ca-Na-SO 4 -Cl (23 samples), Ca-Cl-SO 4 , (4 samples), and Ca-SO 4 -Cl (3 samples), indicating carbonates and gypsum dissolution/ precipitation influences (Kumar 2014).

Irrigation water quality
The palm forms in the Al Uyaynah-Al Jubailah region depend primarily on groundwater for irrigation; therefore, it is important to evaluate the quality and suitability of groundwater for agricultural use. Na%, SAR, PI, KR, and MR were used as parameters for this evaluation (Supplementary Table 2). The Na% is considered an indicator of the soluble sodium content that reacts with the soil to decrease permeability (Janardhana Raju et al. 1992). Values of Na% varied from 8.21 to 47.63%, with an average of 28.97%. Six samples were excellent, 21 were good, and three were in the permissible range, implying that the groundwater is suitable for irrigation.
The SAR ranged from 3.29 to 32.49, with an average of 14.95. The groundwater samples were classified as excellent for irrigation (13 samples), good for irrigation (4 samples), doubtful for irrigation (11 samples), and unsuitable (22 and 25). The average value of Na + (235.59 mg/L) was greater than the permissible limit of WHO standards (2011) for drinking water (200 mg/L). Samples 13 and 25 were of the Ca-Na-SO 4 -Cl facies type and had higher Na + values. On the Wilcox diagram, most SAR values were within the S1 class and few within the S2 class and are of low to medium hazard (Fig. 5).
The EC values ranged from 1554 to 5450 μS/cm, with an average of 3334 μS/cm. Four samples were within permissible limits for irrigation, seven were doubtful for irrigation, and 19 were unsuitable for irrigation. On the Wilcox diagram, most EC values fell within the C4 (very high) and few fell within the C3 (high) salinity zones (Fig. 4), which could be attributed to the reverse ion exchange and lack of recent groundwater recharge (Alharbi 2018). The KR values ranged from 0.09 to 0.93, with an average of 0.48 (Supplementary Table 2). All groundwater samples were safe for irrigation (KR ˂ 1). Moreover, the magnesium ratio (MR) in the study area varied from 2.88 to 9.96%, with an average of 6.27%, indicating that the samples investigated were suitable for irrigation (MR ˂ 50%).   (Table 1). The GWQI is a mathematical application transferring large water quality data into a single number, indicating the suitability of water for drinking (Patel and Vadodaria 2015;Sahu and Sikdar 2008). It varied from 53.64 in sample 3 to 145.05 in sample 25 (Supplementary Table 2 Table 2). All groundwater samples in the study area were categorized as slightly affected (MI = 1.0-2.0). The HPI is a powerful tool used to rank the composite influence of individual HMs on the overall water quality (Rizwan et al. 2011;Sirajudeen et al. 2014;Rezaei et al. 2019). In the study area, HPI values ranged from 7.05 in sample 8 to 17.51 in sample 1, with an average of 11.89 (Supplementary Table 2). As for the MI, all samples were classified as low pollution (HPI < 45), which is attributed to the lower average levels of HMs than the permissible WHO standard limits.

Possible sources of ions and HMs
Q-mode HCA classifies the 30 groundwater samples into clusters based on TDS and ion levels (Fig. 6). Cluster 1 includes samples 1-5, 8, 9, 11-13, 15, 17, 18, 20, and 23 , and SO 2 (sample 29). From the field survey, most samples of cluster 2 are from farms cultivated with orange, pomegranate, lemon, and leafy green plants (e.g., lettuce, arugula, and radish), with or without date palms. Extensive and repeated irrigation of such plant types could dissolve cations and anions through rock-water could be attributed to the extensive use of fertilizers and pesticides (Alshahri and El-Taher 2018). The high vertical permeability of Quaternary loss could facilitate the vertical transport of contaminants into groundwater (Su et al. 2017). For drinking water, 18 groundwater samples had F − levels greater than the permissible limit of 1.5 mg/L (WHO 2011). The possible F − sources are the leaching of minerals rich in F − , industrial emissions, and the extensive use of phosphatic fertilizers (Aswathanarayana et al. 1985; Dissanayake and Chandrajith 2009). R-mode HCA classifies the hydrogeochemical parameters into two clusters (Fig. 7). The first cluster includes EC and TDS, and the second cluster includes the remaining hydrogeochemical parameters and HMs. EC showed a significant positive correlation with TDS, Ca 2+ , Na + , Cl − , and SO 4 2− , indicating a similar origin (Table 2). Ca 2+ is correlated positively with Na + , Cl − , and SO 4 2− , indicating that the dissolution of gypsum is a possible source of the ion levels in groundwater (Li et al. 2016;Zhang et al. 2018;Wu et al. 2020). Moreover, a strong positive correlation exists between SO 4 2− and Ca 2+ , Mg 2+ , Na + , and Cl − , implying a common source of these ions . A positive correlation exists between NO 3 − and HCO 3 − , indicating anthropogenic factors, especially agricultural activities, which are controlled by enhanced weathering (Adimalla and Li 2019). Land use showed a weak and negative correlation with all major anions, cations, and HMs.
PCA was performed to document the possible sources of hydrogeochemical parameters and HMs in groundwater (Wen . Nine principal components including 23.08%, 13.39%, 8.60%, 7.97%, 7.15%, 6.76%, 5.77%, 4.66%, and 4.21% of the total variance (Table 3). PC1 showed a strong association with EC, TDS, Ca 2+ , Mg 2+ , Na + , Cl − , and SO 4 2− , reflecting a natural process of the dissolution/precipitation of silicates, gypsum, and carbonates (Rezaei et al. 2019;Wu et al. 2020). PC3 showed a strong association with NO 3 − and Pb, attributed to an anthropogenic reason because of the excessive use of fertilizers and pesticides Qian 2018, Li and. PC4 had high loadings for Pb and Zn, indicating soil leaching that could originate from fertilizer and pesticide use (Kükrer and Mutlu 2019;Wen et al. 2019;Kahal et al. 2020). PC5 showed a strong association with B, Cd, and Hg, which could be influenced by mixed anthropogenic and natural sources. PC6 had high loadings for Ni and Sb, and PC8 had high loadings for Se.

Conclusions
This study used hydrogeochemical characteristics, pollution indices, irrigation quality parameters, and multivariate statistical analyses to evaluate the groundwater quality and HM contamination in some palm farms in central Saudi Arabia. The following findings were obtained: 1. The average concentrations of TDS, Ca + , Na + , K + , Cl − , SO 4 2− , and F − were higher than the permissible limits of WHO standards for drinking water. The principal mineral phase saturation indices were negative, indicating undersaturated conditions regarding the capacity of groundwater to dissolve more minerals. 2. The most dominant ions, in decreasing order, were SO 4 2− , Ca 2+ , Cl − , and Na + . According to their dominance, groundwater facies types in the study area were Ca-Na-SO 4 -Cl (23 samples), Ca-Cl-SO 4 , (4 samples), and Ca-SO 4 -Cl (3 samples), indicating carbonates and gypsum dissolution/precipitation influences. 3. The irrigation quality parameters (Na%, SAR, KR, and MR) indicated the suitability of most samples for irrigation. However, the unsuitability of most samples for irrigation, based on EC values, could be attributed to the reverse ion exchange and lack of recent groundwater recharge. 4. The average of HM values was lower than the permissible limits of WHO standards for drinking water. The GWQI indicated that half of the groundwater samples were of good quality and half of poor quality. The MI and HPI indicated that all samples were classified as slightly affected and with low pollution, respectively, attributed to increasing values of TDS, Ca + , Na + , K + , Cl − , SO 4 2− , and F − and decreasing HM values.   Data availability All data generated or analyzed during this study are included in this published article and its supplementary information file.

Declarations
Ethical approval and consent to participate Not applicable. The ethical committee does not require permission to work on the collection of specimens.

Consent for publication Not applicable.
Competing interests The authors declare no competing interests.