Health Risk of the Shallow Groundwater and Its Suitability for Drinking Purpose in Tongchuan, China

Studying the quality and health risks of groundwater is of great significance for sustainable water resources utilization, especially in arid and semi-arid areas around the world. The current study is carried out to evaluate the quality and potential health risks of groundwater in the Tongchuan area on the Loess Plateau, northwest China. Water quality index (WQI) and hydrochemical correlation analysis were implemented to understand the status of groundwater quality. Daily average exposure dosages through the oral and dermal contact exposure pathways were taken into consideration to calculate the health risks to the human body. Additionally, graphical approaches such as Piper diagram, Durov diagram and GIS mapping were used to help better understand the results of this study. The WQI approach showed that 77.1% of the samples were of excellent quality. The most significant parameters affecting water quality were NO3−, F−, and Cr6+. The health risk assessment results showed that 27.1% and 54.2% of the samples lead to non-carcinogenic risks through oral intake for adults and children, respectively. In contrast, 12.5% of the groundwater samples would result in carcinogenic risks to the residents. This study showed that the WQI method needs to be supplemented by a health risk evaluation to obtain comprehensive results for groundwater quality protection and management in the Tongchuan area.


Introduction
Groundwater is an important source for drinking and other various purposes for the majority of the population around the world, especially in arid and semiarid regions where precipitation and runoff are rare [1][2][3][4][5][6]. In addition to drinking, groundwater is useful for domestic, industrial and agricultural purposes. Due to the increased demand for groundwater, the groundwater table is subject to fluctuations, and aquifers are becoming contaminated in the context of climate change, rapid population growth, industrial development and urban expansions [7][8][9][10][11][12]. This situation is also aggravated where natural phenomena are controlling the physicochemical parameters of groundwater, such as rock influences, volcanic eruption or marine salt intrusions [13].
There is a critical increase in freshwater demand correlated with the rapid growth of the population all over the world [14] and intensive agriculture activities [15,16]. The increment of the population also leads to the expansion of cities and municipal waste that affect the groundwater quality through organic and inorganic contaminants [17][18][19][20]. Furthermore, industrialization is one of the most significant factors affecting groundwater quality through the effluents released into the nature [21][22][23][24]. Papazotos et al. [25] the skin surface (cm 2 ) were taken into consideration. Geographical information system approaches helped to better understand the results of this study.

Study Area
Tongchuan City is 70 km away from Xi'an City, the capital city of Shaanxi Province ( Figure 1). It belongs to the Chinese Loess Plateau, with longitude between 108 • 35 44" E and 109 • 29 22" E and latitude between 34 • 48 27" N and 35 • 35 23" N. The altitude of Tongchuan City ranges from 900 to 1350 m above mean sea level [39]. The study area is situated in the middle edges of Loess Plateau and adjacent to the Weihe River Valley and Guanzhong Basin [34,40]. Tongchuan lies in the transition zone of semi-humid and semi-arid climate with annual mean rainfall and evaporation of around 540 and 1964 mm, respectively. The annual temperature of Tongchuan City is 8.9-12 • C [34,39]. Precipitation, reservoir leakage and irrigation are the main recharges of groundwater, whereas discharge to some rivers such as the Beiluo River and Juhe River, evaporation and artificial extraction [34] are the main discharge pathways of groundwater. Li et al. [39] estimated the groundwater recharge at 52.8% from precipitation and 40.1% from irrigation infiltration, whereas 37.4% and 44.9% of groundwater were discharged by artificial extraction and the lateral outflow, respectively. Geologically, the study area is dominated by Quaternary loess divided into three landforms, including loess tableland, loess gully and alluvial terrace. Furthermore, this area has several layers from top down [39]: Holocene loess layer and upper Pleistocene loess layer, which are unsaturated. The middle Pleistocene layer is composed of silty clay, which separates the phreatic aquifer and the confined aquifer partially formed by the lower Pleistocene loess layer, alluvial, sand and gravel layers. The phreatic aquifer with a thickness of 25 to 60 m is the main water supply aquifer in this area.
Water 2021, 13, x FOR PEER REVIEW 3 of 21 and graphical approaches were used. The health risk assessment was performed considering daily average exposure dosage through oral pathway per unit weight (mg/(kg.d)) for drinking water intake; and for dermal contact, the exposure dosage of every single event in mg/cm 2 and the skin surface (cm 2 ) were taken into consideration. Geographical information system approaches helped to better understand the results of this study.

Study Area
Tongchuan City is 70 km away from Xi'an City, the capital city of Shaanxi Province ( Figure 1). It belongs to the Chinese Loess Plateau, with longitude between 108°35′44″ and 109°29′22″ E and latitude between 34°48′27″ and 35°35′23″ N. The altitude of Tongchuan City ranges from 900 to 1350 m above mean sea level [39]. The study area is situated in the middle edges of Loess Plateau and adjacent to the Weihe River Valley and Guanzhong Basin [34,40]. Tongchuan lies in the transition zone of semi-humid and semi-arid climate with annual mean rainfall and evaporation of around 540 and 1964 mm, respectively. The annual temperature of Tongchuan City is 8.9-12 °C [34,39]. Precipitation, reservoir leakage and irrigation are the main recharges of groundwater, whereas discharge to some rivers such as the Beiluo River and Juhe River, evaporation and artificial extraction [34] are the main discharge pathways of groundwater. Li et al. [39] estimated the groundwater recharge at 52.8% from precipitation and 40.1% from irrigation infiltration, whereas 37.4% and 44.9% of groundwater were discharged by artificial extraction and the lateral outflow, respectively. Geologically, the study area is dominated by Quaternary loess divided into three landforms, including loess tableland, loess gully and alluvial terrace. Furthermore, this area has several layers from top down [39]: Holocene loess layer and upper Pleistocene loess layer, which are unsaturated. The middle Pleistocene layer is composed of silty clay, which separates the phreatic aquifer and the confined aquifer partially formed by the lower Pleistocene loess layer, alluvial, sand and gravel layers. The phreatic aquifer with a thickness of 25 to 60 m is the main water supply aquifer in this area.

Groundwater Samples
For this study, 48 groundwater samples were collected from the wells and boreholes distributed in the study area. The criteria for the selection of water samples were based on the depth of wells, water purposes and the zone of collection. The sampling locations were recorded by coordinates using a portable GPS device and are shown as were collected in pre-cleaned plastic polyethylene bottles for physicochemical analysis after the wells were pumped for 10 min. Before sampling, all the containers were washed and rinsed thoroughly with the groundwater to be sampled. Water was filtered through 0.45 µm filter during sampling. Sample collection, handling, and preservation complied with the standard procedures recommended by Standard Examination Methods for Drinking Water [30] to ensure data quality and consistency. The water samples were analyzed in the Soil and Water Testing Center of Shaanxi Institute of Engineering Investigation, China.

Chemical Analysis and Data Processing
The samples were analyzed for physical and chemical parameters, including temperature, pH, electrical conductivity (EC), total hardness (TH), total dissolved solids (TDS), major ions (Na + , K + , Ca 2+ , Mg 2+ , Cl − , SO 4 2− , HCO 3 − , NO 3 − , NO 2 − and F − ), and Cr 6+ . Some parameters such as pH, EC and temperature were recorded on the field by portable multi-parameter devices. Drying and weighing approach was used to measure TDS. Na + and K + were determined using flame atomic absorption spectrometer and TH, Ca 2+ , and Mg 2+ were analyzed using EDTA titrimetric methods. Spectrophotometer and ion chromatography were used to determine the enrichment of NO 2 − , NO 3 − , and SO 4 2− , respectively. Standard titration method using AgNO 3 as a reactant solution was used to determine the concentration of Cl − . Traditional titrimetric and ion selective electrode methods were used to determine HCO 3 − , and F − , respectively. Ion chromatographic-colorimetric analytical principle was used to determine Cr 6+ .
The evaluation of water suitability for drinking purposes was based on the concentrations of physical and hydrochemical characteristics of the considered samples compared to the limits of physicochemical parameters recommended by the WHO [14,41,42]. The groundwater quality standards set by the Ministry of Health of the People's Republic of China, and the Standardization Administration of the People's Republic of China [43] were also considered in this study.

Statistical Analysis and Computing
In this study, statistical analysis was conducted by using SPSS 25 for Pearson's correlation. Pearson's correlation coefficient (r) helps to quantify the significance of a relationship between two parameters and was widely used in groundwater quality assessment because it gives a quick correlation value. Its mathematical formula is expressed as follows [44]: where, r xy represents the correlation coefficient between the parameters x and y, n denotes the sample size, x i is the individual value of the parameter x, x is the mean value of the parameter x, y i stands for the individual value of the parameter y, and y denotes the mean value of the parameter y. The values of correlation coefficient can be classified as very strong for r ≥ 0.80, strong for 0.60 ≤ r < 0.80, moderate for 0.40 ≤ r < 0.60, weak for 0.20 ≤ r < 0.40, and very weak for r < 0.20. In addition, the correlation coefficient is evaluated on the basis of p value. The correlation coefficient is statistically considered as highly significant when p < 0.01, marginally significant when p < 0.05, and not significant when p > 0.10 [44].
For various computing and plots, Microsoft Office 2016 (Excel and Word), Origin 2018, and Grapher 12 were used. Parameter analysis, Piper [45] and Durov [46] diagrams plots were executed using AqQA software. Finally, for mapping, ArcMap 10.3 software was used to locate samples and make a water quality distribution map. This map was obtained using Bayesian Kriging method, which is an automatic Geo-statistical interpolation pack-age incorporated in ArcGIS software. The general Kriging equation can be described as follows [47]: where λ i is the Kriging weight; Z*(x p ) estimates the unknown true value.

Water Quality Index (WQI)
To evaluate groundwater quality status in the study area, method of water quality index (WQI) was used to integrate comprehensive information through the analysis of physicochemical parameters [31,[48][49][50][51]. In other words, WQI is a single numerical value obtained by combining a large water quality data [52,53]. First, each chemical parameter is assigned with a weight (w i ), which is determined by affecting the degree of the parameters to groundwater quality. The relative weight (W i ) is computed as: where, W i is the relative weight, w i is the assigned weight of each parameter, n is the number of parameters. The value of w i ranges from 1 to 5 according to the impact of the contaminant on human health.
Then, the quality rating scale (q i ) can be computed by: where, q i is the quality rating scale, C i is the concentration of each chemical parameter in each water sample in mg/L, and S i is the standard for each chemical parameter.
To calculate the WQI, SI i has to be determined with the following equations: where, SI i is the sub-index of the ith parameter and WQI is the water quality index.

Human Health Risk Assessment
The evaluation of drinking water quality needs to be completed by a health risk assessment as polluted water may cause adverse effects on the human body through water intake and dermal contact [1,38,42]. In this study, the potential risks through dermal contact pathway were neglected for non-carcinogenic risk because it is usually low [27,38,39], and water contamination in the study area was not considerably high as listed in Table 1. The risk assessment parameters selected for this study are NH 4 + , NO 3 , NO 2 , F and Cr 6+ , using the models recommended by the Ministry of Environmental Protection of the P.R. China [30], which are also based on the model recommended by the United States Environmental Protection Agency [29,39]. According to the references mentioned above, the non-carcinogenic risk through the oral intake pathway is calculated as follows: HQ oral = Intake oral R f D oral (8) where Intake oral denotes the daily average exposure dosage through oral pathway per unit weight (mg/(kg·d)), C is the concentration of the parameter in water (mg/L), and IR represents the ingestion rate of water through drinking (L/d). EF and ED represent the exposure frequency (d/a) and exposure duration (a), respectively. BW and AT are the average body weight (kg) and average time of non-carcinogenic effects (d), respectively. For this study, the ingestion rate of water used was based on statistical investigation that considers 1.5 L per day for adults and 0.7 L per day for children under 12 years old [38]. For non-carcinogenic risk assessment, EF is 365 days per year for both adults and children. ED is 30 years for adults and 12 years for children. BW is 15.9 kg for children, 56.8 kg for adults [30]. The average time (AT) for non-carcinogenic effects on children is 4380 days, whereas it is 10,950 days for female and male adults. HQ oral and RfD oral represent the hazard quotient and the reference dosage for non-carcinogenic pollutants through the oral exposure pathway (mg/(kg.d)), respectively. This study considered the RfD oral values for NH 4 + , NO 3 − , NO 2 − , F − and Cr 6+ as 0.97, 1.6, 0.1, 0.04 and 0.003 mg/(kg.d), respectively [1,30,38]. HQ with a value exceeding 1 indicates a high potential health risk [21]. In addition, Cr 6+ can also cause carcinogenic risks through drinking water intake and dermal contact. The total carcinogenic risk is the sum of calculated cancer risk through drinking pathway and that of dermal contact and is calculated as follows [1,30]: CR total = CR oral + CR dermal (12) where CR oral represents the carcinogenic risk through the oral exposure pathway. The CR limit is set as 1 × 10 −6 . Intake oral denotes daily average exposure dosage through oral pathway per unit weight (mg/(kg·d)), SF oral is the slope factor for the carcinogenic pollutants (mg/(kg·d) −1 . The SF oral value of Cr 6+ is set as 0.5 (mg/(kg·d)) −1 by the Ministry of Environmental Protection of the P.R. China [30]. ABS gi is the gastrointestinal absorption factor, and its value is 1 for all contaminants except for Cr 6+ , with ABS gi equals 0.025 [1,30,55].
The Intake dermal is calculated as [1,30,38] as in Equations (13)- (15): where DA and SA are the exposure dosage of every single event in mg/cm 2 and the contacting area skin surface (cm 2 ), respectively. EV is the daily exposure frequency of dermal contact set at 1 for this study. ED is the exposure duration for carcinogenic risk, different from non-carcinogenic risk, and is set as 25,550 days for both adults and children.
K is the coefficient of skin permeability (0.001 cm/h), t is the contact duration, which is set as 0.4 h/day for both adults and children [1,38]. CF is a conversion factor that equals 0.001, and H denotes the average height of the population estimated at 165.3 cm for males, 153.4 cm for females and 99.4 for children [1].

Physicochemical Parameters
Groundwater quality data were first checked for reliability and accuracy by calculating the correlation between EC and the sum of cations on one hand and with the sum of anions on the other hand. The results show a good correlation with R 2 > 0.8 (Figure 2a,b. The reliability of groundwater quality data was also checked by the ion charge balance between cations and anions as follows: where, N c and N a denote total concentrations of cations and anions of a sample in meq/L, respectively. The biggest value of E was 3.14%, which indicated that the samples were reliable, as the E value was between −5% and +5%. The physicochemical indices of groundwater samples were statistically analyzed, and the results are listed in Table 1. The pH values in this study ranged from 7.05 to 8.39, which were within the guidelines set by the WHO [42] for drinking water (6.5 to 8.5). Hem [56] concluded that the pH of groundwater was controlled by the equilibrium of CO 3 2− , CO 2 and HCO 3 − , and interpreted the chemical characteristics of natural water. The mean pH value of groundwater samples was 7.77, which was suitable for drinking purpose. Mechenich and Andrews [57] considered the range of pH values from 7.5-8.3 as an ideal values range for drinking water. Thus, it can be assumed that pH values for drinking water in Tongchuan City are good and ideal. However, 12 samples (25% of the total samples) showed slight alkalinity of the drinking water in the study area with pH ranging from 8 to 8.39. Alkalinity is not only associated with high pH values, but also with hardness and excessive TDS [33]. Water 2021, 13, x FOR PEER REVIEW 8 of 21 The physicochemical indices of groundwater samples were statistically analyzed, and the results are listed in Table 1. The pH values in this study ranged from 7.05 to 8.39, which were within the guidelines set by the WHO [42] for drinking water (6.5 to 8.5). Hem [56] concluded that the pH of groundwater was controlled by the equilibrium of CO3 2− , CO2 and HCO3 − , and interpreted the chemical characteristics of natural water. The mean pH value of groundwater samples was 7.77, which was suitable for drinking purpose. Mechenich and Andrews [57] considered the range of pH values from 7.5-8.3 as an ideal values range for drinking water. Thus, it can be assumed that pH values for drinking water in Tongchuan City are good and ideal. However, 12 samples (25% of the total samples) showed slight alkalinity of the drinking water in the study area with pH ranging from 8 to 8.39. Alkalinity is not only associated with high pH values, but also with hardness and excessive TDS [33].
According to the average pH value, the groundwater in the study area can be used as drinking water. However, when comparing the detected TDS and TH values with the drinking water standards, there were two samples (4.2%) with TDS exceeding 1000 mg/L, and five samples (10.4%) with TH exceeding 500 mg/L. At the same time, referring to the drinking water quality guidelines recommended by the Ministry of Health of the People's Republic of China, there were eight samples (17%) whose TH exceeded 450 mg/L. This would be considered as hard water [1]. However, this classification is far different from the drinking water classification early made by Freeze and Cherry [58] (Table 2) based on TH. The groundwater classification on the basis of TDS and TH [14,31,58,59] in Tongchuan are as follows (   According to the average pH value, the groundwater in the study area can be used as drinking water. However, when comparing the detected TDS and TH values with the drinking water standards, there were two samples (4.2%) with TDS exceeding 1000 mg/L, and five samples (10.4%) with TH exceeding 500 mg/L. At the same time, referring to the drinking water quality guidelines recommended by the Ministry of Health of the People's Republic of China, there were eight samples (17%) whose TH exceeded 450 mg/L. This would be considered as hard water [1]. However, this classification is far different from the drinking water classification early made by Freeze and Cherry [58] (Table 2) based on TH. The groundwater classification on the basis of TDS and TH [14,31,58,59] in Tongchuan are as follows ( Table 2): 35.4% and 64.6% of samples were hard water or very hard water; 47.9% were desirable and permissible for drinking; 95.8% were fresh water and 4.2% were brackish. In addition, the TH values of water are the measures of the dissolved Ca 2+ and Mg 2+ content, which are expressed in CaCO 3 mg/L and can be associated EC, which is normally twice the hardness for uncontaminated water [23,57]. Otherwise, if it is higher than that proportion, it provides information on the presence of components such as Na + , Cl − or SO 4 2− [57]. Through the analysis of the physical and chemical indicators of the samples in the study area, the average values of EC and TH were 869.75 µS/cm and 349.94 mg/L, respectively, and the conductivity was greater than two times of the TH, which indicated that slightly high concentrations of Na + , Cl − , and SO 4 2− were in some groundwater samples.
The order of major cations in the groundwater samples from the study area was Ca 2+ > Na + > Mg 2+  Indicated by the detected results of the samples, there was no HN 4 + contamination in the groundwater of the study area because the maximum HN 4 + concentration of the samples (0.13 mg/L) was in the range of natural levels of HN 4 + in groundwater (below 0.2 mg/L), according to WHO [42]. The concentration of HN 4 + in water is an indicator of possible bacterial, sewage, landfill, and animal waste pollution [30]. The concentration of Cl − was not excessive in the analyzed samples from drinking water as it ranged from 2 to 144 mg/L. The WHO [42] has not set a health-based guideline value for Cl − , but a concentration exceeding 250 mg/L can cause the water to be unsuitable for drinking as high Cl − waters have a laxative effect for some people [33,55].
Although there is no health-based guideline value for Na + in potable water according to WHO [42], if its concentration exceeds 200 mg/L, it may taste bad, and excessive intake may cause hypertension [18]. Na + concentrations of four samples (8.3% of the total samples) slightly exceeded that threshold for the present study. A value of K + exceeding 12 mg/L in drinking water gives it a bitter taste [31]. In this study, only two samples (4.2%) exceeded this permissible limit. SO 4 2− was not excessive, except in one sample, where its concentration exceeded (572 mg/L) the SO 4 2− concentration limit proposed by WHO [30] for potable water, which is 500 mg/L.
To check the simultaneous occurrence of NO 3 − and NO 2 − in drinking water, the sum of the ratios of the concentration of each over its guideline value (GV) should not exceed 1 [42]: where C nitrate is the concentration of NO 3 − , C nitrite is the concentration of NO 2 − , and GV nitrate and GV nitrite are the guideline values of NO 3 − and NO 2 − , respectively. The application of this formula reveals that 16.6% of the groundwater samples were in the situation of simultaneous occurrence of NO 3 − and NO 2 − in drinking water. Furthermore, in the presence of microbial contamination, especially due to fecal contamination in drinking water, the health risk to infants is high [42].
In this study, 6.2% of the groundwater samples slightly exceeded the guideline value of permissible concentration in drinking water, which is 0.05 mg/L [43]. Fluoride is important for drinking water, with a concentration ranging from 0.7 to 1.2 mg/L, as it protects against dental cavities and strengthens the bones. When F − concentration exceeds 1.5 mg/L, it causes teeth mottling, fluorosis or discoloration [33,42,60,61] as well as other health problems such as nervous system harm and urinary tract disease [62,63]. Although there were two samples with F − concentration exceeding 1.5 mg/L, most of the samples (83.3%) were associated with low F − concentrations below 0.7 mg/L. Therefore, to ensure the good health of the population in Tongchuan City, F − should be added in drinking water to the majority of wells and be reduced in a few wells to avoid potential health hazards. In addition, 50 mg/L of the guideline value for NO 3 − was established by WHO [42] to protect the most sensitive populations. However, this population must be free of adverse health effects such as methemoglobinemia and thyroid effects at a concentration below 50 mg/L of NO 3 − . This health risk can seriously affect bottle-fed infants when mathemoglobinemia is complicated by the presence of microbial contamination and subsequent gastrointestinal infection that manifests as diarrhea.
Excessive boiling of water for microbiological safety purposes may increase the concentration of NO 3 − . Water for drinking should be heated until it reaches a rolling boil [42]. For NO 3 − , 45.8% of the samples exceeded the limits (20 mg/L) set by the Ministry of Health of the P.R. China [43]. Figure 3 shows the scatter plot of F − , Na + , and NO 2 − concentrations with groundwater level depth. It shows that the water samples were mostly concentrated in the shallow depth (less than 30 m). Fluoride is present in lower concentrations in shallow groundwater than in deep groundwater. This is because the dissolution of F-containing minerals such as fluorite is an important source of F − in groundwater of the study area, and the amount of fluorite is higher in the deep aquifer. The alkaline pH can influence CaF 2 activity and favors the mobilization of F − from soil and weathered rocks into groundwater. This assumption was also formulated by other researchers [64][65][66]. The enrichment of F − can also be influenced by the ratio between HCO 3 − , Na + and Ca 2+ in groundwater, as confirmed by Saxena and Ahmed [67], Rango et al. [68], and Kimambo et al. [64]. Na + concentration is lower in the shallow aquifer, which also supports the phenomenon of low F − in shallow groundwater.

Relationship between Depth of Wells and the Concentration of Physicochemical Parameters
For NO3 , 45.8% of the samples exceeded the limits (20 mg/L) set by the Ministry of Health of the P.R. China [43]. Figure 3 shows the scatter plot of F − , Na + , and NO2 − concentrations with groundwater level depth. It shows that the water samples were mostly concentrated in the shallow depth (less than 30 m). Fluoride is present in lower concentrations in shallow groundwater than in deep groundwater. This is because the dissolution of F-containing minerals such as fluorite is an important source of F − in groundwater of the study area, and the amount of fluorite is higher in the deep aquifer. The alkaline pH can influence CaF2 activity and favors the mobilization of F − from soil and weathered rocks into groundwater. This assumption was also formulated by other researchers [64][65][66]. The enrichment of F − can also be influenced by the ratio between HCO3 − , Na + and Ca 2+ in groundwater, as confirmed by Saxena and Ahmed [67], Rango et al. [68], and Kimambo et al. [64]. Na + concentration is lower in the shallow aquifer, which also supports the phenomenon of low F − in shallow groundwater. Samples with low concentration of NO2 − are usually observed in the shallow aquifer than in the deep aquifer. This may be due to the oxidation environment in the shallow aquifer that favors the transformation of NO2 − to NO3 − . Numerous studies have shown that human activities such as agriculture, industry, domestic sewage, landfills, and household waste influences shallow groundwater quality [1,32,69].

Hydrochemical Types of Groundwater
The Durov diagram depicted in Figure 4b reveals that most of the samples are concentrated in the field of HCO3-Ca type and combined HCO3·SO4-Ca·Mg type. This situation may result from the dissolution of CO3 − minerals and F − [68]. As also discussed by Ravikumar et al. [70] and Lloyd and Heathcote [71], the HCO3-Ca dominant frequently indicates that recharging waters in limestone and sandstone is associated with dolomite. To assess the water quality, a Piper diagram (Figure 4a) was used to characterize the hydrogeochemical facies of groundwater samples from the study area. The Piper plot shows that Ca 2+ , Na + , and Mg 2+ are dominant cations in the region. Conversely, HCO3 − and SO4 2− dominate the facies, while Cl − is quasi-inexistent. The general classification of all samples Samples with low concentration of NO 2 − are usually observed in the shallow aquifer than in the deep aquifer. This may be due to the oxidation environment in the shallow aquifer that favors the transformation of NO 2 − to NO 3 − . Numerous studies have shown that human activities such as agriculture, industry, domestic sewage, landfills, and household waste influences shallow groundwater quality [1,32,69].

Hydrochemical Types of Groundwater
The Durov diagram depicted in Figure 4b reveals that most of the samples are concentrated in the field of HCO 3 -Ca type and combined HCO 3 ·SO 4 -Ca·Mg type. This situation may result from the dissolution of CO 3 − minerals and F − [68]. As also discussed by Ravikumar et al. [70] and Lloyd and Heathcote [71], the HCO 3 -Ca dominant frequently indicates that recharging waters in limestone and sandstone is associated with dolomite. To assess the water quality, a Piper diagram (Figure 4a) was used to characterize the hydrogeochemical facies of groundwater samples from the study area. The Piper plot shows that Ca 2+ , Na + , and Mg 2+ are dominant cations in the region. Conversely, HCO 3 − and SO 4 2− dominate the facies, while Cl − is quasi-inexistent. The general classification of all samples shows 81.25% Ca·Mg-HCO 3 , 8% Ca·Mg-SO 4 ·Cl, 4.1% Na-Cl and 6.25% Na-HCO 3 water type (Figure 4a). The dominant Ca·Mg-HCO 3 type may indicate that the influence of dissolution on groundwater chemistry is more considerable, and it signifies the dominance of alkaline earths over alkalis; weak acids exceed strong acids. This observation was also found by other researchers, notably Xu et al. [72], Ravikumar et al. [70] and Singh et al. [16].
shows 81.25% Ca·Mg-HCO3, 8% Ca·Mg-SO4·Cl, 4.1% Na-Cl and 6.25% Na-HCO3 water type (Figure 4a). The dominant Ca·Mg-HCO3 type may indicate that the influence of dissolution on groundwater chemistry is more considerable, and it signifies the dominance of alkaline earths over alkalis; weak acids exceed strong acids. This observation was also found by other researchers, notably Xu et al. [72], Ravikumar et al. [70] and Singh et al. [16].

Hydrochemical Correlation Analysis of Water Quality
To better understand the major hydrogeochemical processes that control the chemical characteristics, it is necessary to carry out a Pearson's correlation analysis that shows the relationship between each pair of physicochemical indices [39,73]. Table 3 gives the correlation values of physicochemical parameters of water samples. Table 3. Pearson correlation matrix between physicochemical parameters of water samples. Bold number indicates that the correlation is significant at the 0.05 level (two-tailed). Italic number indicates that the correlation is significant at the 0.01 level (two-tailed).
As shown in Table 3, there is a strong correlation, which is explained by ions exchange between TDS and EC with r = 0.980 at the level of p > 0.01, Ca 2+ content and TH

Hydrochemical Correlation Analysis of Water Quality
To better understand the major hydrogeochemical processes that control the chemical characteristics, it is necessary to carry out a Pearson's correlation analysis that shows the relationship between each pair of physicochemical indices [39,73]. Table 3 gives the correlation values of physicochemical parameters of water samples. Bold number indicates that the correlation is significant at the 0.05 level (two-tailed). Italic number indicates that the correlation is significant at the 0.01 level (two-tailed). Table 3, there is a strong correlation, which is explained by ions exchange between TDS and EC with r = 0.980 at the level of p > 0.01, Ca 2+ content and TH with r = 0.916 at the level of p > 0.01, Cland TDS with r = 0.860 at the level of p > 0.01, Cl − and EC with r = 0.857 at the level of p > 0.01, and SO 4 2− correlates with TDS and EC with both r = 0.804 at the level of p > 0.01. In addition, a strong correlation exists between Na + and TDS, SO 4 2− , EC, and Cl − with r = 0.765, 0.745, 0.742, and 0.727, respectively. Furthermore, there is a strong relationship between NO 3 − and TH with r = 0.662 at the level of p > 0.01, Cl − and SO 4 2− with r = 0.623 at the level of p > 0.01, TH and EC with r = 0.607 at the level of p > 0.01, and Na + and F − with r = 0.602 at the level of p > 0.01.

As shown in
Although all the aforementioned correlations between parameters are positive, there is a strong negative correlation between Ca 2+ and pH with r = −0.721 at the level of p > 0.01. Ca 2+ and Mg 2+ are significantly correlated to TH because they contribute to the water hardness.
A strong correlation between Cr 6+ and F − with r = 0.703 at level p > 0.05, which may be due to the oxidation mechanism of Cr 3+ to Cr 6+ in the presence of F − in groundwater, was observed. Finally, a significant correlation between Cr 6+ and both Na + and pH was also noticeable. All these parameters may have triggered the mobilization of Cr in the groundwater system [29]. Table 4 shows the weight assigned to each parameter, and the relative weights are calculated using Formula (3).  Table 5 lists the water quality assessment results. As shown in Table 5, 37 of the samples (77.1%) are of excellent quality, 9 samples (18.7%) are of good quality, and 2 samples (4.2%) are of poor quality. The most significant parameters affecting the water quality in the study area are NO 3 − , F − , and Cr 6+ . Water without excellent quality is dominated by wells with low depth represented by samples TW1-052, TW1-047, TW1-041, TW2-021, TW2-66 and TW2-67 with 2, 3, 2, 3, 10, and 8 m, respectively.

Water Quality Index Assessment
The contamination source of the wells represented by samples TW1-052, TW1-047, TW1-041 might be the ravines situated nearby. These ravines may bring contaminated water that leaks in the phreatic and shallow aquifer. The other concerned wells with low depth were possibly contaminated by human activities, as they are located in residential and agricultural areas.
As depicted in Figure 5, a major part of Tongchuan is dominated by excellent water and can be used for drinking purpose. However, in some towns such as Yuhua, Wangshiao, and Chenlu, for example, groundwater quality is not suitable for drinking. Therefore, water needs pretreatment before drinking, and taking effective measures to prevent groundwater pollution is imperative. Low deep wells should also be drilled deeply to avoid contamination by surface water leakage and pollution caused by human activities. Table 6 presents the calculated health risk to adults and children when they are exposed to the contaminants in groundwater through drinking water intake. The total health risk due to contaminated drinking water intake ranges from 0.21 to 4.71, with a mean of 0.89 for adults. For children, the health risk is evaluated through the hazard quotient ranged from 0.35 to 7.85 with a mean of 1.52. Considering that HQ > 1 for non-carcinogenic risk indicates high potential health risk [1], water from wells represented by samples TW1-008, TW1-009, TW1-037, TW1-041, TW1-047, TW1-049 to TW1-054, TW1-059 to TW1-061, TW2-014 to TW2-067 was not safe, especially for children.  TW1-025 14.05 Excellent TW1-054 59.63  Good  TW2-058 19.68 Excellent  TW1-032 16.13 Excellent TW1-055 25.67 Excellent TW2-066 54.70  Good  TW1-036 21.27 Excellent TW1-058 23.88 Excellent TW2-067 50.18  Good  TW1-037 38.25 Excellent TW1-059 31.14 Excellent TW2-069 15.90 Excellent Water without excellent quality is dominated by wells with low depth represented by samples TW1-052, TW1-047, TW1-041, TW2-021, TW2-66 and TW2-67 with 2, 3, 2, 3, 10, and 8 m, respectively.

Health Risk Assessment
The contamination source of the wells represented by samples TW1-052, TW1-047, TW1-041 might be the ravines situated nearby. These ravines may bring contaminated water that leaks in the phreatic and shallow aquifer. The other concerned wells with low depth were possibly contaminated by human activities, as they are located in residential and agricultural areas.
As depicted in Figure 5, a major part of Tongchuan is dominated by excellent water and can be used for drinking purpose. However, in some towns such as Yuhua, Wangshiao, and Chenlu, for example, groundwater quality is not suitable for drinking. Therefore, water needs pretreatment before drinking, and taking effective measures to prevent groundwater pollution is imperative. Low deep wells should also be drilled deeply to avoid contamination by surface water leakage and pollution caused by human activities.  Table 6 presents the calculated health risk to adults and children when they are exposed to the contaminants in groundwater through drinking water intake. The total health As shown in Figure 6, NO 3 − contributes a considerable amount to non-carcinogenic risk for both adults and children and is followed by F − , Cr 6+ , and lastly, by NO 2 − . The respective HQ mean values for adults are 0.54, 0.31, 0.02, 0.02, and 0.90, 0.52, 0.07, and 0.03 for children. HN 4 + has zero contribution to health risk in this study area for both adults and children. High nitrate health risk is probably due to the anthropogenic activities, especially fertilizers in agriculture [21]. In addition, Wei et al. [34] reported that NO 3 − pollution was a major environmental geological problem in the groundwater for this region. Overall, 27.1% and 54.2% of the samples present a health risk through drinking water intake for adults and children, respectively. In this study, Cr 6+ was also considered as a carcinogenic risk pollutant. Considering the acceptable CR total limit set as 1 × 10 −6 by the Ministry of Environmental Protection of the P.R. China [30], the results shown in Table 7 revealed a critical carcinogenic risk by drinking and daily contact of water from six (12.5%) wells in the study area. As shown in Figure 6, NO3 − contributes a considerable amount to non-carcinogenic risk for both adults and children and is followed by F − , Cr 6+ , and lastly, by NO2 − . The respective HQ mean values for adults are 0.54, 0.31, 0.02, 0.02, and 0.90, 0.52, 0.07, and 0.03 for children. HN4 + has zero contribution to health risk in this study area for both adults and children. High nitrate health risk is probably due to the anthropogenic activities, especially fertilizers in agriculture [21]. In addition, Wei et al. [34] reported that NO3 − pollution was a major environmental geological problem in the groundwater for this region. Overall, 27.1% and 54.2% of the samples present a health risk through drinking water intake for adults and children, respectively.

Health Risk Assessment
In this study, Cr 6+ was also considered as a carcinogenic risk pollutant. Considering the acceptable CRtotal limit set as 1 × 10 −6 by the Ministry of Environmental Protection of the P.R. China [30], the results shown in Table 7 revealed a critical carcinogenic risk by drinking and daily contact of water from six (12.5%) wells in the study area. Table 7. Calculated carcinogenic risk due to Cr 6+ in water intake and dermal contact.

Conclusions
In the present study, water quality index (WQI), statistical analysis and graphical approaches were implemented to understand the status of groundwater quality in the Tongchuan area on the Loess Plateau, northwest China. In addition, GIS approaches helped to map the WQI results of this study. Daily average exposure dosage through oral pathway was taken into consideration to calculate health risks to the human body through drinking water intake. For dermal contact, the exposure dosage of every single event in mg/cm 2 and the skin surface (cm 2 ) were considered. The following conclusions can be achieved:

•
In summary, the results of this study demonstrated that groundwater in the study area is suitable for drinking in general. WQI approach showed that 77.1% of the samples are of excellent quality, nine samples (18.7%) are of good quality, and two samples (4.2%) are of poor quality. • NO 3 − , F − , and Cr 6+ are the most significant parameters affecting water quality in this study; 27.1% and 54.2% of the overall samples present a non-carcinogenic health risk through drinking water intake for adults and children, respectively. The CR total of 12.5% of the samples ranges from 4.18 × 10 −5 to 4 × 10 −4 for adults and from 6.89 × 10 −5 to 8 × 10 −4 for children, which exceeded the acceptable limit (1 × 10 −6 ). • NO 3 − considerably contributes to non-carcinogenic risk for both adults and children and is followed by F − , Cr 6+ and lastly by NO 2 − , with respective mean HQ of 0.49, 0.28, 0.02 and 0.02 for adults. For children, the mean HQ for NO 3 − , NO 2 − , F − and Cr 6+ are 0.95, 0.03, 0.55 and 0.07, respectively. HN 4 + has zero contribution to health risk in this study area for both adults and children. The high concentration of NO 3 − in the study area is due to anthropogenic activities, especially fertilizers in agriculture as also discussed by previous researchers.
• WQI is not enough to conclude whether water is suitable or not for drinking. The assessment of carcinogenic and non-carcinogenic risk on the human body showed that groundwater in Tongchuan was not totally safe. Therefore, water pretreatment before drinking and taking effective measures to prevent groundwater pollution are imperative.
This study will be helpful to local decision makers for implementing measures, policy and strategies to protect groundwater resources and reduce the health risks of residents by groundwater consumption through oral and dermal pathways. It is also useful for international scholars who may find information for similar studies or its improvement. Data Availability Statement: All processed data generated or used during the study appear in the submitted article. Raw data may be provided on reasonable request from the corresponding author.