Introduction

Groundwater plays a significant role in domestic water supply and agricultural irrigation in arid and semi-arid regions due to the limitations of precipitation and surface water (Chen et al. 2019; Li and Gao 2019; Rao et al. 2020; Singh et al. 2020). Hence, understanding the quality of groundwater is of great significance for ensuring the safety of drinking water and human health (Li et al. 2017). Groundwater quality is largely affected by natural conditions (lithology, groundwater velocity, water–rock interactions, quality of recharge water, influence from other aquifers), anthropogenic activities (agriculture, industry, overexploitation of groundwater, leaching of pollutants in soil), and atmospheric precipitation input (Jiang et al. 2009; Li et al. 2017; Peng et al. 2018; Re et al. 2014; Zereg et al. 2018). A comprehensive understanding of groundwater geochemistry plays a critical role in identifying the changes in water quality, protecting groundwater resources, and evaluating the suitability for drinking and irrigation (Arumugam and Elangovan 2009; Cloutier et al. 2008; Fendorf et al. 2010; Yang et al. 2020). Moreover, the hydrogeochemical characteristics of groundwater largely determine the suitability of groundwater for drinking and irrigation (Belkhiri et al. 2010; Kumar et al. 2016).

The multivariate statistical methods (principal component analysis (PCA), hierarchical cluster analysis (HCA), and correlation analysis (CA)) combined with hydrogeochemical analysis have greatly promoted the understanding of the formation and evolution of water chemistry characteristics (Cloutier et al. 2008; Cortes et al. 2016; Heydarirad et al. 2019; Patil et al. 2020). Generally, statistical analysis of the groundwater geochemical data is performed first, and then the hydrogeochemical analysis method is used to identify the hydrogeochemical characteristics and the evolution process controlling groundwater chemistry (Yang et al. 2020). Hierarchical cluster analysis is an effective tool to divide water samples into different clusters based on groundwater chemistry data, representing different hydrochemical facies or water chemical types (Belkhiri et al. 2010). However, a principal component analysis divides various hydrochemical parameters into several principal components according to the relationship between different variables, which represent the factors that control groundwater chemistry (Cloutier et al. 2008; Cortes et al. 2016; Jehan et al. 2020a). The integration of principal component analysis, hierarchical cluster analysis, and correlation analysis is an effective tool in identifying groundwater hydrochemical characteristics and hydrogeochemical evolution processes and promotes the thorough understanding of groundwater quality (Ayed et al. 2017; Monjerezi et al. 2011; Noshadi and Ghafourian 2016; Yang et al. 2015).

The water quality index (WQI), proposed by (Horton 1965), is an effective tool for assessing the overall water quality (Jehan et al. 2020b) and the suitability for drinking based on hydrochemical data and water quality standards. Many previous studies have utilized the WQI method to evaluate groundwater quality in different regions of the world (Adimalla and Qian 2019; Simões et al. 2008; Jianhua et al. 2011; Logeshkumaran et al. 2015; Varol and Davraz 2015; Vasanthavigar et al. 2010). However, in the process of calculating the WQI, the assignment of hydrochemical parameters weights is usually subjective, which will obscure part of the important information of water quality (Amiri et al. 2014). In order to objectively reflect the weight of different hydrochemical parameters in water quality evaluation, the entropy weight method (Shannon 1948) is used to determine the weight of each variables, namely the entropy weight water quality index (EWQI) method. The EWQI has been widely used to evaluate the quality of groundwater for drinking (Adimalla et al. 2018; Gao et al. 2020; Karunanidhi et al. 2021; Wu et al. 2018), which is usually combined with ArcGIS to more intuitively reflect the spatial distribution characteristics and differences of water quality categories (He and Wu 2019; Hossain and Patra 2020; Karunanidhi et al. 2020; Shahid et al. 2014).

The study area is located in the southwest of Ordos Basin, which is a typical loess plateau gully area with fragmented and complex terrain. The study area belongs to the semi-arid climate, which is characterized by low precipitation and large evaporation, leading to water shortages and becoming a crucial issue restricting the development of the region. Therefore, groundwater, especially quaternary phreatic groundwater, plays a vital role in domestic drinking and agricultural irrigation. However, insufficient research and attention on the hydrochemical characteristics and the water quality evaluation of the quaternary phreatic groundwater for drinking and irrigation limits our understanding of the current status of groundwater quality and water resources management in this region. Consequently, there is an increasing significance to identify the hydrochemical characteristics and evolution process of groundwater and evaluate the suitability for drinking and irrigation purposes.

This study utilizes multivariate statistical methods (CA, PCA, HCA) and hydrogeochemical analysis methods (Piper trilinear diagram, Gibbs diagram, binary diagram, saturation index, and chloro-alkaline indices) to analyze the hydrochemical characteristics and identify the hydrogeochemical evolution process of groundwater. Furthermore, the EWQI method was used to evaluate the suitability of groundwater quality for drinking, then the ArcGIS software was employed to establish the evaluation zones. In addition, multiple water quality assessment indicators for irrigation (SAR, %Na, RSC, KR, MH, PI, USSL diagram, and Wilcox diagram) are adopted to evaluate the suitability of groundwater for irrigation in the study area.

Study area

Description of the study area

The study area (E: 106° 21′ 23″–108° 42′ 33″, N: 35° 14′ 31″–37° 9′ 31″) is located in the southwest of the Ordos Basin (Fig. 1), which belongs to a typical loess area. The terrain slopes from the northeast and west to the southeast, with an altitude of 885–2089 m. The study area is located in the gully area of the Loess Plateau in the middle reaches of the Yellow River. The terrain is high on the east, north, and west and low on the middle and southern sides, which forms a non-closed basin with an opening in the southwest. The rainfall gradually decreases from south to north, and the rainfall is mostly concentrated in July to September, accounting for about 60% of the annual precipitation. The study area has crisscross gullies and dendritic water systems, which belong to the Jinghe water system. The main rivers are Puhe River and Malianhe River. In addition, there are two larger secondary tributaries of Huanjiang River and Rouyuanhe River.

Fig. 1
figure 1

Location map of the study area and sampling sites distribution map

Full size image

Geology and hydrogeology

The study area is an integral part of the Ordos Basin and is a large syncline sedimentary basin composed of Paleozoic and Mesozoic with an axis near north to south. The stratigraphic lithology and lithofacies in the study area are relatively complex, mainly including Quaternary alluvial deposits, Quaternary loess, Neogene mudstone, Cretaceous Jingchuan Formation mudstone and sandstone, Cretaceous Luohandong Formation sandstone, and Cretaceous Huanhe Formation mudstone and sandstone from top to bottom (Fig. S1). This study focuses on the quaternary strata, including aeolian loess and river alluvial deposits. According to lithological characteristics and geological age, the quaternary strata can be divided into the following: (1) Pleistocene aeolian loess layer, which is composed of Lower Pleistocene Wucheng Formation, Middle Pleistocene Lishi Formation, and Upper Pleistocene Malan Formation. (2) Holocene alluvial–diluvial strata, which is mainly distributed in the valley terraces, large gullies, and piedmont alluvial fans of major rivers.

According to the water-bearing rock layer and the type of medium, the study area can be divided into karst groundwater system, clastic rock groundwater system, and Quaternary loose layer groundwater system. This study takes the Quaternary loose rock groundwater system as the target aquifer. The overlying Quaternary loose rock groundwater distribution is relatively independent, which is mainly divided into three sub-systems of loess hilly phreatic aquifer, loess plateau phreatic aquifer, and river valley alluvial–proluvial phreatic aquifer. The phreatic aquifers of the loess hills are distributed in the loess hilly areas. The aquifer medium is mainly Lishi loess (Q2), the overlying Malan loess (Q3) is permeable and does not contain water, and the underlying Wucheng loess (Q1) has a high content of clay particles and is relatively water-resistant. However, the phreatic aquifers in the loess plateau are mainly distributed in the 19 loess plateau areas south of Qingcheng (Fig. S2). Each plateau is a relatively independent hydrogeological unit, with the similar hydrogeological conditions. Atmospheric precipitation is the main source of replenishment for the loess phreatic water in the basin. The runoff of the loess phreatic aquifer is mainly affected by topography, and the direction of runoff is changeable, mainly from high terrain to low terrain.

Data and methods

Sampling and test

A total of 67 groundwater samples were collected from motor-pumped wells in the study area in October 2020, as shown in Fig. 1. Most of the groundwater wells in the study area are used for drinking and irrigation, and the groundwater samples are taken from Quaternary phreatic aquifers. All wells must be pumped for at least 15 min before sampling. The groundwater samples are taken into polyethylene bottles after the temperature and pH have stabilized. The portable instruments are used for on-site measurement of temperature, pH, and conductivity (EC). The water samples used to test other general hydrochemical indicators (Na+, K+, Ca2+, Mg2+, HCO3−, SO42−, Cl−, TDS, NO3−, NH4+, F−, and Fe) were put into polyethylene bottles, sealed and stored, and transported to the laboratory for testing. The detection of parameters is recommended by the water quality—determination of 32 elements—inductively coupled plasma optical emission spectrometry (Ministry of Environmental Protection of the People’s Republic of China 2015), standard examination methods for drinking water—water analysis quality control (Ministry of Health of the People’s Republic of China 2006).

Data pre-processing

The test results of hydrochemical parameters inevitably appear as censored data (the measured concentration less than detection limits). In order to utilize the data as much as possible for analysis, the commonly used methods for processing censored data are either to exclude from hydrochemical analysis or to replace it with a specific data less than the detection limit (Barescut et al. 2011; Cloutier et al. 2008; Croghan and Egeghy 2003; Güler and Thyne 2004). This study uses 75% of the detection limit to replace the censored data (VanTrump Jr and Miesch 1977; Yang et al. 2020).

Moreover, in order to facilitate groundwater chemistry analysis, the concentration data (mg/L) of all parameters are converted into milligram equivalent concentration (meq/L). Quality control was carried out by calculating the percent charge balance errors (CBE) to verify the reliability of the chemical analysis results. The formula for calculating CBE is as follows:

$$ \mathrm{CBE}=\frac{\sum \mathrm{cation}-\sum \mathrm{anion}}{\sum \mathrm{cation}+\sum \mathrm{anion}}\times 100 $$
(1)

where all units of chemical data are meq/L.

The data with the CBE < ± 10% (Karunanidhi et al. 2020) were used in this study, while the groundwater samples with the CBE over 10% were excluded from the all data. In this study, the number of the datasets used for subsequent analysis is 66 based on the CBE.

Statistical analysis

Correlation analysis

Correlation analysis is a statistical method to test the degree of intimacy between two variables. In this study, Pearson correlation analysis was employed to discriminate the correlation between different water chemistry parameters. A statistical difference is considered at a 5% significant level (p < 0.05).

Hierarchical cluster analysis

Hierarchical cluster is a clustering method that classifies different datasets according to their similarity, which is determined by calculating the distance between the data points of each category and all the data points. In this study, the Euclidean distance is used to calculate the distance between different types of data points, which is calculated as follows:

$$ D=\sqrt{{\left({x}_1-{y}_1\right)}^2+{\left({x}_2-{y}_2\right)}^2} $$
(2)

Then, Ward’s linkage method (Ward 1963) is utilized to assess the distances between different clusters using the variance analysis (Yang et al. 2020), and the final result is presented in the form of a dendrogram. Moreover, the number of clusters is determined according to the phenon line.

Principal component analysis

Principal component analysis (PCA) is a widely used data dimensionality reduction method. PCA can retain most of the information while reducing the dimensionality of the data (Boonkaewwan et al. 2021). In this study, PCA was employed to verify whether the hierarchical cluster analysis is reasonable.

Hydrochemical analysis

Geochemical modeling

The hydrogeochemical model describes the chemical reactions that occur in the groundwater system and the state of ions in groundwater and minerals. The geochemical model can simulate various equilibrium reactions between groundwater and minerals using the chemical thermodynamic (Charlton and Parkhurst 2011; Karunanidhi et al. 2021). PHREEQC is a geochemical reaction model (Parkhurst and Appelo 1999), which is widely used to simulate the chemical equilibrium between groundwater and minerals by calculating the saturation indices (SI) in this study.

Groundwater quality evaluation

Groundwater quality evaluation for drinking

The entropy weighted water quality index (EWQI) is an improved method based on water quality index (WQI) (Amiri et al. 2014; Singh et al. 2019), which is characterized by the information entropy used to determine the weight of groundwater hydrochemical parameters. EWQI is widely used to quantitatively evaluate groundwater quality for drinking, which can be obtained through the following steps (Gorgij et al. 2017; Rao et al. 2020; Su et al. 2018):

  1. 1.

    According to the groundwater hydrochemistry data, the eigenvalue matrix, X, which is composed of “m” groundwater samples and “n” hydrochemistry parameters. X is shown as follows:

$$ X=\left[\begin{array}{cccc}{x}_{11}& {x}_{12}& \cdots & {x}_{1n}\\ {}{x}_{21}& {x}_{22}& \cdots & {x}_{2n}\\ {}\vdots & \vdots & \ddots & \vdots \\ {}{x}_{m1}& {x}_{m2}& \cdots & {x}_{mn}\end{array}\right] $$
(3)
  1. 2.

    Due to the difference in the dimensions of the hydrochemistry parameters, in order to facilitate the analysis of different parameters, the data is normalized using Eq. (4) and matrix X is converted into a standard-grade matrix Y, which is shown in Eq. (5).

$$ {y}_{ij}=\frac{x_{ij}-{\left({x}_{ij}\right)}_{\mathrm{min}}}{{\left({x}_{ij}\right)}_{\mathrm{max}}-{\left({x}_{ij}\right)}_{\mathrm{min}}} $$
(4)
$$ Y=\left[\begin{array}{cccc}{y}_{11}& {y}_{12}& \cdots & {y}_{1n}\\ {}{y}_{21}& {y}_{22}& \cdots & {y}_{2n}\\ {}\vdots & \vdots & \ddots & \vdots \\ {}{y}_{m1}& {y}_{m2}& \cdots & {y}_{mn}\end{array}\right] $$
(5)
  1. 3.

    The ratio of the index value of the jth parameter at the ith sample and the information entropy of the parameter j (ej) were calculated by Eqs. (6) and (7), which is as follows:

$$ {P}_{ij}=\frac{1+{y}_{ij}}{\sum_{i=1}^m\left(1+{y}_{ij}\right)} $$
(6)
$$ {e}_j=-\frac{1}{\ln m}{\sum}_{i=1}^m{P}_{ij}\ln \left({P}_{ij}\right) $$
(7)
  1. 4.

    The entropy weight of the parameter j (wj) was calculated by the Eq. (8).

$$ {w}_j=\frac{1-{e}_j}{\sum_{j=1}^n\left(1-{e}_j\right)} $$
(8)

Then, the quality rating scale of the parameter j (qj) was calculated by Eq. (9).

$$ \left\{\begin{array}{l}{q}_j=\frac{C_j}{S_j}\times 100\\ {}{q}_{\mathrm{pH}}=\frac{C_{\mathrm{pH}}-7.0}{8.5-7.0}\times 100\end{array}\right. $$
(9)

where Cj, CpH represents the concentration (mg/L) and pH of parameter j, respectively. Sj represents the limitations of parameter j based on the standards for drinking water quality of China.

  1. 5.

    Finally, EWQI is calculated by following Eq. (10):

$$ \mathrm{EWQI}=\sum \limits_{j=1}^m{w}_j{q}_j $$
(10)

According to the classification criteria of EWQI values (Amiri et al. 2014; Gorgij et al. 2017; Islam et al. 2017), groundwater quality is divided into five categories, which are excellent, good, medium, poor, and extremely poor, respectively, as shown in Table 4.

Groundwater quality evaluation for irrigation

The primary goal of evaluating the suitability of water for irrigation is to understand the salinity levels of the water, which affect the soil structure and the crops yield. The following criteria are used to identify the characteristics of water used for irrigation:

Sodium adsorption ratio

Sodium adsorption ratio (SAR) is one of the significant indicators for evaluating irrigation water quality (Ghouili et al. 2018; Srinivasamoorthy et al. 2014), which was used to indicate the degree of sodium hazard (alkali hazard) caused by irrigation water to soil or plants. It can be expressed by the ratio of Na+ concentration to the square root of the sum of Ca2+ and Mg2+ concentration, and the formula is as follows (Karanth 1987):

$$ \mathrm{SAR}=\frac{{\mathrm{Na}}^{+}}{\sqrt{\frac{{\mathrm{Ca}}^{2+}+{\mathrm{Mg}}^{2+}}{2}}} $$
(11)

Soluble sodium percentage (%Na)

Soluble sodium percentage is used to evaluate the degree of sodium damage (Kumar et al. 2016), which is determined by the ratio of Na+ to (Na+ + K+ + Ca2+ + Mg2+). The calculation formula is as follows:

$$ \%\mathrm{Na}=\frac{{\mathrm{Na}}^{+}}{{\mathrm{Na}}^{+}+{\mathrm{K}}^{+}+{\mathrm{Ca}}^{2+}+{\mathrm{Mg}}^{2+}}\times 100 $$
(12)

Residual sodium carbonate (RSC)

The suitability of groundwater for irrigation is affected by the total amount of carbonate and bicarbonate exceeding the total amount of calcium and magnesium in the water (Ghouili et al. 2018; Srinivasamoorthy et al. 2014).

$$ \mathrm{RSC}=\left({{\mathrm{HCO}}_3}^{-}+{{\mathrm{CO}}_3}^{2-}\right)-\left({\mathrm{Ca}}^{2+}+{\mathrm{Mg}}^{2+}\right) $$
(13)

Kelly’s ratio

Kelly’s ratio (KR) evaluates the suitability of water for irrigation by examining the balance between sodium ions, calcium ions, and magnesium ions. KR greater than 1 represents the excess Na+, indicating the unsuitability of water for irrigation. KR is calculated by the following formula (Kelly 1940):

$$ \mathrm{KR}=\frac{{\mathrm{Na}}^{+}}{{\mathrm{Ca}}^{2+}+{\mathrm{Mg}}^{2+}} $$
(14)

Magnesium hazard (MH)

The physical properties of the soil will be affected due to the excessive absorption of magnesium. Therefore, the magnesium hazard, one of the important criteria, was used to assess the suitability of irrigation water (Szabolcs 1964). The MH is defined as the ratio of magnesium concentration to divalent cations (Ca2+ and Mg2+), which is calculated by the following formula:

$$ \mathrm{MH}=\frac{{\mathrm{Mg}}^{2+}}{{\mathrm{Ca}}^{2+}+{\mathrm{Mg}}^{2+}}\times 100 $$
(15)

Permeability index

The permeability of soil is affected by the sodium, magnesium, calcium, and bicarbonate in water (Saha et al. 2019). The permeability index (PI), proposed by Doneen (1964), provides a division scheme for irrigation water quality based on the ratio of sodium and bicarbonate to all cations, indicating the suitability of water used for irrigation. PI is calculated by the following formula:

$$ \mathrm{PI}=\frac{{\mathrm{Na}}^{+}+\sqrt{{{\mathrm{HCO}}_3}^{-}}}{{\mathrm{Ca}}^{2+}+{\mathrm{Mg}}^{2+}+{\mathrm{Na}}^{+}}\times 100 $$
(16)

where all units of the ions are in meq/L.

Results and discussion

Hydrogeochemical characteristics

Statistical analysis of hydrochemistry

Hierarchical cluster analysis

The ward’s linkage method and Euclidean distance are employed in HCA, which divides all groundwater samples into four clusters according to the hydrogeochemical characteristic. The dendrogram is obtained as shown in Fig. 2, representing that the four clusters are composed of 25, 30, 8, and 3 samples, respectively. Each cluster has similar hydrochemistry characteristics, but there are differences between each cluster.

Fig. 2
figure 2

Classification results of groundwater samples using hierarchical cluster analysis

Full size image

The spatial distribution (Fig. S3) of the clustering results and TDS shows that cluster 3 and cluster 4 samples are located in areas where TDS is greater than 1000 mg/L (brackish water to saline water), while most of cluster 1 and cluster 2 samples are located in areas where TDS is less than 1000 mg/L (fresh water). From a spatial perspective, on the whole, clusters 3 and cluster 4 samples are located in the northwest of the study area, representing the groundwater characteristics of brackish water to saline water, while most of clusters 1 and cluster 2 samples are located in the southern and eastern of the study area, indicating that the groundwater is dominated by fresh water.

The box–whisker plot was used to identify the characteristics and differences of various hydrochemical parameters in different clusters. Figure 3 shows that the concentration of Na+, K+, Ca2+, Mg2+, SO42−, Cl−, NO3−, TH, and TDS represent the increasing trend from the cluster 1 to cluster 4. However, the pH value shows a decreasing trend. The average value of pH of the four clusters range from 7.46 (cluster 4) to 7.93 (cluster 1), indicating the weakly alkaline conditions in the groundwater. The weak alkalinity of groundwater may be caused by the reaction of groundwater with silicate minerals (potassium feldspar, albite); in addition, the evaporation and cation exchange (calcium and magnesium ions in groundwater and sodium ions in clay minerals) may have strengthened the weak alkalinity conditions (Dehbandi et al. 2019). The average values of the TDS of the four clusters are 685.4 mg/L, 690.1 mg/L, 3430.1 mg/L, and 7646.7 mg/L, respectively. According to the groundwater classification standard based on TDS (Fetter 1994). Cluster 1 and cluster 2 can be classified as the fresh water (TDS < 1000 mg/L) except for the TDS value of several samples (Q3, Q16, Q34, Q42, Q62, Q63, Q64) slightly greater than 1000 mg/L. Half of the samples in cluster 3 belong to brackish water (1000 mg/L < TDS < 3000 mg/L), while the other half of cluster 3 and all cluster 4 belong to saline water (> 3000 mg/L). Moreover, cluster 4 (Q25, Q65, Q67) has the highest concentration of Cl− and NO3− (Fig. 4), which may be caused by the human activities such as agricultural fertilization (Li and Gao 2019). Therefore, the cluster 4 can be identified as a human activity factor.

Fig. 3
figure 3

Box plot of different parameters based on hierarchical cluster analysis (HCA)

Full size image
Fig. 4
figure 4

Piper diagram of groundwater samples for different clusters

Full size image

Correlation analysis

The correlation analysis reflects the interdependence between different hydrochemical parameters and was used to identify the similar sources of major ions with good correlation (Pant et al. 2018; Yin et al. 2021). As shown in Table 1, there is a strong positive relationship between Na+ and Cl− with the correlation coefficient of 0.78, indicating that the possibility of dissolution of halite. Na+ is significantly correlated with SO42− (correlation coefficient is 0.958), implying that the contribution of the dissolution of Glauber’s salt (Liu et al. 2015), which can be explained by the reaction (R3). Moreover, Ca2+ and SO42− show an excellent correlation with the correlation coefficient of 0.841, suggesting that the gypsum may be responsible for the hydrogeochemistry of groundwater as shown in reaction (R4). Mg2+ and SO42− also show the positive relationship with the correlation coefficient of 0.645, implying the dissolution of sulfur-bearing minerals.

Table. 1 Correlation coefficient between different physical and chemical parameters
Full size table

Principal component analysis

Principal component analysis (PCA) extracted four principal components (PCs) with eigenvalues greater than 1 as shown in Table 2 and Fig. S4, and the cumulative explained variance accounted for 83.78%, indicating that PCA significantly reduced the dimensionality of the original complex hydrochemical datasets. Table 2 shows that there is a positive or negative relationship between most parameters and PCs, although some parameters have no significant correlation with the PCs. PC1 explains 55.71% of the total variance and shows a significant correlated relationship with Ca2+, Mg2+, Na+, K+, SO42−, Cl−, TDS, and TH, implying the factor controlling the increase of the salinity in groundwater (Kim et al. 2020). However, PC1, PC2, and PC3 explain 11.65%, 8.83%, and 7.59% of the total variance, respectively. PC3 and PC4 are positively correlated with F− and NH4+, while PC2 is negatively correlated with Fe. It can be preliminarily inferred that PC2 and PC4 may be controlled by anthropogenic activity factors.

Table. 2 Principal component loadings and explained variance for the first four components
Full size table

It can be seen from Table 2 that the total variance explained by PC3 and PC4 is smaller than that of PC1 and PC2, which represents less importance. Therefore, PC1 and PC2 are used to represent the results of PCA to examine the rationality of using HCA for classification. Figure S4 shows the scores of all groundwater samples based on the HCA classification on the two principal components PC1 and PC2. Furthermore, four-ellipsis spatial cloud map of the four clusters is plotted. The four clusters are reasonably distinguished despite the marginal overlapping, which indicates that the classification of HCA is reasonable (Fig. S4).

Hydrogeochemical facies

The piper trilinear diagram, a graphical representation of groundwater chemistry data, is proposed by Piper (1944) and is widely used to understand the groundwater evolution (Subba Rao 2017). The piper diagram separates the anions and cations of the water samples and represents the dominant anions and cations in two triangles and then maps the combination of anions and cations in a diamond for identifying various hydrogeochemical types and hydrogeochemical facies of groundwater.

Figure 4 illustrates the hydrogeochemical facies and groundwater types. The triangle diagram of cations shows that the sodium is the leading cations in most groundwater samples, nearly half of groundwater samples are dominated by the cations of mixed type, while only few samples have cations dominated by calcium. On the anion plot, most of groundwater samples fall in the lower left corner, indicating that the bicarbonate plays a leading role in groundwater, and some of samples belong to mixed type. Moreover, only few samples are dominated by chloride, which may be due to the effect of halite dissolution and human activities. However, the overall hydrogeochemical characteristics of all samples are the sodium-dominated type in cation and bicarbonate-dominated type, respectively.

Given the characteristics of anions and cations, the overall hydrogeochemical characteristics can be expressed as the groundwater chemistry types. As shown by the central diamond in Fig. 4, the groundwater is characterized by the dominance of the alkaline earth elements (Ca + Mg) over the alkalis (Na + K) and the nearly balance of the weak acids (HCO3) and the strong acids (Cl + SO4). The diamond diagram shows that 37.9% of samples fall into the zone 1, which represents the Ca–HCO3 type. In comparison, one-third of the water samples fall into zone 3 (mixed type) and are divided into Ca–Mg–Cl type (24.2%) and Ca–Na–HCO3 type (9.1%). In particular, cluster 4 samples (Q25, Q65, Q67) are located in zone 4 and belong to Ca–Cl type. Moreover, 16.7% of samples fall into zone 2, representing the Na–Cl type. Only four samples fall into the zone 5 and are subclassified as Na–HCO3 type.

Controlling factors of hydrochemistry characteristics

Hydrochemical evolution process

Gibbs diagram (Gibbs 1970) is an effective tool in identifying the factors that control the hydrochemical characteristics of groundwater and is widely used to understand the formation process of hydrogeochemistry (Marandi and Shand 2018; Nazzal et al. 2014). The Gibbs diagram is represented by the plot of TDS versus (Cl−/Cl− + HCO3−) and TDS versus (Na+/Na+ + Ca2+) and divides the factors controlling the groundwater chemical processes into three aspects, which are evaporation crystallization, rock weathering, and precipitation, respectively. Figure 5 shows that all the cluster 3 and cluster 4 samples fall into evaporation crystallization dominance zone, indicating the possibility of evaporation and dissolution of evaporites, such as halite, gypsum, and mirabilite (Luo et al. 2018). Most of cluster 1 and cluster 2 samples fall within the rock weathering zone with the low TDS (< 1000 mg/L), suggesting that the water–rock interactions are the main hydrogeochemical formation mechanism in areas where cluster 1 and cluster 2 are located. Moreover, some of the cluster 1 and cluster 2 samples are located in evaporation dominance zone, which indicates that the possibility of evaporation or the existence of ancient brackish water (Li et al. 2018).

Fig. 5
figure 5

The plot of TDS versus Cl−/(Cl− + HOC3−) (a) and TDS versus (Na+)/(Na+ + Ca2+) (b)

Full size image

In order to further decipher the major weathering/dissolution mechanisms of the water–rock interactions processes, bivariate diagrams are used to explain which lithological minerals (carbonate, silicate, evaporite) are dissolved (Li et al. 2018; Luo et al. 2018). The sodium normalized ratios represented by Mg/Na, HCO3/Na, and Ca/Na were calculated. The relationships between (Mg/Na) and (Ca/Na), (HCO3/Na), and (Ca/Na) are plotted in the Gaillardet diagram (Gaillardet et al. 1999).

The Gaillardet diagram (Fig. 6) shows that most samples fall near the carbonate–silicate dissolution line and silicates weathering zone, suggesting that the dissolution of silicate plays a dominant role in groundwater chemistry. The albite, a common silicate, has the following weathering process (Ghouili et al. 2018):

$$ 2{\mathrm{Na}}^{+}{\left[{\mathrm{Al}\mathrm{Si}}_3{\mathrm{O}}_8+2{\mathrm{CO}}_2\right]}^{-}+11{\mathrm{H}}_2\mathrm{O}\to 2{\mathrm{Na}}^{+}+2{{\mathrm{H}\mathrm{CO}}_3}^{-}+{\mathrm{Al}}_2{\mathrm{Si}}_2{\mathrm{O}}_5{\left(\mathrm{OH}\right)}_4+{\mathrm{Mg}}^{2+}+4{\mathrm{H}}_4{\mathrm{Si}\mathrm{O}}_4 $$
(R1)
Fig. 6
figure 6

The sodium normalized bivariate plots (a) Mg/Na versus Ca/Na and (b) HCO3/Na versus Ca/Na

Full size image

Moreover, some samples fall on the transition zone of silicate weathering to carbonate weathering and the transition zone from silicates weathering to evaporites weathering, which indicates that the carbonates and the evaporites may be responsible for groundwater chemistry.

Water–rock interactions

Ion exchange

Chloro-alkaline indices (CAI)

Ion exchange is an essential process in water–rock interactions and plays a significant role in groundwater chemistry. In different hydrogeochemical conditions, an aquifer is composed of different minerals, resulting in the unique controlling factors of groundwater hydrochemistry. To better understand the ion exchange process that occurs between groundwater and its host environment during the process of residence or travel (Marghade et al. 2012), the chloro-alkaline indices (CAI-I and CAI-II) proposed by Schoeller (1977) have been widely used to identify the cation exchange process between Na, K and Ca, Mg. The formulas for calculating CAI are as follows:

$$ {\displaystyle \begin{array}{c} CAI-I=\frac{C{l}^{-}-\left(N{a}^{+}+{K}^{+}\right)}{C{l}^{-}}\\ {} CAI- II=\frac{C{l}^{-}-\left(N{a}^{+}+{K}^{+}\right)}{HC{O_3}^{-}+S{O_4}^{2-}+C{O_3}^{2-}+N{O_3}^{-}}\end{array}} $$
(17)

where the units of all ions are meq/L.

The Na+ in groundwater is replaced by the Ca2+ or Mg2+ absorbed on the aquifer medium, indicating the reverse ion exchange (Zaidi et al. 2015). Correspondingly, the CAI-I and CAI-II are all greater than 0, which resulted in the increase of Ca2+ or Mg2+ concentration in groundwater. On the contrary, negative values of CAI-I and CAI-II imply that the Ca2+ or Mg2+ in groundwater replaced by the Na+ absorbed on the aquifer medium, suggesting the ion exchange may cause the increase of Na+ in groundwater. The ion exchange processes are as follows:

  • Ion exchange: Ca2+(Mg2+) + 2Na − X → 2Na+ + Ca(Mg) − X2

  • Reverse ion exchange: 2Na+ + Ca(Mg) − X2 → Ca2+(Mg2+) + 2Na − X

The relationship of CAI-I and CAI-II were plotted to discriminate the ion exchange process, as shown in Fig. 7a. It can be seen that 89.4% of the samples fall on the bottom left of plot, indicating the ion exchange, which leads to the enrichment of Na+ in groundwater. Moreover, 10.6% of samples that falls on the upper right represents the reverse ion exchange.

Fig. 7
figure 7

Binary diagram: a CAI-II versus CAI-I, b (Na+ + Cl−) versus (Ca2+ + Mg2+–HCO3−–SO42−), c Na versus Cl, d Ca versus SO4, e Mg versus SO4, and f (Ca + Mg) versus (HCO3 + SO4), g total cation versus Na + K, and h total cation versus Ca + Mg

Full size image

In order to further identify the ion exchange between groundwater and aquifer materials, the relationship between (Na+–Cl−) and (Ca2+ + Mg2+–HCO3−–SO42−) was established to examine the ion exchange of sodium and calcium. If there is a cation exchange, the slope of the fitted line representing the above relationship should be close to − 1 (Fisher and Mullican III 1997). In this study, the slope of the fitted line of (Na+–Cl−)/(Ca2+ + Mg2+–HCO3−–SO42−) is − 1.297, and 81.8% of the groundwater samples fall on the zone of ion exchange (Fig. 7b), which further confirmed the existence of cation exchange.

Ion ratio

The stoichiometric relationship between different ions can be used to identify the source of the main ions in groundwater (Liu et al. 2018). The relationship between Na+ and Cl− would be located on the 1:1 line if the halite is the only source of the Na+ and Cl−. Figure 7c shows that some of the cluster 1, cluster 2, and cluster 3 samples are near the 1:1 line, indicating the dissolution of the halite is responsible for the groundwater chemistry. However, most of cluster 1, cluster 2, and cluster 3 samples are above the 1:1 line, representing the excess Na+, which may be the contribution of the ion exchange and the silicate weathering (Zhu et al. 2011). The sodium normalized bivariate plots (Fig. 6) indicates that there may be silicate weathering. Moreover, the ion exchange may be the source of excess Na+, which may be responsible for the less Ca2+ over SO42− (Fig. 7d). However, cluster 4 samples are below the 1:1 line, and the chlorine excess may be caused by human activities such as agricultural fertilization (Deshmukh 2013).

The ratio of Ca/SO4 would be 1:1 if there is only the dissolution of gypsum in the groundwater. Most of samples are below the 1:1 line, indicating the dominance of SO42− over Ca2+. Meanwhile, the relationship between Mg2+ and SO42− exhibits the same characteristics (Fig. 7e), that is, excess SO42−. On the one hand, the dissolution of gypsum plays a major role in groundwater chemistry, which can be explained by the negative saturation indices of gypsum. In addition, the bicarbonate-bearing rock is oversaturated, calcite and dolomite will not dissolve to produce Ca2+ and Mg2+, and the Ca2+ forming the dissolution of gypsum tends to precipitate calcite and dolomite, which further reduces the concentration of Ca2+ and Mg2+.

The ratio close to 1 between (Ca2+ + Mg2+) and (HCO3− + SO42−) indicates that the dissolution of bicarbonate-bearing rocks (calcite and dolomite) and sulfate minerals (gypsum) are responsible for groundwater chemistry (Edmunds et al. 2002). Figure 7f represents that most samples are below 1:1 line of (Ca2+ + Mg2+) versus (HCO3− + SO42−) scatter plot with the deficient Ca2+ + Mg2+ over HCO3− + SO42−, indicating that the dissolution of silicate weathering and ion exchange (exchange of Ca2+ and Mg2+ in groundwater with Na+ on clay minerals) may be responsible for groundwater chemistry (Han et al. 2013). However, few samples fall above the 1:1 line, suggesting the possibility of the reverse ion exchange, which can be explained by the (Na+–Cl−) versus (Ca2+ + Mg2+–HCO3−–SO42−) scatter plot (Fig. 7b).

In order to further reflect the controlling effect of silicate minerals weathering on groundwater chemistry, the relationship between Na+ + K+ and TZ+ is used to indicate the weathering process of silicate minerals (Subramani et al. 2010). Figure 7g shows that most of samples are located near the Na + K = 0.5 × TZ line, indicating that the silicate weathering is involved in the geochemical process. However, some of samples are deviated from the that line, which may be the cause of ion exchange.

The formation mechanism of high salinity in groundwater

The groundwater is characterized by high salinity, which may be caused by the dissolution of minerals. Analysis of the correlation between salinity and Ca2+, Mg2+, Na+, Cl−, and SO42− shows that high salinity may come from the dissolution of halite, gypsum, anhydrite, sulfate, and sulfur-containing minerals (Liu et al. 2018).

Figure 8 shows that most of cluster 1 and cluster 2 samples have a high ratio Na+/Cl−, Ca2+/Cl−, Mg2+/Cl−, and Ca2+/SO42−, while they have a low salinity (< 2000 mg/L), suggesting that the mineral weathering (silicate weathering (Fig. 8a, carbonate weathering (Fig. 8d), evaporite dissolution (Fig. 8b), or cation exchange may be responsible for the groundwater chemistry. However, cluster 3 and cluster 4 samples are featured by the high salinity, indicating the possibility of evapotranspiration, halite dissolution (Fig. 8a), or gypsum dissolution (Fig. 8d) in groundwater belongs to cluster 3 and cluster 4.

Fig. 8
figure 8

The relationship between different ion ratios and salinity. a Na+/Cl− versus salinity, b Ca2+/Cl− versus salinity, c Mg2+/Cl− versus salinity, d Ca2+/SO42− versus salinity

Full size image

Mineral dissolution processes

Saturation indices

In order to further examine the contribution of the mineral dissolution process to the groundwater hydrochemistry, the saturation indices (SI) can be used to indicate the saturation status of minerals in groundwater, when the solubility of the minerals in groundwater reaches the limit, and the ions in groundwater precipitate and reattach to the minerals. Moreover, SI can identify whether dissolution or precipitation has occurred, which was calculated by the following formula:

$$ \mathrm{SI}=\log \left(\frac{\mathrm{IAP}}{K_{\mathrm{s}}}\right) $$
(18)

where IAP represents the value of ionic activity product, Ks represents the solubility product of different minerals. When SI is greater than 0, equal to 0 and less than 0, it denotes the oversaturation, equilibrium, and unsaturated state of minerals in groundwater, respectively (Boudot et al. 1996). The unsaturated state means that the minerals still have the potential to release ions into groundwater.

The typical minerals are selected to calculate the SI of various minerals based on the hydrogeochemical characteristics. Figure 9a shows that the SI of calcite and dolomite of all samples are greater than zero, suggesting that groundwater are oversaturated and has a trend of precipitation in term of calcite and dolomite. However, the SI of gypsum, anhydrite, and fluorite are less than zero, indicating that groundwater are unsaturated. In particular, the SI of halite is far below the zero (− 4.06~− 8.6), indicating that groundwater has the capacity to dissolve halite. The dissolution processes of various minerals are as follows (García et al. 2001):

$$ \mathrm{NaCl}\to {\mathrm{Na}}^{+}+{\mathrm{Cl}}^{-} $$
(R2)
$$ {\mathrm{Na}}_2{\mathrm{SO}}_4\cdot 10{\mathrm{H}}_2\mathrm{O}\to 2{\mathrm{Na}}^{+}+{{\mathrm{SO}}_4}^{2-}+10{\mathrm{H}}_2\mathrm{O} $$
(R3)
$$ {\mathrm{Ca}\mathrm{SO}}_4\bullet 2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Ca}}^{2+}+{{\mathrm{SO}}_4}^{2-}+2{\mathrm{H}}_2\mathrm{O} $$
(R4)
$$ {\mathrm{Ca}\mathrm{SO}}_4\to {\mathrm{Ca}}^{2+}+{{\mathrm{SO}}_4}^{2-} $$
(R5)
$$ {\mathrm{Ca}\mathrm{F}}_2\to {\mathrm{Ca}}^{2+}+2{\mathrm{F}}^{-} $$
(R6)
Fig. 9
figure 9

Saturation indices of different minerals in groundwater: a saturation indices versus TDS, b box plot of saturation indices, c gypsum versus (Ca2+ + SO42−), d halite versus (Na+ + Cl−), e fluorite versus (Ca2+ + F−), and f anhydrite versus (Ca2+ + SO42−)

Full size image

Figure 9c–f illustrates the relationship of SI (Gypsum) versus (Ca2+ + SO42−), SI (halite) versus (Na+ + Cl−), SI (fluorite) versus (Ca2+ + F−), and SI (anhydrite) versus (Ca2+ + SO42−). The SI of gypsum, halite, and anhydrite show an increasing trend toward zero from cluster 1 to cluster 4, indicating a tendency toward a saturation state. Moreover, groundwater of cluster 1 and cluster 2 has the significant potential to dissolve gypsum, halite, fluorite, and anhydrite, which cause the hydrogeochemical characteristic.

Evaluation of groundwater quality for drinking

The EWQI method was used to evaluate the groundwater quality and the suitability of quaternary phreatic groundwater for drinking. According to the limits of various indicators in the standards for drinking water quality of China, the value of EWQI of all groundwater samples were calculated using Eqs. (3)–(10). Furthermore, the groundwater quality can be divided into five categories (excellent, good, medium, poor, extremely poor) based on the classification criteria of the EWQI values (Table 3), corresponding to the five ranks (I, II, III, IV, V) of the Quality Standard for Groundwater of China (Table S1, QSGC 2015). Table 4 shows the overall groundwater quality and classification results, and the EWQI values of all samples range from 16.30 to 549.81. The critical limit of EWQI is specified as 150; 6.1% and 10.6% of samples are classified as poor water and extremely poor water, suggesting that these samples are unsuitable for drinking purpose. Moreover, 47.0% and 31.8% of samples are classified as excellent water and good water, representing that these samples are suitable for drinking purpose. Meanwhile, only 4.5% of samples are classified as medium water, indicating the possibility of being used for drinking after additional treatment.

Table. 3 Evaluation results of groundwater for drinking based on the criteria of EWQI
Full size table
Table. 4 The results of criteria used to evaluate the suitability of irrigation purposes
Full size table

The results of groundwater quality categories based on the HCA show that out of 16 and 6 groundwater samples of cluster 1 are categorized as excellent water and good water, representing that the groundwater is suitable for drinking. Only 1 groundwater sample is categorized as poor water and is unsuitable for drinking, while 2 samples are categorized as medium water and can be used to drinking and agricultural purposes with the additional treatment. Moreover, all cluster 2 samples are classified as excellent water and are suitable for drinking. However, out of 7 samples of cluster 3 are categorized as poor or extremely water and are unsuitable for drinking, while only 1 sample is classified as medium water. In addition, all cluster 4 samples are categorized as extremely poor water and are unsuitable for drinking purpose.

In order to understand the spatial distribution characteristics of groundwater quality, the inverse distance weighting method in ArcGIS is used to interpolate EWQI, and the spatial distribution of EWQI is obtained. Then, the spatial distribution map of different hydrochemical parameters were plotted to further explain the spatial relationship between hydrogeochemistry and groundwater quality. Figure 10 shows that the spatial distribution characteristics of groundwater quality is characterized by high value of EWQI in the northwest and low value of EWQI in the east and south, indicating that the overall groundwater quality in the northwest of the study area is poor and not suitable for drinking, while the groundwater quality in the south of the study area is good and suitable for drinking purpose on the whole. According to the analysis of the hydrochemistry characteristics and evolution process of groundwater in the previous sections, it may be caused by the effects of cation exchange, evaporation, and precipitation. In addition, it is also related to the lithofacies paleogeographic conditions of the study area (Ding et al. 2021; Hou et al. 2008).

Fig. 10
figure 10

a–f Spatial distribution of EWQI versus hydrochemical parameters

Full size image

According to the classification limit in the Quality Standard for Groundwater of China (Table S1, QSGC 2015), the Na+, Cl−, SO42−, NO3−, TDS, and TH are classified as five ranks (I, II, III, IV, V). Figure 10 illustrates the relationship of the spatial distribution characteristics between the above six parameters and EWQI values. It can be seen that the samples with the rank 4 and rank 5 groundwater quality are located at the area where EWQI values are greater than the critical limit (150), indicating a good correlation between groundwater hydrochemical characteristics and groundwater quality based on EWQI.

Evaluation of groundwater quality for irrigation

Irrigation water quality parameters

The sodium adsorption ratio (SAR), soluble sodium percentage (%Na), residual sodium carbonate (RSC), Kelly’s ratio (KR), magnesium hazard (MH), and permeability index (PI) were calculated by Eqs. (11)–(16), as shown in Table 4. The RSC in the study area ranges from − 108.61 to 6.70 with the average of − 6.39. It can be seen that most of samples belong to safe category (77.27%), while some samples belong to the category of doubtful (10.61%) and unsuitable (12.12%). The KR values range from 0.10 to 6.75 with the average of 1.15, and most of samples are classified as suitable (68.18%), while the category of marginal suitable and unsuitable account for 19.7 and 12.12%. However, for the Mg hazard indicator, most of samples belong to harmful category (75.76%), while the rest of samples (24.24%) belong to good category. In addition, according to the value of PI, the water quality can be divided into three categories, which are class I (PI > 75%), class II (25% < PI < 75%), and class III (PI < 25%), respectively. The results (Table 4) show that 78.79% of groundwater samples belong to class I category, while 21.21% of samples belong to class II category. Therefore, the evaluation results based on the PI show that all groundwater samples are suitable for irrigation.

Sodium adsorption ratio and USSL diagram

The SAR evaluates the sodium hazard of excessive sodium in the water to the soil. The water with higher SAR is even more unsuitable for irrigation. Long-term use of high SAR water for irrigation will destroy the structure of soil and affect the permeability of soil, resulting in soil compaction (Salifu et al. 2017), thereby affecting crop yields. The water quality grades used for irrigation are divided into four categories according to various range of SAR. The evaluation results of irrigation water quality using SAR indicator show that all groundwater samples are excellent and good for irrigation, which accounted for 89.39 and 10.61%, respectively. Therefore, the sodium hazards are low and groundwater can be used for irrigation of most crops. Furthermore, the US salinity laboratory’s diagram (USSL diagram) is widely used to comprehensively evaluate the quality of irrigation water based on SAR (representing the sodium hazard) and EC (representing the salinity hazard). The USSL diagram is classified into 16 zones according to different classification standards of SAR and EC. Most samples fall on the zone of C3S1 (high-salinity hazard and low-sodium hazard) and C2S1 (medium-salinity hazard and low-sodium hazard), as shown in Fig. 11a, indicating that the necessity of drainage or harm to the soil and the suitability to all plants but drainage should be good, respectively (Madhav et al. 2018).

Fig. 11
figure 11

Suitability evaluation of groundwater for irrigation: a United States Salinity Laboratory (USSL) diagram and b Wilcox diagram

Full size image

Percent sodium and Wilcox diagram

Long-term use of groundwater containing high concentration of sodium will diminish the soil permeability and reduce the crop yield (Marghade et al. 2021). The %Na values range from 8.88 to 86.97 with the average value of 45.69 (Table 4). Water quality classification based on %Na shows that 78.79% of groundwater samples are suitable for irrigation. Furthermore, the classification scheme based on %Na and EC, proposed by Wilcox (1948), was used to categorize groundwater quality for irrigation. The Wilcox diagram is divided into five zones, representing excellent to good, good to permissible, permissible to doubtful, doubtful to unsuitable, and unsuitable, respectively. Figure 11b shows that 15.2% of groundwater samples belong to unsuitable category, while the “doubtful to unsuitable” category accounts for 7.6%, indicating that these two categories are unsuitable for irrigation due to the low agricultural yield irrigated by these types of water (Ramesh and Elango 2012).

Conclusions

In this paper, multivariate statistical methods and hydrogeochemical methods are employed to identify the hydrochemistry characteristics and controlling mechanisms of the phreatic groundwater in the study area. Furthermore, using the EWQI method and multiple characteristic indicators to evaluate the suitability of groundwater for drinking and irrigation purposes, the conclusions are as follows:

  1. 1.

    The hierarchical cluster analysis divides 66 groundwater samples into four clusters, and the principle component analysis is used to further examine the rationality of classification. Cluster 1 and cluster 2 are dominated by fresh water (TDS < 1000 mg/L), cluster 3 and cluster 4 are dominated by brackish water (1000–3000 mg/L)–salinity water (> 3000 mg/L). Furthermore, the Piper diagram determines the main hydrochemical facies: Ca–HCO3 type (cluster 1 and cluster 2), mixed type (mainly cluster 1 and cluster 2), Na–Cl type (mainly cluster 2 and cluster 3) and Ca–Cl type (cluster 4).

  2. 2.

    The Gibbs diagram identifies the main controlling factors, which are evaporation crystallization (cluster 3 and cluster 4) and water–rock interactions (cluster 1 and cluster 2). Moreover, the saturation index combined with chloro-alkaline indices (CAI-I, CAI-II) indicates that silicate and evaporite weathering, dissolution of minerals (halite, gypsum, and anhydrite), and cation exchange are responsible for groundwater hydrochemistry. In addition, human activities may contribute to groundwater chemistry characteristics.

  3. 3.

    The entropy water quality index method is used to evaluate the suitability of groundwater for drinking purpose and combined with ArcGIS to analyze the spatial distribution characteristics. The results reveal that groundwater quality is categorized as the excellent (47.0%), good (31.8%), medium (4.5%), poor (6.1%), and extremely poor (10.6%) water types, respectively. Therefore, 16.7% of samples are unsuitable for drinking. On the whole, the water quality in the southern part of the study area is better than that in the northwest.

  4. 4.

    Multiple characteristic indicators are used to evaluate the suitability of groundwater for irrigation purpose. The USSL diagram shows that most of samples belong to C3S1 (high-salinity hazard and low-sodium hazard) and C2S1 (medium-salinity hazard and low-sodium hazard), indicating that the necessity of drainage or harm to the soil and the suitability to all plants but drainage should be good. Moreover, the Wilcox diagram shows that 22.8% of groundwater samples are doubtful or unsuitable for irrigation due to the high salinity of groundwater. Therefore, it is necessary to strengthen the continuous monitoring of groundwater quality and pay more attention to the variations of groundwater quality in the future for sustainable development and human health.