Human Activity and Hydrogeochemical Processes Relating to Groundwater Quality Degradation in the Yuncheng Basin, Northern China

Groundwater quality degradation has raised widespread concerns about water supplies and ecological crises in China. In this study, hydrogeochemistry, environmental stable isotopes (δ18O, δD), and principal component analysis were conducted together to reveal the mechanism’s response to the hydrogeochemical and quality degradation of groundwater in Yuncheng Basin, Northern China, so that reasonable water resource management strategies can be developed. The study reveals that groundwater faces a tremendous risk of quality decrease during the past decade: (1) the hydrochemical facies of groundwater shows that the bicarbonate and chloride type water was replaced with sulfate type water and the occupying area of SO4·Cl-Na, SO4·HCO3-Na type water expanded dramatically in shallow and intermediate-deep aquifers. (2) Major ion chemistry and hydrogen and oxygen isotope compositions indicate that the major hydrogeochemical processes responsible for groundwater quality deterioration include the dissolution of evaporates (i.e., halite, gypsum, and mirabilite), ion exchange, and evaporation process. Additionally, (3) anthropogenic activities (overutilization of fertilizer) have resulted in nitrate contamination, and have thereby led to groundwater quality degradation.


Introduction
Groundwater is a vital part of the total water resource for domestic, agricultural, and industrial purposes in Northern China, which is one of the world's most water-scarce regions [1][2][3]. The study area in Yuncheng Basin, located in Shanxi province, is a typical semi-arid area in Northern China. Along with the underdeveloped surface water system, groundwater has become the primary water resource. With rapid population growth and economic development, the unreasonable exploitation and utilization of groundwater has brought about a series of environmental and geological issues such as the decrease in the groundwater level, the enlargement of the scope of groundwater salinization, the intrusion of a salt lake, and groundwater quality degradation [4,5]. Thus, a comprehensive understanding of the geochemical evolution and the factors affecting groundwater quality is crucial to guarantee the safety of groundwater consumption and the rational management of groundwater resources.
Basically, groundwater hydrogeochemistry is controlled by water-rock interactions (i.e., mineral dissolution/precipitation and ion-exchange) within the groundwater system, while human activities,

Materials and Methods
A total of 183 groundwater samples, including 51 Table S1). The samples were stored in 250 mL polyethylene bottles that had first been pre-cleaned with deionized water in the laboratory and then with the extracted water at least three times. Before sampling, groundwater was pumped over one hour. During sampling, T, pH, electrical conductivity (EC), and oxidation-reduction potential (ORP) were

Materials and Methods
A total of 183 groundwater samples, including 51 Table S1). The samples were stored in 250 mL polyethylene bottles that had first been pre-cleaned with deionized water in the laboratory and then with the extracted water at least three times. Before sampling, groundwater was pumped over one hour. During sampling, T, pH, electrical conductivity (EC), and oxidation-reduction potential (ORP) were measured using a portable Hanna pH and EC meter. Alkalinity was measured using the Gran titration method on the sampling day. All groundwater samples were filtered through 0.45 µm membranes on site. For cation and trace element analysis, reagent-quality nitric acid (HNO 3 ) was added to the polyethylene bottles until the pH < 2. For anions analysis, the samples were stored without acidification.
Concentrations of major cations and anions were analyzed within two weeks after sampling, using ion chromatography (IC) (ICS2100, Thermo Fisher, Massachusetts, MA, USA). Trace elements were determined by quadrupole-inductively coupled plasma-mass spectrometry (Q-ICP-MS) (Agilent 7500a ICP-MS instrument, Agilent Technologies, Tokyo, Japan). The analytical precision for all ions concentration measurements was indicated by the ionic balance error, which was better than the standard limit of 5%. δD and δ 18

General Hydrogeochemistry
The major properties and hydrogeochemistry of groundwater samples were summarized in Table 1. Groundwater samples had T from 15.8 to 39 • C, with the highest temperature observed in deep groundwater samples (XX-17) from Xiaxian County, which were affected by the deep geothermal fluid [29]. pH values ranged from 6.80 and 9.02, indicating neutral to slightly alkaline conditions. Total dissolved solid (TDS) values varied from 349 and 9590 mg/L and from 205 to 14,051 mg/L for shallow and intermediate-deep groundwater samples, respectively. Over 70% of shallow groundwater and 50% of intermediate-deep groundwater belonged to the brackish or saline water category according to the classification of Reference [30]. The ionic compositions were dominated by Na + (8.28-3638 mg/L), SO 4 2− (11.51-9214 mg/L), and HCO 3 − (15.39-1953 mg/L). According to our previous study [31], the Yuncheng Salt lake water is characterized by SO 4 ·Cl-Na type with a slightly alkaline pH (7.9) and a very high TDS value (11,050 mg/L). The hydrochemical types of both shallow and intermediate-deep groundwater displayed certain patterns along the flow path from recharge areas to flow-through and discharge areas (Supplementary Material Figures S1 and S2). Low TDS groundwater from the mountain front area was mainly HCO 3 -Na, HCO 3 -Na·Mg, and HCO 3 -Ca·Mg type. Groundwater with high TDS values collected from the center of Yuncheng Basin was generally SO 4 -Na, SO 4 ·Cl-Na, SO 4 ·Cl-Na·Mg, and SO 4 -Na·Mg type water. The groundwater from the flow-through areas have variable water types, including HCO 3 -Na, HCO 3 ·SO 4 -Na, SO 4 ·HCO 3 -Na, SO 4 ·HCO 3 -Na·Mg, SO 4 ·Cl-Na, and SO 4 ·Cl-Na·Mg type.
The correlation analysis between major hydrochemical parameters was applied in this study (Supplementary Material Tables S2 and S3) to understand the relationships between different ionic species. EC values did exhibit a positive correlation with TDS, Na + , Ca 2+ , Mg 2+ , Cl − , SO 4 2− , and HCO 3 − , indicating that water-rock interactions played a key role in groundwater mineralization [31][32][33].
The observed well-defined correlation between TDS and Na + , Mg 2+ , Cl − , and SO 4 2− suggested that processes controlling these ion compositions were, in part, related to the controls on salinity.   The distribution pattern of groundwater with different hydrochemical facies is illustrated for the year 2005 [34] and 2015 ( Figure 2). For shallow groundwater, the major water types are HCO 3 -Na, HCO 3 -Na·Ca, HCO 3 ·Cl-Na, and HCO 3 ·SO 4 -Na, which are mainly found in the West and Northeast parts of the Basin in 2005 (Figure 2a). The Cl·SO 4 -Na type water are mainly presented in the surrounding areas of the Yongji city and the Salt Lake. The sulfate type water includes SO 4 ·Cl-Na, SO 4 ·HCO 3 -Na, SO 4 ·HCO 3 -Na·Mg, and SO 4 ·Cl-Na·Mg type, and scattered in the central basin between Yongji and Yuncheng city, and the belt area close to the Yellow river.
The hydrochemical facies of shallow groundwater shows a significant change in 2015, compared to that in 2005. First, the Cl·SO 4 -Na type water was replaced with the sulfate type water. Then, the occupying area of SO 4 ·Cl-Na, SO 4 ·HCO 3 -Na type water extended dramatically, as shown in Figure 2b. Since this sulfate type water all belongs to saline groundwater, the degradation of groundwater quality is obvious. For the intermediate-deep groundwater in Yuncheng basin, after 10 years of consumption, the distribution areas of HCO 3 -Na, HCO 3 ·SO 4 , and Cl·SO 4 type water were reduced, while the sulfate type water, SO 4 ·Cl, SO 4 ·HCO 3 , and SO 4 , expanded its territory (Figure 2c,d).
It is regretful to see that the intermediate-deep groundwater also faces the risk of quality decrease in the area. The distribution pattern of groundwater with different hydrochemical facies is illustrated for the year 2005 [34] and 2015 ( Figure 2). For shallow groundwater, the major water types are HCO3-Na, HCO3-Na·Ca, HCO3·Cl-Na, and HCO3·SO4-Na, which are mainly found in the West and Northeast parts of the Basin in 2005 (Figure 2a). The Cl·SO4-Na type water are mainly presented in the surrounding areas of the Yongji city and the Salt Lake. The sulfate type water includes SO4·Cl-Na, SO4·HCO3-Na, SO4·HCO3-Na·Mg, and SO4·Cl-Na·Mg type, and scattered in the central basin between Yongji and Yuncheng city, and the belt area close to the Yellow river.
The hydrochemical facies of shallow groundwater shows a significant change in 2015, compared to that in 2005. First, the Cl·SO4-Na type water was replaced with the sulfate type water. Then, the occupying area of SO4·Cl-Na, SO4·HCO3-Na type water extended dramatically, as shown in Figure  2b. Since this sulfate type water all belongs to saline groundwater, the degradation of groundwater quality is obvious. For the intermediate-deep groundwater in Yuncheng basin, after 10 years of consumption, the distribution areas of HCO3-Na, HCO3·SO4, and Cl·SO4 type water were reduced, while the sulfate type water, SO4·Cl, SO4·HCO3, and SO4, expanded its territory (Figure 2c,d). It is regretful to see that the intermediate-deep groundwater also faces the risk of quality decrease in the area.

Stable Isotopes Oxygen and Hydrogen
The stable isotopic compositions of hydrogen and oxygen in the groundwater samples from Yuncheng basin are presented in Table 1. Stable isotopic composition of the shallow groundwater samples varied from −10.14‰ to −8.18‰ for δ 18 O and from −75.57‰ to −62.93‰ for δD, respectively. Intermediate and deep groundwater samples showed relatively depleted stable isotopic values, ranging from −11.34‰ to −8.36‰ for δ 18 O and from −81.87‰ to −61.80‰ for δD, respectively.

Stable Isotopes Oxygen and Hydrogen
The stable isotopic compositions of hydrogen and oxygen in the groundwater samples from Yuncheng basin are presented in Table 1. Stable isotopic composition of the shallow groundwater samples varied from −10.14% to −8.18% for δ 18 O and from −75.57% to −62.93% for δD, respectively.
The standard diagram of δ 18 O-δD diagram ( Figure 3) showed the position of all groundwater samples relative to the meteoric water lines. It was observed that all the samples were plotted close to the meteoric water lines, indicating the meteoric origin in the study area. The δ 18 O and δD values for shallow groundwater defined a regression line: δD = 6.22δ 18 O − 11.51. Compared with the meteoric water lines, the relatively lower slope implied that evaporation through a dry surface layer occurred with a low moisture condition. The standard diagram of δ 18 O-δD diagram ( Figure 3) showed the position of all groundwater samples relative to the meteoric water lines. It was observed that all the samples were plotted close to the meteoric water lines, indicating the meteoric origin in the study area. The δ 18 O and δD values for shallow groundwater defined a regression line: δD = 6.22δ 18 O − 11.51. Compared with the meteoric water lines, the relatively lower slope implied that evaporation through a dry surface layer occurred with a low moisture condition.

Extraction of Principal Components by PCA
For the shallow groundwater samples, three components, accounting for 91.6% of the total variance in the dataset, were extracted (Supplementary Material Table S4). The first component (PC1) explained 57% of total variance and was represented by SO4 2− , Cl − , Na + , and Mg 2+ (Figure 4). The associations of these ions contributed to groundwater salinity and were mainly derived from evaporate dissolution. Therefore, PC1 can be seen as the impact of evaporate dissolution on groundwater chemistry. The higher the PC1 scores, the greater the impacts of evaporate dissolution. PC2 accounted for 20.1% of total variance and was characterized by high associations of Ca 2+ and HCO3 − , indicating the influence of carbonate weathering. Component three (PC3) explained 14.5% of total variance and showed a high correlation of K + . The source of K + could be the weathering dissolution of K-bearing silicate minerals, such as K-felspar and/or cation exchange.
For the intermediate-deep groundwater samples, two principle components were extracted to explain 75.83% of the total variance (Supplementary Material Table S5, Figure 4). PC1 accounted for 60.3% of total variance with the associations of Na + , Ca 2+ , Mg 2+ , SO4 2− , and Cl − , reflecting the influence of evaporate and carbonate dissolution. PC2 explained 15.55% of total variance and showed high correlations of K + and HCO3 − , suggesting the controlling role of K-baring silicate mineral dissolution and/or cation exchange on groundwater chemistry.

Extraction of Principal Components by PCA
For the shallow groundwater samples, three components, accounting for 91.6% of the total variance in the dataset, were extracted (Supplementary Material Table S4). The first component (PC1) explained 57% of total variance and was represented by SO 4 2− , Cl − , Na + , and Mg 2+ (Figure 4). The associations of these ions contributed to groundwater salinity and were mainly derived from evaporate dissolution. Therefore, PC1 can be seen as the impact of evaporate dissolution on groundwater chemistry.  Figure 4). PC1 accounted for 60.3% of total variance with the associations of Na + , Ca 2+ , Mg 2+ , SO 4 2− , and Cl − , reflecting the influence of evaporate and carbonate dissolution. PC2 explained 15.55% of total variance and showed high correlations of K + and HCO 3 − , suggesting the controlling role of K-baring silicate mineral dissolution and/or cation exchange on groundwater chemistry.

Mechanisms of Natural Processes Controlling Groundwater Chemistry
The Gibbs plot proved to be a useful representation to assess the natural mechanisms controlling groundwater chemistry [38][39][40][41]. In the diagram, TDS values are shown against the weight ratio of Na + /(Na + + Ca 2+ ) for cations and Cl − /(Cl − + HCO3 − ) for anions ( Figure 6). All the groundwater samples fell into both the evaporation and rock weathering dominant area, indicating the predominant role of evaporation and rock weathering on groundwater chemistry in the study area. Most shallow groundwater samples were positioned in the evaporation dominant zone, suggesting the significant effect of evaporation and/or dissolution of evaporates [42,43] on shallow groundwater hydrogeochemistry.

Mechanisms of Natural Processes Controlling Groundwater Chemistry
The Gibbs plot proved to be a useful representation to assess the natural mechanisms controlling groundwater chemistry [38][39][40][41]. In the diagram, TDS values are shown against the weight ratio of Na + /(Na + + Ca 2+ ) for cations and Cl − /(Cl − + HCO 3 − ) for anions ( Figure 6). All the groundwater samples fell into both the evaporation and rock weathering dominant area, indicating the predominant role of evaporation and rock weathering on groundwater chemistry in the study area. Most shallow groundwater samples were positioned in the evaporation dominant zone, suggesting the significant effect of evaporation and/or dissolution of evaporates [42,43] on shallow groundwater hydrogeochemistry. The observed good, positive correlation between Na + and SO 4 2− (r = 0.87), Cl − (r = 0.74) (Table S3) and the high concentrations of these major ions in shallow groundwater indicated that evaporates (i.e., halite and mirabilite [44]) dissolution was the major process controlling the shallow groundwater chemistry. The well-defined correlation between Na + and SO 4 2− (r = 0.93) in the intermediate-deep aquifers suggested that the dissolution of mirabilite is the major source of these ions in the groundwater. An insignificant correlation between HCO 3 − and TDS contents (Table S3 and S4) was observed for both shallow and deep groundwater. Bicarbonate mainly came from carbonate and/or silicate minerals in natural waters. However, the poor correlation between HCO 3 − and Ca 2+ (Tables S3 and S4, Figure 7d) indicated that Ca 2+ was significantly altered by other geochemical processes. The Quaternary sediments at Yuncheng Basin were made up of aeolian loess, lacustrine clays, fluvial sands and gravels, mixed with halite, mirabilite, and gypsum [4,34,45]. Combined with the low correlation between Ca 2+ and SO 4 2− in groundwater, we can conclude that gypsum was considered as one, but not the only, significant contributor to the Ca 2+ contents. To evaluate the contribution of gypsum to calcium in groundwater, the correlation between (Ca 2+ + Mg 2+ ) and (SO 4 2− + HCO 3 − ) was illustrated in Figure 7f. If the (Ca 2+ + Mg 2+ )/(SO 4 2− + HCO 3 − ) ratios were equal to or lower than 1, those ions were dominantly controlled by Ca-salt (i.e., calcite, dolomite, and gypsum) dissolution. The ratios higher than 1 reflected other sources for Ca, such as reverse cation exchange and/or silicate weathering [46]. As seen in Figure 7f, The observed good, positive correlation between Na + and SO4 2-(r = 0.87), Cl -(r = 0.74) (Table S3) and the high concentrations of these major ions in shallow groundwater indicated that evaporates (i.e., halite and mirabilite [44]) dissolution was the major process controlling the shallow groundwater chemistry. The well-defined correlation between Na + and SO4 2-(r = 0.93) in the intermediate-deep aquifers suggested that the dissolution of mirabilite is the major source of these ions in the groundwater. An insignificant correlation between HCO3 -and TDS contents (Table S3 and S4) was observed for both shallow and deep groundwater. Bicarbonate mainly came from carbonate all the groundwater samples were saturated with calcite and dolomite, and under-saturated with halite ( Figure 8). The majority of groundwater shows a positive value of the saturation index of gypsum. The saturation indices of calcite, dolomite, halite, and gypsum show an increasing trend with increasing TDS concentrations, illustrating the major involvement of the dissolution of carbonates and evaporates. Cation exchange was considered to be another possible factor controlling the hydrochemical compositions of groundwater at Yuncheng Basin. The plot of Na + against Cl − (Figure 7b) showed that some groundwater samples were positioned with a slope of 1 (halite dissolution), illustrating that Na + and Cl − was derived predominantly from halite. The majority of groundwater were placed above the dissolution line, which can be explained by the contribution of Na-containing minerals dissolution and/or cation exchange. The controlling role of ion exchange for the formation of groundwater chemistry can be understood by the bivariate diagram of [(Ca 2+ + Mg 2+ )-(SO 4 2− + HCO 3 − )] against (Na + -Cl − ). The (Na + -Cl − ) is related to the amount of Na + gained or lost from sources other than the dissolution of chloride salts, whereas [(Ca 2+ + Mg 2+ )-(SO 4 2− + HCO 3 − )] showed the sum of Ca and Mg gained or lost relative to that provided by the dissolution of calcite, dolomite, and gypsum. If cation exchange was a significant factor responsible for the groundwater chemistry, the relation between [(Ca 2+ + Mg 2+ )-(SO 4 2− + HCO 3 − )] and (Na + -Cl − ) should be linear with a slope of ± 1.0 [47]. Figure 7g depicted a straight line with a slope of −0.99 for all groundwater samples, clearly pointing out the existence of Na + exchange with Ca 2+ and Mg 2+ . Furthermore, two chloro-alkaline indices, CAI1 (CAI1 = (Cl − -(Na + + K + ))/Cl − ) and CAI2 (CAI2 = (Cl − -(Na + + K + ))/(SO 4 2− + HCO 3 − + CO 3 − + NO 3 − )) [48], were used to indicate the ion exchange between the groundwater and its host environment. The CAI1 and CAI2 values of majority of groundwater samples were negative, suggesting that ion exchange was one of the dominant processes in groundwater (Figure 9).  Cation exchange was considered to be another possible factor controlling the hydrochemical compositions of groundwater at Yuncheng Basin. The plot of Na + against Cl − (Figure 7b) showed that some groundwater samples were positioned with a slope of 1 (halite dissolution), illustrating that Na + and Cl − was derived predominantly from halite. The majority of groundwater were placed above the dissolution line, which can be explained by the contribution of Na-containing minerals dissolution and/or cation exchange. The controlling role of ion exchange for the formation of groundwater chemistry can be understood by the bivariate diagram of [(Ca 2+ + Mg 2+ )-(SO4 2− + HCO3 − )] against (Na + -Cl − ). The (Na + -Cl − ) is related to the amount of Na + gained or lost from sources other than the dissolution of chloride salts, whereas [(Ca 2+ + Mg 2+ )-(SO4 2− + HCO3 − )] showed the sum of Ca and Mg gained or lost relative to that provided by the dissolution of calcite, dolomite, and gypsum. If cation exchange was a significant factor responsible for the groundwater chemistry, the relation between [(Ca 2+ + Mg 2+ )-(SO4 2− + HCO3 − )] and (Na + -Cl − ) should be linear with a slope of ± 1.0 [47]. Figure  7g depicted a straight line with a slope of −0.99 for all groundwater samples, clearly pointing out the existence of Na + exchange with Ca 2+ and Mg 2+ . Furthermore, two chloro-alkaline indices, CAI1 (CAI1 = (Cl − -(Na + + K + ))/Cl − ) and CAI2 (CAI2 = (Cl − -(Na + + K + ))/(SO4 2− + HCO3 − + CO3 − + NO3 − )) [48], were used to indicate the ion exchange between the groundwater and its host environment. The CAI1 and CAI2 values of majority of groundwater samples were negative, suggesting that ion exchange was one of the dominant processes in groundwater (Figure 9).
Evaporation played an important role in the chemistry of groundwater. Judged from the Gibbs diagrams (Figure 6), the chemical compositions of some shallow groundwater were governed by evaporation process. The plot of δ 18 O and δD values further constrains the effect of evaporation on the shallow groundwater chemistry (Figure 3). The relationships between Cl and δ 18 O can also  [29]. Additionally, it is interesting to note that three shallow groundwater show relatively lower Cl concentrations (< 300 mg/L) and depleted δ 18 O values (<−9.4‰). Conferring the sample numbers to their sampling site, we note that they are all collected in the flow-through and discharge areas, indicating the probable influence of up-coning of deep groundwater. Figure 9. Scatter plot of CAI1 versus CAI2 for the groundwater samples in the study area. Figure 9. Scatter plot of CAI1 versus CAI2 for the groundwater samples in the study area.
Evaporation played an important role in the chemistry of groundwater. Judged from the Gibbs diagrams (Figure 6), the chemical compositions of some shallow groundwater were governed by evaporation process. The plot of δ 18 O and δD values further constrains the effect of evaporation on the shallow groundwater chemistry (Figure 3). The relationships between Cl and δ 18 O can also provide the potential for hydrogeochemical processes in the intermediate-deep groundwater [49][50][51][52]. From Figure 10, four trends can be detected for the groundwater samples: (I) changes in Cl contents with no accompanying variation of δ 18 O. Plenty of the deep groundwater were grouped in this group, illustrating the dissolution of evaporate during irrigation leaching [29]. During the process of leaching, little or no changes in stable isotopic composition of the deep groundwater take place due to low evaporation rates and high vertical recharge. Simultaneously, continued evaporite dissolution will lead to elevated Cl concentrations.  [29]. Additionally, it is interesting to note that three shallow groundwater show relatively lower Cl concentrations (< 300 mg/L) and depleted δ 18 O values (<−9.4% ). Conferring the sample numbers to their sampling site, we note that they are all collected in the flow-through and discharge areas, indicating the probable influence of up-coning of deep groundwater. to low evaporation rates and high vertical recharge. Simultaneously, continued evaporite dissolution will lead to elevated Cl concentrations. (II) Simultaneous increase of δ 18 O and Cl values. This group included shallow groundwater samples affected by evaporation. Groundwater Cl and δ 18 O values would be expected to increase during evaporation process (i.e., recharge process or directly from the shallow water tables). (III) Changes in δ 18 O values without variation of Cl contents. The third group included plenty of deep groundwater samples and all the samples from the intermediate aquifers.
The large variation in oxygen isotopic compositions with little change in Cl concentration could be due to mixing with the lateral recharge water characterized by low Cl contents and varying δ 18 O values. (IV) Slight increase in δ 18 O values with low to medium Cl contents. This group consisted of several deep groundwater, indicating the existence of an additional recharge source, probably shallow groundwater [29]. Additionally, it is interesting to note that three shallow groundwater show relatively lower Cl concentrations (< 300 mg/L) and depleted δ 18 O values (<−9.4‰). Conferring the sample numbers to their sampling site, we note that they are all collected in the flow-through and discharge areas, indicating the probable influence of up-coning of deep groundwater.

Anthropogenic Factors Affecting Groundwater Chemistry
Considering that Yuncheng city is one of the major crop yield areas in Shanxi Province, Northern China, agriculture activity is the dominant anthropogenic factor response for groundwater quality

Anthropogenic Factors Affecting Groundwater Chemistry
Considering that Yuncheng city is one of the major crop yield areas in Shanxi Province, Northern China, agriculture activity is the dominant anthropogenic factor response for groundwater quality degradation. Due to the absence of nitrate in natural aquifers, nitrate contamination was largely due to agricultural activities. It can be observed that the NO 3 − concentration of groundwater samples in the study area ranged from 1.16 to 135 mg/L. Over 15% of the shallow groundwater and 10% of intermediate-deep groundwater had NO 3 − concentrations above the WHO drinking guideline (50 mg/L) [53]. Therefore, strict policies and proper management should be applied to reduce the environmental nitrate contamination.

Conclusions
Integrated analysis of hydrochemistry, environmental isotopes, and multivariate statistical analysis are applied in this study to better understand the major geochemical processes response to groundwater quality degradation in the Yuncheng Basin, Northern China.
1. Groundwater TDS values increase gradually from the mountain front recharge area to the discharge area in the central basin. Along the flow path, groundwater hydrochemical types display certain patterns of variability. The major hydrochemical facies in shallow unconfined aquifers are HCO 3 , HCO 3 ·SO 4 , SO 4 ·HCO 3 , SO 4 ·Cl, and Cl·SO 4 -type. The major hydrochemical facies in deep confined aquifers are HCO 3 , HCO 3 ·SO 4 , SO 4 ·HCO 3 , SO 4 , SO 4 ·Cl, and Cl·SO 4 -type. Additionally, the occupying area of SO 4 ·Cl-Na, SO 4 ·HCO 3 -Na type water expanded dramatically in both shallow and intermediate-deep aquifers.
2. Water-rock interactions, including the dissolution of evaporates (halite, gypsum, and mirabilite), ion exchange and evaporation, have a predominant role on groundwater quality degradation. Salt Lake water intrudes into shallow groundwater to some extent. Some deep groundwater near the faults is deteriorated by the mixing of geothermal waters.
3. Overutilization of fertilizer is another important factor response for the deterioration of groundwater quality in the study area. Reasonable strategies for fertilizer usage should be made to protect groundwater resources.
Supplementary Materials: The following are available online at http://www.mdpi.com/1660-4601/17/3/867/s1; Figure S1. Piper diagram of shallow groundwater samples; Figure S2. Piper diagram of deep groundwater samples; Table S1. Groundwater sample numbers, locations and depths; results of groundwater chemical and isotopic analyses; Table S2. Pearson correlation coefficients of major hydrochemical parameter in shallow groundwater; Table S3. Pearson correlation coefficients of major hydrochemical parameter in intermediate-deep groundwater; Table S4. Variance explained by the first three principal components in shallow groundwater samples; Table S5. Variance explained by the first three principal components in intermediate-deep groundwater samples.

Conflicts of Interest:
The authors declare that there is no conflict of interest regarding the publication of this article.