Characteristics and Controlling Factors of Groundwater Hydrochemistry in Dongzhi Tableland Area of the Loess Plateau of Eastern Gansu—A Case Study of Ning County Area, North China

Abstract

Groundwater plays an irreplaceable role in all aspects of the Loess Plateau. In this study, the loess phreatic water (LPW) and bedrock phreatic water (BPW) in the Ning County area (NCA) were sampled and analyzed, and the characteristics and controlling factors of groundwater were determined by using statistical analysis, hydrochemical methods, and hydrogeochemical simulation. The results indicated that the groundwater in the NCA was alkaline as a whole, and the average pH values of LPW and BPW were 8.1 and 7.8, respectively. The mean values of TDS concentrations of LPW and BPW were 314.9 mg/L and 675.3 mg/L, and the mean values of TH contents were 194.6 mg/L and 286.6 mg/L, respectively, which were mainly divided into hard fresh water. The Piper diagram illustrated that the hydrochemical type of groundwater in the NCA was mainly the HCO₃·Ca type. The main recharge source of groundwater was atmospheric precipitation, and it was affected by evaporation to a certain extent. The linear relationships of δ¹⁸O and δ²H of LPW and BPW were δ²H = 6.998δ¹⁸O − 3.802 (R² = 0.98) and δ²H = 6.283δ¹⁸O − 10.536 (R² = 0.96), respectively. Hydrochemical analysis indicated that the groundwater in the NCA was mainly controlled by rock weathering and cation exchange. BPW was affected by the dissolution of gypsum. The possible mineral phases were identified on the basis of the main soluble minerals in the aquifer, and hydrogeochemical reverse simulations were performed. The dissolution of calcite, illite, and hornblende, and the precipitation of dolomite, plagioclase, and microcline occurred on both the LPW and BPW pathways.

1. Introduction

Water resources play a pivotal role in human production and life [1,2,3]. Industrial production, agricultural irrigation, residential life, and ecological maintenance all depend on stable and clean water sources [4,5,6,7,8]. As an important component of water resources, groundwater has the characteristics of stable water quantity and excellent water quality [9]. Groundwater is buried at a certain depth, easy to be exploited and not easily polluted, and has become the preferred source of water for residential and industrial use. Population growth and rapid socio-economic growth have led to increased consumption of groundwater resources and frequent occurrence of water pollution [10,11,12,13,14]. However, the contradiction between supply and demand of water resources is serious in many regions of the world. Problems such as shortage of groundwater resources, deterioration, and imbalance of groundwater resources limit the sustainable and healthy development of society and economy, having caused serious environmental problems in the past few decades, such as land subsidence, water environment pollution, and water ecosystem degradation [15,16,17,18]. To realize the sustainable utilization of regional groundwater resources and promote the coordinated development of the ecological environment and social economy, it is necessary to recognize the hydrochemical characteristics and causes of groundwater resources, as well as to develop and utilize groundwater resources scientifically and rationally [19,20,21].

The hydrochemical components of groundwater are affected by the long-term action of the environment and continuously interact with the surrounding medium during the migration process, resulting in a series of physical and chemical effects [22,23,24]. The chemical composition of water contains information such as climate change and the origin of water bodies. Therefore, the environmental quality of groundwater and its ecological functions can be understood by analyzing the chemical composition of water [25,26,27]. Groundwater under different hydrogeological conditions and human activities has different hydrochemical characteristics. Analyzing the hydrochemical characteristics of groundwater can effectively discriminate the formation process of hydrochemical components and reveal important information such as chemical behavior of elements and chemical weathering intensity in groundwater [28,29]. Therefore, the research on the formation mechanism of groundwater hydrochemistry has important practical significance.

The Chinese Loess Plateau (CLP) is located in the hinterland of China, in the middle reaches of the Yellow River. Affected by the continental monsoon climate, most of its areas belong to arid–semi-arid regions [30]. The CLP is also the region with the most serious soil erosion and the most fragile ecology in China, and the available surface water is very short [31]. The CLP has the widest distribution of loess in the world, and the thickest part of the soil layer is more than 350 m [32]. The deep soil layer forms a huge soil reservoir and groundwater reservoir, providing conditions for the exploitation of groundwater. For a long time, groundwater has become an important source of water supply for agricultural production, industrial development, and residents’ domestic use in the CLP. Many researchers have carried out a large number of discussions on the study of groundwater in the Loess Plateau. For example, Li et al. [33] analyzed the hydrogeological conditions and groundwater characteristics, clarified the types and distribution characteristics of freshwater, and revealed the formation mechanism and influencing factors of freshwater on the CLP. Ling et al. [34] studied the isotopic significance of groundwater recharge, retention time, and hydrogeochemical evolution in the CLP of Eastern Gansu. Wang et al. [35] determined the spatial and seasonal variation characteristics and controlling factors of fluorine in natural water on the CLP and assessed its health risks. Xiao et al. [36] analyzed the characteristics and sources of trace elements in river water and well water in the CLP and evaluated water quality and health risks. Zhang et al. [37] studied the hydrochemical processes between surface water and groundwater in the northeastern CLP: implications for water chemistry and environmental evolution in a semi-arid region.

However, the research on the formation mechanism and evolution law of groundwater in the Ning County area (NCA) is relatively limited. Therefore, the purpose of this study was (1) to analyze the hydrochemical characteristics of groundwater in the area, and (2) to clarify the formation mechanism of groundwater hydrochemistry. Statistical analysis, hydrochemical methods, and hydrogeochemical modeling were comprehensively applied in this study, and the research results can provide some reference for the rational utilization of regional groundwater resources and ecological environment protection in the CLP or other similar regions in the world.

2. Study Area

The study area (Figure 1) is located in the middle reaches of the Malian River on the CLP of Eastern Gansu. Geographical coordinates: 35°30′–35°45′ east longitude, 107°45′–108°00′ north latitude. The region is located in the mid-latitude warm temperate zone, with a typical continental monsoon climate. The annual average precipitation is 525.8 mm, the annual maximum precipitation is 827.7 mm (1975), and the annual minimum precipitation is 357.7 mm (1995). The rainfall is mainly concentrated in July–September (Figure 2), and the spatial distribution of rainfall is extremely uneven after the flood season.

The study area belongs to Longdong Basin, located in the southwest of Ordos Basin, which is a Mesozoic asymmetric syncline basin with a south–north strike, gentle in the east and steep in the west (Figure 3). Its western boundary is the Zhuozi mountain-Pingliang fault zone, the southern part is the Weibei uplift, the eastern boundary is the Ziwu Ridge, and the northern boundary is the Baiyu mountain. Affected by the sedimentation of the CLP and the convoluted structure in Longxi, most of the area is covered by the Quaternary. In addition, the drilling data in the study area indicate that the strata mainly include the Quaternary Pleistocene loose layer, Cretaceous Huanhe Formation sand mudstone interbed, Cretaceous Luohe Formation medium and coarse sandstone, and Jurassic Anding Formation sand mudstone interbed. Loess and valley sand–gravel layer of loose rock pore aquifer and Huanhe formation bedrock weathering crust pore aquifer formation are the main groundwater exploitation layers in the study area.

The study area (Figure 1) is located in the loess gulled-hilly area in the east of Dongzhi tableland. The overall topography shows that the elevation gradually increases to both sides with the Malian River as the boundary. Quaternary loess is widely distributed in the area, with a thickness of 50–180 m. The region has ravines and gullies criss-cross, undulating terrain, and relatively simple landform, which is mainly divided into three geomorphic types: loess remnant plateau, hilly and gully, and river valley.

The flow direction of groundwater is consistent with the flow direction of surface water and generally flows from the tableland area to the nearby valley area. The recharge area is mainly distributed in the loess tableland, loess hills, and loess ridge area; the discharge area is gullies; and the Malian River valley is the final destination of diving discharge. The dynamic change of phreatic water in the loess tableland has obvious regional and seasonal changes, and the variation amplitude of phreatic water lags behind the precipitation time. The changes of phreatic water level and water volume in the river valley are mainly affected by meteorological and hydrological factors, with obvious seasonal changes. The change of water level lags behind the atmospheric precipitation for 1–2 months. The flow and water level dynamics of bedrock fissure pore phreatic water are mainly controlled by seasonality, and the annual variation range is small, generally 0.5–1.5 m, which is relatively stable.

3. Materials and Methods

3.1. Sampling and Testing

A total of 53 groundwater samples were obtained in this study, including 31 loess phreatic water (LPW) samples and 22 bedrock phreatic water (BPW) samples. The sampling time was June 2018, and the sampling position is shown in Figure 1. LPW is distributed in a large area of loess tableland area in the NCA, and the water-bearing medium is mainly “Lishi loess”. BPW is mainly distributed in the river valley area, with few surface outcrops, and the aquifer is the weathering fissure and interlayer fissure in the surface of the Huanhe Formation. The sampling depth of LPW is mainly between 50 and 100 m, and that of BPW is between 5 and 20 m. All groundwater samples were collected in 2.5 L polyethylene plastic bottles. Bottles were rinsed with the water sample to be taken for more than 3 times before actual sampling. After sampling collection, the bottles were then sealed with sealing film. Groundwater samples were kept refrigerated during transport, and samples were sent to the Gansu geological engineering laboratory for water quality testing analysis within one week of completion of groundwater sampling.

The pH value of groundwater samples was determined by the pH potentiometric method, total dissolved solids (TDS) were determined by drying at 180 °C, and total hardness (TH) was determined by disodium EDTA titration method. The content of cations including K⁺, Na⁺, Ca²⁺, and Mg²⁺ was determined by atomic absorption spectrometry; the content of anions (Cl⁻, SO₄²⁻, NO₃⁻, F⁻) was determined by high-performance liquid chromatography; and the concentration of HCO₃⁻ was determined by titration. Hydrogen (δ²H) and oxygen (δ¹⁸O) stable isotopes were measured using a liquid-water isotope analyzer (CSPC-I26), and Vienna Standard Seawater was used for the determination, using the δ²H and δ¹⁸O stable isotope content of Vienna standard mean ocean water as the standard. After the water quality test was completed, the ionic charge balance error (%CBE) calculation was carried out. The calculation results showed that the CBE% of all water samples was within the range of ±5%, which indicated the reliability of the water quality data.

$$\%\mathrm{CBE}=\frac{\sum cation-\sum anion}{\sum cation+\sum anion}\times 100\%$$

(1)

3.2. Geochemical Data Analysis

In this study, descriptive statistics of groundwater hydrochemical data in the Ning County area were firstly carried out by means of statistical analysis, the general characteristics of hydrochemistry were determined, and the hydrochemical types of groundwater were divided by the Piper diagram. Furthermore, using the Gibbs model and the ion ratio relationship, the genetic mechanism of groundwater was revealed. Finally, the possible hydrogeochemical processes on the groundwater flow path and the amount of mineral transfer were determined on the basis of the hydrogeochemical inverse simulation.

3.3. Cation Exchange

The linear relationship between Mg²⁺ + Ca²⁺ − HCO₃⁻ − SO₄²⁻ and K⁺ + Na⁺ − Cl⁻ was used to identify cation exchange processes (Equations (2) and (3)) in groundwater. If the slope of its linear relationship is −1, it indicates that cation exchange is an important reaction affecting the hydrochemical characteristics of groundwater [38]. Further, the cation exchange direction can be identified using the chlor-alkali index (CAI) [38,39]. The calculation methods of CAI-1 and CAI-2 are shown in the Equations (4) and (5).

$${\mathrm{Na}}_2\mathrm{X}+{\mathrm{Ca}}^{2+}\to 2{\mathrm{Na}}^{+}+\mathrm{CaX}$$

(2)

$$2{\mathrm{Na}}^{+}+\mathrm{CaX}\to {\mathrm{Na}}_2\mathrm{X}+{\mathrm{Ca}}^{2+}$$

(3)

$$\mathrm{CAI}-1=\frac{{\mathrm{Cl}}^{-}-{\mathrm{Na}}^{+}-{\mathrm{K}}^{+}}{{\mathrm{Cl}}^{-}}$$

(4)

$$\mathrm{CAI}-2=\frac{{\mathrm{Cl}}^{-}-{\mathrm{Na}}^{+}-{\mathrm{K}}^{+}}{{\mathrm{HCO}}_3^{-}+{\mathrm{SO}}_4^{2-}+{\mathrm{CO}}_3^{2-}+{\mathrm{NO}}_3^{-}}$$

(5)

3.4. Hydrogeochemical Modeling

Hydrogeochemical simulation is an important method to analyze various geochemical interactions in the water–rock interaction system on the basis of equilibrium thermodynamics and chemical kinetics [40,41]. To further analyze the evolution law of groundwater in the study area, two representative groundwater runoff paths were selected, and PHREEQC software (USGS, Virginia State, USA) was used to perform a reverse hydrogeochemical simulation to quantitatively explain the changes of groundwater chemical composition and the main water–rock interactions.

Inverse hydrogeochemical simulations were performed on the basis of PHREEQC software, and the uncertainty of data analysis was set to 0.05. After entering the water quality parameters of the water samples at the initial and terminal points into the software, the SI of minerals and the amount of mineral transfer can be calculated. In the process of operation, considering various factors comprehensively, we adjusted the parameters to obtain the least number of solutions, and finally, we selected the optimal solution.

3.4.1. Saturation Index (SI)

The saturation index [10,26] is used to describe the equilibrium state of minerals and aqueous solutions, referring to the logarithm of the ratio of the ionic activity product (IAP) to the solubility product constant (K_s).

$$\mathrm{SI}=\lg \frac{\mathrm{IAP}}{{\mathrm{K}}_{\mathrm{S}}}$$

(6)

SI = 0, SI > 0, and SI < 0 indicate that the minerals are in the equilibrium state of dissolution–precipitation, supersaturated state, and unsaturated state, respectively.

3.4.2. Mass Balance

The theoretical basis for reverse hydrogeochemical modeling is mass balance. On the basis of the difference of the water chemical composition between the initial point and the terminal point on the groundwater flow path, the possible hydrogeochemical reactions are deduced backward [42,43]. This relationship can be expressed by the mass balance and electron conservation equations, which are as follows:

$${\displaystyle \sum }_{n=1}^p{a}_n{b}_{n,k}=m_{T,k\left(terminal point\right)}-m_{T,k\left(initial point\right)} \left(k=1,2,3\right)\cdots ,j)$$

(7)

where a_n represents the moles of phase transfer (dissolution or precipitation) of the nth mineral phase, b_n,k represents the stoichiometric number of the kth element of the nth mineral phase, and m_T,k represents the total molar concentration of the kth element in the solution (initial or terminal point).

$${\displaystyle \sum }_{n=1}^{p}unan={\displaystyle \sum }_{n=1}^I{u}_i{m}_{i\left(terminal point\right)}-{\displaystyle \sum }_{n=1}^I{u}_i{m}_{i\left(initial point\right)}$$

(8)

where u_n represents the effective valence state in the nth mineral phase, u_i represents the effective valence state of the ith component, and m_i represents the molar concentration of the ith component in the solution (initial or terminal point).

4. Results

4.1. Statistical Characteristics of Hydrochemistry

Groundwater is a complex mixed solution, and statistical analysis of its chemical composition can help identify its source and cause [10]. The statistical results of the main chemical components of groundwater are shown in Figure 4. The average pH values of LPW and BPW were 8.1 and 7.8, respectively, indicating that the groundwater in NCA was weakly alkaline as a whole. On average, the dominant cations in LPW and BWP were Na⁺ and Ca²⁺, and the predominant anion was HCO₃⁻. In terms of mean value, cations in LPW and BPW were characterized by Na⁺ > Ca²⁺ > Mg²⁺ > K⁺, while anions were characterized by HCO₃⁻ > SO₄²⁻ > NO₃⁻ > Cl⁻ > F⁻ and HCO₃⁻ > SO₄²⁻ > Cl⁻ > NO₃⁻ > F⁻ (Figure 4), respectively.

TH and TDS in groundwater are important parameters that reflect the quality of water [39,44]. The TH concentrations of LPW and BPW ranged from 77.6 to 906.7 mg/L and 152.1 to 567.5 mg/L, with mean values of 194.6 mg/L and 286.6 mg/L, respectively. The TDS contents of LPW and BPW were in the ranges of 238.2~1142 mg/L and 349.4~1826 mg/L, with averages of 314.9 mg/L and 675.3 mg/L, respectively.

4.2. Hydrochemical Type of Groundwater

The Piper diagram [45] is often used as an effective graphical representation of the chemical composition of water samples in hydrogeological research and is the most widely used diagram in groundwater hydrochemistry research [24,44]. On the basis of the Piper diagram, water chemical types can be divided into four categories (Figure 5), namely, ① SO₄·Cl-Ca·Mg type, ② HCO₃-Na type, ③ HCO₃-Ca type, and ④ SO₄·Cl-Na type. The 53 groundwater samples in the NCA were drawn into the Piper diagram, as shown in Figure 4. It can be seen that the groundwater samples were mainly distributed in zone ③, followed by zone ②, indicating that the hydrochemical type of groundwater in the study area was mainly HCO₃-Ca type.

4.3. δ²H and δ¹⁸O Composition Characteristics

Analysis of the δ²H and δ¹⁸O composition characteristics of groundwater and comparison with the atmospheric precipitation line in the region can be used to determine the source of groundwater recharge [46,47]. The statistical results indicate that the values of δ¹⁸O and δ²H of LPW were between −10.85~−9.46 and −78.40~−70.20, and the mean values were −9.94 and −73.00, respectively. The values of δ¹⁸O and δ²H of BPW were in the range of −12.70~−8.70 and −92.89~−63.34, and the mean values were −9.88 and −72.95, respectively. In addition, the linear relationships of the hydrogen and oxygen isotopes of LPW and BPW were δ²H = 6.998δ¹⁸O − 3.802 (R² = 0.98) and δ²H = 6.283δ¹⁸O − 10.536 (R² = 0.96), respectively.

4.4. Hydrogeochemical Reverse Simulation

4.4.1. Selection of Simulation Path

The precondition for reverse hydrogeochemical simulation is that the initial point and the terminal point are in the upstream and downstream relationship, namely, they are on the same groundwater flow path [42]. It can be seen from Figure 1 that the groundwater flow direction in the north of the study area was NX32 to NX34, and in the middle valley was along the river (NZ55→NX58). Therefore, according to the location and type of groundwater sampling points, combined with the characteristics of groundwater flow field in the study area (Figure 1), LPW and BPW each selected one profile (LPW path: NX32→NX34, BPW path: NX55→NX58) for simulation.

Table 1. Main chemical components at the initial point and terminal point of different paths.

	LPW Path (NX32→NX34)		BPW Path (NX55→NX58)
	Initial Point	Terminal Point	Initial Point	Terminal Point
pH	7.9	8.0	7.0	7.7
TDS (mg/L)	323.5	318.3	581.7	545.8
K+ (mg/L)	0.8	1.1	2.9	0.5
Na+ (mg/L)	21.1	25.3	97.5	129.2
Ca2+ (mg/L)	55.9	52.3	36.2	28.7
Mg2+ (mg/L)	26.2	26.3	45.4	30.8
HCO3− (mg/L)	278	295	427	393
SO42− (mg/L)	11.0	19.7	64.8	103.7
Cl− (mg/L)	13.5	6.4	51.8	33.3
F− (mg/L)	0.33	0.34	0.6	0.63
NO3− (mg/L)	38.2	21.3	26.6	8.0
H4SiO4 (mg/L)	28.4	29.2	36.7	25.0

illustrates the main chemical components of groundwater samples at the initial and terminal points of different simulation paths.

4.4.2. Selection of Possible Mineral Phases

The selection of possible mineral phases is the key to determining the hydrogeochemical reactions that may occur on the groundwater flow path, and is also the core of the mass balance model of water–rock interaction [43]. Many studies have shown that there are two main factors for selecting possible minerals: one is the analysis of the mineral composition in the aquifer, and the main mineral composition in the aquifer should be given priority; the other is the chemical composition of groundwater and the occurrence conditions of groundwater.

The results of previous studies [48] have indicated that there are differences in the mineral composition and average content of loess soil samples and bedrock rocks in the region. Whole rock mineral X-ray diffraction analysis results showed that the main water-soluble minerals in loess were plagioclase (16.7%), microline (5.8%), illite (10%), chlorite (13.7%), calcite (8.1%) %), dolomite (4.9%), and hornblende (3.44%), while the main water-soluble minerals in bedrock rocks were plagioclase (18.6%), microline (8.8%), illite (10%), chlorite (4.0%), calcite (4.9%), dolomite (6.6%), and gypsum (10%). The possible mineral phases and main reaction equations are shown in

Table 2. Possible mineral phases and their reaction equations.

Mineral Name	Reaction Equation
Calcite	CaCO3 = Ca2+ + CO32−
Dolomite	CaMg(CO3)2 = Ca2+ + Mg2+ + 2CO32−
Gypsum	CaSO4·2H2O = Ca2+ + SO42− + 2H2O
Cation exchange	2NaX + Ca2+ = 2Na+ + CaX2
Chlorite	Mg5Al2Si3O10(OH)8 + 16H+ = 2Al3+ + 6H2O + 3H4SiO4 + 5Mg2+
Illite	K0.6Mg0.25Al2.3Si3.5O10(OH)2 + 11.2H2O = 0.6K+ + 0.25Mg2+ + 2.3Al(OH)4− + 3.5H4SiO4 + 1.2H+ 2.3Al(OH)4− + 3.5H4SiO4 + 1.2H+
Plagioclase	Na0.62Ca0.38Al1.38Si2.62O8 + 5.52H+ + 2.48H2O = 1.38Al3+ + 0.38Ca2+ + 2.62H4SiO4 + 0.62Na+ +0.38Ca2+ + 2.62H4SiO4 + 0.62Na+
Hornblende	Ca2Mg5Si8O22(OH)2 + 14H+ + 8H2O = 2Ca2+ + 8H4SiO4 + 5Mg2+
Microline	KAlSi3O8 + H2O = K+ + Al(OH)4− + 3H4SiO4

.

5. Discussion

5.1. Hydrochemical Characteristics of Groundwater

The order of Na⁺ > Ca²⁺ > Mg²⁺ > K⁺ existed in the cations in the two types of groundwater, while the anions were slightly different. The relationship of anions in LPW was HCO₃⁻ > SO₄²⁻ > NO₃⁻ > Cl⁻ > F⁻, and the order of anions in BWP HCO₃⁻ > SO₄²⁻ > Cl⁻ > NO₃⁻ > F⁻. When we compared the content of main chemical components in LPW and BPW (Figure 4), except for the fact that the pH value in LPW was higher than that in BPW, we found that other components were lower than BPW.

According to the classification [39,44] of TH and TDS (Figure 6a), the groundwater in Ning County was mainly divided into two categories: soft fresh water (18.9%) and hard fresh water (75.5%). In addition, one LPW sample and two BPW samples belonged to hard brackish water. All BWP samples were classified as hard water, while the proportion of LPW samples classified as hard water was 67.7%. The total ionic salinity (TIS) of LPW and BPW were in the range of 8.58–38.86 meq/L and 12.44–54.29 meq/L (Figure 6b), respectively. In addition, the HCO₃-Ca type was the predominant type of groundwater in the study area, indicating that the hydrochemical characteristics of groundwater may be influenced by the dissolution of carbonate rock minerals.

5.2. Hydrochemical Genetic Mechanism

5.2.1. Rock Weathering

The Gibbs diagram [49] divides the main formation mechanisms that control water chemistry into three categories: precipitation, evaporation, and rock weathering. At present, the Gibbs graphical model is widely used to identify the controlling factors of groundwater hydrochemical characteristics [12,29]. The 53 groundwater samples in the NCA were projected onto the Gibbs map, as shown in Figure 7. It can be seen that the water sample points were mainly located in the rock weathering control area, that is, rock weathering was found to be the main formation mechanism of groundwater hydrochemistry in the NCA.

5.2.2. Mineral Dissolution

The main ion ratio relationship in groundwater can be used to further determine the rock weathering source type and mineral dissolution [22,24]. The Na⁺ to Cl⁻ ratio relationship can be used to determine the effect of halite dissolution on groundwater chemistry, since the dissolution of halite produces equal amounts of Na⁺ and Cl⁻ (Equation (9)). Figure 8a illustrates that all groundwater samples in the Ning County area were distributed on the right side of the line Na⁺/Cl⁻ = 1, namely, Na⁺/Cl⁻ > 1 in groundwater, indicating that the dissolution of halite was not the main source of Na⁺ and Cl⁻ in groundwater in the area. In addition, excess Na⁺ may originate from other hydrogeochemical processes such as dissolution of silicate minerals or cation exchange.

$$\mathrm{NaCl}\to {\mathrm{Na}}^{+}+{\mathrm{Cl}}^{-}$$

(9)

The ratio relationship between Ca²⁺, Mg²⁺, and HCO₃⁻ is widely used to determine the influence of the dissolution of carbonate minerals (calcite and dolomite) in groundwater on hydrochemical characteristics [6,39]. As shown in the Figure 8b, some water samples were distributed near the Ca²⁺/HCO₃⁻ = 1/2 line or between the Ca²⁺/HCO₃⁻ = 1/2 line and Ca²⁺/HCO₃⁻ = 1 line, indicating that the chemical characteristics of groundwater in the study area are affected by the dissolution of calcite (Equations (10) and (11)). Furthermore, most of the water samples were distributed near the line Ca²⁺ + Mg²⁺/HCO₃⁻ = 1 or between the line Ca²⁺ + Mg²⁺/HCO₃⁻ = 1 and the line Ca²⁺ + Mg²⁺/HCO₃⁻ = 1/2 (Figure 8c), illustrating that the dissolution of dolomite (Equations (12) and (13)) is an important process affecting the chemical characteristics of groundwater in the study area.

$${\mathrm{CaCO}}_3+{\mathrm{H}}^{+}\to {\mathrm{Ca}}^{2+}+{\mathrm{HCO}}_3^{-}$$

(10)

$${\mathrm{CaCO}}_3+{\mathrm{CO}}_2+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Ca}}^{2+}+2{\mathrm{HCO}}_3^{-}$$

(11)

$$\mathrm{CaMg}{({\mathrm{CO}}_3)}_2+2{\mathrm{H}}^{+}\to {\mathrm{Ca}}^{2+}+{\mathrm{Mg}}^{2+}+2{\mathrm{HCO}}_3^{-}$$

(12)

$$1\mathrm{CaMg}{({\mathrm{CO}}_3)}_2+2{\mathrm{CO}}_2+2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Ca}}^{2+}+{\mathrm{Mg}}^{2+}+4{\mathrm{HCO}}_3^{-}$$

(13)

The ratio of Ca²⁺/Mg²⁺ is often used to reveal the source of Ca²⁺ and Mg²⁺ in groundwater [50]. Ca²⁺ /Mg²⁺ = 1 indicates that it comes from the dissolution of dolomite, while if 1 < Ca²⁺/Mg²⁺ < 2, it means that there is more dissolution of calcite. Ca²⁺/Mg²⁺ > 2 indicates that more Ca²⁺ is provided by the dissolution of silicate minerals or gypsum [50]. As shown in Figure 8d, the LPW samples were mainly distributed between Ca²⁺/Mg²⁺ = 1 and Ca²⁺/Mg²⁺ = 2, indicating that the Ca²⁺ in LPW mainly originated from the dissolution of calcite. In addition, the BPW samples were mainly concentrated below the line of Ca²⁺/Mg²⁺ = 1, indicating that the dissolution of dolomite or Mg-rich carbonate minerals had an important influence on the chemical composition of the BPW.

The dissolution of gypsum may be one of the important sources of SO₄²⁻ in groundwater in a region. If gypsum is the only source of Ca²⁺ and SO₄²⁻ in groundwater, the ratio of Ca²⁺ to SO₄²⁻ should be 1 (Equation (13)). As shown in Figure 8e, most of the BPW samples were distributed close to the Ca²⁺/SO₄²⁻ = 1 line, while the LPW samples deviated from the Ca²⁺/SO₄²⁻ = 1 line, indicating that the dissolution of gypsum had a more significant contribution to the chemical characteristics of the BPW. In addition, the oxidation of pyrite or other sulfur-bearing minerals may also be a source of SO₄²⁻ in groundwater [8]. However, the Cretaceous strata in the study area were generally a set of complex sediments under arid and semi-arid climate conditions, in which the upper part of sandstone and mudstone of the Huanhe formation was mostly sandwiched with a salt marsh facies gypsum layer, with high salt content and being rich in gypsum, sodium sulfate, halite, and other evaporite minerals [51]. Thus, the main source of SO₄²⁻ in the groundwater of the NCA is the dissolution of gypsum and other sulfate minerals. The relationship between Mg²⁺ + Ca²⁺ and HCO₃⁻ + SO₄²⁻ is also often used to explain the contribution of the weathering of carbonate and silicate and the dissolution of gypsum to groundwater hydrochemical characteristics [50]. As shown in Figure 8f, the groundwater sample points were mainly distributed in the upper part of the line (Mg²⁺ + Ca²⁺)/(HCO₃⁻ + SO₄²⁻) = 1, indicating that the weathering of carbonate and the dissolution of gypsum were important processes that affect the chemical composition of groundwater in the NCA.

$${\mathrm{CaSO}}_4·2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Ca}}^{2+}+{\mathrm{SO}}_4^{2-}+2{\mathrm{H}}_2\mathrm{O}$$

(14)

5.2.3. Cation Exchange

In general, cation exchange may also be an important hydrogeochemical reaction in a regional groundwater system that affects the chemical composition of groundwater [16,38]. As shown in Figure 9a, groundwater sample points in Ning County were all distributed along the x/y = −1 line, and only one LPW sample deviated from the x/y = −1 line, suggesting that cation exchange is an important hydrochemical process in the groundwater system in the region. If the CAI value is greater than 0, it indicates that reverse cation exchange has occurred, resulting in the exchange of Na⁺ in groundwater with Ca²⁺ in the aquifer (Equation (2)). If the Ca²⁺ in groundwater is exchanged with Na⁺ in the aquifer, direct cation exchange occurs (Equation (3)), and the value of CAI is less than 0. Figure 9b illustrates that the calculated CAI values of the water samples were all less than 0, indicating that direct cation exchange occurred in the groundwater system in the NCA.

5.3. Analysis of Groundwater Recharge Sources

The δ²H and δ¹⁸O data of groundwater samples in the study area were compared and analyzed with the Global Meteoric Water Line (GMWL) and the Local Meteoric Water Line (LMWL), as shown in Figure 10. Generally, the water sample points were distributed close to the atmospheric precipitation line, indicating that the main source of the water body was atmospheric precipitation. The GMWL (δD = 8δ¹⁸O + 10) in Figure 10 was obtained by Craig [52] in 1961. In addition, the LMWL in this study adopted the atmospheric precipitation line (δD = 8.10δ¹⁸O + 6.97) in the Pingliang area [53], which was close to the study area. It can be seen from Figure 10 that the groundwater sample points in the study area were mainly distributed near the lower part of the GMWL and along the LMWL as a whole, indicating that the main recharge source of groundwater in the region was atmospheric precipitation, which is affected by evaporation to a certain extent.

5.4. Hydrogeochemical Simulation Results

5.4.1. Saturation Index (SI)

The saturation indices and charge balance coefficients of minerals in groundwater at starting points on different simulation paths are shown in

Table 3. Calculated mineral saturation index and charge balance coefficient.

	LPW Path (NX32→NX34)		BPW Path (NX55→NX58)
	Initial Point	Terminal Point	Initial Point	Terminal Point
Calcite	0.68	0.80	−0.24	0.26
Dolomite	1.39	1.65	−0.04	0.90
Gypsum	−2.66	−2.44	−2.16	−2.04
Chlorite	−0.22	0.81	−6.00	−2.15
Illite	−1.23	−1.28	0.50	−1.31
Plagioclase	9.16	9.21	9.72	9.35
Hornblende	7.22	9.49	−3.11	3.52
Microline	−2.21	−2.06	−1.36	−2.61
Charge balance coefficient	−2.5 × 10−4	−5.88 × 10−4	−1.48 × 10−3	−1.35 × 10−3

. Charge balance is a fundamental condition for performing reverse hydrogeochemical simulations.

Table 4. Mineral transfer mass on the LPW path and BPW path.

	Mineral Transfer Mass (mol/L)
	LPW Path	BPW Path
Calcite	8.197 × 10−4	3.667 × 10−3
Dolomite	−2.658 × 10−4	−2.826 × 10−3
Gypsum	/	3.297 × 10−4
Chlorite	−2.170 × 10−4	2.850 × 10−4
Illite	1.231 × 10−4	4.899 × 10−4
Plagioclase	−1.208 × 10−4	−9.721 × 10−4
Microline	−7.311 × 10−4	−3.554 × 10−4
Hornblende	2.127 × 10−4	1.233 × 10−4
NaX	1.150 × 10−3	2.537 × 10−3
CaX2	−5.750 × 10−4	−1.269 × 10−3

illustrates that the charge balance coefficients of the water samples on the groundwater flow path were all close to 0, which indicates that the charges were essentially balanced. Table 3 indicates that there was a difference in the SI between the initial point and terminal point. The SI was less than 0, indicating that these minerals in the groundwater did not reach a saturated state, and that these minerals have a tendency to dissolve. The SI was greater than 0, indicating that it is in a supersaturated state and there is a tendency for precipitation to occur.

5.4.2. Mineral Transfer Mass

By using the PHREEQC software, the amount of mineral transfer mass on the simulated path was calculated. Mineral transfer masses greater than 0 or less than 0 represent mineral dissolution and precipitation, respectively. It can be seen from Table 4 that the dissolution of calcite, illite, and hornblende, as well as the precipitation of dolomite, plagioclase, and microcline occurred on both the LPW and BPW simulation paths. In addition, cation exchange occurred on both paths, and the Na⁺ in the aquifer was replaced with Ca²⁺ in the groundwater. The dissolution of gypsum and chlorite also occurred on the BWP path, while the precipitation of chlorite occurred on the LPW path. On different groundwater flow paths, the hydrogeochemical reactions of mineral transfer mass were different, which was the result of the combined effects of the lithology, geology, and hydrogeological conditions of the aquifer and its surrounding physical, chemical, and biological environments.

5.5. Suggestions for Regional Groundwater Research

This study analyzed the chemical characteristics and controlling factors of groundwater in the NCA, and the research results can provide reference for the development and utilization of regional groundwater resources. However, the data from this study are only from a sampling period, and future studies should carry out sampling for different periods, depths, and types of groundwater in order to obtain more water quality data. In addition, other chemical components in groundwater such as trace elements, heavy metals, and organic matter should also be analyzed in order to enrich the results of regional groundwater-related studies and provide more comprehensive support for the development and utilization of groundwater resources.

6. Conclusions

The study of groundwater hydrochemistry is helpful for the rational development, utilization, and management of groundwater resources in a region, and can also provide a reference for similar regions in the world. To understand the hydrochemical characteristics and formation mechanism of groundwater in the NCA of the Loess Plateau, the LPW and BPW were sampled and analyzed, respectively, and the following main conclusions were obtained:

(1)

The groundwater of NCA was weakly alkaline as a whole, mainly classified as hard fresh water, and HCO₃·Ca was the dominant hydrochemical type. Except for pH, the concentrations of major chemical components in BPW were higher than those in LPW.

(2)

δ²H and δ¹⁸O analysis illustrated that the main recharge source of groundwater in NCA was atmospheric precipitation and was affected by evaporation. The linear relationships of the δ²H and δ¹⁸O of LPW and BPW were δ²H = 6.998δ¹⁸O−3.802 (R² = 0.98) and δ²H = 6.283δ¹⁸O − 10.536 (R² = 0.96), respectively.

(3)

The main controlling factors of LPW and BPW in the study area were rock weathering and cation exchange. Ca²⁺, Mg²⁺, and HCO₃⁻ in groundwater mainly originated from the dissolution of calcite and dolomite. In addition, the dissolution of gypsum also had a significant effect on the water chemical composition of BPW.

(4)

Hydrogeochemical inverse simulations indicated that cation exchange; the dissolution of calcite, illite, and hornblende; and the precipitation of dolomite, plagioclase, and microcline occurred on both the LPW and BPW pathways. In addition, the dissolution of gypsum and chlorite also occurred on the BWP path, and the precipitation of chlorite occurred on the LPW path.