A New Insight into Sulfate Contamination in Over-Exploited Groundwater Areas: Integrating Multivariate and Geostatistical Techniques

Abstract

The issue of sulfate (SO₄²⁻) pollution in groundwater has already attracted widespread attention from scientists. However, at the large-scale regional level, especially in areas with groundwater overexploitation, the pollution mechanisms and sources of sulfate remain unclear. This study innovatively investigates the spatial distribution characteristics and sources of SO₄²⁻ in the groundwater of the Hutuo River alluvial fan area, an understudied region facing significant environmental challenges due to overexploitation. Utilizing a combination of hydrochemical analysis, multivariate statistical methods, and geostatistical techniques, we reveal that the mean concentration of SO₄²⁻ is significantly higher (127 mg/L) in overexploited areas, with an exceedance rate of 5.1%. Our findings uncover substantial spatial heterogeneity in SO₄²⁻ concentrations, with particularly high levels in the river valley plain (RVP) (175 mg/L) and the upper area of the alluvial fan (UAF) (169 mg/L), which we attribute to distinct human activities. A novel contribution of our study is the identification of groundwater depth as a critical factor influencing SO₄²⁻ distribution (p < 0.001). We also demonstrate that the higher proportion of sulfate-type waters in overexploited areas is primarily due to the accelerated oxidation of sulfide minerals caused by overexploitation. Principal component analysis (PCA) and correlation analysis further identify the main sources of SO₄²⁻ as industrial wastewater, domestic sewage, the dissolution of evaporites, and the oxidation of sulfide minerals. By integrating geostatistical techniques, we present the spatial distribution of sulfate pollution sources at a fine scale, providing a comprehensive and spatially explicit understanding of the pollution dynamics. These results offer a novel scientific basis for developing targeted strategies to control sulfate pollution and protect the sustainable use of regional groundwater resources. Our study thus fills a critical knowledge gap and provides actionable insights for groundwater management in similar regions facing overexploitation challenges.

1. Introduction

Sulfate (SO₄²⁻) is one of the common soluble anions in groundwater, characterized by its chemical stability, strong solubility, and ease of migration [1,2]. Sulfate exhibits a certain degree of oxidizing properties, which can influence redox reactions in groundwater and alter the migration and transformation behaviors of other pollutants. Due to its diverse sources and complex migration and transformation processes [3], the distribution characteristics and source identification of sulfate in groundwater have long been a research focus and challenging issue in the field of hydrogeochemistry.

Excessive concentrations of sulfate in groundwater can pose multifaceted threats to ecosystems and human health [4]. For example, they can impact the respiration and reproduction of fish and may even lead to the death of aquatic organisms [5]. Regarding human health, the consumption of groundwater with sulfate levels exceeding the standard may cause gastrointestinal dysfunction, such as diarrhea and bloating, and long-term consumption may also damage organs such as the kidneys [6]. In the industrial sector, groundwater with elevated sulfate concentrations can corrode pipelines and equipment, increasing the difficulty and cost of industrial water treatment [7]. Therefore, studying the distribution characteristics and sources of sulfate in groundwater is of great significance for protecting the ecological environment, safeguarding human health, and the rational utilization of groundwater resources.

The accurate identification of sulfate pollution sources is a crucial step in formulating groundwater pollution prevention strategies and sustainable water resource management plans. Previous studies found that the sources of sulfate in water environments are extensive and diverse, including natural weathering, agricultural activities, industrial emissions, and domestic sewage [8], which greatly increases the complexity of pollution source identification. Current research endeavors to identify sulfate sources within aquatic environments have predominantly concentrated on elucidating the pollution sources of sulfate in surface water systems [8,9], while studies on the sources of sulfate pollution in groundwater are mainly concentrated in industrial zones and karst regions [3,10]. However, there is a relative lack of research on the identification of sulfate sources in large-scale overexploited groundwater areas. Especially in groundwater overexploitation areas where the flow and recharge conditions of the groundwater are complex, making the pollution characteristics and source identification of sulfate more challenging.

The Hutuo River alluvial fan is located in the North China Plain. In recent years, with the rapid increase in water demand for industrial and agricultural purposes, the area has experienced intense groundwater extraction and is considered a typical region of groundwater overexploitation [11]. As the intensity of groundwater extraction has increased, the groundwater level has declined annually, severely affecting the groundwater chemical and hydrodynamic fields. This has intensified the water–rock interactions [12]. Additionally, the region is characterized by well-developed coal-bearing strata and human activities have a significant impact on groundwater. As a result, the concentration of sulfate in groundwater has been increasing [13]. However, the characteristics of groundwater sulfate distribution and its pollution sources in this area have not yet received attention, posing a potential threat to the drinking water safety of local residents.

In this study, the Hutuo River alluvial fan is selected as the research site. By integrating hydrochemical analysis, multivariate statistical methods, and advanced geostatistical techniques, we provide a comprehensive and spatially explicit investigation of the sulfate pollution dynamics. Our study not only elucidates the spatial distribution patterns of sulfate and its primary sources but also reveals the underlying pollution mechanisms in this critical region. These findings fill a significant knowledge gap regarding sulfate pollution in overexploited groundwater systems and offer a novel scientific foundation for developing targeted strategies to mitigate and control sulfate contamination in similar regions worldwide.

2. Materials and Methods

2.1. Study Area Overview

The research area is situated in the central and upper reaches of the Hutuo River alluvial fan (Figure 1), stretching from the Gangnan Reservoir (located within the territory of Pingshan County) in the west to the western boundary of Gaocheng County in the east, covering a drainage area of 2442 km². The Hutuo River basin is characterized by a scarcity of available surface water resources. Consequently, human activities such as domestic use, industrial production, and agricultural practices predominantly depend on groundwater [14]. The topography of the study area gradually slopes from west to east. The region experiences a temperate continental monsoon climate, with average annual precipitation and temperature varying between 400 and 750 mm and 13.3–15.0 °C, respectively. The majority of precipitation occurs during the months of June to September, accounting for more than 70% of the annual rainfall. The average annual temperature is approximately 13 °C, with the lowest temperature occurring in January, averaging 0.6 °C, and the highest temperature occurring in July, averaging 25 °C. The study area experiences significant seasonal variations in evaporation. The highest evaporation occurs in summer, ranging from 200 to 250 mm, which accounts for approximately 45% of the annual evaporation. In contrast, the lowest evaporation occurs in winter, ranging from 50 to 80 mm, representing about 10% of the annual evaporation.

2.2. Hydrogeological Conditions

The aquifer within the study area is part of the thick multi-layered Quaternary aquifer system in the North China Plain. This aquifer is relatively water-bearing, with groundwater flowing from the northwest to the southeast. The depth of the water table varies from 12 to 240 m, while the thickness of the aquifer ranges between 20 and 140 m. Both the water table and the aquifer exhibit favorable characteristics in terms of conductivity and water yield. Today, the acceleration of urbanization has led to persistent over-exploitation of groundwater in this area, resulting in a significant decline in groundwater depth. Specifically, the depth of the groundwater table has increased from 36 m in 2002 to 43 m in 2016 [12]. The dominant types of groundwater are fracture water in bedrock and pore water in loose rock layers. The primary sources of recharge are infiltration from atmospheric precipitation, infiltration from surface water, and return flow from agricultural irrigation, while the main discharge mechanism is artificial extraction. The lithology of the Quaternary sediments mainly comprises silty clay, sandy loam, gravelly sandy clay, and gravel and sand layers of various grain sizes (Supplementary Figure S1). The permeability coefficient of the vadose zone in the study area generally ranges from 37 to 145 m/d.

Based on the geological conditions and aquifer characteristics, the study area is divided into three hydrogeological units (Figure 1): (1) the fracture–pore water unit in the river valley plain (RVP) between the Gangnan Reservoir and Huangbizhuang Reservoir; (2) the pore water unit at the upper of the Hutuo River alluvial fan (UAF); and (3) the pore water unit in the central part of the alluvial fan (CAF). In the RVP, the aquifer thickness ranges from 10 to 20 m, with groundwater depths between 2 and 21 m. In the UAF, the aquifer thickness ranges from 20 to 35 m, with groundwater depths between 12 and 20 m. In the CAF, the aquifer thickness ranges from 25 to 60 m, with groundwater depths between 40 and 50 m.

2.3. Sample Collection and Analysis

2.3.1. Sample Collection

To eliminate the impact of spatial heterogeneity of groundwater samples, this study established sampling points at an interval of one per 20 km², with a total of 118 sampling set ups (Figure 1) and were collected in January 2023. Furthermore, three sewage samples were collected. The groundwater samples were primarily obtained from domestic wells and agricultural irrigation wells, with well depths ranging from 4.0 to 50 m and an average depth of 9.8 m. Prior to sample collection, the wells were purged for 5–10 min until the groundwater pH and electrical conductivity (EC) values stabilized. Subsequently, groundwater samples were obtained through pumping operations from the wells. The sampling bottles were made of polyethylene, with one 1.5 L bottle and one 0.5 L bottle used for analyzing cations and anions, respectively. For cation analysis, the water samples were acidified with 1:1 HNO₃ to adjust the pH to less than 2. Before sampling, the bottles were rinsed three times with distilled water and the original water samples. After collecting, the samples were transported to the laboratory in a portable cooler and tested within 48 h.

2.3.2. Sample Analysis and Testing

The monitoring indicators included the following: (1) on-site measurement indicators: pH, dissolved oxygen (DO); (2) laboratory analysis indicators: total dissolved solids (TDS), K⁺, Na⁺, Ca²⁺, Mg²⁺, NO₃⁻, SO₄²⁻, Cl⁻, HCO₃⁻, and total hardness (TH, expressed as calcium carbonate). The analytical methods for the hydrochemical indicators are detailed in Supplementary Table S1.

2.4. Data Analysis

2.4.1. Data Preprocessing

Due to the requirement of normality in the distribution of each variable for principal component analysis (PCA), an initial assessment of the kurtosis and skewness of all data were conducted prior to PCA. Values of kurtosis and skewness beyond the range of −2 to +2 indicate a significant deviation from normality [15]. The statistical analysis revealed that the distributions of K⁺, Na⁺, and NO₃⁻ were skewed and not normally distributed, with standardized kurtosis and skewness values ranging from 10.1 to 30.7 and 2.84 to 7.29, respectively. Following a logarithmic transformation of these parameters, the kurtosis and skewness values were reduced to the range of 0.592 to 1.31 and −0.87 to 0.65, respectively. Subsequently, all water quality parameters underwent further standardization using the z-scale to mitigate the risk of misclassification arising from the substantial variability in data dimensions.

2.4.2. Statistical Analysis and Spatial Interpolation Analysis

In this research, the non-parametric Kruskal–Wallis and Mann–Whitney U tests were utilized to assess the significant differences in SO₄²⁻ concentrations among various hydrogeological units. Spearman’s analysis was used to investigate the correlation between hydrochemical components. PCA was conducted to pinpoint the primary sources of sulfate pollution. Pollution sources were spatially visualized by projecting the source identification results from PCA onto the study area map through geostatistical interpolation, primarily using inverse distance weighting. All statistical analyses were performed using the R computing platform (version 4.3.2, Auckland, New Zealand), while geospatial analysis was carried out using ArcGIS v10.5 software (ESRI Inc., Redlands, CA, USA).

3. Results and Discussion

3.1. Characteristics of Groundwater Quality

To analyze the characteristics of groundwater quality in the study area, statistical analysis was conducted on the numerical ranges, means, and exceedance rates of various water quality indicators (

Table 1. Statistics of different groundwater parameters in the Hutuo River alluvial fan basin.

Parameters	Min	Max	Mean	SD	RES (%)	Standard
DO	1.04	13.8	4.91	2.66	0	—
pH	6.53	8.47	7.47	0.356	0	6.5–8.5
K+	0.12	25.9	2.29	2.57	0	—
Na+	8.78	172	34.7	26.4	0	200
Ca2+	27.2	367	131	63.8	0	—
Mg2+	10.3	102	32.0	13.8	0	—
Cl−	11.9	379	71.4	59.7	1.7	250
SO42−	12.9	395	127	77.2	5.1	250
HCO3−	109	565	296	72.6	0	—
NO3−	0.24	440	61.8	70.2	22.0	88.6
TH	115	1339	458	198	50.0	450
TDS	174	2091	628	292	8.5	1000

Note: Standard is the grade III standard for groundwater quality in China (GB/T 14848-2017) [16]. The pH value is dimensionless, while the units for all other indices are expressed in mg/L. The sample size for all indicators is 118. RES = rate of exceeded standard; SD = standard deviation. “—” indicates the absence of a corresponding standard.

). The results show that the pH of groundwater ranges from 6.53 to 8.47, indicating neutral to slightly alkaline conditions, with no exceedance of standards. The concentration of DO ranges from 1.04 to 13.8 mg/L, with a mean value of 4.91 mg/L, indicating an overall oxidizing environment.

The average concentrations of cations are ordered as follows: Ca²⁺ (131 mg/L) > Na⁺ (34.7 mg/L) > Mg²⁺ (32.0 mg/L) > K⁺ (2.29 mg/L). For anions, the concentration order is HCO₃⁻ (296 mg/L) > SO₄²⁻ (127 mg/L) > Cl⁻ (71.4 mg/L) > NO₃⁻ (61.8 mg/L). The highest exceedance rate in the study area is for TH, with a rate of 50.0%, which is mainly related to the geological background of the region. Previous studies have shown that carbonate rocks are widely distributed in the study area, and the dissolution of these rocks leads to high concentrations of Ca²⁺ and Mg²⁺ in groundwater, resulting in significant exceedance of TH levels [17]. Additionally, exceedance of groundwater quality standards for NO₃⁻, TDS, SO₄²⁻, and Cl⁻ was observed, with exceedance rates of 22.0%, 8.47%, 5.08%, and 1.69%, respectively. It is worth noting that nitrate, with its high exceedance rate and mean value, indicates the significant impact of human activities on groundwater quality [18].

3.2. Spatial Distribution Characteristics of SO₄²⁻ in Groundwater

3.2.1. Variations in SO₄²⁻ Across Distinct Hydrogeological Units

The spatial heterogeneity in sulfate concentrations observed in the Hutuo River alluvial fan groundwater (Figure 2) provides critical insights into the interplay between human activities and natural hydrogeological conditions. The significantly higher sulfate concentrations in the RVP and UAF (175 mg/L and 169 mg/L, respectively) compared to the CAF (77.3 mg/L) (p < 0.001) highlight the profound impact of human activity on groundwater quality. In the RVP and UAF, especially around the Huangbizhuang Reservoir, there are large-scale power plants, the wastewater discharged from power plants may be an important cause of the elevated sulfate concentrations in the surrounding groundwater [19]. Additionally, the Ye River basin, a tributary of the Hutuo River (Flow into Huangbizhuang Reservoir), has extensive coal-bearing strata and hosts multiple coal mines and coal-washing plants. The discharge of industrial wastewater from these facilities is the other primary cause of elevated sulfate concentrations in the region. This suggests that even with existing regulations, there remains a critical gap in controlling industrial effluents, particularly in areas where groundwater serves as both a receptor and a resource. In addition, in the RVP, the lack of sewage treatment systems in villages leads to the discharge of sewage directly into the Hutuo River, and ultimately infiltrates into groundwater, which is another significant factor contributing to high sulfate concentrations in groundwater. In conclusion, the spatial differences in sulfate concentrations across the Hutuo River alluvial fan groundwater system represent more than a local pollution issue. They illustrate the complex interplay between development pressures, industrial activities, and natural hydrogeological conditions in an overexploited groundwater region.

3.2.2. Variation of SO₄²⁻ with Groundwater Depth

The depth of groundwater burial significantly influences the pathway distance that pollutants must travel to reach the aquifer. In the present study, a significant negative correlation (p < 0.001) was observed between groundwater depth and sulfate concentration (Figure 3). Specifically, sulfate concentration decreased with increasing groundwater depth. These findings indicate that groundwater depth is a critical factor influencing the distribution of sulfate. Further analysis suggests that this relationship can be attributed to the fact that shallow groundwater levels facilitate the infiltration of pollutants, such as sulfate, from the surface, while deeper groundwater levels are less susceptible to such direct contamination due to longer flow paths and the filtering effect of the unsaturated zone [20]. Additionally, the redox conditions in deeper aquifers may also play a role in reducing sulfate concentrations, as the reducing environment may promote the precipitation of sulfate or the reduction in sulfate to sulfide [21]. The findings align with previous studies that have highlighted the importance of groundwater depth in determining the vulnerability of aquifers to contamination.

3.3. Hydrochemical Types of Groundwater

Using the Piper trilinear diagram, the hydrochemical types of groundwater in the three hydrogeological units of the study area were identified (Figure 4). At the same time, we integrated the Kurlov classification method (which classifies and names the types based on the content of anions and cations in the water, considering ions with a percentage greater than 25% of the milliequivalent as significant) to accurately calculate and distinguish the types and proportions of groundwater chemistry in different hydrogeological units. In the RVP and UAF, the dominant hydrochemical type of groundwater is HCO₃·SO₄–Ca, accounting for 65.4% and 39.8% of the sampling points, respectively. The secondary type is HCO₃·SO₄–Ca·Mg, with proportions of 11.5% and 27.8%, respectively. In the CAF, the primary hydrochemical type is HCO₃–Ca·Mg, accounting for 48.2% of the sampling points, followed by HCO₃·SO₄–Ca·Mg and HCO₃·Cl–Ca·Mg types, each with a proportion of 14.3%.

Overall, from the RVP to the CAP, the hydrochemical types of groundwater gradually evolve from HCO₃·SO₄–Ca to HCO₃–Ca·Mg. This transition is mainly related to differences in lithology and human activities [22]. According to the hydrogeological conditions of the study area, coal-bearing strata are widely distributed in the RVP and UAF regions, and there are coal mines and power plants around the Huangbizhuang Reservoir. Both coal-bearing strata and industrial wastewater contribute to elevated SO₄²⁻ concentrations in groundwater, resulting in a higher proportion of SO₄-type water in these regions. In the CAF, the aquifer is primarily composed of quaternary loose sedimentary rocks. This area is mainly a residential zone without large industrial sources, leading to a shift in the hydrochemical type to predominantly HCO₃–Ca·Mg. In the region characterized by overexploitation of the alluvial fan, the prevalence of SO₄-type water significantly surpasses that observed in urban settings (dominated by HCO₃–Ca·Mg-type water) [23], coastal zones (dominated by HCO₃·Cl–Na-type waters) [24], and arid regions (dominated by HCO₃–Ca·Mg-type waters) [25]. These findings imply that the intensification of groundwater extraction may be exacerbating the dissolution of sulfur-bearing minerals within the geological formations.

3.4. Identification of SO₄²⁻ Sources in Groundwater

3.4.1. Identification of SO₄²⁻ Sources Using Hydrochemical Methods

The use of hydrochemical ion ratio diagrams to identify the sources of pollutants in groundwater is a traditional method [26]. In this study, the ratios of [Ca²⁺]/[SO₄²⁻] and [NO₃⁻]/[Ca²⁺] to [SO₄²⁻]/[Ca²⁺] were used to identify the sources of sulfate in the Hutuo River alluvial fan groundwater.

Based on the [Ca²⁺]/[SO₄²⁻] ratio diagram (Figure 5a), a good correlation between Ca²⁺ and SO₄²⁻ was observed in the groundwater of the three hydrogeological units in the study area (p < 0.001), indicating a common source for these ions, such as the dissolution of gypsum in coal-bearing strata. This result further corroborates that the spatial variability of sulfate concentrations is also influenced by natural factors, such as the distribution of coal-bearing strata.

To further clarify the sources of sulfate in groundwater, the relationship between [NO₃⁻]/[Ca²⁺] and [SO₄²⁻]/[Ca²⁺] was used to characterize the impact of human activities on groundwater. Previous studies have shown that industrial activities are characterized by higher [SO₄²⁻]/[Ca²⁺] values, while agricultural activities have higher [NO₃⁻]/[Ca²⁺] values [27]. As shown in Figure 5b, 93.2% of the groundwater samples had higher [SO₄²⁻]/[Ca²⁺] values than [NO₃⁻]/[Ca²⁺] in the study area, indicating that industrial activities are the primary human influence on groundwater in the study area. Specifically, the [SO₄²⁻]/[Ca²⁺] ratios ranged from 0.167 to 1.15, 0.149 to 1.48, and 0.186 to 2.89 in the RVP, UAF, and CAF, respectively. The [NO₃⁻]/[Ca²⁺] ratios ranged from 0.043 to 0.389, 0.022 to 0.731, and 0.009 to 0.985 in the RVP, UAF, and CAF, respectively. The markedly higher [SO₄²⁻]/[Ca²⁺] ratios across all sub-regions provide compelling evidence of the significant role of industrial activities in groundwater sulfate pollution. This conclusion is further supported by the analysis of three sewage samples, which also demonstrated elevated [SO₄²⁻]/[Ca²⁺] ratios, aligning with the observed groundwater patterns. These findings collectively underscore the need for targeted industrial pollution control measures to mitigate sulfate contamination in groundwater resources.

3.4.2. Identification of SO₄²⁻ Sources Using Correlation Analysis

The relationships between groundwater quality indicators can reflect the sources of these parameters [28]. In this study, Pearson’s correlation analysis was conducted to explore the relationships between sulfate and other indicators, which offers crucial insights into the origins of SO₄²⁻ in groundwater. As depicted in Figure 6, indicators with high correlation coefficients to SO₄²⁻ include TDS, TH, Ca²⁺, Na⁺, Mg²⁺, Cl⁻, and NO₃⁻, with correlation coefficients of 0.85, 0.80, 0.80, 0.76, 0.56, 0.61, and 0.61, respectively. The strong correlations between SO₄²⁻ and Cl⁻, Na⁺, and NO₃⁻ suggest common sources for these ions. Prior research has shown that domestic sewage and industrial activities are common sources of Cl⁻ and NO₃⁻ [26], suggesting that SO₄²⁻ in the study area likely originates from industrial activities and domestic sewage. The significant correlation between SO₄²⁻ and Ca²⁺ confirms that gypsum dissolution is one of the sources of sulfate, consistent with the earlier analysis. Additionally, the positive correlation between SO₄²⁻ and TH, Mg²⁺ may be related to enhanced water–rock interactions caused by intensive groundwater extraction [29].

3.4.3. Identification of Sulfate Sources Using PCA

To delve deeper into the sources of sulfate in the groundwater in the area, PCA was conducted on 12 water quality parameters, including DO, pH, K⁺, Na⁺, Ca²⁺, Mg²⁺, Cl⁻, SO₄²⁻, HCO₃⁻, NO₃⁻, TH, and TDS. Prior to the analysis, the Kaiser–Meyer–Olkin (KMO) and Bartlett’s sphericity tests were performed. The results showed that the KMO value was 0.740, and Bartlett’s sphericity test value was 2827 (p < 0.001), indicating that the water quality data met the requirements for PCA. Based on eigenvalues exceeding 1, three principal components were extracted, which collectively explained 78.5% of the total variance.

As indicated in

Table 2. Results of principal component analysis for groundwater quality data in Hutuo River alluvial fan.

Parameters	PC1	PC2	PC3
TH	0.825	0.529	−0.024
NO3−	0.820	0.224	0.178
Ca2+	0.814	0.470	0.064
pH	−0.814	0.157	0.070
HCO3−	0.751	0.225	−0.305
TDS	0.739	0.652	0.009
Mg2+	0.589	0.522	−0.264
K+	−0.112	0.781	0.044
Na+	0.359	0.769	−0.065
Cl−	0.523	0.688	−0.138
SO42−	0.592	0.595	0.222
DO	−0.009	−0.006	0.940
Eigenvalue	6.98	1.30	1.14
Contribution Rate (%)	58.2	10.8	9.46
Cumulative Contribution Rate (%)	58.2	69.0	78.5

, PC1 accounted for 58.2% of the total variance, with high factor loadings for TH, NO₃⁻, Ca²⁺, pH, HCO₃⁻, TDS, SO₄²⁻, and Mg²⁺. Common sources of groundwater sulfate include the oxidation of sulfides, the dissolution of evaporites, industrial wastewater, domestic sewage, and precipitation [8]. The positive correlation between NO₃⁻ and SO₄²⁻ with PC1 suggests a common source for these ions. Previous studies have shown that the NO₃⁻ of groundwater in the area primarily originates from domestic sewage and fertilizers [14]. Given that nitrogen-based fertilizers are predominantly utilized in local agriculture, coupled with the minimal application of sulfur-containing fertilizers, it is reasonable to posit that domestic sewage constitutes a significant source of sulfate in the groundwater.

As discussed in Section 3.3 and Section 3.4.1, there are large-scale power plants around the Huangbizhuang Reservoir in the UAF of the study area, and the groundwater monitoring sites with exceedances are primarily located around the reservoir. A previous study found that the Ye River exhibits a relatively high average SO₄²⁻ concentration of 216 mg/L [30]. Moreover, the mean SO₄²⁻ concentration in the three sewage samples we collected reached 274.5 mg/L. These findings collectively suggest that industrial wastewater discharge from power plants and other industrial facilities, along with domestic sewage, are the predominant contributors to the elevated sulfate concentrations in the region. In summary, PC1 represents that the main sources of sulfate in groundwater are domestic sewage and industrial wastewater.

PC2 explained 10.8% of the total variance, with high factor loadings for K⁺, Na⁺, Cl⁻, TDS, and SO₄²⁻. Moderate correlations were also observed between TH, Mg²⁺, Ca²⁺, and PC2. Since PC1, PC2, and PC3 are independent of each other, the strong correlation between Na⁺ and Cl⁻ is indicative of rock salt dissolution [31]. The good correlation between SO₄²⁻ and Ca²⁺ with PC2 indicated that the evaporites (gypsum) are an important source of SO₄²⁻. Previous studies have shown that the increase in TH in groundwater in this region is caused by intensive groundwater extraction [19]. Indeed, in recent years, the Hutuo River alluvial fan area has experienced significant groundwater extraction, leading to a substantial decline in groundwater levels and enhanced water–rock interactions, which in turn accelerate the dissolution of rock salt and evaporites in the strata [32]. It is important to note that while rock salt dissolution contributes to the increase in Na⁺ and Cl⁻, it does not directly provide sulfate ions. Therefore, the correlation between SO₄²⁻ and Ca²⁺ with PC2, coupled with the presence of evaporites, strongly indicates that the SO₄²⁻ in groundwater originates from the dissolution of evaporites such as gypsum. This conclusion is further supported by the hydrogeochemical processes prevalent in areas with significant groundwater extraction, where enhanced water–rock interactions play a crucial role in shaping groundwater chemistry. Thus, PC2 demonstrates that SO₄²⁻ in groundwater originates from the dissolution of evaporites (gypsum).

PC3 explained 9.46% of the total variance, with high factor loadings for DO, and slight correlations were also observed between SO₄²⁻ and PC3. The mean concentration of DO in groundwater in the study area is 4.91 mg/L, indicating an oxidizing environment. Previous studies have shown that sulfide minerals in the region mainly originate from pyrite in Carboniferous and Permian coal seams [33]. Moreover, large-scale coal mining in the study area has exposed reducing inorganic sulfur minerals, accelerating the processes of weathering, oxidation, and dissolution. These minerals then enter into the groundwater through rainfall runoff, thereby causing an increase in the sulfate concentration of groundwater [34]. The positive correlation between DO and SO₄²⁻, as reflected in PC3, supports the conclusion that the oxidation of sulfide minerals, particularly pyrite, is a significant source of sulfate in the groundwater. This finding aligns with the hydrogeochemical processes typical of mining-impacted regions, where the disruption of coal seams and associated sulfide minerals introduces additional pathways for oxidation and subsequent sulfate enrichment in groundwater. Therefore, PC3 demonstrates that SO₄²⁻ in groundwater originates from the oxidation of sulfide minerals.

3.5. Spatial Distribution of Sulfate Sources in Groundwater

To analyze the spatial distribution characteristics of SO₄²⁻ sources in the Hutuo River alluvial fan groundwater, inverse distance weighting interpolation was performed on the scores of the three factors, resulting in a spatial distribution map of sulfate pollution sources in the Hutuo River basin (Figure 7). We employed cross-validation techniques to assess the accuracy and reliability of our interpolation model. The results showed that the mean prediction errors for PC1, PC2, and PC3 were −0.018, 0.026 and 0.012, respectively, with root mean square errors of 0.938, 1.106, and 0.857. The results of the cross-validation demonstrated that our interpolation model performed well.

As shown in Figure 7, high scores of PC1 are mainly located in the RVP and UAF (around power plants and reservoirs) (Figure 7a), indicating that SO₄²⁻ in groundwater in these regions primarily originates from industrial and domestic wastewater discharges. High scores of PC2 are mainly distributed in the RVP and sporadically in the study area (Figure 7b), suggesting that the dissolution of evaporites also contributes to sulfate in groundwater in these regions. High scores of PC3 are primarily located around the Huangbizhuang Reservoir in the river valley plain (Figure 7c), where coal-bearing strata are widely distributed. The oxidation of sulfide minerals is an important factor affecting SO₄²⁻ concentrations in groundwater in this region.

Based on the above analysis, the elevated SO₄²⁻ concentrations in the Hutuo River alluvial fan groundwater are mainly influenced by industrial activities around the Huangbizhuang Reservoir and coal mining in the Ye River basin. Intensive groundwater extraction has accelerated the dissolution of evaporites and the oxidation of sulfide minerals in the strata. Therefore, to control the continuous increase in SO₄²⁻ concentrations in groundwater in the Hutuo River alluvial fan area, strict measures should be implemented to regulate the discharge of untreated wastewater and prevent the overexploitation of groundwater.

4. The Limitations of This Study

This study provides a comprehensive analysis of the pollution characteristics and mechanisms of sulfate in groundwater in over-exploited areas, which can offer scientific evidence for the prevention and control of sulfate pollution in groundwater in similar regions both domestically and internationally. However, there are some limitations in this study. For example, the identification of sulfate sources in groundwater mainly relied on hydrochemical methods, which can only qualitatively recognize the sources of sulfate in the water environment, without quantifying the contribution rates of each sulfate pollution source. Therefore, future research should focus on improving the accuracy of sulfate source apportionment in over-exploited areas. Additionally, the study area is a thick multi-layered aquifer system in the North China Plain. Due to limitations in research objectives and funding, this study only collected shallow groundwater samples once and conducted a preliminary exploration of the pollution mechanisms of sulfate in groundwater in over-exploited areas. Future research should involve high-frequency sampling of groundwater at different depths and times in this region to deeply analyze the pollution mechanisms and evolution mechanisms of sulfate in over-exploited areas.

5. Conclusions

This study provides a comprehensive analysis of the spatial distribution and sources of sulfate (SO₄²⁻) in the groundwater of the Hutuo River alluvial fan, an intensively exploited region where groundwater quality is influenced by complex natural and anthropogenic interactions. Our innovative integration of multivariate statistical techniques, including principal component analysis (PCA), with advanced geostatistical methods has revealed unprecedented insights into the heterogeneity of SO₄²⁻ distribution across distinct hydrogeological units. A key novel finding is the demonstrated relationship between groundwater depth and SO₄²⁻ concentration, establishing groundwater depth as a critical factor in sulfate distribution. This discovery enhances our understanding of how groundwater dynamics influence solute transport and distribution in overexploited aquifers.

Our spatially explicit analysis identifies industrial activities as the primary driver of elevated sulfate levels in the river valley plain and upper alluvial fan, contrasting with the central alluvial fan where natural processes dominate. Beyond anthropogenic sources, we have uncovered the significant contribution of natural processes—specifically evaporite dissolution and sulfide mineral oxidation—to groundwater sulfate pollution. The methodological framework developed in this study offers a cutting-edge approach for identifying and mapping pollution sources in complex hydrogeological settings. By combining PCA with geostatistical techniques, we created detailed spatial distributions of sulfate sources, providing a clearer picture of pollution dynamics than previously possible. These results highlight the need for a paradigm shift in groundwater management, emphasizing that effective pollution control requires both source-specific mitigation and systemic hydrogeological considerations.

The outcomes of this research not only advance the scientific understanding of sulfate pollution dynamics in groundwater systems but also deliver actionable knowledge for policymakers and water resource managers. By providing a robust scientific basis for sustainable groundwater management, this study serves as a model for addressing groundwater pollution challenges in other regions facing similar threats from overexploitation and contamination. This work represents a significant step forward in the development of science-based solutions for groundwater protection and sustainable water resource management.