Chemical Characterization and Source Analysis of Shallow Groundwater in a Typical Area of Huaihe River Basin

Abstract

In this paper, water quality samples were collected from 215 sampling sites in Huaibin County, Xinyang District, Huaihe River Basin, in May 2024, and 11 key indicators of groundwater quality were analyzed. On the basis of hydrochemical statistics and water quality analysis to determine the water quality categories and characteristic pollutants, principal component analysis (PCA) was used to summarize the main driving factors affecting water quality, and it was combined with the absolute principal component score-multiple linear regression receptor model (APCS-MLR model) to further quantify the degree of influence of anthropogenic and natural factors on groundwater quality in the basin. The results showed that about 52% of the groundwater exceeded the Class III water standard of Groundwater Quality Standard (GB/T 14848-2017). Four types of principal component factor affecting the water quality were extracted by principal component analysis (PCA), which were dissolved filtration, migration enrichment (37.39%), agricultural surface pollution (15.52%), leaching and agricultural surface pollution (11.07%) and industrial pollution factor (10.24%). The APCS-MLR model was used to further quantify the effects of various anthropogenic and natural factors on water quality. The average contributions of the five factors to the groundwater quality in the basin were 66.51%, 51.66%, 19.61% and 78.13%, respectively, and the average fitting coefficient of the measured and predicted values of each index was 0.74. This method is highly relevant to the calculation of the allocation of the groundwater pollution sources, and it is suitable for the analysis of the groundwater pollution sources.

1. Introduction

Groundwater pollution is a complex and challenging environmental problem, mainly due to its hidden nature and slow pollution process, and it is difficult to manage. Huaibin County is located in the middle and upper reaches of Huaihe River, with a warm and humid climate, abundant water resources, fertile land and mainly agricultural production. In recent years, Huaibin County has been vigorously developed for its agriculture, and it relies on the port resources provided by the Huaihe River to actively develop the port economy. It has formed a “one textile and two manufacturing” industrial system mainly focusing on fashionable textiles, modern boats and green food. The rapid development of industrialization, urbanization and agricultural modernization has also brought pollution risks to the shallow groundwater in Huaibin County, and most of the groundwater in Huaibin County is Class VI and Class V water, the cause of which may be the massive use of pesticides and chemical fertilizers brought about by agricultural activities, and which may also be the cause of industrial wastewater discharges, etc. The region has not systematically carried out groundwater pollution evaluation studies, so the sources of groundwater pollution remain unclear. Therefore, there is a lack of a comprehensive and systematic understanding of the sources of groundwater pollution, and insufficient research on the potential hazards of groundwater pollution. Against the backdrop of accelerated modernization, the deterioration of groundwater environment caused by industrial pollution, agricultural non-point source pollution and domestic sewage discharge has become a global environmental governance challenge, urgently requiring the prevention and control of groundwater pollution. In groundwater pollution prevention and control strategies, the identification of groundwater pollution sources is the core foundation for groundwater environmental management and pollution prevention and control, with pollution source control being a top priority. Given the invisibility of pollutant migration and the lag effect in groundwater systems, accurately identifying pollution sources and analyzing their spatio-temporal distribution characteristics, has become a key link in the control of groundwater pollution sources. Therefore, systematically carrying out the identification and analysis of pollution sources provides an important decision-making basis for preventing and controlling groundwater pollution and achieving sustainable groundwater management. Therefore, how to accurately identify the pollution sources affecting regional groundwater quality and calculate the relative contribution is of great significance to the development of local production and life, as well as the exploitation and utilization of groundwater and its prevention and treatment.

The quantitative identification of factors affecting groundwater quality belongs to the category of groundwater pollution source analysis [1], and the main methods include inventory analysis, diffusion modeling and receptor modeling [2]. Among them, inventory analysis and diffusion modeling are based on the emission of each pollutant and the status quo of the surrounding environment [3], while receptor modeling is mainly based on the characteristics of the pollutant receptor to reverse analyze the pollution source [4]. In contrast, receptor modeling is a widely used and effective source analysis technique because it does not need to provide data on the emission conditions of the pollutant sources and the surrounding environment, nor does it need to track the migration and transformation patterns of the pollutants [5,6]. Among the source analysis methods of receptor modeling, isotope tracing, principal component analysis combined with multiple linear regression (PCA-MLR) [7,8] and absolute principal component score-multiple linear regression (APCS-MLR) are more frequently used [9,10]. The isotope tracing method focuses on specific pollutants, such as nitrate and sulfate pollution [11,12], while the regression modeling method focuses on the integrated regional pollution problem [13]. Compared with the APCS-MLR model, the principal component score matrix obtained in the PCA-MLR model does not represent the real principal component scores [14], while the introduction of the absolute zero factor scores solves this problem [15], and the pollutant contribution rate can be obtained after multiple linear regression calculation [16]. Therefore, the PCA-APCS-MLR model was used to quantitatively identify the factors affecting groundwater quality in the Huaihe River Basin, which can provide a more comprehensive and in-depth understanding of the groundwater environmental conditions in the study area.

In summary, this paper intends to take the groundwater in Huaibin County of Huaihe River Basin as the research object, and, based on the evaluation of groundwater quality, firstly adopt the principal component analysis (PCA) method to analyze the main indexes affecting the quality of groundwater by dimensionality reduction, and extract the main environmental factors, and then quantify the contribution of various types of environmental factors to the influence of groundwater chemical indexes in the basin by using the APCS-MLR model, which is not only expected to fill the gap of quantitative identification of groundwater quality and influencing factors, but also provide important data support for the formulation of groundwater pollution control measures in this area. The results of this study are not only expected to fill the gap of quantitative identification of groundwater quality and influencing factors in the Huaihe River Basin in Huaibin County, but also to provide important data support for the formulation of measures to prevent and control groundwater pollution in this area.

2. Overview of the Study Area

Huaibin County, a county under the jurisdiction of Xinyang City, Henan Province, is located in the northeastern part of Xinyang City, between longitude 115°11′–115°35′ E and latitude 32°15′–32°38′ N, with a total area of 1209 square kilometers. Huaibin County is located in the transition area between the northern subtropical zone and the warm temperate zone, belonging to the warm temperate humid climate, with an average annual temperature of 15.5 °C. The river surface area of the Huai, Hong, Bai and Lu Rivers in Huaibin County is about 23.7 km², accounting for 2% of the total area of the county. The main stream of the Huai River passes through the county from west to east, and the Hong River, Bailu River and Lv River flow into the Huai River in the territory, together with Wulong Harbor and the Happiness River, which are rivers in the county, with a river area of 36,000 mu. The average multi-year precipitation in Huaibin County is 938.0 mm, the average runoff depth is 274.0 mm, the average runoff volume is 331 million m³, and the shallow groundwater reserve is 5.37 billion m³, with an average annual extraction volume of 156 million m³. Total water resources total 480.5 million m³, 990 m³ per capita, slightly lower than the provincial level.

Huaibin County is connected to the Dabie Mountain gentle hills in the south and belongs to the Huanghuai Plain in the north, which is a transition area from hills to plains, and is an integral part of the southern edge of the Huanghuaihai Plain of Henan Province. The overall terrain is relatively flat, tilting from west to east, north of the Huaihe River, and gradually decreasing from northwest to southeast, south of the Huaihe River. Controlled by the geological structure, the landforms are roughly divided into three types: granite, plains and depressions. The study area is located in the north of the Zhushan-Gushi rupture in the regional geological structure, and belongs to the southern edge of the North China Plain, which is covered by the Paleoproterozoic, Neoproterozoic and Quaternary strata, with a thickness of 200 m–1000 m (thin in the southeast and thick in the northwest). Below the cover is mainly Mesozoic Jurassic, with quartz sandstone and conglomerate. In the south, there is part of Paleozoic Cambrian and Ordovician exposed, with gray rock as the main lithology. Groundwater in the study area is mainly pore-pressurized water of loose rocks of the fourth system, which is stored in fine sand and widely distributed, with powdery loam constituting the top plate of its water insulation, and powdery clay constituting the bottom plate of its water insulation. According to the storage, transportation and discharge characteristics of groundwater, pressurized water is mainly recharged by lateral runoff, and discharged mainly by lateral runoff and artificial mining.

3. Sample Collection and Testing

3.1. Layout of Sampling Points

The deployment of monitoring wells should consider the distribution of pollution sources and the diffusion of pollutants in groundwater. In order to deeply study the hydrogeochemical characteristics of groundwater in Huaibin County, according to the “Technical Specification for Groundwater Environmental Monitoring” (HJ/T164), combining the different hydrogeological conditions of the study area and the functions of groundwater monitoring wells, as well as the actual situation of pollution sources and pollutant discharge, the point-face combination of the distribution method is adopted, and the groundwater sampling section is laid out in the perpendicular direction of the flow direction of the groundwater in the study area, and the principle of face-control point and local sampling is adopted. Based on the principle of surface control points and local sampling, 41 key double-source investigation objects were selected for the environmental conditions of groundwater sources in the county, monitoring wells were set up, and 215 groups of shallow groundwater samples were collected, with sampling wells of less than 100 m, most of which were within the range of 0 m to 50 m, and the distribution of the specific sampling points was as shown in Figure 1.

3.2. Sample Testing

Groundwater sampling in the study area, as well as measurement methods, were operated with reference to the “Environmental Protection Industry Standard of the People’s Republic of China-Technical Specification for Environmental Monitoring of Groundwater” (HJ/T164-2004) [17]. Groundwater level (buried water level) and well washing were measured before sampling, and each time the water level was measured, it was recorded whether the monitoring wells had been pumped and whether they were subject to pumping from nearby wells, and the well washing fulfilled the relevant requirements of the Technical Guidelines for Environmental Monitoring of Sites. Wastewater is measured in the field using portable water quality meters, and field monitoring items include water level, water temperature, pH, conductivity, turbidity, redox potential, color, odor and taste, and visibility to the naked eye, as well as measuring air temperature, describing weather conditions and collecting recent precipitation and recording latitude and longitude of the site. According to the sampling requirements of different test indicators, sampling needs to be in a certain order: add the appropriate protective agent and, after sampling and sealing, sealed samples will be sent to the testing center laboratory in a timely manner for the detection of relevant data. The monitoring indexes are analyzed by all the indexes stipulated in the Groundwater Quality Standard (GB/T 14848-2017) [18], and the basic program of analytical methods is composed of Inductively Coupled Plasma Mass Spectrometry (ICP-MS), with Gas Chromatography Mass Spectrometry (GC-MS) as the main body, supplemented by other advantageous analytical methods such as Spectrophotometry, Iodine Amount Method and so on [19,20].

3.3. Data Quality Monitoring

In order to guarantee the scientific validity and accuracy of the water quality test results, it is necessary to verify the credibility of the test data obtained in the laboratory [21]. According to the principle of ionic charge balance, the total charge of cations and anions in the water body should maintain a dynamic balance. Based on this theory, quality control can be implemented by comparing the molar concentrations of anions and cations [22]. When the absolute value of the relative deviation between the anion molar concentration and the cation molar concentration in the water sample test result does not exceed 5%, it indicates that the test result meets the quality control standard; for the outliers exceeding this threshold, data traceability verification and nullification are required, so as to reduce the interference of the experimental systematic error or exogenous pollution factors on the subsequent study of the characterization of the chemical components of water. The formula is as follows:

$$EBC\left(\%\right)=\left[{\displaystyle \frac{\sum c^{+}-\sum c^{-}}{\sum c^{+}+\sum c^{-}}}\right]\times 100\%$$

(1)

where c⁺, c⁻ represent the molar concentration of cations and anions in the water sample taken, unit mmol/L. When E < 5% indicates that the water sample test results meet the requirements, otherwise it should be removed.

3.4. Methods of Analysis

PCS-MLR has significant benefits over PCA, which can simplify the data by dimensionality reduction and explore the potential relationship, but is unable to quantify the contribution of pollutant sources, unlike APCS-MLR, which not only has the advantage of simplifying the data by PCA, but also can accurately calculate the proportion of the influence of each pollutant source on the pollutant concentration through multiple linear regression. This is crucial when analyzing the causes of pollution and formulating treatment plans, and can help in precise policy making.

Researchers in many parts of the world have utilized these two methods to conduct groundwater contamination studies. In Jiangsu, China, researchers refer to “Characterization and Source Analysis of Organic Pollution in Shallow Groundwater Surrounding a Chemical Park in Jiangsu”, where PCA was used to screen key pollutants and APCS-MLR to quantify the sources of pollution, and chemical production was identified as the main source of pollution. In Tehran, Iran, scholars in the study “Assessment of groundwater quality and identification of pollution sources in the southern plain of Tehran, Iran” utilized PCA to identify pollution patterns and APCS-MLR to assess the contribution of each source. MLR to assess the contribution of each source and provide support for local groundwater protection. These studies show that APCS-MLR and PCA are powerful tools for analyzing groundwater pollution in different regions of the world.

The core of principal component analysis (PCA) is dimensionality reduction, which maps high-dimensional data to a low-dimensional space through linear transformation, retains the maximum variance information to achieve data simplification, identifies the main components using eigenvalue decomposition of the covariance matrix and extracts the principal controlling factors reflecting the common variability characteristics of the water quality parameters. Each of these principal components is a linear combination of the original variables, and each principal component is uncorrelated with each other, which makes the principal components have superior performance than the original variables. The specific steps include: ① standardize the original data to eliminate the influence of the scale; ② establish the correlation coefficient matrix R between the variables; ③ calculate the eigenvalues and eigenvectors of the correlation coefficient matrix R; and ④ extract the principal components (PCs) with eigenvalues greater than 1 and calculate the composite score.

The PCs can be expressed as:

Z_pn = a_p1·x_1n + a_p2·x_2n + … + a_pm·x_mn

(2)

where: Z—principal component score; a—measured value of each indicator parameter; p—number of principal components; n—number of samples; m—number of variables; x_mn-_m×n centered data Matrix.

The Absolute Factor Score Multiple Linear Regression (APCS-MLR) is based on the premise that the principal components of water quality indicators are obtained from PCA analysis, and the standardized principal component scores are further transformed into Absolute Principal Component Scores (APCS), and then multiple linear regression (APCS-MLR) analysis is performed.

$${{\displaystyle u}}_{ji}={\displaystyle \frac{\left|{{\displaystyle \mathrm{a}}}_{ij}{{\displaystyle \overline{s}}}_{ji}\right|}{{\displaystyle \sum _j\left|{{\displaystyle \mathrm{a}}}_{ij}{{\displaystyle \overline{s}}}_{ji}\right|+\left|{{\displaystyle b}}_i\right|}}}$$

(3)

$${\mathrm{c}}_i={\displaystyle \frac{\left|b_i\right|}{{\displaystyle \sum _j\left|{\mathrm{a}}_{ij}{\overline{s}}_{ji}\right|+\left|b_i\right|}}}$$

(4)

where u_ji is the average contribution of principal component j to pollution factor i, c_i is the average contribution of unknown sources to pollution factor i, and ${\overline{s}}_{ji}$ is the average of the j absolute principal component scores of pollution factor _i. a_ij is the regression coefficient of principal component j to pollution factor i; b_i is the constant term in the multivariate linear regression equation of pollution factor i, which is generally considered to be the contribution of the unidentified sources. The average contribution of principal component j to pollution factor i can be calculated by Equation (3), and the contribution of unidentified sources can be calculated by Equation (4).

The contribution of each factor to the water quality index was calculated to quantitatively characterize the contribution of each pollution source to the quality of the water body, and the regression model predicted concentrations were linearly fitted to the measured concentrations to test the correctness of the results of the analysis of pollution sources [23,24].

4. Results and Discussion

4.1. Groundwater Chemical Characterization and Quality Assessment

4.1.1. Groundwater Chemical Characterization

The test results of the water samples in the study area are shown in

Table 1. Water chemical composition statistics.

	Maximum Values	Minimum Value	Average Value	(Statistics) Standard Deviation	Coefficient of Variation
pH	9.09	6.32	7.251	0.421	0.06
TDS	908.00	25.00	400.71	126	0.32
K+	116.00	0.24	1.75	8.63	4.92
Na+	140.00	9.40	43.76	20.94	0.48
Mg2+	47.80	6.75	19.75	7.94	0.40
Ca2+	175.00	25.50	78.41	28.96	0.37
Cl−	236.00	2.67	37.48	35.80	0.96
SO42−	198.00	0.86	25.76	24.88	0.97
HCO3−	650.00	67.80	320.04	85.72	0.27
Fe	7.32	0.01	0.162	0.647	4.00
Mn	4.32	0.004	0.697	0.840	1.20
Pb	0.028	0.0004	0.002	0.004	2.00
As	0.035	0.0003	0.002	0.005	2.50
F	8.58	0.155	0.400	0.623	1.56
NH4+	0.914	0.04	0.179	0.138	0.77

Note: Water quality indicators are in mg/L; pH dimensionless.

, the range of pH in the water samples was 6.32–9.09, and the mean value was 7.25, which indicated that the water samples in the study area were weakly alkaline, in which the coefficient of variation of pH was the smallest, 0.06, which indicated that the spatial distribution of pH values was balanced, and the influence on spatial variations was relatively small. The content of TDS ranged between 25 and 908 mg/L, and the mean value was 400.71 mg/L. The mean values of cation mass concentration were Ca²⁺ > Na⁺ > Mg²⁺ > K⁺ in descending order; the mean values of anion mass concentration were HCO₃⁻ > Cl⁻ > SO₄²⁻ in descending order, and the cation Ca²⁺ had the highest content, while the anion HCO₃⁻ had the highest content.

Ion correlation analysis can reflect the similarity, dissimilarity, and consistency and difference in the source of groundwater aqueous chemical constituents. The correlation analysis of the groundwater sample data in the study area was carried out using origin software to derive the correlation coefficients between the major anions and cations in the groundwater. As can be seen in Figure 2, among the cations, TDS was more strongly correlated with Ca²⁺ and Mg²⁺, with correlation coefficients of 0.87 and 0.9, respectively. Among the anions, TDS was more strongly correlated with Cl⁻ and SO₄²⁻, with correlation coefficients of 0.75 and 0.65, respectively. Of these, the strongest correlation was observed between TDS and Ca²⁺ (R = 0.9, p < 0.001), indicating that Ca²⁺ is the main contributor to TDS in the water samples from the study area. Ca²⁺ correlates strongly with Mg²⁺ and Cl⁻, with correlation coefficients greater than 0.7; Mg²⁺ correlates strongly with Cl⁻ and SO₄²⁻, with correlation coefficients greater than 0.65, and Na⁺ correlates strongly with Mg²⁺ and HCO₃⁻. The correlation coefficients were around 0.5. From the above results, it can be inferred that carbonate minerals such as calcite, sulfate minerals, and the weathering and leaching of rock salt are the main sources of ions in the groundwater of the study area.

4.1.2. Groundwater Quality Assessment

Through the Nemero index formula and combined with

Table 2. Standard for the classification of groundwater quality.

Norm	Class I	Class II	Class III	Class IV	Class V
Cl−	≤50	≤150	≤250	≤350	>350
SO42−	≤50	≤150	≤250	≤350	>350
TDS	≤300	≤500	≤1000	≤2000	>2000
NO3−	≤2	≤5	≤20	≤30	>30
NO2−	≤0.01	≤0.1	≤1	≤4.8	>4.8
NH4+	≤0.2	≤0.5	≤1	≤1.5	>1.5
F−	≤0.5	≤1	≤1.5	≤2	>2
As	≤0.001	≤0.002	≤0.01	≤0.05	>0.05
Fe	≤0.1	≤0.2	≤0.3	≤2	>2
Mn	≤0.04	≤0.05	≤0.1	≤1.5	>1.5
COD	≤1	≤2	≤3	≤10	>10
total coliform bacteria	≤3	≤3	≤3	≤100	>100

, the evaluation map of shallow groundwater quality in the study area was derived, and it can be seen from Figure 3 that the polluted portion of the water samples from the water sampling points in the study area accounted for about half of the total, of which 24% of the groundwater was uncontaminated, 14% was mildly contaminated, 10% was moderately contaminated, and the proportion of more severely contaminated and heavily contaminated were 16% and 36%, respectively.

Firstly, the monitoring data were categorized into components based on the Class III limit value of Groundwater Quality Standard, and the baseline scores of each index were obtained through the grading assignment rules shown in Table 2; then, the comprehensive score value F was calculated by using the formula of the Nemero index, and the comprehensive index grade of groundwater contamination was classified in accordance with

Table 3. Grading table of the comprehensive index of groundwater pollution.

Pollution Category	I (Unpolluted)	II (Light Pollution)	III (Moderate Pollution)	IV (Heavier Pollution)	V (Heavy Pollution)
Index range	<1	1–2	2–3	3–5	>5

, and its mathematical expression was:

$$F=\sqrt{{\displaystyle \frac{F_{ave}^2+F_{max}^2}{2}}}$$

(5)

where F_ave represents the arithmetic mean of the individual pollution indices of all pollutants; F_max represents the maximum value of the individual pollution indices of all pollutants.

In this paper, based on the grading standards for each item in the Environmental Quality Standards for Groundwater (GBT14848-2017), based on the preliminary monitoring results of the shallow groundwater samples in the study area and combining the natural and anthropogenic influencing factors, a total of 12 items, including TDS, Cl⁻, SO₄²⁻, Mn, NO₃⁻, NO₂⁻, NH₄⁺, Fe, As, F⁻, COD, and total Escherichia coli, were selected as water quality evaluation indicators that were used as water quality evaluation indexes. Through the Nemero index calculation formula and combined with Table 3, the shallow groundwater quality evaluation map of the study area was derived, and it can be seen from Figure 3 that the polluted part of the water samples from the water sampling points in the study area accounted for about half of the total, in which the groundwater was not polluted accounted for 24%, mildly polluted accounted for 14%, moderately polluted accounted for 10%, and more seriously polluted and seriously polluted accounted for 16% and 36%, respectively.

4.1.3. Types of Groundwater Chemistry

In order to accurately show the relative content of each ion in the groundwater and the chemical type of groundwater, Piper’s trilinear diagram, one of the most widely used methods for analyzing water chemistry types, was chosen [25,26]. Piper trilinear diagram is based on the standardization technology of triangular coordinate system. It uses the main anions and cations in water as coordinate axes, combines mathematical transformations with geochemical mechanisms to draw a trilinear diagram. The red circles in the diagram represent sampling points, and the chemical type of groundwater can be intuitively displayed based on the projected positions of the sampling points in the diagram [27,28].

From the cation analysis (see Figure 4), it can be seen that the shallow groundwater is mainly dominated by Ca-Na-K type water, and from the anion analysis, it can be seen that the shallow groundwater is mainly dominated by bicarbonate. Analyzing the Piper trilinear map, it is found that the sample points are mainly distributed in areas 2, 3, 5 and 6. Based on the type classification, it can be determined that the water chemistry type in the groundwater in this region is HCO₃-Ca type water [29,30].

4.2. Analysis of Groundwater Pollution Sources and Spatial Distribution Characteristics

Based on the classification standards of each item in the Environmental Quality Standards for Groundwater (GBT14848-2017), and combining the preliminary monitoring results of shallow groundwater samples in the study area with natural and anthropogenic influencing factors, a total of 11 items were selected as water quality evaluation indicators. Eleven parameter indicators such as TDS, Mg²⁺, As, Cl⁻, HCO₃⁻, NO₃⁻-N, Mn, NO₂⁻-N, NH₄⁺-N, Ca²⁺ and K⁺ were selected for the downscaling of the groundwater sample data of the study area by using PCA, and the main control factors were extracted based on the correlation between the indicators [31,32]. Firstly, KMO and Bartlett’s sphericity test in SPSS27 software were applied to test the applicability of the standardized dataset, and the results showed that the KMO value was 0.652 > 0.5, the approximate chi-square of Bartlett’s sphericity test was 1132.51, the result of the df test of the degree of freedom was 55, and the probability of significance p was 0.000. The synthesis showed that the data among the 215 groups of groundwater samples collected in the current study had a certain correlation and can be used to identify the number of sources via the PCA method.

Next, principal components with eigenvalues greater than 1 were extracted, and a total of four principal components were extracted by PCA, with a cumulative variance contribution rate of 74.21% [33], indicating that the four factors are more concentrated in reflecting 74.21% of the information of the influencing factors [34], and the total variance and rotated factor loading matrices explained by the factor analyses are shown in

Table 4. Total variance explained.

Ingredient	Initial Eigenvalue			Extract the Sum of Squares and Load			Rotate the Sum of Squares to Load
	Add Up the Total	Variance (%)	Cumulative (%)	Add Up the Total	Variance (%)	Cumulative (%)	Add Up the Total	Variance (%)	Cumulative (%)
1	4.113	37.388	37.388	4.113	37.388	37.388	3.824	34.766	34.766
2	1.707	15.515	52.903	1.707	15.515	52.903	1.885	17.137	51.903
3	1.217	11.065	63.968	1.217	11.065	63.968	1.282	11.659	63.562
4	1.127	10.244	74.211	1.127	10.244	74.211	1.171	10.649	74.211
5	0.943	8.575	82.787
6	0.672	6.111	88.898
7	0.518	4.709	93.607
8	0.377	3.423	97.03
9	0.199	1.807	98.837
10	0.081	0.736	99.572
11	0.047	0.428	100

and

Table 5. Rotated factor loading matrix.

Norm	F1	F2	F3	F4
K+	0.134	−0.1	0.896	0.19
Ca2+	0.924	0.093	−0.095	0.034
Mg2+	0.907	0.178	0.064	0.011
HCO3−	0.664	−0.25	−0.559	0.243
Cl−	0.781	0.331	0.169	−0.204
Mn	0.216	0.659	−0.04	0.003
NO3−	0.333	0.665	0.243	−0.117
NO2−	−0.016	0.469	0.117	0.454
NH4+	−0.027	0.742	−0.184	0.186
As	0.03	0.04	0.067	0.879
TDS	0.959	0.12	0.114	0.08

.

The eigenvalue of the common factor F1 was 4.11 with a contribution rate of 37.39%, mainly including Ca²⁺, TDS, Cl⁻ and Mg²⁺ water quality indicators [35]. As the primary factor affecting groundwater quality in the study area, F1 includes six of the eight major ions in groundwater, with TDS having the highest loading coefficient. The study area is regionally geologically and tectonically located north of the Chesan-Gushi rupture, which belongs to the southern edge of the North China Plain, and is covered by Paleoproterozoic, Neoproterozoic and Quaternary strata, with a thickness of 200 m to 1000 m (thin in the southeast and thick in the northwest). Below the cover layer is mainly Mesozoic Jurassic, with the lithology of quartz sandstone and conglomerate, and in the south there is a part of Paleozoic Cambrian and Ordovician exposed, with the lithology of chert as the main one [36]. The aquifer is characterized by loose particles, pore development, good water permeability and relatively good runoff conditions. The strong alternating action causes the dissolution and filtration of calcium and magnesium compounds into ions in the rock layer, which increases the concentration of soluble ions, and thus influences the rise of total dissolved solids (TDS). According to the spatial distribution of the factor 1 scores (see Figure 5), the distribution pattern of the principal component F1 scores is high in the north and low in the south, which is due to the fact that the topography of the study area north of the Huaihe River is a granite, and south of the Huaihe River is a plain, and thus the migratory-enrichment effect is obvious. The range of the score 1.55–5.09 is mainly distributed in the northeast of Zhaojizhen and Sankongqiao Township. In the direction of Zhaozhi Township to Fanghu Township in the study area, there are folds and fractures in the strata, thus forming local uplift and downlift, and the stratigraphic direction is basically the same as that of the regional tectonic line, with a tendency to the northeast, and the thickness of the aquifer is larger than that of the surrounding area [37,38], so the migration-enrichment effect on the groundwater is more obvious, and therefore F1 can be identified as the factor of dissolution filtration-migration-enrichment effect that affects the quality of the groundwater.

The eigenvalue of the male factor F2 was 1.71, with a contribution rate of 15.52%, mainly including NO₃⁻-N, Mn, NO₂⁻-N and NH₄⁺-N water quality indexes, among which NH₄⁺-N was the largest, and the main sources of ammonia nitrogen in groundwater could be divided into human activity discharge and natural sources. According to the spatial distribution of Factor 2 (see Figure 6), the factor scores on both sides of the Huaihe River are low, while the factor scores in Zhaojizhen, Sankongqiao Township, the southwestern part of Luji Township and the southeastern part of Wangdian Township are high. Eutrophication: under the leaching effect of rainfall, the phosphorus and nitrogen elements in the randomly piled manure will be transformed into nitrate and phosphate after entering the soil, causing groundwater pollution [39]. According to the land use planning and land use status quo in the study area, 80% of the land in the study area is designated as permanent basic farmland, and the study area vigorously promotes agricultural development, so the application of chemical fertilizers in the process of agricultural production, and the infiltration of agricultural irrigation water will cause surface pollution of groundwater. In addition, the untreated discharge of domestic sewage and industrial wastewater can also lead to excessive nitrogen content in groundwater [40]. Therefore F2 reflects that groundwater quality is affected by agricultural production activities and is an agricultural surface pollution factor [41].

The eigenvalue of the common factor F3 is 1.22, with a contribution of 11.07%, and consists mainly of K⁺ and HCO₃⁻ water quality indicators.

Among them, K⁺ is the largest, and the main lithology below the overburden in the study area is quartz sandstone and conglomerate, with part of the lithology in the south being dominated by tuff [42]. K⁺ is mainly derived from the dissolution of sedimentary rocks containing potassium salts, and we know that according to the spatial distribution of Factor 3 (see Figure 7), the highest-scoring area is mainly located in the western part of Zhangzhuang Township, which is located on both sides of the Huaihe River and is vigorously developing its agriculture, and has been influenced by the policy in recent years. Green ecological agriculture actively implements soil testing, formula fertilization and organic fertilizers, implements pesticide reduction and pest control technology, and applies a large amount of nitrogen, phosphorus and potash fertilizers to the main crops, which is also the reason for the high content of K⁺, so Factor 3 can be regarded as a factor of dissolved filtration and agricultural surface source pollution [43].

The eigenvalue of the common factor F4 is 1.13 with a contribution rate of 10.24%, mainly As. The high levels of nitrite and arsenic in factor 4 may be attributed to the fact that, under natural conditions, the input of nitrogen into the groundwater environment is mainly dependent on atmospheric nitrogen deposition and biological nitrogen fixation processes [44]. However, with the rapid development of modern industrial and agricultural systems, the intervention of human activities in the nitrogen cycle has increased significantly. Nitrate nitrogen is mainly derived from the end products of nitrification, organic or ammonia nitrogen is completely converted to nitrate by nitrifying bacteria, or nitrite is further converted to nitrate by nitrosating bacteria. Nitrogen in the form of nitrate in water is mainly derived from agricultural fertilizers, industrial wastewater and nitrogen oxides in atmospheric deposition. Nitrogen loss due to excessive application of fertilizers in agriculture and excessive application of fertilizers leads to elevated nitrate and nitrite concentrations in soil and water bodies. Most of the study area is zoned for agriculture and excessive application of fertilizers may trigger eutrophication of water bodies, acid rain and soil degradation. Groundwater table depth in the study area is <10 m in most areas, where nitrification prevails and NO₃⁻ is easy to directly infiltrate [45]. In recent years, Huaibin County has put forward the strategy of “Industrial Strengthening County”, in which the textile and garment industry is the leading industry in Huaibin County. Wastewater from textile printing and dyeing processes (e.g., dyeing, rinsing) contains a large amount of nitrogenous organics (e.g., dyes, auxiliaries) and ammonia nitrogen (NH₄⁺), and industrial activities (e.g., fertilizer factories, leather processing) may indirectly affect NO₃⁻ through NH₄⁺ inputs, but their contribution is usually lower than that of agricultural sources. The pH of the study area ranges from 6.5 to 9.09 under neutral to weakly alkaline conditions. As Under neutral to alkaline conditions (pH > 6.5), arsenic exists mainly in the form of arsenate (As⁵⁺), which is highly soluble and mobile, and readily disperses with groundwater or surface water. A weakly alkaline environment (pH 7–9) may promote the dissolution of arsenic-containing minerals (e.g., iron arsenate and calcium arsenate) and increase the activity of arsenic in the water body [46]. The industrial agglomeration in the study area is dominated by textile industrial parks, and the mixing of alkaline wastewater from textile factories (pH 9–10) and acidic wastewater from chemical factories may promote the re-dissolution of arsenic-containing precipitates if they are neutralized to pH 8–9 [47], which is presumed to be caused by wastewater discharge (see Figure 8).

4.3. Calculation and Analysis of the Contribution of Pollution Sources

On the basis of PCA analysis to determine the composition and spatial distribution characteristics of the main pollution sources in the study area, the absolute principal component score multiple linear regression receptor model (APCS-MLR) was used to calculate the contribution rate of the public factors to the water quality indicators and the predicted results were linearly fitted to the measured results [48], and the fitted curves of the typical representative pollutants were selected, as shown in Figure 9, Figure 10, Figure 11 and Figure 12, which were used to further verify the results of the factor analysis [49]. The results are shown in Table 5.

The analysis results show that most of the linear fitting R² of the predicted and measured values of the APCS-MLR model are greater than 0.6, and the average fitting coefficient is 0.74, indicating that the two are in good agreement. The calculation of the pollution contribution rate of each factor is more accurate, which indicates that the APCS-MLR model has good applicability for the calculation of the allocation of groundwater pollution sources. The representative indicators of each factor were selected to be plotted as shown in Figure 9, Figure 10, Figure 11 and Figure 12, and the results of factor analysis are shown in Table 5.

As shown in

Table 6. Contribution of factors to each indicator.

	Lysofiltration Migration Enrichment Factor	Agricultural Surface Pollution Emission Factors	Dissolved Filtration and Agricultural Surfacepollution Factors	Geoenvironmental Factors	Other Sources	R2
K+	8.45	6.46	56.80	11.88	16.41	0.87
Ca2+	76.96	7.09	7.93	2.80	5.22	0.88
Mg2+	75.52	15.33	5.57	0.80	2.78	0.85
HCO3−	32.54	12.23	27.33	11.90	16.00	0.87
Cl−	45.95	19.21	9.78	12.05	13.01	0.79
Mn	22.22	67.90	4.12	0.37	5.39	0.46
NO3−	22.42	44.83	16.14	7.63	8.98	0.62
NO2−	1.70	42.44	10.19	40.75	4.93	0.43
NH4+	2.00	51.46	12.49	12.49	21.56	0.61
As	0.43	0.62	0.49	78.13	20.33	0.78
TDS	67.60	8.95	8.01	5.76	9.68	0.96

, the contribution rate of leaching migration enrichment factor F1 to Ca²⁺, TDS, Cl⁻ and Mg²⁺ indicators was 76.96%, 67.60%, 45.95% and 75.52%, respectively, and the contribution rate to other indicators ranged from 0.43% to 32.54%. The pollution of the groundwater environment by the agricultural surface pollution emission factor F2 is mainly reflected in the nutrient elements, and the contribution rates to the indicators of NO₃⁻-N, Mn, NO₂⁻-N and NH₄⁺-N are 44.83%, 67.90%, 42.44% and 51.46%, respectively, which shows that the agricultural pollution has a certain polluting effect on the nutrient elements, metal ions and organic matter in the groundwater. This shows that agricultural pollution has an effect on nutrient elements, metal ions and organic matter in groundwater quality. The contribution of dissolved filtration and agricultural surface pollution factor F3 to K⁺ and HCO₃⁻ indicators were 56.80% and 27.33%, respectively. The reason for this is that, in recent years, Huaibin County vigorously develops green ecological agriculture, actively promotes soil testing and formula fertilization and organic fertilizer, implements pesticide reduction and pest control technology, and applies a large amount of nitrogen, phosphorus and potassium fertilizer to the main crops, which is the reason for the high K⁺ content. The contribution of other indicators ranged from 0.49% to 16.14%. Because of the industrial wastewater discharge in the study area, its influence on the groundwater quality is mainly reflected in the change of As concentration, with a contribution rate of 78.13%, and the contribution rate of other indicators is basically less than 13%. In summary, the contribution of public factors from four types of pollution sources to groundwater quality pollution is 37.39%, 15.52%, 11.07% and 10.24%, respectively. Therefore, the dissolution and filtration of groundwater, migration and enrichment, and agricultural surface pollution, have a great influence on groundwater quality.

5. Conclusions

(1)

The four main sources of groundwater pollution in the Huaihe River basin in Huaibin County are F1 leaching, migration and enrichment, with a variance contribution of 37.39%, and the pollution areas are mainly concentrated in the north-eastern part of Zhaojizhen and Sankongqiao Township. F2 agricultural surface pollution, with a variance contribution of 15.52%, is mainly found in the south-western part of Zhaojizhen, Sankongqiao Township, Lujizhen Township and the southeastern part of Wangdian Township. F3 leaching and agricultural surface pollution factor, the variance contribution rate is 11.07%, the pollution area is mainly in the western part of Zhangzhuang Township; F4 industrial pollution factor, the variance contribution rate is 10.24%, the pollution area is mainly concentrated in the eastern part of Wangjiagang Township and Qisi Township.

(2)

The APCS-MLR model was used to calculate and analyze the contribution of each factor to the chemical indicators of groundwater in the basin, and the average contribution of the four factors to the characteristic indicators was 66.51%, 51.66%, 19.61% and 78.13%, respectively, which resulted in the average fit coefficient of 0.74 between the measured and predicted values of the indicators, which indicated that the results of the calculations could accurately analyze the contribution of the indicators to the pollution in the study area. Indicator contribution.