Surface Water Quality Evaluation and Pollution Source Analysis at the Confluence of the Wei River and Yellow River, China

Abstract

Water quality is a critical aspect of environmental health, affecting ecosystems, human health, and economic activities. In recent years, increasing pollution from industrial, agricultural, and urban sources has raised concerns about the deterioration of water quality in surface water bodies. Therefore, this study investigated the spatio-temporal distribution of water elements, human health risks of surface water, and pollutant sources at the confluence of the Wei River and the Yellow River. Using 80 samples collected during both wet and dry seasons, the content of the 22 water chemistry indicators was tested. A statistical analysis, Piper diagram, and entropy water quality index were employed to analyze the chemistry indicator content, hydrochemical composition, and water environmental quality of the surface water in the area. Moreover, the health risk assessment model was utilized to evaluate the carcinogenic and non-carcinogenic health risks associated with heavy metal elements in surface water. Finally, correlation heatmaps and a principal component analysis were used to identify potential pollution sources in the study area. The results indicated that $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ and ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N}$ were the main pollutants during the wet season, while surface water quality during the dry season was mainly influenced by ${\mathrm{F}}^{-}$. The hydrochemical type in the study area was mainly ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$. The health risk assessment revealed a high carcinogenic risk in the study area, with $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ being the primary heavy metal element contributing to health risks. The correlation and principal component analysis results show that the surface water environment in the study area was influenced by soil characteristics (soils containing ${\mathrm{F}}^{-}$ in the Dalí region, soils containing heavy metals in the Tongguan region), native geological environment (mineral resources and terrain conditions), and industrial activities (ore smelting). This study identified the key pollution indicators, the priority control areas, and the extent of the human health impact of the surface water at the confluence of the Wei River and the Yellow River, guiding targeted management of surface water environments.

1. Introduction

Surface water is a crucial water resource for agricultural irrigation and drinking water [1]. With the acceleration of industrialization and urbanization in China, a substantial amount of toxic pollutants is discharged into surface water, posing both direct and potential threats to the ecological environment and human health [2]. Therefore, to effectively protect water resources, scholars are increasingly focusing on river water quality assessment and identification of major pollution sources [3,4,5].

The confluence of the Wei River with the Yellow River is an area abundant in surface water resources, extensively used for agricultural irrigation and reservoir storage. This plays a vital role in human well-being and socio-economic development. However, due to the rich mineral resources in the research area, substantial tailings are generated from ore extraction in the eastern Qinling Mountains. Previous studies [6,7] have indicated that the development of the Tongguan Gold Mine in the 1980s resulted in a large amount of inadequately processed waste occupying land. Other studies [8,9,10] revealed elevated levels of heavy metals, including Pb, Cd, and Hg, in the surface soil of Tongguan County, exceeding national standards. A principal component analysis and spatial analysis suggested that heavy metals such as Hg, Cd, Pb, Cu, and Zn primarily originate from industrial sources such as mining and tailings, while Cr and Ni mainly come from natural sources such as parent material [11]. The accumulation of elements in river sediment can indirectly reflect the environmental quality of the river [12], and studies have shown that the heavy metals in the sediment of the Taiyu River and Taiyu Reservoir near the Qinling mining area correspond to heavy metal elements in the mining area [13,14]. These findings indicate that the soil and river sediment in the study area are severely polluted with heavy metals, posing a significant risk to the water quality of surface rivers.

Additionally, changes in land use in the research area can also impact surface runoff water quality [15,16]. In recent years, some farmland in the Dali area has been converted into ponds. The accumulation of residual bait and chemical drugs during the aquaculture process may lead to changes in various physicochemical factors and sediment environments [17,18,19].

In conclusion, the sources of surface water pollution in the study area are complex, and the degree of pollution is unclear. A systematic assessment of surface water quality at the confluence of the Wei River with the Yellow River is needed to quantify the major pollution sources and the impact of heavy metals on human health in the area.

Therefore, this study collected water samples during both wet and dry seasons from upstream, midstream, and downstream in the study area. The basic water quality indicators were tested according to the “Environmental quality standards for surface water” (GB 3838-2002; Environmental quality standards for surface water. China Environment Publishing House, China, 2002) [20]. Statistical methods and the Piper diagram were employed to analyze the hydrochemical characteristics and spatiotemporal variability of water quality in the study area [21,22]. The entropy-weighted water quality index (EWQI) was used to assess surface water quality [23,24]. To explore the impact of heavy metals on human health, the USEPA-recommended health risk assessment model was used to conduct health risk assessments for both adults and children [25,26]. Simultaneously, a principal component analysis (PCA), was applied to clarify the sources of surface water pollution in the area [27,28]. The main novelty of this study lies in combining traditional water quality assessment methods with advanced statistical techniques and health risk assessment models. This comprehensive approach allows for a more thorough understanding of surface water pollution in the confluence area of the Wei and Yellow Rivers, providing new insights and scientific references crucial for developing effective strategies to prevent and control surface water pollution.

2. Materials and Methods

2.1. Study Area

The study area is located at the confluence of the Shaanxi, Shanxi, and Henan provinces, specifically at the mouth of the Wei River into the Yellow River, with longitude ranging from 109°50′ to 110°25′ east and latitude from 34°25′ to 34°50′ north, as shown in Figure 1. The region falls within the warm temperate zone of the monsoon region, characterized by a continental semi-arid climate [8]. The climate features distinct seasons: ample sunshine and an annual rainfall of approximately 666.2 mm. The precipitation from June to October accounts for over 75% of the annual average, totaling around 500.3 mm. The annual average temperature is 14.3 °C. Wind speeds vary from 3 to 15 m/s, predominantly from the east. The annual evaporation is estimated to be within the range of 1300 to 1700 mm [8,29].

The topography in the study area is divided into five categories from south to north: the Qinling mid–low Mountains, the alluvial plain at Piedmont, the loess tableland, the Wei River alluvial plain, and the river wetland [30]. The terrain exhibits significant undulations, forming a stepped pattern. The terrain of the basin in the study area is characterized by short sources and rapid flows, and a small water collection area [31]. The Qinling Mountains are rich in minerals, including gold, brass ore, pyrite, and galena [32]. In the 1980s, unregulated mining activities were prevalent in these mountains, leading to indiscriminate dumping of tailings along riverbanks. The alluvial plain at Piedmont was also dotted with smelting plants, tailings ponds, cyanide pools, and tailings piles without containment. Erosion by surface river flow has caused severe degradation in the loess tableland area, forming loess pillars and cliffs with depths ranging from 10 to 30 m [33]. The Wei River alluvial plain, which is distributed along the banks of the Yellow and Wei Rivers, has a gently sloping terrain [34]. The river wetland is mainly found on the right bank of the Yellow River, characterized by lush vegetation and species richness. The loess tableland and river alluvial plain areas are predominantly cultivated with wheat and corn. Additionally, large areas of the river wetland have been developed with extensive lotus ponds for growing lotus roots and breeding crayfish [35,36].

2.2. Sample Collection and Analysis

To accurately reflect the water quality conditions in the study area, this study reasonably distributed sample points based on different types of water bodies and their potential impact on water quality. Specifically, 27 samples were collected from the upstream, midstream, and downstream sections of the rivers, representing the flow path and potential pollution sources along the river. Additionally, 11 samples were collected from river wetlands to capture the influence of wetland ecosystems on water quality, and 2 samples were collected from reservoirs to assess the impact of stored water. In total, 80 samples were collected during both the wet season (October 2022) and the dry season (March 2023). The sampling locations were recorded by coordinates using Real-time Kinematic (RTK) positioning and are shown in the Table S1 in the Supplementary File. The sample collection and storage strictly followed the guidelines outlined in “Water quality-Guidance on sampling techniques” (HJ 494-2009; Water quality-Guidance on sampling techniques, China Environment Publishing House, China, 2009) and the “Water quality sampling—technical regulation of the preservation and handling of samples” (HJ 493-2009; Water quality sampling—technical regulation of the preservation and handling of samples, China Environment Publishing House, China, 2009) [37,38]. The 22 water chemistry indicators were tested, including ${\mathrm{F}}^{-}$, ${\mathrm{C}\mathrm{l}}^{-}$, ${\mathrm{S}\mathrm{O}}_4^{2-}$, ${\mathrm{C}\mathrm{O}}_3^{2-}$, ${\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}$, ${\mathrm{N}0}_3\text{-}\mathrm{N}$, ${\mathrm{B}\mathrm{r}}^{-}$, $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$, ${\mathrm{N}\mathrm{a}}^{+}$, ${\mathrm{K}}^{+}$, ${\mathrm{C}\mathrm{a}}^{2+}$, ${\mathrm{M}\mathrm{g}}^{2+}$, Fe, ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N}$, $\mathrm{C}\mathrm{u}$, Hg, $\mathrm{C}\mathrm{d}$, $\mathrm{P}\mathrm{b}$, Mn, Ni, $\mathrm{A}\mathrm{s}$, and Ag.

2.3. Methodology

2.3.1. Entropy-Weighted Water Quality Index (EWQI)

The EWQI combines the information entropy method with the Water Quality Index (WQI), using information entropy to calculate the weights of indicators [39]. This method reduces the subjectivity of traditional weight allocation, resulting in more objective outcomes [40]. In the study, based on the Water Quality Standards for Surface Water (GB 3838-2002), seven water quality indicators, including ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N}$, $\mathrm{P}\mathrm{b}$, $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$, $\mathrm{C}\mathrm{d}$, $\mathrm{A}\mathrm{s}$, ${\mathrm{F}}^{-}$, and $\mathrm{C}\mathrm{u}$, were selected for water quality assessment. The calculation steps are as follows:

Assuming there are $m (i=\mathrm{1,2},\cdots ,m)$ samples, each with $n (j=\mathrm{1,2},\cdots ,n)$ water quality indicators, an initial matrix $X_{(m\times n)}$ is established. The initial matrix $X_{(m\times n)}$ is then standardized to eliminate the dimensionality effect, resulting in the standardized evaluation matrix ${Y=\left(y_{ij}\right)}_{(m\times n)}$.

Next, the information entropy $e_j$ and entropy weight $w_j$ of each indicator are calculated using the information entropy method. To compute the ratio $P_{ij}$ for the ith indicator, a correction parameter of ${10}^{-6}$ is introduced. The specific formula is as follows:

$$P_{ij}={\displaystyle \frac{y_{ij}+{10}^{-6}}{\sum _{i=1}^m{(y}_{ij}+{10}^{-6})}}$$

(1)

$$e_j=-{\displaystyle \frac{1}{lnm}}{\sum }_{i=1}^m{P}_{ij}ln{P}_{ij}$$

(2)

$$w_j={\displaystyle \frac{1-e_j}{{\sum }_{j=1}^n(1-e_j)}}$$

(3)

The EWQI value for each sample is calculated using the concentration ratio $q_j$ of water quality indicator j in each sample, where $C_j$ is the measured concentration of the sample, and $S_j$ is the concentration value listed in the three water quality standards categories specified in the “Environmental quality standards for surface water” (GB 3838-2002). The specific calculation formula is as follows:

$$q_j={\displaystyle \frac{C_j}{S_j}}\times 100$$

(4)

$$\mathrm{E}\mathrm{W}\mathrm{Q}\mathrm{I}={\sum }_{j=1}^n{w}_j{q}_j$$

(5)

Based on the EWQI values, water quality can be categorized into five levels, where Grade I: Excellent (EWQI < 25), Grade II: Good (25 ≤ EWQI < 50), Grade III: Fair (50 ≤ EWQI < 100), Grade IV: Poor (100 ≤ EWQI < 150), and Grade V: Very Poor (EWQI ≥ 150).

2.3.2. Human Health Risk Assessment

The USEPA-recommended health risk assessment model was selected to assess human health risks for both adults and children [41,42,43]. Heavy metal in surface water pose health risks through two main pathways: oral ingestion and dermal contact. The model includes both carcinogenic and non-carcinogenic risk assessment models. The expression for the oral ingestion pathway risk assessment model is as follows:

$$D_i=C_i\times A\times EF\times ED/(B\times AT)$$

(6)

$$R^c=\sum R_i^c=\sum D_i\times q_i$$

(7)

$$R^n=\sum D_i/{RfD}_i$$

(8)

where $D_i$ is the unit weight daily exposure dose of metal element i (mg/(kg·d)); $A$ is the average daily water intake per person, 2.2 L/d for adults and 1.0 L/d for children; $C_i$ is the measured concentration of heavy metal i (mg/L); $EF$ is the exposure frequency ($\mathrm{d}·{\mathrm{a}}^{-1}$), 350 for carcinogens and 350 for non-carcinogens; $ED$ is the exposure duration (a), 70 for carcinogens and 35 for non-carcinogens; B is the average weight per resident, 56 kg for adults and 22 kg for children; $AT$ is the average exposure time (d), 70 × 365 for carcinogens and 35 × 365 for non-carcinogens; $R_i^c$ is the individual annual average health risk value for carcinogenic metal element i via oral ingestion; $q_i$ is the carcinogenic coefficient for metal element i via oral ingestion ($(\mathrm{k}\mathrm{g}·\mathrm{d})/\mathrm{m}\mathrm{g}$); $R_i^n$ is the individual annual average health risk value for non-carcinogenic metal element i via oral ingestion; and ${RfD}_i$ is the reference dose for non-carcinogenic metal element i via oral ingestion ($\mathrm{m}\mathrm{g}/(\mathrm{k}\mathrm{g}·\mathrm{d})$).

The expression for the dermal contact pathway risk assessment model is as follows:

$$D_d=I_d\times A_{sd}\times FE\times EF\times ED/(B\times AT)$$

(9)

$$I_d=k\times c_d\times TE$$

(10)

$$R^p=\sum R_d^p=\sum D_d\times q_d$$

(11)

$$R^f=\sum R_d^f=\sum D_d/{RfD}_d$$

(12)

where $D_d$ is the unit weight daily exposure dose of metal element d ($\mathrm{m}\mathrm{g}/(\mathrm{k}\mathrm{g}·\mathrm{d})$); $I_d$ is the adsorption amount of pollutant d per unit area for each bathing session (mg/(cm²·session)); $A_{sd}$ is the body surface area (cm²), 16,000 cm² for adults and 6660 cm² for children; $FE$ is the bathing frequency (session$·{\mathrm{d}}^{-1}$), with a value of 0.3; $k$ is the skin adsorption parameter ($\mathrm{c}\mathrm{m}·{\mathrm{h}}^{-1}$), with a value of 0.001; $c_d$ is the mass concentration of heavy metal d (mg/L); $TE$ is the bathing time (h), with a value of 0.56; $R_d^p$ is the individual annual average health risk value for carcinogenic metal element d via dermal contact; $R_d^f$ is the individual annual average health risk value for non-carcinogenic metal element d via dermal contact; $q_d$ is the carcinogenic coefficient for metal element d via dermal contact ($(\mathrm{k}\mathrm{g}·\mathrm{d})$/mg); and ${RfD}_d$ is the reference dose for non-carcinogenic metal element d via dermal contact (mg/$(\mathrm{k}\mathrm{g}·\mathrm{d})$). The values of the above parameters refer to r relevant technical guidelines and guide-lines for risk assessment [44,45,46,47].

The overall formula for calculating the total health risk $R_T$ for the water environment is as follows:

$${R_T=R^c+R^n+R}^p+R^f$$

(13)

The United States Environmental Protection Agency (USEPA) categorizes carcinogenic and non-carcinogenic risks in health risk assessments into different levels, as shown in

Table 1. Risk levels in USEPA health risk assessment.

Risk Type	Risk Level	Description
Carcinogenic Risk	High Risk	${10}^{-3}<R^p$
	Moderate Risk	${10}^{-4}<R^p<{10}^{-3}$
	Low Risk	${10}^{-6}<R^p<{10}^{-4}$
	Negligible Risk	$R^p<{10}^{-6}$
Non-Carcinogenic Risk	High Risk	$100<R^f$
	Moderate Risk	$10<R^f<100$
	Low Risk	$1<R^f<10$
	No Significant Risk	$R^f<1$

.

2.3.3. Principal Component Analysis (PCA)

The PCA is a commonly applied multivariate method which finds the relationship between different variables in a large dataset. In this study, the PCA tool in the SPSS v26.0 software was employed [48,49]. A total of 7 indicators (${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N}$, $\mathrm{P}\mathrm{b}$, $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$, $\mathrm{C}\mathrm{d}$, $\mathrm{A}\mathrm{s}$, ${\mathrm{F}}^{-}$, and $\mathrm{C}\mathrm{u}$) were considered for the calculation of the PCA for identifying pollution sources. The maximum variance method was used for factor rotation, and the Kaiser–Meyer–Olkin (KMO) measure and Bartlett’s test of sphericity were utilized to assess the validity of the PCA results. The PCA results were considered valid when the KMO test value exceeded 0.5, and the significance probability (P) in Bartlett’s test was less than 0.05. The factor loadings were classified into strong (>0.75), moderate (0.5~0.75), and weak (0.3~0.5) corresponding to the absolute loading values. Depending upon the eigen-values which are more than 1 is considered as the principal components.

3. Results and Discussion

3.1. Hydrochemical Characteristics

To clearly assess the seasonal variation of the surface water in the study area, mathematical–statistical methods were employed to explore various physico-chemical parameters (

Table 2. Summary of physico-chemical parameters in the study area.

Parameters	The Dry Season						The Wet Season
	Min	Max	Mean	STD	CV (%)	ES (%)	Min	Max	Mean	STD	CV (%)	ES (%)
${\mathrm{N}\mathrm{a}}^{+}$ (mg/L)	3.22	527	118.66	111.5	93.97		4.08	956	163.6	212	129.59
${\mathrm{K}}^{+} (\mathrm{m}\mathrm{g}/\mathrm{L})$	1.6	20.6	6.88	4.14	60.14		2.04	23.3	8.41	5.08	60.42
${\mathrm{C}\mathrm{a}}^{2+} (\mathrm{m}\mathrm{g}/\mathrm{L})$	26.7	139	70.89	24.62	34.73		17.1	382	100.47	66.6	66.29
${\mathrm{M}\mathrm{g}}^{2+} (\mathrm{m}\mathrm{g}/\mathrm{L})$	3.87	77.9	33.79	19.18	56.75		3.84	62.4	31.21	17.44	55.89
${\mathrm{C}\mathrm{l}}^{-}$ (mg/L)	3.51	366	137.34	103.3	75.22		2.07	1776	139.26	275.86	198.09
${\mathrm{S}\mathrm{O}}_4^{2-}$ (mg/L)	40.3	973	251.96	181.2	71.92		32.8	1968	238.2	338.87	142.26
${\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}$ (mg/L)	28.43	525.09	164.91	107.8	65.37		11.1	108	45.45	19.67	43.29
${\mathrm{C}\mathrm{O}}_3^{2-}$ (mg/L)	0	16.43	5.53	5	90.35		0	1.52	0.08	0.34	441.44
${\mathrm{N}\mathrm{O}}_3\text{-}\mathrm{N}$ (mg/L)	0	0.3	0.05	0.06	119.24		0	1.43	0.17	0.28	166.48
${\mathrm{F}}^{-}$ (mg/L)	0.27	2.26	0.72	0.36	49.49	15	0.15	2.23	0.63	0.36	57.51	5
${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N} (\mathrm{m}\mathrm{g}/\mathrm{L})$	0.16	0.92	0.34	0.21	62.11		0.04	0.72	0.27	0.22	81.54	7.5
Mn (µg/L)	0.52	773	123.23	145.81	115.79		31	800	241.12		91.48
Ni (µg/L)	1.17	12.5	5.47	2.7	49.81		2.21	18.3	7.83		56.74
$\mathrm{C}\mathrm{u}$ (µg/L)	1.58	19.3	5.9	3.64	61.63		8.17	54.5	15.94		55.17
$\mathrm{C}\mathrm{d}$ (µg/L)	<0.2	0.33	/	0.05	15.81		0.2	0.44	0.31		49.59
$\mathrm{P}\mathrm{b}$ (µg/L)	0.38	25.1	2.74	3.93	143.64	2.5	0.38	29.1	3.41		156.44	5
Hg (µg/L)	<0.16	<0.16	/	/	/		<0.16	<0.16	/	/	/
As (µg/L)	<1.20	16.92	/	3.82	86.8		1.13	81.1	18.53		110.02	10
Ag (µg/L)	<0.16	<0.16	/	/	/		<0.16	0.9	/	/	15.81
Fe (mg/L)	<0.07	0.85	/	0.2	73.15		<0.07	1.49	/	/	95.52
$\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ (mg/L)	<0.004	0.03	/	0.01	65.28		0.14	54.8	19.4		60.95	95

Note(s): Min: minimum value; Mean: average value; Max: maximum value; STD: standard deviation; CV: the coefficient of variation; ES: exceeding standards.

). During the dry season, the coefficient of variation for ions such as ${\mathrm{N}\mathrm{O}}_3\text{-}\mathrm{N}$, Mn, and Pb ranged from 115% to 145%. In the wet season, the coefficient of variation for ions including ${\mathrm{N}\mathrm{a}}^{+},{\mathrm{C}\mathrm{l}}^{-}$, ${\mathrm{S}\mathrm{O}}_4^{2-}$, ${\mathrm{C}\mathrm{O}}_3^{2-}$, ${\mathrm{N}\mathrm{O}}_3\text{-}\mathrm{N}$, Pb, and $\mathrm{A}\mathrm{s}$ ranged from 110% to 442%. These indicate a relatively spatial variability for these ions, with high local enrichment levels, suggesting significant influence from human activities. The high-value samples of ${\mathrm{N}\mathrm{O}}_3\text{-}\mathrm{N}$ during both wet and dry seasons were located in the midstream region of the Changjianhe River, near Huashan City. The high-value samples of Pb were concentrated in the midstream region of the Luofuhe River during the dry season, near the Huashan Hot Spring Resort. In the wet season, the high-value samples of Pb were concentrated in the Jingouhe basin, and the high-value samples of ${\mathrm{N}\mathrm{a}}^{+}$, ${\mathrm{C}\mathrm{l}}^{-}$, ${\mathrm{S}\mathrm{O}}_4^{2-}$, and ${\mathrm{C}\mathrm{O}}_3^{2-}$ were concentrated in the midstream region of the Luofuhe River and the downstream region of the Liuyehe River. Relatively high-density dwellings are distributed across this river basin.

Based on the concentration values listed in the Class III water quality standards of the “Environmental quality standards for surface water” (GB 3838-2002) as the criteria, the degree of exceedance of water indicators during both wet and dry seasons was analyzed (Table 2). The results reveal that 15% and 2.5% of the samples, respectively, exceeded the allowable limit of Pb and ${\mathrm{F}}^{-}$ in the dry season. In the wet season, the exceeded ions were As, ${\mathrm{F}}^{-}$, ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{H}$, and $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$, with the degree of exceedance as follows: $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$> ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{H}$ = $\mathrm{A}\mathrm{s}$ > ${\mathrm{F}}^{-}$, and other ions. Among them, the $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ contents in 95% of the samples surpass the permissible limit (0.05 mg/L). This significant exceedance of Cr(VI) and other heavy metals in the wet season may be attributed to illegal mining activities in the Qinling Mountains and improper disposal of mining tailings. The heavy precipitation during the wet season likely facilitates the leaching of these pollutants from soil and river sediments into surface water. Future research will focus on a detailed investigation to accurately trace the specific sources of these pollutants.

During both the wet and dry seasons, the cationic dominance in the surface water can be described in the following order: ${\mathrm{N}\mathrm{a}}^{+}>{\mathrm{C}\mathrm{a}}^{2+}>{\mathrm{M}\mathrm{g}}^{2+}>{\mathrm{K}}^{+}$, with higher cation concentrations in the wet season compared to the dry season. The main anion concentration relationships in surface water during the wet season can be described in the following order: ${\mathrm{S}\mathrm{O}}_4^{2-}>{\mathrm{C}\mathrm{l}}^{-}>{\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}>{\mathrm{C}\mathrm{O}}_3^{2-}$, while during the dry season, the main anion concentration relationships are ${\mathrm{S}\mathrm{O}}_4^{2-}>{\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}>{\mathrm{C}\mathrm{l}}^{-}>{\mathrm{C}\mathrm{O}}_3^{2-}$. In the wet season, the concentrations of ${\mathrm{S}\mathrm{O}}_4^{2-}$ and ${\mathrm{C}\mathrm{l}}^{-}$ are higher than in the dry season, whereas ${\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}$ and ${\mathrm{C}\mathrm{O}}_3^{2-}$ are higher in the dry season compared to the wet season.

The Piper trilinear diagram is an important tool for the analysis of hydrochemical types. Figure 2a shows the Piper trilinear diagram for the wet season, data samples concentrated in zones 1 (28 samples) and 2 (12 samples), indicating that the hydrochemical types of surface water during the wet season are mainly of the ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$ and ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{N}\mathrm{a}$. For the Piper trilinear diagram during the dry season (Figure 2b), most of the samples are mostly concentrated in zone 1, indicating that the hydrochemical type of surface water during the dry season is primarily of the ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$. Additionally, 11 samples and 2 samples are located in zone 2 and zone 4, representing the ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{N}\mathrm{a}$ and ${\mathrm{H}\mathrm{C}\mathrm{O}}_3\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$ types, respectively. The overall hydrochemical type (${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$) in the study area during the both wet and dry seasons is related to the hydrolysis of minerals containing calcium (Ca) and magnesium (Mg). In comparison, between wet and dry seasons, during the wet season, the hydrochemical type of some samples is ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{N}\mathrm{a}$, which is more related to the dissolution of minerals containing sodium (Na), while the hydrochemical type of some samples (${\mathrm{H}\mathrm{C}\mathrm{O}}_3\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$) during the dry season may be related to the dissolution of carbonate minerals.

Analyzing the different basins, both the Beiluohe River and the Weihe River undergo a mutual transformation between the ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$ type (wet season) and the ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{N}\mathrm{a}$ type (dry season). In the Beiluohe River, ${\mathrm{S}\mathrm{O}}_4^{2-}$ is the dominant anion during the wet season, while ${\mathrm{N}\mathrm{a}}^{+}$ is the dominant cation during the dry season. For both seasons in the Wei River, there is no obvious dominant cation, and ${\mathrm{S}\mathrm{O}}_4^{2-}$ is the dominant anion. The hydrochemical type of the Huanghe River is ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$, with ${\mathrm{C}\mathrm{a}}^{2+}$ and ${\mathrm{S}\mathrm{O}}_4^{2-}$ as the dominant cation and anion during the wet season, respectively. In the dry season, there is no obvious dominance of any ion. The hydrochemical type of the river wetland is mainly of the ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$ type, with abundant ${\mathrm{N}\mathrm{a}}^{+}$ and ${\mathrm{S}\mathrm{O}}_4^{2-}$ during the wet season, while no ion has a clear dominance during the dry season.

For small basins, the hydrochemical type of the Jingouhe River is ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$ during both wet and dry seasons, with ${\mathrm{C}\mathrm{a}}^{2+}$ as the dominant cation and ${\mathrm{S}\mathrm{O}}_4^{2-}$ as the dominant anion. In the Changjianhe River, the hydrochemical type is ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$. The hydrochemical type varies significantly from upstream to downstream in the Liuyehe River, with the concentration of ${\mathrm{N}\mathrm{a}}^{+}$ and ${\mathrm{S}\mathrm{O}}_4^{2-}$ gradually increasing. The hydrochemical types of the Luofuhe River are ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$ and ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{N}\mathrm{a}$, with higher ${\mathrm{S}\mathrm{O}}_4^{2-}$ concentrations among the anions. Compared with the hydrochemical type in small basins, the hydrochemical type of the Jinguhe River shows no significant difference between wet and dry seasons, which means that the basin has no strong traces of human activities. For the reservoir, the Xiaopuyu Reservoir is of the ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$ type, with ${\mathrm{C}\mathrm{a}}^{2+}$ and ${\mathrm{S}\mathrm{O}}_4^{2-}$ as the dominant cation and anion, respectively. During the wet season, the Tiegou Reservoir is of the ${\mathrm{H}\mathrm{C}\mathrm{O}}_3\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$ type, with ${\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}$ as the dominant anion.

3.2. Evaluation of Surface Water Quality and Health Risk Assessment

The entropy weight water quality index (EWQI) was employed to calculate the water quality data for the sampling points during both wet and dry seasons in the study area. Seven water quality indicators, including ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N}$, $\mathrm{P}\mathrm{b}$, $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$, $\mathrm{C}\mathrm{d}$, $\mathrm{A}\mathrm{s}$, ${\mathrm{F}}^{-}$, and $\mathrm{C}\mathrm{u}$ were considered. The entropy weights of each water quality indicator are presented in Figure 3, and the water quality grades for each sampling point are shown in

Table 3. Water quality distribution in the study area.

The Location of the Sample		Name	The Dry Season	The Wet Season	The Location of the Sample		Name	The Dry Season	The Wet Season
Huanghe	upstream	W-35	I	V	Liuyehe	upstream	W-12	I	V
	midstream	W-36	I	V		midstream	W-13	I	V
	downstream	W-37	I	II		downstream	W-14	I	V
Beiluohe	upstream	W-40	I	V	Luofuhe	upstream	W-15	I	V
	midstream	W-39	I	V		midstream	W-16	I	V
	downstream	W-38	I	V		downstream	W-17	I	V
Weihe	upstream	W-18	I	V	Jingouhe	upstream	W-01	I	II
		W-19	I	V			W-05	I	V
	midstream	W-20	I	V		midstream	W-02	I	V
	downstream	W-21	I	V		downstream	W-03	I	V
		W-22	I	V	Changjianhe	upstream	W-07	I	V
The southern part of the river wetland		W-23	II	V			W-09	I	V
		W-24	I	V		midstream	W-08	I	V
		W-25	I	V			W-10	I	V
		W-26	I	V		downstream	W-11	I	V
		W-27	II	V	The northern part of the river wetland		W-31	I	V
		W-28	II	V			W-32	II	V
		W-29	I	V			W-33	I	V
		W-30	II	V			W-34	I	V
Xiaopuyu Reservoir		W-06	I	V	Tiegou Reservoir		W-04	I	V

.

During the dry season, $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ had the highest entropy value, followed by Cu. Parameters with higher entropy values typically have a more significant impact on water quality, while other ion weights are relatively smaller. This reflects the enrichment of $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ and Cu during the dry season in the study area. According to the water quality grades determined by EWQI, the overall surface water quality reached a good level during the dry season and is suitable for direct consumption. Grade I water quality constituted 90% of the samples, with Grade II comprising 10%. In contrast, during the wet season, ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N}$ had a higher entropy value, indicating a more substantial impact on water quality. The water quality grades revealed that, except for the downstream region of the Yellow River and the upstream region of the Jingou River, classified as Grade II and Grade III, respectively, the water quality of other samples during the wet season was Grade V, constituting 95%. This indicates poor water quality, falling into a completely non-drinkable category. It can be observed that there are significant differences in surface water quality between the wet and dry seasons in the study area, with relatively small variations between different river basins. This phenomenon can be attributed to the washing down of pollutants by the increased river flow during the wet season. Specifically, in the Qinling Mountains area, tailings and waste piles accumulated along the riverbanks are likely to be eroded by the higher water levels during the wet season. This erosion results in chromium and other heavy metal pollutants being washed into the rivers, leading to poorer water quality during the wet season.

Furthermore, through the analysis of ion concentrations in the water samples, it was found that the levels of heavy metals in the area are relatively high. Therefore, the health risk assessment model recommended by the US-EPA was employed to assess the health risks associated with potential toxicological indicators with high detection rates in the local surface water. The evaluated indicators include carcinogenic substances such as $\mathrm{C}\mathrm{d}$, As, $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$, and non-carcinogenic substances such as Ni, $\mathrm{C}\mathrm{u}$, $\mathrm{P}\mathrm{b}$. The personal annual average health risks and total health risks caused by oral intake and skin contact pathways were calculated based on the health risk assessment model and the measured concentrations of each heavy metal element. The detailed results are presented in

Table 4. The results of the health risk assessment for adults and children in the study area (${\mathrm{a}}^{-1}$).

Risk Pathway		The Wet Season		The Dry Season
		Adults	Children	Adults	Children
Carcinogenic risk	The oral ingestion	$120.9\times {10}^{-4}$	$121.3\times {10}^{-4}$	$1.37\times {10}^{-4}$	$1.58\times {10}^{-4}$
	The dermal contact	$122.1\times {10}^{-4}$	$122.3\times {10}^{-4}$	$2.15\times {10}^{-4}$	$2.28\times {10}^{-4}$
Non-carcinogenic risk	The oral ingestion	$2.89\times {10}^{-9}$	$3.34\times {10}^{-9}$	$1.57\times {10}^{-9}$	$1.82\times {10}^{-9}$
	The dermal contact	$4.56\times {10}^{-9}$	$4.83\times {10}^{-9}$	$2.48\times {10}^{-9}$	$2.63\times {10}^{-9}$
$R_T$		$243.1\times {10}^{-4}$	$243.6\times {10}^{-4}$	$3.52\times {10}^{-4}$	$3.84\times {10}^{-4}$

.

The human health risk assessment indicates that the total health risks for adults in the study area range from $3.52\times {10}^{-4}~243.1\times {10}^{-4}{\mathrm{a}}^{-1}$, while, for children, the total health risks range from $3.84\times {10}^{-4}~243.6\times {10}^{-4}{\mathrm{a}}^{-1}$. The magnitude of risk is consistent between adults and children, but the carcinogenic and non-carcinogenic risks for children are slightly higher than those for adults. The carcinogenic risk in the study area is in the range of ${10}^{-4}~{10}^{-2}$. The health risk rankings for carcinogenic pollutants are $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ > As > $\mathrm{C}\mathrm{d}$, with $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ posing a significantly higher carcinogenic risk than other heavy metal pollutants, making it the primary carcinogenic factor. The non-carcinogenic risk is in the range of ${10}^{-9}$, and the health risk rankings for non-carcinogenic pollutants are $\mathrm{C}\mathrm{u}$ > Ni > $\mathrm{P}\mathrm{b}$. The health risks caused by carcinogenic heavy metal elements are much higher than those caused by non-carcinogenic elements. The carcinogenic risk values are significantly greater than $1\times {10}^{-3}$, indicating a high health risk that requires attention. On the other hand, the non-carcinogenic risks from non-carcinogenic elements are less than 1, indicating no significant health risk.

The total health risk values caused by heavy metals in surface water are relatively high, with the health risks from the dermal exposure pathway being greater than those from oral ingestion. In comparison, between the wet and dry seasons, the total health risk during the wet season is much higher than during the dry season. In summary, whether for adults or children, $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ is the major heavy metal element causing the highest health risk in the study area. Therefore, $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ should be a priority for effective control in the utilization of surface water resources.

3.3. The Analysis of the Pollution Source

(1)

Correlation Analysis

Correlation analysis has a basic distinguishing effect on exploring the sources of different elements and can simply identify the correlation between different elements. In this paper, the correlation between the elements was analyzed using correlation heatmaps (Figure 4).

During the dry season, a significant positive correlation (0.94) was observed between $\mathrm{P}\mathrm{b}$ and $\mathrm{C}\mathrm{d}$, indicating that the pollution of $\mathrm{P}\mathrm{b}$ and $\mathrm{C}\mathrm{d}$ during the dry season may have a certain homogeneity. $\mathrm{C}\mathrm{u}$ showed a weak positive correlation (0.34) with $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$, while the correlation coefficients between the other parameters were less than 0.3. The results show that the pollution sources may be different during the dry season.

During the wet season, a significant positive correlation (0.65) was observed between ${\mathrm{F}}^{-}$ and $\mathrm{A}\mathrm{s}$. Weak positive correlations (0.2~0.4) were observed between $\mathrm{P}\mathrm{b}$ and $\mathrm{C}\mathrm{d}$, between $\mathrm{A}\mathrm{s}$ and $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$, and between $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ and ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{H}$. The results show that ${\mathrm{F}}^{-}$ and $\mathrm{A}\mathrm{s}$ are more likely to come from the same polltion source. In addition, there is no correlation between other parameters during the wet season, and there may be independent pollution sources.

(2)

The result of PCA

The correlation analysis indicated weak correlations among various elements, suggesting a complex origin of the river’s elements in the study area. Therefore, the PCA was employed to analyze the possible sources of seven elements in the study area.

During the dry season, the KMO test value was 0.624 (>0.5), and the Bartlett sphericity test statistic was 0, indicating that the data met the test conditions of the PCA. Three principal components (DPC1, DPC2, DPC3) were extracted via the principal component analysis and could explain 89.454% of the total variance.

During the wet season, the KMO test value was 0.574 (>0.5), the Bartlett sphericity test statistic was 0, and the significance test value of the Bartlett sphericity test was p < 0.05, thus indicating that the data met the test conditions of the PCA. Three principal components were extracted via the principal component analysis, which were WPC1, WPC2, and WPC3, and could explain 70.128% of the total variance. The principal component loads contained in each component are shown in

Table 5. The main calculation results of principal component analysis.

Parameters	The Dry Season			The Wet Season
	DPC1	DPC2	DPC3	WPC1	WPC2	WPC3
${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N}$	−0.083	0.032	−0.707	0.160	0.012	0.739
${\mathrm{F}}^{-}$	0.062	−0.630	0.413	0.843	0.211	0.197
$\mathrm{C}\mathrm{u}$	−0.032	0.473	0.634	−0.632	0.191	0.518
$\mathrm{C}\mathrm{d}$	0.984	−0.043	−0.018	0.037	0.879	0.078
$\mathrm{P}\mathrm{b}$	0.970	0.106	0.122	−0.028	0.782	−0.141
$\mathrm{A}\mathrm{s}$	−0.193	−0.605	−0.060	0.816	−0.153	0.299
$\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$	−0.100	0.713	0.255	0.230	−0.459	0.619

.

DPC1, explaining 28.782% of the total variance, relies mainly on Cd (0.984) and Pb (0.970). Combining the spatial distribution characteristics of elements, the high-value samples of Cd and Pb are found in the midstream region of the Luofuhe River, near the Huashan Royal Hot Spring Resort (Figure 5a). Zong et al. pointed out that Huashan has abundant hot spring resources, and Cd and Pb are enriched in the hot springs [50,51]. The pollution source of the DPC1 may result from the combined effects of underground water–rock interactions and human activities leading to the discharge of hot springs.

DPC2, accounting for 22.924% of the total variance, exhibits moderate positive loadings on Cr(VI) (0.713), and moderate negative loadings on ${\mathrm{F}}^{-}$ and As. The high-value samples of $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ are found in the river near the Dali area, which may be affected by anthropogenic industrial sources. The high-value samples of ${\mathrm{F}}^{-}$ and $\mathrm{A}\mathrm{s}$ are distributed in the river wetland of Dali area, where ${\mathrm{F}}^{-}$ exceeds the concentration listed in Class III water quality standards during the dry season. Previous studies indicated the groundwater of the Dali area was rich in element ${\mathrm{F}}^{-}$, and As-heavy metals were also abundant in the farmland soil [52]. Agricultural activities involving water pumping and reservoir storage have led to the accumulation of the elements ${\mathrm{F}}^{-}$ and As in the pond. Therefore, the pollution source of the DPC2 may be influenced by both agricultural and industrial activities in the Dali area.

DPC3, representing 14.505% of the total variance, and significant loadings were observed on ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N}$ and $\mathrm{C}\mathrm{u}$. The high-value samples of ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N}$ are mainly distributed in the river wetland, where there is animal manure and feed residue from grazing activities. The high-value samples of Cu are located in the middle and lower reaches of the Liuyehe River. The result may be that the tailings generated by the smelting of chalcopyrite in factories by the river cause pollutants ($\mathrm{C}\mathrm{u}$) to enter the soil and surface water due to precipitation. Therefore, DPC3 may be associated with grazing activities in the river wetland and industrial activities in the Liuyehe basin.

WPC1, explaining 30.964% of the total variance, exhibits substantial strong positive loadings on ${\mathrm{F}}^{-}$ (0.843) and $\mathrm{A}\mathrm{s}$ (0.816), and medium negative loadings on $\mathrm{C}\mathrm{u}$ (−0.632). Combined with the correlation analysis of the elements, ${\mathrm{F}}^{-}$ and $\mathrm{A}\mathrm{s}$ have a strong correlation during the wet season (0.65), with high-value samples distributed in the river wetland of the Dali area (Figure 5e). The high-value samples of $\mathrm{C}\mathrm{u}$ are distributed in the middle and lower reaches of the Liuyehe River. The distribution of high-value samples of ${\mathrm{F}}^{-}$, $\mathrm{A}\mathrm{s}$, and $\mathrm{C}\mathrm{u}$ is consistent with the dry season. Therefore, WPC1 may be influenced by agricultural activities in the Dali area and industrial activities in the Liuyehe basin.

WPC2 accounted for 22.857% of the total variance and exhibits significant strong positive loadings on Cd and Pb. The high-value samples of Cd and Pb are situated in the upstream region of the basin (Jingouhe, Liuyehe, Changjianhe) (Figure 5b,d). Leaching of minerals is a common natural process because of the water–rock interactions, and this easily leads to elevated heavy metal content in the upstream water environment. Therefore, WPC2 can be considered as a composite effect of native geological environment and water–rock interactions.

WPC3, accounting for 16.307% of the total variance, exhibits moderate positive loadings on NH₃-N (0.739), Cr(VI) (0.619), and Cu (0.518). The high-value samples of elements are distributed in the river wetland and the Huanghe River (Figure 5e), and the content is influenced by local agricultural activities. The pollution source of the WPC3 may result from agricultural activities.

In summary, the variation of physicochemical parameters in surface water is influenced by a combination of factors: local soil characteristics (such as soils containing fluoride in the Dali region and soils containing heavy metals in the Tongguan region), the native geological environment of the Qinling Mountains (including mineral resources and terrain conditions), and industrial activities (particularly ore smelting).

4. Conclusions

This study analyzed the surface water quality at the confluence of the Wei River with the Yellow River, including hydrochemical characteristics, water quality assessment, human health risk assessment, and potential pollution sources identification. The main conclusions are as follows:

During the dry season, surface water quality is mainly influenced by ${\mathrm{F}}^{-}$, particularly in the river wetland of the Dali area, where the ${\mathrm{F}}^{-}$ concentrations exceed Class III standards. The EWQI results indicate generally good water quality suitable for direct consumption. During the wet season, $\mathrm{C}\mathrm{r}\left(\mathrm{VI}\right)$ and ${\mathrm{N}\mathrm{H}}_3\text{-}\mathrm{N}$ were the main pollutants. The EWQI results reveal extremely poor water quality during this season.

The hydrochemical type in the study area was mainly ${\mathrm{S}\mathrm{O}}_4\mathrm{C}\mathrm{l}\text{-}\mathrm{C}\mathrm{a}\mathrm{M}\mathrm{g}$, associated with the hydrolysis of minerals containing $\mathrm{C}\mathrm{a}$ and $\mathrm{M}\mathrm{g}$. The health risk assessment indicates a high carcinogenic risk from Cr(VI), exceeding the acceptable threshold, while non-carcinogenic risks remain lower.

Furthermore, the principal component analysis (PCA) and the correlation analysis suggest that local soil characteristics (such as fluoride in Dali and heavy metals in Tongguan), the geological environment of the Qinling Mountains, and industrial activities (such as ore smelting) are the main drivers of variations in surface water physicochemical parameters.

The findings of this study underscore the critical need for effective management and mitigation strategies to address the identified pollutants, particularly Cr(VI). The seasonal variations in water quality highlight the dynamic nature of pollutant sources and the need for continuous monitoring.

Based on the findings of this study, future research should conduct detailed investigations in the study area to trace the specific sources of pollutants, particularly from industrial and agricultural activities. This will enable the development of more effective source control strategies. Additionally, advanced remediation techniques should be explored and developed to reduce the concentrations of hexavalent chromium and other heavy metals in surface water. These efforts will provide relevant guidelines for protecting surface water resources and public health in the study area and beyond.