ABSTRACT
Groundwater is a vital water source for human consumption and irrigation. Understanding its fluoride content and health implications is crucial for water resource management. This study investigated the quaternary aquifer in Suzhou, China, collecting and analyzing 49 groundwater samples. Thermodynamic simulation, multivariate statistical analysis, and health risk assessment models were employed to determine fluoride concentration characteristics, hydrochemical controlling mechanisms, and noncarcinogenic risks. Results revealed an average fluoride concentration of 0.89 mg/L, with 26.5% of samples exceeding the Grade III groundwater quality standard. High-fluoride groundwater (>1 mg/L) exhibited spatial heterogeneity and was predominantly of the Na-Mg-HCO3 hydrochemical type. Multivariate analysis and thermodynamics simulations indicated that water–rock interactions (e.g., silicate mineral weathering and fluorite dissolution) governed groundwater hydrochemistry. Fluoride primarily originated from fluoride-bearing mineral dissolution, while negative cation exchange and precipitation of calcite and dolomite enhanced fluoride enrichment. pH had minimal impact on fluoride concentrations under weakly alkaline conditions. Health risk assessment suggested that fluoride in shallow groundwater posed a higher noncarcinogenic risk to children than adults via ingestion. These findings provide valuable insights for regional groundwater resource management.
HIGHLIGHTS
High-fluoride groundwater was predominantly of the Na-Mg-HCO3 hydrochemical type.
Negative cation exchange and precipitation of calcite and dolomite enhanced fluoride enrichment.
Fluoride in shallow groundwater posed a higher noncarcinogenic risk to children than adults.
INTRODUCTION
Fluoride is a widely distributed element in nature and an essential trace element for the human body (Zuo et al. 2018). However, excessive fluoride in the environment poses significant health risks to animals, plants, and humans (Borgnino et al. 2013; Dehbandi et al. 2018). Studies have established a strong positive correlation between the incidence of dental fluorosis, fluorosis, and carotid atherosclerosis in adults and fluoride concentration in drinking water (Liu et al. 2014, 2019). As a vital component of terrestrial water resources, groundwater provides the daily water supply for approximately one-third of the global population. Its water quality directly affects various aspects, including production, livelihood, and ecological water supply (Mandal et al. 2022; Nikolenko et al. 2024). Therefore, elucidating the distribution characteristics of fluoride in groundwater and its associated health risks is crucial for ensuring regional water supply safety and effective water resource management.
Multivariate statistical analyses and geostatistical methods are commonly used to characterize the spatial distribution and potential sources of fluorine (Abba et al. 2023; Yassin et al. 2024). Numerous studies by domestic and international scholars have shown that fluoride ions in groundwater are mainly related to factors such as the weathering of fluoride-containing minerals and rocks, long-term hydrogeochemical evolution, industrial and agricultural activities, and overexploitation of groundwater (Zheng et al. 2020; Su et al. 2021). Brindha et al. (2016) found that in addition to the weathering of fluoride-rich minerals in rocks, evaporation and precipitation are the main contributors of high fluoride in weathered rock aquifers in southern India, and fluoride concentration increases with the increase of the groundwater level. A study by Emenike et al. (2018) on fluoride in tap water in southern Nigeria showed that approximately 91% of infants in the study area were at noncarcinogenic risks of fluoride through dermal contact and direct ingestion pathways. In China, research on high-fluoride groundwater has mainly focused on arid and semi-arid climatic zones and plain and basin areas. For example, Wu et al. (2018) suggested that the primary source of high fluoride concentration in groundwater in the Yanchi endorheic basin in northwestern China is the dissolution of fluorite and the desorption of exchangeable F−.
According to the China Health Statistics Yearbook, over 12.4 million individuals in China suffered from dental fluorosis in 2019. Notably, Anhui Province exhibited a significantly higher prevalence of dental fluorosis (5.62%) compared with the national average (0.98%). Endemic fluorosis is particularly prevalent in the northern Anhui Plain. Previous studies have indicated that fluoride concentrations in groundwater within this region can reach 3.6 mg/L, posing a potential threat to public health (Gao et al. 2013; Hu et al. 2021). Suzhou City, situated in the northernmost portion of the northern Anhui Plain, relies heavily on groundwater as its primary water source. Although prior researches have examined the hydraulic connection between shallow and deep groundwater and conducted basic water quality assessments (Chen et al. 2022; Ma et al. 2022), the formation mechanisms and human health risks associated with high-fluoride groundwater in the study area remain poorly understood. This study aims to address these knowledge gaps by investigating the following: (1) the concentration and spatial distribution of fluoride in groundwater; (2) the source and control mechanisms of fluoride; and (3) the noncarcinogenic risks of fluoride posed to children and adults. The findings of this study are anticipated to provide valuable insights into the management of shallow groundwater in the study area and other similar plain regions.
MATERIALS AND METHODS
Study area
Suzhou city is situated in the northernmost part of Anhui Province, China, with geographical coordinates ranging from 116°09′01″ to 118°12′08″ east longitude and 33°16′52″ to 34°38′05″ north latitude (Figure 1). Covering an area of 9,787 km2, Suzhou city experiences a warm temperate monsoon semi-humid climate, with an average annual temperature of 14.0–14.5 °C and an average annual precipitation of 850 mm. The study area is predominantly characterized by plains, which constitute approximately 91% of the city's total area. These plains are primarily distributed in the western and southeastern regions of Suzhou city, with altitudes ranging from 14 to 53 m.
The study area exhibits a comprehensive stratigraphic sequence, with the exception of the Qingbaikou System of the upper Archean–upper Proterozoic Eon and the upper Ordovician–lower Carboniferous systems. The exposed strata are predominantly of Sinian-Ordovician age, while Carboniferous and Permian strata are sporadically distributed. Based on groundwater storage media and water-bearing pore types, the water-bearing rock groups in the study area are categorized as follows: loose rock pore water-bearing rock group, carbonate rock fissure karst water-bearing rock group, clastic rock pore fissure water-bearing rock group, and igneous rock fissure water-bearing rock group. The loose rock pore water-bearing rock group is widely distributed throughout the region, with aquifers composed of Quaternary and Pliocene Series sediments. These sediments include clay loam, sandy loam, silt, fine sand, and medium-coarse sand. Based on their burial depth, the aquifers are further divided into shallow (less than 50 m) and deep aquifer groups. The shallow aquifer group comprises Holocene Series, middle Pleistocene Series, and upper Pleistocene Series strata, characterized by silty soil, sandy loam, clay loam with fine sand, and limited medium-coarse sand. Atmospheric precipitation serves as its primary recharge source, providing water for rural communities. The deep aquifers consist primarily of the middle Pleistocene Series, lower Pleistocene Series, and Pliocene Series, and are composed of clay loam with fine sand, medium-coarse sand, silt, and sandy loam. The cumulative thickness of the sand layer ranges from 10 to 100 m. The single-well yield is typically between 1,000 and 4,000 m3/d, and this aquifer system serves as the primary water source for centralized water supply in urban areas. Artificial pumping and irrigation are the dominant anthropogenic activities impacting shallow groundwater in the study area.
High fluoride concentrations in groundwater in the study area are a result of regional geological conditions. The study area is partly mountainous and hilly, with metamorphic and carbonate rocks as the dominant geological formations. These rocks serve as sources of fluoride in the groundwater. In addition, the study area contains an extensive alluvial plain formed by the deposition of paleochannels and the erosion and deposition of the Yellow River during its numerous floods. The silt and clay minerals remaining after the floods often contain adsorbed fluoride, contributing to the accumulation of fluoride in the groundwater.
Sampling and analyses
This study collected 49 sets of shallow groundwater samples in Suzhou. The samples were collected onsite in sterile polyethylene bottles, with their longitude and latitude recorded. In the laboratory, the samples were filtered through 0.45 μm filter membranes and analyzed for pH, total dissolved solids (TDS), bicarbonate (), carbonate (
), fluoride (F−), sodium (Na+), potassium (K+), magnesium (Mg2+), calcium (Ca2+), chloride (Cl−), and sulfate (
). pH and TDS were measured onsite with a portable water quality tester, while
and
were determined in the laboratory by acid–base titration. Other ions were analyzed using a Thermo Fisher Scientific ion chromatograph (ICS-600-900). The anion–cation balance error of all samples was within ±5%, indicating the reliability of the test results.
Statistical methods
The statistical analysis was conducted using Microsoft Excel 2021. The spatial distribution of fluoride in groundwater was analyzed using the inverse distance weighted interpolation method in ArcGIS 10.7. A Piper diagram was constructed using AqQA software to illustrate the hydrochemical types of groundwater. The saturation indices (SIs) for various rock types were calculated using the PHREEQC software. In addition, correlation and principal component analyses were performed using the R statistical programming language.
Correlation analysis
where n is the number of ranks and di is the difference between the ranks of the ith sample for the two variables.
Principal component analysis
Principal component analysis (PCA) is a widely used dimensionality reduction technique that transforms complex multidimensional data into fewer dimensions while preserving relevant information. In hydrogeochemical studies, the loadings of hydrochemical parameters on each principal component are commonly used to infer potential hydrochemical processes occurring in aquifers (Koh et al. 2016). Researchers suggest classifying loading values into three categories: low, moderate, and high, corresponding to values less than 0.5, between 0.5 and 0.75, and greater than 0.75, respectively (Egbueri & Agbasi 2022).
Human health risk assessment model
Reference values for noncarcinogenic health risk assessment model parameters
Model parameters . | Unit . | Groups . | |
---|---|---|---|
Children . | Adult . | ||
C (concentration of fluoride) | mg L−1 | Analyzed value | Analyzed value |
IR (ingestion rate) | L·d−1 | 1.5 | 2.2 |
EF (exposure frequency) | d·a−1 | 365 | 365 |
ED (exposure duration) | a | 6 | 30 |
BW (body weight) | kg | 15 | 65 |
AT (averaging time) | d | 365 × ED | 365 × ED |
DRf (reference dose of fluoride) | mg kg−1 d−1 | 0.06 | 0.06 |
Model parameters . | Unit . | Groups . | |
---|---|---|---|
Children . | Adult . | ||
C (concentration of fluoride) | mg L−1 | Analyzed value | Analyzed value |
IR (ingestion rate) | L·d−1 | 1.5 | 2.2 |
EF (exposure frequency) | d·a−1 | 365 | 365 |
ED (exposure duration) | a | 6 | 30 |
BW (body weight) | kg | 15 | 65 |
AT (averaging time) | d | 365 × ED | 365 × ED |
DRf (reference dose of fluoride) | mg kg−1 d−1 | 0.06 | 0.06 |
RESULTS AND DISCUSSIONS
Characterization of water quality parameters
![](https://iwa.silverchair-cdn.com/iwa/content_public/journal/ws/24/9/10.2166_ws.2024.202/2/m_watersupp-d-24-00275if06.gif?Expires=1730809958&Signature=fzGOhKk2sGRI2szcCMeigcUlOm129HumkzMKpiL2wYBDZ8mlmfDPYvfLYoDZ8AKCJ15TjzDBvlaol-Ush8oniAhjsIt596IwE9KVNmxHp8WEn25SSKpakplv~DKgJLSKAtVTXOFXMK4L5ciPdp8q0qnJUaCoF7KAUQYBby7qWHSwsVV0WW290a~X2puvCnvPcTVeL~jAaYrpYmBAbmWvU-btywbcfnyVWKgx7n4GizDHULuMl1gHDYVZ3m0KWaxXizfieW57H9d0zUxtQAF62uUBY2UHD5Gm-nfNm008bwAugNHZaL8RQRAbZMVA01-0dCUQA9jhFcqBfpC5BeXTOw__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
![](https://iwa.silverchair-cdn.com/iwa/content_public/journal/ws/24/9/10.2166_ws.2024.202/2/m_watersupp-d-24-00275if07.gif?Expires=1730809958&Signature=Y9lajZquCT33xcAYkiTW0uOl3MxoykRMaiVvV24GY~NQqRElrgJ1aV-bZOgt4E1kfPZtBZLdUba8e8UemfWiifl2y2PZaesYr1i9twu8LOsDvEw1jK7Osy0UwPc1qxhHZbvVWEYn81aLfsca2xYZ8PXq85oT37ox1NGJ4pANbCOIO7jSLBTCQ3qh4EmMuqTeaSgsqE4917zs1F26ZuiCjEURZ3vapzRsDus~7XuSoGzMaf6N55ZC~S8yzFs4qHvuZODcgndVyJgzqEfiEe7N2u71Q37tJ2jLjAIIkr2CQ3JzdYPflFZK9j2Tot-oh3uHIVtGrBpX8NnFD0x8BrATUw__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
![](https://iwa.silverchair-cdn.com/iwa/content_public/journal/ws/24/9/10.2166_ws.2024.202/2/m_watersupp-d-24-00275if08.gif?Expires=1730809958&Signature=P5QjXCcdvvKtJSEnoJJGbGOmFXUhvK8XFqJPYOr34CLgox7x32y9rPNFS2ej1ASvmbc~6vuxbpTgk5PVKYA7hpGsPpgoQKWcMbzWQ77aWvTNT77uCBOtMAFnorYfDftqxWlRxBGbY0M8gCOO~NtoDOCz2mK~VzH2Tv0zTO61bptPfbO1N6w4r2e~y7CD8HiwVg9K9yXUAej43m-V0IRum7vqxSY2NsvaZYayaO9UELDy3wY5DlppGM89ux3f-~UxLASMVXW-Ag~Vqu3dylMbSPMGutsUQe362Y2vlIu6PcKMzhrLCQXTD82nY3JDDT5dECoDm0FmMMFHhOSVzmdVeQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Basic statistics of water quality parameters of shallow groundwater
Statistics . | Na+ (mg/L) . | K+ (mg/L) . | Ca2+ (mg/L) . | Mg2+ (mg/L) . | Cl−(mg/L) . | ![]() | ![]() | F−(mg/L) . | pH . | TDS (mg/L) . |
---|---|---|---|---|---|---|---|---|---|---|
Permissible limit | <200a | <75b | <150b | <250a | <250a | <1.0a | 6.5–8.5a | <1,000a | ||
Min | 16.3 | 0.13 | 26.2 | 24.7 | 16 | 9.1 | 233.7 | 0.19 | 6.98 | 349 |
Mean | 91.0 | 1.8 | 83.8 | 55.9 | 85.4 | 86.5 | 523.1 | 0.89 | 7.6 | 693.9 |
Max | 696.2 | 45.3 | 211.8 | 172.7 | 442.1 | 1,048.9 | 897.8 | 2.33 | 8.32 | 2,996 |
SD | 25.2 | 0.1 | 8.2 | 3.44 | 19.5 | 7.7 | 22.4 | 0.57 | 0.11 | 60.5 |
Rf | 2% | 43% | 8% | 4% | 27% | 2% |
Statistics . | Na+ (mg/L) . | K+ (mg/L) . | Ca2+ (mg/L) . | Mg2+ (mg/L) . | Cl−(mg/L) . | ![]() | ![]() | F−(mg/L) . | pH . | TDS (mg/L) . |
---|---|---|---|---|---|---|---|---|---|---|
Permissible limit | <200a | <75b | <150b | <250a | <250a | <1.0a | 6.5–8.5a | <1,000a | ||
Min | 16.3 | 0.13 | 26.2 | 24.7 | 16 | 9.1 | 233.7 | 0.19 | 6.98 | 349 |
Mean | 91.0 | 1.8 | 83.8 | 55.9 | 85.4 | 86.5 | 523.1 | 0.89 | 7.6 | 693.9 |
Max | 696.2 | 45.3 | 211.8 | 172.7 | 442.1 | 1,048.9 | 897.8 | 2.33 | 8.32 | 2,996 |
SD | 25.2 | 0.1 | 8.2 | 3.44 | 19.5 | 7.7 | 22.4 | 0.57 | 0.11 | 60.5 |
Rf | 2% | 43% | 8% | 4% | 27% | 2% |
Note: Rf is the percentage of the water samples exceeding the permissible limit.
aThe Chinese groundwater quality limit (GAQS & IQPRC 2017).
bThe World Health Organization (WHO) drinking water limit (WHO 2017)
Spatial distribution of fluoride concentration in shallow groundwater of the study area.
Spatial distribution of fluoride concentration in shallow groundwater of the study area.
Hydrochemical types
![](https://iwa.silverchair-cdn.com/iwa/content_public/journal/ws/24/9/10.2166_ws.2024.202/2/m_watersupp-d-24-00275if09.gif?Expires=1730809958&Signature=ekmBvsTbtFOpHY-P8Nt-GeuWapXDrXvOBU0o7kiScP7cv2j95fUOlTr5xKZKBF7auIpmAfpokpAC8ZajpCxpXOoV8wwurX136eK5~305o8c1CtIU8cxpzBJ2wNAoiXBVbgkx6byhlBMCZwJf2hcPm~4~URMDICJp1sPZp9HVy0UA9kAadIgUu~Mwl2V7huKKoQ6pH7JO20ZZ8zLt1jEeCXUuS5x0pm37uhrOrrAWMv1MMRBkeWA0FkO7ibqbWhVr04N9s22WQJdJzCjmL2nYLFgq~Q7Hieo8AH2oIO~aIELjvCdw7n9V2kiL2g0xvLgMGzcZIAkZxj3XukK38Vv5kQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
![](https://iwa.silverchair-cdn.com/iwa/content_public/journal/ws/24/9/10.2166_ws.2024.202/2/m_watersupp-d-24-00275if10.gif?Expires=1730809958&Signature=i00NNZYSBYbRGWB2jrX9A-Jn~pD9iGEldMi0cDFqTQ7zDGIp6QsAU8~yJzuFq7r08MK2fbzfbXjJ~lMESUsuTNWUg79pq7UHurVcnorC0aNNRVaUWi5vuTg1xHmDG74I55VNDDm9OZn7k6ojtPSObcbRc1mXsG4g3Z2s4jfuSyNlJY~IAou9pdBFiwU3oNMFMpVGddKkwL4ILxp5LXPOgDetSs7CNFFVYlsMp7bCukhbXcL3c0L7vGRLTtY1RGv0NhsvlTGCJ0NWNj1Nf2CtNN2tkn0qqPe-EaNlntzuQSTDfdCxVRipvH1i5jjJicP4ekYrjG7rubZRv-HgEsyhxQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
![](https://iwa.silverchair-cdn.com/iwa/content_public/journal/ws/24/9/10.2166_ws.2024.202/2/m_watersupp-d-24-00275if11.gif?Expires=1730809958&Signature=FYHJthfCKzg~IdIwxyY~Dvu9c6q0g1C867LvWh9WH4V3Ehd~KnFmI2cMelwlana41Df9WOz6VF67I2Hde5rwG-YimSphZ10rtMk1Odub4YGpuFgXfRWBjUMcazIkqZ1KFNrWCgaMqggJofz5P6hnJO5Q86dtcm8JJcaE8NTHwjisb6xG5cWz6BtgSwVRGBGkjg0EoSoBsgetMSzWBDQAjUx4vA-M285VMh8f5-x11WOW2WCoKwEBXlAa73OreNyxYGE9mZJaA7KPsY7l2XaAqfDegrf37x8eYxym0oB79vBFdu6ExkS9H9A4Ca6VlUHSrIvu4AKvg-5OSLYmsW4AzA__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
![](https://iwa.silverchair-cdn.com/iwa/content_public/journal/ws/24/9/10.2166_ws.2024.202/2/m_watersupp-d-24-00275if12.gif?Expires=1730809958&Signature=HMaZDY1bV~b3c~a28swBfgqV46e3Wtn914V2U2R~X~Fj-RYmzZ-CIpJF73AbOvtCE5PRY3r-SlG9~04rZqpgv3ZppwAUOjJRX0rkT1L~7Acde1npOQN-voMywKea6zOnAeJtCQaAAXQD7stywOSkftASTSgSXqajzFhV~BkkqUungXrCqeu01OfhHGaANz1EodLuwJrlcJ1CYVAdcyxWUe47OkNhl7uDckca-sq4QQnzhwsvJIIO7BcRxmnSwaQrpQ~HXpG16bIEtztdqyzFI36QBjw-CcQgJ5WH8~X7kaKiHv3rLnAT04utV2reUz6MNmrtka8HAA4OlcmHi7jw0Q__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
For the 13 groundwater samples with F− concentrations exceeding 1 mg/L, the dominant hydrochemical type is Na-Mg-HCO3 (six samples), followed by Mg-Na-HCO3 and Mg-Ca-HCO3 (three samples each) and Ca-Mg-HCO3 (one sample). This distribution is strongly influenced by water–rock interactions within the aquifer. F− in groundwater primarily originates from the dissolution of fluorine-bearing minerals (e.g., biotite, muscovite, fluorapatite, and fluorite). For instance, the dissolution of fluorite releases both Ca2+ and F− into the aqueous environment. Consequently, F− and Ca2+ in groundwater exhibit a mutually inhibitory relationship. Groundwater with elevated Ca2+ concentrations typically has lower F− concentrations, and vice-versa.
Sources of fluoride
The relationship between F− and pH
Cation exchange
Figure 4(b) illustrates the distribution of CAI-I and CAI-II values. CAI-I ranged from −12.9 to 0.79, with a mean of −1.45, while CAI-II ranged from −0.67 to 1.29, with a mean of −0.1. Notably, 79.6% of the samples exhibited negative values for both CAI-I and CAI-II, suggesting that reverse cation exchange was the dominant process in the groundwater system. This exchange process involved the replacement of Ca2+ and Mg2+ in the groundwater by Na+ and K+ from the surrounding rocks, leading to an increase in Na+ concentration and a decrease in Ca2+ concentration. The reduced Ca2+ concentration facilitated the dissolution of fluoride minerals. Interestingly, the only sample with a fluoride concentration exceeding 1 mg/L had positive CAI-I and CAI-II values, indicating that reverse cation exchange may have contributed to fluoride enrichment in the groundwater of the study area.
Saturation index
To elucidate the precipitation, equilibrium, and dissolution states of minerals during water–rock interaction, the mineral SIs of shallow groundwater samples in the study area were calculated using PHREEQC software. SI values greater than 0 indicate mineral precipitation, values equal to 0 indicate saturation, and values less than 0 indicate undersaturation (Liu et al. 2021). The simulation results showed that SIFluorite ranged from −2.28 to −0.34, with an average of −1.39, indicating that fluorite is undersaturated in the shallow groundwater. SICalcite ranged from 0.06 to 1.18, with an average of 0.68, suggesting that calcite is oversaturated. SIDolomite ranged from 0.76 to 2.32, with an average of 1.55, also indicating supersaturation of dolomite. Finally, SIGypsum ranged from −2.86 to −1.08, with an average of −2.02, indicating that gypsum is undersaturated.
Scatter plot of (a) F− versus SIFluorite, (b) SIFluorite versus SICalcite, (c) SIFluorite versus SIDolomite, and (d) SIFluorite versus SIGypsum.
Scatter plot of (a) F− versus SIFluorite, (b) SIFluorite versus SICalcite, (c) SIFluorite versus SIDolomite, and (d) SIFluorite versus SIGypsum.
Correlation analysis
Table 3 summarizes the statistical results of Spearman's correlation coefficient analysis between ions, computed using Equation (1). For the 49 samples, the critical correlation coefficient value at a significance level of 0.01 (two tailed) is 0.365. In groundwater geochemistry, variables exhibiting correlation coefficients exceeding this critical value typically originate from the same source. As indicated in Table 3, demonstrates significant positive correlations with both Na+ and K+, with correlation coefficients of 0.579 and 0.458, respectively. This suggests that the three ions may share a similar origin, potentially linked to the weathering of silicate minerals (e.g., feldspar) within the aquifer's matrix. Conversely, F− exhibits a significant negative correlation with Ca2+ (correlation coefficient = −0.565), implying the occurrence of fluorite dissolution.
Spearman correlation of the major ions in groundwater
. | Na+ . | K+ . | Ca2+ . | Mg2+ . | Cl . | SO4 . | HCO3 . |
---|---|---|---|---|---|---|---|
K+ | 0.405a | ||||||
Ca2+ | −0.217 | 0.206 | |||||
Mg2+ | 0.624a | 0.361 | 0.235 | ||||
Cl− | 0.427a | 0.256 | 0.429a | 0.518a | |||
![]() | 0.632a | 0.229 | 0.189 | 0.685a | 0.454a | ||
![]() | 0.579a | 0.458a | 0.317 | 0.771a | 0.266 | 0.628a | |
F− | 0.229 | −0.310 | − 0.565a | 0.116 | −0.187 | 0.089 | 0.083 |
. | Na+ . | K+ . | Ca2+ . | Mg2+ . | Cl . | SO4 . | HCO3 . |
---|---|---|---|---|---|---|---|
K+ | 0.405a | ||||||
Ca2+ | −0.217 | 0.206 | |||||
Mg2+ | 0.624a | 0.361 | 0.235 | ||||
Cl− | 0.427a | 0.256 | 0.429a | 0.518a | |||
![]() | 0.632a | 0.229 | 0.189 | 0.685a | 0.454a | ||
![]() | 0.579a | 0.458a | 0.317 | 0.771a | 0.266 | 0.628a | |
F− | 0.229 | −0.310 | − 0.565a | 0.116 | −0.187 | 0.089 | 0.083 |
Note:aThe two ions are significantly correlated at the 0.01 level.
Principal component analysis
![](https://iwa.silverchair-cdn.com/iwa/content_public/journal/ws/24/9/10.2166_ws.2024.202/2/m_watersupp-d-24-00275if14.gif?Expires=1730809958&Signature=4fWU-W0Yib-IuiwU9ObsXWYDwEzI14utTr5bed8uguW6PKd0t3v7e6mBuNInXpCHayMAJalHxMBLCvLVx8sIBtZv1v8TPGXXprj8h8UrTKvrkvICLLfVG3DqzVYVY~p3K286Nj0ZZVD9gXlI1OsivO0hRibB~BBFLiW46P9s4vrNY13RUjd1m2kLehnE9xQ9mHe7yKejQ6D5qKKQzujKZSLgdtKTzPTlj8Ez3sarTyPc4IBpUV8t-jF-qoFC550y0HL425dcIzwKaUzxK9GfrKpbod8FHgTLTP4EwDWo7mhZFa-drWqWmnt72dmPfdypMiom6PhtUr7OwZNygu2EiQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
![](https://iwa.silverchair-cdn.com/iwa/content_public/journal/ws/24/9/10.2166_ws.2024.202/2/m_watersupp-d-24-00275if15.gif?Expires=1730809958&Signature=aU0CfNwhbhQCu36274wZFnuWaaq8Upzhkb0lDhoC6ghxHdVHHPf9lwaBN0MREzByDo6tSVaOnYsD-3O3HiK~bYUO3JgJ77ssoHYNPQKunsRs~D2XC0JeyyyDgcrlz2JhR3pQ3ZuzK~HFgxqEuyOshFH0vT5iy3h7h6WfIH512zGarZqmBDTtaOX~veNoVi6Biccp~TjIDGUlR9Gs5eQ6WVnJ68immj07IJxrj1ygc~q4LoqsAPgioR7cMKcZceQFcHMN8-W2WZacsM89JgaKXSuOEcoreKSuLSc2~PFgZwDhjnpidb17K3BOwBYVtYi1z20UyD-uNluiGMSMd89NVQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Statistics of variables loading on principal components
Species . | PC1 . | PC2 . | PC3 . |
---|---|---|---|
Na+ | 0.889 | −0.291 | −0.039 |
K+ | −0.062 | 0.282 | 0.861 |
Ca2+ | 0.196 | 0.875 | −0.021 |
Mg2+ | 0.880 | −0.016 | 0.065 |
Cl | 0.698 | 0.491 | −0.313 |
SO4 | 0.905 | −0.159 | −0.031 |
HCO3 | 0.691 | −0.038 | 0.405 |
F | 0.067 | − 0.813 | 0.087 |
Eigenvalue | 3.38 | 1.86 | 1.02 |
Variance explained | 42.3 | 23.2 | 12.7 |
Species . | PC1 . | PC2 . | PC3 . |
---|---|---|---|
Na+ | 0.889 | −0.291 | −0.039 |
K+ | −0.062 | 0.282 | 0.861 |
Ca2+ | 0.196 | 0.875 | −0.021 |
Mg2+ | 0.880 | −0.016 | 0.065 |
Cl | 0.698 | 0.491 | −0.313 |
SO4 | 0.905 | −0.159 | −0.031 |
HCO3 | 0.691 | −0.038 | 0.405 |
F | 0.067 | − 0.813 | 0.087 |
Eigenvalue | 3.38 | 1.86 | 1.02 |
Variance explained | 42.3 | 23.2 | 12.7 |
(a) Histogram of percentage of variables explained by principal components and (b) two-dimensional expression of variables in the first two principal components.
(a) Histogram of percentage of variables explained by principal components and (b) two-dimensional expression of variables in the first two principal components.
Human health risk assessment
According to Equation (1), the noncarcinogenic hazard quotients (HQs) of fluoride (F−) in shallow groundwater due to direct ingestion were calculated for children and adults. The results indicate that HQs range from 0.32 to 3.88 for children (average: 1.48) and 0.11 to 1.31 for adults (average: 0.51). This finding is similar to reports in the Al-Hassa region, KSA and the southwest plain of Shandong Province, North China (Liu et al. 2021; Yassin et al. 2024). Notably, 67.3 and 6.1% of the samples exceeded HQ values of 1 for children and adults, respectively. Children's higher noncarcinogenic risk is primarily attributable to disparities in water intake and body weight. Relative to their body weight, children consume more water than adults. In addition, children's lower body weight results in a higher concentration of fluoride in their bodies following equivalent fluoride exposure. Therefore, the pathogenic risk of F− in shallow groundwater warrants particular attention for children.
CONCLUSIONS
This study analyzed the physicochemical compositions of shallow groundwater in Suzhou, China, using thermodynamics simulation, multivariate statistical analysis, and health risk assessment models. The following major conclusions were obtained:
(1) The shallow groundwater in the study area was weakly alkaline, with most water samples having TDS <1,000 mg/L. F− concentrations ranged from 0.19 to 2.33 mg/L, averaging 0.89 mg/L. Fluoride concentrations in 26.5% of the samples exceeded the Grade III standard for groundwater quality in China. F− concentrations exhibited point-like distribution in the area. The Piper diagram indicated that the groundwater in the area was predominantly of the Ca-Mg-HCO3 type, while water samples with F− concentrations >1 mg/L were primarily of the Na-Mg-HCO3 type.
(2) The results of multivariate statistical analysis suggested that water–rock interactions, including dissolution of silicate minerals and fluorite, were the primary controlling factors for the hydrochemical compositions of groundwater. Thermodynamics simulations indicated that F− mainly originated from the dissolution of fluorine-containing minerals. In addition, retrogressive cation adsorption and precipitation of calcite and dolomite contributed to the enrichment of F− in groundwater. However, pH had a minimal influence on F− concentrations in groundwater under weakly alkaline conditions.
(3) Health risk assessment models revealed noncarcinogenic risks to children and adults from F− in shallow groundwater through ingestion. Therefore, groundwater treatment is necessary prior to drinking to ensure the health and safety of local residents.
Future research should sample from a wider range of locations and depths, and monitor fluoride concentrations over a longer period to comprehensively understand the spatiotemporal variations of fluoride concentrations in shallow groundwater in the study area. In addition, further research is needed to assess the health effects of fluoride exposure to better understand the potential risks to human health.
ACKNOWLEDGEMENTS
The work was funded by the Research and development fund project of Suzhou University (2021fzjj32) and the Integration and Innovation of Precise Geological Over Detection Technology for Coal Mines Based on Artificial Intelligence (SZKJXM202309).
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.