During sampling and analysis of Sultanpur groundwater, the quality control measures were exercised to avoid impurity and guarantee reliability. Standards used for analysis were prepared by using analytical-grade reagents in Milli-Q water. pH and EC were again measured in the laboratory using an electrochemistry meter (Thermo Scientific Orion Versastar Pro). The analysis of the samples was also repeated using different methods, such as Na⁺ and K⁺, which were also determined using a flame photometer (Labtronics); the argentometric titration method was used for Cl⁻ determination. The results were compared, and a deviation within 5% was found. A charge balance error (CBE) was also calculated for all major ions and was found to be <10%. For trace element analysis with Q-ICPMS, commercial multi-element standard solution and Hg single element reference standard (procured from Inorganic Ventures) were used as calibrating standards. For measurement and data quality assessment, a multi-element standard solution was measured at intervals of every 20 samples as an unknown, and the precision of all these measurements was found to be better than 5% relative standard deviation (RSD) for all elements. The elemental concentration for each unknown sample was derived from the average of three replicate analyses per measurement.

The Pearson correlation coefficient is a statistical technique commonly used to show the dependency of one variable on the others. It measures the strength of the linear relationship between two variables (Berman, 2016). Depending on their magnitude, correlations are classified as very strong (>0.80), strong (0.60–0.79), moderate (0.40–0.59), weak (0.20–0.39), and very weak (<0.20) (Wuensch and Evans, 1996; Papageorgiou, 2022). Apart from the correlation coefficient technique, principal component analysis (PCA) is among the most used multivariate statistical tools and is broadly used to gain helpful information on water quality data (Benkov et al., 2023; Franco et al., 2021; Ibrahim et al., 2023). Large data sets are transformed into small components (principal components, PCs), which are obtained from linear combinations of the original variables in such a way that the first PC tries to denote most of the variation in the original variable (Jollife and Cadima, 2016; Wold et al., 1987). These new sets of variables are also known as principal components (PCs), which are weighted linear combinations of the original variables. Eigenvalues help to discern the significance of the component. The components with eigenvalues ≥1.0 are considered significant in the studies (Kim et al., 2005). The methodology is valid when the Kaiser-Meyer-Olkin (KMO) coefficient value is greater than 0.5 (Liu et al., 2003; Sajil Kumar et al., 2014). Factor loading is classified into three classes: strong (>0.75), moderate (0.75–0.50) and weak (0.50–0.30) (Liu et al., 2003). The correlation analysis and PCA were used to identify the relations among the dissolved elements in the groundwater and their possible sources in the area.

Hydrogeochemical approaches (i.e., Piper diagram, Gibbs diagram, molar ratios, scatter plots) were used to understand the hydrogeochemical facies and mechanisms controlling groundwater chemistry of the study area. Furthermore, chloro-alkaline indices were used to understand the processes of ion exchange and reverse ion exchange between groundwater and its surroundings (host-rock) (Schoeller, 1977). The chloro-alkaline indices are expressed as: (1) CAI − 1 = [ Cl − − ( Na + + K + ) ] / Cl − (2) CAI − 2 = [ Cl − − ( Na + + K + ) ] / ( SO 4 2 − + HCO 3 − + NO 3 − )

All values are expressed in meq/L.

Different water quality indices are used to assess the overall quality of water for several usages, including drinking and irrigation. The water quality index (WQI) provides information about water quality in a single value. It may be defined as a reflection of the cumulative influence of all analysed water parameters on the overall quality of water (Horton, 1965). WQI of the analysed groundwater samples was calculated using 13 analysed parameters (pH, TDS, Turbidity, TH, Na⁺, K⁺, Ca²⁺, Mg²⁺, F⁻, Cl⁻, HCO₃⁻, NO₃⁻, and SO₄²⁻). These parameters have been assigned a certain weight from 2 to 5 based on their importance in water quality assessment (Vasanthavigar et al., 2010; Tiwari et al., 2017; Bharat and Singh, 2024). WQI is calculated by using the expressions in the following steps;

Step 1:

Assign a weight to each water hydrochemical parameter (i.e., 5 is given to TDS, F⁻ and NO₃⁻; 4 is given to pH, SO₄²⁻ and Cl⁻; 3 is given to TH and Na⁺; and 2 is given to Ca²⁺, Mg²⁺ K⁺ and turbidity).

Step 2: (3) W i = w i / ∑ i = 1 n w i where,

W i  = relative weight, w i = weightage of each parameter, and n = number of parameters.

Step 3: (4) q i = ( C i / S i ) × 100 where,

q i is the quality rating, see (Vasanthavigar et al., 2010; Tiwari et al., 2017) for more detail, and C i = concentration of each chemical parameter in each water sample. Si is the BIS (2012) and the WHO (1997), for sodium and potassium, standards of each parameters.

Step 4: (5) SI = W i × q i where, SI is the sub-index of each parameter.

Step 5: (6) WQI = ∑ SI i

Where, SI i is the sub-index of the ith parameter.

Based on the calculated WQI values, water is classified into five classes, i.e., (a) Excellent (<50), (b) Good (50–100), (c) Poor (100–200), (d) Very poor (200–300), and (e) Unfit (>300) (Neogi et al., 2023; Tiwari et al., 2017; Vasanthavigar et al., 2010).

Apart from assessing water quality for drinking purposes, several water quality indices are utilised to estimate the suitability for irrigation purposes, such as Sodium Adsorption Ratio (SAR), Residual Sodium Carbonate (RSC), Percent Sodium (%Na), Magnesium Hazard (MH), Kelley Index (KI) (Richards, 1954; Shainberg and Oster, 1979; Todd and Mays, 2004; Wilcox, 1955).

These indices were estimated by using the following equations: (7) SAR = Na / [ ( Ca + Mg ) / 2 ] 1 / 2 (8) RSC = ( C O 3 + HC O 3 ) − ( Ca + Mg ) (9) % Na = ( Na + K ) / ( Ca + Mg + Na + K ) × 100 (10) MH = Mg / ( Ca + Mg ) × 100 (11) KI = Na / ( Ca + Mg )

All parameter concentrations are in meq/L.

Carcinogenic materials (such as trace elements) can enter the human body through three main pathways: direct ingestion, dermal absorption through the skin, and inhalation through the nose and mouth. However, among the above-discussed pathways, the ingestion pathway is crucial for drinking groundwater from exposure to trace elements (Giri and Singh, 2015; O'Rourke et al., 1999; USEPA, 2004). The dose received by ingestion for the present study was calculated by the following equation: (12) ADD = ( C w × IR × EF × ED ) / ( BW × AT ) where,

ADD is an abbreviation for average daily dose (μg/kg/day),

Risk characterisation (for non-cancer health risk) is reflected by the hazard quotient (HQ). HQ was calculated by using the following equation: (13) HQ = ADD / R f D where,

R_fD is the reference dose as per the USEPA risk-based concentration table (USEPA, 2011).

For risk assessment of multiple trace elements, the HQs of all the trace metals are summed up, which gives a hazard index (HI). (14) HI = ∑ i n HQ i

Drinking water with HI > 1 shows a potential for an adverse impact on human health, requiring further study (USEPA, 2004).

The pH of the analysed groundwater samples of the study area varied from 7.53 to 8.22, with an average value of 7.83, inferring a slightly alkaline nature (Table 1). The lowest and highest pH were observed at GW-2 (7.53) and GW-51 (8.22), respectively. Electrical conductivity (EC) is an essential parameter for assessing the quality of water for drinking and irrigation utilisations and is influenced by the concentrations of dissolved ions (Tutmez et al., 2006). EC varied from 306 μS/cm (GW-17) to 1,373 μS/cm (GW-2), with an average value of 691 μS/cm (Table 1). The total dissolved solids (TDS) of the groundwater samples varied from 235 mg/L (GW-17) to 859 mg/L (GW-2), with an average value of 496 mg/L (Table 1). Figure 2 shows that all analysed of the groundwater samples belongs to fresh water categories (Freeze and Cherry, 1979). The turbidity of the groundwater samples varied from 0.01 NTU to 69.2 NTU (highest at GW-25) with an average value of 9.0 NTU. The high turbidity values in the groundwater samples may result from several factors: surface recharge, dissolution and/or weathering processes, and the corroded nature of iron pipes of hand pumps (Thakur et al., 2024; Ullah et al., 2023; Prasad et al., 2014).

The anion chemistry of the groundwater samples showed that HCO₃⁻ was the dominant anion, followed by SO₄²⁻, Cl⁻, NO₃⁻, and F⁻. The concentration of HCO₃⁻ varied from 154 mg/L (GW-17) to 482 mg/L (GW-15), with an average value of 340 mg/L (Table 1). HCO₃⁻ accounted for 91.4% of total anions (TZ⁻) in the groundwater. HCO₃⁻ in groundwater is due to the dissolution of carbonate, weathering of alumino-silicate minerals, oxidation of organic matter, and respiration by roots in the unsaturated zone (Macpherson, 2009; Njitchoua and Ngounou, 1997). Higher concentrations of HCO₃⁻, compared to other anions, indicated primary weathering of silicate minerals dominated by alkaline earth elements (Rose, 2002). SO₄²⁻ was the second dominant anion after HCO₃⁻ and varied from BDL (detection limit, 0.01 mg/L) to 133 mg/L with an average value of 15.6 mg/L. The average contribution of SO₄²⁻ in the groundwater was 3.9% of the total anions (TZ⁻). Sulphate concentration in groundwater may be primarily due to the dissolution of sulphate and sulphide minerals and anthropogenic sources (Torres-Martínez et al., 2020). F⁻, Cl⁻, and NO₃⁻ anions are generally used as tracers for water resource contamination (Loizidou and Kapetanios, 1993; Su et al., 2021; Abascal et al., 2022). Both geogenic and anthropogenic sources contribute to Cl⁻ in groundwater. Geogenic sources include the weathering of evaporite and halite minerals, the dissolution of salt deposits and atmospheric precipitation. Anthropogenic sources may include domestic and municipal discharge, farm and fertiliser discharge, industrial effluents, and leachates from landfills (Venkatesan and Swaminathan, 2009). Cl⁻ concentration ranged from 1.2 mg/L to 92.1 mg/L, with an average of 12.1 mg/L (Table 1). The maximum NO₃⁻ value was recorded as 38.0 mg/L, with an average of 5.7 mg/L (Table 1). A multitude of sources can contribute NO₃⁻ to groundwater. It includes nitrogenous fertilisers, organic manures, industrial effluents, animal and human wastes and biochemical activities (Suthar et al., 2009; Kumar et al., 2019). Sultanpur district is situated in the fertile Ganga Plain, and intensive agricultural activities are practised in about 68% of the total geographical area (CGWB, 2023). Farmers of the district generally used chemical fertilisers to increase crop productivity. The elevated NO₃⁻ level in the groundwater samples may be due to agricultural fertiliser or the leaching of human and animal wastes. The contamination of F⁻ in groundwater is generally due to both natural and anthropogenic activities (Li et al., 2019). The common natural source is soluble fluoride-bearing minerals (Li et al., 2018). Meanwhile, anthropogenic sources are industrial and agricultural activities (Kumar et al., 2018; Kalpana et al., 2019). The concentration of F⁻ varied from 0.26 mg/L to 1.71 mg/L, with an average value of 0.62 mg/L (Table 1). The spatial distribution map for F⁻ shows that the majority of the groundwater from the district has a concentration less than 1, except at locations GW-12 and GW-20 (Figure 3). Excess of HCO₃⁻ ions in the groundwater at higher pH could release F⁻ ions into groundwater by the dissolution of fluorite (CaF₂) minerals (Dey et al., 2012; Guo et al., 2007), as shown in equations 15, 16: (15) CaF 2 + 2NaHCO 3 = CaCO 3 + 2Na + + 2F − + H 2 O + CO 2 (16) CaF 2 + 2HCO 3 − = CaCO 3 + 2F − + H 2 O + CO 2

Fluoride is also dissolved in groundwater by the displacement of F⁻ with OH⁻ from minerals such as muscovite, biotite and amphibole in alkaline conditions (Guo et al., 2007; Mukherjee and Singh, 2018), as shown in equation 17. (17) KAl 2 [ AlSi 3 O 10 ] F 2 + 2OH − = KAl 2 [ AlSi 3 O 10 ] [ OH ] 2 + 2F −

Thus, a higher pH value (alkaline nature) produces favourable conditions for mobilising groundwater fluoride. Additionally, the percolation of phosphatic fertilisers from agricultural runoff may contribute to fluoride in nearby water resources (Prabhu et al., 2023; Hossein et al., 2024).

The mean abundance of major cations in the groundwater of the study area was Na⁺ > Ca²⁺ > Mg²⁺ > K⁺, respectively. The concentration of Na⁺ varied from 7.6 mg/L to 140 mg/L, with an average value of 41.1 mg/L (Table 1), contributing 34% of the total cationic mass balance (TZ⁺). In general, the possible sources of Na⁺ in groundwater can be dissolution of salt deposits, weathering of silicate, halite and evaporite, sewage and industrial effluents (Priyadarshi, 2004). Ca²⁺ concentration varied from 15.1 mg/L to 60.6 mg/L, with an average value of 38.3 mg/L (Table 1), contributing 32.2% of the total cationic charge balance (TZ⁺). The concentration of Mg²⁺ varied from 8.6 mg/L to 87.3 mg/L, with a mean value of 34.3 mg/L (Table 1) and contributing 29% of the total cation charge balance (TZ⁺). The possible sources of Ca²⁺ and Mg²⁺ are the weathering of calcite, dolomite and silicate minerals. At some locations, the concentration of Mg²⁺ exceeds Ca²⁺, possibly due to the weathering of silicate minerals and the precipitation of Ca²⁺ ions (Mayo and Loucks, 1995; Brindha et al., 2017; Nasher and Ahmed, 2021). The concentration of K⁺ varied from 3.1 mg/L to 20.0 mg/L, with a mean value of 5.5 mg/L (Table 1), contributing 4.8% of TZ⁺. The possible sources of K⁺ are potassium-rich fertilisers, potassium-bearing rocks, leachates from landfill sites and industrial discharges (Handa, 1975; Sayyed and Arjun, 2011; Banerjee and Prasad, 2020).

The Piper trilinear diagram (Piper, 1944) is extensively used to determine the chemical character of water by showing the similarities and dissimilarities of dominant anions and cations. It is also used to identify hydrochemical facies and ionic types of water (Tarawneh et al., 2019). Water samples of the study area were plotted in the Piper diagram (Figure 4). The triangular cationic diagram reveals that the majority of samples (74%) were plotted in the no-dominant zone. However, the triangular anionic diagram reveals that all the samples were plotted in the HCO₃⁻ dominant zone, which indicates the dominance of weak acids over strong acids (Figure 4). The triangular cationic diagram field revealed that about 93% of groundwater samples fall in subdivision 1, suggesting the dominance of alkaline earths (Ca²⁺ + Mg²⁺) exceeds alkalies (Na⁺ + K⁺). Also, 93% of groundwater samples fall in sub-division 5, indicating the carbonate hardness (secondary alkalinity exceeded 50%). About 5% of the samples fell in subdivision 8, signifying primary alkalinity and carbonate alkali water. Based on the dominance of different ions in the groundwater, the hydrochemical facies of the Sultanpur district were primarily Ca–Mg–HCO₃ with a few Na–K–HCO₃–Cl water types. The dominance of Ca–Mg–HCO₃ geochemical facies suggests that rock weathering primarily controls the area’s groundwater chemistry.

Table 2 presents the calculated correlation coefficients among 13 water quality parameters. TDS was very strongly associated with HCO₃⁻ (0.81), strongly associated with Na⁺ (0.70) and Mg²⁺ (0.66), and moderately correlated with Cl⁻ (0.57), and SO₄²⁻ (0.47). TH was very strongly associated with Mg²⁺ (0.89) and moderately associated with Ca²⁺ (0.45) and Cl⁻ (0.46), suggesting that the dominating factor for TH is Mg²⁺ ions, with contributions from Ca²⁺ ions and Cl⁻ ions. The moderate correlation with Na⁺ and HCO₃⁻ (0.45) indicates their contribution from the same process, possibly due to silicate weathering (Rogers, 1989). The strong correlation with Cl⁻ and SO₄²⁻ (0.66) indicates that Cl⁻ and SO₄²⁻ ions were partially or entirely derived from anthropogenic sources (Bakshe et al., 2024).

The principal component loadings and the percent of variance explained by each component are shown in Table 3. PC I has a strong loading of HCO₃⁻ (0.93), Mg²⁺ (0.77), TDS (0.83) and EC (0.72), which suggests that HCO₃⁻ and Mg²⁺ have a similar source/process (Vesković et al., 2024). PC II has a strong positive loading of SO₄²⁻ (0.89) and Cl⁻ (0.88) with weak loading of NO₃⁻ (0.47), indicating the infiltration of irrigation return flow and wastewater discharge (Elumalai et al., 2020; Nethononda et al., 2019). PC II also suggests a long evaporation history (Qu et al., 2024). SO₄²⁻ and Cl⁻ also signified the presence of anthropogenic sources (agricultural and domestic) for groundwater contamination (Cao et al., 2022). PCIII has a loading of Ca²⁺ (0.77) and TH (0.67), indicate that Ca²⁺ contribution to TH. PC III also suggest that the possible calcium-sodium ion exchange process in the study area. PC IV had a strong loading on F⁻ (0.92), indicating that the dissolution of fluoride-rich minerals (which escalated in alkaline conditions) and the use of phosphatic fertilisers can be potential factors for F⁻ in the groundwater (Kom et al., 2023; Shaji et al., 2024).

The chemical composition and solute acquisition processes of groundwater are controlled by inputs from weathering, ion exchange, meteoric inputs, adsorption/desorption, and anthropogenic sources (Xiao et al., 2012; Tiwari and Singh, 2014). The relationship between the dissolved ions can reveal the origin of solutes and the processes that yield the observed composition of water (Kumar et al., 2008). For example, the relationship between Na⁺ and Cl⁻ helps to reveal the mechanism for procuring water salinity in dry regions and quantifying the atmospheric influence (Sarin et al., 1989). Theoretically, the molar Na⁺/ Cl⁻ ratio dissolution of halite is equal to one, and if the ratio is greater than 1, it usually indicates silicate weathering (Meybeck, 1987). About 56 groundwater samples have a Na⁺/Cl⁻ ratio greater than 1, indicating the weathering of silicate minerals and anthropogenic sources (Meybeck, 1987; Tiwari and Singh, 2014). Only two groundwater samples had a Na⁺/Cl⁻ ratio of less than 1, possibly due to the exchange of Na⁺ for Ca²⁺ and Mg²⁺ from clay (Meybeck, 1987; Tiwari and Singh, 2014). Sodium is the most dominant cation, and bicarbonate is the most dominant anion in the groundwater samples of the study area, indicating that silicate weathering is the possible source of sodium (Rogers, 1989). It is probably due to the reaction of feldspar minerals with carbonic acid (H₂CO₃) in the presence of water, which releases the bicarbonate (Rajmohan and Elango, 2004).

Gibbs diagram is widely used to identify influencing factors of dissolved chemical constituents such as rock-weathering, evaporation or precipitation (Gibbs, 1970). It uses the ratio of Na⁺/(Na⁺ + Ca²⁺) and Cl⁻/(Cl⁻ + HCO₃⁻) as a function of TDS. The Gibbs diagram indicates the dominance of rock weathering with a minute influence from evaporation (Figure 5). Hence, water-rock interactions were the predominant natural process in determining groundwater hydrogeochemistry.

The 1:1 scatter plot between different ions is an important tool for identifying the sources of groundwater chemistry. The scatter plot also determines the ion-exchange/reverse ion-exchange process (Fisher and Mullican, 1997). Generally, the weathering of silicate, carbonate and sulphide minerals and the dissolution of evaporites are the primary lithogenic sources of dissolved ions in groundwater (Singh et al., 2013b). The scatter plot between Ca²⁺ + Mg²⁺ versus HCO₃⁻ + SO₄²⁻ suggests the dissolution of calcite, dolomite and gypsum when the samples fall close to 1:1 equiline (Cerling et al., 1989; Paul et al., 2019). Figure 6a shows that the majority of plotted points fell below the theoretical 1:1 equiline, showing the dominance of HCO₃⁻ + SO₄²⁻ over the Ca²⁺ + Mg²⁺. It also indicates the contribution by non-carbonate sources and demands that the excess anions (HCO₃⁻ + SO₄²⁻) be balanced by Na⁺ + K⁺. Only two of the groundwater samples fell above the theoretical 1:1 equiline, indicating some extra sources of Ca²⁺ + Mg²⁺ cations. These extra Ca²⁺ + Mg²⁺ cations should be balanced by extra anions (Cl⁻ and NO₃⁻). The 1:1 scatter plot between Ca²⁺ + Mg²⁺ and HCO₃⁻ may be crucial for identifying carbonate minerals dissolution if the plotted points are close to the theoretical 1:1 equiline (Figure 6b). The plot showed that 85% of the groundwater samples fell below the equiline, indicating that the excess of HCO₃⁻ should be balanced by Na⁺ + K⁺. The balancing Na⁺ + K⁺ is provided by the weathering of Na–K silicate minerals. In 15% of samples, excess Ca²⁺ + Mg²⁺ should be balanced by SO₄²⁻ + Cl⁻. Figure 6c is plotted between Na⁺ + K⁺ and TZ⁺ and shows that groundwater samples fell below the equiline, indicating that cations are contributed in a higher ratio via silicate weathering (Stallard and Edmond, 1983). Also, Na⁺ and K⁺ showed a better correlation with TZ⁺, suggesting that calcite and gypsum were not major contributors due to the dominance of silicate weathering in aquifer media (Nasher and Ahmed, 2021). The 1:1 scatter plot between HCO₃⁻ and SO₄²⁻ + Cl⁻ shows that HCO₃⁻ dominated the anion chemistry (Figure 6d). The abundance of HCO₃⁻ in the groundwater explains the silicate weathering in the area (Rogers, 1989). When feldspar minerals react with carbonic acid in the water, they release bicarbonate ions (Lakshmanan et al., 2003) as shown in equation 18 below. (18) 2 KAlSi 3 O 8 + 2 H 2 CO 3 + H 2 O ( Feldspar ) ( Carbonic Acid ) = Al 2 Si 2 O 5 ( OH ) 4 + 2 K + + HCO 3 − + 4 SiO 2 ( Kaolinite ) ( Bicarbonate ion )

Furthermore, a molar ratio (Na⁺ + K⁺)/TZ⁺ is 0.28, and the (Ca²⁺ + Mg²⁺)/(Na⁺ + K⁺) ratio is 3.31, indicating that groundwater chemistry was largely controlled by silicate weathering, followed by the contribution from carbonate mineral dissolution. The bivariate plot of Mg²⁺/Na⁺ versus Ca²⁺/Na⁺ and HCO₃⁻/Na⁺ versus Ca²⁺/Na⁺. Figure 7 suggests that the chemical composition of groundwater is primarily controlled by the weathering and dissolution of silicate and carbonate minerals (Gaillardet et al., 1999). The sand fraction of the Ganga Plain sediments mainly comprises quartz, muscovite, biotite, plagioclase, and orthoclase (Tripathi et al., 2006; Yadav et al., 2024; Srivastava et al., 2025). Examples of the reaction of silicate weathering are: (19) 2 NaAlSi 3 O 8 + 9 H 2 O + 2 H 2 CO 3 ( Albite ) = Al 2 Si 2 O 5 ( OH ) 4 + 2 Na + + 2 HCO 3 − + 4 H 4 SiO 4 ( Kaolinite ) (20) CaAl 2 Si 2 O 8 + 2 CO 2 + 3 H 2 O ( Anorthite ) = Al 2 Si 2 O 5 ( OH ) 4 + Ca 2 + + 2 HCO 3 − ( Kaolinite ) (21) Mg 2 SiO 4 + 4 H 2 CO 3 ( Olivine ) = 2 Mg 2 + + 4 HCO 3 − + H 4 SiO 4 (22) 2 KMg 3 AlSi 3 O 10 ( OH ) 2 + 14 CO 2 + 15 H 2 O ( Biotite ) = Al 2 Si 2 O 5 ( OH ) 4 + 2 K + + 6 Mg 2 + + 14 HCO 3 + 4 H 4 SiO 4 ( Kaolinite )

The values of CAI-1 and CAI-2 are either negative or positive, depending on whether Na⁺ and K⁺ are exchanged/reverse-exchanged for Ca²⁺ and Mg²⁺ from the host environment (Singh et al., 2018). The value will be positive if Na⁺ and K⁺ in water replace Ca²⁺ and Mg²⁺ of the aquifer material, indicating reverse ion exchange (Rao et al., 2012; Srinivasamoorthy et al., 2013). The value will be negative, indicating chloroalkaline disequilibrium and the reaction as a cation-anion exchange reaction. The negative value also suggests that host rocks are groundwater’s primary source of dissolved ions (Srinivasamoorthy et al., 2013; Mohamed et al., 2022). The Schoeller indices of the samples were calculated, and it was found that CAI-1 and CAI-2 varied from −68.8 to −0.27 with an average value of −12.09 and −0.78 to −0.05 with an average value of −0.26, respectively (Supplementary Table S1). The negative values CAI-1 and CAI-2 signify the chloro-alkaline disequilibrium or cation-anion exchange reaction in the study area.

The evidence for cation exchange can be verified by plotting the geochemical data on a bivariate plot between Ca²⁺ + Mg²⁺ − HCO₃⁻ − SO₄²⁻ (meq/L) and Na⁺ + K⁺ − Cl⁻ (meq/L). The axis Ca²⁺ + Mg²⁺ − HCO₃⁻ − SO₄²⁻ represents the amount of Ca²⁺ + Mg²⁺ gained or lost relative to that provided by minerals such as calcite, dolomite and gypsum. The axis Na⁺ + K⁺ − Cl⁻ represents the amount of Na⁺ and K⁺ gained or lost relative to that provided by Cl⁻ salts. The relationship between Ca²⁺ + Mg²⁺ − HCO₃⁻ − SO₄²⁻ (meq/L) and Na⁺ + K⁺ − Cl⁻ (meq/L) should be linear with a slope of ‘−1’ when the geochemical process of groundwater is dominated by cation exchange (Fisher and Mullican, 1997; McLean et al., 2000). Figure 8 shows that the Sultanpur district groundwater plotted points are close to a straight line (with R² = 0.78) with a slope of −0.93, suggesting the ion exchange reactions (García et al., 2001).

The result of the trace elements (Al, Cr, Mn, Fe, Ni, Cu, Zn, As, Se, Sr, Cd, and Pb) analysis in the groundwater samples of the Sultanpur district is provided in Table 1. The Ni, Cr, Cu, Zn, As, Se, Cd, and Sr concentrations did not exceed the acceptable limits of the BIS (2012) (for Sr, USEPA drinking water regulation was used). In contrast, Al, Mn, Fe, and Pb concentrations were higher in some samples and exceeded BIS’s acceptable drinking water limit (BIS, 2012) (Table 1). Al concentration varied from BDL (detection limit, 0.001 μg/L) to 85.7 μg/L, exceeding the acceptable limit (30 μg/L) at two sites (GW-14, GW-39). Fe concentration varied from 10.2 μg/L to 5674 μg/L (Figure 9). The spatial variation map also shows that 4 locations (GW-5, GW-16, GW-27, and GW-37) in the district have very high Fe contamination (Figure 9). About 28% of the samples exceeded the acceptable iron limit (300 μg/L). Mn concentrations varied from 2.0 μg/L to 486 μg/L, and exceeded the acceptable limit (100 μg/L) in 17% of the samples (Figure 10). The principal component loading of trace elements is shown in Table 4. The table shows that PC I have a strong loading to Cd (0.85), Pb (0.82), and Cu (0.75), and a moderate loading to Al (0.68) and Cr (0.67), explaining about 26.5 percent of the variance. The source of these elements may be geogenic. PC II has strong loading to Fe (0.90) and Mn (0.86) with moderate loading to Cr (0.54), indicating the local host rock may be responsible for Fe and Mn in groundwater samples (Table 4). Geogenic processes that often contribute Fe and Mn in groundwater in the alluvial plains involve the weathering of iron and manganese-bearing rocks and minerals (Arshad and Umar, 2023). Organic matter-rich alluvial soil promotes microbial activities and thus creates reducing conditions that favour Fe and Mn dissolution in groundwater (Weng et al., 2007). Furthermore, the formation of ferromanganese nodules is an important soil-forming process in the Ganga alluvium (Srivastava et al., 2025). Prolonged dry spells and monsoonal wet seasons contribute to the weathering and redistribution of elements. Upper horizons typically lack mottling and gleying, indicating limited waterlogging, while lower horizons show such features, suggesting waterlogging that aids the mobilisation and precipitation of Fe and Mn. The alluvial sediments demonstrate moderate chemical weathering, breaking down aluminosilicate minerals to release essential elements for the nodule formation (Kumar et al., 2024). Additionally, chromium and other elements were also mobilised and concentrated in these nodules. PC III has a moderate loading to Zn (0.72) and Ni (0.53), indicating possible anthropogenic sources for both trace elements (Table 4). Ni in the groundwater could be due to vehicular emissions, landfill pollution and waste discharge. Agricultural activities could be a major source of Zn in the groundwater. PC IV has a strong loading to As (0.76) (Table 4), indicating the geogenic processes may be a possible source of it.

The analysed data is compared with the BIS (2012) drinking water recommendations to determine their suitability for drinking purposes (Table 1). The data is also compared with several ratios/indices to check their suitability for irrigation purposes.

The pH of the Sultanpur district groundwater samples (7.53–8.22) was well within the safe limit (6.5–8.5) prescribed for drinking purposes. Turbidity is a crucial physical water parameter that depends on the presence of suspended solids in water. Drinking turbid water (exceeding acceptable limits) can cause diseases associated with gastrointestinal illness (Mann et al., 2007). Turbidity of the groundwater was found between 0.01 NTU and 69.2 NTU. Around 50% of the water samples exceeded the acceptable limit (1 NTU) as prescribed by BIS (2012). TDS in the district varied from 235 mg/L to 859 mg/L, and 53% of the samples exceeded the acceptable limit (500 mg/L) (Table 1). Based on total hardness (TH) values, water is classified as soft (0–75 mg/L as CaCO₃), moderately hard (75–150 mg/L as CaCO₃), hard (150–300 mg/L as CaCO₃) and very hard (>300 mg/L as CaCO₃) (Sawyer and McCarty, 1967). The TH of the groundwater of the study area varied from 100 mg/L to 464 mg/L, and 81% of samples exceeded the BIS acceptable limit of 200 mg/L. Fluoride prevents tooth decay and strengthens tooth enamel when taken at the prescribed level. According to the WHO, the minimum requirement of F⁻ in water is 0.5 mg/L. However, excess fluoride can cause serious health issues, including dental fluorosis, skeletal fluorosis, osteoporosis, arthritis, and bone damage (Solanki et al., 2022). F⁻ concentration varied from 0.26 mg/L to 1.71 mg/L. In about 5% of the groundwater samples, F⁻ concentration exceeded the BIS acceptable limit of 1.0 mg/L (Table 1). Nitrate, chloride and sulphate concentrations were within the acceptable limit of 45 mg/L, 250 mg/L, and 200 mg/L, respectively, in all the groundwater samples (Table 1).

Excess sodium consumption can raise blood pressure issues (Scheelbeek et al., 2016). The concentration of Na⁺ varied from 7.6 mg/L to 140 mg/L (Table 1). Ca²⁺ and Mg²⁺ are essential ions for bone health, cell development, muscle function and heart regulation (Baker and Worthley, 2002; Fouhy et al., 2023). However, excess intake of calcium and magnesium may cause nausea, kidney stones, risk of heart disease, low blood pressure, and difficulty in breathing (Anderson and Klemmer, 2013; Fouhy et al., 2023). The concentration of Ca²⁺ was well within the BIS acceptable limit of 75 mg/L. However, the concentration of Mg²⁺ exceeded the BIS acceptable limit of 30 mg/L in 69% of the groundwater samples of the study area (Table 1).

Fe is essential in various metabolic processes, including immune response, oxygen transport and energy production (McDowell et al., 2024). However, excess intake may cause liver cancer, hemochromatosis, diabetes, heart disease, organ damage, joint pain, hemosiderosis and infertility (Rahman et al., 2024; Chaturvedi et al., 2014). About 28% of samples exceeded the drinking water limit prescribed by the BIS (2012) (Table 1). Mn is another trace element that exceeded the prescribed BIS (2012) limit in about 17% of samples in the district. Excess intake of Mn can cause serious health risks, including neurotoxicity (conditions like manganism, a disease similar to Parkinson’s disease) and anaemia (Baj et al., 2023; Shaffer et al., 2023). Lead is a poisonous element that affects almost every part of the human body, especially the renal, reproductive, and nervous systems (Wani et al., 2015). Pb may also affect children’s brain and intellectual development (Mani et al., 2019). The lead concentration exceeded the drinking water limit at one location (GW-46) (Table 1).

WQI was calculated from the analysed parameters and varied from 35.9 to 106.5, with a mean value of 58.1 (Supplementary Table S1). Based on the WQI categorisation, groundwater is considered as ‘excellent’ (0–50), ‘good’ (50–100) and ‘poor’ (>100) (Vasanthavigar et al., 2010). As per this categorisation, 65% of samples belonged to ‘good’, 28% to ‘excellent’ and 7% to ‘poor’ category. The spatial distribution of WQI of the Sultanpur district is shown in Figure 11. Based on the spatial distribution of WQI, Mirdaspur, Jajjour, Bhatpura, and Shivgarh villages have poor water quality.

Health risk assessment for adults and children was also estimated due to the intake of trace elements via drinking water. Table 5 and Supplementary Table S1 present the calculated Hazard Quotient (HQ) and Hazard Index (HI) for adults and children of all the groundwater samples, respectively. Calculated HQ values varied from 0 to 2.38 with an average value of 0.075, and 0–1.49 with an average value of 0.049 for children and adults, respectively. The result shows that manganese and arsenic have the highest HQs, advocating that these two trace elements can be high-potential health risk pollutants. The calculated HI values ranged from 0.33 to 3.28 with an average value of 0.89, and 0.2–2.58 with an average value of 0.58 for children and adults, respectively (Supplementary Table S1). The calculated HIs for children have values exceeding 1 at 15 locations (around 26%) (GW-3, GW-4, GW-5, GW-6, GW-9, GW-17, GW-18, GW-25, GW-31, GW-35, GW-36, GW-38, GW-39, GW-44, and GW-45), suggesting a potential health risk to the children due to consuming drinking water at these locations (Supplementary Table S1 and Figure 12). However, for adults, the HIs value is greater than 1 at seven locations (around 12%) (GW-3, GW-4, GW-6, GW-17, GW-31, GW-39, and GW-45) (Supplementary Table S1 and Figure 13). Moreover, 3 samples (GW-1, GW-12, and GW-41) have HIs for children close to 1, and 3 samples (GW-9, GW-25, and GW-35) have HIs for adults close to 1, advocating that such samples were not suitable for direct consumption for children and adults, respectively. Hence, the calculated HI values for children and adults suggest that the groundwater needs appropriate treatment and management to protect human health at several locations in the study area.

Different classifications based on several water quality indices, such as %Na, RSC, SAR, KI, and MH, were used to determine the suitability of irrigation water.

The US Salinity Laboratory proposed a diagram to assess the suitability of water for irrigation purposes (Richards, 1954). In the USSL diagram, the groundwater for irrigation purposes is classified based on the EC and the SAR values, as shown in Figure 14. The diagram explains the combined effect of sodium hazard and salinity hazard while classifying irrigation water. Figure 14 shows that the majority of the samples fall in the C2S1 and C3S1 zones. C2S1 denoted medium salinity and low alkali water, indicating that most of the soil is at low risk of exchangeable sodium and salinity. About 25% of samples fell in the C3S1 zone, which denoted high salinity and low alkaline water. Due to high salinity, water in this zone requires special management for salinity control.

Richards (1954) proposed the classification of water resources for irrigation purposes based on the SAR value. The water quality SAR measures the amount of sodium relative to the amount of calcium and magnesium in water. The value of SAR varied from 0.26 to 15.8, with a mean value of 1.46 (Supplementary Table S1). The irrigation water is classified into four categories based on SAR values, i.e., ‘low’ (<10), ‘medium’ (10–18), ‘high’ (18–26), and ‘very high’ (>26). Based on this classification, most groundwater samples are classified as ‘low’ (SAR < 10) category.

Percent sodium (%Na) is another parameter used to evaluate the suitability of water for irrigation utilisation. Based on Na% values, the groundwater can be classified into five categories, i.e., ‘excellent’ (<20), ‘good’ (20–40), ‘permissible’ (40–60), ‘doubtful’ (60–80), and ‘unsuitable’ (>80) (Wilcox, 1955). Based on this classification, groundwater was classified into ‘excellent’ (31%), ‘good’ (60%), ‘permissible’ (5.6%), and ‘doubtful’ (3.4%) (Wilcox, 1955). Irrigation with high sodium concentrations in water can cause the exchange of Na⁺ ions in water for Ca²⁺ and Mg²⁺ ions in soils, which reduces the permeability of the soils and causes poor internal drainage (Collins and Jenkins, 1996). Wilcox (1955) used %Na and EC to evaluate groundwater suitability for irrigation, as shown in Figure 15. The plot indicated that 88% of the groundwater samples belonged to the ‘excellent to good’ category.

Excess of bicarbonate and carbonate compared to alkaline earth (Ca²⁺ + Mg²⁺) impacts the suitability of groundwater for irrigation usage because it may cause complete precipitation of Ca²⁺ and Mg²⁺ ions as carbonate (Karanth, 1987). RSC of the groundwater samples varied from −2.39 meq/L to 5.66 meq/L with a mean value of 0.88 meq/L (Supplementary Table S1). According to Richards (1954) water classification, the suitability of groundwater for irrigation usage is categorised into ‘good’ (<1.25), ‘moderate’ (1.25–2.5) and ‘unsuitable’ (>2.5). Based on this classification, 94.8% of groundwater samples belonged to the ‘good’ category and 5.2% belonged to the ‘unsuitable’ category.

Classification of water for irrigation purposes can also be done by the Kelley index (KI) (Kelley, 1963). KI value (≥1) is considered ‘unsuitable’, and KI value (<1) is considered ‘suitable’ for irrigation usage. By definition, KI evaluates the proportion of sodium concerning calcium and magnesium, signifying sodium-related issues such as reduced permeability and soil infiltration (Kelley, 1963). Based on the KI classification, around 93% of the groundwater samples were ‘suitable’ for irrigation purposes. The remaining 7% of the groundwater samples were ‘unsuitable’ for irrigation usage. Magnesium hazard (MH) is another method for classifying water for irrigation purposes. It indicates the degree of damage caused by magnesium to the soil structure (Tahmasebi et al., 2018). Excess magnesium is absorbed between the clay particles, reducing the soil’s infiltration capacity. It adversely impacts crop growth, resulting in low production (Ravikumar et al., 2011). The value of MH varied from 26.83 to 77.38, with a mean value of 58.82 (Supplementary Table S1). A value of MH > 50 denotes that water is unsuitable for irrigation, while values of MH < 50 denote that water is suitable for irrigation usage. Based on the MH classification, 14% and 86% of groundwater samples belonged to ‘suitable’ and ‘unsuitable’ categories, respectively.

The present study advocates implementing mitigation measures to reduce the concentration levels of major and trace elements in the area’s groundwater, which can pose a risk to the local population. Furthermore, calculated irrigation indices suggested that the study area needed an adequate drainage and water management plan. Awareness and education about groundwater quality and its impact on human health and a water management plan for irrigation practices are strongly recommended for the district. The groundwater sampling and analysis were carried out during the post-monsoon season, which shows the limitations of the present study. Hence, this study recommends a seasonal groundwater sampling and analysis in the district to understand the seasonal variation in the major and trace elements chemistry and their potential impact on the local population’s health. Furthermore, the present study contributes to baseline data generation on major and trace elements of groundwater and offers scope for understanding and comparative studies in future.