Hydrogeochemistry, solute source identification, and health risk assessment of groundwater of cancer-prone region in India

Groundwater is a very useful primary resource for drinking, irrigation, and daily needs such as cooking and washing (Adimalla et al. 2018). Globally, approximately 80–95% of rural drinking water, more than 65% of irrigation water, and 50% of urban drinking water are sourced from groundwater (Adimalla 2020). Poor groundwater quality is a threat to the health of humans. The groundwater quality deterioration may be either from natural or anthropogenic sources or both (Barbieri et al. 2019). The sources of the trace metals like B, Cu, Co, Ni, Pb, Be, As, Se, Hg, U, and Cr, in groundwater, are generally geogenic (Kumar et al. 2018; Ricolfi et al. 2020), and anions like NO₃, SO₄, and NH₄ are generally anthropogenic from the application of fertilizers used in agriculture practices, wastewater infiltration, unmanaged aquifer recharge, etc. (Huang et al. 2013; Adimalla 2020). Further, the presence of certain anions like NO₃ and NH₄ in the groundwater reflects information about the anthropogenic sources of pollution as a result of increased human interaction with the groundwater resources (Reza et al. 2019). Nitrogen is the most abundant gas in the atmosphere and its recycling in the environment depends on both biotic and abiotic factors. Higher concentrations of nitrogen in the water can occasionally be due to geological deposits of nitrate salts and decay of forest residues but are usually due to improper disposal of human wastes, industrial wastes, and agricultural practices (Barbieri et al. 2019; Shukla & Saxena 2020).

Major sources of groundwater contamination are organic and inorganic pollutants carried by the rainwater/wastewater during infiltration, and geochemically originated toxic metals as a result of leaching and secondary reactions of aquifer minerals. Climate change is having an impact on the intensity and duration of rainfall resulting in a change in the surface runoff and infiltration patterns, ultimately leading to a change in groundwater storage and quality (Shrestha et al. 2020; Barbieri et al. 2021). The change in metal concentration in dry and wet periods was observed to be affected by a change in environmental conditions, evapotranspiration, and runoff (Barbieri et al. 2019; Ricolfi et al. 2020). The dissolution of minerals and solute acquisition is also affected by the change in the redox conditions of the aquifer. The oxidizing aquifer promotes the dissolution of trace metals that are soluble in an oxidized state and the reducing aquifer promotes the dissolution of trace metals soluble in a reduced state (Stuyfzand 1999; Malakar et al. 2022). The anthropogenic pollutants (NO₃, NH₄) play a major role in the dissolution of the geogenic element in sub-surface water (Weber et al. 2006; Anoubam et al. 2016; Kumar et al. 2021). Intake of groundwater contaminated with trace metals and organic pollutants results in chronic diseases such as cancer and organ failure (Garg et al. 2014; Sar et al. 2018; Malyan et al. 2019; Saleh et al. 2019). Further, poor quality groundwater used for irrigation causes a threat to sustainable agricultural practices (Srinivasan & Reddy 2009) and food quality, and subsequently poses a threat to human health (Wcislo et al. 2002; Rashed 2010; Chotpantarat et al. 2011).

The multivariate statistical methods, such as principal component analysis (PCA), factor analysis (FA), cluster analysis (CA), hierarchical cluster analysis (HCA), and correlation coefficient matrix, are very helpful in interpreting the complex environmental data matrices, identification of possible influencing factors/sources, and management solutions to the environmental problems (De Andrade et al. 2008; Goyal et al. 2021; Pant et al. 2021; Samadi 2022). Usually, CA is carried out to reveal specific links between sampling points, while FA/PCA is used to identify the ecological aspects of pollutants in environmental systems (Varol & Sen 2009; Maurya & Kumari 2021). PCA is a more effective multivariate tool for large data reduction by transforming a given set of interrelated variables into a new set of variables/groups that may be considered as having similar behavior and common origin (Edokpayi et al. 2018; Kormoker et al. 2021; Proshad et al. 2021). The low co-variance (CV) (<50) of the principal component (PC) is an indicator of geogenically originated pollutants and CV > 50 represents anthropogenic impact in water (Wang & Lu 2011; Zhang et al. 2016).

The Bathinda district of Punjab state in India has been reported with the highest per capita number of cancer patients and the reasons highlighted for the high cancer incidences are excessive pesticide usage, toxic metals in the groundwater, and the coal-fired thermal power plant's wastes (Baker 2010; Aggarwal 2015; Duggal & Rani 2018; Singh et al. 2018). Thakur et al. (2015) analyzed trace metals, pesticides, and other relevant parameters in some major drains, water samples (surface as well as groundwater), fodder, vegetable, and blood samples, and concluded that these samples contained harmful contaminants above safe levels. Sharma et al. (2021) analyzed the water samples from Bathinda and Faridkot districts and observed elevated concentrations of U, Fe, As, Hg, and Pb. Sharma et al. (2019) investigated the groundwater samples in the Ropar wetland and observed Cd and Cr exceeding the drinking water limits. The habitation of the Bathinda district is mostly agrarian and dependent on groundwater for agricultural activities and domestic needs including drinking due to the lack of rivers in the district. In recent years, a significant increase in the district's population has been observed and the usage of land also continues to vary. The agricultural practices have also changed and in turn the usage of water and agrochemicals. The district houses more than 4,000 micro, small, and large industrial enterprises. As a result of the developmental activities, a change in the status of groundwater, the primary source of domestic supply, is expected and, therefore, necessitates the assessment of water quality status and associated health risks to the residents.

Therefore, this study has been carried out in the Bathinda district of the Malwa region of Punjab with the objectives: (a) evaluation of groundwater hydrochemical parameters, (b) sources of solute acquisition, (c) human health hazard and suitability for irrigation, and d) way forward to minimize the impact on human health. The outcomes of the study will be helpful for policymakers and planners in devising measures to mitigate groundwater deterioration and improve human health.

The water level depth in the area ranges from 2.24 to 20.76 meters below ground level (mbgl); (Singh et al. 2018). The general slope of the water table is towards the southwest from the north, east, and southeast. The aquifers embody several granular layers alternating with clay lenses. The unconfined aquifer in the region ranges from 7 m to 71 m depth and a clay layer of 8–10 m separates the first aquifer from the second aquifer which extends up to 139 m. The first and second aquifer are mainly composed of fine to coarse grained sand. The specific yield value for the unconfined aquifer is 0.072 (CGWB 2017).

The groundwater samples were collected from 20 locations on a grid basis (10 × 10 km) from Bathinda district, Punjab during pre-monsoon (April 2019) and post-monsoon (December 2019), and the coordinates of the sampling locations were recorded with the global positioning system (Figure 1). The groundwater samples were collected from the hand pumps and bore wells (depth: 30–150 m) after purging for 10–15 minutes and achieving constant temperature in polypropylene bottles. The samples for major ion analysis were collected in polypropylene (PP) bottles (500 mL) after filtration through a 0.45 μm membrane filter, and for trace metals in PP bottles (125 mL) after acidification to pH < 2 with HNO₃. The samples were transported in an ice box to the laboratory and kept at 4 °C in a cold cabinet till the analysis of samples. All the chemicals used for preservation and analysis were of analytical grade (Merck).

The pH, conductivity, and temperature of the samples were measured in situ with the help of a multi-parameter analyzer (Thermo Scientific Orion Star A329) calibrated in the field before sampling. The major cations (Li, NH₄, Ca, Mg, K, and Na and anions (F, Cl, NO₃, NO₂, and SO₄) were analyzed using ion chromatograph (Metrohm 930 Compact IC Flex) with a conductivity detector and the alkalinity (HCO₃) was determined by the acid titration method (APHA 2017). The trace metals (B, Cu, Co, Ni, Pb, Be, As, Se, Hg, U, and Cr) were analyzed by inductively coupled plasma mass spectrometry (ICP-MS) (Agilent 7850 ICP-MS). The ion chromatograph and ICP-MS were calibrated for the analyte of interest using certified reference materials (CRMs) traceable to NIST (Merck) and the standards/blank were also run after a periodic interval during the analytical run for the continuing calibration verification (CCV). The analysis run was accepted if the percentage recovery in the CCV run was within ±10%. All the parameters were analyzed in triplicate for quality control and the analysis correctness was checked by anion-cation balance and the analyses were accepted with the difference of ±5%.

The descriptive analysis for minimum, maximum, range mean, standard deviation, and Pearson's correlation coefficient (r) for different chemical and physical parameters and the principal component analysis (PCA) in the study was conducted by SPSS version-22 software.

Pearson's correlation analysis was used for revealing and highlighting the relationship among the parameters (Egbueri et al. 2019). Correlation coefficients <0.5 are supposed to exhibit poor correlation. A correlation coefficient of 0.5 is termed a good correlation and >0.5 is termed to have excellent correlation (Kaiser 1958). Further, p values <0.01, <0.05, and <0.1 indicate a strong, significant, and moderate correlation among the parameters respectively (Goyal et al. 2021).

PCA is a multivariate statistical analysis for reducing the dimensionality of large datasets that are often difficult to interpret, increasing the interpretability without losing information. The PCA has helped in the identification of the major variable factors responsible for pollution (Edet et al. 2013; Vasanthavigar et al. 2013; Singaraja et al. 2014; Varol & Davraz 2015; Ismail et al. 2016; Zhang et al. 2016; Chabukdhara et al. 2017; Sharma et al. 2021). Factor loading helps to arrive near the significant factor and the Kaiser Normalization scheme is used for the interpretation of the factor score on varimax rotation. The significance of a factor is deduced from the Eigen value and the factor with the highest Eigen value is most significant. Eigen value ≥1 is considered significant (Demirel & Guler 2006; Singaraja et al. 2014).

The hydrochemical facies represents the dominancy of the major cations and anions in the groundwater (Adimalla 2020) and is presented by the trilinear diagram (piper chart) which was prepared using Grapher software version-14. Water quality indices (WQI) for drinking and irrigation usage and health risk assessment were computed using MS-Excel 2016.

WQI is used to classify the status of water for different designated usage by comparing the observed parameters with the guideline value for the designated uses (Ahada & Suthar 2018; Jaswal et al. 2021; Shalumon et al. 2021).

WQI values in the range 0–25, 26–50, 51–75, 76–100, and >100 were classified as excellent, good, medium, poor, and unsuitable for use (Mgbenu & Egbueri 2019; Pant et al. 2021). The guideline values for computing the WQI were considered from BIS (2012), BIS (1986), WHO (2017), and Singh et al. (2021).

The health risk assessments of both overall health risk and cancer risk (CR) were carried out for males, females, and children, for both pre-and post-monsoon period samples. The chronic daily intake (CDI), hazardous quotient (HQ), hazardous index (HI), and cancer risk (CR) were calculated by the following equations (Equations (6)–(10)) (EPA 1989; Duggal & Rani 2018; Ravindra & Mor 2019; Adimalla 2020; Ghosh et al. 2020; USEPA 2022) with input parameters provided in Table 1.

The spatial distribution of HI and CR for males, females, and children in pre- and post-monsoon has been done by using IDW (inverse distance weighting) (Fallahzadeh et al. 2017), ARC GIS10.7.1.

The mean, minimum, maximum, standard deviation, and % samples exceeding the prescribed limits for drinking water for the observed parameter are given in Table 2. The pH of the groundwater samples ranged from circumneutral to alkaline in the range of 7.3–8.2 and 7.6–8.4 in pre- and post-monsoon respectively. During the pre-monsoon period, pH has excellent significant negative correlation with Ca (−0.820) and Mg (−0.779) and a strong positive correlation with fluoride (0.531); however, in the post-monsoon period, a good significant negative correlation was observed with U (−0.499) and NH₄ (−0.41) (Table 3). The anions and cations present in the samples were in the order SO₄ > Cl > NO₃ > F and Na > Ca > Mg > NH₄ respectively. Fluoride concentration exceeded the prescribed limit in 35% and 45% of samples in pre-monsoon and post-monsoon respectively. The F ions were observed to have a significant positive correlation with Co (0.731) and a strong positive correlation with pH (0.531) during the pre-monsoon period indicating the dissolution of fluoride ions into the groundwater from Co and F-bearing minerals with increasing pH. However, no significant correlation of F with other parameters was observed in the post-monsoon period. A low concentration of Ca in the groundwater may be also responsible for the dissolution of F in alkaline groundwater (Adimalla et al. 2018). The electrical conductivity (EC) of groundwater samples was observed to be lower in post-monsoon samples indicating a positive impact of groundwater recharge and can be one of the solutions for mitigating the high salinity problem of the area. Nitrate concentration also exceeded the prescribed limits in 35% of samples and although the mean concentration reduced for post-monsoon samples, the number of samples exceeding the limit increased by around 15% indicating input through groundwater recharge and agricultural activities. This was further confirmed by the significant positive correlation with NH₄ ions (0.594) for post-monsoon samples (Table 3). Further, NH₄ ions showed a strong negative correlation with pH (−0.471) indicating a decrease in pH with the conversion of NH₄ ions to NO₃ ions. Higher sulfate concentrations in drinking water are mostly associated with laxative effects and more than 50% of samples have sulfate concentrations exceeding the limits. Ca, Mg, and NH₄ were observed higher than the acceptable limit in 15%, 70%, and 70% of samples in pre-monsoon and in 35%, 80%, and 35% of the post-monsoon samples respectively. The concentration of Ca and Mg increased in post-monsoon while the NH₄ (0.10–1.71 mg/L) concentration decreased in post-monsoon (0.2–0.84 mg/L), due to the percolation of oxygenated rainwater into the aquifer. Widespread agricultural practice in the district is responsible for the nitrogenous pollutants entering the groundwater as a result of urea and DAP (diammonium phosphate) application. NO₃ was observed in the range 1–336.7 mg/L and 1.11–126.6 mg/L in pre- and post-monsoon groundwater samples respectively. The unused nitrogenous compounds applied in agriculture infiltrate the groundwater and oxidize to NO₃ (Zhang et al. 2014). Further, NO₃ is responsible for oxidizing conditions below the earth's surface and mobilization of U and other ions (Konhauser 2007; Nolan & Weber 2015). The trace metal concentration in the groundwater samples also exceeded the prescribed limits for drinking water with U concentration exceeding limits in 90% of samples during the post-monsoon period. Almost all the trace metals showed a significant positive correlation with each other except U during the post-monsoon period indicating similar sources and processes resulting in their dissolution. U ions were observed to be strongly correlated with B (0.515), Ca (0.487), and pH (−0.499). The solubility of the U ions is enhanced with the increase in Ca ions due to the formation of zero-valent Ca-U-CO₃ complexes and a reduction in pH (Ulrich & Jakob 2019).

The chemical composition of groundwater is influenced by many factors such as recharge water properties, mineral composition of the aquifer, groundwater residence time, evaporation and dissolution, and anthropogenic activities in the watershed. Geochemical modeling helps in predicting the reactive mineralogy of the subsurface through mineral equilibrium calculations from the groundwater quality data. The modeling using PHREEQC indicated 13 potential solid phase minerals, namely anhydrite, aragonite, calcite, chalcedony, chrysotile, dolomite, fluorite, gypsum, halite, quartz, sepiolite, sylvite, and talc, participating in the geochemical process and influencing the groundwater chemistry of the study area (Table 4). Most of the samples were observed to be oversaturated with the minerals and have a tendency to precipitate.

Further, the delineation of the solute contribution from the atmosphere or weathering/erosion is evaluated from the ratio of elements to Cl. The Na/Cl and K/Cl molar ratio equivalent to 0.8517 and 0.01767 respectively indicate a contribution from atmospheric precipitation (Zhang et al. 1995). In this case, Na/Cl and K/Cl ratios in the groundwater were observed in the range 1.30–12.09 (Avg. 3.82) and 0.005–0.849 (Avg. 0.127) respectively, significantly higher than those of seawater, suggesting a limited contribution from atmospheric precipitation and major contribution from weathering of minerals. Na/Cl ratio greater than unity also indicates the prevalence of silicate weathering.

The water quality indices for drinking water purposes were computed with the reference value prescribed by BIS (2012) for drinking water and for the values for parameters not mentioned in BIS (2012), the values prescribed by WHO (2017) were considered. The water quality indices of groundwater for drinking purposes ranged from 15.56 to 85.61 and 39.59 to 1,596.66 during pre- and post-monsoon respectively (Table 5). During the pre-monsoon period, 15%, 60%, 15%, and 10% of samples were observed in excellent, good, poor, and very poor categories respectively. The groundwater quality deteriorated during the post-monsoon period with 95% of samples unfit for drinking and only 5% of samples in the good category. The parameters responsible for the poor water quality were both carcinogenic (Ni, Pb, As, Se, U, and Hg) and non-carcinogenic (NO₃, SO₄, NH₄, B, Cu, and Fe). Most of the samples were observed to be unfit for drinking purposes in post-monsoon sampling due to increased U concentration.

Table 5

Water quality indices of groundwater in Bathinda district for drinking purposes

Sample location .	Pre-monsoon .		Post-monsoon .
	.	Water quality status .	.	Water quality status .
B1	39.59	Good	39.59	Good
B2	28.95	Good	712.61	Unfit
B3	29.10	Good	1,102.95	Unfit
B4	45.29	Good	1,596.66	Unfit
B5	85.61	Very poor	1,547.39	Unfit
B6	40.56	Good	593.28	Unfit
B7	46.31	Good	491.20	Unfit
B8	62.68	Poor	410.02	Unfit
B9	44.76	Good	542.67	Unfit
B10	17.49	Excellent	745.72	Unfit
B11	39.57	Good	625.14	Unfit
B12	32.07	Good	375.20	Unfit
B13	81.56	Very poor	467.08	Unfit
B14	61.56	Poor	854.49	Unfit
B15	45.23	Good	789.00	Unfit
B16	34.73	Good	418.17	Unfit
B17	46.89	Good	504.81	Unfit
B18	62.37	Poor	547.72	Unfit
B19	25.26	Excellent	551.50	Unfit
B20	15.56	Excellent	546.33	Unfit

Sample location .	Pre-monsoon .		Post-monsoon .
	.	Water quality status .	.	Water quality status .
B1	39.59	Good	39.59	Good
B2	28.95	Good	712.61	Unfit
B3	29.10	Good	1,102.95	Unfit
B4	45.29	Good	1,596.66	Unfit
B5	85.61	Very poor	1,547.39	Unfit
B6	40.56	Good	593.28	Unfit
B7	46.31	Good	491.20	Unfit
B8	62.68	Poor	410.02	Unfit
B9	44.76	Good	542.67	Unfit
B10	17.49	Excellent	745.72	Unfit
B11	39.57	Good	625.14	Unfit
B12	32.07	Good	375.20	Unfit
B13	81.56	Very poor	467.08	Unfit
B14	61.56	Poor	854.49	Unfit
B15	45.23	Good	789.00	Unfit
B16	34.73	Good	418.17	Unfit
B17	46.89	Good	504.81	Unfit
B18	62.37	Poor	547.72	Unfit
B19	25.26	Excellent	551.50	Unfit
B20	15.56	Excellent	546.33	Unfit

View Large

The suitability of the groundwater for irrigation purposes was evaluated based on the guidelines prescribed by BIS (1986) in relation to electrical conductivity, sodium adsorption ratio (SAR), residual sodium carbonate (RSC), and boron (Table 6). The residual sodium carbonate is the deciding parameter due to which 10% and 15% of samples were observed in the very high hazardous class during pre- and post-monsoon respectively. The high content of residual sodium carbonate in irrigation water impacts the soil texture and structure, and reduces the porosity and permeability of the soil (Rawat et al. 2018; Panneerselvam et al. 2021). The Na% in the groundwater samples ranges from 16.5% to 90.9% with average values 61.85 ± 19.6%. High Na in the irrigation water causes deflocculating and reduces the perviousness of the soils (Pant et al. 2021). The Wilcox plot (1955) indicates that the groundwater of Bathinda district is in the excellent to good (37.5), good (20%), good to permissible (20%), and poor category (22.5%) (Figure 4(a)). The SAR and EC values of groundwater plotted on the US Salinity Laboratory (USSL) diagram (Figure 4(b)) with BIS (1986) recommended values elucidates that most of the samples (65%) were in the C1S1 and C2S1 group indicating low salinity and low to medium alkaline waters that are suitable for irrigation. Some of the samples were in C2S2 (7.5%), C2S3 (12.5%), and C3S2 (5%) indicating medium to high salinity and medium to high alkaline water that should be used for crops tolerant to salinity. Few samples were in the C3S3 and C4S4 groups and irrigating the land with this water should be avoided. From the USSL and Wilcox plot, it can be further concluded that more deterioration in the irrigation water quality was observed in the pre-monsoon period as compared to the post-monsoon period.

In this study, principal component (PC) extraction was done with a minimum acceptable Eigen value of 1 (Kaiser 1958), and the components with an Eigen value <1 were discarded as noise (Yong & Pearce 2013; Ledesma et al. 2015). The priority to define the informative factor was based on the factor loading: strong (>0.75), moderate (0.75–0.50), and weak (<0.50) (Varol & Davraz 2015). In this study, B, Cu, Co, Ni, Pb, Be, As, Se, Hg, U, Cr, pH, EC, F, Cl, NO₃, SO₄, NH₄, K, Ca, Mg, and Na were considered for PCA. The PCA analysis revealed two principal components for pre-monsoon samples and four principal components for post-monsoon samples.

In pre-monsoon, extracted principal component, PC1, with Eigen value >1 account for 82.32% of the total variance with strong loading of NO₃ and NH₄ (Tables 9 and 10). High variance is an indicator of sporadic distribution and anthropogenic origin (Zhang et al. 2016), and the same was observed in this case indicating the source of NO₃ and NH₄ as anthropogenic. The main contributor to these ions are fertilizers, human waste, and so on (Adimalla & Venkatayogi 2017; Adimalla 2020). Component 2, PC2, has 13.25% of covariance of the total with strong loading of SO₄ and Cl, moderate loading of B, and weak loading of Cr, F, Ca, and Mg. The low variance is an indicator of even distribution and originating from natural sources; therefore, SO₄, Cl, Cr, F, Ca, and Mg in the groundwater may have originated from the dissolution of aquifer sediment (Wang & Lu 2011; Zhang et al. 2016).

The principal components PC1, PC2, PC3, and PC4 for the post-monsoon samples account for 45.25%, 27.73%, 15.44%, and 10.46% variance of the total respectively. PC1 was influenced by the strong loading of Cu, Fe, Co, Ni, Pb, Be, As, Se, and Hg, and weak loading of B, U, pH, NO₃, SO₄, and Na. PC2 had strong loading for EC, Cl, SO₄, K, Ca, Mg, and Na, and moderate loading for U and Cr. PC3 had weak loading for B, U, and NH₄. PC4 had strong loading of B and weak loading of Fe, U, and Ca (Tables 9 and 10). The CV of all the PCs was less than 50% indicating even distribution and a geogenic origin. During the post-monsoon period, the anthropogenic pollutants were diluted and, therefore, a reduction in their loading was observed.

The district is occupied by Indo-Gangetic alluvium, the principal aquifer in the study area is alluvium and the major aquifers are Older Alluvium and Aeolian Alluvium. Most of the stratum age is Holocene, having yellowish-brown loose sand with and without kankar and in some locations oxidized silt-clay and micaceous sand sediments in lithose. The sediments typically consist of fine- to medium-grained sand (Anoubam et al. 2016). The calcareous arid brown soil is observed in the east of the Bathinda and siezoram soil white in color with accumulated calcium carbonate is found in the southwest part of the district. The overall soil of the Bathinda district is clay-sandy with a high amount of calcium carbonate (CGWB 2017). Poor fertility is the main problem of this soil requiring the application of a significant amount of fertilizers. The concentration of Fe and SO₄ may be due to the presence of iron sulfite minerals generally present in alluvial sediments (Mattos et al. 2018). The F in the groundwater of Bathinda district may be due to the dissolution of mica minerals in the aquifer sediments (Giri et al. 2021). The Indo-Gangetic plain sediments contain a significant amount of As associated with metal oxides/hydroxides, adsorbed on the mineral surface and/or incorporated in the crystal lattice which is mobilized into the groundwater due to the change in the environmental conditions in the aquifer influenced by both biotic and abiotic factors (Alam et al. 2016; Kumar et al. 2021; Malakar et al. 2022). The presence of uraninite mineral in the aquifer and the development of aerobic/oxidizing conditions due to vertical recharge in the monsoon period may be responsible for the high concentration of U in groundwater during the post-monsoon period (Nolan & Weber 2015; Yadav et al. 2020).

In the study, the groundwater of the cancer-prone Bathinda district was investigated and the groundwater was observed to be alkaline with the dominancy of Na-HCO₃ in pre-monsoon and Na-SO₄/Cl in post-monsoon. The water quality was observed to be more deteriorated in the post-monsoon period due to the anthropogenic inputs and dissolution of minerals. Residual sodium carbonate was observed to be the deciding factor for the non-suitability of the groundwater for irrigation and around 15% of samples were observed to be in a very high hazardous class. The WQI for drinking water usage indicated 15%, 60%, 15%, and 10% of samples were in the excellent, good, poor, and very poor categories respectively during pre-monsoon, which changed to 95% in the unfit category and 5% in the good category during post-monsoon period. The parameters responsible for the poor water quality were both carcinogenic and non-carcinogenic and U was responsible for the poor quality of most of the samples. The order of HQ for the contaminants in the water was U > As > NO₃ > F > B > Se > Cr > Fe > Hg > Cu = Ni > Pb during the pre-monsoon period and U > As > F > NO₃ > Hg > Pb > B > Se > Cr > Ni > Fe > Cu during the post-monsoon period, and the HI value ranges from 6.71 to 238.44 in pre-monsoon period and 8.47 to 239.67 in post-monsoon period, indicating high non-cancer health risk to consumers. The average CR value due to the presence of carcinogenic metals, As, Ni, Cr, and Pb, for males, females, and children was 1643 × 10⁻⁶, 1415 × 10⁻⁶, and 3066 × 10⁻⁶ during pre-monsoon and 2091 × 10⁻⁶, 1802 × 10⁻⁶, and 3904 × 10⁻⁶ during the post-monsoon period respectively. The PCA indicated that NO₃ in the groundwater is of anthropogenic origin and other contaminants were of geogenic origin. Eliminating U, As, NO_3, and F from the groundwater will result in a low health risk category and safeguard the health of people. Although further investigations are required to understand the surface and groundwater interaction through isotope tracers and characterization of biotic and abiotic factors, including irrigation flow, facilitating the mobilization of U, As, Pb, Hg, and Se into the groundwater for the in-situ mitigation, it is certain that the anthropogenic inputs are triggering the mobilization of these elements. Further, the study area practices intensive agriculture necessitating the monitoring of pesticides in the groundwater for evaluating the comprehensive health risk to the consumers. The mutagenicity of water samples through cell line culture study along with epidemiological study is also needed to understand the overall cancer risk to the consumers.

This work has been finically supported by the National Hydrology Project under Purpose Driven Study number NIH-14_2017_24. The authors are highly thankful to the Director, National Institute of Hydrology, Roorkee, Uttarakhand, India for providing all facilities required for this study. We also thank Dr M. K. Sharma, Dr Sandeep Singh, Dr Sandeep Malyan, Ms Meenakshi Rawat, and Mr Rakesh Goyal for their valuable suggestions, encouragement, and support during the field visit and laboratory analysis. The authors express their sincere gratitude to the anonymous reviewers for helping to improve the standard of the manuscript.

This work was supported by the National Hydrology Project sponsored by Ministry of Jal Shakti, Govt. of India (Project No. NIH-14_2017_24) received by R.S. Author K.S. has received research support from Madan Mohan Malaviya University of Technology, Gorakhpur.

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Kaptan Singh: Investigation, writing – original draft. Rajesh Singh: Project administration; conceptualization, funding acquisition, writing – reviewing and editing. Govind Pandey: Writing – reviewing and editing

All relevant data are included in the paper or its Supplementary Information.