A comprehensive assessment of groundwater for seasonal variation in hydro-geochemistry, quality, contamination and human health risk from Deccan Basaltic region, Western India

Groundwater occurrence in hard rock basaltic terrains is restricted to weathered and fractured zones and pockets wherein slow movement of groundwater, prolonged rock-water interactions and higher residence time alter the natural chemistry of groundwater raising water quality issues. The qualitative geochemical analysis, contamination levels and human health risk assessment (HHRA) of groundwater is an integral step in groundwater management in the Deccan Plateau basalt ow region of India. Representative groundwater samples (68) collected from the Shivganga River basin area during pre-monsoon (PRM) and post-monsoon (POM) seasons in 2015 were analyzed for major cations and anions. According to World Health Organization (WHO) EC, total dissolved solids, hardness, bicarbonate, calcium and magnesium surpassed the desirable limit. Boron and uoride content exceeded the prescribed desirable limit of the WHO. The pollution and drinking suitability were assessed by computing pollution index of groundwater (PIG), groundwater quality index (GWQI), and HHRA particularly for boron and uoride toxicity. PIG values inferred that about 6% of groundwater has moderate, 24% has low, and 70% has insignicant pollution in the PRM season; while, only 1 sample (3 %) showed high pollution, 6% showed low, and 91% showed insignicant pollution in the POM season. GWQI results indicate that 27% and 15% samples are within the poor category, and only 15% and 18% of the samples fall within the excellent water quality category in the PRM and the POM season, respectively. Total hazard index (THI) revealed that 88% of children, 59% of adults, and about 38% of infants are exposed to non-carcinogenic risk, as THI values (> 1) were noted for the PRM season; while, 62% of children, 47% of adults and 24% of infants, are vulnerable to non-carcinogenic health hazard during the POM period. the seasonal variation in hydro-geochemistry of groundwater and identify the inuencing parameters altering groundwater quality. 2) To assess the groundwater contamination and degree of pollution level through GWQI and PIG methods and prepared a spatial interpolation maps to understand the seasonal variation in groundwater quality. 3) To evaluate human health risk from groundwater suitability perspective and recognize the processes controlling groundwater composition in the study area. In sum, outcomes of studies like this will provide scientic data on source and history of groundwater contaminants. This information helps governing bodies, water planners and resource managers to develop basin management plans in semi-arid western parts of the DVP. well as for identication and validation of land use land cover study. Also, these location coordinates were used for the preparation of spatial interpolation maps. The potential of Hydrogen ion (pH), electrical conductivity (EC) and total dissolved solid (TDS) were measured on-site using calibrated digital handheld pH, EC and TDS meters. The separate 100 ml of samples were also collected in acidied bottles with 0.5 ml nitric acid for reduced the precipitation of major salt. The major ions such as calcium (Ca 2+ ), magnesium (Mg 2+ ), bicarbonates (HCO 3- ), and chlorides (Cl - ) were analyzed by standard titrimetric methods. Sodium (Na + ) and potassium (K + ) ions were estimated by ame photometric method (ELICO CL 3610). Boron (B 3+ ) and uoride (F - ) were analyzed by HPIC (Dionex make DX-600). Also, nitrate (NO 3- ) and sulphate (SO 4- ) were measured by spectrophotometer (Shimadzu UV-800) as per the standard procedures of water and wastewater analysis (APHA, 2005). The charge balance errors (CBE) were observed within the allowable limit of ± 10% (Berner and Berner 1987). The analyzed results were presented on piper tri-linear diagrams using Aqua-chem 4.0 software to understand the dominant hydro-geochemical facies and Gibbs diagrams were prepared to evaluate the quality regulatory mechanism. The analyzed sample results were compared to the World Health Organization (WHO, 2017) drinking standards for drinking and household purposes. The analyzed samples were also used in the computation of the Groundwater Quality Index (GWQI) and Pollution Index of Groundwater (PIG). To understand the health impact on inhabitants, the human health risk assessment (HHRA) is calculated for oral and dermal exposure pathways for B and F by following USEPA guidelines. MS Excel was used for statistical analysis. excellent water quality, if GWQI value (<50); good, if GWQI values of (50 to 100); poor, if GWQI values (100 to 200); very poor, if GWQI values (200 to 300) and Unsuitable water quality, if GWQI values is (>300). This classication method is used to categorize the groundwater quality in the study area. The calculated GWQI values are ranges from 35.34 to 166.53 and 38.78 to 194.47, for the PRM and the POM seasons, respectively, indicating that groundwater quality is poor to excellent for drinking. This variation may be due to the inputs of domestic and/or agricultural discharges. The POM samples exhibit poor to excellent categories of groundwater for drinking, plausibly due to agriculture return ow causing increase in boron and uoride. The PRM has a maximum 27% samples (numbers 1, 5, 18-20, 23-25, 31) and the POM season has 15% (numbers 16, 20, 23, 29, 33) in poor category. However, only 15% of the PRM samples (numbers 9, 12-14, 16), and 18% of the POM samples (numbers 12-15, 25, 26) represent excellent (Fig. 11 a, b). The central and south parts in the study area have few poor samples in the PRM season. However, in the POM season, local anthropogenic inputs resulted in poor water quality. The overall interpretation is that groundwater quality declines in the PRM season. Sample number 20 exhibited poor water quality in both the seasons and may due the proximity to agricultural eld and brick kiln activities. (< 1); low pollution (1 to 1.5); moderate pollution (1.5 to 2); high pollution (2 to 2.5) and very high pollution (> 2.5). This classication method has been used to categorize the groundwater samples in the study area. The PIG values are varies with 0.44 to 1.87 with an average value of 0.83 in the PRM season; while, 0.32 to 2.06 with (average 0.66) in POM season. In the present study, 34 samples for each season were studied for PIG and out of that, 2 samples (numbers 20, 31) (about 6%) found moderate pollution, 8 samples i.e. 24% (numbers 1, 5, 18-19, 23-25, 28) were identied as low pollution and rest of the 24 samples i.e. (about 70%) are showing insignicant pollution in the PRM season. In the POM season, only 1 sample (number 16) shows high pollution, 2 samples (numbers 20, 33) (about 6%) show low pollution and the rest of the 31 sample i.e. (about 91%) are imply insignicant pollution. This study showed that 18% of the samples are found to be unt by PIG classication. PIG variation maps depict that southern and central parts of the study area comprise the low polluted samples in the PRM season (Fig. 12 a). However, in the POM season, a few samples are found at periphery of the area having low pollution values and only one sample (number 16) showed pollution (Fig. 12 b).


Introduction
Groundwater is the primary source of freshwater for drinking, irrigation and industrial use in most developing nations. In India, during the last ve decades of population growth and economic expansion, the groundwater-based agriculture sector contributed nearly 46% of the gross national product and played an important role in nancial growth of the country (Singh 1983, CGWB 2010. Due to rapid increase in the irrigation-related agricultural activities, groundwater in many parts of the country is under severe stress and resulted depletion in groundwater quality as well as a reduction of groundwater quantity particularly in hard rock terrains Sonkamble et al. 2012; Thomas et al. 2015;Sethy et al. 2016). Therefore, sustainable use of groundwater resources and its protection are of vital importance to public health and the economy of India (Subba ) and in particular, this region of the state of Maharashtra. Generally, the chemical concentration in groundwater is in uenced by natural factors including water-rock contacts, groundwater residence time and ion exchange processes. Anthropogenic factors include agricultural practices using chemicals and fertilizers, industrial waste processes, and mining activities, etc., Adimalla et al. 2018;Simsek and Gunduz 2007). Several researchers have investigated common pollutants in groundwater in the area such as toxic metals such as arsenic /trace elements such as uoride and compounds such as nitrate (Subbarao et al. 2019;Gaikwad et al. 2020a). In recent years some of the pollutants like uoride, boron and nitrate have widely impacted groundwater and cause carcinogenic and non-carcinogenic impacts on human health (Subba Rao 2017; Adimalla et al. 2019a, b;Wagh et al. 2020a;kadam et al. 2019). Human health related risk assessment (HERA) developed by the United States Environmental Protection Agency (USEPA 1980). It is used to evaluate the nature and potential undesirable effects on health of people due to drinking contaminated water (Li et al. 2016;Zhang et al. 2018). Within the Indian state of Maharashtra, cities like Chandrapur, Gadchiorli, Nagpur, and Nanded, for example, are prone to uoride problems. The possible sources of uoride in these regions are leaching uoride containing minerals from granitic exposures in a semi-arid condition with alkaline waters (Kadam et al. 2019;Pandith et al. 2017;Panaskar et al. 2017). Thus, it is a crucial and challenging task for water researchers to detect the origin and occurrence of uoride and their possible health impacts for e cient groundwater resource management.
Groundwater pollution in hard and fractured rock topographies like the Deccan Plateau basalt provinces is a continuous but variable process due to lithologic heterogeneity and comparatively uneven ow of rainwater in fractured and weathered basalt aquifers. There is also the preponderance of lesser amount of subsurface groundwater movement resulting in a high rock-water reaction time.
Moreover, in some isolated areas where large fractures within the basalts do not intersect, the groundwater travels slowly through the aquifer media, unconnected in the phreatic zone to other more regional aquifers. In several studies, under natural situations, this mechanism is favorable for raising the pollutant content due to prolonged rock-water interactions. Furthermore, the geochemistry of groundwater in phreatic aquifers of the hard rock volcanic provinces is potentially vulnerable by anthropogenic inputs such as unrestricted industrialization; metropolitan land ll use and stormwater ow from the agricultural sector  geochemical studies are necessary to distinguish the discrepancies in ionic concentrations for utilization in various sectors; therefore, various hydro-geochemical methods were exercised to evaluate groundwater composition in the respective regions (Marghade et al. 2011;Ledesma-ruiz et al. 2016). Accordingly, several new studies have been included for the interpretation of the cations, anions and heavy/trace metals in the groundwater, to distinct the natural and anthropogenic origins that alter groundwater quality and their associations within the aquifer system (Kumar et al. 2006;Tian et al. 2015;Brindha et al. 2017;Wagh et al. 2018). The previous works carried out in the Deccan Volcanic Provinces -Western Ghat (DVP-WG) have mainly focused on the hydro-geochemical evolution in the groundwater Vincy et al. 2015;Wagh et al. 2019b). However, several studies were addressed the crisis of groundwater pollution due to agriculture activities, industrialization, and urban growth in DVP-WG region (Pawar and Shaikh 1995;Wagh et al. 2016 a, b).
In the case of the Shivganga River basin, studies related to the suitability assessment and hydro-geochemistry of groundwater is relatively sparse until recently. Usually, in the alluvium plain, the shallow aquifers are more susceptible to alteration than the deeper basaltic aquifers because of high transmissivity and porosity of soil and rocks. In the studied region, dug wells commonly provide water for domestic and irrigation needs; therefore, the groundwater quality is intimately linked with local inhabitants. Therefore, frequent monitoring and suitability appraisal of groundwater quality is an essential to avoid further human health deterioration in portions of the study area having elevated pollution levels. Hydro-geochemical studies are required to recognize the mechanism of natural and anthropogenic processes concerned with alteration of groundwater quality. In ltrated recharged water interacts with soil including physical, chemical, biological processes, and mineral dissolution takes place, impacting the chemical constituents of the groundwater. In general, in this study, hydro-geochemical processes which are mainly accountable for changing the chemical composition of groundwater varied with respect to time and space. Therefore, the objectives of the study are 1) To assess the seasonal variation in hydro-geochemistry of groundwater and identify the in uencing parameters altering groundwater quality. 2) To assess the groundwater contamination and degree of pollution level through GWQI and PIG methods and prepared a spatial interpolation maps to understand the seasonal variation in groundwater quality. 3) To evaluate human health risk from groundwater suitability perspective and recognize the processes controlling groundwater composition in the study area. In sum, outcomes of studies like this will provide scienti c data on source and history of groundwater contaminants. This information helps governing bodies, water planners and resource managers to develop basin management plans in semi-arid western parts of the DVP.

Geo-environmental outline
The Shivganga River basin (latitudes 18°13'36" and 18°24'7; longitudes 73°44'1"and 73°56'17") ows within a small basin of about 176 km 2 , situated on the easterly sloping face of the Western Ghat region, Maharashtra state, India. The area is drained by a fth order, 27 km long Shivganga River (Fig. 1). The study area is encompassing undulating topography with elevated hill ranges originates from Deccan Trap basaltic ows; the highest peak being 1316 m above mean sea level (amsl) and lowest at mouth of the river 590 m amsl.
The area represents a tropical monsoon climate and obtains yearly rainfall of ~900 mm from the southwest monsoon in the month of June and September. Also, a wide range of temperature variation occurs in winter (10°C) and summer (39°C). The two main cropping seasons with irrigation-based agriculture Kharif (July-October) and Rabi (October-January), are assigned to rice paddies and jowar (sorghum) elds for the major crop cultivation in the area. The study area shows mainly ve land use and land cover classes: irrigated crop land (45.14%), residential and paved land (6.38%), surface water reservoir (0.57%), forest and plantation land (27.75%), and barren land (19.16%) (Kadam et al. 2018). Forest cover is observed at high elevation, in the peripheral areas representing river catchment. Due to major consumption of groundwater for agricultural and domestic needs, groundwater levels are going down very quickly in the area and are not rebounding. Also, the study area is restraining growth of 58 large villages having around 0.07 million of population (Census of India 2011). This lack of available water resources exerts pressure on excess groundwater pumping, resulting in overexploitation and contamination of groundwater in study area (Kadam et al. 2018). The study region consists of little primary porosity in the form of basalt vesicles (air bubbles) which do not support transmission of groundwater as they are not interconnected. Hence, the groundwater potential in this area is primarily governed by permeability and increased of secondary porosity by compressional, tensional forces and rapid cooling of magma leading to weathering and fracturing in the basaltic terrain. In the area, the depth of weathered rocks is (1-16m) and fractured rocks (12-60m) below ground level (CGWB 2013). The weathered zone comprises black cotton alluvium soil derived from basalts and clayey soil derived from red inter ow horizon, which is a break between two successive lava ows. As clay is characterized by micro porosity, the in ltration capacity of water is negligible, which in turn retards groundwater movement. So, in the weathered zone, groundwater occurs in water table phase and fractured/jointed aquifers under semi-con ned conditions. The water table depth varies from less than 4m to more than 12m below ground level (bgl), based on the varied topographic features. As per the cross-section of the hydrogeological conditions drawn in the upstream part to downstream part side based on the dug well section, bore well drilling and previous unpublished work as well as geophysical survey work are presented for this study area (Fig. 2), the depth of surface soil changes from 1m to 5m from the ground surface from the peripheral part of sample number 12 to the mouth of the river at sample number 27. Subsequently, the depth of highly to moderately weathering host rock, which primarily decreases the permeability, the weathering thickness of the rock ranges from 1m to 24m. This condition is categorized by low porousness owing to isolated minute voids in the weathering basalt rock. The lack of connected porosity restricts liquid movements inside these rock layers. The next layer down is the fractured basalt having depth ranges between 10 m to 35 m, which is characterized by high porousness, since there are continuously connected of the pore spaces in the rock fractured zone that allow the groundwater movement to ow easily from one point to next point. The lithology in this zone is shows vesicles in the form of primary porosity which acts as a storage for the groundwater. As these vesicles are not interconnected, the ow and movement of groundwater is partially restricted. The compact massive basalt is occurred beneath the vesicular basalts and lacking in capable of water storage and block the movement of groundwater. The above lithologic characteristics are generally observed in the aquifers in the study area.

Material And Methods
The survey of India (SOI), topographic sheets (numbers 47F/15 and 47F/16) on 1.50,000 scales has been used for preparation of a base map of the study area including major features such as the road network, streams, and settlements. In the study area, a groundwater comprehensive assessment was performed using a collection of sixty-eight dug well water samples in the year of 2015 during pre-monsoon and post-monsoon seasons (May and November) with respect to land use, lithology, landform and use in drinking based on random sampling method. The groundwater has been sampled in pre-washed half liter polytene bottles and kept in temperature below 4 0 C in the laboratory to avoid further reaction. The coordinates of each sampling location were collected by GARMIN GPS for preparation of base map as well as for identi cation and validation of land use land cover study. Also, these location coordinates were used for the preparation of spatial interpolation maps. The potential of Hydrogen ion (pH), electrical conductivity (EC) and total dissolved solid (TDS) were measured on-site using calibrated digital handheld pH, EC and TDS meters. The separate 100 ml of samples were also collected in acidi ed bottles with 0.5 ml nitric acid for reduced the precipitation of major salt. The major ions such as calcium (Ca 2+ ), magnesium (Mg 2+ ), bicarbonates (HCO 3 -), and chlorides (Cl -) were analyzed by standard titrimetric methods. Sodium (Na + ) and potassium (K + ) ions were estimated by ame photometric method (ELICO CL 3610). Boron (B 3+ ) and uoride (F -) were analyzed by HPIC (Dionex make DX-600). Also, nitrate (NO 3 -) and sulphate (SO 4 -) were measured by spectrophotometer (Shimadzu UV-800) as per the standard procedures of water and wastewater analysis (APHA, 2005). The charge balance errors (CBE) were observed within the allowable limit of ± 10% (Berner and Berner 1987). The analyzed results were presented on piper tri-linear diagrams using Aqua-chem 4.0 software to understand the dominant hydro-geochemical facies and Gibbs diagrams were prepared to evaluate the quality regulatory mechanism. The analyzed sample results were compared to the World Health Organization (WHO, 2017) drinking standards for drinking and household purposes. The analyzed samples were also used in the computation of the Groundwater Quality Index (GWQI) and Pollution Index of Groundwater (PIG). To understand the health impact on inhabitants, the human health risk assessment (HHRA) is calculated for oral and dermal exposure pathways for B and F by following USEPA guidelines. MS Excel was used for statistical analysis.

Computation of Groundwater quality index (GWQI)
GWQI is widely used technique to categories the groundwater quality as: excellent, good, poor, very poor and unsuitable for drinking. It is based on the rank and weights given to the analyses parameter and it is one of the most trusted indices for the quality assessment of groundwater. For the comprehensive assessment of groundwater quality and suitability, the water quality parameters such as pH, TDS, Ca ++ , Mg ++ , TH, Na + , K + , HCO 3 -, Cl -, SO 4 -, PO 4 -, NO 3 -, B 3+ and Fwere considered. The GWQI calculated by the following steps.
Step 1: Allotted the weight (AW) to each analyzed parameter considering its importance to overall body growth (Table 1); Step 2: computation of relative weight (RW) for each water quality parameter (Eq. 1).
Where, W is sum of all allotted weight; n is number of groundwater quality parameter Step 3: Computation of quality rating scale (QRS) of each groundwater quality parameter (Eq. 2) QRS = (CP/ SWHO) × 100 (2) Where, the CP stands for the content of each parameter in groundwater and SWHO stands for standard limit of the WHO of the respective parameters.
Step 4: The calculation of the GWQI, which the summation of sub-index (Sbi), Where, Sbi is the multiplication of relative weight (RW) by quality rating scale (QRS) of each groundwater quality parameters (Eqs. 3 & 4), (Table 1).

Computation of pollution index of groundwater (PIG)
PIG is a numerical expression for rating the quantifying range of contamination by considering parameters such as pH, TDS, major cations and anions, boron and uoride based on their relative importance in de ning groundwater quality (Rao and Chaudhary 2019; Wagh et al., 2020b). PIG was based on considering WHO drinking standards; also, computed by following the methodology proposed by Subba Rao (2012).
To generate the index values, rstly a relative weight (Rw) was assigned to each of the water variables. The Rw values ranges between 1 to 5; where, 1 is having the least importance in health risk due to low pollution; while, 5 having the highest importance for health risk due to high pollution. Potassium has an assigned Rw value of 1 and calcium and magnesium has an assigned Rw value of 2. Bicarbonate has an assigned Rw value of 3. Chloride has an assigned RW value of 4, and the assigned maximum values of 5 are for pH, TDS, boron and uoride ( Table 2).
The weight parameter (Wp) is calculated by the ratio of relative weight (Rw) of each parameter to the sum of all relative weight, presented by following equation (5): Further, statues of concentration (Sc) are calculated with equation (6): Where, C is the concentration in each groundwater samples; Ds is drinking water quality standard.
The overall water (Ow) quality is multiplication function of weight parameter (Wp) with statues of concentration (Eq. 7): Finally, PIG is calculated by summation of overall water quality (Ow) computed by equation (8): HHRA is computed based on daily intake, dermal contact and inhalation (He et al. 2020). Moreover, consumption of water with elevated boron and uoride concentrations may result a non-carcinogenic risk to inhabitants. Thus, three age groups were considered: infants 6 months (0.5 years); children up to age 6 year and adults (≥16 years) (Kadam et al. 2019;Narsimha and Rajitha 2018).
The value of estimated daily intake (EDI) of the above age groups for F and B content in groundwater was computed by Eq. (9) (USEPA, 1989;Zango, et al. 2019). (F, B) x Cd}/ Bw (9) Where, EDI having unit mg/kg/day; C (F, B) is content of F or B in groundwater; Cd indicate every day average ingestion of water and Bw denotes body weight. Cd value for adult is 3L/day; children 1.5 L/day and infants 0.250 L/day (Vetrimurugan et al. 2013). However, Bw for adult is 57.5 kg; children 18.7 kg and infants 6.9 kg ICMR Expert Group (1990) The hazard quotient (HQ) was determined for groundwater by F and B exposure to individuals was projected from Eq. (10) (USEPA, 1989), and it is the ratio of EDI to the reference dose (Rfd) HQ = EDI /Rfd (10) Where, the RfD value of F (0.06 mg/kg/day) was considered from the US Environmental Protection Agency (USEPA, 2014) guideline; while, the RfD of B (0.13 mg/kg/day) was obtained from WHO (2009).

EDI = {C
Finally, Total Hazard Index (THI) calculated for the human health risk of F and B; it is summation of hazard quotient of uoride plus hazard quotient of B.

Results And Discussion
Seasonal variation in hydro-geochemistry The statistical summary of physicochemical parameters for the pre-monsoon (PRM) and post-monsoon (POM) seasons of year 2015 and its comparison with the WHO drinking standards (2017) is illustrated (Table 1 and Table 2). The measured pH value in PRM season ranges from 7.44 to 8.38 due to the bicarbonate form of liqui ed carbonate in water. However, during POM season it varies from 6.85 to 7.51 showing the natural range due to rainwater dilution of rainwater alkalinity. Accordingly, groundwater shows moderately alkaline nature in both seasons of the study area. The groundwater samples of both seasons are within the threshold limit (6.5-8.5) of the WHO (2017) drinking norms (Table 2). Groundwater had a slight decrease in pH from PRM to POM season which indicates a good rock-water interface. The electrical conductivity (EC) values are within the ranges of 296 to 1070 µS/cm (avg. 635µS/cm) and 240 to 980 µS/cm (avg. 589µS/cm) in the PRM and POM groundwater samples respectively (Table 1). EC was increased during the PRM season may be due to the evaporation of soil moisture from the phreatic zone as well as prolonged rockwater reaction and manmade contamination by concentration, causing increases in ionic content ). According to WHO standards, 80 and 68 percent samples are higher than the recommended range of 500μS/cm for both the seasons (Table 2). TDS values representing total positive and negative ionic contents in water varies from 196 to 832 mg/L (avg. 404 mg/L) and 164 to 591 mg/L (avg. 383 mg/L) in PRM and POM seasons, respectively (Table 1). As compared to WHO standards, desirable limit (DL) is 500 mg/L of TDS, (24%) in PRM and (15%) in POM samples are above the DL (Table 2). Also, the high content of TDS is possibly related to root exudation of aquifer media salts from the surface soil and certain manmade actions (Mukate et al. 2019). However, during the POM season, the concentration of these parameters has decreased due to dilution by fresh rainwater recharge in the aquifer system. High content of EC and TDS were detected in the groundwater samples in the lower reach of the study area owing to accumulation of salt from agricultural activities. Total hardness (TH) content ranges from 52 to 604 mg/L with (average 234 mg/L), and as per WHO standards, 95% samples above DL of (500 mg/L) in the PRM season. TH content for the POM season varied from 96 to 384 mg/L with average value of (248 mg/L); 97% samples were detected above the DL ( Table 2). The standard deviation values of TH were also very high in both seasons representing the local effect on the groundwater quality which re ected in the minimum and maximum value of groundwater samples. Sample number 5 had a TH content over the PL threshold (>500 mg/L), due to the salt deposition on the inner lining on the well due the long rock-water interaction.
TH values obtained in groundwater were divided in four classes following Sawyer and McCarty (1967) Figure 3a, b.
In the study area, the POM samples represent the cation dominance in decreasing order of Ca 2+ > Mg 2+ > Na + > K + due to dissolution of aquifer minerals with rainwater. Whereas, Ca 2+ > Na + > Mg 2+ > K + in the PRM seasons are the result of evaporation dominance, anthropogenic inputs and irrigation practices. The Ca 2+ contents are found within the range of 8 to 120 mg/L averaging of 50 mg/L and 13 to 90 mg/L (average 46 mg/L) in both PRM and POM periods, respectively (Table 1). The calcium content varies with the monsoon; if rainfall decreases in the pre-monsoon season, the calcium concentration increased signi cantly. Thus, ~ 18% samples of the PRM season and only 6% samples in the POM period are beyond the DL (75 mg/L) of the WHO (Table 2). Moreover, Mg 2+ occurred in ranges of 8 to 74 mg/L (average 26 mg/L); and 3 to 55 mg/L (average 23 mg/L) in the PRM and the POM seasons, respectively (Table 1). High content of Mg 2+ is most probably due to the higher rate of irrigation return ow, which increases the dissolution of evaporite minerals and subsequently increase concentration of magnesium in groundwater within the shallow aquifers especially, in low-lying regions of the watershed (Haritash et al. 2008). As per WHO drinking standards, all the samples are suitable for drinking; however, only 3% samples surpass the DL in the PRM and the POM seasons (Table 2). Also, diminutive content of Na found with an average value of 27 mg/L and 14 mg/L in the PRM and the POM seasons (Table 1). As compared to other common cationic constituents occurring in water, the potassium concentration is low due to the high resistance of this element in the clay mineral structure (Srinivas et al. 2017). Generally, potassium content in natural hydrological cycle varies from 0.1 ppm in rainwater to a few ppm in surface water and groundwater (Matthess, 1982). In the area, average K + value is 0.93 mg/L and 0.99 mg/L in the premonsoon and the post monsoon seasons (Table 1). It is observed that all the samples having content of sodium and potassium are within threshold limit of the WHO.
The anion abundance was observed in order of HCO 3 -> Cl -> SO 4 2-> NO 3 -> PO 4 2in the pre-monsoon season; while, in the postmonsoon season HCO 3 -> SO 4 2-> Cl -> NO 3 -> PO 4 2-. It is observed that HCO 3 content varies from 30 to 320 mg/L with an average value of (168 mg/L) and 100 to 360 mg/L (average 240 mg/L) in the PRM and the POM seasons (Table 1). Moreover, it is inferred that due to alkaline condition of water, carbonate species be present in the form of bicarbonate. Results indicated that 30% and 76% of groundwater samples exceeded the permissible limit of the WHO in the PRM and the POM seasons, respectively ( Table 2). The elevated content of HCO 3 in few groundwater samples is due to agricultural runoff as well as from basaltic host rock (Locsey and Cox, 2003). SO 4 2content varies with ranges of 12 to 151 mg/L with an average value of (4.17 mg/L) and 10 to 85 mg/L (avg. 32.74) in the PRM and the POM seasons, respectively. The excessive concentration of SO 4 2in the POM season is due to the addition of manmade as activities involving detergents and fertilizers. Chloride content varies from 19 to 249 mg/L and 4 to 62 mg/L with an average values of (78 mg/L and 17 mg/L) in the PRM and the POM seasons, respectively. It is pragmatic that high content of chloride present in the PRM season; however, low chloride content in the POM season is due to dissolution phenomenon. High chloride content is attributed to secondary sources like domestic sewage including human fecal material, decomposition of carbon-based substances and agrarian surface ow Kumar et al. 2008). NO 3 content ranges of 6 to 19 mg/L (avg. 13 mg/L), and 0.12 and 14 mg/L (avg. 3 mg/L) during the PRM and the POM seasons, respectively (Table 1). According to WHO speci cations, all groundwater samples are suitable for drinking. PO 4 2values vary within ranges of 0.02 to 0.32 mg/L (avg. 0.05 mg/L) in the PRM season and below detection limit to 2.90 mg/L (avg. 0.11 mg/L) in the POM season (Table 1). However, high content in the POM season is due to agricultural return ow from the irrigation elds (Vetrimurugan et al. 2013).
Boron content varies from 0.08 to 12.45 mg/L (avg. 3.97 mg/L) during the PRM season; conversely, 0.34 to 14.33 mg/L with (avg. 3.15 mg/L) in the POM season (Table 1). As compared with WHO standards (Table 2), 70% and 51% of samples exceed the DL (0.5 mg/L) for the PRM and the POM seasons, respectively; while, 58% and 39% of the samples are above the PL (1 mg/L) in the PRM and the POM seasons, respectively. The spatial variation maps (Fig. 6a, b) demonstrate that the PRM season samples have higher concentrations of boron as compared to the POM season samples. The possible source of boron in groundwater is from the rockwater interaction, sewage e uent and fertilizer application (Bhat et al. 2018). The high concentration of boron in study area is due to agricultural runoff, excessive use of herbicides and manures, inputs from poultry farming and animal excreta (kadam et al. 2019). The surplus content of Boron is toxic to human health and crops; it also reduces the soil productivity (Ahmad et al. 2012;Buszka et al. 2007;USEPA 2008).
Generally, uoride is one of the primary trace element in sub-surface water, which is essential for human health; but, when it exceeds the allowable content, uoride poses serious human health hazards (Kale and Pawar 2017;Nag 2017;Panaskar et al. 2017). The host rock with F bearing minerals acts as a source and is accountable for the elevated F content groundwater (Kale et al. 2010). F content varies from 0.10 to 1.84 mg/L with an average value of (0.84 mg/L) during the PRM season; while, 0.02 to 1.48 mg/L (avg. 0.63 mg/L) in the POM season (Table 1). As per the WHO drinking speci cations (Table 2), uoride content in the PRM (35%) and the POM (18%) seasons showed groundwater samples above the DL (1 mg/L). However, only (6%) of the groundwater samples exceed the PL (1.5 mg/L) in the PRM season; thus, elevated F restricts drinking water use in two locations (numbers 13 and 28). However, those sampling locations are beyond the PL (1.5 mg/L) for F and may lead to dental uorosis and skeletal deformities in the study area; so, groundwater in those locations is un t for human consumption. The elevated content of F in the PRM season is due to semi-arid condition with evaporation, alkaline water in the study area is more favorable for dissolution of uorite mineral (Chen et al. 2017;Kadam et al. 2019). Also, uoride-rich minerals like uorite, muscovite, biotite, topaz, apatite, and hornblende, from host rocks in the area and are the possible sources of leaching of F ion into the groundwater Narsimha and Li, 2019).
Based on equation (1), when groundwater charged with CO 2 reacts with biotite minerals, the ions such as K, Mg, HCO 3 , F and SiO 2 enter into the groundwater from aquifer matrix. Thus, the high bicarbonate leached from aquifer matrix in groundwater facilitates release of large concentrations of uoride ions from the host rock into groundwater. The scatter plot of Mg+K vs HCO 3 +F (Fig. 4) shows noteworthy positive correlation (r= 0.54) con rming the process in (Eq.1). Based on the (Eq. 2) the reaction suggested that rock mineral deposits of hornblende with subsurface water (H 2 O) and atmospheric CO 2 reacts with ions such as calcium, magnesium, sodium and bicarbonate. F and SiO 2 are released into groundwater from the host rock. The Scatter plot of Ca+Mg+Na vs HCO 3 +F (Fig.   5) shows strong positive correlation (r= 0.72) which suggests that the weathering of hornblende mineral is related to the increased content of ions including uoride ions in groundwater. Both equations 1 and 2 show that Deccan Plateau rain water is highly alkaline, having a pH above 7.5. Generally, rainwater leaches surface minerals and enters the subsurface as groundwater and carrying a high content of bicarbonate and sodium ions which released comparatively more hydroxyl ions in groundwater. A further exchange of ions occurs in the subsurface and results in the creation of favorable conditions to leaches uoride ions from host rocks into groundwater.
The silica in the groundwater also increases with leaching of uoride. Hornblende mineral dissociates in Na and HCO 3 -rich groundwater which in turn enables leaching of uoride from aquifer matrix. The use of agricultural fertilizers, insecticides, domestic waste water and high withdrawal of subsurface water are another source of uoride pollution in the aquifer system (EPA, 1997). The spatial variation maps (Fig. 7a, b) show that the PRM season has wide dispersion of elevated F concentrations with comparison of the POM season samples, having less. Furthermore, low concentrations of the F ion are observed at the upper reaches of the study area where, high precipitation resulted into high dissolution rates as well as a high dilution of the groundwater near the host basaltic geology (Kale and Pawar, 2017). The downstream part of the study area has low rainfall but the presence of arable farmland results in intense agriculture with signi cant use of fertilizers. The fertilizers are the main cause of high concentration of F ion in that portion of the study area.

Sources of ions in the groundwater
Generally, a number of factors like natural processes, anthropogenic factors, geology and mineral composition of the area, types of weathering etc are responsible for determining groundwater quality. Thus it is crucial to recognize the positive and negative assimilation within cation and anions and their combined in uence on overall water quality . The ionic plot of Ca + Mg vs HCO 3 and Mg vs HCO 3 (Fig. 8 a, b) represents a signi cant association which is indicative for weathering of olivine and pyroxene minerals (Ca + Mg-HCO 3 ) from host rock. The plot of Ca + Na vs HCO 3 (Fig. 8c) illustrates a positive association in both seasons which indicate some contribution is due to plagioclase dissolution. However, in the post-monsoon season, the Ca + Mg vs HCO 3 association are high due to greater rock-water contacts ).
Ca vs HCO 3 (Fig. 8d) plot demonstrated a good correlation and is generally used to point out calcium sources in the groundwater. The Ca is attributed to dissolution of minerals like calcite and dolomite from carbonate weathering. The geochemical plots of Ca + Mg: Cl + SO 4 (Fig. 8e) signify a positive relationship, suggestive that the groundwater may have a preference for ion pairs formation Gaikwad et al. 2020b). The plot of Ca vs Mg (Fig. 8f) indicates that about 80% of the samples are having a Ca/Mg ratio between 1 and 2, which is evidence for calcite as the main mineral (Subramani et al. 2010). Na vs HCO 3 plot indicates that majority of samples shifting towards HCO 3 representing groundwater derived from basalt rocks (Fig. 8g). The plot shows a good correlation between Na + K vs Cl + SO 4 (Fig. 8h) , representing anthropogenic pollution which appears to alter the natural groundwater composition. The bivariate plot of Ca/Na: HCO 3 /Na (Fig. 8i) signi es that these ions are attributed to weathering of a silicate mineral, due to the high solubility of Na over Ca, increasing the sodium content ). In the study area, there is little information available on groundwater geochemistry and their in uence factors. However, Pawar et al. (2008) reported that the aquifers are exemplify as basalt, weathered basalt and doleritic dyke restraining the Ca+Mg-HCO 3 water type; whereas, aquifers from alluvial parts of the study area are distinguished by the Ca+Mg+Na-HCO 3 type. The inputs of ions are mainly attributed to the weathering of silicates minerals like (olivine, augite and plagioclase feldspar), and there is also a slight input from zeolites ). Groundwater quality varies in the region, due to rainwater is charged with Cl, SO 4 , NO 3 , Na, Ca, Mg, and small quantities of HCO 3 (Das et al. 2005). The precipitation in uence in the groundwater was eliminated from the acquired geochemical data and recti ed values were considered for the analyses.

Hydro-geochemical facies
Piper's trilinear diagram was used to recognize the geochemical progression in the groundwater of study region (Piper, 1944). The plot represents that mixed water type Ca-Mg-Cl is dominant in most of the groundwater samples from both the seasons (Fig. 9). The plot exempli es that, alkaline earths (Ca 2+ and Mg 2+ ) signi cantly go beyond the alkalis (Na + and K + ); weak acids (HCO 3 − and CO 3 2− ) and exceed the strong acids (Cl − and SO 4 2− ). The POM season shows the Ca+Mg-HCO 3 water type, indicating that wells are present in weathered basaltic, and dolerite dyke aquifers ). The PRM samples show Ca -HCO 3 as well as mixed Ca + Na -HCO 3 water types that correspond to host basaltic rock lithology and anthropogenic inputs ). Gibb's diagram is used to evaluate processes like precipitation, rock and evaporation dominance which control the groundwater composition in the aquifer (Gibbs, 1970). It is inferred that rock dominance processes in uence the groundwater quality in the studied region (Fig. 10a, b).

Groundwater quality index (GWQI)
GWQI is a widely used technique to categorize the groundwater quality as: excellent, good, poor, very poor and unsuitable for drinking (Subba Rao et al. 2019;2020). It uses the rank and weights given to the analyses parameter to calculate the groundwater quality, as it is one of the most trusted indices for the quality assessment of groundwater. GWQI values were further classi ed into 5 categories: excellent water quality, if GWQI value (<50); good, if GWQI values of (50 to 100); poor, if GWQI values (100 to 200); very poor, if GWQI values (200 to 300) and Unsuitable water quality, if GWQI values is (>300). This classi cation method is used to categorize the groundwater quality in the study area. The calculated GWQI values are ranges from 35.34 to 166.53 and 38.78 to 194.47, for the PRM and the POM seasons, respectively, indicating that groundwater quality is poor to excellent for drinking. This variation may be due to the inputs of domestic and/or agricultural discharges. The POM samples exhibit poor to excellent categories of groundwater for drinking, plausibly due to agriculture return ow causing increase in boron and uoride. The PRM has a maximum 27% samples (numbers 1, 5, 18-20, 23-25, 31) and the POM season has 15% (numbers 16, 20, 23, 29, 33) in poor category. However, only 15% of the PRM samples (numbers 9, 12-14, 16), and 18% of the POM samples (numbers 12-15, 25, 26) represent excellent (Fig. 11 a, b). The central and south parts in the study area have few poor samples in the PRM season. However, in the POM season, local anthropogenic inputs resulted in poor water quality. The overall interpretation is that groundwater quality declines in the PRM season. Sample number 20 exhibited poor water quality in both the seasons and may due the proximity to agricultural eld and brick kiln activities.

Pollution index of groundwater (PIG)
PIG is a numerical expression for rating the quantifying range of contamination by considering numerous parameters based on relative importance in de ning groundwater quality such as pH, TDS, major cations and anions, boron and uoride. Several authors proposed PIG to derive the extent of contamination of surface and/or groundwater (Rao and Chaudhary 2019;Wagh et al., 2020b;Egbueri, 2020;Marghade et al., 2020). The PIG values were further classi ed into 5 classes namely, insigni cant pollution if PIG value (< 1); low pollution (1 to 1.5); moderate pollution (1.5 to 2); high pollution (2 to 2.5) and very high pollution (> 2.5). This classi cation method has been used to categorize the groundwater samples in the study area. showed that 18% of the samples are found to be un t by PIG classi cation. PIG variation maps depict that southern and central parts of the study area comprise the low polluted samples in the PRM season (Fig. 12 a). However, in the POM season, a few samples are found at periphery of the area having low pollution values and only one sample (number 16) showed pollution (Fig. 12 b).

Human Health risk assessment (HHRA) of uoride and boron
In the study area, F content in groundwater above the DL (35 and 18%) in the PRM and the POM seasons and (6%) exceed the PL of WHO standards in the PRM season; thereby, restricting drinking water use at a few locations. B content in (70 and 51%) surpasses the DL and 58 and 39% samples exceed the PL of the WHO in the PRM and the POM seasons, respectively. Therefore, to ascertain the human health risk, B and F was considered. The EDI values of F for different age groups varies for inhabitants show that infants (less than 0.5 years) are having high ingestion groundwater with high F content as compared with children and adults as the body mass will be less for children. It is exemplify that (41%) of the samples in the PRM season and (48%) of the samples in the POM season are above the safe limit of F, 0.03 mg/kg/day for infants. The safe EDI value for children is 0.13 mg/kg/day for F, where, all samples are below than that, except 1 sample in the PRM season. While, adult has a safe limit of 0.05 mg/kg/day; 29% of the samples in the PRM season and 12% of the samples in the POM season exceeded this limit. If, HQ value of F greater than 1 signi es that people are exposed to non-carcinogenic health risks associated with high F content in the drinking water (Table 3 and 4). In the study area, HQ values of F varies from 0.09 to 1.60; 0.13 to 2.46 and 0.06 to 1.11 in the PRM season and 0.02 to 5.75, 0.03 to 8.84 and 0.01 to 3.99 in the POM season for adult, children and infants, respectively. The children age group shows more risk than infants and adults. The probable cause for the HHRA for children is the modest body weight (Bw) as comparison with other age groups (Zango et al., 2019).
HHRA results con rm that non carcinogenic health risk of F is in order of children > adults > infants. The children age group is highly vulnerable to potential of dental and skeletal deformities in the future.
In the study area, EDI value of B ranges from 0.0 to 0.65, 0.0 to 1.00 and 0.0 to 0.45 for adult, children and infants, respectively in the PRM season (Table 3). Moreover, 0.0 to 0.75, 0.0 to 1.15 and 0.0 to 0.52 for adult, children and infants, respectively in the POM season (Table 4). The results show that the POM season has a lower health risk as B concentration decreases due to dilution phenomenon in the POM season as compared with the PRM season. Infants and children are having high ingestion rates of groundwater with high B content as compared with adults as the body mass of the younger population will be less. Also, 68% samples from the PRM season and 53% from the POM season are above the safe limit of 0.01 mg/kg/day of B for infants (Table 3 and 4). The PRM season (38%) and the POM season (32%) samples are having values higher than that safe EDI value of (0.16 mg/kg/day) for children in both seasons. If, HQ value for B is greater than 1; signifying that the people are more vulnerable to non-carcinogenic health related problems (Table 3 and 8). The HQ values ranges from 0.00 to 5.00, 0.0 to 7.68 and 0.0 to 3.47 in the PRM season for adult, children and infants, respectively (Table 3). Although, HQ values in the POM season varies from 0.00 to 5.75, 0.0 to 8.84 and 0.0 to 3.99 for adult, children and infants, respectively (Table 4). The highest HQ values were found in children, compared to lower value associated with adults and infants.
Total hazard index (THI), is computed from summation of the hazard quotient (HQs) of B and F; if, THI is more than 1 it means, the water probably caused health problems in relation to a non carcinogenic risk; while, if the value is less than 1 it indicates there are no symptoms of non-carcinogenic risk (USEPA 2014). THI value was calculated separately for the adult, children and infants, given in the Tables 3 and 4. THI values ranged from 0.21 to 6.13 (avg. 1.86), 0.27 to 8.89 (avg. 2.55) and 0.14 to 4.25 (avg. 1.29) for adults, children and infants, respectively in the PRM season (Table 3). Children (88%), adults (59%) and infants (38%) possess noncarcinogenic risk as THI values (>1) in the PRM season (Fig. 13) children, 47% adults and 24% of infants are possessed to non-carcinogenic risk in the POM season (Fig. 13). It is shown that children face higher non-carcinogenic health risk than infants and adults and small body mass compared to adults.
The spatial variation maps of THI for infants in the PRM and POM seasons are shown (Fig.14 a, b). Figure 14a shows the complete northern part of the study area is identi ed under the no risk category; however, central and south parts of the study area fall in the higher risk category (except sample numbers 22, 28, 31). Figure 14b shows that in POM season, the locations for sample numbers 2, 16,20,23,24,30,31 and 34 have a potential for a health risk. It is inferred that infants are more vulnerable to health risk in the PRM season. The spatial extent of THI for children ( Fig. 15a and b) demonstrate that, in the PRM season (samples number 10, 15, 31) and the POM season (numbers 6, 12, 14, 21, 26, 29,31,32) are con rm their tness for drinking and remaining samples from both seasons are un t for drinking use. For adult risk, many of the samples showed a potential for exposed risk except sample numbers (4, 9, 10, 15, 22, 28, 31, and 34) in the PRM season. However, in the POM season the samples (numbers 4, 8, 12-15, 21, 25, 28, 29, 31, 32) fall in no risk category; while, other samples showed potential health risk for adults ( Figs. 16a and b). The POM season groundwater quality is comparatively better than the PRM season. HHRA results showed that children have a greater non-carcinogenic risk than adults and infants in both seasons.

Conclusions
The study is summarized with following conclusions, inferences drawn from hydro-chemical analysis, GWQI, PIG and HHRA from the Shivganga River basin of Western India. Hydro-chemical interpretation reveals that the groundwater quality is slightly alkaline with hard to very hard water types. As per the WHO drinking standards, the parameters like pH, Na + , K + , Cl − , SO 4 − and NO 3 − are within threshold limits. Besides, EC (80 and 68%), TDS (24 and 15%), TH (95 and 97%), Ca 2+ (18 and 6%), Mg 2+ (3%), HCO 3 − (30 and 76%) surpass the desirable limits in the PRM and the POM seasons, respectively. High content of EC and TDS were observed in the downstream part of the study area due to accrual of salt. TH is increased in the POM season due to dissolution of minerals, thereby, groundwater is unsuitable for drinking. F content in groundwater above the DL (35 and 18%) in the PRM and the POM seasons and (6%) exceed the PL of WHO standards in the PRM season. B content in (70 and 51%) surpasses the DL and 58 and 39% samples exceed the PL of the WHO in the PRM and the POM seasons respectively, thus, restricted drinking water use at a few locations. The enrichment of B content in groundwater is due to sewage e uent, fertilizers application, agricultural runoff, overuse of herbicides and poultry waste in the basin. The excessive F is due to semi-arid condition which increased the F ion leaching from the host lithology.
Also, the minerals like uorite, muscovite, biotite, etc., are the main contributors for leaching of F ion in the groundwater. In addition, the slight alkaline nature of groundwater is more favourable to dissolve the uorite mineral. GWQI classi cation exempli ed that, (27 and 15%) samples fall in poor category and only (15 and 18%) samples into excellent category from PRM and the POM seasons, respectively. PIG results classi ed groundwater samples as 6% moderate pollution, 24% low pollution and 70% insigni cant pollution in the PRM season. Also, 3% signify high pollution, 6% low pollution and 91% insigni cant pollution in POM season. Consequently, 18% of groundwater samples are un t for drinking. The HQ of F inferred that children have a higher risk than adults and infants. HHRA results corroborate that non carcinogenic risk of F in order of children > adults > infants. Therefore, children are more susceptible to non carcinogenic risk with deformities of dental and skeletal related F problem than other age groups. The average HQ of B shows the order of impacts with children > adults > infants suggesting that children are the most vulnerable age group. THI results show (88 and 62%) children, (59 and 47%) adults and (38 and 24%) infants in the PRM and POM seasons respectively possess non-carcinogenic risk as THI values (> 1). Thus, the remedial measures like use of safe drinking water, de-uoridation techniques, intake of calcium and phosphorous-rich food, least use B-herbicides and manures, poultry waste management, and public awareness on health risk of F and B contamination are recommended to reduce the health problems in the study area.
Declarations Figure 1 Shivganga watershed with groundwater sample stations Figure 2 Page 21/28 Cross-section pro le show the major litho units in the study area Correlation between the concentration of Mg+K vs HCO3+F Page 22/28

Figure 5
Correlation between the concentration of HCO3+F vs Ca+Mg+Na

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Supplementarymaterial.docx