The water samples were collected from 56 sampling sites in clean polypropylene bottles from Sonitpur and Biswanath Chariali districts, in the central Brahmaputra floodplain region, Assam, India (Figures 1, 2). The water samples collected were from sources used for drinking and irrigation requirements to evaluate their quality concerning their application in irrigation and for drinking. The Global Positioning System (GPS) was used for recording the coordinates of the sampling sites. Physicochemical parameters [electrical conductivity (EC), dissolved oxygen (DO), pH, turbidity, salinity, and total dissolved solids (TDS)] were determined in the field using the multiparameter probe (Systronics). The bottles were sealed airtight and stored following the APHA (2012) protocols. Various parameters such as total alkalinity (TA), total hardness (TH), chloride (Cl⁻), sulfate (SO42-), phosphate (PO43-), bicarbonate (HCO3-), calcium (Ca²⁺), magnesium (Mg²⁺), sodium (Na⁺), potassium (K⁺), F⁻, Fe, inorganic arsenic (iAs) along with calculation of irrigation water quality indices like sodium percentage (Na%), sodium absorption ratio (SAR), permeability index (PI), Kelly's index (KI), magnesium adsorption ratio (MAR), and Irrigation Water Quality Index (IWQI) were calculated to assess the suitability of water for irrigation. Spectrophotometric methods were used to estimate Cl⁻, F⁻, SO42-, and PO43- content, and titrimetric methods were used to determine TA, TH, and HCO3-. Na⁺ and K⁺ were estimated using the flame photometer. Ca²⁺ and Mg²⁺ were determined through the atomic absorption spectrophotometry method. As and Fe were estimated through atomic absorption spectroscopy (AAS) (Thermo scientific ICE 3000) with hollow cathode lamps (iCE™ 3000 Series) implementing Atomic Spectroscopy Standards for As and Fe estimation. Determination of various irrigation water quality parameters was done as follows.

The concentration of dissolved constituents in water determines the water quality, and its suitability for agriculture is termed irrigation water quality. Good irrigation water quality and adequate water management practices along with good soil can maximize the yield.

The GW suitability for irrigation was assessed based on nine irrigation water quality parameters like sodium adsorption ratio (SAR), percentage sodium (Na%), Kelly's index (KI), permeability index (PI), and magnesium adsorption ratio (MAR).

Sodium hazard represented as SAR was calculated as follows (Richard, 1954):

where Na⁺, Ca²⁺, and Mg²⁺ are expressed in meqL⁻¹.

Na% is calculated using the given formula and expressed in meqL⁻¹ (Eaton, 1950):

KI was estimated by the following equation (Kelly, 1940):

The ions are expressed in meqL⁻¹.

Doneen (1964) forwarded the formula for computing permeability index (PI):

MAR was calculated using the equation:

The concentrations of ions are expressed in meqL⁻¹.

The parameters EC, SAR, Na⁺, Cl⁻, and HCO3- are considered for unweighted arithmetic water quality index (WQI_UA) calculation to evaluate groundwater irrigation suitability and are estimated using Equations (8) and (9) (Farrag, 2005; El Tahlawi et al., 2016):

where q_i is the quality rating for ith parameter, V_i is the observed value for ith parameter, and S_i is the water quality standard value.

where n is the number of parameters.

SPSS (21) was used for the multivariate statistical analysis to investigate and interpret the data sets. The physicochemical parameters were used to obtain Pearson's correlation matrix and identify the association among the various parameters and determine the control of the water ion chemistry. The dataset dimensionality is reduced using the principal component analysis (PCA) and hierarchical agglomerative cluster analysis (HCA) and helps characterize the variance of a substantial set of inter-related variables (Kumar et al., 2017). PCA helps to identify the probable causes of pollution and the factors influencing them. The normalized data were applied with Ward's linkage method in HCA, and a dendrogram was obtained (Kumar et al., 2017; Elsayed et al., 2020). Piper and Gibbs plots were implemented to evaluate the groundwater hydrogeochemical evolution of the study area. The piper trilinear diagram illustrates the hydrogeochemical facies of the aquifer system (Piper, 1944). The diamond-shaped central region of the diagram displays the groundwater classification, and the major cations (Ca²⁺, Mg²⁺, Na⁺, and K⁺) and anions (HCO3-, CO32-, Cl⁻, and SO42-) are expressed as a percent of meqL⁻¹ and plotted in the two basal triangles. Similar water samples will be inclined toward the same group (Todd, 1980; Saha et al., 2019). The Gibbs plot sheds early insight into the governing hydrogeochemical processes of the study area (Gibbs, 1970).

The electrical conductivity (EC) represents the conductivity of water, and the total dissolved species is represented by TDS. The results of EC and TDS varied from 105 to 980 μScm⁻¹ and 56 to 934 mgL⁻¹ having average values of 540.96 μScm⁻¹ and 216.22 mgL⁻¹, respectively. TDS values for 96% of samples were found within the acceptable limit of 500 mgL⁻¹. Consuming water with TDS values >500 mgL⁻¹ can cause gastrointestinal irritation (BIS, 2012). In natural aquifer systems, dissolved oxygen (DO) is a significant parameter that determines the health of the system. DO ranges between 2.00 and 3.30 mgL⁻¹.

The chemical analysis revealed that the mean concentrations (mgL⁻¹) of major cations varied in the order Na⁺>Ca²⁺>Mg²⁺>K⁺, and for anions, the order was HCO3->Cl⁻>SO42->PO43- (Table 1). The Na⁺, K⁺, Ca²⁺, and Mg²⁺ were present in the range of 17.40–62.30 mgL⁻¹, 0.10–45.00 mgL⁻¹, 11.37–32.22 mgL⁻¹, and 2.10–17.40 mgL⁻¹, respectively. The most abundant alkali metal is observed to be Na⁺, and sources like the weathering of halite and silicate minerals may contribute to it. However, by-products of agricultural activities may increase the sodium concentration of the groundwater (Saha et al., 2019). Content of Na⁺>K⁺ because the rate of Na⁺ dissolution is much higher than that of K⁺. The use of potassium-based fertilizers and the decomposition of animal or waste products are considered potential sources of K⁺ (Saha et al., 2019). Ca²⁺ was observed as the most abundant alkaline earth metal, and Mg²⁺ can be derived from sources such as the weathering of dolomite and ferromagnesium minerals (Saha et al., 2019). All samples were well within the permissible limit of WHO for drinking and irrigation concerning Na⁺, K⁺, Ca²⁺, and Mg²⁺.

Cl⁻ in the groundwater can be derived from rocks, seawater intrusion, or pollution by domestic sewage and industrial waste. Lower Cl⁻ content represents low salinity in the groundwater of the region (Saha et al., 2019). The organic matter present in the water system oxidizes to produce carbon dioxide facilitating mineral dissolution and contributing to HCO3- enrichment (Khashogji and El Maghraby, 2013). Weathering of silicate minerals also contributes to HCO3- enhancement (Gastmans et al., 2010; Saha et al., 2019). The groundwater is enriched with Ca²⁺, Mg²⁺, and HCO3- ions due to weathering activities. HCO3- was observed between 35.98 and 485.98 mgL⁻¹ and Cl⁻ between 22.36 and 115.23 mgL⁻¹, respectively. Cl⁻ results were observed within the WHO permissible limit for drinking purposes. Cl⁻ was observed as a principal contributor of TDS and EC enrichment in the aquifer system. SO42- was observed in the range of 0.67–4.36 mgL⁻¹ and well within the WHO and BIS permissible limits for drinking and irrigation. However, 48% samples exceeded the WHO permissible limit for HCO3-. PO43- values ranged between BDL-0.08 mgL⁻¹ (average- 0.01 mgL⁻¹), and all samples were well within the permissible limit of WHO for PO43- for drinking and irrigation.

F⁻, naturally occurring in the earth's crust, has a remarkable effect on human physiology (Lanjwani et al., 2021). F⁻ is essential for developing dental health; however, higher concentrations will result in toxic effects such as dental and skeletal fluorosis [(World Health Organization (WHO)., 2017; Gogoi et al., 2021)]. F⁻ values ranged between 0.06 and 0.44 mgL⁻¹ with an average value of 0.19 mgL⁻¹ indicating the groundwater is safe for drinking and irrigation concerning F⁻ concentrations.

However, Fe values were observed to be significantly high ranging from 0.05 to 32.20 mgL⁻¹ with an average value of 7.04 mgL⁻¹. About 95% (n = 53) samples surpassed the WHO permissible limit of Fe in drinking water, thereby raising concerns for the region. Although Fe is an essential element for living cells, its long-term exposure can result in toxicological concerns (Goswami et al., 2020). High levels of Fe cannot be disposed by the body and can cause severe damage to the internal organs (Chaturvedi et al., 2014; Goswami et al., 2020).

As values ranged between BDL-0.08 mgL⁻¹ (average-0.03 mgL⁻¹) which indicates As contamination of the groundwaters in the region. Consuming unsafe amounts of As on a regular basis may lead to chronic As poisoning eventually resulting in carcinogenic risks, hyperkeratosis, cardiovascular diseases, birth effects, diabetes, and neurotoxicity (Flanagan et al., 2012; Kumar et al., 2017; Goswami et al., 2020, 2022; Shaji et al., 2021). The potential of As as a threat to human health is intense in highly populated regions primarily dependent on the groundwater as the chief drinking water source and for use in irrigation (Gillispie et al., 2015; Goswami et al., 2020). Food and Agriculture Organization (FAO) recommended 0.10 mgL⁻¹ as the maximum arsenic concentration for irrigation assuming the rate of water application to be 10,000 m³, following adequate irrigation practices. However, if the rate of water application largely surpasses 10,000 m³, then the permissible limit reduces. The existence of Archean and gneissic rocks in the region (Central Ground Water Board, 2013) may be a significant contributor to its rich mineralogical composition. The shallow aquifers of the Holocene epoch in the Brahmaputra floodplains are enriched with high As contents as reported by several researchers (Goswami et al., 2014; Shah, 2015; Kumar et al., 2016a,b; Patel et al., 2019). As concentrations >10μgL⁻¹ were also reported in most of the shallow tube well waters of Bangladesh (Zhao et al., 2010). This may be due to the re-depositional and sediment mixing during high-flow duration of the Brahmaputra River (Goswami et al., 2020). Using the contaminated water for irrigation purposes may result in As accumulation in the agricultural soils, gradually affecting the growth and yield of the crops and causing adverse health effects on humans on the consumption of As-contaminated food (Brammer and Ravenscroft, 2009; Gillispie et al., 2015; Goswami et al., 2020). Moreover, even if the source of the contaminated water that is consumed is the same, still As bioaccumulation in the residents of the same region may vary. The contamination level is also reliant on the storage methods of the water for domestic use. Storage in earthen pots reduces the As content by facilitating As oxidation and precipitation (Goswami et al., 2020).

The hydrochemical facies of the groundwater is interpreted by implementing the Piper diagram (Piper, 1944). Samples having similar qualities tend to fall under the same group (Ramesh et al., 2014; Saha et al., 2019). Two triangles represent the major cations and anions, and a diamond-shaped plot represents the combination of both. The latter helps to interpret and classify the hydrogeochemical characteristics of the groundwater (Shil et al., 2019). Figure 3 indicates the water of the study area belonged to the mixed type of waters namely Ca²⁺-Na⁺-HCO3- and Ca²⁺-Mg²⁺-SO42-. The dominant cation is observed to be Na⁺, and the plot suggests higher content of Ca²⁺ is being exchanged with Na⁺ than Mg²⁺. HCO3- is the major anion represented by the diagram which can be attributed to the presence of calcite, dolomite, rhyolite, and basalt minerals.

Gibbs diagram helps in the assessment of the source of ions in the groundwater by evaluating the ratio among the cations and anions and TDS. The Gibbs ratio is estimated by the following equations:

The distribution of the cations and anions and TDS in the diagram was used to interpret rock–water interaction, evaporation, and precipitation dominance (Lanjwani et al., 2021). Figure 4 indicates the dominance of rock–water interactions pinpointing that rock geochemistry was the principal source in the groundwater of the region. The abundance of the varied dissolved ions in the groundwater relies on their abundance in the parent rock and their rate of solubility (Saha et al., 2019).

HCO3- vs. Na⁺ was plotted to investigate the weathering processes. Samples on the left of the equiline represent carbonate weathering and, on the right, represent silicate weathering (Lanjwani et al., 2021). Figure 5A shows that the samples were dispersed on both sides of the equiline indicating the prevalence of carbonate and silicate weathering promoting the release of HCO3- and Na⁺. The Ca²⁺ vs. HCO3- (Figure 5B) indicates a majority of samples on the left of the equiline represent the dominance of silicate weathering (carbonic acid weathering) over carbonate weathering. Samples plotted along the trendline indicate calcite dissolution (Lanjwani et al., 2021).

In the Na⁺ vs. Cl⁻ scatter plot (Figure 5C), samples along the equiline (Na⁺/Cl⁻ ratio equivalent to 1) represent halite dissolution. If Na⁺/Cl⁻ ratio> 1, it signifies silicate weathering (reverse ion exchange) whereas Na⁺/Cl⁻ ratio <1 indicates the ion exchange process (Egbueri et al., 2019; Lanjwani et al., 2021). Figure 5C implies the prevalence of reverse ion exchange and forward ion exchange processes. A few samples also represented halite dissolution. Ca²⁺ vs. SO42- plot (Figure 5D) shows silicate weathering as a predominant mechanism.

The effects of the distribution of sulfate and carbonate minerals are evaluated by the Ca²⁺+Mg²⁺ vs. HCO3-+SO42-. Figure 5E ascertains the prevalence of reverse ion and forward ion exchange processes in the aquifer systems. Sample points along the equiline indicate sulfate and carbonate weathering. Points above the equiline indicate reverse ion exchange and carbonate–sulfate dissolution, whereas samples plotted below the equiline represent silicate weathering (Lanjwani et al., 2021). Evaporation processes may contribute to the enrichment of species in the water. The Na⁺/Cl⁻ ratio remains unchanged if the evaporation process is dominant (Jankowski and Acworth, 1997; Lanjwani et al., 2021). Figure 5F reveals the reverse ion exchange process to be predominant in the region.

Mineral equilibrium estimations, using MINTEQ 3.1 geomodeling tool, for groundwater aid in the prediction of reactive mineral presence in the aquifer system which provide a saturation index (SI) for the minerals interacting in the groundwater system (Goswami et al., 2020, 2022). SI is the logarithm of the ratio of the ion activity product to the mineral thermodynamic equilibrium constant, adjusted for the temperature of the sample, and given as SI=log(IAPKSP). Saturated (equilibrium, SI = 0), undersaturated (dissolution, SI < 0.5), and oversaturated (precipitation, SI > 0.5) saturation states are characterized on the basis SI value. SI for anhydrite, arcanite, bischofite, carnallite, chloromagnesite, epsomite, fluorite, gypsum, halite, kainite, melanterite, mirabilite, pentahydrite, sylvite, tachyhydrite, and thenardite is calculated and is displayed in Table 2. The bulk of samples are undersaturated with the selected minerals, according to SI values. Fluorite and gypsum (an evaporitic mineral) showed the potential for increase, indicating that evaporitic mineral dissolution has a substantial influence on the region's groundwater chemistry. Low-to-high undersaturated forms of anhydrite, halite, and epsomite along with other minerals can be found. The saturation indices are greater due to higher recharge and shifts in the aquifers with reduction potential and weathering.