Mechanisms controlling major ion chemistry and its suitability for irrigation of Narmada River, India

The quality of surface water is driven by several natural and anthropogenic factors, and is one of the most susceptible resources across the globe (Diamantini et al. 2018). The chemistry of surface water quality is controlled by evaporation, precipitation, chemical weathering, geological composition, inputs from its tributaries and human intervention (Pant et al. 2018; Bhat et al. 2021). Therefore, studies on physicochemical characteristics of surface water, especially major cations and anions, have broad implications on the understanding of water quality (Meybeck 2003; Khan & Srivastava 2008).

The main source of water for domestic and agricultural purposes is from surface water like rivers, streams, and ponds (Gupta et al. 2020). The ions present in surface water disclose its potential sources, whether it is natural or anthropogenic and water suitability for human use. Accordingly, Ca⁺⁺, Mg⁺⁺, and HCO₃⁻ ions are considered to be of natural origin, on the other hand, NO₃⁻, SO₄⁻ and Cl⁻ ions are said to be contributed from anthropogenic activities (Acharya et al. 2020). Therefore, estimation of these ions may be helpful in revealing its source, whether it is due to natural interactions such as climate, hydrochemistry, geology, rock weathering (Ca⁺⁺/HCO₃⁻), evaporation, atmospheric deposition (Na⁺/Cl⁻); and anthropogenic activities (Dinka et al. 2015; Samanta et al. 2019).

Studies on water chemistry emphasizing on ions have been conducted globally for determining water quality and its controlling factors (Sahoo et al. 2019; Ayogu et al. 2020; Li et al. 2020; Liu et al. 2020; Ogwueleka & Christopher 2020). Similarly in India, physicochemical studies of rivers have been conducted by several researchers (Sharma & Subramanian 2008; Gupta et al. 2011; Ghosh et al. 2019; Patel et al. 2020). Recent studies on water quality of river Narmada have been carried out by Gupta et al. (2017, 2020). However, all of these researchers have focused on exploring studies on drinking water quality of Narmada River. Only a few studies have been carried out of major ions of Narmada River (Sharma & Subramanian 2008; Gupta et al. 2011; Sharma & Subramanian 2008) reporting that the water of the Narmada was slightly alkaline, and the river was characterized by silicate and carbonate weathering. Gupta et al. (2011) studied the fluvial geochemistry of river Narmada, and indicated that carbonates and salinity obtained from basaltic source influence its water chemistry. However, to date, studies on suitability of Narmada River water for irrigation purposes have not yet been carried out. The suitability of surface water for irrigation can also be analyzed from its physicochemical characteristics using several indices such as Kelly index (Ki), permeability index (Pi), magnesium hazard (MH), sodium percentage (Na%), sodium adsorption ratio (SAR) as well as Wilcox diagram, and USSL's diagram.

Therefore, the current approach focuses: (i) to discern the physicochemical characteristics of Narmada River during pre- and post-monsoon on the basis of major ions, (ii) to identify the mechanisms controlling water chemistry for drawing the inferences on the sources of these ions, and (iii) to recognize the suitability of surface water for irrigation.

The study area covers the entire stretch of the Upper Narmada River in Madhya Pradesh, from its origin point at Amarkantak to Jabalpur (Figure 1). Narmada basin in central India is the fifth-longest river in the Indian subcontinent. It is also known as ‘Life Line of Madhya Pradesh’ bounded within latitude 21°20′N to 23°45′N and longitude 72°32′E to 81°45′E. Narmada River originates at Amarkantak and it flows westwards over a length of 1312 km and discharges in the Arabian Sea at Bharuch in Gujarat (Gupta et al. 2020). Narmada flows in the northern part of Deccan plateau, with its catchment area of 98,796 km² it flows through Madhya Pradesh, Gujarat, and Maharashtra states of India. It is a perennial river, which flows in a rift valley, between the Vindhyan range in the north and Satpura range in the south. Narmada basin has a humid tropic zone with both hilly as well as plain regions. In summer and winter the temperature reaches an average of 42 °C and 9 °C respectively. The upper basin of Narmada is covered by hilly and mountainous regions. It receives an annual rainfall of 1400 mm. Nearly, 80–85% of annual rainfall is contributed by monsoon rainfall (during June–October) (Pandey et al. 2021).

Almost 35% of the Narmada basin is covered by forests, 60% by cultivable land, and 5% by grasslands (Gupta & Chakrapani 2005). Narmada drains a significant load from forests and agricultural lands which discharge into the Arabian Sea. The maximum water discharge flow of the Narmada River at Garudeshwas was about 53,749 m³/s.

The Narmada flows through the Deccan Traps. The basin is subjected to monsoon and has a tropical climate. The Deccan Traps are a major basaltic region on the earth covering approximately 500,000 km² (Gupta et al. 2011). The Narmada graben is a deep-seated, precambrian grofracture (Biswas 1987), and a component of the 1800×200 km² SONATA lineament zone (Shanker 1991). During the monsoon season, the average discharge from Narmada ranges from about 10,000 to 60,000 m³/s (Kale et al. 1994). The underlying lithology of the Narmada basin is dominated by basaltic terrain (Sant & Karanth 1993). These basaltic regions are covered by a tholeiitic composition mainly constituted of phenocrysts derived from olivine, plagioclase, clinopyroxene, opaque minerals (Vigier et al. 2005). Apart from this, some granite and the Gondwana shale including sedimentary rocks in the hills and plains with alluvial gets deposited near the river courses. Alluvial plains about 500 km length and 35–45 km width stretch (from Jabalpur to Barwani) overlying the Deccan traps and Gondwana on both banks of the river.

Surface water samples from 13 sites were collected in pre-acid washed clean PET bottles (1-liter capacity) at a depth of 1–1.5 m below the water surface. Water samples (n=104; in triplicates) were collected twice a year covering pre- and post-monsoon seasons for a period of four years (2017–2020). Separate water samples were collected in 50 mL bottles for Na⁺ and K⁺ analysis. Physical parameters (temperature, pH, electrical conductivity, and total dissolved solid) were analyzed on-site using a portable Oakton Multi parameter tester. The analysis of parameters such as chloride (Cl⁻), bicarbonate (HCO₃⁻), total hardness and calcium hardness were conducted according to the standard protocol prescribed by APHA 2012. Na⁺ and K⁺ was estimated by the flame photometric method. Nitrate (NO₃⁻), sulphate (SO₄⁻), and phosphate (PO₄^—) were estimated by the spectrophotometric method (Thermo Fisher Scientific, Evolution 201 UV-Vis spectrophotometer). Ca⁺⁺ and Mg⁺⁺ ions were calculated according to Equations (1) and (2) respectively.

The conversion factor 2.5 for calcium ions and 4.11 for magnesium ions come from the fact that the total hardness in water is due to the presence of Ca and Mg hardness as CaCO₃ (molecular weight of CaCO₃=100) contributed by both Ca and Mg hardness as CaCO₃. Therefore, the concentration of Ca⁺⁺ and Mg⁺⁺ ions is calculated by dividing Ca hardness and Mg hardness by the factor 2.5 (i.e. molecular weight of CaCO₃/molecular weight of Ca) and 4.11 (i.e. molecular weight of CaCO₃/ molecular weight of Mg) respectively (Ghosh et al. 2019).

The hydrochemical facies reflects the chemical reaction processes going on with the lithology and the surface water. It is broadly used to understand the composition of surface water and its classification into different chemical types which is occurring in hydrological systems. Therefore, for hydrochemical facies, Piper diagrams were plotted (using software GW Chart developed by USGS; Winston 2020). A Piper diagram is a trilinear graphical representation of the chemistry of surface water. It basically represents the cations and anions separately by ternary plots (Figure 4). In addition, Gibb's plot has also been executed to understand the ionic relationships of the water samples.

If present in high concentrations, water quality parameters such as EC, TDS, Na⁺, Ca⁺⁺, Mg⁺⁺, and HCO₃⁻ have adverse impacts on the soil-water-plant relationship. Therefore, the suitability of water used for irrigation should be determined using different indices such as the sodium percentage (Na%), sodium adsorption ratio (SAR), Kelly index (Ki), permeability index (Pi) and magnesium ratio (MAR).

Irrigation water with high Na⁺ content is not suitable for agricultural use as it deteriorates the soil physical structure (causes swelling and dispersion of soil) and restricts plant growth. As a result, permeability of soil becomes poor. Therefore, it is necessary to measure the sodium hazard of irrigation water. It is usually estimated by sodium adsorption ratio (SAR) and sodium percentage (Na %). SAR is the ratio of Na⁺ to Ca⁺⁺ and Mg⁺⁺. It helps to know the strength of water entering into cation-exchange reaction in soils, it is calculated as {SAR=Na⁺/[(Ca⁺⁺+Mg⁺⁺)/2]^1/2} (USSL 1954). High sodium percentage (Na %) is calculated by using the formula {Na %=[(Na⁺+K⁺)×100]/([Ca⁺⁺+Mg⁺⁺+Na⁺ +K⁺]} (Maharana et al. 2015). The infiltration capacity of a soil depends upon its permeability, structure and soil type. The irrigation water which limits the permeability of soil due to high concentrations of Na, Ca, Mg and HCO₃⁻ in water is called permeability hazard and is estimated as permeability index (PI). Permeability index (PI) is calculated as {PI=(Na⁺+√HCO₃⁻)/(Ca⁺⁺+Mg⁺⁺+Na⁺) ×100} (Doneen 1964). Kelly ratio (Ki) is evaluated as the ratio of sodium ion to calcium and magnesium ion {Ki=Na⁺/(Ca⁺⁺+Mg⁺⁺)}. If the Ki value of water is <1 then it is considered to be suitable but when Ki >1 it is unfit for irrigation (Kelly 1963). Magnesium ratio (MAR) is calculated as magnesium ion to calcium and magnesium ion {MAR=[Mg⁺⁺/(Ca⁺⁺+Mg⁺⁺)]×100} (Paliwal 1972). When irrigation water has MAR<50 it is said to be suitable for irrigation but, if MAR >50 it is said to be unsuitable for irrigation use.

The suitability of Narmada water for irrigation purpose has also been assessed through Wilcox diagram and USSL's diagram.

The detailed statistics (minimum, maximum, mean, and standard) of the water samples was calculated using MS Excel 2013. The mean difference of ions for pre- and post-monsoon were determined by one-way analysis of variance (ANOVA) for homogenous and heterogeneous variances at significance level p<0.05. Pearson's correlation analysis was performed in R environment using RStudio software (RStudio Team 2020). Graphs were made in SPSS v.20 software (IBM, 213 USA).

The summary of Narmada River water samples analyzed for different physicochemical parameters for pre- and post-monsoon season is shown in Table 1. The temperature ranges from 22.9 to 36 °C throughout the year in the upper Narmada basin (Figure 2(a)). During pre- and post-monsoon, the average temperature was 30.3 and 27.4 °C respectively. The mean value of pH in pre- and post-monsoon monsoon is ∼7.9, indicating alkaline water like other Indian rivers (Sarkar & Islam 2019; Setia et al. 2021), and it is within the acceptable limit for drinking as per (BIS 1991) (Table 1; Figure 2(b)).

The specific conductivity (EC) of water is its ability to conduct electricity. It directly depends upon the dissolved ions present in water. It is an important parameter because several indices related to irrigation water quality depends upon EC concentration. The mean value of EC was higher in pre-monsoon (322.8 μS/cm) and lower in post-monsoon (312.1 μS/cm) (Figure 2(c)). The mean concentration of total dissolved solids (TDS) in the Narmada River was maximum in post-monsoon (213.4 mg/L) and minimum during pre-monsoon (203.4 mg/L) (Figure 2(d)). Since all water samples have TDS value <500 mg/L, the water is said to be fresh water and can be used for domestic purposes (WHO 2011; Sharma et al. 2021).

The anions variations were in the order of HCO₃⁻ > Cl⁻ > SO₄⁻ > NO₃⁻ > PO₄^— for both pre- and post-monsoon season (Table 1). All the seasons had exceptionally high HCO₃⁻ in the ranges of 13.4–158 mg/L. The maximum concentration of HCO₃⁻ was estimated in pre-monsoon with a mean value of 51 mg/L followed by post-monsoon having a mean concentration of 62.9 mg/L (Figure 2(j)). As the pH of the Narmada was in the range of 7.2–8.8 (Figure 2(b)) i.e. slight alkaline, HCO₃⁻ concentration may have influenced its pH level. The high concentration of HCO₃⁻ indicates chemical weathering inferred from silicate and carbonate weathering rocks present in the river basin (Nisha et al. 2021). In the river catchment area CO₂ dissolves in the surface water through natural gas exchange from atmosphere, respiration of riparian plants and microbial activity in sediments. This dissolved CO₂ forms different ions, i.e. CO₃⁻ and HCO₃⁻ etc. These species are mainly accountable for rocks weathering, particularly carbonate rocks and aluminosilicate minerals (Cole & Prairie 2009).

The concentration of Cl⁻ was found to be maximum in post-monsoon with a mean value of 33.6 mg/L followed by pre-monsoon with a mean value of 29 mg/L respectively (Figure 2(i)). Dissolution of halites and surface runoff sources from alkali/alkaline lithology in the river basin have contributed Cl⁻ as maximum anion after HCO₃⁻ in the river basin (Sharma & Subramanian 2008; Dinka et al. 2015). The source of high Cl⁻ concentration in water samples may be due to the precipitation of Na⁺ in the arid and semi-arid region (Gaillardet et al. 1999), similar results have been reported by Maharana et al. 2015 for Son River which also lies in the Deccan trap region of central India. However, the Cl⁻ concentration in all seasons did not exceed the permissible limit defined by WHO (2011).

NO₃⁻ concentration in Narmada water samples ranged from 0 to 2.5 mg/L (Figure 2(k)) for all seasons and is considered to be in the acceptable range for drinking purposes as per BIS (1991) norms. The major source of NO₃⁻ at some sites could be the inputs from agricultural runoffs or untreated domestic sewage, and degradation of organic matter (Maharana et al. 2015; Pant et al. 2018). The phosphate concentration in the range of 0–0.8 mg/L has been recorded in the Narmada River and was within the permissible limit (WHO 2011; BIS 1991) for both the seasons (Figure 2(l)).

The concentration of sulphate ranged from 0.3 to 24 mg/L; the maximum mean concentration of 8.7 mg/L was recorded during pre-monsoon and minimum in post-monsoon with a mean concentration of 3.3 mg/L (Figure 2(m)). Among the nutrients such as SO₄⁻, PO₄^—, and NO₃⁻ the concentration of SO₄⁻ is found to be maximum in all the seasons, this can be due to oxidation of pyrite related to Gondwanas (Maharana et al. 2015) and the pyrites present in the black shales of Vindhyan range (Banerjee et al. 2006).

The cations concentration in the Narmada River indicates that the dominant cation is Ca⁺⁺ followed by Na⁺> Mg⁺⁺> K⁺ for all seasons (Table 1). The concentration of Ca⁺⁺ in surface water samples was found to be maximum in pre-monsoon with a mean value of 30 mg/L followed by post-monsoon with a mean value of 27.6 mg/L (Figure 2(g)). The concentration of Na⁺ ranges from 6.9 to 29.1 mg/L; the maximum mean concentration of 15.7 mg/L was recorded during pre-monsoon and minimum in post-monsoon with a mean concentration of 14.9 mg/L (Figure 2(e)). The major source of Ca⁺⁺& Na⁺ is weathering of Ca⁺⁺ & Na⁺-bearing minerals like pyroxene present in the Narmada River basin (Srivastava et al. 2019). The cause of the occurrence of sodium in the surface water can be attributed from mineral rocks. The lithology containing salts including sodium, leaches and seepage through springs, precipitation, irrigation, sewage effluents, and infiltration of leachate from landfills or industrial sites also contributes ions (Dwivedi & Pandey 2002; Chaurasia & Pandey 2007). It is said to be that approximately 40% of Na in fluvial water is derived from the leaching of halite and sedimentary or evaporite deposits. Weathering of silicate dominated by sodic plagioclase feldspar is a major source of sodium leaching and contributes to surface water (Priyadarshi 2005). High Na⁺ and Cl⁻ concentration in post-monsoon season may be due to the accumulation of leached salts from small tributaries or springs in the catchment area. Similar types of results, i.e. high Na⁺ and Cl⁻ concentration, were observed in Gandaki river basin during post-monsoon season by Pant et al. (2018).

The average concentration of Mg⁺⁺ was detected maximum in pre-monsoon, i.e. 9.9 mg/L, when compared with mean concentration of post-monsoon, i.e. 9.2 mg/L (Figure 2(h)). Presence and weathering of dolomite rock may be the major source of Mg⁺⁺ in the study area. Moreover, the K⁺ concentration was found to be lowest 0.2–8.6 mg/L among all the cations, with an average value of 3 and 2 mg/L in pre- and post-monsoon season respectively (Figure 2(f)). The weathering of carbonate rocks in the river basin supports the high concentration of Ca⁺⁺ and Mg⁺⁺. Weathering of dolomite, calcium-magnesium silicates are the chief sources of these cations.

Therefore, to check the variation in surface water quality for pre- and post-monsoon season, one-way ANOVA at significant level p<0.05 was applied (Table 2). The ANOVA findings indicated that significant difference was found only for temperature, EC and SO₄⁻. However, for the rest of the parameters only slight difference was observed but this was not statistically significant.

A Pearson correlation analysis between the water quality parameters for both pre- and post-monsoon season was conducted to understand the interrelationship of the parameters. The correlation matrix for pre-monsoon season (Figure 3(a)) shows a good correlation between HCO₃⁻ & Na⁺, SO₄⁻ & temp, K⁺ and Na⁺ (0.7). However, a weak positive correlation was observed between Mg⁺⁺ and pH, SO₄⁻ and pH, EC and NO₃⁻, K⁺ and HCO₃⁻, Ca⁺⁺ and PO₄^— (0.6) whereas, a weak negatively correlation was also observed between Cl⁻ and Na⁺, Ca⁺⁺ and Cl⁻, Cl⁻ and PO₄^— (−0.6). The correlation matrix (Figure 3(b)) for the post-monsoon season showed a positive correlation between K⁺ and Na⁺, HCO₃⁻ and EC (0.6). A significant positive correlation was observed between PO₄^— and temp (0.9), Ca⁺⁺ and pH (0.8). A negative correlation was noted between Cl⁻ and NO₃⁻, Cl⁻ and temp (−0.6). The correlation between NO₃⁻ and Cl⁻ was weak (0.3) for pre-monsoon and negative (−0.64) for post-monsoon. Thus, it is recommended that synthetic fertilizers runoff from agriculture lands may be the main source of nitrate in the river water (Roy et al. 2021).

Hydrochemical facies illustrate the overall scenario of the interaction of surface water and lithology. It has been broadly used to understand the likeness and variances in the composition of surface water and its classification into certain chemical types. The concentration of cations and anions was plotted on a Piper diagram (Piper 1944) to examine the geochemical development of Narmada River water. The typical classification of hydrochemical facies for the Narmada River is shown in Figure 4.

The Piper plot for the Narmada River during the pre-monsoon season (Figure 4(a)) suggests that the major hydrochemical facies in water samples are Ca-Mg-HCO₃ and mixed Ca-Mg-Cl type. The cationic composition indicates Ca-type and anionic composition indicates majorly HCO₃⁻ type, Na-K-type as well as no dominant type in the two ternary plots, whereas for post-monsoon (Figure 4(b)) the hydrochemical faces of surface water samples revealed Ca-Mg-HCO₃ and Cl⁻ type. Overall one major facies, i.e. Ca-Mg-HCO₃ type, is dominant throughout all the seasons. Ternary plot discloses that alkaline earth (Ca⁺⁺ and Mg⁺⁺) exceeds alkalies (Na⁺ and K⁺) in both seasons. Similar types of results, i.e. (Ca⁺⁺ and Mg⁺⁺) exceeding (Na⁺ and K⁺) were also reported in west flowing rivers by Reddy et al. (2021). Dominance of alkaline earth metals may be due to weathering of deccan basaltic rocks present in this region (Ghose 1971).

Gibbs plot was made to understand the hydrochemical measures such as rock dominance, precipitation dominance and evaporation dominance for the surface water chemistry of Narmada River. Gibbs (1970) validated the plot of TDS against Na⁺/(Na⁺+Ca⁺⁺) and TDS against Cl⁻/(Cl⁻+HCO₃⁻) in comprehending the mechanism controlling the hydrochemistry of Narmada. According to Gibb's plot, as shown in Figure 5(a) and 5(b), all the water samples for all the season are located in the rock dominance section. Thus, we can say that the surface water chemistry of Narmada is controlled mainly by the rock-water interaction.

Gibbs plot indicated a rock dominance section (Figure 5(a) and 5(b)) in the Narmada River. Therefore, plots of HCO₃⁻/Na⁺ against Ca⁺⁺/Na⁺ (Figure 6(a)) and Mg⁺⁺/Na⁺ against Ca⁺⁺/Na⁺ (Figure 6(b)) are likewise used for better understanding of rock weathering processes such as silicate or carbonate or evaporate end-members weathering. The plot in Figure 6 shows seasonal variations in water chemistry with carbonate-silicate-evaporates end-members. The plot of Mg⁺⁺/Na⁺ against Ca⁺⁺/Na⁺ (Figure 6(b)) shows that the data lies in the silicate weathering section in both seasons. However, data is scattered towards silicate-carbonate end-members when data was plotted for HCO₃⁻/Na⁺ against Ca⁺⁺/Na⁺ (Figure 6(a)). Therefore, both the plots in Figure 6(a) and (b) revealed that both silicate and carbonate weathering is governing the hydrochemical processes in the Narmada River.

Broadly, CAI (Schoeller 1965) is used to appraise and interpret the ion exchange reactions between the surface water and its lithological structures by using two chloro-alkaline indices CAI-1 and CAI-2. Positive CAI-1 and CAI-2 explains the exchange of Ca⁺⁺ and Mg⁺⁺ found in the rocks with Na⁺ and K⁺ present in the water, i.e. it influences the base-exchange reactions. The CAI-1 and CAI-2 is said to be negative when Ca⁺⁺ and Mg⁺⁺ ions present in water get exchanged with Na⁺ and K⁺ ions found in the rock-bed. Calculation of CAI's was carried out using Equations (3) and (4).

In this study, CAI-1 and CAI-2 were largely positive except for a few samples showing negative CAI-1 & CAI-2 for pre-monsoon season (at sites 10–13) and at site 11 for post-monsoon (Figure 7(a) and 7(b); Table 3). This positive CAI-1 and CAI-2 indicates that the rate of cation exchange between Na⁺ and K⁺ present in the surface water got exchanged with Ca⁺⁺ and Mg⁺⁺ ions present on the surfaces of rocks at sites 1–9 however, it was vice-versa, i.e. negative CAI-1 & CAI-2, for some of the sites (10–13) during pre-monsoon season.

Site 10 is known as Sangam ghat, located in Mandala where Banjar nadi (one of the major tributaries of upper Narmada River) meets Narmada River. Sites 11–13 are located in the Jabalpur region, Jabalpur is also known as the marble city of India because of the presence of marble rock in this region. As CAI-1 and CAI-2 is negative at these sites it hints that Ca⁺⁺ and Mg⁺⁺ loaded water travelling from Maikal hills from Amarkantak, Dindori and Mandala exchanges its Ca⁺⁺ and Mg⁺⁺ with the Na⁺ and K⁺ of the surfaces of rocks. The source of Ca⁺⁺ and Mg⁺⁺ in loaded water may be from water coming from Banjar nadi and meeting Narmada River at Sangam ghat (note: studies on water quality of Banjar River are beyond the scope of this study).

Na/Cl ratios are calculated to determine the probable source of Na⁺ ion in water. Na/Cl ratio >1 indicates silicate weathering, Na/Cl ratio <1 suggests possibility of Na⁺ ion exchanged with Ca⁺⁺ and Mg⁺⁺ ion present in rocks (Wagh et al. 2019). Na/Cl ratios calculated for pre-monsoon at sampling sites 10–13 were 1.35, 1.42, 2.11 & 1.45 respectively (Table 3). Thus, silicate weathering is dominant at these sites (Figure 2(e)).

The river water chemistry is chiefly controlled by erosion and weathering processes occurring in the catchment area. The weathered constituent of the rocks, minerals, and soils decides the flux of dissolved load carried by rivers. Therefore, the estimation, characterization, and quantification of a dissolved constituent are necessary to understand river chemistry. The dissolved constituent of a river is governed by its lithological chemical signatures, vegetative structure, and climatic condition of the catchment. Besides, atmospheric and human interventions also contribute to the chemistry of river water (Meybeck 2003). However, the Narmada River also receives groundwater and geothermal water (Minissale et al. 2000) which also contributes some major ions through chemical weathering from those aquifers (contribution of ions from aquifers has not been discussed here, it is beyond the scope of this study).

The ratio of Ca⁺⁺/HCO₃⁻ was found to be 0.5 and 0.44 for the pre- and post-monsoon season respectively. The ratio of Ca⁺⁺+Mg⁺⁺/HCO₃⁻ was found to be 0.776 and 0.599 for pre- and post-monsoon season respectively. The high value of Ca⁺⁺/HCO₃⁻ and Ca⁺⁺+Mg⁺⁺/HCO₃⁻ during pre- and post-monsoon season directs carbonate weathering and silicate weathering in the basin. The ratio of (Na⁺+K⁺)/total cations are used to discern the input of ions from silicate weathering (Stallard & Edmond 1987). Its calculated value was found to be 0.312 for pre-monsoon and 0.314 for post-monsoon season suggesting a contribution from aluminosilicates weathering (Sharma & Subramanian 2008).

Since carbonate weathering is 10–100 times faster than silicate weathering (Meybeck 1987), the high (Ca⁺⁺+Mg⁺⁺)/(Na⁺+K⁺) ratio favours release of Ca⁺⁺+Mg⁺⁺ ions. The (Ca⁺⁺+Mg⁺⁺)/(Na⁺+K⁺) ratio for Narmada River was 2.44 during pre-monsoon and 2.2 for post-monsoon season. This result indicates that during pre-monsoon there is maximum weathering of carbonates rocks. The rate of weathering in Narmada is significantly higher than the ratio of world river average (2.2) however, the weathering rate of Narmada is lower than the average ratio of 2.5 of Indian River. Thus, we can say in the Narmada River carbonate weathering dominates the silicate weathering.

The assessment of surface water suitability for irrigation was accomplished by using various indices such as the sodium percentage (Na %), sodium adsorption rate (SAR), Kelly index (Ki), permeability index (PI), magnesium hazard (MH), Wilcox diagram, and USSL's diagram. The results of several agricultural water qualities are presented in Table 4. The results revealed that the water quality standards were within the recommended standards for irrigation purposes. Here, Na% values ranged under excellent (7.69%), good (76.93%), and permissible (15.38%) for pre-monsoon whereas under good category (100%) for post-monsoon season. The SAR values were found to be ranged under excellent category for all seasons. Thus, Kelly's ratio (KR), permeability index (PI), and magnesium hazard (MH) collectively indicated that the surface water of Narmada River is suitable for irrigation purposes for different adjacent agroecosystems.

The United States Soil Laboratory Staff (USSL's) diagram (Richards 1954), i.e. SAR vs. EC plot as shown in Figure 8, recommends the usage of Narmada water for agricultural purposes because it falls under the good category and low-medium risk zone (C2S1) in all seasons. The Wilcox plot (Wilcox 1955) was plotted using the Na% and EC of Narmada River (Figure 9) for understanding water usage for agricultural purposes pictorially. The study displays (Figure 9) that water in all seasons falls in the excellent region. Thus, we can recommend its use for agricultural purposes.

This paper highlights the physicochemical characteristics of the Narmada River and its suitability for irrigation activities using the major ions concentration of river water. The overall analysis showed a maximum concentration of HCO₃⁻ in surface water. Among the nutrients, SO₄⁻ concentration was found maximum in all seasons when compared with PO₄^— and NO₃⁻. The Piper plot for Narmada River indicated Ca-Mg-HCO₃ and mixed Ca-Mg-Cl type water in the pre-monsoon season whereas in post-monsoon it is of Ca-Mg-HCO₃ and Cl-type. Gibbs plot suggested that both silicate and carbonate weathering is governing the water chemistry of the river. The rate of cation exchange between Na⁺ and K⁺ in the water and Ca⁺⁺ and Mg⁺⁺ on the rocks was maximum in both the seasons. The Na%, SAR, Ki, PI, MH indices related to water quality standards for irrigation suggested that the surface water of Narmada River is suitable for irrigation use. Although the current study provides detailed information about the hydrochemistry and suitability of irrigation water quality, still we suggest hydrochemical studies on tributaries of Narmada should be carried out for detailed understanding of the source of ions in Narmada River. We further recommend detailed studies of toxic pollutants such as heavy metals and pesticides for better in-depth revelations about the environmental condition of the basin.

We are grateful to the two anonymous reviewers and responsible editors of the journal for their valuable suggestions and comments which helped to improve this manuscript.

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