This study details the hydrochemical characterization and human health risk assessment of groundwater in the Narmada River Basin. The study was performed based on data collected from 305 groundwater sample stations in the Narmada River Basin. Hydrochemical evaluation illustrated that cationic ions in the upper and middle Narmada Basin were dominated by Ca2+; however, in the lower basin it was dominated by Na+ ions. Similarly, anionic ions were dominated by HCO3 throughout the basin. A Chadha plot drawn from the collected data inferred that most groundwater belonged to the recharge water category (Ca-Mg-HCO3 type). Base-exchange indices of the collected data confirmed the presence of Na+-SO42− type of groundwater. Meteoric genesis indices indicated deep meteoric percolation groundwater. Further, Gibbs plots categorized groundwater samples in the rock dominated section, while chloro-alkaline indices confirmed direct as well as reverse ion-exchange reactions governing groundwater quality. Water Quality Index values showed that groundwater ranged from excellent to very poor. Human health risk of the Narmada River confirmed the non-carcinogenic risk for Nitrate (NO3) and Fluoride (F) ions. However, several indices justified that groundwater was ideal for irrigation. However, groundwater treatment is recommended before direct consumption such as drinking.

  • Both reverse and direct ion exchange reactions and rock weathering govern the groundwater quality.

  • Dominance of Na+-SO42− type and deep meteoric percolation types of groundwater.

  • The Water Quality Index (WQI) ranges from excellent to very poor categories.

  • A non-carcinogenic risk due to nitrate and fluoride contamination was confirmed.

Graphical Abstract

Graphical Abstract
Graphical Abstract

Groundwater is a clean, safe, and portable source of water in comparison to surface water (Şener et al. 2017). Surface water deficiency in arid and semi-arid areas can be complemented by groundwater to meet drinking and irrigation demands (Khosravi et al. 2016). Population growth, urbanization and industrialization during the past few decades have accelerated water demand in different sectors (Gupta et al. 2020a, 2020b). The constantly increasing pressure of waste discharges from households and industries, and runoff from agricultural lands to surface water have all imposed pressure on groundwater resources and has led to their depletion and degradation (Kaviarasan et al. 2016; Gnanachandrasamy et al. 2020). In the past few decades, an increasing trend in groundwater exploitation has resulted in poor recharge, and enhanced contamination through rock water interaction (McArthur et al. 2011). Groundwater quality alters with time and space, depending upon lithology, weathering of rocks, ion-exchange reactions, evaporation and dissolution (Narany et al. 2015; Gnanachandrasamy et al. 2020). Extreme weather events such as floods and droughts etc. triggered due to climate change have also impacted groundwater quality. Occurrence of a drought in an area can affect aquifer ‘recharge/discharge’ balance and further alter the water quality of aquifers (Barbieri et al. 2021). Groundwater quality also gets degraded by anthropogenic inputs such as discharge of untreated industrial and municipal wastewater, agriculture runoff and leaching (Mondal et al. 2010; Matta et al. 2017; Barbieri et al. 2019).

The hydrochemical characterization of groundwater is necessary because it helps in revealing the processes involved in governing its quality to meet the various standards for drinking and irrigation. Several techniques have been used for geochemical identification of groundwater, such as a Chadha plot, chloro-alkaline indices, and saturation indices (Kumar et al. 2022). Moreover, determination of overall water quality through different indices like the Arithmetic Water Quality Index (AWQI) (Gupta et al. 2020a, 2020b), Comprehensive Water Quality Indices (CPI) (Gupta et al. 2020a, 2020b), Canadian Council of Ministers of the Environment Water Quality Index (CCMEWQI) (Akkoyunlu & Akiner 2012), Entropy Water Quality Index (EWQI) (Rao et al. 2020) and Oregon Water Quality Index (OWQI) (Abbasi 2002) have also been applied to evaluate the quality of water. Water quality assessment for irrigation, using different indices, e.g. Sodium Percentage (Na%), Kelly Index (Ki), and Sodium Adsorption Ratio (SAR), has also been promising (Gupta et al. 2022) in assessing groundwater conditions.

Several diseases such as dental fluorosis and methemoglobinemia (blue baby syndrome) are caused by the consumption of water contaminated with fluoride and nitrate (Rao et al. 2018). Thus, methods of assessment of human health hazards concerning ingestion of groundwater with high levels of fluoride and nitrate need to be identified for the protection and safeguarding of human health (Shukla & Saxena 2021).

Groundwater quality in the Narmada basin has been evaluated by several researchers and found to be contaminated with many pollutants. Jha et al. (2021) studied the water quality in the Mandala region located in the Narmada basin finding that aquifers in the low lying region (below 500 m mean sea level) were contaminated by fluoride and are harmful to human health. Gawle et al. (2021) studied the groundwater of the Dindori region situated in the Narmada basin and concluded that the majority of groundwater samples were not fit for direct drinking. Thus, the groundwater resource of the Narmada basin might be of poor quality and may be hazardous to human health especially concerning fluoride and nitrate contamination. Considering the pressure on groundwater resources and importance of its quality assessment, we performed this study with the objectives (i) to assess and determine the important geochemical processes and factors impacting the groundwater chemistry in different sub-basins of the Narmada River, and (ii) to identify its suitability for consumption and irrigation.

General characteristics of study area

Narmada River Basin (NRB) extends between longitude 72°38′ to 81°43′ E and latitude 21°27′ to 23°37′ N, covering an area of 92,672.42 km2 which accounts for nearly 3% of the total geographical area of India (Figure 1). The NRB lies largely in Madhya Pradesh (82,094.94 km2), Gujarat (8,326.34 km2) and a small part of Maharashtra (1,580.07 km2) and Chhattisgarh (675.7 km2). The basin is divided into three sub-basins i.e., the Upper Narmada Basin (UNB), Middle Narmada Basin (MNB) and Lower Narmada Basin (LNB). Typically, the basin has four seasons i.e., cold, hot, south-west monsoon and post-monsoon, with a tropical humid type climate. High rainfall occurs in the upper and the lower (coastal) region of the river basin with moderate-to-low rainfall in the middle plain areas of the basin. The average annual temperature during summer swings from 30 °C to 32.5 °C, but sometimes maximum temperature reaches ∼48 °C. The basin receives an average rainfall of about 1,200 mm annually.
Figure 1

Narmada River Basin map showing geological features as well as groundwater sampling stations.

Figure 1

Narmada River Basin map showing geological features as well as groundwater sampling stations.

Close modal

Hydrogeology and topography of study area

The NRB is a part of a volcanic province i.e., the Deccan Trap, covered by heterogeneous geological features. The main lithology of the study area consist of Archaean, alluvial sedimentary formations of Gondwanas and Vindhyan groups, with the major portion covered by Deccan Trap basalt (Kumar et al. 2005). Topographically, UNB terrain is hilly consisting of flat Proterozoic sedimentary rock, whereas the terrains of the MNB and LNB are fertile plains dominated by limestone and sandstones and are suitable for cultivation. However, the western part of LNB lies in the coastal region of the Arabian Sea and is of an alluvial nature. The basin is dominated and covered by black soil.

Granite rock, medium to coarse grained sandstone and limestone bearing rocks, and basaltic rock are the dominant rocks found in this region. Granitic aquifers, mostly unconfined sandstone supporting phreatic aquifers, basaltic aquifers of a confined to semi-confined nature and alluvium aquifers are found in the basin (Jha et al. 2021). The developed porosity in Gondwana rocks, vascularity in the basalt rocks, and presence of joints and fractures in the rocks along with varying degrees of weathering and secondary porosity developed naturally due to joints and fractures play a crucial role in groundwater yield and its movement (CGWB 2017). The basin is mostly covered with basaltic aquifers and yields about 10–50 m3 day−1 (CGWB 2017), with low hydraulic conductivity and high inhomogeneity. The depth of water level in the study area varies from 5 to 20 m below ground level (CGWB 2018).

Data collection and preparation

The India Water Resources Information System (India WRIS) is a web-based database developed under the National Hydrology Project in 2016 with the prime goal of gathering information on the water resources of India. India-WRIS has groundwater quality monitoring stations throughout the nation. The groundwater quality data for 2018 for the NRB from 305 monitoring stations (132 samples from UNB; 122 samples from MNB; and 51 samples from LNB) were collected from India-WRIS. The locations of the above mentioned 305 sampling stations are shown in Figure 1. The data obtained represent 12 parameters i.e., pH, total dissolved solids, electrical conductivity, sodium, potassium, calcium, magnesium, chloride, bicarbonate, nitrate, sulphate, and fluoride.

All data were checked for normalized charge balance using Equation (1).
(1)

The values of the normalized charge balance were in the range of ±15%, justifying the balance in ions (cations and anions), and confirmed the reliability of data for further geochemical studies.

Groundwater chemistry and geochemical modeling

Chemometric analysis of groundwater quality data was performed through different plots such as a Chadha plot, biplot, scatter plot, Gibbs plot, end-member diagram and chloro-alkaline indices (CAI) etc. The hydrochemical classification of groundwater relating the base-exchange of ions was calculated. Thus, r1 < 1 (Na+-SO42− type) and r1 > 1 (Na+-HCO3 type) (r1 values were computed using Equation (2)). The values of r2 obtained from Equation (3) are classified as r2 < 1 (deep meteoric percolation) and r2 > 1 (shallow meteoric percolation).
(2)
(3)

Inverse geochemical modeling

Inverse geochemical modeling has been extensively used to understand the geochemical processes liable for the factors governing groundwater composition (Singh et al. 2017). It is a mass balance simulation. Therefore, to evaluate an inverse geochemical model, PHREEQC version 3, a freely available computer program launched by USGS, was used. The carbonate minerals include calcite, dolomite, and aragonite, whilst anhydrite, and gypsum are considered to be sulphate minerals. The Saturation Index (SI) for both carbonate and sulphate minerals was calculated using Equation (4). The values obtained from the SIs specify/reveal the hydrogeochemical reactions responsible for the alterations in the chemical characteristics i.e., dissolution or precipitation of minerals in groundwater. SI = 0 indicates an equilibrium condition, of minerals in solution; SI > 0 indicates oversaturation condition, having tendency of precipitation of the minerals; and SI < 0 suggests an under saturation condition having tendency of dissolution of the minerals until equilibrium is attained.
(4)
where IAP = ion activity product, and KSP = solubility product.

Water quality index and irrigation water quality indices

The suitability of groundwater with respect to the Indian Standard (BIS 2012) for drinking purposes was addressed by a Water quality index (WQI). To compute the WQI, specific weights (swi) were assigned according to the importance of each hydrochemical parameter ranging from 2 to 5. The WQI is computed using the following equations:
(5)
(6)
(7)
where RWi is the relative weight, QRi = quality rating, coni = measured concentration, stdi = standard values (proposed by BIS 2012) of the respective parameter. “n” is the total number of hydrochemical parameters. The water quality index (WQI) was categorized into five categories as follows: excellent (WQI < 50); good (50< WQI < 100); poor (100 < WQI < 200); very poor (200 < WQI < 300); and unsuitable (WQI > 300).

The appropriateness of groundwater for irrigation purposes was computed and classified as per sodium percentage (Na %), Kelly's index (Ki), sodium adsorption ratio (SAR), Permeability index (Pi), Magnesium ratio (MAR), Residual Sodium Carbonate (RSC), Wilcox diagram (Wilcox 1955) and US Salinity Laboratory (USSL's) diagram (Table 8).

Human health risk assessment

Non-carcinogenic risks caused due to direct consumption of groundwater with chemical parameters exceeding permissible limits were computed using health risk assessment equations (USEPA 2004). In this study, NO3 and F were the two parameters considered to compute non-carcinogenic health risk for two target groups i.e., children and adults. The chronic per day intake of a particular pollutant is computed based on Equation (8):
(8)
where, CDI = chronic per day intake (mg/kg/day); CGW = chemical concentration in groundwater, here being nitrate or fluoride concentration (mg/L); IR = ingestion (consumption) rate (In the Indian context daily IR is 2.8 liters (adults) and 0.95 liters (children); EF = exposure frequency (365 days)); ED = exposure duration (64 years (adults) and 12 years (children)) (Shukla & Saxena 2021); ABW = average body weight in kg (65 (adults) and 20 (children)); AET = average exposure duration in days (23,360 (adults) and 4,380 (children)).
Further, to measure non-carcinogenic chronic hazards of a particular contaminant, a reference dosage (RfD) is used and is computed as the hazard quotient (HQ) using Equation (9):
(9)

Here, RfD = 0.04 and 1.6 mg/kg/day for fluoride and nitrate, respectively (USEPA 2014).

Lastly, a total hazard index (THI) for non-carcinogens (i.e., combined effect of NO3 and F) is estimated (Equation (10))
(10)

According to (USEPA 2014), THI > 1 indicates non-carcinogenic health risk to its consumers, whereas THI < 1 confirms no risk to its consumers (Haghnazar et al. 2022).

Statistical and spatial analysis

The detailed statistics (mean, standard deviation, minimum, maximum) of different physicochemical parameters were calculated using MS Excel 2013. Pearson's correlation coefficient was calculated in R-4.2.0 software. Spatial analysis using the interpolation technique (IDW)) was performed in Quantum gis (https://www.qgis.org).

General physicochemical characteristics of groundwater

The physicochemical characteristics of the groundwater of UNB, MNB and LNB are summarized and compared with BIS (BIS 2012) and World Health Organization (WHO) (WHO 2017) standards in Table 1.

Table 1

Descriptive statistics of hydrochemical parameters of groundwater in the Narmada River Basin

ParameterBIS (2012) AND WHO (2017)
Upper Narmada Basin
Middle Narmada Basin
Lower Narmada Basin
Acceptable limitPermissible limitminmaxmeanSDminmaxmeanSDminmaxmeanSD
pH 6.5–8.5 – 6.79 8.69 7.54 0.37 6.86 8.42 7.37 0.25 6.22 8.36 7.85 0.39 
EC (μmhos/cm) – 500 210 1,742 800.64 332.3 365 2,320 1,067.25 428.41 363 5,530 1,248.9 957.65 
TDS (mg/L) 500 2,000 136.5 1,132.3 520.41 215.99 237.25 1,508 693.71 278.47 243.21 3,705 834.4 641.3 
Ca2+ (mg/L) 75 200 12 178 59.37 29.5 20 312 80.82 41.73 28 218 77.42 41.88 
Mg2+ (mg/L) 30 100 1.49 153.25 32.22 24.39 2.45 137.59 38.15 23.51 277 49.85 48.07 
Na+ (mg/L) – 200 321 50.19 39.04 14 289 70.99 54.27 777 110.31 137.26 
K+ (mg/L) 12 300 0.1 1.46 1.39 0.1 86.1 7.19 14.02 0.1 38.4 4.47 7.79 
HCO3 (mg/L) 120 600 49 656 297.91 135.26 24 766 366.24 130.03 49 659 314.44 154.21 
Cl (mg/L) 250 1,000 260 63.42 52.04 10 404 96.75 76.89 14 1,454 189.94 252.21 
SO42− (mg/L) 200 400 195 24.55 22.46 185 28.57 25.38 414.37 73.3 88.55 
NO3 (mg/L) 45 100 115 27.88 23.39 217 51.82 39.73 250 34.88 46.46 
F (mg/L) 1.5 0.01 2.19 0.36 0.36 0.08 3.7 0.57 0.52 0.12 0.87 0.5 
ICB% (±15%)   − 9.5 5.26 − 2.06 2.23 − 9.9 1.93 − 1.61 2.3 − 14.2 9.59 − 4.1 4.69 
Anhydrite   − 3.73 − 2.01 − 2.79 0.35 − 3.48 3.03 − 2.62 0.64 − 3.50 − 1.27 − 2.44 0.52 
Aragonite   − 15 1.06 0.00 1.39 − 0.67 1.03 0.15 0.33 − 1.38 1.39 0.52 0.45 
Calcite   − 1.38 1.21 0.28 0.44 − 0.53 1.17 0.30 0.32 − 1.24 1.53 0.65 0.46 
Dolomite   − 2.83 2.29 0.52 0.96 − 1.99 2.50 0.49 0.82 − 2.43 3.18 1.40 0.97 
Fluorite   − 5.29 0.31 − 2.43 0.92 − 3.16 1.45 − 1.74 0.74 − 2.45 − 0.23 − 1.37 0.41 
Gypsum   − 3.42 − 1.71 − 2.49 0.35 − 22.77 − 1.33 − 2.52 1.89 − 3.19 − 0.97 − 2.13 0.52 
Halite   − 8.51 7.20 − 7.20 1.37 − 8.31 − 5.61 − 7.00 0.59 − 8.23 − 4.62 − 6.76 0.84 
Sylvite   − 10.1 − 7.07 − 8.45 0.57 − 9.94 − 5.64 − 7.96 0.86 − 9.27 − 6.06 − 7.85 0.83 
ParameterBIS (2012) AND WHO (2017)
Upper Narmada Basin
Middle Narmada Basin
Lower Narmada Basin
Acceptable limitPermissible limitminmaxmeanSDminmaxmeanSDminmaxmeanSD
pH 6.5–8.5 – 6.79 8.69 7.54 0.37 6.86 8.42 7.37 0.25 6.22 8.36 7.85 0.39 
EC (μmhos/cm) – 500 210 1,742 800.64 332.3 365 2,320 1,067.25 428.41 363 5,530 1,248.9 957.65 
TDS (mg/L) 500 2,000 136.5 1,132.3 520.41 215.99 237.25 1,508 693.71 278.47 243.21 3,705 834.4 641.3 
Ca2+ (mg/L) 75 200 12 178 59.37 29.5 20 312 80.82 41.73 28 218 77.42 41.88 
Mg2+ (mg/L) 30 100 1.49 153.25 32.22 24.39 2.45 137.59 38.15 23.51 277 49.85 48.07 
Na+ (mg/L) – 200 321 50.19 39.04 14 289 70.99 54.27 777 110.31 137.26 
K+ (mg/L) 12 300 0.1 1.46 1.39 0.1 86.1 7.19 14.02 0.1 38.4 4.47 7.79 
HCO3 (mg/L) 120 600 49 656 297.91 135.26 24 766 366.24 130.03 49 659 314.44 154.21 
Cl (mg/L) 250 1,000 260 63.42 52.04 10 404 96.75 76.89 14 1,454 189.94 252.21 
SO42− (mg/L) 200 400 195 24.55 22.46 185 28.57 25.38 414.37 73.3 88.55 
NO3 (mg/L) 45 100 115 27.88 23.39 217 51.82 39.73 250 34.88 46.46 
F (mg/L) 1.5 0.01 2.19 0.36 0.36 0.08 3.7 0.57 0.52 0.12 0.87 0.5 
ICB% (±15%)   − 9.5 5.26 − 2.06 2.23 − 9.9 1.93 − 1.61 2.3 − 14.2 9.59 − 4.1 4.69 
Anhydrite   − 3.73 − 2.01 − 2.79 0.35 − 3.48 3.03 − 2.62 0.64 − 3.50 − 1.27 − 2.44 0.52 
Aragonite   − 15 1.06 0.00 1.39 − 0.67 1.03 0.15 0.33 − 1.38 1.39 0.52 0.45 
Calcite   − 1.38 1.21 0.28 0.44 − 0.53 1.17 0.30 0.32 − 1.24 1.53 0.65 0.46 
Dolomite   − 2.83 2.29 0.52 0.96 − 1.99 2.50 0.49 0.82 − 2.43 3.18 1.40 0.97 
Fluorite   − 5.29 0.31 − 2.43 0.92 − 3.16 1.45 − 1.74 0.74 − 2.45 − 0.23 − 1.37 0.41 
Gypsum   − 3.42 − 1.71 − 2.49 0.35 − 22.77 − 1.33 − 2.52 1.89 − 3.19 − 0.97 − 2.13 0.52 
Halite   − 8.51 7.20 − 7.20 1.37 − 8.31 − 5.61 − 7.00 0.59 − 8.23 − 4.62 − 6.76 0.84 
Sylvite   − 10.1 − 7.07 − 8.45 0.57 − 9.94 − 5.64 − 7.96 0.86 − 9.27 − 6.06 − 7.85 0.83 

The pH values ranged from (6.79–8.69), (6.86–8.42), (6.22–8.36) for UNB, MNB and LNB respectively, with an average value of ∼7.58 for the NRB, showed a neutral to slight alkaline nature of groundwater. However, 1.5% of samples in UNB and 1.96% of samples in LNB exceeded the recommended pH value of 6.5–8.5 (BIS 2012). The electrical conductivity (EC) had a wide range of variation from (210–1,742), (365–2,320), (363–5,530) with a mean value of 800.64, 1,067.25, and 1,248 μmhos/cm for UNB, MNB and LNB, respectively. According to WHO standards, EC value should not exceed 1,500 μmhos/cm, but 3%, 13.9% and 29.4% of the groundwater samples were far beyond the permissible limit for UNB, MNB and LNB, respectively. The TDS values in groundwater of UNB, MNB, and LNB varied from 136.5–1,132.3 mg/L, 237.25–1,508 mg/L, and 243.21–3,705 mg/L with mean values of 520.41, 693.71, and 834.4 mg/L, respectively. About 52.3%, 72.1%, and 66.7% of groundwater samples of UNB, MNB, and LNB, respectively, surpassed the prescribed values of acceptable limits of TDS i.e., 500 mg/L (BIS 2012).

Ion concentration in groundwater

Cation concentration indicated the dominance of Ca+ followed by Na+ > Mg2+ > K+ in the upper and middle Narmada basin. However, in LNB the cation concentration was considerably dominated by Na+ followed by Ca2+ > Mg2+ > K+. The mean value of Ca2+ in groundwater was found to be highest in MNB (80.82 ± 41.73 mg/L), followed by LNB (77.4 ± 41.88 mg/L) and lowest in UNB (59.37 ± 29.5 mg/L). The obtained results suggested that 2% of samples from UNB and 1.6% samples from LNB exceeded the permissible range of 200 mg/L of Ca in consumable water (BIS 2012). However, the Ca2+ values in UNB water samples were within the given standard. Sodium values were found to be highest in the LNB at 110.31 mg/L, followed by MNB at 54.27 mg/L and lowest in UNB at 50.19 mg/L; almost 17.6% (in LNB), 3.3% (in MNB) and 1.5% (in UNB) of groundwater samples surpassed the permissible standard of 200 mg/L of Na+ rcommended by WHO (WHO 2017). The values of Mg2+ ranged from 1.49 to 277 mg/L throughout the basin, with mean concentrations of 49.85, 38.15, and 32.22 mg/L in LNB, MNB, and UNB respectively. The observed values shows that 7.8%, 3.3%, and 1.5% of samples in LNB, MNB, and UNB respectively were above the recommended standard of 100 mg/L in groundwater (BIS 2012). The maximum value of K+ in groundwater was detected in MNB with mean value of 7.19 mg/L and lowest in UNB having mean value of 1.46 mg/L; 17.2% of MNB and 11.8% of LNB groundwater samples were found above the permissible limit of 12 mg/L for K+ concentration (WHO 2017) whereas, all the groundwater samples of UNB were within the permissible limit. The concentration of anions indicated the dominance of HCO3 followed by Cl > NO3 > SO42− > F in UNB and MNB. However, in LNB the anion concentration followed the trend of HCO3 > Cl > SO42− > NO3 > F. The values of HCO3 vary from 24 to 766 mg/L throughout the basin with an maximum average value of 366.24 mg/L in MNB, followed by an average value of 314 mg/L in LNB, whilst a minimum concentration of HCO3 was found in UNB with mean value of 297.91 mg/L. Cl concentration was detected to be maximum in LNB with mean value of 189.94 mg/L and had a minimum value of 63.42 mg/L in UNB. The value of SO42− ranged from 4 to 414.37 mg/L and was found to be highest in LNB. The Cl and SO42− concentrations of all the groundwater samples were below the permissible limt of <1,000 mg/L (Cl) and 400 mg/L (SO42−) in UNB and MNB. However, about 2% of water samples exceeded the prescribed permissible limits for Cl and SO42− concentration in LNB. The value of NO3 ranged from 1 to 250 mg/L and was found highest in MNB and lowest in UNB. Almost 48.4%, 23.5% and 16.7% of groundwater samples surpassed the prescribed range of 45 mg/L in MNB, LNB and UNB respectively (Figure 2). The major sources of nitrate in the groundwater are from leaching of applied fertilizers in agricultural fields, effluent leaching from landfills, and mobilization of nitrogen from soils by an ion-exchange process etc (Gao et al. 2020; Sarkar et al. 2021). F value varied from 0.01 to 3.7 mg/L throughout the basin, and was found highest in LNB with an average value of 0.87 mg/L whereas the lowest average value of 0.36 mg/L was found in UNB; about 6.6%, 5.9%, and 1.5% of groundwater samples were above the prescribed standard of 1.5 mg/L in MNB, LNB and UNB respectively (Figure 2). The alkaline conditions of high HCO3 concentration in the groundwater is liable for the dissolution of F in groundwater via water-rock-interaction (Rao et al. 1993). The weathering of fluoride bearing rocks like fluorite, apatite phosphate rocks is the major source of fluoride through leaching in the groundwater (Singh et al. 2019; Jha et al. 2021).
Figure 2

Spatial distribution of fluoride and nitrate concentration in groundwater.

Figure 2

Spatial distribution of fluoride and nitrate concentration in groundwater.

Close modal

Correlation analysis of groundwater samples

The correlation analysis of the major ions concentration is depicted in Figure 3. The correlation plot showed an approximately similar trend of correlation among several parameters in UNB, MNB and LNB. The strong and significant positive correlation of TDS with Mg2+ (UNB 0.74, MNB 0.7, LNB 0.84) implies that the groundwater is mostly influenced by Mg2+ salts. A strong correlation of TDS with HCO3 (UNB 0.83, MNB 0.77) was observed in the upper and middle basin, and the presence of the HCO3 ion may be due to silicate mineral weathering (Gastmans et al. 2010). A strong correlation between TDS and Cl (MNB 0.89, LNB 0.97), and between Na and TDS (MNB 0.63, LNB 0.91) in the respective sub-basins might be due to saltwater intrusion and its interference with groundwater quality (Egbueri et al. 2019). Strong positive correlation between Na and Cl (0.9), and Mg and Cl (0.81) in LNB confirms Cl ions in groundwater by dissolution of halite and dolomite respectively. A good correlation between Mg-HCO3 in UNB (0.72) and MNB (0.64) infers Mg2+ source in groundwater by dissolution of carbonate minerals.
Figure 3

Correlation matrices of different physicochemical parameters.

Figure 3

Correlation matrices of different physicochemical parameters.

Close modal

Hydrogeochemistry of groundwater

The lithology, climatic condition and movement of groundwater controls the chemical composition of subsurface water (Gopinath & Seralathan 2006). As a result, a hydrochemical scatter plot (Figure 4) developed by Chadha 1999 was used to comprehend the hydrochemical processes. The cations and anions present in the groundwater were represented as [(Ca2+ + Mg2+)-(Na+ + K+)] and [(CO3 + HCO3)-(SO42− + Cl) respectively. Chadha (1999), divided the graph into four quadrants: (1) recharging water (Ca-Mg-HCO3 type), (2) base ion-exchange water (Na-HCO3 type), (3) sea water (Na-Cl type), and (4) reverse ion-exchange water (Ca-Mg-Cl type) (Figure 4).
Figure 4

Representation of a Chadha plot showing different types of groundwater.

Figure 4

Representation of a Chadha plot showing different types of groundwater.

Close modal

From the scatter plot (Figure 4) we can observe most groundwater samples of NRB falling in the recharge water (Ca-Mg-HCO3 type) section, meaning that during the groundwater recharge process, it might have carried and dissolved the carbonates in HCO3 form and Ca2+ ions in groundwater. Few groundwater samples of UNB and MNB were influenced by base ion-exchange reactions i.e., dissolution of Na+ ions in water and precipitation of Ca2+ ions. About four LNB groundwater samples were found to be of seawater type which indicates the mixing of seawater and groundwater, which is generally restricted to the coastal region (Ravikumar & Somashekar 2017). However, NRB groundwater samples were also controlled by reverse-ion exchange mechanisms where Ca + Mg ion concentration exceeds Na + K concentration owing to release of Ca and Mg dissolution and subsequently precipitation of Na+ ions on minerals.

The classification of groundwater via base-exchange indices (r1) using Equation (2) categorized 84.85% of water samples as Na+-SO42− type (r1 < 1) and 15.15% of samples as Na+-HCO3 type (r1 > 1) in UNB. In MNB about 86.04% and 13.93% of water samples were of Na+-SO42− type and Na+-HCO3 type respectively. However, in LNB 94.12% of water samples was categorized as Na+-SO42− type (r1 < 1) and only 5.88% of water samples belonged to Na+-HCO3 type (r1 > 1). A meteoric genesis index (r2) was calculated to classify groundwater samples of NRB using Equation (3). The obtained results showed that 82.58%, 78.69% and 96.08% of groundwater samples belonged to deep meteoric water percolation type (i.e., r2 < 1) for UNB, MNB, and LNB respectively, whilst the rest of the groundwater samples i.e., 17.42% (UNB), 21.31% (MNB) and 3.92% (LNB) were categorized as shallow meteoric water percolation type. Therefore, results from Equations (2) and (3) infer that the basin is dominated by Na+-SO42− type and deep meteoric percolation type of groundwater.

The contributing source of ions in water can be identified through weathering and rock-water interaction processes (Figures 5, 6(a) and 6(b)). The Gibbs plot ratios of different ions i.e., Na +/(Na+ + Ca2+) and Cl/(Cl + HCO3) present in subsurface water were plotted against respective TDS values (Gibbs 1970). The Gibbs plots showed that 3.03% (in UNB), 11.47% (in MNB), and 29% (in LNB) of groundwater samples were in the evaporation dominated section and the remaining 96.97% (in UNB), 88.53% (in MNB), and 71% (in LNB) of groundwater samples were categorized in the rock dominated part (Figure 5). Therefore, The Gibbs plots indicated rock weathering as major and evaporation as minor responsible factors in governing the groundwater chemistry of NRB.
Figure 5

Identifying the source of ions in groundwater by Gibbs plot (a) TDS against Na/(Na + Ca) (b) TDS against Cl/(Cl + HCO3).

Figure 5

Identifying the source of ions in groundwater by Gibbs plot (a) TDS against Na/(Na + Ca) (b) TDS against Cl/(Cl + HCO3).

Close modal
Figure 6

Identification of different types of weathering through an end-member diagram plot (a) HCO3/Na+ against Ca2+/Na+ (b) Mg2+/Na+ against Ca2+/Na+.

Figure 6

Identification of different types of weathering through an end-member diagram plot (a) HCO3/Na+ against Ca2+/Na+ (b) Mg2+/Na+ against Ca2+/Na+.

Close modal

To examine the insights of lithology in weathering processes like silicate/carbonate/evaporate weathering were identified using different plots of HCO3/Na+ v/s Ca2+/Na+ and Mg2+/Na+ v/s Ca2+/Na+ also known as end-member diagrams. The plot of HCO3/Na+ v/s Ca2+/Na+ (Figure 6(a)) shows that silicate weathering was occurring in NRB, whereas, few sampling locations of LNB were governed by evaporate weathering. However, a plot of Mg2+/Na+ v/s Ca2+/Na+ reveals that the groundwater of NRB follows a silicate-evaporate mixing trend (Figure 6(b)), mainly governed by silicate weathering to attain thermodynamic equilibrium.

The SO42− v/s Ca2+ scatter plot indicated dissolution of gypsum or anhydrite. Since most NRB water samples were below the equiline (Figure 7(a)) i.e., reverse ion exchange process was dominant, Ca2+ ions enters the groundwater through dissolution and Na+ ions was mineralized on the rock surface.
Figure 7

Bi-plot of (a) Ca against SO4 (b) Ca + Mg against SO4 + HCO3 (c) HCO3 against Ca + Mg and (d) Na+ against Cl.

Figure 7

Bi-plot of (a) Ca against SO4 (b) Ca + Mg against SO4 + HCO3 (c) HCO3 against Ca + Mg and (d) Na+ against Cl.

Close modal

The SO42− + HCO3v/s Ca2+ + Mg2+ scatter plot (Figure 7(b)) showed that the groundwater samples of UNB, MNB and LNB fall both above and below the equiline, thus indicating that weathering of carbonate and silicate minerals is contributing HCO3 ions to the groundwater.

The scatter diagram between Ca2+ + Mg2+ v/s HCO3 (Figure 7(c)) confirmed dolomite and calcite as main sources of Ca2+ and Mg2+ ions in groundwater. However, around 27% of groundwater samples of LNB were above the equiline. Therefore, the source of Ca2+ and Mg2+ ions in subsurface water might be due to dissolution of minerals like gypsum and cation-exchange between Ca2+ and Na+ during dissolution and precipitation processes (El Mountassir et al. 2021).

A Ca2+/Mg2+ molar ratio as represented in Figure 8 reveals the dissolution of calcite and dolomite in groundwater: if Ca2+/Mg2+ ratio ≤ 1 (dissolution of dolomite rocks); 1 < Ca2+/Mg2+ ratio < 2 (dissolution of calcite rock); Ca2+/Mg2+ ratio > 2 (silicate weathering contributing Ca2+ and Mg2+). Overall 38.52%, 27.87% and 39.22% of groundwater samples of UNB, MNB, and LNB respectively indicated dissolution of dolomite. In the groundwater samples of NRB, maximum dissolution of calcite was found in MNB (42.62%) followed by LNB (41.18%) and UNB (31.15%). However, the contribution of Ca2+ and Mg2+ in groundwater by silicate weathering was found to be highest in UNB (30.33%) > MNB (29.51%) > LNB (19.60%). Thus, from the Ca2+/Mg2+ molar ratio we can conclude that dissolution of dolomite in UNB, and dissolution of calcite in both MNB and LNB, was influencing the composition of dissolved minerals.
Figure 8

Representation of Ca2+/Mg2+ molar ratio in groundwater.

Figure 8

Representation of Ca2+/Mg2+ molar ratio in groundwater.

Close modal

The obtained Na+ v/s Cl scatter plot (Figure 7(d)) clearly depicts the origin of Na+ ions in groundwater samples from halite dissolution. According to Loni et al. (2015), where Na+ ions are in abundant against Cl ion there is a possibility of silicate weathering being dominant during the weathering process. However, some proportion of samples were above the equiline which clearly showed the dominancy of Cl over Na+. In LNB, Cl dominance might be an indication of seawater intrusion ultimately influencing groundwater mineralization (El Mountassir et al. 2021).

The chemical properties of subsurface water is controlled by interaction with rock leading to ion-exchange processes. The chloro-alkaline indices (CAI-1 and 2) developed by Schoeller (1965), explain the base-ion exchange reactions between subsurface water and rock. The computed values of chloro-alkaline indices as shown in Table 2 and Figure 9, indicate that about 63.64%, 55.74% and 82.35% of groundwater samples were positive for UNB, MNB and LNB respectively. Thus 36.36%, 44.26% and 17.65% of groundwater samples were negative for UNB, MNB and LNB respectively. The obtained results show that both ‘reverse ion-exchange” and ‘direct ion-exchange” reactions were altering the chemical properties of subsurface water in the study area. Most groundwater samples had positive CAI values i.e., reverse ion occurrence being dominant with (Ca + Mg)-(Cl-SO4) type of groundwater facies. Positive CAI values indicate the presence of excess Cl ions against Na+ ions in groundwater, which might be from anthropogenic pollution sources (Zaidi et al. 2019).
Table 2

Classification of groundwater samples as direct or reverse ion exchange through CAI-1 and CAI-2 values

ParametersRange% of samples
UNBMNBLNB
CAI-1 (Schoeller, 1965)

 
< 0 (Direct ion exchange) 36.36% 44.26% 17.65% 

>0 (Reverse ion exchange) 

63.64% 

55.74% 

82.35% 
CAI-2 (Schoeller, 1965)

 
< 0 (Direct ion exchange) 36.36% 44.26% 17.65% 

> 0 (Reverse ion exchange) 

63.64% 

55.74% 

82.35% 
ParametersRange% of samples
UNBMNBLNB
CAI-1 (Schoeller, 1965)

 
< 0 (Direct ion exchange) 36.36% 44.26% 17.65% 

>0 (Reverse ion exchange) 

63.64% 

55.74% 

82.35% 
CAI-2 (Schoeller, 1965)

 
< 0 (Direct ion exchange) 36.36% 44.26% 17.65% 

> 0 (Reverse ion exchange) 

63.64% 

55.74% 

82.35% 

UNB, Upper Narmada Basin; MNB, Middle Narmada Basin; LNB, Lower Narmada Basin.

Figure 9

Biplot of CA1-1 and CAI-2.

Figure 9

Biplot of CA1-1 and CAI-2.

Close modal

Source identification and apportionment of minerals

The average SI values for carbonate minerals such as SIcalcite, SIdolomite and SIaragonite, respectively, were found to be 0.28, 0.52, and 0.002 for UNB; 0.3, 0.49, and 0.15 for MNB; and 0.65, 1.4, and 0.52 for LNB. Therefore, a plot of SIcalcite against TDS values, showed that 77.27%, 86.1% and 92.2% of water samples of UNB, MNB and LNB were oversaturated; a plot of SIdolomite against TDS, showed that 77.28%, 79.5% and 94.2% of water samples of UNB, MNB and LNB were oversaturated and a plot of SIaragonite against TDS values, showed that 62.87%, 69.68% and 94.2% of water samples of UNB, MNB and LNB were oversaturated, all due to carbonate minerals (Figure 10 and Table 3).
Figure 10

Plots of SIcalcite, SIdolomite, SIaragonite, SIanahydrite and SIgypsum, SIsylvite SIhalite and SIflourite against TDS values.

Figure 10

Plots of SIcalcite, SIdolomite, SIaragonite, SIanahydrite and SIgypsum, SIsylvite SIhalite and SIflourite against TDS values.

Close modal

The mean SI values for SO42− minerals i.e., SIanahydrite & SIgypsum were found to be −2.79 and −2.49, respectively, for UNB; −2.62 and −2.52 for MNB; −2.45 and −2.13 for LNB. The average SI values for potassium minerals, sodium mineral and flouride mineral such as SIsylvite, SIhalite and SIflourite were found to be −8.45, −7.2 and −2.43 for UNB, respectively; −7.96, −7 and −1.74 for MNB; and −7.85, −6.76 and −1.37 for LNB. Therefore, plots of SIanahydrite and SIgypsum against TDS values (Figure 10 and Table 3), showed that all groundwater samples was under saturated with sulphate minerals. Similarly, plots of SIsylvite SIhalite and SIflourite against TDS values (Figure 10 and Table 3), were found to be under saturated. Therefore, plots of SI of different minerals against TDS values point out minimal or absent soluble sulphate, potassium, and sodium or fluoride minerals. Carbonate minerals are said to be reactive minerals; therefore, oversaturation condition of carbonate minerals in the basin indicates regular dissolution of carbonate minerals with continuous contribution of Ca2+, Mg2+ and HCO3 ions in the groundwater (Batabyal 2018).

Table 3

Classification of groundwater samples as precipitation, dissolution, and equilibrium through saturation indices

% of sample
Upper Narmada Basin
Middle Narmada Basin
Lower Narmada Basin
SI > 0SI < 0SI = 0SI > 0SI < 0SI = 0SI > 0SI < 0SI = 0
Calcite 77.27% 21.96% 0.77% 86.1% 13.9% – 92.2% 5.8% 2% 
Dolomite 77.28% 22.72% – 79.5% 20.5% – 94.2% 5.8% – 
Aragonite 62.87% 36.36% 0.77% 69.68% 30.32% – 94.2% 5.8% – 
Anhydrite – 100% – 0.82% 99.18% – – 100% – 
Gypsum – 100% – – 100% – – 100% – 
Sylvite – 100% – – 100% – – 100% – 
Halite 0.77% 99.23% – – 100% – – 100% – 
Fluorite 0.77% 99.23% – 2.5% 97.5% – – 100% – 
% of sample
Upper Narmada Basin
Middle Narmada Basin
Lower Narmada Basin
SI > 0SI < 0SI = 0SI > 0SI < 0SI = 0SI > 0SI < 0SI = 0
Calcite 77.27% 21.96% 0.77% 86.1% 13.9% – 92.2% 5.8% 2% 
Dolomite 77.28% 22.72% – 79.5% 20.5% – 94.2% 5.8% – 
Aragonite 62.87% 36.36% 0.77% 69.68% 30.32% – 94.2% 5.8% – 
Anhydrite – 100% – 0.82% 99.18% – – 100% – 
Gypsum – 100% – – 100% – – 100% – 
Sylvite – 100% – – 100% – – 100% – 
Halite 0.77% 99.23% – – 100% – – 100% – 
Fluorite 0.77% 99.23% – 2.5% 97.5% – – 100% – 

SI > 0 (precipitation); SI < 0 (dissolution); SI = 0 (equilibrium).

Appraisal of groundwater quality for drinking purposes using a water quality index (WQI)

To compute the WQI, the relative weights (RWi) of the major chemical parameters were calculated and shown in Table 4, and further processed using Equations (6) and (7). WQI values were categorized as shown in Table 5, and spatial variation of the WQI is shown in Figure 11. The obtained results for the WQI indicated that in UNB, the WQI values varied from 27.75 to 195.4 with an average value of 71.82; WQI values in MNB varied from 36.87 to 187.8 with an average value of 97; and in LNB, WQI values varied from 44.94 to 289.5 with a mean value of 103.71. As per the WQI categories, in UNB the WQI values were excellent (23.5%), good (63.6%), and poor (12.9%); in MNB the WQI values were excellent (54.1%), good (52.1%), and poor (43.8%); whereas in LNB the WQI values were in the excellent (9.85%), good (49%), poor (37.3%) and very poor (3.9%) categories. The high WQI values were influenced by high concentrations of parameters such as TDS, EC, F, NO3, and HCO3 etc. (Yadav et al. 2018). Therefore, the groundwater samples whose WQI values were greater than 100 were not fit for consumption. Thus, their proper treatment is required before use for drinking purposes.
Table 4

Relative weights of major parameters

Chemical parametersAcceptable limits (BIS 2012; WHO 2017)Weight (wi)Relative weight (RWi)
pH 8.5 0.114285714 
TDS (mg/L) 500 0.085714286 
Ca2+ (mg/L) 75 0.057142857 
Mg2+ (mg/L) 30 0.057142857 
Na+ (mg/L) 200 0.057142857 
K+ (mg/L) 12 0.057142857 
HCO3 (mg/L) 120 0.085714286 
Cl (mg/L) 250 0.114285714 
SO42− (mg/L) 200 0.085714286 
NO3 (mg/L) 45 0.142857143 
F (mg/L) 0.142857143 
  ∑wi = 35 ∑RWi = 1 
Chemical parametersAcceptable limits (BIS 2012; WHO 2017)Weight (wi)Relative weight (RWi)
pH 8.5 0.114285714 
TDS (mg/L) 500 0.085714286 
Ca2+ (mg/L) 75 0.057142857 
Mg2+ (mg/L) 30 0.057142857 
Na+ (mg/L) 200 0.057142857 
K+ (mg/L) 12 0.057142857 
HCO3 (mg/L) 120 0.085714286 
Cl (mg/L) 250 0.114285714 
SO42− (mg/L) 200 0.085714286 
NO3 (mg/L) 45 0.142857143 
F (mg/L) 0.142857143 
  ∑wi = 35 ∑RWi = 1 
Table 5

Classification of ground water quality based on WQI values

Water Quality Index (WQI)
% of sample
RangeWater classUNBMNBLNB
(WQI < 50) excellent 23.5% 4.1% 9.85% 
(50 < WQI < 100) good 63.6% 52.1% 49.0% 
(100 < WQI < 200) poor 12.9% 43.8% 37.3% 
(200 < WQI < 300) very poor – – 3.9% 
Water Quality Index (WQI)
% of sample
RangeWater classUNBMNBLNB
(WQI < 50) excellent 23.5% 4.1% 9.85% 
(50 < WQI < 100) good 63.6% 52.1% 49.0% 
(100 < WQI < 200) poor 12.9% 43.8% 37.3% 
(200 < WQI < 300) very poor – – 3.9% 

UNB, Upper Narmada Basin; MNB, Middle Narmada Basin; LNB, Lower Narmada Basin.

Figure 11

Spatial distribution of WQI throughout the Narmada River Basin.

Figure 11

Spatial distribution of WQI throughout the Narmada River Basin.

Close modal

Human health risk assessment due to presence of NO3 and F

The presence of NO3 and F in groundwater at higher concentration poses a threat to the people of developing and under-developed nations because groundwater is their primary source for drinking. The consumption of groundwater contaminated with NO3 and F leads to health risks (USEPA 2014). Since, the study area is contaminated with NO3 and F as shown in Figure 2, we computed the non-carcinogenic human health risk for children (age < 12) and adults (age > 19) through a drinking water ingestion pathway. Further, we integrated a total hazard index (THI) (Figure 12) for these groups of children and adults.

The obtained results after computing THI are represented in Tables 6 and 7 and Figure 12. The THI values for children ranged from 0.12 to 3.91 with an average value of 1.26, 0.32 to 6.98 with an average value of 2.22, and 0.20 to 8.25 with an average value of 2.07 in UNB, MNB and LNB respectively. Similarly the THI values for adults in UNB ranged from 0.11 to 3.54 with a mean value of 1.14, in from 0.29 to 6.33 with an average value of 2.02 in MNB, and from 0.18 to 7.48 with an mean value of 1.88 in LNB.
Table 6

Total Hazard Index (THI) indicting non-carcinogenic health risk (CR) and no risk (NR) among different population groups (adults and children)

Percentage of sample
Upper Narmada Basin
Middle Narmada Basin
Lower Narmada Basin
< 1 (NR)> 1 (CR)< 1 (NR)> 1 (CR)< 1 (NR)> 1 (CR)
THI: Adult (age > 19) 50.8% 49.2% 17.2% 82.8% 23.5% 76.5% 
THI: Children (age < 12) 46.2% 53.8% 13.9% 86.1% 19.6% 80.4% 
Percentage of sample
Upper Narmada Basin
Middle Narmada Basin
Lower Narmada Basin
< 1 (NR)> 1 (CR)< 1 (NR)> 1 (CR)< 1 (NR)> 1 (CR)
THI: Adult (age > 19) 50.8% 49.2% 17.2% 82.8% 23.5% 76.5% 
THI: Children (age < 12) 46.2% 53.8% 13.9% 86.1% 19.6% 80.4% 

USEPA (2014), THI > 1 indicates non-carcinogenic health risk (CR) to its consumers whereas THI < 1 confirms no risk (NR).

Table 7

Descriptive statistics of CDI (chronic daily intake) and HQ (hazard quotient) for NO3 and F-, and THI (total hazard index)

Nitrate
Fluoride
THI
CDI
HQ
CDI
HQ
AdultsChildrenAdultsChildrenAdultsChildrenAdultsChildrenAdultsChildren
UNB Min 0.04 0.05 0.03 0.03 0.00 0.00 0.01 0.01 0.11 0.12 
Max 4.95 5.46 3.10 3.41 0.09 0.10 2.36 2.60 3.54 3.91 
Average 1.20 1.32 0.75 0.83 0.02 0.02 0.39 0.43 1.14 1.26 
MNB Min 0.09 0.10 0.05 0.06 0.00 0.00 0.09 0.10 0.29 0.32 
Max 9.35 10.31 5.84 6.44 0.16 0.18 3.98 4.39 6.33 6.98 
Average 2.25 2.48 1.40 1.55 0.02 0.03 0.61 0.68 2.02 2.22 
LNB Min 0.09 0.10 0.05 0.06 0.01 0.01 0.13 0.14 0.18 0.20 
max 10.77 11.88 6.73 7.42 0.13 0.14 3.23 3.56 7.48 8.25 
average 1.50 1.66 0.94 1.04 0.04 0.04 0.94 1.03 1.88 2.07 
Nitrate
Fluoride
THI
CDI
HQ
CDI
HQ
AdultsChildrenAdultsChildrenAdultsChildrenAdultsChildrenAdultsChildren
UNB Min 0.04 0.05 0.03 0.03 0.00 0.00 0.01 0.01 0.11 0.12 
Max 4.95 5.46 3.10 3.41 0.09 0.10 2.36 2.60 3.54 3.91 
Average 1.20 1.32 0.75 0.83 0.02 0.02 0.39 0.43 1.14 1.26 
MNB Min 0.09 0.10 0.05 0.06 0.00 0.00 0.09 0.10 0.29 0.32 
Max 9.35 10.31 5.84 6.44 0.16 0.18 3.98 4.39 6.33 6.98 
Average 2.25 2.48 1.40 1.55 0.02 0.03 0.61 0.68 2.02 2.22 
LNB Min 0.09 0.10 0.05 0.06 0.01 0.01 0.13 0.14 0.18 0.20 
max 10.77 11.88 6.73 7.42 0.13 0.14 3.23 3.56 7.48 8.25 
average 1.50 1.66 0.94 1.04 0.04 0.04 0.94 1.03 1.88 2.07 

USEPA (2014), THI > 1 indicates non-carcinogenic health risk whereas THI < 1 confirms no risk.

Note: Adults (age > 19); Children (age < 12).

UNB, Upper Narmada Basin; MNB, Middle Narmada Basin; LNB, Lower Narmada Basin.

Table 8

Classification of groundwater quality through different indices for irrigation purposes

ParametersRangeWater class% of sample
UNBMNBLNB
Na% (Maharana et al. 2015)

 
<20 Excellent 18.9% 14.8% 21.6% 
20-40 Good 44.7% 45.1% 35.3% 
40-60 Permissible 30.3% 27.9% 25.5% 
60-80 Doubtful 4.5% 10.7% 15.7% 
80-100 Bad 1.5% 1.6% 2% 
SAR (USSL 1954)

 
<10 Excellent 76.5% 64.8% 54.9% 
10-18 Good 18.2% 24.6% 21.6% 
18-26 Doubtful 3% 5.7% 11.8% 
>26 Unsuitable 2.3% 4.9% 11.8% 
Kelly's index (Ki) (Kelly 1963)

 
Ki < 1 Suitable 84.1% 83.6% 70.6% 

Ki > 1 

Unsuitable 

15.9% 

16.4% 

29.4% 
Permeability index (Pi) (Doneen 1964)

 
Pi < 60% Suitable 52.27% 67.21% 62.75% 

Pi > 60% 

Unsuitable 

47.72% 

32.79% 

37.25% 
Magnesium ratio (MAR) (Paliwal 1972)

 
MAR < 50 Suitable 82.6% 89.3% 76.5% 

MAR > 50 

Unsuitable 

17.4% 

10.7% 

23.5% 
RSC (USSL 1954)

 
<1.25 Good 90.9% 89.3% 96% 
1.25 < RSC > 2.5 Marginally suitable 6.1% 9% 2% 
>2.5 Unsuitable 3% 1.6% 2% 
ParametersRangeWater class% of sample
UNBMNBLNB
Na% (Maharana et al. 2015)

 
<20 Excellent 18.9% 14.8% 21.6% 
20-40 Good 44.7% 45.1% 35.3% 
40-60 Permissible 30.3% 27.9% 25.5% 
60-80 Doubtful 4.5% 10.7% 15.7% 
80-100 Bad 1.5% 1.6% 2% 
SAR (USSL 1954)

 
<10 Excellent 76.5% 64.8% 54.9% 
10-18 Good 18.2% 24.6% 21.6% 
18-26 Doubtful 3% 5.7% 11.8% 
>26 Unsuitable 2.3% 4.9% 11.8% 
Kelly's index (Ki) (Kelly 1963)

 
Ki < 1 Suitable 84.1% 83.6% 70.6% 

Ki > 1 

Unsuitable 

15.9% 

16.4% 

29.4% 
Permeability index (Pi) (Doneen 1964)

 
Pi < 60% Suitable 52.27% 67.21% 62.75% 

Pi > 60% 

Unsuitable 

47.72% 

32.79% 

37.25% 
Magnesium ratio (MAR) (Paliwal 1972)

 
MAR < 50 Suitable 82.6% 89.3% 76.5% 

MAR > 50 

Unsuitable 

17.4% 

10.7% 

23.5% 
RSC (USSL 1954)

 
<1.25 Good 90.9% 89.3% 96% 
1.25 < RSC > 2.5 Marginally suitable 6.1% 9% 2% 
>2.5 Unsuitable 3% 1.6% 2% 

UNB, Upper Narmada Basin; MNB, Middle Narmada Basin; LNB, Lower Narmada Basin.

Figure 12

Spatial distribution of total hazard index (THI) for different age groups along the basin.

Figure 12

Spatial distribution of total hazard index (THI) for different age groups along the basin.

Close modal

Over 53.8%, 86.1% and 80.4% of groundwater samples of UNB, MNB and LNB, respectively, were greater than the acceptable limit (>1) and found to cause non-carcinogenic risk for children, whilst for adults, about 49.2%, 82.8% and 76.5% of groundwater samples of UNB, MNB and LNB, respectively, were identified to cause non-carcinogenic risk due to ingestion of groundwater. Tables 6 and 7 reveal that the THI risk assessment reveals more adverse health effects on children compared with adults.

Appropriateness of groundwater for irrigation and agriculture needs

To discern the appropriateness of groundwater for irrigation, six indices were calculated: Na%, SAR, Ki, Pi, MAR and RSC. The results of these indices are given in Table 8.

According to USSL (1954) classification, SAR categorised water class from excellent to unsuitable (Table 8) for irrigation purposes. In UNB, SAR values ranged from 1.3 to 86.57 with an average value of 8.24. In MNB, SAR values ranged from 1.4 to 62.87 with a mean value of 9.92. In LNB, SAR values ranged from 0.23 to 69.9 with a mean value of 13.3. The biplot of SAR against EC also known as United States Salinity Laboratory Staff (USSL's) diagram (USSL 1954), as shown in Figure 13(a), revealed that maximum samples of UNB and MNB were concentrated in C2S1 (low to medium), C2S2 (medium), C3S1 (low to high), and C3S2 (medium to high). However, few groundwater samples of LNB fall in the C4S4 (very high) category thus it is not fit for irrigation use.
Figure 13

(a) USSL's diagram, (b) Wilcox diagram, representing the suitability of the groundwater samples for irrigation use.

Figure 13

(a) USSL's diagram, (b) Wilcox diagram, representing the suitability of the groundwater samples for irrigation use.

Close modal

In UNB, Na% values ranged from 10.58 to 92.12 with a mean value of 35.68; in MNB they ranged from 7 to 88.15 with an average value of 37.6; and in LNB, Na% values ranged from 1.46 to 87.47 with an average value of 38.33.

The groundwater samples are identified for agriculture purposes with Na% falling in excellent to bad categories (Table 8). The biplot of Na% against EC was plotted to generate a Wilcox plot (Wilcox 1955) (Figure 13(b)) to identify the water usage for irrigation needs. The Wilcox plot showed that mostly groundwater samples were in a excellent to good water quality category and only a few water samples of MNB and LNB were in a doubtful to unsuitable region.

Kelly (1963) classified groundwater into two categories: if Ki < 1 then water is suitable but when Ki > 1 then water is considered to be unsuitable for irrigation use due to it being an alkali hazard (Table 8). In UNB, Ki values ranged from 0.09 to 11.68 with an average value of 0.74; in MNB, Ki values ranged from 0.05 to 7.3 with a mean value of 0.74; and Ki values ranged from 0.01 to 6.9 with a mean value of 0.88 in LNB. As shown in Table 8, 84.1%, 83.6% and 70.6% of groundwater samples of UNB, MNB, and LNB, respectively, were found to be suitable for irrigation use.

Doneen (1964) developed a formula (Table 8) to produce a permeability index (Pi) to understand the mobility capacity of groundwater in soil. Pi was categorised into two classes i.e., Pi < 60% (suitable) and Pi > 60% (unsuitable). In UNB, Pi values ranged from 25 to 120 with an average value of 58, whereas in MNB Pi values ranged from 19.13 to 107.6 with an average value of 55.29, and in LNB, Pi values ranged from 22.35 to 100.67 with an average value of 53.29. As shown in Table 8, 52.27%, 67.21%, and 62.75% of groundwaterwater samples of UNB, MNB, and LNB, respectively, were found to be fit for irrigation.

The presence of Mg2+ ions in high concentration degrade the soil quality, which ultimately promotes low yield. Therefore, it is very important to determine magnesium hazards. MAR was calculated according to the formula given in Table 8: if the MAR value of water samples is less than 50 then it is said to be suitable and when it exceeds 50 then the water is considered to be unfit for irrigation purposes. In UNB, MAR values ranged from 5.05 to 69.9 with average value of 33.54, and 82.6% of groundwater samples fell in the suitable category. The MAR values of groundwater samples in MNB ranged from 2.01 to 64.9 with an average value of 31.9 and 89.3% of water samples being categorised suitable. In LNB, MAR values ranged from 4.76 to 85.23 with an average value of 36.38 and only 76.5% of groundwater samples were found to be suitable for agricultural purpose.

Residual Sodium Carbonate (RSC) refers to sodium carbonate and sodium bicarbonate present in water. Therefore, a high concentration of carbonate and bicarbonate in water leads to the precipitation of calcium and magnesium in the soil, thus altering soil structure. According to USSL (1954), RSC can be calculated using the formula given in Table 8, and it is classified into three categories i.e., RSC < 1.25 (good), 1.25 < RSC > 2.5 (marginally suitable) and RSC > 2.5 (unsuitable). In UNB, RSC values ranged from 0.3 to 6 with an average value of 1 and about 90.9% of groundwater was categorised in the good category. In MNB, the RSC values ranged from 0.06 to 4.9 with an average an value of 0.9 and around 89.3% of groundwater samples were categorised in the good category. However, in LNB, most groundwater samples (96%) ranged in the good category with RSC values ranging from 0.13 to 3.4 with an average value of 0.8.

This is the first research of its kind which covers the hydrochemical study, human health risk assessment, geochemical modelling, water quality index, and irrigation water suitability of groundwater in the Narmada River Basin. The hydrochemical composition of subsurface water in the NRB is driven by natural processes such as rock weathering, as well as some anthropogenic activities. The results show that most water quality parameters exceeded the permissible limits. An overall high WQI value was present in the study area indicating the water to be unsuitable for human consumption. Human health risk assessments concluded that there was a ‘non-carcinogenic’ risk due to consumption of groundwater. However, groundwater is suitable for irrigation uses. This study highlights the need for detailed proper planning and policy making for water resources management. Thus, the current study recommends the treatment of polluted water before its consumption. Additionally, the government and policymakers of Madhya Pradesh should create awareness, and set up and supply treated water for drinking purposes. In general, this study will assist in the prevention of health risks, and also safeguard human well being, environmental sustainability and prosperity.

The authors wish to extend their gratitude towards India Water Resources Information System (India -WRIS) for providing the online data on their official website.

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

The authors declare there is no conflict of interest.

Abbasi
S. A.
2002
Water Quality Indices. State of the Art Report, National Institute of Hydrology, Scientific Contribution No. INCOH/SAR-25/2002. INCOH, 73, Roorkee
.
Akkoyunlu
A.
&
Akiner
M. E.
2012
Pollution evaluation in streams using water quality indices: a case study from Turkey's Sapanca Lake Basin
.
Ecological Indicators
18
,
501
511
.
https://doi.org/10.1016/j.ecolind.2011.12.018
.
Barbieri
M.
,
Ricolfi
L.
,
Vitale
S.
,
Muteto
P. V.
,
Nigro
A.
&
Sappa
G.
2019
Assessment of groundwater quality in the buffer zone of Limpopo National Park, Gaza Province, Southern Mozambique
.
Environmental Science and Pollution Research
26
(
1
),
62
77
.
https://doi.org/10.1007/s11356-018-3474-0
.
Barbieri
M.
,
Barberio
M. D.
,
Banzato
F.
,
Billi
A.
,
Boschetti
T.
,
Franchini
S.
,
Gori
F.
&
Petitta
M.
2021
Climate change and its effect on groundwater quality
.
Environmental Geochemistry and Health
1
12
.
https://doi.org/10.1007/s10653-021-01140-5
.
Batabyal
A. K.
2018
Hydrogeochemistry and quality of groundwater in a part of Damodar Valley, Eastern India: an integrated geochemical and statistical approach
.
Stochastic Environmental Research and Risk Assessment
32
(
8
),
2351
2368
.
https://doi.org/10.1007/s00477-018-1552-y
.
Bureau of Indian Standards (BIS)
2012
Drinking Water – Specification (Second Revision). ICS 13.060.20. BIS, New Delhi, India
.
CGWB
2017
Groundwater Scenario of India
.
Report Central Groundwater Board (CGWB), Ministry of Water Resources, Government of India
.
CGWB
2018
Groundwater Scenario of India
.
Report Central Groundwater Board (CGWB), Ministry of Water Resources, Government of India
.
Chadha
D. K.
1999
A proposed new diagram for geochemical classification of natural waters and interpretation of chemical data
.
Hydrogeology Journal
7
(
5
),
431
439
.
https://doi.org/10.1007/s100400050216
.
Doneen
L. D.
1964
Notes on Water Quality in Agriculture. Water Science and Engineering, Paper 4001, Department of Water Sciences and Engineering, University of California, Davis
.
Egbueri
J. C.
,
Mgbenu
C. N.
&
Chukwu
C. N.
2019
Investigating the hydrogeochemical processes and quality of water resources in Ojoto and environs using integrated classical methods
.
Modeling Earth Systems and Environment
5
(
4
),
1443
1461
.
https://doi.org/10.1007/s40808-019-00613-y
.
El Mountassir
O.
,
Bahir
M.
,
Ouazar
D.
,
Chehbouni
A.
&
Carreira
P. M.
2021
Temporal and spatial assessment of groundwater contamination with nitrate using nitrate pollution index (NPI), groundwater pollution index (GPI), and GIS (case study: Essaouira basin, Morocco)
.
Environmental Science and Pollution Research
1
18
.
https://doi.org/10.1007/s10653-020-00546-x
.
Gao
Y.
,
Qian
H.
,
Ren
W.
,
Wang
H.
,
Liu
F.
&
Yang
F.
2020
Hydrogeochemical characterization and quality assessment of groundwater based on integrated-weight water quality index in a concentrated urban area
.
Journal of Cleaner Production
260
,
121006
.
https://doi.org/10.1016/j.jclepro.2020.121006
.
Gastmans
D.
,
Chang
H. K.
&
Hutcheon
I.
2010
Groundwater geochemical evolution in the northern portion of the guarani aquifer system (Brazil) and its relationship to diagenetic features
.
Applied Geochemistry
25
(
1
),
16
33
.
https://doi.org/10.1016/j.apgeochem.2009.09.024
.
Gawle
S.
,
Pateria
K.
&
Mishra
R. P.
2021
Physico-chemical analysis of groundwater during monsoon and winter season of Dindori district, India
.
Groundwater for Sustainable Development
12
,
100550
.
https://doi.org/10.1016/j.gsd.2021.100550
.
Gibbs
R. J.
1970
Mechanisms controlling world water chemistry
.
Science
170
(
3962
),
1088
1090
.
https://doi.org/10.1126/science.170.3962.1088
.
Gnanachandrasamy
G.
,
Dushiyanthan
C.
,
Jeyavel Rajakumar
T.
&
Zhou
Y.
2020
Assessment of hydrogeochemical characteristics of groundwater in the lower Vellar river basin: using Geographical Information System (GIS) and Water Quality Index (WQI)
.
Environment, Development and Sustainability
22
(
2
),
759
789
.
https://doi.org/10.1007/s10668-018-0219-7
.
Gopinath
G.
&
Seralathan
P.
2006
Chemistry of groundwater in the laterite formations of the Muvattupuzha River Basin, Kerala
.
Journal of Geological Society of India
(Online archive from Vol 1 to Vol 78)
68
(
4
),
705
714
.
Gupta
D.
,
Shukla
R.
,
Barya
M. P.
,
Singh
G.
&
Mishra
V. K.
2020a
Appraisal of river water quality based on field observations: a case study on Narmada river
.
Indian Journal of Ecology
47
(
4
),
897
901
.
Gupta
D.
,
Shukla
R.
,
Barya
M. P.
,
Singh
G.
&
Mishra
V. K.
2020b
Water quality assessment of Narmada River along the different topographical regions of the central India
.
Water Science
34
(
1
),
202
212
.
https://doi.org/10.1080/11104929.2020.1839345
.
Gupta
D.
,
Kaushik
S.
,
Shukla
R.
&
Mishra
V. K.
2022
Mechanisms controlling major ion chemistry and its suitability for irrigation of Narmada River, India
.
Water Supply
22
(
3
),
3224
3241
.
https://doi.org/10.2166/ws.2021.408
.
Haghnazar
H.
,
Johannesson
K. H.
,
González-Pinzón
R.
,
Pourakbar
M.
,
Aghayani
E.
,
Rajabi
A.
&
Hashemi
A. A.
2022
Groundwater geochemistry, quality, and pollution of the largest lake basin in the Middle East: comparison of PMF and PCA-MLR receptor models and application of the source-oriented HHRA approach
.
Chemosphere
288
,
132489
.
https://doi.org/10.1016/j.chemosphere.2021.132489
.
Jha
S.
,
Sinha
S.
&
Hazra
S.
2021
Hydrochemical evolution and assessment of groundwater quality in fluorosis-affected area, Mandla District, central India
.
Groundwater for Sustainable Development
14
,
100614
.
https://doi.org/10.1016/j.gsd.2021.100614
.
Kaviarasan
M.
,
Geetha
P.
&
Soman
K. P.
2016
GIS-based ground water quality monitoring in Thiruvannamalai District, Tamil Nadu, India
. In:
Proceedings of the International Conference on Soft Computing Systems
.
Springer
,
New Delhi
, pp.
685
700
.
https://doi.org/10.1007/978-81-322-2671-0_65.
Kelly
W. P.
1963
Use of saline irrigation water
.
Soil Science
95
(
6
),
385
391
.
Khosravi
H.
,
Karimi
K.
,
Nakhaee Nezhad Fard
S.
&
Mesbahzadeh
T.
2016
Investigation of spatial structure of groundwater quality using geostatistical approach in Mehran Plain, Iran
.
Pollution
2
(
1
),
57
65
.
Kumar
M. D.
,
Ghosh
S.
,
Singh
O. P.
,
Ranade
R.
&
Ravindranath
R.
2005
Changes in groundwater ecology and its implication for surface flows: studies from Narmada River Basin, Madhya Pradesh, India
.
IWMI-Tata Water Policy Research Highlight
.
Kumar
S.
,
Kumar
A.
,
Prashant
V. N. J.
,
Sahood
S. K.
&
Ranjan
R. K.
2022
Groundwater quality and its suitability for drinking and irrigational purpose in Bhojpur district: middle Gangetic plain of Bihar, India
.
Water Supply
22 (9), 7072–7084. https://doi.org/10.2166/ws.2022.317
.
Logeshkumaran
A.
,
Magesh
N. S.
,
Godson
P. S.
&
Chandrasekar
N.
2015
Hydro-geochemistry and application of water quality index (WQI) for groundwater quality assessment, Anna Nagar, part of Chennai City, Tamil Nadu, India
.
Applied Water Science
5
(
4
),
335
343
.
https://doi.org/10.1007/s13201-014-0196-4
.
Loni
O. A.
,
Zaidi
F. K.
,
Alhumimidi
M. S.
,
Alharbi
O. A.
,
Hussein
M. T.
,
Dafalla
M.
,
AlYousef
K. A.
&
Kassem
O. M.
2015
Evaluation of groundwater quality in an evaporation dominant arid environment; a case study from Al Asyah area in Saudi Arabia
.
Arabian Journal of Geosciences
8
(
8
),
6237
6247
.
http://dx.doi.org/10.1007/ s12517-014-1623-4
.
Maharana
C.
,
Gautam
S. K.
,
Singh
A. K.
&
Tripathi
J. K.
2015
Major ion chemistry of the Son River, India: weathering processes, dissolved fluxes and water quality assessment
.
Journal of Earth System Science
124
(
6
),
1293
1309
.
Matta
G.
,
Srivastava
S.
,
Pandey
R. R.
&
Saini
K. K.
2017
Assessment of physicochemical characteristics of Ganga Canal water quality in Uttarakhand
.
Environment, Development and Sustainability
19
(
2
),
419
431
.
https://doi.org/10.1007/s10668-015-9735-x
.
McArthur
J. M.
,
Nath
B.
,
Banerjee
D. M.
,
Purohit
R.
&
Grassineau
N.
2011
Palaeosol control on groundwater flow and pollutant distribution: the example of arsenic
.
Environmental Science & Technology
45
(
4
),
1376
1383
.
https://doi.org/10.1021/es1032376
.
Mondal
N. C.
,
Singh
V. S.
,
Puranik
S. C.
&
Singh
V. P.
2010
Trace element concentration in groundwater of Pesarlanka Island, Krishna Delta, India
.
Environmental Monitoring and Assessment
163
(
1
),
215
227
.
https://doi.org/10.1007/s10661-009-0828-6
.
Narany
T. S.
,
Ramli
M. F.
,
Aris
A. Z.
,
Sulaiman
W. N. A.
&
Fakharian
K.
2015
Groundwater irrigation quality mapping using geostatistical techniques in Amol–Babol Plain, Iran
.
Arabian Journal of Geosciences
8
(
2
),
961
976
.
https://doi.org/10.1007/s12517-014-1271-8
.
Paliwal
K. V.
1972
Irrigation with Saline Water, Monogram No. 2 (New Series)
.
IARI
,
New Delhi
.
Rao
N. S.
,
Sunitha
B.
,
Adimalla
N.
&
Chaudhary
M.
2020
Quality criteria for groundwater use from a rural part of Wanaparthy District, Telangana State, India, through ionic spatial distribution (ISD), entropy water quality index (EWQI) and principal component analysis (PCA)
.
Environmental Geochemistry and Health
42
(
2
),
579
599
.
https://doi.org/10.1007/s10653-019-00393-5
.
Rao
N. S.
,
Sunitha
B.
,
Rambabu
R.
,
Rao
P. V.
,
Rao
P. S.
,
Spandana
B. D.
,
Sravanthi
M.
&
Marghade
D.
2018
Quality and degree of pollution of groundwater, using PIG from a rural part of Telangana State, India
.
Applied Water Science
8
(
8
),
1
13
.
https://doi.org/10.1007/s13201-018-0864-x
.
Rao
N. V. R.
,
Rao
N.
,
Surya Prakash Rao
K.
&
Schuiling
R. D.
1993
Fluorine distribution in waters of Nalgonda district, Andhra Pradesh, India
.
Environmental Geology
21
(
1
),
84
89
.
https://doi.org/10.1007/BF00775055
.
Ravikumar
P.
&
Somashekar
R. K.
2017
Principal component analysis and hydrochemical facies characterization to evaluate groundwater quality in Varahi river basin, Karnataka state, India
.
Applied Water Science
7
(
2
),
745
755
.
https://doi.org/10.1007/s13201-015-0287-x
.
Sarkar
S.
,
Mukherjee
A.
,
Duttagupta
S.
,
Bhanja
S. N.
,
Bhattacharya
A.
&
Chakraborty
S.
2021
Vulnerability of groundwater from elevated nitrate pollution across India: insights from spatio-temporal patterns using large-scale monitoring data
.
Journal of Contaminant Hydrology
243
,
103895
.
https://doi.org/10.1016/j.jconhyd.2021.103895
.
Schoeller
H.
1965
Geochemistry of groundwater: An international guide for research and practice
.
UNESCO, Paris, France
;
Chapter 15, 1–18
.
Şener
Ş.
,
Şener
E.
&
Davraz
A.
2017
Assessment of groundwater quality and health risk in drinking water basin using GIS
.
Journal of Water and Health
15
(
1
),
112
132
.
https://doi.org/10.2166/wh.2016.148
.
Shukla
S.
&
Saxena
A.
2021
Appraisal of groundwater quality with human health risk assessment in parts of Indo-Gangetic alluvial plain, North India
.
Archives of Environmental Contamination and Toxicology
80
(
1
),
55
73
.
https://doi.org/10.1007/s00244-020-00771-6
.
Singh
C. K.
,
Kumar
A.
,
Shashtri
S.
,
Kumar
A.
,
Kumar
P.
&
Mallick
J.
2017
Multivariate statistical analysis and geochemical modeling for geochemical assessment of groundwater of Delhi, India
.
Journal of Geochemical Exploration
175
,
59
71
.
https://doi.org/10.1016/j.gexplo.2017.01.001
.
Singh
R.
,
Deolia
D. K.
&
Tignath
S.
2019
A comparative study on usage of piper and durov diagrams to interpret hydrochemical processes in groundwater of Bichhiya Tehsil, Mandla District, Madhya Pradesh, India
.
Research Journal of Recent Sciences
8
(
2
),
44
51
.
USEPA
2004
USEPA Claims to Meet Drinking Water Goals Despite Persistent Data Quality Shortcomings. Report no. 2004-P-0008
.
USEPA
2014
Surface water standards and guidance
. In:
USEPA (eds)
,
Water Quality Standards Handbook
.
US Environmental Protection Agency
,
Washington, DC.
USSL (United States Salinity Laboratory)
1954
Diagnosis and improvement of saline and alkali soils
. In:
US Department of Agriculture
,
Agricultural Handbook, Vol. 60
.
US Government Printing Office
,
Washington
. p.
160
.
WHO
2017
Guidelines for Drinking Water Quality, 4th Edition Incorporating the First Addendum
.
World Health Organization
,
Geneva
,
Switzerland
.
Wilcox
L. V.
1955
Classification and Use of Irrigation Waters
.
Circular 969. US Department of Agriculture (USDA)
,
Washington, DC
,
USA
.
Zaidi
F. K.
,
Salman
A.
,
Hag-Elsafi
S.
&
Alfaifi
H. J.
2019
Assessment of hydrological processes operating in a multi-layered sedimentary aquifer system in Saudi Arabia using integrated chemical and statistical approach
.
Environmental Monitoring and Assessment
191
(
7
),
1
18
.
https://doi.org/10.1007/s10661-019-7597-7
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).