The results of descriptive statistical analysis on the quality data are shown in Fig. 2. The pH of the groundwater ranged from 3.06 to 8.35, with an average value of 7.08, suggesting that it is acidic to alkaline in nature (SI Table 1. Statistics for analysed parameter). Groundwater samples may be categorized as acidic (pH < 6.5, 12%, N = 7), neutral (pH 6.5–7.5, 44%, N = 25), or alkaline (pH > 7.5, 44%, N = 25). Groundwater samples with high acidic pH include S6 (3.06), S16 (4.40), S43 (4.00), S47 (4.10), and S48 (3.60). The lower pH might be attributed to the high rain fall and low water table condition, which favors slightly acidic water in shallow aquifer systems, especially lateritic terrains like the samples above12,13,22,24). The EC ranged from 50 to 970 μS/cm, with an average of 266 μS/cm. EC < 250 μS/cm is found in around 67% of the samples (N = 38). TDS was determined from EC using the formula TDS (mg/L) = EC (µS/cm) × 0.64, and the average value ranged from 32 to 621 mg/l. Results showed that the TDS levels were found to be < 100 mg/l in 26 samples (46%) and < 250 mg/l in 46 samples (81%). The low TDS is a result of the high rainfall received here during the monsoon season. Because of this intense rainfall and steep topographic gradient actively flushes the products of weathering leading to low dissolved solids in ground water. According to WHO,66, the results of the total hardness (TH) classification show that there are different classes of groundwater in the study area, including soft groundwater (TH < 60 mg/l, N = 34), moderately hard groundwater (60 > TH < 120 mg/l, N = 13), hard groundwater (120 > TH < 180 mg/l, N = 5), and very hard groundwater (TH > 180 mg/l, N = 5).Figure 2Box pot showing the chemical composition of groundwater in study area.EC, TDS, and TH have corresponding standard deviations of 231 µS/cm, 148 mg/l, and 69 mg/l. The descriptive statistical analysis of the major cations and anions are shown in SI Table 1. The standard deviation is an excellent tool for determining groundwater features in any location15,60,70. High standard deviations in TDS, EC, and TH were reported in this research, indicating that the sources of chemical components in this aquifer come from diverse geochemical processes.High enrichments of Fe and Mn are reported in this study area. The Fe ranges between < 0.05 mg/l and 2.83 mg/l, with an average of 0.32 mg/l. (Fig. 2).The samples with N = 19 show Fe > 0.3 mg/L, which is above the acceptable limit as prescribed by BIS7, for drinking purpose, while the samples with N = 37 show 0.1 > Fe < 0.3 mg/l. Similarly, the Mn ranges between < 0.02 mg/l and 6.05 mg/l, with a mean of 0.27 mg/l. (Fig. 2). The samples with N = 16 show Mn > 0.1 mg/L, which is beyond the acceptable limit as recommended by BIS for drinking purpose, whereas the samples with N = 9 show 0.05 > Mn < 0.1 mg/l.Water quality index (WQI)Based on the WQI20,31,35, groundwater quality is graded as follows: excellent quality (< 25), good quality (25–50), average quality (51–75), poor quality (76–100), and unsuitable quality (> 100). The computed values of the WQI in the present study area is ranged from 14 to 621 in the groundwater samples, with an average value of 60. Among these samples, 42% (N = 24) were classified as having good quality (WQI 26–50), while 26% (N = 15) were categorized as having excellent quality (WQI < 25). Furthermore 19% (N = 11) were categorized as having average quality (WQI 51–75). Interestingly, only a small share of the samples, specifically 5% (N = 3), fell into the Poor Quality category (WQI 76–100). These findings highlight the overall favourable water quality conditions observed in the samples analysed. It is worth noting that only a small proportion of the samples, specifically 7% (N = 4), were classified as belonging to the category of unsuitable quality, as indicated by a WQI > 100. Thus, the aforementioned results underscore the overall good water quality characteristics seen in the studied samples.Using the spatial distribution of WQI as depicted in Fig. 3A, it is possible to evaluate the variation of WQI classes throughout the study area. About half of the study area, including 12,735 square kilometres, is classified as Excellent to Good Quality and this classification encompasses mainly Kollam, Kozhikode, Kannur and Wayanad districts. About 39% (9,971 sq. km) and 3% (885 sq. km) of the total area fell into the categories of Average Quality and Poor Quality, respectively, with the majority of these areas located in the districts of Palakkad, Kottayam, Thiruvananthapuram and Idukki. Subsequently, the unsuitable drinking water quality category encompasses about 8% (1953 sq. km) of the overall study area, mostly concentrated in the south part of Idukki district and a north portion of Pattanamtitta district. Additionally, south eastern part of Wayanad district also fall in this category. The high WQI value in this region is due to excessive concentration of Fe and Mn in the groundwater, which primarily derives from the geogenic source (Fig. 3B, C).Figure 3Spatial distribution of Iron (A), Manganese (B), and Water Quality Index (C) in study area. This map was generated using ESRI’s ArcGIS 10.7 for Desktop software (http://www.esri.com), version 10.7.1.Hydrochemical faciesLithospheric composition, groundwater movement, and other natural activities all affect the hydrochemical facies in the groundwater. In this study area, ions that prevail in groundwater are Na+  > Ca2+  > Mg2+  > K+ and HCO3− > Cl− > NO3− > SO42−. The Chadha diagram10 (Fig. 4D) are used to investigate the hydrochemical facies of the study region. The Chadha diagram reveals that the groundwater samples were scattered across three major zones, with the exception of four samples. The samples have been clustered together as follows: CaHCO3 facies (46%) > CaMgCl facies (30%) > NaCl facies (17%) > NaHCO3 facies (7%). The facies of NaCl and CaMgCl of groundwater show the effect of ion exchange processes, evaporation, and prolonged storage/residence but the mineral weathering process is only expected to have an impact on groundwater with CaHCO3 facies14,45. The CaHCO3 facies is observed in the majority of the samples from the district of Wayand, Pathanamthitta, Palakkad and Thrissur, and it exhibits the Recharging water features. Furthermore, CaMgCl facies indicate the reverse ion exchange properties of water, as found in the majority of samples from districts of Kazikhode, Malappuram, and Kottyayam. However, the NaCl facies is only seen in the majority of Trivandrum district samples. This facies represents salt water as a result of a high EC. Subsequently, only four samples, two from the Idukki district and two from the Kollam district, exhibit NaHCO3 facies, which is indicative of the base ion exchange properties of the water. The samples from the districts of Idukki and Kollam display mixed hydrochemical facies with an asymmetrical trend. SI Table 2 (Hydrochemical facies distribution) displays the hydrochemical facies distribution across district.Figure 4Factors influencing the chemistry groundwater chemistry. (A) Gibbs plot for cation, (B) Gibbs plot for anion, (C) Relation between Na/Cl and EC, (D) Chadha diagram.Influencing hydrogeochemical processesTo distinguish the influential hydrogeochemical processes in this aquifer, Gibbs and bivariate plots were used.Gibbs plotsGibbs employed the TDS and ionic ratios as a means of discerning the primary geochemical mechanisms that govern the chemical composition of water. In the cation plot and the anion plot (Fig. 4A, B), the TDS is plotted against the Na+  + K+/Na+  + K+  + Ca2+ ratio and against the Cl−/Cl− + HCO3− ratio, respectively. Evaporation, mineral weathering, and rainfall input may all be readily distinguished using these graphs. Ca2+ and HCO3− will be removed from the system during the process of evaporation by carbonate precipitation, which will then be followed by TDS enrichment in the aquifer45. As a result, the concentrations of Na+ and Cl− in groundwater will rise. In the current study, around 10% of the samples are fall in precipitation zone and rest of the samples cluster in a zone of mineral weathering. No samples are depicted on the evaporation zone. Mineral weathering is the dominant process fallowed by precipitation in the area of study.EvaporationAs reported by Jankowski and Acworth23, the plot of EC against Na+/Cl− will not change with increasing salinity (EC) if evaporation is the dominating process in water chemistry. When Na+/Cl− is plotted against EC, as shown in Fig. 4C, only a few wells follow the trend of evaporation. This demonstrates that evaporation is not the dominant factor that influences the chemistry of wate44. Figure 4C shows that the Na+/Cl− ratios are enriched and depleted, indicating that silicate dissolution and the ion exchange mechanism are primarily controlling the water chemistry36,44.Carbonate dissolution on water chemistryIonic ratios and bivariate plots are used in order to evaluate the influence of carbonate dissolution on chemistry water. Results of carbonate mineral dissolution in the groundwater yield of Ca2+, Mg2+, and HCO3−33,44,46. In this study, the average Ca2+/Mg2+ molar ratio is 2.6, which varied from 0.19 to 9.7 (one sample is eliminated because of its exceptional value of 24.5), and it is greater than 1 in 74% of the samples (Fig. 5). About 40% of the samples have a Ca2+/Mg2+ ratio greater than 2. According to Maya and Loucks33, dolomite dissolution will retain a ratio of 1, whereas calcite dissolution will result in a ratio greater than 1. The higher ratio of Ca2+/Mg2+ (> 2) is due to silicate weathering in the system43,46. In the present study area, water samples that are close to Ca2+/Mg2+ the = 1 line indicates that dolomite dissolution, while clusters that are between 1 and 2 reveal that the dissolution of calcite. The existence of an impact from silicate weathering in this aquifer is indicated by higher ratios, (Ca2+/Mg2+  > 2).Figure 5Influence of mineral weathering and ion exchange processes on groundwater chemistry. (A) Plot of HCO3 Vs Ca2+/Mg2+ for Carbonate dissolution, (B) Plot of HCO3− + SO42− Vs Na+/Cl− for Silicate weathering, (C) Plot of Ca2+/Na Vs HCO3−/Na+ for distinguish the effects of weathering, (D) Plot of HCO3− Vs Na+ + K+ for elucidation of ion exchange processes, (E) Plot of Cl− over Na+ + K+ for elucidation of hydrochemical process, (F) Plot of Na+ + K+–Cl− Vs Ca2+ + Mg2+–Alk–SO42 for weathering and ion exchange processes.Silicate weatheringAs previously stated, silicate weathering influenced the major ion chemistry in this aquifer43. Furthermore, the Na+/Cl− ratio in this aquifer ranges from 0.05 to 3.58, with an average of 1.45 (Fig. 5). The dissolution of halite will maintain a 1:1 Na+/Cl− ratio. However, the weathering of silicates and cation exchange resulted in a rise in the system’s sodium content, which led to a ratio of sodium to chloride greater than one. In this study area, around 81% of samples indicate a higher ratio (Na+/Cl− > 1), and the sodium content in these wells is obtained through silicate weathering and cation exchange. In order to confirm the impact of mineral weathering and ion exchange reaction, the relationship between Ca2+  + Mg2+ and HCO3− + SO42− is exercised. If the Ca2+ and Mg2+ resulted as a consequence of the weathering of carbonate, sulphate (gypsum, anhydrite), and silicate minerals, then the HCO3− + SO42− will balance both of those elements8,34. The water samples plotted on or near the 1:1 line in Fig. 5B illustrate the influence of mineral weathering on water chemistry in the current study. However, samples plotted above or below the equiline indicate the existence of alternative processes. Similarly, to investigate and distinguish the effects of weathering of silicates, carbonates, and evaporites/evaporation in the aquifer, molar Ca2+/Na+ is plotted versus molar HCO3−/Na+18,46. The samples were distributed in a cluster between the evaporation zone and the silicate weathering zone, as shown in Fig. 5. These elements are mostly derived through silicate weathering. The precipitation of carbonate minerals and the subsequent increase in Na+ concentration in the aquifers are explained by samples plotted on or near the evaporation zone. As a result of the weathering of soda feldspar (albite), potash feldspar, orthoclase, and microcline, Na+ and K+ ions are released into the aquifer15,39.Ion exchange reactionsCation exchange or silicate weathering are accountable for the enriching of Na+ over Cl−, while reverse ion exchange events are accountable for the depleting of Na+ over Cl−. The ion exchange process is further confirmed by the plot of Ca2+  + Mg2+ vs HCO3− + SO42− in Fig. 5B. The presence of a high Ca2+  + Mg2+ concentration over HCO3− + SO42− in about 20% of the samples suggests a reverse ion exchange process. Ca2+  + Mg2+ depletion over HCO3− + SO42− confirms that Ca2+  + Mg2+ is withdrawn from groundwater by cation exchange reaction or Na+/K+ silicates weathering should have increased HCO3−45. Similarly, the ratio of Na+ + K+ to HCO3− demonstrates an elevation of Na+ + K+ in relation to HCO3− as a result of cation exchange activities (Fig. 5D). Schoellar, described the chloroalkaline indices (CAI 1 and CAI 2), which were used in several studies to elucidate ion exchange processes in groundwater50.The negative values of CAI 1 and CAI 2 indicate cation exchange process whereas positive values indicate reverse ion exchange. High concentrations of Na+ + K+ over Cl− result in negative values in these calculations, indicating cation exchange processes, whereas high concentrations of Cl− over Na+ + K+ result in positive values, indicating reverse ion exchange reactions. The majority of the samples (91%) had negative values and show a negative trend, as seen in Fig. 5E. Therefore, cation exchange reaction plays a crucial role in controlling water chemistry, whereas only five samples exhibit positive values indicative of reverse ion exchange reaction. In the case of a cation exchange reaction, the plot of Ca2+ + Mg2+–Alk–SO42− versus Na+ + K+–Cl− will exhibit a negative slope and a significant negative correlation. Figure 5F depicts a negative slope with a significant negative correlation, proving that cation exchange reactions in addition to mineral weathering influence the water chemistry in the study area.Geochemical modellingCommonly, aqueous geochemical calculations are often used to examine the interactions between rocks and minerals employing a thermodynamic database. The identification of silicate weathering and carbonate mineral dissolution is facilitated by these calculations11. Based on available data, the PHREEQC algorithm was used to determine the ionic strength and saturation indices for selected minerals40,41. Ionic strength in the current study ranged from 0.0009 to 0.014, with an average value of 0.004. The ionic strength has a significant correlation with EC as seen in Fig. 6. Further, the average value of the partial pressure of CO2 (pCO2) was found to be 2.88, and it ranged from − 9.24 to 0.03. At the location of the study, the pCO2 levels in the water samples are mainly greater than the atmospheric pCO2 (− 3.4), with the exception of six wells. Wells with more salinity water have pCO2 higher values, which are strongly associated with the atmosphere, and the relationship between EC and pCO2 also shows a positive tendency (Fig. 6). The pCO2 numbers may be related to the groundwater level, and deep wells tend to have low EC and pCO2. However, groundwater level data is not used to conform this information for sampling wells.Figure 6Saturation indices of carbonate minerals and their relation with salinity in the groundwater.According to SI calculations and Fig. 6, about 20% of groundwater sample wells are saturated with carbonate minerals such as calcite (CaCO3), aragonite (CaCO3), dolomite (CaMg(CO3)2), siderite (FeCO3), and rhodochrosite (MnCO3), indicating that dissolution of carbonate minerals regulates groundwater chemistry. The fact that the groundwater has high concentrations of Ca2+, Mg2+, and HCO3−, as well as a neutral to alkaline pH, is further evidence that carbonate reactions are crucial for preserving the hydrochemistry of the groundwater in multilevel wells. The dolomite saturation index values range widely across the sample wells (− 10.5 to 3.1). The existence of carbonate minerals and their wide variation are crucial because ground level variations in the sample wells may have caused calcite and aragonite to dolomitize as a result of an interaction with groundwater that contains magnesium.Groundwater tends to be supersaturated with hematite (Fe2O3), goethite (FeOOH), and ferrihydrite (Fe(OH)), as shown in Fig. 7. This indicates that there is a suitable thermodynamic environment for the precipitation of Fe(III) phases from groundwater. Meanwhile, when Fe(II) and Mn(II) concentrations increase, correspondingly increases the SI of siderite and rhodochrosite. Fe(II) and Mn(II) concentrations in groundwater are constrained by adsorption onto or co-precipitation with CaCO3, according to the saturation of groundwater with calcite, siderite, and rhodochrosite5. CaCO3 may be partially substituted with Mn, which can lead to the creation of rhodochrosite37,65. Therefore, these minerals have the ability to act as a sink by removing Fe(II) and Mn(II) ions from anoxic groundwater that contains a high concentration of HCO3−. Such processes can accelerate the rate at which Fe and Mn oxyhydroxide are reduced by shifting the equilibrium of redox towards the reduced phases of Fe (II) and Mn (II), respectively.Figure 7Saturation indices of Iron minerals and their relation with salinity in the groundwater.Multivariate statistical analysisPearson correlation analysisPearson correlation analysis was performed in this study and results are presented in Fig. 8A. Variables namely pH, Fe and Mn does not correlate with other variables. Additionally, pH shows significant negative correlation with NO3− and K+. EC and TH express strong positive correlation with all variables except NO3−, K+ and pH. NO3− shows the significant correlation with only Na+, K+ and Cl− and rest of the parameter are not correlated with it whereas K+ shows correlation with only NO3−. The positive correlation between NO3− and K+ (r 2 > 0.59) suggest that both are derived from anthropogenic sources in this aquifer. Sulphate and chloride minerals are highly undersaturated in this aquifer. Therefore, Cl− and SO42− are good indicator of anthropogenic input. Variables correlated with Cl− and SO42− are partially or completely derived from anthropogenic sources. Similarly, with the exception of pH, K+, and NO3−, HCO3− has a substantial positive correlation with all variables. Also, none of the relationships between the concentration of metals (Fe and Mn) in water samples were significant. This means that the Fe and Mn in the water samples were not related to other parameter of water and may have come from different sources in the study area.Figure 8Multivariate statistical analysis. (A) Results of Pearson correlation analysis, (B) Results of Principle component analysis, (C) Results of Hierarchical cluster analysis in polar heat-map. The polar heat-map was generated using Originlab’s, Origin 2022 for Desktop software (https://www.originlab.com/origin), version Origin Pro 2022.Principal component analysis (PCA)One effective approach for identifying the origins of variables and geochemical processes is the PCA analysis. In PCA, the water analysis data is reduced and simplified into components. Tables 2 and 3 depicts the number of obtained factors with Eigen values greater than one, as well as variance and cumulative variance. Similarly, a graphical illustration is shown in Fig. 8B. In the current study, three factors account for about 77.1% of the variance. Factor 1 has a substantial positive loading on EC, TH, Na+, Mg2+, Cl−, HCO3−, and SO42−, and it accounts for 50.2% of the variance. The main source of chemical constituents in this aquifer is mineral weathering which to some extent includes gypsum and halite for the contributing Cl− and SO42− in the aquafer. Therefore, elevated HCO3− loading implies those processes. Therefore, factor 1 validates the impact of geogenic processes. Factor 2 accounts for 17.9% of total variance and has a high positive loading of K+ and NO3−. This factor illustrates the anthropogenic NO3− inflow into this aquifer. There is no correlation between NO3− and Cl−, SO42−, or other variables. Consequently, nitrification is the most probable source of NO3− in this aquifer. The application of fertilizers and manures in agricultural practices results in an increased presence of NO3− in groundwater due to infiltration in the vadose zone. Factor 3, which accounts for 9.0% of the total variance, exhibits positive loadings only for Fe and Mn. This indicates that the lack of correlation between Fe and Mn and other water characteristics suggests that their presence may be due to the weathering of local geological formations rich in these minerals, such as Laterite and Archaeans.Table 2 Results of total variance in PCA.Table 3 Loading of Component matrix in PCA.Hierarchical cluster analysis (HCA)The results of HCA presented in the form of polar heat-map (Fig. 8C) with vertical dendrogram (Q-mode) represent the cluster of water parameters and horizontal dendrogram (R- mode) represent the cluster of sample location. There are four groups shown in Q-mode. Group 1 is associated with the elements EC, TH, Na+, Mg2+, Cl−, HCO3−, and SO42−, which suggests that silicate and carbonate weathering, as well as gypsum and halite, are present to a certain extent. Group 2 consists of two variables, including NO3− and K+, and these variables indicate anthropogenic origins. Only pH exists in Group 3, and it is not affected by weathering or anthropogenic factors. Group 4 includes metals like Fe and Mn, which may indicate a potential independent origin such as the weathering of Laterite and Archaeans. Furthermore, in R-mode, there are three clusters with 75% samples (n = 43) fall under cluster 1, which is less mineralized, and the EC values with an average of 160 μS/cm which suggest that groundwater is located nearby the aquifer recharge area. Cluster 2 accommodates 17% (n = 10) samples with an average of 470 μS/cm. This cluster indicates that waters that were recharged relatively long ago and have flown a certain distance within the aquifer are moderately mineralized. The locations included in this cluster are mainly from Malappuram district and some parts of Idukki district. Only 4 samples (3 from Palakkad district and 1 from Malappuram district) are grouped in cluster 3. Groundwater in these wells is highly mineralized and EC, with an average value of 900, indicating high mineralization.
