Statistical analysis of the hydrogeochemical evolution of groundwater in hard and sedimentary aquifers system of Gadilam river basin, South India

3.1 Data analysis

Each sampling site is characterized by a large number of chemical and physical variables, making the regional hydrogeochemical study a multivariate problem. The multivariate statistical analysis is a quantitative and independent approach of groundwater classification allows to group of groundwater samples and the making of correlations between chemical parameters and groundwater samples. In this study correlation analyses and factor analyses were applied using Statistical Package of Social Studies (SPSS) version 10.

Data obtained from the laboratory analysis were used as variable inputs for factor analysis. Preparation of a correlation matrix of the data from which initial factor solutions were extracted by the principal component analytical method. Factor extraction was done with a minimum acceptable Eigen value as 1 (Kaiser, 1958; Harmann, 1960). Orthogonal rotation of these initial factors to terminal factor solutions was done with Kaiser’s varimax scheme. This method maximizes the variance of the loadings on the factors and hence adjusts them to be either near ±1 or near 0 (Davis, 1973). Factor score coefficients are derived from the factor loadings. Factor scores are computed for each sample by a matrix multiplication of the factor score coefficient with the standardized data. The value of each factor score represents the importance of a given factor at the sample site. A factor score >+1 indicates intense influence by the process. Very negative values (<−1) reflect areas virtually unaffected by the process while near zero scores reflect areas with only moderate effect of the process. The spatial distribution of the factors can be assessed by a contour of the factor scores representing each factor.

4.1 Correlation analyses

The correlation analysis for groundwater in different litho units of all seasons was studied in detail.

4.2 Alluvium

In SUM, good correlation exists between Cl–Ca, Cl–Mg, Cl–Na, HCO₃–Ca, HCO₃–Na, HCO₃–K, HCO₃–Mg, Ca–Mg, Ca–Na and Mg–Na (Fig. 3). Poor correlation exists between SO₄, PO₄ and H₄SiO₄ with all other ions. Cl shows good correlation with Ca, Mg and Na indicates dissolution and leaching of secondary salts and a significant correlation of HCO₃ with Ca, Mg, Na and K indicates chemical weathering. Poor positive correlation of SO₄, PO₄ and H₄SiO₄ shows the lesser possibility of anthropogenic influence into the system. Ions like Cl and HCO₃ show significant correlation with Ca, Mg, Na and K indicating the predominance of chemical weathering along with leaching of secondary salts during this season. For SWM, good to excellent correlation is obtained between Cl–HCO₃, Cl⁻–Ca, Cl⁻–Na⁺, Cl–Mg, HCO₃–Mg, HCO₃–Ca, HCO₃–Na, PO₄–H₄SiO₄, PO₄–Ca, PO₄–Mg, Ca–Mg and Ca–Na indicating the dominance of weathering and leaching processes. PO₄ shows good correlation with H₄SiO₄ indicates anthropogenic impact from agricultural practices. Opaline Silica, from the rice plants provides a readily soluble source of Silica and frequent oxidation and reduction cycle of paddy soil may accelerate weathering of Smectite, thereby making silica more soluble (Wang et al., 1993). Poor correlation exists between SO₄ with other ions, might be because of the effect of dilution. During monsoon period, the recharge of fresh meteoritic water with lesser ions dilutes the existing groundwater with higher ion concentrations. The intense dilution or dissolution process may reduce the concentration of the SO₄ in the groundwater system (i.e. under saturation condition of sulphates in the system), which is reflected poor correlation with other ions (Prasanna, 2008). Generally weathering along with anthropogenic impact are the dominant processes during this season. In NEM, good correlation exhibits between Cl–Ca, Cl–Mg, Cl–Na, HCO₃–K, PO₄–H₄SiO₄, Ca–Mg, Ca–Na and Mg–Na. Poor correlation exists between HCO₃, SO₄ and PO₄ with all other ions except K and H₄SiO₄. Cl shows good correlation with Na, Ca and Mg indicates leaching of secondary salts. PO₄ is well correlated with H₄SiO₄ indicating anthropogenic impact from agricultural practices as similar as in SWM. HCO₃ shows good correlation to poor correlation with K and Na indicates weathering of potash feldspar. In POM, good correlation exhibits between Cl–Ca, Cl–Mg, Cl–Na, H₄SiO₄–HCO₃, PO₄–K, Ca–Mg, Ca–Na and Mg–Na. Poor correlation exhibits between SO₄ and H₄SiO₄ with other ions might be because of dilution. SO₄, PO₄ and K shows good to poor positive correlation indicates anthropogenic activities from fertilizers. The contribution of HCO₃, H₄SiO₄ and Na indicates albite weathering from the source rock. Cl shows good correlation with Mg, Ca and Na indicates leaching of secondary salts. Generally chemical weathering, leaching of salts and anthropogenic impact are the major contribution for this terrain.

Close

Figure 3

Correlation between the ionic species in groundwater of Alluvium formation.

Export to PPT

4.2.1

4.2.1 Tertiary

In SUM, good correlation exhibits between Cl–Mg, Cl–Na, Cl–K, HCO₃–Ca, SO₄–Mg, SO₄–Na, SO₄–K, PO₄–Ca, Mg–Na and Na–K (Fig. 4) Poor correlation exhibits between H₄SiO₄ with other ions indicates lesser influence of silica during this season. SO₄ shows good poor positive correlation with Mg, Na and K indicates the leaching of marchasite present in this region (Sivalingam, 2005). HCO₃ is well correlated with Ca indicates weathering of Carbonate minerals. In SWM, good to excellent correlation is obtained between Cl⁻–SO₄, Cl⁻–Na⁺, Cl⁻–K, ${\text{HCO}}_3^{-}''–''\text{Ca}$ , SO₄–Na, PO₄–Ca, PO₄–H₄SiO₄, PO₄–Mg, H₄SiO₄–Ca, H₄SiO₄–Mg and Ca–Mg. Poor correlation exists between HCO₃ and SO₄ with other ions except Ca and Na. Cl shows good correlation with SO4, Na and K indicates leaching of secondary salts as in SUM season. Good correlation between HCO₃, Na and H₄SiO₄ indicates albite weathering. H₄SiO₄ is well correlated with Ca and Mg indicating intensive weathering reaction enhancing H₄SiO₄. PO₄ is well correlated with H₄SiO₄ indicating anthropogenic impact from agricultural practices. In NEM, good correlation exhibits between Cl–Na, SO₄–Na, H₄SiO₄–Ca, H₄SiO₄–K and Ca–K, indicating chemical weathering and leaching of secondary salts. Poor correlation exhibits between HCO₃ and PO₄ with other ions. In POM, good correlation exhibits between Cl–SO₄, Cl–Ca, Cl–Na, Cl–K, HCO₃–Ca, HCO₃–Mg, SO₄–Na, H₄SiO₄–Mg and Ca–Na, indicates leaching of secondary salts with anthropogenic impact. Poor correlation exhibits between PO₄ with other ions indicating lesser influence of anthropogenic impact might be the effect of dilution. HCO₃ is well correlated with Ca and Mg indicating weathering of carbonate minerals. Cl shows good correlation with SO₄, Ca, Na and K indicating leaching of salts. In general, SO₄ is well correlated with Na in all the seasons indicates high sodium sulfate concentration is derived from the local fertilizer plants at (NLC) which is located in this terrain. Generally weathering and leaching of secondary salts are the dominant processes during this season.

Close

Figure 4

Correlation between the ionic species in groundwater of Tertiary formation.

Export to PPT

4.2.2

4.2.2 Archaean

In SUM, good correlation exhibits between Cl–Ca, Cl–Mg, Cl–Na, Cl–K HCO₃–Na and Na–K indicates the leaching of secondary salts (Fig. 5). Poor correlation exhibits between SO₄, PO₄ and H₄SiO₄ with other ions indicates lesser influence of anthropogenic impact. Cl shows good correlation with Ca, Mg, Na and K indicates secondary leaching of salts. HCO₃ is well correlated with Na indicating weathering of sodic feldspar from the source rock. In SWM, good correlation exists between Cl–Ca, Cl–Na, Cl–K, HCO₃–Na, PO₄–H₄SiO₄, Ca–Mg and Na–K. Poor correlation exhibits between SO₄ and H₄SiO₄ with other ions. Good to poor positive correlation of HCO₃, Na, K and H₄SiO₄ indicates albite weathering from the source rock. Cl shows good correlation with Ca, Na and K indicating secondary leaching of salts. H₄SiO₄ is well correlated with PO₄ indicates anthropogenic impact from the agricultural practices. In NEM, good correlation exhibits between Cl–Ca, Cl–Na, SO₄–Na and H₄SiO₄–Ca indicates secondary leaching of salts. Good to poor positive correlation between H₄SiO₄, Ca, Na and K indicates potash feldspar weathering from the source rock. Poor correlation exists between HCO₃ and PO₄ with other ions. In POM, good correlation exhibits between Cl–HCO₃, Cl–SO₄, Cl–Ca, Cl–Na, HCO₃–SO₄, HCO₃–Ca, HCO₃–Na, SO₄–Na and Ca–Na, indicating the leaching of salts along with weathering process. Poor correlation exhibits between PO₄ and H₄SiO₄ with other ions indicates lesser influence of anthropogenic impact. Cl shows good correlation with Ca, Na and SO₄ indicates leaching of secondary salts. Generally weathering and leaching of ions are the prominent processes during this season.

Close

Figure 5

Correlation between the ionic species in groundwater of Archaean formation.

Export to PPT

In general Ca, Mg, Na, Cl, HCO₃ and SO₄ remain to be the chief ionic species throughout the seasons and other play a less dominant role with respect to seasons. Hence correlation study reveals Ca, Mg, Na, Cl, HCO₃ and SO₄ form the spinal species (Chidambaram et al., 2007) and others are seasonal. This also shows that spinal species are the chief ions, which are playing a key role in controlling water chemistry of the region.

4.3 Factor analysis

After computation of the correlation matrix and the correlation coefficient, measures of interrelationship for all pairs of constituents are determined. Then by the examination of the factor loadings, factors are interpreted by rotation. Factor analysis is a widely used statistical technique in groundwater studies because it reduces the number of variables and enables the detection of structure in the relationships between variables. In this study principal component method is used as a parameter estimation method to transform a set of observed independent variables into an orthogonal set of variables called principal components. The first principal component accounts as much as possible of variance of the observed variants. The second principal component accounts for as much as possible of the residual variances not accounted by the first principal component. Each succeeding principal component accounts for as much as possible for the residual variance not accounted for by all the previous principal components. Initial factor loadings unlikely to reveal the underlying structure of the observed variants, because of certain mathematical conditions such as the variance properties of the principal components has to be understood in detail. To have a more clear view of this structure, common factors associated with initial set of loadings are linearly transformed into a new set of common factors, associated with new set of loadings, factor rotation. Though numerous schemes have been proposed, Kaiser Scheme called varimax rotation is used in this study. Varimax rotation yields set of loadings such as the variances of loadings are maximum. The factor analyses for four seasons were carried out and the results (Table 2) reflect complexities in chemical nature of the region.

In SUM, three factors were extracted with 78.62% of total data variability (TDV). Factor 1 was represented by Cl⁻, HCO₃, Ca, Mg and Na indicating leaching of secondary salts (Prasanna et al., 2009). The concentration of Na and Cl can be ascribed to the intrusion of seawater into the aquifer system which increases the concentrations of these ions. The presence of HCO₃, Ca and Mg reflects the signatures of natural water recharge and rock–water interaction. Surface water charged with atmospheric and biogenic CO₂ infiltrates into the subsurface and aggressively attack aluminosilicates including feldspars and micas present in the formation liberating cations such as Ca and Mg into the water and leaving residues of clay minerals. A consequence of this incongruent dissolution is a rise in pH and in HCO₃ concentration of the water (Freeze and Cherry, 1979). Factor 2 represented by PO₄ and K indicating the anthropogenic impacts from the agricultural practices. Factor 3 represented by HCO₃ and Ca indicating water–soil/rock interaction.

In SWM, three factors were extracted with 76.90% of total data variability (TDV). Factor 1 was represented by Cl, HCO₃, Ca and Na follow the same trend of factor 1 in SUM. Factor 2 representing PO₄, H₄SiO₄, Ca and Mg indicating anthropogenic impact from the agricultural practices. Factor 3 represented by SO₄ and K also indicating anthropogenic impact.

In NEM, three factors were extracted with 73.96% of total data variability (TDV). Factor 1 representing Cl, Ca, Mg and Na indicating leaching of secondary salts. Factor 2 representing HCO₃, SO₄, Na and K indicating intensive weathering. Factor 3 represented by PO₄ and H₄SiO₄ indicating anthropogenic impacts from the agricultural practices (Prasanna et al., 2009).

In POM, three factors were extracted with 77% of total data variability (TDV). Factor 1 represented by Cl, Ca, Mg and Na follow the same trend of factor 1 in NEM. Factor 2 represented by SO₄, PO₄ and K indicating anthropogenic impacts from fertilizers. Factor 3 represented by HCO₃ and H₄SiO₄ indicating weathering of silicate minerals.

In general leaching of secondary salts, weathering and anthropogenic impacts are the dominant controlling factors in the study area.

4.4 Factor scores

The factor score were also estimated to find out the spatial variation of the factor representation and to identify the zone of its representation. They are commonly obtained by two approaches weighted least square method and the regression method. The regression method is used in the study to compute the factor scores. The positive zones indicate the dominance of that factor (Hydrogeochemical regime). The spatial variation by using the factor score values of each sampling points were plotted by SURFER software. The spatial representation for factor 1 (Fig. 6) irrespective of seasons fall along the coast with higher industrialized region, where the influence of SIPCOT industry is preferably noted. The areal distribution of the factor score show that the intensity of salt water activity is well noted along the coastal area. This suggests that salt water ingress into the aquifer was predominantly related to water recharge from this tide influenced river. The high score area is the most densely populated and consequently witnesses higher groundwater abstraction. This probably establishes local freshwater depression cones, which induces saltwater infiltration into this region. The spatial representation of factor 2 (Fig. 7) is mainly represented in the Archaean region and along the contact of lithology with few in Alluvium region. In SUM and SWM seasons, higher scores are located in the upstream side of Archaean region. They represent leaching of secondary salts and anthropogenic impact by agricultural practices. In NEM, they fall in the region with prominent lithological contact and structural controls and a small patch in the SE side of the region. In POM, factor 2 is represented in the NE side of the region influenced by anthropogenic impacts. The spatial representation of factor 3 (Fig. 8) is mainly represented in the upstream side of Archaean formation with few along the contact of lithology and in Tertiary formation. In SUM, higher values are located in the upstream side of Archaean formation indicating the effect of leaching of secondary salts with some anthropogenic impacts. In SWM, they fall in the NW side and central part of the study area similar to SUM season. In NEM, factor 3 is well represented in the regions covered by sulphide ore bodies (Sivalingam, 2005) in Tertiary formation and agricultural practices. In POM, factor 3 is well represented by the lithology and structural control.

Close

Figure 6

Spatial distribution of Factor 1 in groundwater.

Export to PPT

Close

Figure 7

Spatial distribution of Factor 2 in groundwater.

Export to PPT

Close

Figure 8

Spatial distribution of Factor 3 in groundwater.

Export to PPT

As the hydrogeochemical system is complex and it is not possible to delineate and reason out all possible mechanism controlling the water chemistry of the region. The dominant factors responsible are extracted and their factor scores alone are used for the representations. To study the dominant hydrogeochemical regime of the study area all the factor score representation were overlaid for all seasons (Fig. 9). The overlay indicates the dominant hydrogeochemical regimes are distributed in the upstream side, contact of lithology and along the coast. In the upstream side, Archaean formation is dominantly influenced by leaching process and in the central part contacts of lithology and structural control are the dominant factors. In the NE side, saline water intrusion into the coastal aquifers, anthropogenic activities from industry and agricultural practices are the major controlling factors in the study area.

Close

Figure 9

Spatial distribution of dominant hydrogeochemical regime in the study area.

Export to PPT

4.5 Ionic ratio

Table 3 shows the variation of cations in different litho units of the study area. In Alluvium and Tertiary formations, Na/(Ca + Mg) ratios are >3 irrespective of seasons indicating the dominance of alkalies over alkali earth. In Archaean formation, ratios fall between >1 and <3 category indicating their approach towards equilibrium. Na/Cl ratio in Alluvium formation shows that the dominance of chloride indicates the saline water intrusion into the coastal aquifers (Chidambaram et al., 2005). In Tertiary, ratios fall in >1–<3, indicating their approach towards equilibrium during SUM and SWM and dominance of chloride is well noted in NEM and POM seasons. In Archaean same trend was followed as in Alluvium, indicating the dominance of leaching process. In Alluvium formation, Cl/HCO₃ ratio indicates most of the samples fall below one during SWM and NEM and between >1 and <3 category during SUM and POM seasons indicates the impact of saline water intrusion (Prasanna et al., 2008b). In Tertiary formation, most of the samples irrespective of seasons fall in >1–<3 ratio indicates the mobilization of Cl ions during SUM, SWM and POM seasons. In Archaean formation, most of the samples irrespective of seasons fall in <1 indicates the dominance of bicarbonate due to intensive weathering process. In Alluvium formation, Cl/SO₄ ratio indicates majority of samples irrespective of seasons ranges between in >1 and <3 category indicates the equilibrium of both ions due to the mixing of saline water composition into the groundwater. In Tertiary formation, ratios show lesser value indicating the dominance of sulphate due to the reduction process of sulphides present in the formation (Sivalingam, 2005). In Archaean formation, most of the samples irrespective of seasons fall in >3 indicating the dominance of Cl ions due to the leaching of secondary minerals and longer residence time in the aquifer matrix (Sinivasamoorthy et al., 2008). In (Ca + Mg)/HCO₃ ratio samples are fall below one and range from >1 to <3 categories indicating the dominance of bicarbonate and equilibrium state of both ions i.e. (Ca + Mg) and HCO₃. In Ca + Mg/TZ⁺ ratio of Alluvium formation, majority of samples irrespective of seasons shows lesser values indicates the dominance of total cations. In Na + K/TZ⁺ ratio, most of the samples irrespective of seasons show higher values indicate the dominance of alkalies. Same trend was followed by Tertiary formation. In Ca + Mg/TZ⁺ ratio of Archaean formation, most of the samples fall between <0.3 and >0.6 category indicating their approach towards equilibrium and in Na + K/TZ⁺ ratio, samples show higher values indicate the dominance of alkalies. Generally, NaCl is the dominant facies with a chief factor of rock water interaction was observed in the study area. Overall, the ionic ratios are well substantiated with the results of correlation and factor analysis.