Authors Copy

Provided for research and development purpose.

Not for replication, advertising or commercial use.
Evaluation of the groundwater quality along coastal stretch
between Vembar and Taruvaikulam, Tamil Nadu, India: a
statistical approach

S. Selvakumar, N. Chandrasekar, Y. Srinivas,

T. Simon Peter & N.S. Magesh

This article is published in Journal of Coastal Sciences hoisted by Centre for Geotechnology, MSU. The appended
print is provided to the author for research & development and education use, including for instruction at the
authors institution and sharing with colleagues. Other uses, including replication and distribution, or selling are
prohibited. The soft copy of the article is permitted to post in their personal website or institutional repository.

J O U R N A L

O F

C O A S T A L

S C I E N C E S

JOURNAL OF COASTAL SCIENCES

Journal homepage: www.jcsonline.co.nr
ISSN: 2348 6740

Volume 1 Issue No. 1 - 2014

Pages 22-26

Evaluation of the groundwater quality along coastal stretch between

Vembar and Taruvaikulam, Tamil Nadu, India: a statistical approach
S. Selvakumar*, N. Chandrasekar, Y. Srinivas, T. Simon Peter, N.S. Magesh
Centre for Geotechnology, Manonmaniam Sundaranar University, Tirunelveli, Tamil Nadu 627 012, India

`A B S T R A C T

ARTICLE INFO

The present study investigates the groundwater quality from Vembar to Taruvaikulam coastal area to provide an
overview of the relationship and sources of physicochemical parameters. A total of 23 groundwater samples were
collected and analyzed for various physico-chemical parameters such as pH, electrical conductivity (EC), total
dissolved solids (TDS), calcium (Ca2+), magnesium (Mg2+), sodium (Na+), potassium (K+), bicarbonate (HCO32-) ,
chloride (Cl-) and sulphate (SO42-). The ionic concentrations confirm most of the groundwater samples belong to order
of (Na+) > (Mg2+) > (Ca2+) > (K+) and (Cl-) > (SO42-) > (HCO32-). Spearmans correlation matrix, principal component
analysis (PCA) was used to distinguish the main contamination sources in the study area. The significant correlation
(r=>0.5) of EC, TDS, sodium, magnesium and chloride indicates saline intrusion, insignificant correlation (r=<0.5) on
calcium and bicarbonate ions attribute to a fresh water source. Four varimax components were extracted from PCA,
which explained 55.3, 19.20, and 0.98% of the variation in the groundwater quality in present study area. The
components obtained from PCA indicated that seawater intrusion, industrial pollution, geological activities and
anthropogenic activities exists in the study area.
*Corresponding author, E-mail address: [email protected]
Phone: +91 9751616732
2014 Journal of Coastal Sciences. All rights reserved

Received
20 November 2013
Accepted
3 March 2014
Available online
6 March 2014
Keywords
Groundwater quality
Correlation analysis
Principal Component
Analysis
Seawater intrusion

1. Introduction
As groundwater is mainly used for drinking, irrigation, industrial and
other domestic purposes, a detailed geochemical analysis is
extremely important, as although certain constituents became toxic
beyond particular concentrations, although they may be beneficial at
lower amounts. The natural chemical quality of groundwater
depends on geological as well as geographical arrangement in the
region. The contamination caused by seawater intrusion is a major
problem in the coastal aquifers, which occurs due to over extraction
of groundwater resulting in seawater infiltration into the aquifer
through the unsaturated zone (Sarath Prasanth et al. 2012;
Krishnakumar et al. 2012; Chandrasekar et al. 2013).
Many studies have been reported on groundwater quality
analysis. In this context, statistical analysis is very useful due to its
relative importance in evaluating the combination of large chemical
variable dataset. Principal component analysis (PCA) and Speramans
correlation matrix have been frequently applied to analyze the
pollution parameters in groundwater. Senthilkumar et al. (2008)
proposed multivariate analysis is very useful due to its large
chemical variable dataset and they are used as an analytical tool to
reduce and organize large hydro-geochemical datasets into groups
with similar characteristics. Abdul Halim et al. (2010) proposed PCA
to classify groundwater and identify major mechanisms influencing
the groundwater chemistry. The principal component analysis has
been successfully applied to sort out various hydro-geochemical
processes from commonly collected groundwater quality data
(Hitchon et al. 1971). Mrklas et al. (2005) have used principal
component analysis (PCA) for simplifying the analysis of complex
22

chemical relationships. Moreover, PCA is used as a tool to organize

the data and to differentiate between different chemical
environments in a region. In comparison, PCA is frequently employed
in hydrochemical studies, geology and hydrogeological applications
(Akbal et al. 2011; Najar et al. 2012). Elkarail and Obied (2013) have
used Speramans correlation matrix to identify the groundwater
quality in the Red-sea coast in Sudan. They have evaluated the
processes govern changes in the groundwater composition mainly
associated with over abstraction, biodegradation, carbonate
saturation and seawater intrusion. These analyses are useful for
interpreting the groundwater quality data and relating them to
specific changes in hydrogeological process. Hence, an application of
the principal component analysis can be used as an important tool in
the hydro-geochemical analysis of any region and helps to plan
suitable management measures in time to control and improve the
groundwater quality.
The objective of this study is to investigate the spatial variations
in groundwater quality and to identify the source of pollution.
Correlation matrix and principal component analysis were used to
evaluate the information concerning the similarities between the
sampling stations and to ascertain the contribution of potential
factors or pollution sources among various physico-chemical
parameters at 23 different sampling stations of the coastal region
between Vembar to Taruvaikulam. Based on the information
obtained, a holistic interpretation of the results and the use of
selected parameters as source tracers for contamination were
improved.
ORIGINAL

ARTICLE

J O U R N A L

2. Study area
The present study area is located in the southern part of Tamil Nadu,
which lies between north latitudes 889 to 9 08 and east longitude
7816 to 7836 and has a coastal length of 42 km (Figure 1). Sandy
beaches are perceived along Vembar, Vaippar and Taruvaikulam
coastal areas. The backshore of Kallar, Veppallodai and Taruvaikulam
segment largely consists of salt pans. Back water zone are
extensively present between Vembar and Kalaignapuram area. The
dune patterns are developed by vigorous aeolian processes resulting
in the migration of dunes with frequent changes in their shape and
pattern from time to time, but generally trends almost parallel to the
coastline. The sand sheet in the coastal zone extends from Vembar to
Kalaignapuram and Taruvaikulam. The sand sheet is implanted with
rich black sand deposits. Patches of Teri sand deposits are observed

O F

C O A S T A L

S C I E N C E S

time of sampling. Care was taken to avoid sample contamination by

external means. Each bottle was labeled with its corresponding
sampling station and time of sampling. The location of the sampling
stations was fixed using Garmin E-trex 12 channel Global Positioning
System (GPS). Hanna portable water quality meter (Model No.
HANNA HI-9828) were used to conduct in situ measurements of
physical parameters such as pH, electrical conductivity (EC), total
dissolved solids (TDS).The samples were immediately filtered with
0.45m cellulose acetate membrane filter (Whatman Milipores) after
being transported to the laboratory. The collected samples were
preserved in a refrigerator at 40C to avoid microbial growth,
flocculation, and reduce any adsorption on the polythene bottles.
Samples were preserved according to standard methods for the
examination of water and waste water (APHA 2005) for following
analysis.

Fig. 1 Location map of the study area

in the north of Melmandai and Vembar coastal area. The drainage

pattern of the area is mainly controlled and influenced by the
presence of perennial rivers like Vembar, Vaippar and Kallar. The
Vaippar river basin extends for about 6255 sq.km and the Kallar is a
small river and receives water supply mostly from NE monsoon. The
climate of the area experiences humid tropical and the annual
average rainfall is 662.2 mm with a wind velocity varies from 2.8 to
35.5 km/hr.

3.2. Chemical analysis

3. Materials and Methods

Calcium(Ca2+) and magnesium (Mg2+) concentrations were

determined by complexation reaction using EDTA (Ethylenediamine-teracetate) with ammonium purpurate as an indicator for
the determination of Ca2+ content alone, and Eriochrome Black-T was
used as an indicator for both Ca2+ and Mg2+content. Sodium (Na+) and
Potassium (K+) were estimated by Flame photometer. Total hardness
of the groundwater was calculated using the method proposed by
Hem (1985) and Ragunath (1987).

3.1. Field sampling and preservation

TH ( asCaCO 3 ) = Ca 2 + + Mg 2+ meq / l 50

Groundwater samples were collected from 23 open dug wells and

Chloride (Cl-) content was determined with the standard silver
bore wells of different locations and were homogenized. During nitrate (0.01N) titration method and in the presence of 1mL of
sampling, the pre cleaned acid washed HDPE (High density potassium chromate (5%) as an indicator. The determination of
polyethylene) bottles were normalized by rinsing thoroughly at the carbonate and bicarbonate concentrations was carried out by
23

ORIGINAL

ARTICLE

J O U R N A L

titration using 0.01N sulfuric acid and in the presence of

phenolphthalein indicator for the rst and methyl-orange for the
second. Sulfate (SO4) estimation was done using UVspectrophotometer. The reaction error (E) of all the 23 water
samples was less than the accepted limit of 5 % and samples.
3.3. Statistical analyses
All statistical analyses were performed using the IBM SPSS Statistics
19. PCA was applied to the datasets that had been pre-clustered by
extracting the eigen values and eigen vectors from a square matrix
produced by multiplying the data matrix. The most significant
components were extracted to reduce the contribution of variables
with minimum significance. Then, the obtained components were
further subjected to varimax rotation to generate varimax factors
and maximize the differences between the variables, thus facilitating
easy interpretation of the datasets.
The positive scores (>+1) reflect areas highly affected and
extreme negative scores (<-1) reflect areas essentially unaffected by
the process. PCA can be expressed using following equation:

O F

C O A S T A L

S C I E N C E S

Parameters
Min
Max
Mean
SD
WHO 2004
pH (Range)
7.45
8.5
7.917
0.26
6.5-8.5
EC (S/cm)
288
8976
2537.7
2553.2
1500
TDS (mg/l)
184.32
5744.6
1624.2
1634.1
1500
Ca (mg/l)
26.4
369
103.4
84.2
200
Mg (mg/l)
28.8
622.6
139.3
135.3
150
Na (mg/l)
88.05
557.25
289.1
131.5
200
K (mg/l)
1.26
102.7
28.3
29.8
12
HCO3 (mg/l)
36.6
196.6
89.8
40.8
500
Cl (mg/l)
306.52
3021.3
888.3
636.3
600
SO4 (mg/l)
37.8
173.5
112.3
42.8
250
Table. 1 Statistical summary of hydro chemical parameters of the study
area

The evaluation of groundwater and its relationship between

rock types and water composition has been evaluated by the Ternary
diagram, which reveals that the water type has Na-Cl facies (Fig.2)
and ionic concentration of sodium, magnesium and chloride are the
dominant cations and anion respectively.
Generally, calcium and bicarbonate are major ions in fresh
water, while chloride and sodium are most dominant in seawater.

Yji = fjlzil + fj2zi2 +..... + fjm+ zim + eij

Where y is the measured variable, f is the factor control loading,
z is the factor score, e is the residual term accounting for errors or
other sources of variation, i and j is the simple number and m is the
total number of factors.
Spearmans correlation matrix is a statistical measure of the
strength of a monotonic correlation between paired data which can
be employed to study the degree of relation between two variables
and it is simply measure to demonstrate how well one variable
predicts the other. The interpretation of the data r2 value has >0.19 is
very weak, 0.20 to 0.39 is weak, 0.40 to 0.59 is moderate, 0.60 to 0.79
is strong and 0.80 to 1.0 the very stronger. The monotonic Fig. 2 Groundwater Chemistry (Ternary plot)
relationship or the r2 value has >1.0 is perfect correlation, 0 to 1 is
the two variables tend to increase or decrease together, 0.0 is the two
Therefore, the higher concentration of chloride and sodium ions
variables do not very together at all, -1 to 0 is one variable increases in the coastal groundwater attribute to significant effect of saltwater
as the other decreases and <-1.0 have perfect negative or inverse intrusion.
correlation.

4. Results and discussion

4.1. Hydrochemical characteristics
Statistical summary of hydrochemical parameters of the
groundwater samples are shown in Table 1. The groundwater
samples of the study area have pH values ranging from 7.45 - 8.5
with an average value of 7.92, which indicates that the groundwater
is slightly alkaline in nature and found within the permissible limit
suggested by WHO (2004). The electrical conductivity (EC) values
ranged 288 - 8,976 S/cm with an average value of 2537.7S/cm.
The total dissolved solids values vary between 184 to 5745 mg/l and
average value of 1624 mg/l. The abnormal high value of EC and TDS
indicates the possibility of saline intrusion in the coastal area. The
order of abundance of the major cations is Na >Mg > Ca > K. In
sodium, 70% of the samples exceeded the desirable limit of 200
mg/l, magnesium 61% of the samples exceeded the permissible limit
of 150 mg/l and calcium, potassium was mostly found within the
desirable limit. The abundance of the major anions is Cl > SO4> HCO3
and almost 87% of the samples exceeded the desirable limit of Cl
(600 mg/l), but the sulfate and bicarbonate concentrations are all
below WHO guidelines (WHO, 2004).
24

4.2. Principal Component Analysis (PCA)

The PCA was applied on the water quality dataset to identify the
spatial sources of pollution within the southern coastal aquifer. With
reference to the eigenvalues (>1) four components were extracted
and explained 91% of the total variance (Table 2)
In table 2, PC 1 accounted for 55% of the total variance. This
component showed high loading of EC, TDS, Cl, Mg, Ca, Na and SO4.
The electrical conductivity and total dissolved solids are generally
regarded as indicator for the presence of dissolved ions; these two
concentrations increases at the same dissolved ions will also
increase. Salinity will also increase at the time of increase in
dissolved salt concentration (Elder 1988). This statement is also
supported by the strong positive component loadings for EC and TDS.
The high absolute loadings for the major ions such as Cl, Mg, Ca, Na
and SO4 probably shows the result of mineral water reactions and it
indicates that the seawater intrusion still influence the
hydrochemistry of the study area. It may also be explained by the
mixing condition between freshwater and seawater (Aris et al. 2012),
and occurrence of dolomitic and evaporite minerals in the aquifer. PC
2, which accounts for 19.20% of the total variance, contains high
loadings for pH, K, and HCO3. Its represents the major geochemical
processes taking place in the study area, which is most probably
ORIGINAL

ARTICLE

J O U R N A L

O F

C O A S T A L

S C I E N C E S

due to the presence and dissolution of some carbonate, activities. Insignificant correlation (r=<0.5) on calcium and
agricultural, industrial and anthropogenic activities. PC3 and PC4 bicarbonate attribute to a fresh water source.
which explain about 0.9 and 0.8%, of the variance of the
concentration of the samples, extreme negative scores reflect areas 5. Conclusion
essentially unaffected by the process.
The result of the statistical analysis including Spearmens
Parameters
Component
correlation coefficient and principle component analysis, as applied
PC1
PC2
PC3
PC4
to the hydrochemical data set of groundwater in this coastal area
pH
-0.078
0.715
-0.002
0.675
provides an insight into the underlying factor controlling hydroEC
0.956
0.125
-0.043
0.083
geochemical processes in the region. The evaluation of groundwater
TDS
0.956
0.125
-0.043
0.083
and relationship between rock types and water composition has
Ca
0.749
-0.565
0.188
0.065
been evaluated by the ternary diagram. It reveals that the water type
Mg
0.833
-0.492
0.067
0.151
has dominant Na-Cl facies which indicates possibility of seawater
Na
0.726
0.215
-0.528
0.062
intrusion in the study area. Principal component analysis identified
K
0.504
0.539
0.498
-0.268
by several intrinsic factors responsible for groundwater pollution
either from natural or anthropogenic inputs.
HCO3
0.593
0.440
-0.382
-0.454
Four components were extracted with a total variance of 91%
Cl
0.899
-0.357
-0.031
0.122
and
high
loading of EC, TDS, Cl, Mg, Ca, Na and SO4. The PCA results
SO4
0.707
0.400
0.386
-0.044
recommended
that seawater intrusion, industrial pollution,
Initial Eigen values
5.529
1.921
0.867
0.794
agricultural and anthropogenic activities. The Spearmens
Percent of variance
55.293
19.2013
8.667
7.943
correlation indicate strong, positive monotonic correlation between
Cumulative percent
55.293
74.506
83.172
91.115
EC, TDS with major ionic concentration of Mg2+, Na+, Cl- and SO42-.
Table. 2 Summarized the PCA results including the loadings and the This has indicated seawater influence on the groundwater salinity in
eigenvalues of each PC
the coastal areas. This study indicates that there is an ongoing
problem of salt water intrusion in the study area.
4.3 Correlation matrix

Acknowledgements

Based on Spearmans correlation, it was found that the correlation

between EC (Salinity) and TDS with the major components of sea
water (Na, Mg, Cl, SO42-) showed significant positive correlation
(TDS-Mg2+, r=0.756; TDS-Na+, r=0.565; TDS-Cl-, r=0.828 and TDSSO42- , r=0.712 (Table 3) that reflects signs of sea water influence on
the groundwater salinity.
Parameters
pH
EC
TDS

pH
1
-0.035
-0.035

EC

TDS

1
0.999

Ca
Mg

-0.382
-0.124

0.514
0.756

Na
K
HCO3
Cl
SO4

0.078
-0.031
-0.205
-0.157
0.129

0.565
0.721
0.398
0.828
0.712

The authors would like to thank the Department of Science and

Technology (DST) (WAR for WATER scheme) Govt. of India, New
Delhi. This work forms a part of the major research project (Project
No: DST/WAR-W/SWI/010/2010) granted to Prof. N. Chandrasekar.

Ca

Mg

0.514
0.756

1
0.701

0.565
0.721
0.398
0.828
0.712

0.109
0.457
0.130
0.574
0.440

0.368
0.386
0.354
0.780
0.628

Na

HCO3

Cl

SO4

1
0.275
0.497
0.782
0.428

1
0.368
0.509
0.682

1
0.393
0.414

1
0.609

Table. 3 Spearmans correlation matrix of the groundwater variables (values higher than 0.6 bold are of significant correlations)
The significant correlation of Cl- ion with Na+ (Cl-- Na+, r=0.782)
and Cl- with Mg2+ (Cl-- Mg2+, r=0.780) reflects the seawater intrusion
to groundwater in the study area. The variation of these
relationships may also be inferred to indicate the complexity of the
hydrochemical components of groundwater where natural water
always contain dissolved and suspended substances of mineral origin
and the processes govern changes in the groundwater composition
are mainly associated with over abstraction, biodegradation,
carbonate saturation and seawater intrusions (Elkarail and Obied
2013).
Positive correlation of Mg2+ with SO42- (Mg2+-SO42-, r=0.628) has
undergone reduction, precipitation, solution and concentration as
they traverse through the sedimentary rocks such as gypsum and
anhydrite and K+ ion with SO42- (K+- SO42-, r=0.682). Groundwater is
deteriorated due to using fertilizers in agriculture and anthropogenic
25

References
Abdul Halim, M.D., Mujamder, R.K., Nessa, S.A., Oda, K., Hiroshiro, Y., Jinno, K.,
2010. Arsenic in shallow aquifer in the eastern region of Bangaladesh:
insights from principal component analysis of groundwater compositions.
Environmental Monitoring and Assessment 161, 453-472.
Akbal, F., Grel, L., Bahadr, T., Gler, ., Bakan, G., Bykgngr, H., 2011.
Multivariate statistical techniques for the assessment of surface water
quality at the mid Black sea coast of Turkey. Water Air and Soil Pollution
216(1), 21-37.
APHA, 2005. Standard methods for the examination of water and wastewater.
American Water Works Association, Water environment Federation,
Washington DC.
Aris, A.Z, Praveena, S.M, Abdullah, M.H, Radojevic, M., 2012. Statistical
approaches and hydrochemical modeling of groundwater system in small
tropical island. Journal of Hydroinform 14(1): pp 206-220. Doi:
10.2166/hydro.211.072
ORIGINAL

ARTICLE

J O U R N A L

O F

C O A S T A L

S C I E N C E S

Chandrasekar, N., Selvakumar, S., Srinivas, Y., John Wilson, J.S., Simon Peter, T.,
Magesh, N.S., 2013. Hydrogeochemical assessment of groundwater quality
along the coastal aquifers of southern Tamil Nadu, India. Environmental
Earth Sciences doi 10.1007/s12665-013-2864-3
Elder, J.F., 1988. Metal biochemistry in surface water systems: a review of
principles and concepts. In: U.S. Geological Survey circular, Issue 1013.
Development of the Interior, U.S. Geological Survey, Washington, D.C
Elkarail, A.B., Obied, B.A., 2013. Hydrochemical characterization and
groundwater quality in Delta Tokar alluvial plain, Red sea coast, Sudan.
Arabian Journal of Geosciences 6(8), 3133-3138.
Hem, J.D., 1985. Study and interpretation of the chemical characteristics of
natural water, USGS water supply Paper-2254, p 264l
Hitchon, B., Billings, G.K., Kolvam, J.E., 1971. Geochemistry and origin of
formation waters in the western Canada sedimentary basin. III factor
controlling chemical composition. Geochemica, Cosmochimica Acta 35,
680-687.
Krishna Kumar, S., Chandrasekar, N., Seralathan, P., Prince S. Godson., Magesh,
N.S., 2012. Hydrogeochemical study of shallow carbonate aquifers,
Rameswaram Island, India. Environmental Monitoring and Assessment
184, 41274138
Mrklas, O., Bentley, L.R., Lunn, S.R.D., Chu, A., 2005. Principal Component
Analyses of groundwater chemistry data during enhanced bioremediation.
Journal of Water, Air, and Soil Pollution 169: 395411
Najar, I., Khon, A., 2012. Assessment of water quality and identification of
pollution sources of three lakes in Kashmir, India using multivariate
analysis. Environmental Earth Sciences 66(8), 2367-2378.
Ragunath, H.M., 1987. Groundwater, 2nd edn. Wiley Eastern Ltd, New Delhi, p
563
Sarath Prasanth, S.V., Magesh, N.S., Jitheshlal, K.V., Chandrasekar, N.,
Gangadhar, K., 2012. Evaluation of groundwater quality and its suitability
for drinking and agricultural use in the coastal stretch of Alappuzha
District, Kerala, India. Applied Water Science 2, 165175
Senthilkumar, G., Ramanthan, A.L., Nainwal, H.C., Chidambaram, S., 2008
Evaluation of the hydro geochemistry of groundwater using factor analysis
in the Cuddalore coastal region, Tamil Nadu, India. Indian Journal of Marine
Sciences 37(2), 181-185.

26

ORIGINAL

ARTICLE