Water Science

ScienceDirect
Water Science 31 (2017) 52–66

journal homepage: www.elsevier.com/locate/wsj

Evaluation of heavy metal contamination using environmetrics and

indexing approach for River Yamuna, Delhi stretch, India
Richa Bhardwaj, Anshu Gupta, J.K. Garg ∗
University School of Environment Management, Guru Gobind Singh Indraprsatha University, Sector 16-C, Dwarka 110078, New Delhi, India
Received 11 September 2016; received in revised form 15 January 2017; accepted 20 February 2017
Available online 18 April 2017

Abstract
The objective of the present study is to investigate the current status of heavy metal pollution in River Yamuna, Delhi
stretch. The concentrations of Nickel, Cadmium, Chromium, Copper, Iron, Lead, and Zinc in water samples have been stud-
ied during December 2013–August 2015. The overall mean concentration of heavy metals was observed in the following order
Fe > Cu > Zn > Ni > Cr > Pb > Cd. Correlation analysis formed two distinct groups of heavy metals highlighting similar sources. This
was further corroborated by results from principal components analysis that showed similar grouping of heavy metals (Ni, Zn, Fe,
Pb, Cd) into PC1 having one common source for these heavy metals and PC2 (Cu, Cr) having another common source. Further, our
study pointed out two sites i.e. Najafgarh drain and Shahdara drain outlet in river Yamuna as the two potential sources responsible
for the heavy metal contamination. Based on heavy metal pollution index value (1491.15), we concluded that our study area as a
whole is critically polluted with heavy metals under study due to pollutant load from various anthropogenic activities.
© 2017 National Water Research Center. Production and hosting by Elsevier B.V. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: River Yamuna; Heavy metal pollution; Principal component analysis; Correlation analysis; Heavy metal pollution index; Industrial
effluents

1. Introduction

River water has been cradles of many civilizations and is responsible for supporting and maintaining various forms
of life. However, rapid urbanization with associated economical and industrial development has exerted tremendous
pressure on this vital resource leading to the impairment of water quality and various ecosystem services. River Yamuna,
one of the major rivers of India, is facing serious challenges for its very survival mainly because of the absence of
ecological water flow and various other anthropogenic activities. This severely polluted river has an extensive catchment
area covering several states and is being widely utilized for various domestic, agricultural, and industrial uses. In addition
to this, the river also acts as a sink for sewage, industrial effluents and agricultural runoff. However, lack of common
policy for discharge of waste into the river as well as lack of integrated approach by these states towards the restoration
of current state of river, has left this river unsuitable for providing its designated ecological use.

∗ Corresponding author. Fax: +91 11 2530 2111.

E-mail addresses: richa.em@ipu.ac.in, mirabilia.du@gmail.com (R. Bhardwaj), anshurcy@yahoo.com (A. Gupta), gargjk@gmail.com,
gargjk113@gmail.com (J.K. Garg).

http://dx.doi.org/10.1016/j.wsj.2017.02.002
1110-4929/© 2017 National Water Research Center. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 R. Bhardwaj et al. / Water Science 31 (2017) 52–66 53

The situation of River Yamuna worsens when the river leaves National Capital Territory of Delhi. Despite of the
fact that this region has only 1% of the river’s total catchment area, it contributes more than 50% of the pollutants
found in the Yamuna (Sehgal et al., 2012). Delhi, being the capital of India is an important centre for commerce, trade
and industry in northern India and has urbanized at a very fast rate with its population increasing from 13.9 million in
2001 to 16.8 million in 2011 (Census of India, 2011). The River Yamuna in Delhi stretch is an only natural resource for
sustaining all forms of life in this city. However, the unplanned development and perennial increase of population in
the National Capital Territory has placed tremendous pressures of water supply and sanitation on river Yamuna (Kaur,
2007).
Rapid urbanization and population growth in fast growing cities leading to industrialization poses a major threat of
Heavy metal pollution for India’s rivers flowing through these cities (Sundaray et al., 2006; Karbassi et al., 2007; Akoto
et al., 2008; Ahmad et al., 2010). These heavy metals can enter river systems from either natural or anthropogenic
sources (Akoto et al., 2008). The main anthropogenic sources are disposal of untreated and partially treated industrial
effluents and sewage containing toxic metals, as well as metal chelates from different industries and indiscriminate
use of heavy metal-containing fertilizers and pesticides in agricultural fields (Reza and Singh, 2010; Abbasi et al.,
1998). Several researchers have studied heavy metal contamination in various Indian rivers with respect to industrial,
municipal, and domestic pollution (Sundaray et al., 2012; Aktar et al., 2010; Sundaray, 2009; Prasad et al., 2006; Jain
and Sharma, 2006; Patil & Shrivastava, 2003; Nayak et al., 2001; Manjunatha et al., 2001; Jameel, 2001).
The water quality monitoring of River Yamuna has indicated significant presence of several heavy metals in its
water (Rawat et al., 2003; CPCB, 2006; Jain, 2009; Kaur & Mehra, 2012; Sehgal et al., 2012; Malik et al., 2014). The
Yamuna River stretch in Delhi has been found to be contaminated with moderate to high levels of heavy metals at
various sites which could be related to the untreated industrial discharge, lead battery-based industrial units, vehicular
pollution, sewage discharge and surface run-off from contaminated areas. Rawat et al. (2003) concluded that there is
high concentration of most of the heavy metals suspended in wastewater originating from industrial area, indicating
that these industries could be one of the primary sources of heavy metals in the Yamuna, along with fertilizers and
pesticides laden agricultural runoff.
The catchment area of River Yamuna in Delhi is highly urbanized and is networked by several drains. Najafgarh
and Shahdara drains are the major drains that discharge a heavy load of pollutants into the river. Although, Sewage
treatment plants (STPs) have been constructed in various parts of Delhi in order to improve the water quality of the
river, the treated, or partially treated sewage from these STPs is continuously being discharged directly or through
the carrier drains into the river (CSE India, 2007). Also, many a times untreated sewage goes directly into the river
at few locations due to non-operationality of STPs because of power failures, mechanical problems or maintenance
issues, which further compounds water quality issues (CPCB, 2006). The accessibility for disposal of wastewaters
from various sources into this river makes it vulnerable to heavy metal pollution besides loading with other pollutants.
Monitoring of heavy metal contamination is important because increased concentration of heavy metals in potable
water increases the threat to human health and to the environment due to biological magnification. Assessment of
monitoring data using multivariate statistical techniques, like principal component analysis coupled with metal con-
centration analysis and correlation analysis could be used to determine the causes for the deterioration of water quality
and to identify highly polluted stretches within a given river system (Nair et al., 2010; Yalcin et al., 2010). In addition
to this, the sitewise disposition of heavy metal by producing heavy metal pollution index can be helpful in identifying
and quantifying trends in water quality (Reza & Singh, 2010; Prasad & Kumari, 2008) and can provide an accumulated
assessment of overall water quality in a form that could be utilized by policy makers for regulation and control of
pollution.
Majority of the heavy metal studies conducted on river Yamuna have merely presented the concentrations of various
heavy metals in the Yamuna River. However, in our study we have integrated various statistical techniques in order to
evolve and validate major sources of heavy metals in Delhi stretch. This will also help in understanding the impacts
of various industrial and domestic activities in and around Delhi on heavy metal pollution in the river Yamuna. We
also prepared the most recent heavy metal pollution index of River Yamuna, to determine the most critically polluted
stretches along the length of river.
54 R. Bhardwaj et al. / Water Science 31 (2017) 52–66

Table 1
Description of sampling sites.
Site code Site name Location Description

S1 Wazirabad Barrage 28◦ 43 23 N Latitude and This site is the entry point of River Yamuna in Delhi
77◦ 14 40 E Longitude
S2 Najafgarh Drain 28◦ 42 40 N Latitude and It is the major polluting drain of River Yamuna in Delhi
(Outlet in the river) 77◦ 13 55 E Longitude
S3 Old Yamuna Iron Bridge 28◦ 39 36 N Latitude and This is a place where there are many inhabitants residing
in the riverbank and it reflects the impact of faecal
77◦ 14 59 E Longitude discharge
S4 ITO Barrage 28◦ 37 49 N Latitude and It is the intermediate point having high amount of
77◦ 15 20 E Longitude pollution
S5 Nizamuddin Bridge 28◦ 36 20 N Latitude and This is a place where there are few inhabitants with a
77◦ 15 44 E Longitude visible presence of small houses
S6 Okhla Barrage 28◦ 32 44 N Latitude and It is the exit point of River Yamuna from Delhi
77◦ 18 57 E Longitude
S7 Shahdara Drain 28◦ 31 44 N Latitude and It is the second major polluting drain of River Yamuna
(Outlet in the river) 77◦ 16 57 E Longitude in Delhi located in the east of Okhla Barrage

2. Materials and methods

2.1. Study area

The study was conducted at designated sampling points representative of the River Yamuna in the Delhi stretch. On
its passage, River Yamuna transverses a distance of about 22 km as it pass through the city. Seven sampling sites were
chosen which receive waste from various point and non-point sources including municipal and domestic sewage. The
sites Najafgarh drain intermixing zone (S2) and Shahdara drain intermixing zone (S7) were chosen to assess changes
in water quality due to impact of drains into the river. Fig. 1 gives locational details of the study area while Table 1
provides the geographical setting and description of sampling sites.

2.2. Sampling period

Sampling was conducted for Pre-Monsoon (February–May), Monsoon (June–September) and Post-Monsoon
(October–January) in the years 2013–2014 and 2014–2015. The seasonal reversal of winds leading to rainfall gives
rise to Monsoon (Southwest Monsoon, to be specific) over India from June to September (Indian Meteorological
Department, 2017). Samples were collected in different seasons in view of the fact that the seasons affect the level and
fate of the contaminants in the river to a great extent.

2.3. Sample collection and preservation

The samples were collected using grab sampling technique from midstream of the river wherever possible or from
well-mixed zone at all the sampling points from a depth of about 0.3 m in triplicate in 0.5 L high-grade polyethylene
bottles. They were then mixed to have a composite sample. The sampling bottles used were previously soaked and
rinsed in 10% HNO3 overnight and the collected unfiltered samples were acidified by adding 2 mL of conc. HNO3 per
litre of the sample to avoid precipitation of heavy metals. The bottles were capped tightly and stored at 4 ◦ C to prevent
evaporation (Singh et al., 2005; Sehgal et al., 2012).

2.4. Determination of temperature and total dissolved solids (TDS)

The temperature values of water samples were recorded using a potable thermometer (HACH-HQ30d) and the TDS
were estimated by using Gravimetric analysis and Filtration method following standard methodologies (APHA, 2012).
 R. Bhardwaj et al. / Water Science 31 (2017) 52–66 55

Fig. 1. Map of River Yamuna representing sampling sites.

56 R. Bhardwaj et al. / Water Science 31 (2017) 52–66

2.5. Heavy metal analysis

All chemicals used in the study were of analytical grade and obtained from Merck, India. De-ionized water was
used in various water quality protocols followed in the study. For heavy metal estimation, 50 mL of well-mixed, acid
preserved samples were taken in an acid-washed beaker and 10 mL conc. HNO3 was added to it. The mixture was
digested on a hot plate at 90 ◦ C till the volume got reduced to 10–20 mL. Final volume was made up to 50 mL by
addition of de-ionized water that was followed by filtration using Whatman no. 42 filter paper. The digested filtrates were
used for the metal quantification using Atomic Absorption Spectrophotometer (Shimdzu AA-6300), with setting of
different characteristic wavelengths of metals using hollow cathode lamps and directly aspirating the digested samples
into air–acetylene flame. The instrument was calibrated by analyzing known concentration of heavy metals. Standard
solutions (1000 mg L−1 ) procured from Merck were serially diluted to obtain the desired concentrations and for each
metal, a multi-point calibration graph was prepared. During the analysis, a blank was run after every 10 samples to
examine the instrument’s performance for minimization of any errors. The concentrations of each heavy metal in every
sample were determined three times, and the results were expressed as the mean concentration of heavy metal in the
given sample (Singh et al., 2005; Sehgal et al., 2012).

2.6. Statistical data analysis

Analysis of Variance (ANOVA) and post hoc (Tukey T-test) were used to analyze the significant variation in
heavy metal concentration among different sampling sites and different seasons for all studied heavy metals (normally
distributed).
Pearson correlation matrix was applied to identify the relationship between the seven elements (significance level
0.05 and 0.01).
Principal component analysis (PCA) was employed to infer the hypothetical source of heavy metals. It has emerged
as a useful tool for better understanding of the relationships among the variables (metal concentration in the present
study) and for revealing groups that are mutually correlated within a data body assisting in the identification of sources
of various pollutants. To determine the suitability for conducting PCA, Kaiser-Meyer-Olkin (KMO) and Barlett’s tests
of sphericity have to be initially performed on the water quality data. KMO test is a measure of sampling adequacy and
its value lies between 0 and 1. Smaller values close to 0 indicate inappropriateness of conducting the PCA, values more
than 0.6 are considered satisfactory, while values close to 1 increase the reliability of PCA. Barlett’s test of sphericity
is used to determine whether the correlation matrix is an identity matrix. If the correlation matrix is found to be an
identity matrix then all correlation coefficients become zero and variables become unrelated. In this scenario PCA
becomes inappropriate and unsuitable for data analysis. P < 0.05 is considered significant for the Barlett’s test (Nair
et al., 2010).
The data were statistically analyzed using the statistical package, SPSS 16.0 (SPSS, USA).

2.7. Data evaluation

Heavy metal pollution index (HPI) is a used to determine the aggregate influence of individual heavy metal on the
overall quality of water. It is a rating approach that assigns weightage (Wi ) to every parameter, reflecting the relative
importance of individual quality considerations in a composite way or Wi can be assessed by making values inversely
proportional to the recommended standard (Si ) for the corresponding parameter (Horton, 1965; Mohan et al., 1996).
The value lies between zero and one. The highest tolerant value for drinking water (Si ) refers to the maximum allowable
concentration in drinking water in the absence of any alternate water source. The maximum desirable value (Ii ) indicates
the standard limits for the same parameters in drinking water. For computing HPI, the Bureau of Indian Standards
(BIS) IS: 10500 (2012) for drinking water for each heavy metal (chemical parameter) in ␮g L−1 was considered. Water
quality and its suitability for drinking purpose can be examined by HPI model (Mohan et al., 1996; Prasad and Kumari,
2008; Prasad and Mondal, 2008).
The HPI model (Mohan et al., 1996) is given by
n
i=1 WiQi
HPI =  n (1)
i=1 Wi
 R. Bhardwaj et al. / Water Science 31 (2017) 52–66 57

Table 2
Seasonal statistics of heavy metal concentration in River Yamuna.
Parameters Post-Monsoon (2013 and 2014) Monsoon (2014 and 2015) Pre-Monsoon (2014 and 2015)
Mean ± standard deviation Mean ± standard deviation Mean ± standard deviation

Temperature (◦ C) 17.46 ± 1.09 25.41 ± 1.77 29.90 ± 1.92

Electrical conductivity (S M−1 ) 684.51 ± 634.47 536.77 ± 388.61 1260.60 ± 545.70
Nickel (mg L−1 ) 0.0428 ± 0.0559 0.2324 ± 0.3507 0.8515 ± 0.9549
Zinc (mg L−1 ) 0.3915 ± 1.0693 0.5923 ± 0.6942 3.5185 ± 7.2573
Copper (mg L−1 ) 0.0650 ± 0.0713 1.5996 ± 1.0606 4.7911 ± 4.7757
Chromium (mg L−1 ) 0.0181 ± 0.0298 0.0606 ± 0.0858 0.3627 ± 0.6744
Cadmium (mg L−1 ) 0.0063 ± 0.0034 0.0265 ± 0.0184 0.1101 ± 0.1462
Iron (mg L−1 ) 3.7834 ± 5.2949 8.2134 ± 5.7059 19.4672 ± 12.4086
Lead (mg L−1 ) 0.0333 ± 0.0301 0.0606 ± 0.0281 0.2553 ± 0.3635

where, Qi is the sub-index of the ith parameter. Wi is the unit weightage of the ith parameter and n is the number of
parameters considered.
The unit weightage (Wi ) of the parameter is calculated by

W i = k/S i (2)

where, Si the maximum allowable recommended standard for ith parameter and k is the constant of proportionality.
The sub-index (Qi ) of the parameter is calculated by

|Mi − Ii|
Qi = ni=1 × 100 (3)
Si − Ii

where, Mi is the monitored value of heavy metal of ith parameter, Ii is the desirable value of the ith parameter and Si
is the standard maximum allowable value of the ith parameter in ␮g L−1 . The quantity [Mi −Ii ] indicates numerical
difference of the two values, ignoring the algebraic sign that is the absolute value. Generally, the critical pollution
index of HPI value for drinking water is 100 (Prasad and Bose, 2001). Using Eqs. (1)–(3) HPI values for River Yamuna
overall water quality as well as for each individual site were determined.

3. Results and discussion

3.1. Seasonal distribution of heavy metals

Seasonal variation in the concentrations of heavy metals in water systems, apart from the anthropogenic factors, is
primarily influenced by pH and temperature (Sundaray et al., 2012) that further influences the values of total dissolved
solids (TDS) as well as electrical conductivity. In our study, a similar trend in the average concentrations of heavy
metals vis a vis average values of temperature and TDS seasonally has been observed (Table 2). All these parameters
reported highest average values in Pre-Monsoon, followed by Monsoon and lowest values in Post-Monsoon.
For instance, seasonally, Nickel recorded highest average concentration for Pre-Monsoon (2014 and 2015)
0.8515 ± 0.9549 mg L−1 , followed by Monsoon (2014 and 2015) 0.2324 ± 0.3507 mg L−1 and lowest for Post-
Monsoon (2013 and 2014) 0.0428 ± 0.0559 mg L−1 . Similar pattern was recorded for the remaining heavy metals.
Results indicated a significant (p < 0.05%, ANOVA, post-hoc) difference within mean concentration values of all
heavy metals for Pre-Monsoon, Monsoon and Post-Monsoon for two years. The significant difference was reported
between Pre-Monsoon and Monsoon as well as between Pre-Monsoon and Post-Monsoon. However, no significant
difference was obtained between Monsoon and Post-Monsoon concentrations. Highest values during Pre-Monsoon are
attributed to the high evaporation rate of surface water followed by high temperature and low flow condition of the
river leading to the accumulation of the heavy metals. Lohani et al. (2008) and Kumar et al. (2013) have also reported
high concentrations of heavy metals in Pre-Monsoon for Gomti, and Sabarmati rivers that corroborates our results.
58 R. Bhardwaj et al. / Water Science 31 (2017) 52–66

3.2. Overall distribution of heavy metals

The presence of heavy metals was observed in all the samples throughout the sampling period along the stretch of the
River in Delhi. The three essential micronutrients viz. Fe, Cu and Zn have overall mean values of 10.4880, 2.1518 and
1.5007 mg L−1 with concentrations within the range (minimum and maximum) 0.8785–53.9401, 0.0184–17.6424 and
0.0154–17.6424 mg L−1 , respectively (Table 3). The overall mean value of Fe and Cu exceeded the highest permissible
limits for drinking water prescribed by BIS IS: 10500 (2012). However, for Zn, it was within the limits. As per CPCB
standards for Inland surface water, Fe exceeded the permissible limits where as Cu and Zn were within the permissible
limits. Besides, the overall mean values of four toxic heavy metals viz. Pb, Cd, Cr and Ni were recorded 0.1164, 0.0476,
0.1471, and 0.3755 mg L−1 , respectively in the range (minimum and maximum) of 0.0068–1.1121, 0.0017–0.4330,
0.0026–1.9830, and 0.0014–2.7481 mg L−1 , respectively. In this case, the overall mean values for all these toxic metals,
exceeded well beyond the permissible limits for drinking water prescribed by BIS IS: 10500 (2012). However, as per
CPCB standards for inland surface water, all except Pb, were within the permissible limits.
The above mentioned minimum values for Ni, Zn, Fe and Pb were observed at Wazirabad (S1), for Cu at Nizamuddin
(S5), for Cr at ITO (S4) and for Cd at Lohapul (S3). Likewise the maximum value for Ni, Zn, Fe and Pb were observed at
Shahdara drain intermixing zone (S7), for Cu and Cr at Najafgarh drain intermixing zone (S2), and for Cd at ITO (S4).
The overall mean concentration of heavy metals was observed in the following order Fe > Cu > Zn > Ni > Cr > Pb > Cd.
Fe is present in abundance and Cd is found to be slightest. Similar results were reported by Sundaray et al. (2012) for
Mahanadi river system and by Salah et al. (2012) for Euphrates river system.
River Yamuna in the Delhi stretch has been found to contain higher concentrations of most of the heavy metals when
compared with the concentrations reported for the other Indian rivers (Table 4). However, the concentration of Fe in
the river Yamuna was lower than the observed values for River Kali and River Achankovil. Similarly, the concentration
of Pb was lower than the values observed for River Ganga, River Cauvery.

3.3. Sitewise distribution of heavy metals

The concentration of each heavy metal was also calculated sitewise (average of six values of each heavy metal for
each site). The minimum average concentration of Ni (0.0535 ± 0.04744 mg L−1 ) and Pb (0.0310 ± 0.0206 mg L−1 )
during the sampling period was recorded at Wazirabad (S1). While, the minimum average concentration of
Cd (0.0126 ± 0.0096 mg L−1 ) and Cr (0.0172 ± 0.0101 mg L−1 ) was recorded at Lohapul (S3), on the other
hand, the minimum average concentration of Zn (0.4136 ± 0.4629 mg L−1 ), Fe (5.7029 ± 4.3871 mg L−1 ) and Cu
(0.7463 ± 1.0320 mg L−1 ) were observed at Okhla (S6), ITO (S4), and Shahdara drain intermixing zone (S7), respec-
tively.
The maximum mean values of Zn (5.4129 ± 11.3498 mg L−1 ), Cd (0.1279 ± 0.1414 mg L−1 ), and
Pb (0.3969 ± 0.5504 mg L−1 ) were observed at Shahdara drain intermixing zone (S7) while for Cu
(5.9397 ± 7.7344 mg L−1 ), Cr (0.7563 ± 0.9063 mg L−1 ) and Fe (21.5796 ± 5.8594 mg L−1 ), they were observed
at Najafgarh drain intermixing zone (S2). Maximum average concentration of Ni (1.0456 ± 0.8972 mg L−1 ) was
observed at ITO (S4).
Significant (p < 0.05%, ANOVA) difference exists in mean concentration values of all heavy metals amongst sites
during the sampling period. Sitewise variation in the concentrations of various heavy metals in water system reflected
the impact of anthropogenic activities that are of diverse nature and contaminants are irrevocably added without any
limiting measure. In our study, we found that the overall maximum values and maximum average concentrations of
heavy metals were observed predominantly at S2 and S7 (Najafgarh and Shahdara drain intermixing zones) and were
relatively higher than concentrations measured at other sites (Table 5).
Site S2, had both overall maximum value as well as maximum average concentration for Cu and Cr and this can
be attributed to the discharges carried by Najafgarh drain. This drain receives effluents from electroplating, steel and
pulp industries that are predominant in Jhajjar, Bahadurgarh, Rohtak and Gurgaon area of Haryana and Wazirabad and
Mangolpuri area of Delhi. The major contaminants Cu and Cr along with other heavy metals are added to Najafgarh
drain via supplementary drains arising from these areas and finally the contents of this drain are emptied into the river
at S2. Moreover, it receives domestic sewage and run off from extensive farmed areas that are source of copper.
Site S7 reported overall maximum values for Ni, Zn, Fe and Pb as well as maximum average concentration for Zn
and Pb. This signifies that apart from the cumulative effect of these metals being present at all the sites downstream the
Table 3
Summary of descriptive statistics for heavy metal concentrations in mg L−1 of River Yamuna water.

R. Bhardwaj et al. / Water Science 31 (2017) 52–66

Minimum Maximum Mean (42 Std. Skewness Kurtosis BIS IS: 10500 (for drinking CPCB (for
samples) deviation water) inland surface
water)
Heavy Statistic Statistic Statistic Std. error Statistic Statistic Std. error Statistic Std. error Acceptable Highest Permissible
metal (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 ) permissible (mg L−1 )
(mg L−1 )

Ni 0.0014 2.7481 0.3755 0.1036 0.6717 2.337 0.365 4.665 0.717 0.02 NRa 3.0
Zn 0.0154 28.5209 1.5007 0.6780 4.3940 5.951 0.365 37.157 0.717 5 15 5.0
Cu 0.0184 17.6424 2.1518 0.5246 3.3999 3.398 0.365 13.015 0.717 0.05 1.5 3.0
Cr 0.0026 1.9830 0.1471 0.0638 0.4134 4.050 0.365 16.051 0.717 0.05 NRa 2.0
Cd 0.0017 0.4330 0.0476 0.0146 0.0946 3.029 0.365 8.506 0.717 0.003 NRa 2.0
Fe 0.8785 53.9401 10.4880 1.6379 10.6151 2.075 0.365 5.856 0.717 0.3 NRa –
Pb 0.0068 1.1121 0.1164 0.0353 0.2290 4.115 0.365 16.560 0.717 0.01 NRa 0.1
a NR: no relaxation.

59
60 R. Bhardwaj et al. / Water Science 31 (2017) 52–66

Table 4
Comparison of dissolved metal concentration with other Indian rivers (␮g L−1 ).
Rivers Metals (␮g L−1 ) References

Ni Zn Cu Cr (total) Cd Fe Pb

Mahanadi 15.75 29.28 8.39 3.67 – 189.38 19.13 Sundaray et al. (2012)
Ganga 140 60 10 – 5 800 120 Aktar et al. (2010)
Brahamani 24.78 31.56 6.67 10.89 – 481.78 1.67 Sundaray (2009)
Achankovil – 415 224 – 6.0 11858 72 Prasad et al. (2006)
Hindon 24.0 58.0 6.6 15.0 – 226 37.0 Jain and Sharma (2006)
Godavari 0.14 0.66 0.92 – 0.07 7.50 5.60 Patil & Shrivastava (2003)
Baitarani 3.9 272.3 3.45 9.6 – 100.50 3.45 Nayak et al. (2001)
Kali 0.80 10.03 1.34 – 0.13 51,000 0.70 Manjunatha et al. (2001)
Cauvery 38.3 81.6 6 48.8 0.43 9.90 136 Jameel (2001)
Yamuna 375.50 1500.70 2151.8 147.10 47.60 10,488 116.40 Present study

Table 5
Heavy metal concentration of River Yamuna at different sampling sites.
Heavy S1 mean S2 mean S3 mean S4 mean S5 mean S6 mean S7 mean
metal (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 ) (mg L−1 )

Ni 0.0535 0.2244 0.0610 1.0455 0.3195 0.0635 0.8612

Zn 0.4494 2.3159 0.4891 0.7151 0.7091 0.4136 5.4129
Cu 2.2552 5.9397 2.0625 1.7959 1.1801 1.0833 0.7463
Cr 0.0205 0.7563 0.0172 0.0269 0.0275 0.0243 0.1572
Cd 0.0174 0.0165 0.0126 0.1276 0.0155 0.0156 0.1279
Fe 11.2652 21.5796 8.4319 5.7029 5.9626 6.5462 13.9277
Pb 0.0310 0.1177 0.0406 0.0486 0.0820 0.0978 0.3969

entry point of River Yamuna in Delhi, the Shahdra drain is discharging significant amount of domestic and industrial
effluents at S7, that leads to the observation of maximum values of Ni, Zn, Fe and Pb at this point in comparison to other
sites. The Shahdara drain is the major drain that receives partially or untreated effluents and wastes from neighboring
industrial units as well as discharges of minor drains carrying untreated wastewater (CPCB, 2011). Apart from this, it
also carries municipal sewage and domestic waste from Delhi and joins upstream of Okhla barrage. Primarily, it carries
effluents from the steel processing and electroplating, industries located at Jhilmil, Wazirpur, and Badli area (Sehgal
et al., 2012) and responsible for maximum values of Ni, Fe, Zn. It also carries effluents of lead battery based units that
are a common source of Ni and Pb. Similar reasons are responsible for maximum average values for Zn and Pb reported
at S7. Also, this site reported maximum average value for Cd, which can be again attributed to effluents carried and
discharged by Shahdara drain arising from metal refineries, battery and dye making industries that are source of Cd in
addition to Ni. Though the maximum value of Ni was reported at S7, its maximum average was reported at S4 (ITO),
which may be due to large number of idol immersion observed at this site that contains multicolored paints with high
concentration of Ni. The water at S4 and S5 is widely used for agricultural activities that can lead to runoff carrying
metal-based fertilizers into the river.
Based on the results of the overall as well as sitewise distribution of the heavy metals, we observed an interesting
finding, indicating the overall minimum values of Ni, Zn, Fe and Pb at S1 (relatively less contaminated site), whereas
the overall maximum values of the same set at S7 (shown encircled in Table 6). This observation led us to assume that
the sources of pollution at S7 (as discussed above) have minimal influence at S1, which is responsible for the above
finding. Similarly, the sources of pollution at S2 have led to the maximum observed values for Cu and Cr.

3.4. Correlation analyses and principal component analysis

Two major groups emerged signifying possible similar sources of pollution for each group; one comprising Ni, Zn,
Fe, Pd, Cd and other comprising Cu and Cr (attributed to the sources S7 and S2 explained above, respectively). To
substantiate this, correlation analysis between heavy metal concentrations and principal component analysis (PCA)
 R. Bhardwaj et al. / Water Science 31 (2017) 52–66 61

Table 6
Sitewise allocation of minimum and maximum values and means of heavy metals.
Minimum value Minimumaverage Maximum value Maximumaverage
Sites (Observed for following heavy metals at following sites)

S1 Ni, Pb – –
S2 – – Cu, Cr Cu,Fe, Cr
S3 Cd Cd, Cr – –
S4 Cr – Cd Ni
S5 Cu Fe – –
S6 – Zn – –

S7 – Cu Zn, Cd, Pb

Table 7
Correlation coefficient between elements in the Yamuna River water samples.
Ni Zn Cu Cr Cd Fe Pb

Ni 1 0.616** 0.123 0.128 0.895** 0.444** 0.721**

Zn 1 0.118 0.221 0.521** 0.755** 0.735**
Cu 1 0.900** 0.075 0.542** 0.116
Cr 1 0.073 0.517** 0.254
Cd 1 0.383* 0.640**
Fe 1 0.569**
Pb 1
* Correlation is significant at the 0.05 level (two-tailed).
** Correlation is significant at the 0.01 level (two-tailed).

for river water samples was performed to further statistically prove the existence of two similar groups and to validate
our findings for similar sources of pollution for two groups.
Pearson’s correlation coefficients of’ heavy metals have been summarized in Table 7. Correlation analysis showed
very strong correlation between Ni–Pb (r = 0.721), Pb–Zn (r = 0.735), Zn–Fe (r = 0.755), Ni–Cd (r = 0.895), Ni–Zn
(r = 0.616), Zn–Cd (r = 0.521), Ni–Fe (r = 0.444), Pb–Fe (r = 0.569) and Cd–Pb (r = 0.640) at P < 0.01 level forming
one group of Ni–Zn–Fe–Pd–Cd. Another group represented by Cu–Cr also displayed a significant strong correlation
(r = 0.900, P < 0.01). Heavy metals showing very high correlation indicates same source. Cu–Cr finds its main source
from effluents carried by Najafgarh drain emerging from electroplating, steel and pulp industries as discussed earlier.
Cu and Cr did not show any significant correlation with other heavy metals of previous group, except Fe, thereby
suggesting that these two metals have different sources of contamination. Fe also showed positive correlations with
Cu (r = 0.542) at P < 0.01 and Cd (r = 0.383) at P < 0.05 level indicating its relationship with both groups. Fe comes
mainly from industrial activities/effluents though untreated domestic sewage discharges.
PCA with VARIMAX rotation of the studied heavy metal concentrations was performed. In our study KMO value
was found to be 0.616, which is a good measure of sampling adequacy. Barlett’s test of sphericity produced a highly
significant value of P < 0.001, proving that correlation matrix was not an identity matrix and significant relationships
exist between the variables. These tests (Table 8) confirmed the suitability of water quality data for the PCA.
Tables 8 and 9 show the percentage of variance by different components extracted and the factor loadings of the
different variables, respectively. Eigenvalues indicate the significance of the components. The component with the
highest Eigenvalue is taken to be the most significant and should be one or greater for proper considerations during
PCA. Factor loadings values of >0.75, between 0.75–0.5 and 0.5–0.3 are classified as strong, moderate and weak
respectively, based on their absolute values (Nair et al., 2010).
Two principal components (PC1 and PC2) having Eigenvalues >1 was obtained after applying PCA on the water
quality parameters displaying a cumulative variance of 80.81% (Fig. 2).
PC1 accounts for 48.20% of the total variance (Table 8) and is dominated by the strong factor loadings for Ni
(0.912), Zn (0.823), Cd (0.868) and Pb (0.866), while moderate factor loading for Fe (0.597). Cu and Cr demonstrate
62 R. Bhardwaj et al. / Water Science 31 (2017) 52–66

Table 8
Total variance explained (extraction method: principal component analysis), KMO and Bartlett’s test.
Component Initial Eigenvalues Extraction sums of squared loadings Rotation sums of squared loadings

Total % of Cumulative Total % of Cumulative Total % of Cumulative

variance % variance % variance %

1 3.769 53.839 53.839 3.769 53.839 53.839 3.374 48.205 48.205

2 1.889 26.979 80.818 1.889 26.979 80.818 2.283 32.613 80.818
3 0.717 10.248 91.066
4 0.329 4.701 95.767
5 0.152 2.171 97.937
6 0.098 1.402 99.340
7 0.046 0.660 100.000
Kaiser-Meyer-Olkin measure of sampling adequacy 0.616
Bartlett’s test of Approx. Chi-Square 255.787
sphericity Degrees of freedom 21
Significance 0.000

Table 9
Factor loadings of the principal components extracted.
Parameters Original component Rotated components Communalities

1 2 1 2

Ni 0.823 −0.393 0.912 0.027 0.832

Zn 0.831 −0.184 0.823 0.217 0.724
Cu 0.450 0.845 0.013 0.957 0.916
Cr 0.501 0.810 0.074 0.949 0.906
Cd 0.756 −0.430 0.868 −0.036 0.755
Fe 0.817 0.284 0.597 0.627 0.749
Pb 0.842 −0.257 0.866 0.157 0.775

Extraction method: principal component analysis. 2 components extracted.

Fig. 2. Scree plot for the Eigenvalues indicating two components.

 R. Bhardwaj et al. / Water Science 31 (2017) 52–66 63

Fig. 3. Biplot for the components in rotated space.

low factor loadings in this case indicating their independency within this group (Table 9). PC2 accounts for 32.61% of
the variance (Table 8) and is dominated by Cu and Cr having strong factor loadings of 0.957 and 0.949 respectively
(Table 9). Also, it has moderate factor loading for Fe (0.627). The moderate factor loading for Fe in both the PC1 and
PC2 shows that Fe is present at the source of contamination for both the principal components indicating that it is a
predominant pollutant in the Yamuna River. This is in good agreement with the findings in descriptive analysis and
correlation analysis. The above-mentioned arguments are further illustrated in Fig. 3.
Based on the analyses of descriptive statistics, correlation coefficients and PCA, these seven elements can be clustered
together for their common source and controlling factors. The above trend (Fig. 3) clearly indicates the anthropogenic
influence on the water with respect to the studied heavy metals. The Delhi and the adjoining areas have many small and
mid scale industrial units located nearby the Yamuna River that use these elements for various value added products
and discharge their partly and untreated effluents/wastes into the river. Apart from it, two major drains (Najafgarh and
Shahdara) are the major source of these toxic heavy metals as they are the ones that carry effluents from numerous
supplementary drains (passing through most of the industrialized belts of Haryana, Delhi, Uttar Pradesh, Noida and
carrying industrial and domestic waste) that ultimately drain into the Yamuna River increasing the heavy metals load.
Sewage and other domestic wastes are directly discharged into the river through storm water drains. Traffic sources
and atmospheric depositions also contribute to these elements.

3.5. Heavy metal pollution index overall and sitewise

From the above results, S2 and S7 were identified as the two prime sites responsible for addition of substantial
amount of heavy metals into the river. Discharge of heavy metals from these two sites as well as minor addition of
heavy metals from other sites has rendered the river water unsuitable for major uses, including drinking. Therefore,
we applied heavy metal pollution index (HPI) method, which is an effective tool to characterize the surface water
pollution, and show the composite influence of individual heavy metal on the overall quality of water (Reza and Singh,
2010). In order to calculate the HPI of the water, the mean concentration value of the selected metals (Ni, Fe, Zn, Pb,
Cd, Cu, Cr) have been taken into account (Prasad and Mondal, 2008).
64 R. Bhardwaj et al. / Water Science 31 (2017) 52–66

Table 10
Mean HPI of Yamuna River.
Heavy metals Mean value Standard Highest Unit weightage Sub index (Qi ) Wi × Qi
(␮g L−1 ) Mi permissible desirable value (Wi )
value (␮g L−1 ) (␮g L−1 ) Ii
Si

Fe 10488.00 300 – 0.0033 3496.00 11.53

Cu 2151.80 1500 50 0.00067 144.95 0.097
Zn 1500.70 15,000 5000 0.000067 34.99 0.002
Cd 47.60 3 – 0.34 1586.67 539.46
Pb 116.40 10 – 0.1 1164.00 116.40
Ni 375.50 20 – 0.05 1877.50 93.87
Cr 147.10 50 – 0.02 295.20 5.90
 
Wi = 0.514, Wi Qi = 0.767.25, HPI = 1493.

Table 11
HPI values at different sampling sites.
Serial No. Sampling site HPI Mean deviation % Deviation with
mean HPI value

1 S1 494.96 −996.18 −66.80

2 S2 811.25 −679.89 −45.59
3 S3 405.98 −1085.16 −72.77
4 S4 3419.87 1928.72 129.34
5 S5 673.77 −817.37 −54.81
6 S6 581.98 −909.16 −60.97
7 S7 4050.26 2559.11 171.62

Mean HPI = 1491.15.

Table 10 details the calculations of HPI with unit weightage (Wi) and standard permissible value (Si) as obtained in
the presented study. Though the Fe has highest concentration value in the River Yamuna, nevertheless, the weightage
(Wi) given to Fe is very less, therefore, in evaluation of HPI of river water in total, this parameter do not contribute
much on HPI value. Heavy metals like Pb, Cd, Ni and Cr have been given no relaxation in drinking water standard and
they have been given high weightage (Wi) value in HPI calculation. Hence even their smaller concentration present in
water samples makes the water of poor quality and gives high values in HPI calculations. Concentration of Zn has not
been found at any sampling point and in any season higher than the highest desirable limit of drinking water standard.
The overall heavy metal pollution index value of Yamuna river was found to be 1492 that is far above the critical
pollution index value of 100, above which the overall pollution level should be considered unacceptable (Prasad and
Kumari, 2008; Prasad and Mondal, 2008). This indicates the water is critically polluted with respect to heavy metals.
Such high value of HPI is also reported by a study conducted for River PovPov, Nigeria by Ameh and Akpah, 2011.
Percentage deviation and mean deviation with mean HPI value are also calculated for each sampling site (Table 11).
At all sampling points, HPI calculated is far above the critical index limit of 100 but at sampling sites S4 and S7, the
index value is much higher when compared to other sampling sites. All other sampling sites indicate lower value of
HPI as compared to mean HPI of 1491.15 and also the percentage deviation is negative side. More negative deviation
shows better quality of river water and more positive deviation shows deteriorated quality of river water, relative to
other sites.

4. Conclusion

Among the heavy metals investigated in the River Yamuna water, Fe was found to be most abundant whereas Cd
was reported to be the least among all seven sampling sites. The order of the mean concentrations of tested heavy
metals was recorded to be Fe > Cu > Zn > Ni > Cr > Pb > Cd. The correlation analysis of mean concentrations showed
good to strong positive correlations among Ni, Zn, Fe, Pb, Cd and among Cu and Cr, suggesting that these metals have
common sources.
 R. Bhardwaj et al. / Water Science 31 (2017) 52–66 65

The sitewise distribution pattern indicates large inputs possibly from two major drains (Najafgarh and Shahdara
drain) along with municipal sewage as well as agricultural runoff. The study resulted in the identification of two evident
sites of significant contribution of heavy metals owing to intermixing of outputs of Najafgarh drain at S2 and Shahdara
drain at S7 that carry partially or untreated industrial and domestic effluents from various industrial and residential
areas of Delhi and adjoining states. Heavy metal contamination at other sites was also found out to be significant and
attributed to intermixing of heavy metals added from S2 in addition to the effluents from various STP’s throughout the
length of the river. Along with this, agricultural runoff carrying fertilizers having heavy metals, storm water runoff,
immersion of idols, laundry waste and many more activities could have contaminated the river with heavy metals.
Performing PCA on heavy metals identified two principal components, controlling heavy metal elements variability
in the Yamuna River water. Ni, Zn, Fe, Pb, and Cd (PC1) are thought to be related to one common source, and PC2,
including Cu and Cr, are considered to have originated from another common source. HPI values indicated that River
Yamuna (Delhi stretch) not only as a whole but also sitewise is critically polluted with respect to all the seven heavy
metals studied in the present research.
The comparison of results of various statistical techniques as well as HPI values identified S2 and S7 as major
polluting sites of river Yamuna. This study highlights the need to enhance sewage and industrial treatment mechanisms
and control on the discharge of treated or untreated effluents not only at various sites of River Yamuna, but also for
the effluents draining into Najafgarh and Shahdara drain that ultimately gets discharged into the river. It is therefore
recommended that the river water monitoring agencies like Central Pollution Control Board and Delhi Pollution
Control Committee should enforce more stringent and regular monitoring. Also, measures to regulate the indiscriminate
dumping of domestic waste as well as agricultural and raw sewage runoff into the river should be adhered more
rigorously.
CPCB (2000).

Acknowledgement

This research was supported by Guru Gobind Singh Indraprastha Research Fellowship, New Delhi, India.

References

APHA, AWWA, WEF, 2012. Standard Methods for examination of water and wastewater, 22nd ed. American Public Health Association, American
Water Works Association (AWWA) and Water Environment Federation (WEF), Washington.
Abbasi, S.A., Abbasi, N., Soni, R., 1998. Heavy Metals in the Environment. Mittal Publications, New Delhi.
Ahmad, M.K., Islam, S., Rahman, S., Haque, M.R., Islam, M.M., 2010. Heavy metals in water, sediment and some fishes of Buriganga River,
Bangladesh. Int. J. Environ. Res. 4 (2), 321–332.
Akoto, O., Bruce, T.N., Darko, G., 2008. Heavy metals pollution profiles in streams serving the Owabi reservoir. Afr. J. Environ. Sci. Technol. 2
(11), 354–359.
Aktar, M.W., Paramasivam, M., Ganguly, M., Purkait, S., Sengupta, D., 2010. Assessment and occurrence of various heavy metals in surface water
of Ganga river around Kolkata: a study for toxicity and ecological impact. Environ. Monit. Assess. 160 (1–4), 207–213.
Ameh, E.G., Akpah, F.A., 2011. Heavy metal pollution indexing and multivariate statistical evaluation of hydrogeochemistry of River PovPov in
Itakpe Iron-Ore mining area Kogi State, Nigeria. Adv. Appl. Sci. Res. 2, 33–46, Pelagia Research Library.
2011. Census of India. www.census2011.co.in/census/state/delhi.html.
CPCB, 2006. Water Quality Status Of Yamuna River (1999–2005). Central Pollution Control Board, Ministry Of Environment & Forests, Assessment
and Development of River Basin Series: ADSORBS/41/2006-07.
CPCB, 2011. Control of Urban Pollution. http://www.cpcb.nic.in/oldwebsite/New%20Item/Maintinance-STPS/AK-Sinha-STP2.doc.
CSE India, 2007. Sewage Canal: How to Clean the Yamuna. CSE India.
Horton, R.K., 1965. An index-number system for rating water quality. J. Water Pollut. Control Fed. 37, 300–306.
IMD, 2017. Frequently asked questions. In: Monsoon. Indian Meteorological Department,
http://www.imd.gov.in/pages/monsoon main.php?adta=FAQ&adtb=monsoon (Accessed January 10 2017).
Jain, P., 2009. Sick Yamuna, Sick Delhi—Searching a Correlation. PEACE Institute Charitable Trust, Delhi, India.
Jain, C.K., Sharma, M.K., 2006. Heavy metal transport in the Hindon river basin, India. Environ. Monit. Assess. 112, 255–270.
Jameel, A.A., 2001. A study on the distribution of organic matter and toxic metals in sediments of river Cauvery at Tiruchirapalli. Indian J. Environ.
Prot. 21 (4), 302–304.
Karbassi, A.R., Nouri, J., Ayaz, G.O., 2007. Flocculation of trace metals during mixing of Talar river water with Caspian Seawater. Int. J. Environ.
Res. 1 (1), 66–73.
Kaur, S., Mehra, P., 2012. Assessment of heavy metals in summer and winter seasons in River Yamuna segment flowing through Delhi, India. J.
Environ. Ecol. 3 (1).
66 R. Bhardwaj et al. / Water Science 31 (2017) 52–66

Kaur, R., 2007. Regional resource characterization through remote sensing & GIS for effective decision making—A Case Study of NCR. A note
prepared for an Indo-US workshop on Innovative E-technologies for Distance Education, Extension/Outreach in Efficient Water Management,
March 2007, ICRISAT, Patancheru, AP, India.
Kumar, R.N., Solanki, R., Kumar, J.I.N., 2013. Seasonal variation in heavy metal contamination in water and sediments of river Sabarmati and
Kharicut canal at Ahmedabad, Gujarat. Environ. Monit. Assess. 185 (1), 359–368.
Lohani, M.B., Singh, A., Rupainwar, D.C., Dhar, D.N., 2008. Seasonal variations of heavy metal contamination in river Gomti of Lukhnow city
region. Environ. Monit. Assess. 147, 253–263.
Malik, D., Singh, S., Thakur, J., Singh, R.K., 2014. Review article heavy metal pollution of the Yamuna River: an introspection. Int. J. Curr.
Microbiol. Appl. Sci. 3 (10), 856–863.
Manjunatha, B.R., Balkrishna, K., Shankar, R., Mahalingam, T.R., 2001. Geochemistry and assessment of metal pollution in soils and river
components of a monsoon dominated environment near Karwar, Southwest coast of India. Environ. Geol. 40, 1462–1470.
Mohan, S.V., Nithila, P., Reddy, S.J., 1996. Estimation of heavy metal in drinking water and development of heavy metal pollution index. J. Environ.
Sci. Health A-31, 283–289.
Nair, I.V., Singh, K., Arumugam, M., Gangadhar, K., Clarson, D., 2010. Trace metal quality of Meenachil River at Kottayam, Kerala (India) by
principal component analysis. World Appl. Sci. J. 9 (10), 1100–1107.
Nayak, B.B., Panda, U.C., Panigrahy, P.K., Acharya, B.C., 2001. Dynamics of heavy metals in Dhamara Estuary of Orissa state in India. Chem.
Environ. Res. 10 (3–4), 203–218.
Patil, P.R., Shrivastava, V.S., 2003. Metallic status of river Godavari—a statistical approach. Indian J. Environ. Prot. 23 (6), 650–653.
Prasad, B., Kumari, S., 2008. Heavy metal pollution index of ground water of an abandoned open cast mine filled with fly ash: a case study. Mine
Water Environ. 27 (4), 265–267.
Prasad, B., Mondal, K.K., 2008. The impact of filling an abandoned opencast mine with fly ash on ground water quality: a case study. Mine Water
Environ. 27 (1), 40–45.
Prasad, M.B.K., Ramanathan, A.L., Shrivastav, S.K., Anshumali Rajinder, S., 2006. Metal fractionation studies in surfacial and core sediments in
the Achankovil river basin in India. Environ. Monit. Assess. 121 (1–3), 77–102.
Prasad, B., Bose, J.M., 2001. Evaluation of heavy metal pollution index for surface and spring water near a limestone mining area of the lower
Himalayas. Environ. Geol. 41, 183–188.
Rawat, M., Moturi, M.C.Z., Subramanian, V., 2003. Inventory compilation and distribution of heavy metals in waste water from small-scale industrial
areas of Delhi, India. J. Environ. Monit. 5 (6), 906–912.
Reza, R., Singh, G., 2010. Heavy metal contamination and its indexing approach for river water. Int. J. Environ. Sci. ’Technol. 7 (4), 785–792.
Salah, E.A.M., Zaidan, T.A., Al-Rawi, A.S., 2012. Assessment of heavy metals pollution in the sediments of Euphrates River, Iraq. J. Water Resour.
Prot. 4 (12), 1009.
Sehgal, M., Garg, A., Suresh, R., Dagar, P., 2012. Heavy metal contamination in the Delhi segment of Yamuna basin. Environ. Monit. Assess. 184
(2), 1181–1196, http://dx.doi.org/10.1007/s10661-011-2031-9.
Singh, V.K., Singh, K.P., Mohan, D., 2005. Status of heavy metals in water and bed sediments of River Gomti-a tributary of the Ganga River, India.
Environ. Monit. Assess. 105, 43–67.
Sundaray, S.K., Nayak, B.B., Kanungo, T.K., Bhatta, D., 2012. Dynamics and quantification of dissolved heavy metals in the Mahanadi River
estuarine system, India. Environ. Monit. Assess. 184 (2), 1157–1179, http://dx.doi.org/10.1007/s10661-011-2030-x.
Sundaray, S.K., 2009. Application of multivariate statistical techniques in hydro-geochemical studies-a case study: Brahmani–Koel River (India).
Environ. Monit. Assess. 164 (1–4), 297–310.
Sundaray, S.K., Panda, U.C., Nayak, B.B., Bhatta, D., 2006. Multivariate statistical techniques for the evaluation of spatial and temporal variation
in water quality of Mahanadi river-estuarine system (India). A case study. Environ. Geochem. Health 28 (4), 317–330.
Yalcin, M.G., Tumuklu, A., Sonmez, M., Erdag, D.S., 2010. Application of multivariate statistical approach to identify ’heavy metal sources in
bottom soil of the Seyhan River (Adana), Turkey. Environ. Monit. Assess. 164 (1–4), 311–322.