Environ Monit Assess (2021) 193: 117

https://doi.org/10.1007/s10661-021-08873-x

Spatio-temporal trends in the flow and water quality:

response of river Yamuna to urbanization
Shweta Lokhande & Vinod Tare

Received: 17 April 2020 / Accepted: 11 January 2021 / Published online: 10 February 2021
# The Author(s), under exclusive licence to Springer Nature Switzerland AG part of Springer Nature 2021

Abstract Urban rivers are major victims of pollution significant correlation between DO and biochemical
and the river Yamuna is a notable example. Effective oxygen demand (BOD) was not observed throughout
human intervention requires an understanding of the the river stretch. High values of DO reported by CPCB
water quality response of rivers to urbanization. To this indicate the presence of photosynthetic activity in the
end, the time-series data from the Central Water Com- river; hence diurnal DO measurement is suggested for
mission (CWC) and the Central Pollution Control Board validation. For the rejuvenation of river Yamuna, the
(CPCB) was analyzed. The dataset included 44 param- focus of treatment should be shifted to coliform, DO,
eters for twelve stations on Yamuna mainstream from chemical oxygen demand, and nutrients. The present
1978 to 2015. Statistical tests for analysis of trends study analyzed the response of the riverine ecosystem
revealed decreasing monsoon flows over the past de- to altered flow regimes and changes in river water
cades. Furthermore, increasing non-monsoon flows quality, and the findings can serve as a basis for
from Delhi to Agra was indicative of a significant con- decision-makers engaged in river restoration and con-
tribution from wastewater discharge to the river. servation efforts.
Groundwater parameters such as electrical conductivity,
hardness, and sodium content were found to increase in
Keywords Temporal trends . Spatial variation .
the river over the years. This suggests the use of ground-
Seasonal variation . DO . BOD . Nutrients N and P
water that gets converted into domestic wastewater
flowing in surface drains discharging into the river
Abbreviations
resulting in the deterioration of water quality of river
AN Ammonical nitrogen
Yamuna. Dissolved oxygen (DO) and ammonical nitro-
ANOVA Analysis of variance
gen values from Delhi to Agra stretch do not support
As Arsenic
indigenous aquatic life. A positive correlation between
Avg Average
total and fecal coliform in this river stretch indicates the
BIS British Indian standard
dominance of domestic sewage. The expected
BOD Biochemical oxygen demand
Ca Calcium
Cd Cadmium
S. Lokhande
Centre for Urban Science and Engineering, Indian Institute of
CH Carbonate hardness
Technology Bombay, Powai, Mumbai 400076, India COD Chemical oxygen demand
Conc Concentration
V. Tare (*) CPCB Central Pollution Control Board
Indian Institute of Technology Kanpur, Kanpur, Uttar Pradesh
208016, India
Cr Chromium
e-mail: vinod@iitk.ac.in Cu Copper
117 Page 2 of 14 Environ Monit Assess (2021) 193: 117

CWC Central Water Commission lost their waste assimilation capacity, riparian beauty,
D/S Downstream and ecological habitats as a result of flood control mea-
DO Dissolved oxygen sures (Woo 2010). Polluted rivers are emerging as a
EC Electrical conductivity crucial challenge in developing countries since sewage
EYC Eastern Yamuna Canal treatment is not given the required priority and attention.
FC Fecal coliform The global water crisis was initially viewed as a water
GAP Ganga action plan quantity problem. However, due to deteriorating condi-
GRBMP Ganga River Basin Management Plan tions of water bodies, water quality is now being con-
Hg Mercury sidered of crucial importance (VishnuRadhan et al.
IS Indian standard 2017). Although the green revolution facilitated in-
K Potassium creased agricultural production, it also resulted in the
MK Mann Kendall indiscriminate use of pesticides and fertilizers contain-
Max Maximum ing toxic metals. Some metals of geological origin are
Mg Magnesium also known to enter the river system due to the natural
Min Minimum weathering of rocks. Additionally, increasing anthropo-
Mon Monsoon genic activities such as mining, industrial activities,
Na Sodium traffic emissions, and sewage discharge are predomi-
NCH Non-carbonate hardness nantly responsible for the emerging contaminants in
Ni Nickel river bodies. Thus, freshwater has lately become a pre-
NM Non-monsoon cious commodity.
NMCG National Mission for Clean Ganga As far as India is concerned, river Yamuna gains
P-Alk Phenolphthalein alkalinity significance as a perennial river of the north Indian
Pb Lead plains, a major tributary of the national river Ganga,
pH Potential hydrogen and the lifeline of the country’s capital city, Delhi. The
PM Post-monsoon livelihood of an estimated population of 57 million is
Pre-Mon Pre-monsoon dependent on the Yamuna river (Sharma and Chaudhry
PW Post-winter 2013), as almost 70% of the source of water supply to
Q Discharge Delhi is from Yamuna (Jaiswal et al. 2019). The WHO
SAR Sodium absorption rate reports the death of about 1.8 million infants annually on
T-Alk Total alkalinity account of diseases occurring due to contaminated water
TC Total coliform (Parween et al. 2017). Large-scale cultivation activities
TDS Total dissolved solids on the fertile Yamuna plains have led to a depletion in
TH Total hardness the groundwater table (Kumar et al. 2019). The trans-
TKN Total Kjeldahl nitrogen formation of groundwater with high salt concentration
U/S Upstream into wastewater which ultimately finds its way into the
Win Winter Yamuna river has been reported as the main cause of
WYC Western Yamuna Canal increased chloride content downstream of Delhi
YAP Yamuna Action Plan (Bhargava 1985).
Zn Zinc A study of land use and land cover during 1985–
2005 reported a rapid rise in built-up and cropland area
of the Yamuna river basin, along with diminishing forest
areas. An escalation of 90% in the urban population of
Introduction Yamuna basin states has been reported in this decade as
against 36% and 38.4% during the decades 1991–2001
Water being the elixir of life, rivers have been the and 2001–2011 respectively (Kumar et al. 2019). Sig-
cradles of all historic civilizations. With an increasing nificant reduction in the flood plains of Yamuna in Delhi
population and an exponential rate of development, has been reported due to a 7% increase in the built-up
surface water bodies have been undergoing changes area between 1977 and 2009 (Joshi et al. 2011). Multi-
both in terms of quantity and quality. The rivers have ple urban centers like Delhi, Agra, Mathura,
Environ Monit Assess (2021) 193: 117 Page 3 of 14 117

Yamunanagar, Sonepat, and Panipat are situated on the (DO), biochemical oxygen demand (BOD), and chem-
banks of river Yamuna. These centers are hubs for pulp ical oxygen demand (COD), monitored by the Central
and paper mills, tanneries, steel plants, glass, chemicals, Pollution Control Board (CPCB), were analyzed along
food processing, and rubber industries, all of which with the Central Water Commission (CWC) data to
yield a lot of chemical effluents in river Yamuna. assess the variability in measurements reported by dif-
The degradation of groundwater quality caused by ferent organizations. Using CWC data, the present study
industrial pollution in the suburbs of Agra was exam- aims to understand how a surface water body, such as
ined by Yadav and Keshari (2017). The stretch of Ya- the river Yamuna, has undergone changes in terms of
muna through Delhi is critically polluted by heavy quality and quantity from 1978 to 2015. The primary
metals arising from anthropogenic pollution loads objective of this study is to analyze the temporal and
(Bhardwaj et al. 2017). A positive correlation between spatial trends and patterns in the flow and water quality
potassium and nitrates (both of which are used in fertil- parameters of river Yamuna. This will provide a scien-
izer manufacturing) suggests that agricultural practices tific basis for understanding the effect of anthropogenic
actively influence the Yamuna river basin (Parween activities as well as human interventions carried out on
et al. 2017). Humans have been responsible for releasing the river.
75% more phosphorus in the soil than what would have
existed naturally. Phosphorus is found to be the primary
limiting nutrient in freshwater bodies as evidenced by a Materials and methods
strong positive correlation between algal biomass in
lakes and total phosphorus (Kaur and Singh 2012). Considering the deteriorating conditions of river Yamu-
The potential of excess nutrient load to cause eutrophi- na and its importance for India’s economy, the spatial
cation was suggested by Sharma et al. (2017). stretch of mainstream Yamuna from its source to its
Several water quality programs are initiated with the confluence with river Ganga is the target area for the
objectives of identifying the sources for raw water sup- current analysis.
ply and framing policies to restore the ecological bal-
ance of the river’s polluted stretches. Numerous studies Study area
are being conducted by researchers from time to time at
specific sites to understand the influence of particular River Yamuna originates at Yamunotri glacier in the
cities on the river water quality. However, such studies Mussorie range of the lower Himalayas. With a catch-
only represent the recent water quality, thereby ment area as large as 345,847 km2, the Yamuna basin
restricting our inferences to knowledge based on limited constitutes approximately 32% of the Ganga river basin
data. To evaluate the decadal changes in river water in Indian territory. The river traverses a length of
quality coupled with urbanization, long-term data needs 1376 km from an elevation of about 6387 m from mean
to be analyzed. Substantial literature is available on sea level and joins river Ganga at Allahabad at an
water quality monitoring of river Yamuna. Almost all elevation of 100 m above mean sea level. Out of the
of the studies carried out have been clustered in the total catchment area, nearly 70.9% is contributed by
national capital region of Delhi. Very few studies are sub-basin tributaries, whereas the mainstream accounts
known to have covered the entire spatial stretch of river for only 29.1% of the drainage. Alluvial soil constitutes
Yamuna from source to confluence, but the monitoring 42% of the Yamuna basin and agriculture is practiced on
period is only a year or two. Due to the short analysis a large scale in the river basin’s fertile plains. The north
period, no study has addressed the statistical trends in Indian states namely Uttar Pradesh, Rajasthan, and
water quality parameters. Seasonal variation is mostly Madhya Pradesh contributing to almost 20, 30, and
studied based on one or two samples per season, which 40% of the catchment area of the Yamuna basin are
may not be a true representation of seasonal water reported to have almost 68, 44, and 50% actual cultivat-
quality. Moreover, very few studies have analyzed ed land respectively (CPCB 2006). River Yamuna is
flows in the river Yamuna. Therefore, in this study, we one of the dominant streams in the northern plains, with
analyzed historical trends of water quality parameters as the majority of its water being used for irrigation. About
well as the river flows in Yamuna and assessed the 49% of the irrigated land in the Yamuna basin meets its
pollution loads. The data for coliform, dissolved oxygen water demand from surface water (Mehra 2012). The
117 Page 4 of 14 Environ Monit Assess (2021) 193: 117

CWC and CPCB stations along the mainstream of river Yamuna river stations for a 7-year period from 1999 to
Yamuna are shown in Fig. 1. For analysis, the river was 2005. During this analysis, 19 locations on the main-
divided into three stretches of four stations each, as stream of the river were analyzed. Besides this, 16
depicted in Table 1. Najafgarh and Shahdara are the drains from Delhi which contributed to the pollution of
two drains severely contributing to pollution of the river the river were also monitored. Out of these, 12 locations
in Delhi and the stretch of 22 km between Okhla and as seen in Fig. 1 have been considered for water quality
Wazirabad barrages constitutes over 50% of the pollu- analysis of parameters indicating organic pollution.
tion load of Yamuna (D. Sharma & Kansal 2011).
Almost all freshwater before Delhi is abstracted at bar- Data processing
rages directly or through canals to meet the ever-
increasing agricultural as well as domestic and industrial The CPCB data has been analyzed the reports of which
demands. are available online. However, the flow and water qual-
ity data monitored by CWC have not yet been analyzed.
Organizational details The availability of data and the methodology adopted
for analysis are detailed out in Fig. 2. The flow data is
In India, the Central Pollution Control Board (CPCB) classified data; hence, normalized values have been
and the Central Water Commission (CWC) are two presented throughout the analysis. For studying season-
leading organizations monitoring water quality for de- al variation, initially, five seasons namely monsoon
cades. These government agencies regularly collect a (July and August), post-monsoon (October and Novem-
huge amount of data on multiple water quality parame- ber), winter (December and January), post-winter (Feb-
ters through continuous monitoring programs at differ- ruary and March), and pre-monsoon (April and May)
ent locations. The Central Water Commission monitors were considered. Analysis of variance (ANOVA) for
flows in river Ganga at about 283 locations (hydrolog- flows showed an insignificant difference between five
ical observation sites) under an organizational structure seasons at all stations. Thereafter, only two seasons
of 11 CWC divisions. Out of the total 283 stations namely monsoon and non-monsoon were considered
reported by CWC, water quality is observed at around for analysis. Based on visual observation of monthly
106 stations. The state of river water quality differs at flows, July to September were included in monsoon,
various stages of the river and is mainly affected by whereas November to June were included in non-mon-
population density and land use pattern in its catchment soon. The month of October having irregular flow
area. Some stations on the upstream side of the river values was omitted from the analysis.
show less variation in water quality measures so it is not
economical to monitor the quality at intervals similar to The methodology for temporal and spatial trend analysis
those stations located relatively on the downstream side.
Based on the frequency of monitoring, CWC water The analysis of variance gives a statistical base to con-
quality stations are classified as shown in Table 2. clude about significant changes as against mere visual
The water quality samples are collected on the first representation and is used to study seasonal variation
date of every month and are sent to the nearest testing between parameters. The analysis was carried out for a
lab for analysis. There are three types of labs which are 95% confidence level with a significance value of 0.05.
classified based on water quality parameters analyzed: Furthermore, the non-parametric (univariate) Mann
tier I (measures 6 physical parameters), tier II (25 phys- Kendall trend test, which accounts for missing values
icochemical + bacteriological), and tier III labs (41 in the dataset, was used for trend analysis (Rozemeijer
parameters including pesticides and heavy metals). et al. 2014). The significance value of 0.05 was consid-
Along with policy-making, an important objective of ered, and the M.K. statistic helped to infer whether there
water quality monitoring is to assess the impact of is a monotonic increase, a monotonic decrease, or no
human intervention and the success of the river action change in the parameters over the entire time series
plans. CPCB was entrusted with the responsibility of (Hirsch & Slack 1984). The Theil-Sen slope estimator
water quality monitoring of river Yamuna during 1977. test is a further addition to the Mann Kendall trend test
After the implementation of the Yamuna Action Plan - which gives the rate at which a particular parameter has
phase 1, CPCB has carried out water quality analysis on changed over the years. The Pearson correlation is
Environ Monit Assess (2021) 193: 117 Page 5 of 14 117

Fig. 1 CWC and CPCB stations

on river Yamuna

CWC monitoring stations CPCB monitoring stations

another tool commonly used by researchers to establish Patterns in flow

the relationships between water quality parameters. For
the Pearson correlation, the relation was considered to The extreme hydrological and meteorological events
be statistically significant for p value < 0.05. such as droughts and floods are also known to affect
the quality of water along with quantitative changes,
thereby impacting aquatic life directly or indirectly
(Hrdinka et al. 2012). Spatial variation of flow (median
Results and discussion flows for four decades) during monsoon and non-
monsoon seasons is represented in Fig. 3. Average
The results are discussed in seven categories namely (i) flows reduced in the upper Yamuna stretch as the river
patterns in flows, (ii) seasonal variation of water quality travels from Paonata to Palla. No significant tributary
parameters, (iii) suitability of river water for various joins the upper Yamuna stretch and water is abstracted
beneficiary uses, (iv) spatio-temporal trends in water at the barrages directly or through canals. The flows in
quality parameters, (v) inter-relationship between water middle Yamuna stations were comparable throughout
quality parameters, (vi) assessment of anthropogenic the year as seen in Fig. 3. Analysis of the proportionate
influence on river Yamuna by evaluating heavy metal flows of the tributary and mainstream stations showed
concentrations, and (vii) understanding the effect of that Chambal constituted almost up to 70–80% whereas
human intervention on water quality of river Yamuna. Betwa constituted up to 20–30% of the flow in river
Yamuna after confluence. The seasonal flow in Cham-
bal during monsoon was almost 4.4 times that of the
mainstream Yamuna.
Table 1 Classification of monitoring stations on river Yamuna
Based on CWC data, approximately 74% of the
Sr. no. River stretch Stations discharge in river Yamuna was observed to occur during
July, August, and September. Results of the trend test
1 Upper Yamuna Paonata, Kalanaur, Mawi, Palla
carried out for monsoon and non-monsoon seasons are
2 Middle Yamuna Delhi, Mohana, Mathura, Agra
detailed in Table 3. Reduced average flows were ob-
3 Lower Yamuna Etawah, Auraiya, Hamirpur, Pratappur
served at Paonata and the lower Yamuna stations after
117 Page 6 of 14 Environ Monit Assess (2021) 193: 117

Table 2 Categorization of CWC stations on the basis of frequency of measurement

S no Type of Frequency of monitoring Description

station

1 Base One sample in 2 months Location where the water quality is relatively free from the influence of human activities
2 Trend Monthly The location is designed to understand how a particular point on a water course
varies over time normally, under anthropogenic influence
3 Flux Thrice monthly The extent of pollution due to a geological feature or human activities is considered,
Toxic and trace metal: monthly which helps in analyzing the impact of pollution control measures adopted

the confluence with river Chambal. Constant flows from additions during the lean flow period. Field investiga-
Kalanaur to Delhi might be due to the structural inter- tions reveal the tendency of industries to dispose of
vention in the form of Hathnikund barrage constructed untreated effluents in high flows during summer,
in 1998, which might have led to a controlled flow resulting in the pollution of rivers (Sharma et al. 2015).
throughout the years. Increasing non-monsoon flows The results of the correlation analysis between pre-
from Mohana to Auraiya are indicative of wastewater cipitation and flow are shown in Table 4. A strong

Fig. 2 Methodology adopted for

flow and water quality analysis
Environ Monit Assess (2021) 193: 117 Page 7 of 14 117

Fig. 3 Spatio-temporal variation

of flows over the four decades 6000 1978-1985

Normalized flow (cumec)

Monsoon 1986-1995
4500
1996-2005
3000 2006-2015

1500
0
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300
Distance of station from source (in km)

Pratappur
Hamirpur
Kalanaur

Mathura
Mohana

Auraiya
Paonata

Etawah
Mawi

Delhi
Palla

Agra
400 1978-1985
Non-Monsoon

Normalized flow (cumec)

1986-1995
300
1996-2005
200 2006-2015

100

0
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300
Distance of station from source (in km)

Pratappur
Hamirpur
Kalanaur

Mathura
Mohana

Auraiya
Paonata

Etawah
Mawi

Delhi
Palla

Agra
positive correlation between the two was observed only The amplified impervious surface areas resulting
at Kalanaur and Mawi during the monsoon season. from urbanization result in early occurrence of and a
From Delhi onwards, a moderate positive correlation sharp increase in peak flows and augmented runoff.
was observed between the two variables. During the However, the magnitude of monsoon flow peaks in river
non-monsoon season, flow and precipitation were mod- Yamuna was seen to reduce over the years. The average
erately correlated in the upper Yamuna stretch whereas rates of decreasing monsoon flows were found to be 4,
weakly correlated after Delhi. A lack of correlation 12, and 38 m3/s/year for the upper, middle, and lower
between precipitation and river flow during the non- Yamuna stretch, respectively. The non-monsoon flows
monsoon season is indicative of anthropogenic influ- approximately increased by 0.5 to 1 m3/s/year for the
ence. Furthermore, the trend of precipitation in non- middle stretch. Thus, the rate of decrease in monsoon
monsoon seasons was analyzed using one-way flows is substantial in comparison to the rate of increase
ANOVA. The results show insignificant variation in in non-monsoon flows, indicating that decrease in fresh-
the precipitation over the years for all nine stations (for water during monsoon over the years is a critical issue
which precipitation data is available). Thus, precipita- for surface water bodies. In contrast to the current find-
tion in the non-monsoon season has remained constant ings, Kumar et al. (2019) reported an increase in the
over the years but river flows have increased indicating flow of river Yamuna during monsoon season, with an
wastewater contribution to the flows. increased frequency of 5 lakh m3/s and above discharges

Table 3 Trend analysis of flow (stations 1 and 12 denote the first and last station respectively)

Season Staon 1 2 3 4 5 6 7 8 9 10 11 12
Min - --
Monsoon Max ---
Avg - - - - ---
Min --- - +++ +++ ++++++ ++ ---
Non
Max --- ++ - +++ +++ ++++++ +++ ---
Monsoon
Avg --- +++ +++ +++ ---
Increasing Trend No Trend Decreasing Trend
+++ ++ + - -- ---
0<p<0.001 0.001<p<0.01 0.01<p<0.05 p>0.05 0.01<p<0.05 0.001<p<0.01 0<p<0.001
117 Page 8 of 14 Environ Monit Assess (2021) 193: 117

Table 4 Correlation between precipitation and flow

Station Kalanaur Mawi Delhi Mathura Agra Etawah Auraiya Hamirpur Pratappur

Monsoon +++ +++ ++ ++ + ++ ++ ++ ++

Non-mon ++ ++ ++ + + + + + +

+++ 0 < p < 0.001; ++ 0.001 < p < 0.01; + 0.01 < p < 0.05

in the Yamuna river basin in the last 37 years in com- and water quality monitoring agencies. IS 2296: 1982
parison with the data from the last 102 years. These classifies the tolerance limits for water quality parame-
contradictory results might be because of the difference ters in five designated uses namely drinking, bathing,
in the type of flow data analyzed. While average flows drinking after conventional treatment, fish culture, and
were used in the current analysis, peak flows in mon- irrigation (BIS 1982). The section below discusses the
soon were reported by Kumar et al. (2019). suitability of the river with the permissible limits as
specified in IS 2296: 1982. This comparison is based
Seasonal variation of water quality parameters on the median values of parameters over four decades.

Among the water quality parameters, iron, nitrate, ni- Suitability for drinking purpose
trite, boron, DO, COD, and coliform showed insignifi-
cant variation at all the stations. The chemical parame- The pH values of water at Mathura have exceeded 9
ters such as bicarbonate, sulfate, chloride, hardness, Ca, from 1992 to 1995. Alkalinity was in the range of 75–
Mg, and Na do not undergo seasonal variation at 90, 130–180, and 110–130 mg/L in the upper, middle,
Paonata and Kalanaur. This indicates that the chemical and lower stretches, respectively. During 1996–2005,
composition of the river near its source remains nearly the river at Mathura was observed to be unsuitable for
the same throughout the year and this stretch of river is drinking, while during 2006–2015, hardness at addition-
relatively free from human influence. Various forms of al three stations Mohana, Agra, and Etawah exceeded
nitrogen like nitrate, nitrite, and indicators of organic the permissible limit of 300 mg/L for the same. The
pollution such as COD and coliform do not undergo chloride concentration at Agra and Etawah which was
significant seasonal variation suggesting regular dis- above 200 mg/L during 1996–2005 is however below
charge of domestic and agricultural wastes in the rivers. the drinking water limit of 250 mg/L throughout the
Insignificant seasonal variation for coliform indicates years.
that the dilution during the monsoon season has no
effect on coliform and that the count is significantly Suitability for fisheries
high even during monsoon. The sudden increase in
BOD and COD load during the months of July– In the initial years during monsoon, Agra was reported
October can be attributed to the additional pollution load to have low DO values. The survival of fish is very
from the catchment area in the form of vegetation and sensitive to ammonia in rivers. During 1986–1995,
defecation wastes carried in the river by the monsoon ammonical nitrogen exceeded the limit of 1.2 mg/L in
showers. Similarly, an increased count of total and the monsoon season at Agra, making it unsuitable for
thermotolerant coliform is reported in the post- supporting indigenous aquatic life. From 2006 onwards,
monsoon season which has been attributed to the mixing the stretch from Delhi to Mathura is neither suitable for
of surface runoff with raw sewage thereby augmenting fisheries in monsoon nor during non-monsoon as up to
the bacterial load (Kapilesh & Indrani 2018). 70% of data points exceeded the permissible limit of
ammonical nitrogen. The EC during the non-monsoon
Suitability of river water for various beneficial uses season breached the permissible limit. Moreover, even
during monsoon the EC values from Delhi to Etawah
Assessing the suitability of the river for beneficial pur- were seen to increase in more recent years and are
poses like drinking, fisheries, irrigation, and bathing is approaching the permissible limit of 1000 μs/cm for
one of the prime objectives of pollution control boards fisheries.
Environ Monit Assess (2021) 193: 117 Page 9 of 14 117

Fig. 4 Temporal variation of

Coliform varia
total and fecal coliform (plotted
8

Log coliform ( x10^6 MPN/100 ml)

on log scale) for sample station y = 0.0142x + 4.8206 Total coliform
7 R² = 0.1478
Etawah for 5 years from 1999 to
6
2003 Fecal coliform
5
4
3 Linear (Total
coliform)
2
y = 0.03x + 3.0083 Linear (Fecal
1
R² = 0.3632 coliform)
0
0 12 24 36 48 60
Serial number of month (for 5 years)

Suitability for irrigation groundwater parameters like hardness and electrical

conductivity were found to increase in surface waters.
Delhi has the practice of using treated wastewater for It likely points to groundwater abstraction and its dis-
irrigating field crops. One example is that of the effluent charge in river Yamuna after usage. The Agra canal uses
of sewage treatment plants near Keshopur and Okhla surface water for water supply and irrigation. The use of
being used for irrigating vegetables cultivated over water with higher EC and hardness concentrations is of
1700 ha of land area in Delhi (Parween et al. 2017). serious concern since it results in the loss of fertility of
Thus, it is of utmost importance to ensure the suitability soil thereby turning vast stretches of agricultural land
of river water for irrigation. The presence of sodium, in barren. Soil degradation issues due to alkalinity and
combination with chloride and bicarbonate, forms saline salinity have already affected 60% of the crop area in
and alkaline soil, respectively, which is harmful to the Haryana, one of the primary wheat-producing states in
plants and causes a decline in the yield. Sodium content the Yamuna river basin (Kumar et al. 2019).
(Na%) is a chemical index, dependent on the relative
cation concentration of sodium, and is important from Suitability for bathing
the point of view of irrigation. Though Na% lies within
the limit of 60% for irrigation, it was observed that Na% The coliform variation was plotted on a log scale. The
lies between 50 and 60% from Delhi to Etawah, and positive slope of the regression line for sample station
hence is critical. Also, hardness was found to increase Etawah in Fig. 4 (based on CPCB data) indicates the
over the years during monsoon. The values of monotonic increase in coliform count over 7 years. The

Table 5 Correlation between DO-BOD and TC-FC

DO vs TC vs DO vs TC vs
Sr Station Sr Station
BOD FC BOD FC
1 Hathnikund 10 Mathura US
2 Kalanaur 11 Mathura DS mid
3 Sonepat 12 Mathura DS Quarter
4 Palla 13 Agra US
5 Niz mid 14 Agra DS mid
6 Niz Quarter 15 Agra DS Quarter
7 Mazawali 16 Bateswar
8 Agra canal mid 17 Etawah
9 Agra canal Quarter 18 Juhika
+ + + + - - - -
0.7 < r <
p > 0.05 0 < r < 0.5 0.5 < r < 0.7 0.7 < r < 1 p>0.05 p > 0.05 0 < r < 0.5 0.5 < r < 0.7
1
117 Page 10 of 14 Environ Monit Assess (2021) 193: 117

slope of fecal coliform was higher than that of total Inter-relationship between water quality parameters
coliform, indicating an increase in pollutant loads
due to domestic sewage. If compared to the stan- It is pertinent to study the inter-relationship between
dards, total coliform is found to exceed the maxi- various water quality parameters measured so as to
mum permissible limit of 500 MPN per 100 mL at validate the long-term data monitored. Also, the corre-
all the stations, which indicates that river Yamuna is lation between a specific set of parameters is helpful in
not fit for bathing as far as coliform standards are analyzing the possible contamination sources.
considered. The monotonic rise in coliform count
over the years despite human interventions suggests Physicochemical parameters
the incapability of river action plans to restore mi-
crobiological water quality. Phenolphthalein alkalinity and carbonate showed a pos-
itive correlation throughout the river. The Yamuna river
at Kalanaur receives water mainly through groundwater,
Spatio-temporal trends in water quality parameters and calcium being a significant groundwater component
explains the positive correlation between calcium and
The extent to which anthropogenic influences and nat- total hardness at this station. A positive correlation of
ural processes affect river water quality individually is calcium with both calcium hardness and total hardness
unclear. Understanding the spatial and temporal varia- indicates the dominance of calcium in constituting hard-
tions in water quality contributes to reducing this uncer- ness. Carbonate hardness is the main constituent of total
tainty, resulting in rational decision-making (Wang et al. hardness up to Mohana. The moderate correlation be-
2013). Various anions like bicarbonate, chloride, phos- tween BOD and COD at Mawi, Delhi, and Mohana
phate, and silicate are observed to decrease in the entire possibly indicates the dominance of industrial pollution
upper Yamuna stretch. Nitrate concentration at in these areas in comparison to organic pollution. Soaps
Kalanaur was seen to have an increasing trend through- and detergents being a prime source of phosphate, the
out the year. The non-monsoon load at Palla followed an positive correlation between COD and phosphate at
increasing trend for all parameters, which might be Delhi is indicative of the urban areas responsible for
attributed to increased anthropogenic inputs in the river. emitting phosphate in the river. Out of all anions, non-
Boron concentration at Kalanaur was observed to in- carbonate hardness is found to be associated with sul-
crease throughout the year, which might be due to fate. A positive correlation of chloride with electrical
mining activities carried out in the nearby areas. In the conductivity in the polluted stretch of middle Yamuna is
middle Yamuna stretch, non-monsoon flows were possibly due to domestic wastes finding their way in the
found to increase. For the majority of the parameters, river. Sodium and chloride show a positive correlation
trends in concentration and loads were observed to be with each other at Mawi, Agra, and Auraiya. The pos-
the same, suggesting that concentration was affecting itive correlation of chloride with EC and sodium sug-
the pollution load. Delhi and Agra are the stations where gests the discharge of effluents from tanneries, which
almost all parameters showed an increase in concentra- are known to use substantial amounts of salts for the
tion and pollution load during the non-monsoon season. preservation of animal hides.
Wastewater disposal in the river during the non-
monsoon period may be a cause for the increase in the Indicators of organic pollution
flow. Among the lower Yamuna stations, Etawah and
Auraiya were worst affected as far as the pollution load The correlation between DO-BOD and TC-FC based on
of all chemical parameters is considered. The nutrient CPCB data is presented in Table 5. No uniformity in the
loads were observed to reduce only after the Yamuna correlation between DO and BOD is observed through-
traverses a significant distance after its confluence with out the river. However, the lack of correlation between
Chambal. Similar results in spatial variation have been DO-BOD in 7-year CPCB data analyzed is not particu-
reported in the literature. The quality of water was larly surprising given the fact that DO and BOD are two
reported as good at Hathnikund in the upstream; de- independent parameters and that BOD is not the sole
clined at Delhi, Mazawalli, and Agra downstream; and parameter responsible for depleting the DO levels. Poor
increased at Juhika (Jaiswal et al. 2019). DO levels might be attributed to other water quality
Environ Monit Assess (2021) 193: 117 Page 11 of 14 117

Table 6 Pearson correlation between heavy metals

Heavy metal pairs Paonata Kalanaur Mawi Palla Delhi Mohana Mathura

Moderate 0.5–0.7 + Nil Cd-Cu Cd, Cu-Cr; As-Pb Cd-Cr, Zn As-Hg Cr-Hg Nil
Strong 0.7–1 + Cr, Cu-Hg Pb-Hg-Zn Nil Nil Nil Cd-Hg Nil
Moderate 0.5–0.7 – Nil Nil Nil Nil Cd, Zn-Hg, Cr; Nil As-Zn

parameters such as nitrates (Jaiswal et al. 2019). The Bacteriological parameters (total and fecal coliform)
interaction between aerobic and nitrification processes
may be responsible for the negative correlation between Except for Hathnikund, all the areas upstream of Delhi
DO and nitrates. As per CPCB data, a significant but and downstream of Agra showed an insignificant corre-
weak inverse correlation exists between DO and BOD at lation between total coliform (TC) and fecal coliform
Delhi, Mohana, and Mathura. An insignificant correla- (FC), indicating the absence of FC in TC. So TC is
tion between DO and BOD is observed at all other probably constituted by other natural sources (such as
stations. animal wastes) in these stretches. A strong correlation
With an average BOD level of 93 mg/L, the DO between TC and FC in Delhi to Agra stretch indicates
levels were observed to be less than 4 mg/L, with DO that it is FC which mainly constitutes TC. The highest
even dropping to zero at certain locations in the Delhi correlation was observed at Delhi (0.92 < r < 0.94)
NCR river stretch (Parmar and Singh 2015). The high which gradually reduced from Delhi to Agra. Results
DO values ranging almost up to 20 mg/L as reported similar to those of CPCB were observed in the case of
by CPCB indicate the prevalence of supersaturated CWC data. TC and FC showed a strong correlation at all
conditions in the river. It might be due to the photo- stations. At the upper Yamuna stations namely Kalanaur
synthetic activity occurring in the river, which is and Palla, CPCB data did not show a significant corre-
responsible for the release of oxygen during light lation between TC and FC whereas CWC data showed a
hours of daytime when DO is being measured. Had strong positive correlation. This variation in results
diurnal sampling been conducted for measuring DO, might be attributed to the different timelines for both
very low values might have been obtained at night. data sets. It might be possible that after 2005, the sewage
The occurrence of photosynthetic activity has been discharged into the rivers has increased as there is a high
reported based on the highest and lowest DO values proportion of FC in TC.
at 4 pm and 4 am respectively, as well as marginal
diurnal variation in alkalinity (Tare et al. 2003). One- Assessment of anthropogenic influence on river
time DO sampling thus fails to capture the actual Yamuna by evaluating the heavy metal concentrations
state of the river. Also, the declined DO levels at
Delhi, Mohana, and Agra have been attributed to A relatively high concentration of heavy metals was
the decomposition of organic matter by microbes observed at Kalanaur, indicating the rise of an industrial
and photosynthetic activity of aquatic life (Jaiswal belt in the upper Yamuna catchment area. Mohana was
et al. 2019). If the spatial variation is considered, reported to have the highest concentrations of heavy
both CWC and CPCB datasets report Delhi to Agra metals. Despite the increasing concentrations of heavy
as a highly polluted stretch. However, CWC has not metals, they were found to be within the permissible
reported DO values as high as 20 mg/L at Delhi, limits of drinking water. These findings are in contra-
Agra, and Mathura during the entire period of study. diction to results reported by Parween et al. (2017), who
Despite the difference in sampling locations and stated that the concentration of heavy metals is higher
analysis precision being the reasons for this variabil- than the acceptable limits for drinking water as per BIS
ity, there is a significant difference in DO ranges guidelines. CPCB reported the absence of heavy metals
reported by both organizations. DO is an important in the Yamuna in 1978, and it is alarming that within
on-site measurement parameter and can instantly in- 35 years, heavy metals were observed to be in detectable
dicate the overall health of the ecosystem. Hence, it ranges and have even increased with time. Insignificant
must be monitored accurately and with caution. seasonal variation was observed for heavy metals. No
117 Page 12 of 14 Environ Monit Assess (2021) 193: 117

critical season was observed in particular for either of exceeding the maximum permissible limit of 10^3/
the heavy metals except mercury. The concentrations 100 mL (Walia et al. 2011).
not being affected by seasonal flows are a clear sign of Furthermore, the presence of nitrates in limited quan-
industrial pollution. Based on Pearson correlation, the tity in surface water bodies is desirable as it indicates
results of which are shown in Table 6, no pair of heavy autotrophic conditions. However, increasing ammonical
metal showed uniform correlation throughout a particu- nitrogen tends toward the heterotrophic condition of the
lar stretch, which indicates a variation in the sources of river. In spite of the implementation of YAP, nitrate
heavy metals at all stations. Since copper, chromium, concentration is seen to be increasing whereas DO
cadmium, and zinc are associated with electroplating levels continue to drop low. Increasing nitrate is only
industries, a significant correlation between these metals favorable if the river has a significant amount of flow
from Kalanaur to Delhi indicates the presence of and is free from algal blooms and water hyacinth. How-
electroplating industries in these regions. A strong cor- ever, the stretch of the river from Delhi to Agra is devoid
relation between Cu-Cr is attributed to electroplating of freshwater in the lean season. In the circumstances of
and steel, pulp, and paper mill effluents carried by a stagnant river, even a slight increase in nitrate and
Najafgarh drain (Bhardwaj et al. 2017). The significant phosphate concentrations might be responsible for caus-
correlation between cadmium and chromium at Mawi ing eutrophication.
and Delhi can be due to the manufacture of paints.

Conclusion
Understanding the effect of human intervention
on water quality of river Yamuna Understanding the temporal trends in river water quality
and seasonal flows is of utmost importance for the
The Yamuna Action Plan (YAP) was initiated by the effective management of any riverine ecosystem. The
Government of India in 1993. The YAP was carried out present study is a base for understanding the long-term
in phases with 2003 and 2012 signifying the completion changes in the river, thereby predicting the effects of
of phase I and phase II respectively. YAP mainly anthropogenic wastes in the future.
targeted at trapping and cleaning the open drains, laying The prominent observation after analyzing the
sewers, and setting up decentralized sewage treatment temporal-seasonal variation is that of sewage being
plants to clean the sub-drains. BOD removal has been discharged in the rivers during non-monsoon season.
reported to be the basis of all treatments initiated in YAP The decreasing monsoon flows in the river signify a
(D. Sharma & Kansal 2011). Despite this, it violates the reduction in freshwater availability and hence is of
permissible limits for surface water. BOD is a parameter severe concern. It is of utmost importance to adopt reuse
of great significance in the context of choosing a partic- and recycling practices so as to reduce the dependency
ular source of water for drinking water supply. Howev- on freshwater for non-potable uses such as irrigation.
er, judging the condition of rivers solely based on BOD The growing concentrations of nutrients and heavy
is not appropriate. The micro-organisms constituting metal in the river Yamuna, particularly in the middle
BOD, in fact, support the food chain in the river and Yamuna stretch, are indicative of inputs from agricul-
contribute to the waste assimilation capacity of the river. tural runoff and industrial effluents to the river. The
In our study, we found the DO levels from Delhi to concentration of heavy metals in effluents from indus-
Mathura too poor to sustain ecological life. These find- tries adjacent to the river stretches should be monitored
ings are in support of the work by Parmar and Singh stringently. The middle Yamuna stretch from Delhi to
(2015), who do not find the river to be in good condition Agra does not support indigenous aquatic life. The
after 20 years of implementation of YAP. monotonic rise in coliform count over the years,
Moreover, the choice of treatment technology breaching the permissible limit for bathing in the middle
adopted plays a key role. Up-flow anaerobic sludge Yamuna stretch, indicates the necessity of tertiary treat-
blanket reactor (which is ineffective for coliform reduc- ment and implementation of in-situ treatment technolo-
tion) was adopted for sewage treatment plants construct- gies to achieve coliform reduction.
ed under YAP. TC count in effluent ponds was reported The Yamuna Action Plan has not been completely
to range between 10^6 to 10^9 MPN/ 100 mL thereby successful in improving the water quality of the river,
Environ Monit Assess (2021) 193: 117 Page 13 of 14 117

possibly because of the selection of inappropriate treat- Conflict of interest The authors declare that they have no con-
flict of interest.
ment technologies. Coliform, COD, DO, and nutrients
should be the parameters targeted for action plans to
restore life in the middle Yamuna stretch. The river
action plans should have a holistic approach of restoring
the ecological niches in the river, rather than the con-
References
struction of local STPs. Also, the operation and mainte-
nance of existing sewage treatment infrastructure should
be assigned to competent authorities. Bhardwaj, R., Gupta, A., & Garg, J. K. (2017). Evaluation of
heavy metal contamination using environmetrics and
Policies restricting the use of fertilizers and pesticides indexing approach for river Yamuna, Delhi stretch, India.
should be framed and effective enforcement of the same Water Science, 31(1), 52–66. https://doi.org/10.1016/j.
should be practiced. Standards for phosphates need to be wsj.2017.02.002.
incorporated to enable nutrient removal in the stretch of Bhargava, D. S. (1985). Water quality variations and control
low flows of rivers. To prevent the occurrence of eutro- technology of Yamuna River. Environmental Pollution.
Series A, Ecological and Biological, 37(4), 355–376.
phication in the river, it is important to limit the incom- https://doi.org/10.1016/0143-1471(85)90124-2.
ing nutrients. Economics of nutrient removal in waste- BIS. (1982). Tolerance and Classification. Bureau of Indian
water vs. costs of removal of algal blooms should be Standards, New Delhi, 1–9. Retrieved from http://wqaa.
studied to provide an economic basis to initiate tertiary gov.in/WriteReadData/UserFiles/Documents/Tolerance and
Classification.pdf.
treatment.
Hirsch, R. M., & Slack, J. R. (1984). A nonparametric trend test
Furthermore, analysis of water quality parameters for seasonal data with serial dependence. Water Resources
like turbidity and suspended solids can be done to Research, 20(6), 727–732.
understand the effects of sand mining on the suitability Hrdinka, T., Novický, O., Hanslík, E., & Rieder, M. (2012).
of river water for beneficiary uses. The water quality Possible impacts of floods and droughts on water quality.
Journal of Hydro-Environment Research, 6(2), 145–150.
analysis in the identified critical stretch from Delhi to https://doi.org/10.1016/j.jher.2012.01.008.
Agra must be based on high-frequency sampling. Also, Jaiswal, M., Hussain, J., Gupta, S. K., Nasr, M., & Nema, A. K.
diurnal measurement of DO should be practiced instead (2019). Comprehensive evaluation of water quality status for
of single daytime measurement. The statistical water entire stretch of Yamuna River, India. Environmental
Monitoring and Assessment, 4, 191–208. https://doi.
quality analysis should be extended to include the
org/10.1007/s10661-019-7312-8.
CWC stations in the Ganga river basin. Moreover, mul- Joshi, R. R., Warthe, M., Dwivedi, S., Vijay, R., & Chakrabarti, T.
tivariate statistical analysis tools such as principal com- (2011). Monitoring changes in land use land cover of
ponent analysis or cluster analysis can be used to iden- Yamuna riverbed in Delhi: a multi-temporal analysis.
tify the sources of pollution. This will help in a better International Journal of Remote Sensing, 32(24), 9547–
9558. https://doi.org/10.1080/01431161.2011.565377.
understanding of the response of the riverine ecosystem
Kapilesh, J., & Indrani, J. (2018). Assaying coliform removal in
and the interpretation of trends in the flow and water waste stabilisation ponds system through the dispersed flow
quality. regime. International Journal of Environmental Technology
and Management, 21(1–2), 64–76. https://doi.org/10.1504
/IJETM.2018.092567.
Acknowledgments The authors would like to acknowledge the Kaur S., Singh I. (2012). Accelerated phosphate and nitrate level:
support of the National Mission for Clean Ganga (NMCG) and factors to blame for eutrophication in Yamuna River, Delhi,
Center for Ganga River Basin Management and Studies, IIT India. International Journal of Plant , Animal and
Kanpur (cGanga) in acquiring data from the Central Water Com- Environmental Sciences, 183–187.
mission (CWC). The first author is grateful to Kalyani Deshpande, Kumar, M., Sharif, M., & Ahmed, S. (2019). Impact of urbaniza-
Mayuresh Bakare, Nikita Kakwani, Dr. Ravinder Dhiman, and Dr. tion on the river Yamuna basin. International Journal of
Kohinoor Darda for their inputs and support at various stages. The River Basin Management, 0(0), 1–15. https://doi.
authors would like to express their gratitude to the anonymous org/10.1080/15715124.2019.1613412, 18.
reviewers whose comments have helped in the manuscript Mehra, P. (2012). Assessment of water quality of river Yamuna at
revision. major irrigation sites in Delhi for its suitability for agriculture.
Global J. of Mod. Biol. & Tech., 2(1), 28–30.
Compliance with ethical standards The authors declare that Parmar, S., & Singh, V. (2015). Water quality parameters of river
they have no conflict of interests. No human participants have Yamuna in Delhi after 20 years of the Yamuna Action Plan.
been a part of this study. Statistical methods have been used to 6(4), 662–669. Retrieved from http://www.scienceandnature.
analyze the data. org/IJSN_Vol6(4)J2015/IJSN-VOL6(4)15-22.pdf
117 Page 14 of 14 Environ Monit Assess (2021) 193: 117

Parween, M., Ramanathan, A., & Raju, N. J. (2017). Waste water Water Research, 37(1), 67–77. https://doi.org/10.1016
management and water quality of river Yamuna in the mega- /S0043-1354(01)00385-2.
city of Delhi. International journal of Environmental Science VishnuRadhan, R., Zainudin, Z., Sreekanth, G. B., Dhiman, R.,
and Technology, 14(10), 2109–2124. https://doi.org/10.1007 Salleh, M. N., & Vethamony, P. (2017). Temporal water
/s13762-017-1280-8. quality response in an urban river: a case study in peninsular
Pollution, C., & Board, C. (2006). Water quality status of Yamuna Malaysia. Applied Water Science, 7(2), 923–933. https://doi.
River (1999–2005). org/10.1007/s13201-015-0303-1.
Rozemeijer, J. C., Klein, J., Broers, H. P., van Tol-Leenders, T. P., Walia, R., Kumar, P., & Mehrotra, I. (2011). Performance of
& van der Grift, B. (2014). Water quality status and trends in UASB based sewage treatment plant in India: polishing by
agriculture-dominated headwaters; a national monitoring net- diffusers an alternative. Water Science and Technology,
work for assessing the effectiveness of national and European 63(4), 680–688. https://doi.org/10.2166/wst.2011.333.
manure legislation in the Netherlands. Environmental Wang, Y., Wang, P., Bai, Y., Tian, Z., Li, J., Shao, X., Mustavich,
Monitoring and Assessment, 186(12), 8981–8995. L. F., & Li, B. L. (2013). Assessment of surface water quality
https://doi.org/10.1007/s10661-014-4059-0. via multivariate statistical techniques: a case study of the
Sharma, D., & Kansal, A. (2011). The status and effects of the Songhua River Harbin region, China. Journal of Hydro-
Yamuna Action Plan (YAP). Proceedings of the TERI–Yale Environment Research, 7(1), 30–40. https://doi.org/10.1016
Workshop on the Yamuna River “The Yamuna river: a /j.jher.2012.10.003.
confluence of waters, a crisis of need, (November). Woo, H. (2010). Trends in ecological river engineering in Korea.
Sharma, M., & Chaudhry, S. (2013). Impact of industrial pollution Journal of Hydro-Environment Research, 4(4), 269–278.
on human health in Yamuna Nagar, Haryana. International https://doi.org/10.1016/j.jher.2010.06.003.
Journal of Science and Research, 14(3), 2319–7064.
Yadav, H., & Keshari, A. K. (2017). Assessment of groundwater
https://doi.org/10.13140/RG.2.1.3632.8401.
pollution from industries in Agra region. International
Sharma, M., Kansal, A., Jain, S., & Sharma, P. (2015).
Journal of Renewable Energy Research, 7(4), 2068–2074.
Application of multivariate statistical techniques in determin-
ing the spatial temporal water quality variation of ganga and
Yamuna Rivers present in Uttarakhand State, India. Water Publisher’s note Springer Nature remains neutral with regard to
Quality, Exposure and Health, 7(4), 567–581. https://doi. jurisdictional claims in published maps and institutional
org/10.1007/s12403-015-0173-7. affiliations.
Sharma, S., Jha, P. K., R, . M. R., Singh, U. K., & J, T. (2017).
Water quality monitoring of Yamuna river by using GIS The research has been carried out by Ms. Shweta Lokhande at IIT
based water quality index in Delhi, India. International Kanpur as part of M. Tech thesis, under the guidance of Prof.
Journal of Current Microbiology and Applied Sciences, Vinod Tare (Faculty, Civil Department). The data for this study
6(2), 1249–1263. https://doi.org/10.20546 was acquired with the support of cGanga, IIT Kanpur, and the
/ijcmas.2017.602.141. National Mission for Clean Ganga. The data is collected by the
Tare, V., Yadav, A. V. S., & Bose, P. (2003). Analysis of photo- leading water quality monitoring organization of India namely the
synthetic activity in the most polluted stretch of river Ganga. CWC (Central Water Commission).