Content uploaded by Anurag Ohri
Author content
All content in this area was uploaded by Anurag Ohri on Sep 27, 2017
Content may be subject to copyright.
IDC’s 31st National Conference of Sustainable Development of Smart Cities, 2017
22-23rd September 2017, New Delhi
[1]
Land Use Classification and Watershed Analysis of Assi
River (Varanasi, U.P.)
Mohit Kumar Srivastava1, Arun Goel2, Anurag Ohri3
1. P.G. student, Deptt. of Civil Engineering, National Institute of Technology, Kurukshetra, Haryana
srivastava.mohit0720@gmail.com
2. Professor, Deptt. of Civil Engineering, National Institute of Technology, Kurukshetra, Haryana
drarun_goel@yahoo.co.in
3. Associate Professor, Deptt. of Civil Engineering, Indian Institute of Technology (BHU), Varanasi, U.P.
aohri.civ@iitbhu.ac.in
ABSTRACT
Watershed analysis is essential for planning development activities or improving the features of
a terrain. It gives an idea for various features like – aspect, elevation, slope, drainage, urban
distribution, etc. in the area. This study is done either by field survey or with the help of various
software tools. Varanasi has been selected as one of the cities to be developed into a smart city.
But being one of the oldest cities of the world, a proper sustain planning is really essential to
make this a reality. In the present study, river Assi (a tributary of Ganges) , geographically
located between - 25°16'59.0" N & 83°00'35.3" E, in Varanasi district of Uttar Pradesh (India),
has been considered. Continuous dumping of waste, heavy encroachments and improper
planning has reduced this river into nothing but just a drain. Being a tributary of Ganges, all of
this waste further reaches Ganges water, depleting its water quality too. Softwares like ArcGIS,
ERDAS Imagine 2016 and SWAT has been used for the study of the watershed of this Assi River.
The overall classification accuracy from the Land Use map for the 3rd order watershed has been
computed as 89.32% with Kappa Coefficient being 0.7751. A digitized map of the watershed is
prepared to compute the percentage of various features like - settlement, water bodies, cultivated
land, etc. in the area of the watershed. Through SWAT, watershed has been divided into various
sub-watersheds, which enables to identify key drains of the river. This study will thus not only
help in identifying urban pattern of the area, but will also help in identifying key aspects that are
to be answered in order to remediate Assi back to its river form through proper planning of
development activities around Assi River without affecting its ecology.
Keywords: Watershed Analysis, GIS, Land Use, sub-watersheds, SWAT
1. INTRODUCTION
Land cover refers to the physical
characteristics of earth’s surface i.e. the
distribution of vegetation, water, soil and
other physical features of the land, including
those created solely by human activities e.g.,
settlements. Land-use implies the way in
which land has been used by humans and
their habitat. The land use/cover pattern of a
region is the result of natural and socio-
economic factors and their utilization by
IDC’s 31st National Conference of Sustainable Development of Smart Cities, 2017
22-23rd September 2017, New Delhi
[2]
man in time and space. Information on land
use/cover and possibilities for their optimal
use is essential for the selection, planning
and implementation of land use schemes to
meet the increasing demands for basic
human needs and welfare. This information
also assists in monitoring the dynamics of
land use resulting out of changing demands
of increasing population.
There have been numerous studies of land
cover change analysis on different
watershed by various elite researchers. This
change analysis is essential to formulate
effective management strategies for the
watershed. With various modern techniques
like - remote sensing and Geographical
Information System (GIS), land use/cover
mapping has helped in deciding the selection
of areas appropriate for agricultural, urban
and/or industrial areas of a region.
Varanasi With the rapid urbanization in last
few decades, the population of city has
grown to a tremendous extent. Due to
inappropriate planning of sanitation and
sewerage system across the city, the
enormous dump that is generated by people
is dumped in these tributaries without any
treatment. This raw waste not only pollutes
these tributaries but also Ganges. Since, Assi
River flows through a major portion of the
city, it has been affected the most. Today,
Assi has almost lost its river form and has
been reduced to a Nala (drain).
2. STUDY AREA
The present study area is located in the
Varanasi district of Uttar Pradesh (Fig 1).
Geographical coordinates of Varanasi is
25.28°N latitudes and 82.96°E longitudes.
Varanasi is located at an elevation of 80.71
m (264.8 ft) in the centre of the Ganges
valley of North India, in the Eastern part of
the state of Uttar Pradesh, along the left
bank of the Ganges, averaging between 15
m (50 ft) and 21 m (70 ft) above the river. It
lies at the eastern extent of the state and
encompasses an approximate area of
approximately 1535 sq. km. It has a
population of around 1.202 million, as per
2011 census. Being located in the Indo-
Gangetic Plains of North India, the land is
very fertile because low level floods in the
Ganges continually replenish the soil.
Varanasi is located between the Ganges
confluences with two rivers: the Varuna
River on the north and the Assi River on the
south.
Fig 1: Study area
IDC’s 31st National Conference of Sustainable Development of Smart Cities, 2017
22-23rd September 2017, New Delhi
[3]
3. ASSI RIVER
The Assi River surfaces from
Karmadeshwar Mahadev Kund at
Ghamahapur (25°16'5.81"N &
82°57'30.01"E) and after a brief course of
around 7.7 km through the city (Fig 2),
empties itself into the River Ganges at Assi
Ghat (25°16'58.47"N & 83°0'35.11"E).
Along its course, there are numerous small
industrial and domestic drains that empty its
waste into the river.
Fig 2: Assi River as visible in Google
Satellite image of Varanasi
4. ASSI WATERSHED
The delineated watershed from the ArcGIS
covers an approximate area of 13.5 km2 for
a length of 7.7 km of the Assi River, the
outlet being at Assi Ghat in the River
Ganges. The basin (Fig 3) has been
identified as a third order basin and the
average discharge of the river is
approximately 30 MLD. The drainage of the
basin was of dendritic nature and hence
lacked structural control. The average relief
was found to be approximately 24 m, basin
perimeter as 24.7 km, the drainage density
as 1.59/km. The watershed showed a terrain
which was mostly flat in nature indicting
low runoff and therefore high infiltration
capacity.
Fig 3: Delineated watershed of the Assi
River
5. METHODOLOGY
An ASTER-DEM (Advanced Space borne
Thermal Emission and
Reflection Radiometer - Digital Elevation
Model) for the concerned area is obtained
from USGS (United States Geological
Survey) database of 30 m resolution and is
then processed in ArcGIS 10.1 to obtain the
watershed using pour-point method. The
present work aims to study the watershed of
this Assi River on the basis of its Land Use
and Land Cover (LULC). The delineated
watershed from ArcGIS for Assi River is
classified using ERDAS Imagine 2016 and
its various features like settlement, water
bodies, cultivated land, fallow lands, etc. are
identified using a suitable classification
method.. From the LULC map thus
obtained, an overall classification accuracy
percentage is computed which gives an idea
of the accuracy of the map and Kappa
coefficient is computed. Thereafter, the
watershed is divided into various smaller
sub-watersheds using SWAT tool of
ArcGIS. This enables the identification of
key sub-watersheds that contribute most to
the Assi River flow.
5.1 Types of classification
ERDAS Imagine is used as a remote sensing
system for the extraction and classification
of multispectral image data in predefined
land cover categories (Table 2), so that it
IDC’s 31st National Conference of Sustainable Development of Smart Cities, 2017
22-23rd September 2017, New Delhi
[4]
can be used in further information
distribution.
In ERDAS Imagine there are mainly two
types of image classification – supervised
and unsupervised classification.
In the supervised classification approach a
group of training pixels are selected that are
representative for the specific land cover
units. This training dataset forms the basis
for classification of the total satellite image,
by using the maximum likelihood classifier.
The Maximum Likelihood Classifier applies
the rule that the geometrical shape of a set of
pixels belonging to a class often can be
described by an ellipsoid.
In unsupervised classification approach,
isodata clustering is used, in which clusters
of pixels - based on their similarities in
spectral information - are automatically
classified into the desired number of LULC
categories. When performing an
unsupervised classification it is necessary to
find the right number of classes that are to
be found. Too many, and the image will not
differ noticeable from the original, too few
and the selection will be too coarse.. In areas
where ground control or ground truth is
completely absent, this can be a useful
approach. After an unsupervised
classification it is a challenge to “find out”
or “discover” what these classes mean in
reality and what the position of boundaries
mean in reality.
5.2 Data Accuracy Assessment
Visual comparison of two classifications is
not precise enough and can be subjective. In
modern mapping an accuracy assessment is
necessary. The classification of pixel data is
checked using Accuracy assessment tool of
ERDAS IMAGINE. This tool helps in
understanding of how well and accurate the
classification is in the map. Accuracy
assessments (Table 1) determine the quality
of the information derived from remotely
sensed data.
Table 1: Accuracy Totals
Class name
Reference
totals
Classified
totals
Number
correct
Producer’s
accuracy (%)
User’s
accuracy (%)
Kappa
pastures
3
3
2
66.67
66.67
0.6564
water bodies
3
3
2
66.67
66.67
0.6564
trees
3
5
3
100.00
60.00
0.5876
fallow land
9
9
8
88.89
88.89
0.8779
built up
73
73
69
94.52
94.52
0.7971
roads
12
10
8
66.67
80.00
0.7727
103
103
92
5.3 Terminologies used
1. Error/Confusion/Contingency matrix: It
lists the values for known cover types of the
reference data in the columns and for the
classified data in the rows. The main
diagonal of the matrix lists the correctly
classified pixels.
2. User’s Accuracy: If correct classified
pixels in a class are divided by total number
of pixels that were classified in that class,
this measure is called user’s accuracy.
IDC’s 31st National Conference of Sustainable Development of Smart Cities, 2017
22-23rd September 2017, New Delhi
[5]
User’s accuracy (%) = 100% − error of commission (%)
3. Error of Commision: It refers to the
classes which show a different land-cover on
the ground than predicted on the map.
4. Producer’s Accuracy: The producer’s
accuracy is derived by dividing the number
of correct pixels in one class divided by the
total number of pixels as derived from
reference data. The producer’s accuracy
measures how well a certain area has been
classified.
Producer’s accuracy (%) = 100% − error of omission (%)
5. Error of Ommission: It refers to the
proportion of observed features on the
ground, which are not classified in the map.
6. Kappa Coefficient: The Kappa coefficient
is a measure of overall agreement of a
matrix. The Kappa statistic incorporates the
off-diagonal elements of the error matrices
and represents agreement obtained after
removing the proportion of agreement that
could be expected to occur by chance.
Developed by Cohen[5] (1960) the Kappa
coefficient measures the proportion of
agreement after chance agreements have
been removed from considerations. Kappa
increases to one as chance agreement
decreases and becomes negative as less than
chance agreement occurs. A Kappa of zero
occurs when the agreement between
classified data and verification data equals
chance agreement. According to Gwet[7],
(2002) and Vierra and Garret[9], (2005) -
Kappa Coefficient can be calculated using
the formula:
- (1)
where,
P(A) = number of times k raters agree,
P(E) = number of times k raters are expected
to agree only by chance
According to Anderson et al.[2] (1976), the
minimum level of interpretation accuracy in
the identification of land use and LULC
categories from remote sensing data should
be at least 85%. The classification done
shows an overall accuracy of 89.32% which
is an acceptable percentage for the current
classification. Average Producer’s accuracy
is 80.57% and average User’s accuracy is
76.13%. Overall Kappa value is 0.7751.
Henceforth, a Land Use and Land Cover
map (Fig 4) is prepared for the watershed
and various features are identified (Table 2).
Table 2: Feature Percentage in Assi watershed
Feature
Area (km2)
Percentage (%)
Settlement (Buildings, parks)
5.100
37.78
Water Bodies (Lakes, Ponds, river)
0.936
6.93
Cultivated Land
0.346
2.56
Fallow Land
1.683
12.47
Others (Trees, unmarked settlements, empty lands)
5.540
41.04
IDC’s 31st National Conference of Sustainable Development of Smart Cities, 2017
22-23rd September 2017, New Delhi
[6]
5.4 SWAT (Soil and Water Analysis Tool)
watershed delineation
SWAT as an extension of ArcGIS, has been
used for division of the Assi watershed into
various sub-watersheds. The watershed
delineator of SWAT uses the entire Digital
Elevation Map (DEM) of the area to make a
flow accumulation and a flow direction map
of the area. Later an outlet is selected, which
further aids in delineation of the watershed
of the Assi River. Once the watershed is
delineated, SWAT divides the Assi
watershed into 15 smaller sub-watersheds
(Fig 5), longest flow paths of each of these
watersheds is known too. This enables to
identify the watersheds that affect the most
to the flow of Assi River.
There is a 97% overlap between
thewatersheds obtained from ArcGIS (for
LULC map) and that obtained from SWAT.
This happens due to small variation in the
outlet point taken, but since the drainage
pattern is identical in both the cases hence,
they and their results (for analysis puposes)
can be considered identical to each other.
Topographic Report (Table 3) generated
from SWAT analysis shows a watershed
map of the area with 15 smaller sub-
watersheds. These sub-watersheds each are
analyzed for their respective elevation and
longest flow path values. The sub
watersheds (SW5, SW6, SW8, SW9, SW12)
closer to the main river stream, shows longer
Fig 4: LULC map of the Assi watershed from
LISS-III image as obtained from ERDAS
Imagine
Fig 5: Various sub-watersheds of the Assi
watershed as obtained from SWAT tool
IDC’s 31st National Conference of Sustainable Development of Smart Cities, 2017
22-23rd September 2017, New Delhi
[7]
flow paths indicating that their contribution
to the main drainage is maximum and the most significant of all.
Table 3: Topographic report showing characteristics of various sub-watersheds
Minimum
Elevation (m)
Maximum
Elevation (m)
Mean Elevation
(m)
Standard
Deviation (m)
Longest
Path
(m)
SW 1
8.0
31.0
18.6150
2.9557
1973.7617
SW 2
9.0
28.0
18.3819
2.3573
2520.1513
SW 3
10.0
30.0
17.4692
2.6491
1270.3843
SW 4
8.0
31.0
16.5894
3.1646
1265.7135
SW 5
10.0
25.0
17.2457
2.3666
2181.1442
SW 6
11.0
28.0
18.2464
2.1907
2222.8172
SW 7
10.0
32.0
17.0031
2.7118
1792.9822
SW 8
8.0
31.0
18.8796
2.6363
2390.8244
SW 9
10.0
25.0
18.0145
2.0078
2284.0496
SW 10
14.0
21.0
16.9687
1.7257
694.2175
SW 11
14.0
29.0
18.5938
1.8868
1958.4348
SW 12
13.0
31.0
19.4397
2.6103
2540.7687
SW 13
14.0
25.0
17.2727
1.7447
1031.9849
SW 14
12.0
23.0
17.6325
1.6793
1653.4374
SW 15
12.0
25.0
18.2682
2.1114
1658.5464
(SW – Sub Watershed)
6. CONCLUSION
In the present work hence, an attempt is
made to study the watershed of this Assi
River. The watershed that has been
delineated for the river has an approximate
area of 13.5 sq. km. around it, outlet being at
Assi Ghat – the confluence Point of Assi
River with Ganges. The basin (or, the
watershed) is observed to be of dendritic in
nature and is found to be of 3rd order. The
terrain around the river has been found to
have low relief indicating lower runoff and
high infiltration capacity. Hence, the river
flow also affects the ground water to some
extent.
Thereafter, a supervised classification using
ERDAS Imagine 2016 is done for this
watershed, which indicated an acceptable
overall classification accuracy of 89.32%
and Kappa coefficient of 0.7751. It indicates
that the classification done is accurate and
acceptable, as according to Anderson[2] et al.
(1976), the minimum level of interpretation
accuracy in the identification of land use and
LULC categories from remote sensing data
should be at least 85%.
It can also be inferred that the percentage of
settlement in this watershed is maximum,
i.e. 37.78 %. This clearly indicates that the
increased population of the city has
maximum affect on the topology of the area.
The increased population and unplanned
settlement has led to increased waste
production and increased encroachments
around the area, this also affects the flow of
the river.
Further, with the use of SWAT’s watershed
delineator tool has been used too for
delineation of the catchment too. SWAT
analysis breaks the watershed into 15
smaller watersheds whose various physical
parameters like elevation, longest flow path
IDC’s 31st National Conference of Sustainable Development of Smart Cities, 2017
22-23rd September 2017, New Delhi
[8]
is studied. This longest flow path helped in
identification of key sub-watersheds having
the maximum longest flow paths, as these
are the sub-watersheds which affect most to
the flow of the Assi River.
The Assi River is undergoing rapid
degradation due to continuous dumping of
wastes, encroachments and unplanned
settlements around it. It needs an immediate
attention, before it solely loses its 'river'
form and becomes a 'nala' (or a drain) for
forever.
7. SCOPE OF FURTHER STUDY
This study can help in making a planned
effort to control or measure the flow of the
river even in extreme cases like flood. The
watershed gives us an idea of how the
stream is being affected by its surrounding
smaller streams. Since this Assi River,
ultimately empties its waste into the Ganges
River (at the outlet of the watershed), thus it
may also helps in monitoring the sewage
entering the Holy Ganges. By extending this
study to construction of a hydrologic model
of the watershed, this study can help in
getting values like discharge, infiltration,
runoff, etc.Ganges is the main source of
Irrigation and hydraulic power source for
Varanasi, this study can help in
understanding the amount of discharge of
waste or sediment that flows into the
Ganges.
Assi River is close to extinction, and all the
efforts made for cleaning of Ganges would
all go waste unless appropriate attention is
given to Assi River (or so, reluctantly called
'Assi Nala'). This work is an attempt to bring
out a solution to this problem so that it could
be extended to other similar programs of re-
mediating a river or a stream.
REFERENCES
[1] Alqurashi A. F., Kumar L. (2013),
“Investigating the Use of Remote
Sensing and GIS Techniques to
Detect Land Use and Land Cover
Change: A Review”, Advances in
Remote Sensing, Volume 2, Pg. 193-
204.
[2] Anderson J. R., Hardy E. E.,
Roach J. T., Witmer R. E. (1976),
“A land use and land cover
classification system for use with
remote sensor data”, Geological
Survey Professional Paper.
[3] Biswas S., Sudhakar S. & Desai
V.R. (1999), “Prioritization of Sub-
watersheds based on morphometric
analysis of drainage basin: a remote
sensing and GIS approach”, Journal
of Indian Society of Remote Sensing
(JISR), Volume 27, No. 3, Pg. 155-
166.
[4] Butt A., Shabbir R., Ahmad S. S.,
Aziz N. (2015), “Land use change
mapping and analysis using Remote
Sensing and GIS: A case study of
Simly watershed, Islamabad,
Pakistan”, The Egyptian Journal of
Remote Sensing and Space Sciences
(EJRSS) Volume 18, Pg. 251 – 259.
[5] Cohen, Jacob (1960), “A coefficient
of agreement for nominal
scales”, Educational and
Psychological Measurement,
Volume 20, No. 1, Pg. 37–46.
[6] Congalton, R.G. (1991), “A review
of assessing the accuracy of
classification of remotely sensed
data”, Remote Sensing of
Environment, Volume 37, Pg. 35-46.
[7] Gwet K. (2002), “Inter-Rater
Reliability: Dependency on Trait
Prevalence and Marginal
IDC’s 31st National Conference of Sustainable Development of Smart Cities, 2017
22-23rd September 2017, New Delhi
[9]
Homogeneity”, Statistical Methods
for Inter-Rater Reliability
Assessment, Volume 2, Pg. 1–10.
[8] Rawat J.S., Kumar M. (2015),
Monitoring land use/cover change
using remote sensing and GIS
techniques: A case study of
Hawalbagh block, district Almora,
Uttarakhand, India, The Egyptian
Journal of Remote Sensing and
Space Sciences (EJRSS) 18, Pg. 77–
84.
[9] Viera, Anthony J., Garrett, Joanne
M. (2005), “Understanding
interobserver agreement: the kappa
statistic”, Family Medicine.
Volume 37, No. 5, Pg. 360–363.
[10] Waikar M. L. & Nilawar A. P.
(2014), “Morphometric Analysis of a
Drainage Basin using Geographical
Information System : A Case Study”,
International Journal of
Multidisciplinary and Current
Research (IJMCR), Volume 2, Pg.
179 – 184.
