Spatiotemporal Analysis of Noida, Greater Noida and Surrounding Areas (India) Using Remote Sensing and GIS Approaches

Abstract

Spatiotemporal analysis refers to an analysis having both spatial extension and temporal duration. Remote sensing has provided a great tool to quantify changes using satellite data in our area of interest. Population of Indian cities is growing rapidly. That is why human need for shelter is also growing day by day, resulting in fast urbanization. This study uses remote sensing and GIS approaches to delineate the urban growth resulting in decrease of agricultural land. In this study, it is analyzed as to how much change has taken place in a time span of two decades and five years, mainly emphasizing on change and growth of residential, commercial and industrial structures on agricultural land. The use of a time-series of Landsat data to classify the urban footprints since 1986 has enabled detection of spatial and temporal urban sprawl and urban development in the explosively growing large urban agglomerations of the two metropolitan cities of Noida and Greater Noida. In this study, a spatiotemporal analysis using three Landsat scenes of 25 years of time span, aims at the detection of the changes and the urban footprints in Noida and Greater Noida and its surroundings. It is also assessed as to how much urban expansion has taken place during the last 25 years. With the help of socioeconomic data, demographic profile and increase in number of households are also assessed.

Full text

Journal of Remote Sensing & GIS
Volume 3, Issue 3, December 2012, Pages 42-51
ISSN: 2230 -7990© STM Journals 2012. All Rights
Reserved
Page
42
Spatiotemporal Analysis of Noida, Greater Noida and Surrounding Areas
(India)
Using Remote Sensing and GIS
Approaches
Gopal Krishna*
Indian Agricultural Research Institute, New Delhi, India
ABSTRACT
Spatiotemporal analysis refers to an analysis having both spatial extension and temporal duration.
Remote sensing has provided a great tool to quantify changes using satellite data in our area of interest.
Population of Indian cities is growing rapidly. That is why human need for shelter is also growing day
by day, resulting in fast urbanization. This study uses remote sensing and GIS approaches to delineate
the urban growth resulting in decrease of agricultural land. In this study, it is analyzed as to how much
change has taken place in a time span of two decades and five years, mainly emphasizing on change and
growth of residential, commercial and industrial structures on agricultural land. The use of a time -series
of Landsat data to classify the urban footprints since 1986 has enabled detection of spatial and temporal
urban sprawl and urban development in the explosively growing large urban agglomerations of the two
metropolitan cities of Noida and Greater Noida. In this study, a spatiotemporal analysis using three
Landsat scenes of 25 years of time span, aims at the detection of the changes and the urban footprints in
Noida and Greater Noida and its surroundings. It is also assessed as to how much urban expansion has
taken place during the last 25 years. With the help of socioeconomic data, demographic profile and
increase in number of households are also assessed.
Keywords: Remotely sensed change detection, ratio transformation, geo-spatial and temporal approach,
development and environment, urban expansion, Noida, Greater Noida, India
*Author for Correspondence E-mail: rsgis.gkr@gmail.com, Tel: +91-9873843929
1. INTRODUCTION
The rate of urbanization is very fast in
developing countries especially in the Asian
continent. There are more than 170 urban
areas having population of over 750,000
inhabitants in India and China alone (United
Nations Population Division, 2001). Once
India was called the country of villages, but
now India no longer lives in villages because
more than 285 million people are living in
urban areas [1]. During the last 50 years, the
population of India (today 1.21 billion) has
grown two-and-a-half times, but the urban
population has grown nearly five times. The
number of Indian mega cities will increase
from the current four (Mumbai, Delhi, Kolkata
and Chennai) to six by the year 2021 (new
additions will be Bangaluru and Hyderabad),
when India will have the largest concentration
of mega cities in the world [2]. Statistics show
that India’s urban population is the second
largest in the world after China, and is higher
than the total urban population of all countries,
put together barring China, the United States
and Russia. In 1991, there were 23
metropolitan cities in India, which increased to
35 in 2001 [1]. It shows that the number of
cities is increasing day by day. Research in
detecting urban change and development using
satellite data has a long tradition in geographic
research and planning. Noida and Greater
Noida are getting populous mainly due to
availability of different means of earning
livelihood because Noida has emerged as one
of the most preferred industrial area in the

Journal of Remote Sensing & GIS
Volume 3, Issue 3, December 2012, Pages 42-51
ISSN: 2230 -7990© STM Journals 2012. All Rights
Reserved
Page
43
National Capital Region of Delhi. This area
proves a good place for living and attracts
huge masses due to availability of various
necessary facilities for cozy living including
means of communication, educational
facilities, shopping complexes, malls, utilities
and services, parks and other recreational
facilities. The reason for urbanization in this
area is preference of many individuals to live
in or near a city like Noida and/or Greater
Noida, which are known as commercial hub of
the country. Multitemporal remote sensing
data has become very important in the analysis
of such changes.
2. MAIN OBJECTIVES
This study delineates urban population growth
in the study area in two decades and five years
(25 years) using geospatial techniques. The
specific research objectives of this study are as
follows:
(i). To assess loss of agricultural land due
to increase in built up area.
(ii). To assess spatial and temporal patterns
of urban change.
(iii). To assess the land use/land cover
of Noida during 1986−2011.
(iv). To visualize urban growth by ratio
transformation.
(v). To assess the demographic profile of
Noida and increase in households too.
3. STUDY AREA
Coordinates
77°28'23. 677"E 28°30'32. 542"N
Area 53,000 ha (204 sq mile)
DistrictGautam Buddha Nagar
State, Country Uttar Pradesh, India
Population 750,057 (Census 2011)
Elevation 200 m (656 ft)
Noida (77°22'46.444"E, 28°35'13. 017"N)
acronym for New Okhla Industrial
Development Authority came into existence
on 19 April 1976. This city is considered as
one of the most modern suburbs of Delhi in
the National Capital Region. The city is
located in Gautam Buddha Nagar district of
Uttar Pradesh. Noida is situated in northern
part of India and shares boundary with Delhi.
River Yamuna binds Noida from the west and
southwest and river Hindon from the east and
southeast. Delhi shares boundary with Noida
from north and northwest side while
Ghaziabad from the northeast side. Noida
comes under the catchment area of Yamuna
river and is based on Yamuna’s old riverbed
[3].
Greater Noida (77°30'46.366"E,
28°27'51.508"N) is the proud city to host the
first Indian grand prix at Buddha International
Circuit. This city is developed as an extension
part of Noida due to its proximity to a few
industrial towns of Uttar Pradesh on the east.
The city is a result of intense pressure on
national capital of Delhi and its periphery.

Journal of Remote Sensing & GIS
Volume 3, Issue 3, December 2012, Pages 42-51
ISSN: 2230 -7990© STM Journals 2012. All Rights
Reserved
Page
44
This city shares its boundary with Ghaziabad
on the northern side and of course with Noida
on the western side. It is broadly bounded by
the main national highway, that is called the
G. T. Road (NH-24) and river Hindon binds it
from the western side. Unlike one of the minus
points of Noida it is connected with Indian
railway line on the eastern side (Figure 1)
[4, 5].
Their corresponding Industrial development
authorities govern both Noida and Greater
Noida cities.
Fig. 1: The Geographical Location of Study Area in India.
4. DATA AND METHODOLOGY USED
The main analysis depends on the Landsat
scenes of path row 146 040. Out of these three
scenes, the study area was subseted using a
shape file of area of interest. Then these two
scenes were rectified using survey of India
toposheet, 40 GCPs using second order
polynomial were marked across the scene of
study area and resampled to 30 m spatial
resolution. After that, radiometric correction
and image enhancement techniques were
applied (Table I).

Journal of Remote Sensing & GIS
Volume 3, Issue 3, December 2012, Pages 42-51
ISSN: 2230 -7990© STM Journals 2012. All Rights
Reserved
Page
45
Table I: Datasets Used in Analysis.
S. No. Name of Data Used Scale/Resolution Year
1.
2.
3.
4.
5.
6.
Landsat
Landsat
Survey of India Toposheet
Quickbird Scene
City Guide Map
Socio Economic Data
TM 5 30 m
MSS 57 m
1:50000
0.6 m
Not scaled
Yearly and Decadal
2000 and 2011
1986
1999, 2009
2011
2011
1991, 2001 to 2011
Without applying radiometric correction, it is
not possible to get good results from
comparison of temporal imageries, even
acquired on the same date (different years) and
generated by the same sensor. A few factors
responsible for poor or incorrect results are
atmospheric conditions (presence of clouds),
variation in solar illumination conditions,
change in sensors radiometric performance
over time, atmospheric scattering and
absorption. So, it is necessary to apply
radiometric correction on the imagery [6].
If any two datasets are to be used for
quantitative analysis based on radiometric
information, as in the case of multi-date
analysis for detecting surface changes, they
ought to be adjusted to compensate for
radiometric divergence. Two approaches were
used in the study area to quantify the loss of
agriculture due to urbanization. The second
approach is for visualization purpose only;
still, it represents a good picture of change.
First Approach: Change Detection Analysis
In this approach, change detection analysis
was done using supervised classification
(maximum likelihood method). Landsat scenes
were classified to generate five classes. These
classes are:
• Water bodies: covering all type of water
classes, i.e., river, pond, seasonal water,
drainage, shallow water and canal
• Vegetation: deciduous forest, mixed forest
lands, palms, conifer, scrub and others
• Cultivated land: agricultural area and
fallow lands
• Built-up area: residential, commercial,
industrial, settlement, village abadi, mixed
urban, minor roads and other urban
• Bare
soil/Landfill sites: open area,
exposed soils, barren land, landfill sites,
major roads and areas of active excavation
Using the maximum likelihood classifier, all
three images were classified.
Then the
classified data was converted into the data
with sieve classes. Sieve classes method
removes isolated pixels occurring
in
classification images using blob grouping.
After applying sieve classes method, every
unclassified pixel was assigned to its native
class using GIS approaches [7].

Journal of Remote Sensing & GIS
Volume 3, Issue 3, December 2012, Pages 42-51
ISSN: 2230 -7990© STM Journals 2012. All Rights
Reserved
Page
46
Accuracy Assessment: “Accuracy measures
the arrangement between a standard (assumed
to be correct) and a classified map. This
represents the correctness of the classified
map. If the final map corresponds closely to
the standard, the classified map is thought to
be accurate”. The classified Landsat scene of
the year 2000 has 88.6254% accuracy with
kappa coefficient = 0.8771 and the Landsat
scene of year 2011 has 82.6984% accuracy
with kappa
Fig. 2: Methodology Used in Study.
coefficient = 0.8161. Accuracy of classified
image of 1986 cannot be assessed due to lack
of reference data. Various approaches were
used to achieve that much accuracy. With
these resulting images, change detection
analysis was performed to perfectly identify
the change (Figure 2).
Second Approach: Ratio Transformation
Ratio transformations can be applied on
remotely sensed data to reduce the effects of
environment and to get information from the
desired layers.
According to Landsat science band 3 that lies
in between 0.63 to 0.69 µm wavelength region
is a strong reflectance region for soils and

Journal of Remote Sensing & GIS
Volume 3, Issue 3, December 2012, Pages 42-51
ISSN: 2230 -7990© STM Journals 2012. All Rights
Reserved
Page
47
keeps sensing in strong chlorophyll absorption
region. It has the capability to discriminate
between soils and vegetation. This band
highlights urban areas, barren lands and street
patterns. On the other hand, band 4 that lies in
between 0.76 to 0.90 µm wavelength region
proves its usefulness in crop identification
because this band is operational in the best
spectral region to distinguish vegetation
varieties and conditions. This band also
highlights water bodies but barren land, urban
areas and major roads have not been
highlighted.
So, according to James W. Quinn [8], applying
the ratio of band third and fourth of Landsat
TM 5 defines barren land and urban area
uniquely.
Landsat TM 5 Band3/Landsat TM 5 Band 4 = Bare Soil and Urban Area
Fig. 3: Landsat TM 5 “Ratio Transformation,” One Layer Image Depicting Built-up Area and Bare
Soil Pixels Only.
This approach was applied using image-
processing software to get output as a single-
layered image having built-up area and bare
soil only. This approach is for visualization
purpose only in this study (Figure 3).
5. RESULTS AND DISCUSSION
According to change detection analysis, a very
wide change has been observed. The LULC
classes those have changed drastically are
cultivated land and built-up area. Ultimately,
out of these two classes built-up area has
increased while cultivated land has decreased.
In the year 1986, agriculture (cultivated area)
was the dominating land cover class.
Agriculture was spread in more than 50% of
total area while built-up has occupied 19% of
the total area.
In the year 1986, vegetation was all time high
(during present study period) having 15% of
its share in total area that decreased by 11% in
the year 2000 and increased a little (2%) in the

Journal of Remote Sensing & GIS
Volume 3, Issue 3, December 2012, Pages 42-51
ISSN: 2230 -7990© STM Journals 2012. All Rights
Reserved
Page
48
year 2011 due to planning of recreational area
by town development authorities. In the year
2000, agriculture was the dominating class
though it had decreased by 1% and built-up
got an increase of 4%. Year 2000’s image
analysis shows that the class bare soil/landfill
sites had grown up from 9% (in 1986) to 21%
(in 2000). Water bodies got a decrease of 4%
in the year 2000. Analysis shows that class
bare soil/landfill sites have their best share in
the year 2000 with 21% of total area
(Table II).
Table II: LULC Classes Generated after Classification Showing Area in Hectares during Different
Years.
Change in LULC Categories (Area in Hectares)
Class/Year 1986 2000 2011
W
ater
B
o
d
ies
2605
.
4
0
0780
.
8
9
1200
.
0
6
Ve
g
etati
o
n
7833
.
5
6
2349
.
9
3
3368
.
9
7
Cu
l
ti
v
a
ted
L
a
n
d
275
5
1
.
3
6
263
5
6
.
1
9
158
4
0
.
8
1
B
u
ilt
-
u
p
A
r
ea
100
9
0
.
6
8
123
8
5
.
1
7
249
9
6
.
6
3
Bare Soil Landfill Sites 4919.936 11128.59 7594.169
Fig. 4: Landsat Classified Scenes of Study Area during Year 1986, 2000, 2011, Increase in Built-up,
Open Land and Loss of Agricultural Land Can Be Seen Easily.

Journal of Remote Sensing & GIS
Volume 3, Issue 3, December 2012, Pages 42-51
ISSN: 2230 -7990© STM Journals 2012. All Rights
Reserved
Page
49
Fig. 5: Change in LULC Categories Fig. 6: Regression between Population and Built-up.
(Area in Hectares) during 25 Years.
In year 2011, the dominating class was built-
up area with 48% share of the total area,
having 29% of huge change in comparison of
built-up in year 1986 and cultivated area
shows a huge loss of 22% in comparison of
year 1986. In 1986, cultivated land was
27,551.36 ha that has shrunk to 15,840.81 ha
in 2011. The above records (Figures 4 and 5)
mirror that urbanization is the most
widespread cause of the loss of arable land,
decline in natural vegetation cover as well as
habitat destruction.
Generally, if population increases built-up area
will definitely increase. Regression analysis
between population and built-up has very
strong R
2
(.9697) and provides strength to the
above fact. It shows population is proportional
to built-up area (Figure 6).
5.1. Demographic Profile
As per 1981 census, the population of Noida
was 36,972 and during 1991 census, it was
declared a “Census Town” with a population
of 146,514. The population of Noida grew by
nearly 300% during 1981–1991 and it became
one of the fastest growing towns in the
country. In the decade of 1991-2001,
population grew with a growth of 108.21%
(Figure 7). The town contained 68% of the
total urban population of Gautam Buddha
Nagar district completely overshadowing other
towns like Dadri, Dankaur and Jewar. The
1991–2001 decade witnessed slowing down of
growth,
which was 108.21%. The decade 1981–1991
was the take off stage of the new township
(Noida Development Authority).
According to census of year 2001[1], the
population reached 305,058 and recent census
of year 2011 shows a drastic change in
Noida’s population by an increment of 111%
with 642,381 people. During 1981–1991, the
population increased by a difference of
110,973 inhabitants, in 1991–2001 decade by
158,544 inhabitants and in 2001–2011 decade
by 337,323 inhabitants. Greater Noida had
107,671 inhabitants in the year 2011.

Journal of Remote Sensing & GIS
Volume 3, Issue 3, December 2012, Pages 42-51
ISSN: 2230 -7990© STM Journals 2012. All Rights
Reserved
Page
50
2001
2011
2021
Proposed
Resi
d
en
t
i
a
l
2917
.
0
0
6240
109
9
4
Industrial 2688.96 4496 6035
Commercial 0135.74 1380 1724
Institutional 1141.63 2182 3744
Recreational 1569.90 3540 6776
Transportation 1150.32 2582 5244
Table III: Landuse Distribution in Noida and
Greater Noida (Area in Hectares).
(Source: Noida and Greater Noida Development
Authorities)
Fig. 7: Increase in Population of Noida City
between Three Decades.
5.2. Landuse Distribution in Noida and
Greater Noida
According to the respective development
authorities of the two towns, following records
of landuse were provided in the draft of master
plan for 2011 and revised draft of master plan
for 2021 (Figures 8 and 9) [3, 4]. Table III
shows the combined landuse pattern
distribution for Noida and Greater Noida.
Table facts reveal that main emphasis for
development is on residential and industrial
categories under built-up class with an
increment in recreational and transportation
categories also.
Fig. 8: Landuse Distributions in Noida and Greater Fig. 9: Proposed Landuse Distributions
Noida Area in Year 2001, 2011, and 2021 proposed. (Ha) of year 2021.
(Source: Noida and Greater Noida Development Authorities)
6. CONCLUSIONS
This study explains LULC changes and the
extent of urban expansion in Noida andGreater
Noida, India, using remote sensing satellite
data in conjunction with socio-economic
dataset. Urban expansion and loss of
agriculture was quantified for the last 25 years

Journal of Remote Sensing & GIS
Volume 3, Issue 3, December 2012, Pages 42-51
ISSN: 2230 -7990© STM Journals 2012. All Rights
Reserved
Page
51
using the post-classification comparison
technique. The study area was found to have
experienced rapid changes in LULC,
particularly in built-up/urban areas and
obviously in cultivated land. Change detection
analysis reveals that the built-up area has
increased by 29% (from 10090.68 ha to
24996.63 ha) and cultivated land has
decreased by 22% (from 27,551.36 ha to
15,840.81 ha) during 1986 to 2011.
Results of ratio transformation and post-
classification reveal that only two categories,
i.e., built-up area and bare soil/landfill sites,
have occupied 32,590.80 ha out of 53,000 ha
(61% of total area). According to census of
India [1], district Gautam Buddha Nagar used
to have 204,302 households in year 2001 that
increased by 111,976 becoming 316,278 in
2011. This fact also shows a huge rise in
urbanization in the last decade. The conversion
of agricultural land, vegetation and open area
to built-up area/urban land has caused varied
environmental degradation with loss of fertile
land and its main negative outcomes are
directly associated with urban expansion.
Over all, this study reveals decadal change in
various LULC classes using RS, GIS and
socio-economic data. Being the most
influential national capital region towns,
Noida and Greater Noida are rapidly running
towards urban expansion. Few months ago in
Greater Noida, many hectares of agricultural
land was acquired from farmers by the state
government to develop industrial areas and
residential townships. So, in the near future,
this area will definitely show a great change in
terms of urbanization.
REFERENCES
1. Census of India. Provisional Population
Totals. Office of Registrar General of India,
Government of India, New Delhi. 2001 and
2011
2. Taubenböck H. et al. Spatiotemporal
Analysis of Indian Mega Cities. The
International Archives of the
Photogrammetry, Remote Sensing and
Spatial Information Sciences. XXXVII(B2).
Beijing 2008.
3. Master Plan for Noida-2011, Year 2004.
New Okhla Industrial Development
Authority.
4. City profile of Greater Noida. Greater Noida
Development Authority. 2006.
5. Master Plan for Greater Noida-2011, Year
2006. Greater Noida Development
Authority.
6. Bhatta B. Remote Sensing and GIS. New
Delhi. Oxford University Press. 2008. 385–
387p, 394–395 p.
7. Environment for Visualization of Images
(ENVI). Users Guide.
8.
Quinn W. James. Ratio Transformations.
2001. Available at:http://web.pdx.edu/~emc
h/ip1/bandcombinations.html.