See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/332294194

Modelling the growth of two rapidly urbanizing Indian cities

Article  in  Journal of Geomatics · October 2017

CITATIONS READS
5 231

4 authors:

Dr. Bharath H Aithal Chandan M C

Indian Institute of Technology Kharagpur The National Institute of Engineering, Mysore
253 PUBLICATIONS   2,151 CITATIONS    40 PUBLICATIONS   188 CITATIONS   

SEE PROFILE SEE PROFILE

Vinay Shivamurthy T V Ramachandra

SR University Warangal Indian Institute of Science
173 PUBLICATIONS   679 CITATIONS    1,108 PUBLICATIONS   12,440 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

ISRO-IISC Space Technology Cell- Biofuel ISTC/BES/TVR/389 View project

APN：The impact of spatial parameters on GHG emission: a comparative study between cities in China and India View project

All content following this page was uploaded by Chandan M C on 09 April 2019.

The user has requested enhancement of the downloaded file.

 149

Journal of Geomatics Vol 11 No. 2 October 2017

Modelling the growth of two rapidly urbanizing Indian cities

H.A. Bharath1,4, M.C. Chandan4, S. Vinay 1 and T.V. Ramachandra1,2,3
1Energy & Wetlands Research Group, Center for Ecological Sciences [CES]
2Centre for Sustainable Technologies (astra)
3Centre for infrastructure, Sustainable Transportation and Urban Planning [CiSTUP]
4RCGSIDM, Indian institute of Technology Kharagpur, West Bengal-721302, India

Indian Institute of Science, Bangalore, Karnataka, 560 012, India

Email: bhaithal@iitkgp.ac.in

(Received: May 16, 2017; in final form: Sep 26, 2017)

Abstract: Sustainable urban planning requires an understanding of the spatial and temporal patterns of urbanization
with insights to the sprawl. Indian cities have been experiencing a rapid urbanization consequent to globalization and
relaxations in market economy during the past three decades. This has posed challenges to the policy makers and
necessitated appropriate urban policies taking into account sensitiveness of a region and dynamic changes. Availability
of remote sensing data at regular intervals with Geo-informatics has proved to be highly efficient in identifying,
measuring and quantifying spatio-temporal patterns of urban growth. This communication quantifies the changes in
two rapidly urbanizing landscapes in India. This involved analysis of (i) land cover and land use changes, (ii) spatial
patterns of urbanisation through zone wise density gradients. Tier I cities namely Hyderabad and Chennai, emerging
IT giants in India were chosen for the analysis. Computation of Shannon’s entropy and spatial metrics aided in
assessing the sprawl and spatial patterns of urbanisation in a landscape. Modelling and prediction of likely changes in
the land use was done using an integrated approach with fuzzy, analytical hierarchical process (AHP), cellular
automata (CA) and Markov chain. Results of land use analysis revealed an increase in urban area from1.46% (1991)
to 18.81% (2013) for Chennai region and 1.75% (1989) to 22.19% (2014) for Hyderabad region. The spatial analysis
through prioritized eight metrics reveal a fragmented or dispersed growth in the outskirts and compact growth in the
core area. Modelling was performed considering a set of agents and constraints with two scenarios- implementation
of city development plan (CDP) and without CDP. Modelling of urban growth in 2025 reveals urbanized landscapes
of 36.6% and 51% in Chennai and Hyderabad respectively. Periphery of the major roads and outside the city
jurisdiction limits are favorable areas of urban growth with land use changes from agriculture or others category to
built-up category. Modelling and visualisation of urban growth equip regional decision making to provide basic
amenities and appropriate infrastructure considering the likely demand with urbanisation.

Keywords: Urbanisation, Cellular automata, Markov chain, AHP, Fuzzy

1. Introduction with increasing urbanisation trend, which is expected

to contribute to 75% of India’s GDP with serious
Urban growth is a form of metropolitan growth occurs erosion in food production by 2030.
when local patches of settlement agglomerate in
response to various economic, social, and political Unplanned urbanisation has impacted on the regional
forces and to the physical geography of an area. Urban environment evident from fragmentation of natural
area constitute predominantly of built up or paved landscape due to changes in landscape patterns with a
surfaces with transitions from forest patch, agricultural complex irreversible socio-economic phenomenon
fields or rural landscapes (Dupont, 2005). Earlier (Jat et al., 2008; Bharath et al., 2012). Unprecedented
civilizations consisted of small colonies adjacent to and irreversible urban growth and concentration of
rivers. This eventually led to the formation of villages, urban region in the city core, has led to escalation in
towns, cities and at present many are settled in land prices leading to sprawl in peri-urban regions
complex urban ecosystems also reflected by across the city boundary. The buffer region or zone of
population growth. Thus urban areas are formed transition exists between a rural landscape and urban
through progressive concentration of population. pockets (Pryor, 1968). Prior visualization of this
Statistics indicate that today 54% of the global dynamic zone of rural-urban continuum aid in the
population and 34.16% of Indian population reside in effective decision making. Sprawl thus refers to
urban areas (World Urban Prospects, 2014). Urban dispersed or the scattering of new developments on
areas are expected to house 40% of India’s population isolated tracts, separated from other land uses

© Indian Society of Geomatics

 150

Journal of Geomatics Vol 11 No. 2 October 2017

(Ottensmann, 1977). Also referred as the pattern of Information System (GIS) aids in capture, store,
low-density expansion near urban areas, mainly into query, analyze and display geo-spatial data (Chang,
the surrounding rural regions having urban rural 2006). Remote sensing is cost effective and
transitions. Urbanisation in these regions are patchy, technologically reliable, and is therefore, increasingly
scattered and strung out, with discontinuity and lack being used for urban sprawl analysis (Bharath, H.A. et
basic amenities such as treated water, sanitation, etc., al., 2014; Ramachandra et al., 2014; Vishwanatha et
These kind of unplanned growth have a tendency of al., 2015). Availability of temporal data acquired
attenuating natural resources as a consequence of large through space borne sensors drives remote sensing
land use change (conversion of green lands, water techniques better for its ability to characterize
bodies etc.) affecting directly on human health and spatiotemporal trends of urban sprawl that forms a
quality of life (Alberti, 2005; Ramachandra et al., basis for projecting future urbanization processes.
2009). Government agencies and planners often
neglect rural-urban transition land which leads to Spatial metrics aid in assessing the spatial patterns of
unsustainable development. Sprawl regions in most urbanisation through spatial heterogeneity of patches,
metropolitan regions have been posing serious classes of patches, or entire landscape mosaics of a
challenges with respect to electricity, water, sanitation, geographic area (O’Neill et al., 1988, Herold et al.,
waste management and other basic amenities, 2005). There are numerous metrics to quantify spatial
necessitating prior visualization of spatial patterns of patterns and the selection of spatial metrics depends
urban growth. Agents of urban growth are geography upon the study region (Irwin and Bockstael, 2007;
of a region, economic progression, population growth Furberg and Ban, 2012) and earlier studies (Wu, 2006;
and migration, industrialization, transportation, way of Hepinstall-Cymerman et al., 2013). Zone wise (based
living etc. (Barnes et al., 2001; Yang and Lo, 2003; on directions), gradient analysis of a particular region
Bruckner and Kim, 2003). helps in viewing the growth scenario at micro scale
and also helps to identify drivers or catalysts of
The growth of Indian urban centers in an urbanisation. Gradient analysis, earlier implemented
unprecedented rate has often led to deterioration of to analyze vegetation (Whittaker, 1975), has been used
balance in natural ecology, while impacting ambient to study the effects of urbanisation on plant
environment due to spurt in greenhouse gas (GHG) distribution (Kowarik, 1990; Sukopp, 1998), green
emissions, leading to global warming and consequents spaces (Kong and Nakagoshi, 2006) and ecosystem
changes in the climate (Ramachandra and Kumar, properties (Zhu and Carreiro, 1999). This
2010; Ramachandra et al., 2015). Unplanned communication focuses on combining temporal
urbanisation is resulting in urban sprawl with escalated remote sensing data, GIS with spatial metrics analyses
vehicle and traffic density (Ewing et al. 2002), impacts along density gradients helps to understand urban
on the biodiversity, environment and ecosystem (Xian land-use changes at local levels.
et al., 2007; Li et al., 2010), land use fragmentation,
human-animal conflicts (Hotton, 2001) and most Prediction of likely land uses is essential to provide
importantly the rapid changes in hydrological cycle vital inputs for urban planning, which will help to
with changing rainfall patterns and flooding regimes ensure sustainability and balance in the natural
(McCuen et al., 2003). Mitigation of the consequences ecosystem. Modelling refers to the data acquired to
of climate change and environmental degradation calibrate, validate, verify and predict future urban
necessitates an understanding of spatial patterns of trends (Batty, 1997, 1998). Various models available
urbanisation, quantification and visualization of urban for analyzing urban growth based on allocation of
growth and sprawl. different land use activities within a region are cellular
automata (CA), Markov chain, analytical hierarchical
Sustainability of natural resources entails planning and process (AHP), slope, land use, exclusion, urban
stewardship in management by the government and extent, transportation and hill shade (SLEUTH),
other agencies considering population growth and artificial neural network (ANN) and decision making
urban expansion. This is possible only with the tool such as multi criteria evaluation (MCE).
inventorying, mapping and monitoring of urbanisation
process through land use and land cover dynamics Recently, the Government of India (GoI) has
analysis (Ramachandra et al., 2013). Recent embarked on ‘Smart City’ concept to boost economy,
advancements in remote sensing technologies and infrastructure and improve quality of living in
Geoinfomatics have further boosted efforts to analyze emerging urban regions in India. The objective needs
growth (Bharath S, et al., 2012; Ramachandra et al., to be towards enabling E-governance for efficient
2014a). Space borne sensors assists in inventorying, management of natural resources, including urban
mapping and monitoring earth resources. Geographic mobility and housing, waste management, etc. to
 151

Journal of Geomatics Vol 11 No. 2 October 2017

ensure sustainability at the same time maintaining spatial metrics and (ii) predict land use dynamics in
ecological balance. GoI programme of 100 smart cities 2025 for Chennai and Hyderabad through an
(Smart cities, 2015) includes Chennai and Greater integrated modelling framework considering
Hyderabad, two rapidly growing metropolitan cities. geographic, topographic and socio-economic factors.
Chennai also figures in one among 35 global mega
cities (population greater than 10 million people). 2. Study area and data
Advance visualisation of urban growth help in this
regard to identify growth poles and provide Chennai is capital city of Tamilnadu state, India. It is
appropriate infrastructure and basic amenities. Models located between two major rivers i.e. Coovum and
based on CA and Markov chain aided by analytical Adayar and at the eastern coast - Coromandel Coast
hierarchal process (AHP) and fuzzy to account agents line also known popularly as “Gateway to South
with the weightages of influences. This involved India”. Chennai is known as “Detroit of India” due to
estimation of Eigen vectors or priority vectors the presence of a wide array of automobile industries.
followed by measure of consistency using consistency Chennai has tropical wet and dry climate with
ratio (Khwanruthai and Yuji, 2011). Decision support temperatures ranging from 15˚- 40˚C. The jurisdiction
tool MCE is adapted to evaluate choice between of the Chennai (city) Corporation was expanded from
alternative factors. This process is necessary for CA 174 sq. km (2001) to 426 sq.km in 2011. Chennai
models to generate site suitability maps for future land Metropolitan Area (CMA) has an area of 1189 km2
use predictions. CA is a discrete two dimensional comprising Chennai city district and partially
dynamic systems with local interactions among extending to two districts Kancheepuram and
components generate global changes in space and time Tiruvallur. Chennai is presently fourth most populous
(Wolfram, 2002). CA follows a “Bottom-up” city in India with 4.68 million (2011), whereas CMA
approach, in which the future state of the pixel depends population shows an increase of 1.86 million
on its past and current state with a set of specified considering 2001 and 2011 census.
transition rules. Finally, CA-Markov chain analysis
provides the transition probability matrix and Hyderabad is a capital of Telangana state and Andhra
transition area matrix. CA are thus not just a Pradesh (after partition in 2014). The city is located
framework for dynamic spatial modelling but provide along the banks of river Musi and surrounded by many
insights about complex spatial-temporal phenomena lakes like Himayat Sagar, Hussain Sagar, etc. It has
and constitute an experimental laboratory for testing very old history since 1500’s under Nizam’s rule.
ideas. Predication of urban dynamics using CA model Hyderabad is the largest contributor to the gross
is flexible due to easy integration with GIS (Wagner, domestic product. With creation of special economic
1997). CA has been adopted earlier to simulate land zones at Gachibowli, Pocharam, Manikonda etc.
use changes (Lau and Kam, 2005; Stevens and dedicated to have encouraged companies from across
Dragicevic, 2007) and also by considering spatial India and around the world to set up operations.
agents (Loibl and Toetzer, 2003), transition rules Erstwhile Hyderabad urban development authority
(Almeida et al. 2005), neighborhood functions (White (HUDA) was expanded in 2008 to form Hyderabad
and Engelen, 2000; Yuzer, 2004) and mapping urban metropolitan area (HMA) covering 7100 km2 and
and non-urban states (Cheng and Masser, 2004; He et population of 7.74 million (2011). HMA covers a total
al. 2006; Li et al. 2008). CA coupled with Markov of 5 districts namely Hyderabad, Rangareddy, Medak,
chain helped to demonstrate quantify the states of Mehaboobnagar and Nalgonda. Chennai and
conversion between land-use types, especially from Hyderabad are at the verge of attaining “Mega city”
forest, agriculture, wetland and other landuse status (urban agglomerations greater than10 million
categories to urban landuse (Mukunda et al. 2012; inhabitants), while India already has 3 mega cities
Praveen et al. 2013; Hossein and Marco, 2013). namely Mumbai, Delhi and Kolkata (United Nations,
2012).
The main objectives of this research are (i) quantify
urbanization and urban sprawl process with the help of The geographical bounds of the two study cities are
temporal remote sensing data, density gradient and given in table 1.

Table 1: Geographical extents of Chennai and Hyderabad metropolitan area

City Latitude values (N) Longitude values (E)

Chennai 12˚51’04’’ 13˚17’29’’ 79˚59’45’’ 80˚20’16’’
Hyderabad 17˚12’51’’ 17˚42’26’’ 78˚12’34’’ 78˚45’29’’
 152

Journal of Geomatics Vol 11 No. 2 October 2017

3. Data acquisition buffer zone of 10 km from an administrative boundary

(centroid as central business district) as in figure 1 was
Temporal data of Landsat 5 (TM), Landsat 7 (ETM+) considered to account for likely growth or sprawl at
and Landsat 8 (OLI-TIRS) satellites were downloaded outskirts or peri-urban regions. Data used land cover
from the public archive of USGS Earth Explorer. IRS and land use analyses and for transition rules are listed
LISS-III data was procured from the National Remote in table 2.
Sensing Centre (NRSC) for the year 2012. A circular

Figure 1: Location of Chennai and Hyderabad

Table 2: Data used for the analysis

DATA YEAR PURPOSE

Landsat 5 TM (30m) 1989,1991 and Land use and land cover analysis
1999
Landsat 7 ETM+ (30m) 2000 and 2009
Landsat 8 OLI-TIRS (30m) 2013 and 2014
IRS LISS-III (23.5m) 2012
Aster GDEM 2010 Extraction of drainage lines, slope analysis
Google earth Geo-correction, classification and validation.
Collection of point, line and polygon data
Boundary maps and raster layers To create agents and constraints data sets based
on city development plans
Survey of India Topographic Base layers of the administrative boundary
maps, online portals (Google and
Bhuvan data)
Field data – using GPS Geo-correction, training data and validation data

4. Method cropped corresponding to study regions with ten km

buffer. Co-ordinates of known locations such as road
Preprocessing: This involved geo-referencing of data, intersections, edges of huge permanent structures like
done with the help of known location points (compiled dams, bridges etc. were compiled using GPS and
from the Survey of India topographic maps and also online high resolution data (Google-earth) at
from field using pre-calibrated GPS – Global inaccessible places. Further, resampling was
Positioning System). Remote sensing data were performed to maintain the spatial resolution
 153

Journal of Geomatics Vol 11 No. 2 October 2017

uniformity across temporal remote sensing data. distributed across the entire study area. These
Histogram equalization was performed wherever polygons and its coordinates with GPS and attribute
enhancement was necessary to maintain the dynamic information is compiled with respect to corresponding
range. Landsat and IRS LISS-III images were co- land use type (ground truth data). Training polygons
registered to WGS 1984 and UTM zone 44. were supplemented with the data available at Google
earth for classification. 60% of these training polygons
Land cover analysis: Land cover refers to the original were used for classification purpose while the rest
earth surface features that are formed naturally in the 40% for validation and accuracy assessment.
form of vegetation, water body, etc. (Ramachandra et. Supervised Gaussian maximum likelihood
al., 2013). Land cover analysis helps to understand the classification (GMLC) was employed to assess
changes of the vegetation cover over the study area at quantitatively land uses in the region. GMLC
different time periods. It is obtained by performing algorithm considers cost functions as well as
normalized difference vegetation index (NDVI). probability density functions and proved to be efficient
NDVI value ranges from -1 to +1. Values consisting among other classifiers (Duda et al., 2000). It
of -0.1 and below indicates soil, barren land, rocky evaluates both variance and co-variance of the
outcrops, built up/urban cover, whereas water bodies category while classifying an unknown pixel
are indicated by zero values. Low density vegetation (Lillesand et al., 2012). Land use classification under
is indicated in the range +0.1 to +0.3 while high four categories (table 3) using GRASS.
density vegetation or thick forest canopy is given in
the range +0.6 to +0.8. Accuracy assessment: Possible errors during spectral
classification are assessed by a set of reference pixels
Land use analysis: Land use analysis starts with collected by ground data collection. Based on the
generation of false colour composite (FCC)of 3 bands reference pixels, statistical assessment of classifier
(Green, Red and NIR). Creation of FCC directly helps performance including confusion matrix, kappa (κ)
in identifying heterogeneous patches in the landscape statistics and producer and user's accuracies were
(Ramachandra et al., 2014). Training polygons are calculated. These accuracies relate solely to the
digitized based on the distinguishable heterogeneous performance of spectral classification. Entire method
features in FCC, covering at least 15% and uniformly followed has been summarized in figure 2.

Figure 2: Method adopted to understand, quantify and model urban growth

 154

Journal of Geomatics Vol 11 No. 2 October 2017

Table 3: Land use categories

Category Features involved
Builtup Houses, buildings, road features, paved surfaces etc.
Vegetation Trees, gardens and forest
Water body Sea, lakes, tanks, river and estuaries
Others Fallow/barren land, open fields, quarry site, dry river/lake basin etc.

Density gradient and zonal analysis: Earlier Ramachandra et al., 2015) six metrics to characterize
investigations of spatial patterns of urbanisation were urban growth.
restricted to political boundaries (Taubenbock et al.,
2009; Deng et al., 2009; Sadhana et al., 2011). In order Modelling: Urban growth during 2025 is predicted
to understand the growth at local levels, specific to considering agents with constraints (listed in table 5)
neighborhood, the entire study area was divided based and base layers of historical land uses (based on the
on directions into four zones (i.e. North East (NE), classified temporal remote sensing data). Data values
North West (NW), South East (SE) and South West were normalized (between 0 and 255) through
(SW) and concentric circle with the central business fuzzyfication wherein 255 indicates maximum
district as centroid and incrementing radii of 1km. The probability of land use changes. Fuzzy outputs thus
zone wise concentric circle based analyses was derived are then taken as inputs to AHP for different
performed helped to interpret, quantify and visualize factors into a matrix form to assign weights. Each
forms of urban sprawl pattern (low density, ribbon, factor is compared with another in pair wise
leaf-frog development) and agents responsible in comparison followed by enumeration of consistency
urbanization at local levels spatially (Ramachandra et ratio which are to be <0.1 for the consistency matrix
al., 2014b). to be acceptable (Saaty, 1980).

Spatial patter analysis: Shannon's entropy (Hn) is Constraints were assigned considering city
computed (equation 1) to determine whether the development plan (CDP), Digital elevation model and
growth of urban areas is compact or dispersed growth. slope data. Drainage lines were delineated using
Dispersed growth is also known as ‘urban sprawl’ This ASTER DEM and a buffer of 30m from drains were
analyses gives a better understanding of degree of assigned constraint to restrict development as per the
spatial concentration or dispersion of geographical guidelines of the regional metropolitan development
variables among “n” concentric circles across four authority. Constraints and factors were fed to multi
direction zones. Also, the regions undergoing sprawl criteria evaluation (MCE) (Table 5). The MCE
needs decision makers’ attention to provide approach combines various criteria into a single index
appropriate infrastructure and adequate basic that indicates the site suitability of specific land use of
amenities. each location in the study area. Markovian transition
estimator provided bi-temporal land use data to
Hn = -∑𝑛𝑖=1 𝑃i log (Pi) …………… (1) estimate transition and predict future likely land uses.
Probability distribution map was developed through
where, Pi is the proportion of the built-up in the ith Markov process. First-order Markov model based on
concentric circle. Shannon’s Entropy, values ranges probability distribution over next state of the current
from 0 to log n. 0 if the distribution is maximally cell that is assumed to only depend on current state.
concentrated whereas log n indicates sprawl. CA was used to obtain a spatial context and
distribution map. Transition suitability areas and
Spatial metrics: Metrics pertaining to spatial matrix, iterations to be performed and filter
heterogeneity of patches, classes of patches, or entire stipulations are carried out by CA coupled with
landscape mosaics of a geographic area (O’Neill et al., Markov chain to predict future land use. Validation of
1988, Herold et al., 2005) give quantitative description predicated data was done by comparing reference
based on the composition and configuration of the (classified) image versus the predicted image for the
urban pixels in a landscape. Spatial metrics were same year (in this case 2014). Various kappa indices
computed for urban class through FRAGSTATS of agreement and related statistics were calculated.
(McGarigal and Marks, 1995) for each zone and Further, Land use is predicted for 2025.
density gradients. Table 4 lists prioritized (based on
our earlier work (Ramachandra et al., 2012,
 155

Journal of Geomatics Vol 11 No. 2 October 2017

Table 4: Spatial metrics used

Indicators formula Range

1 Number of patches- NP = n (no. of patches in landscape) NP>0
NP

2 Normalised ei−min ei
NLSI =
max ei – min ei
landscape shape
index-NLSI ei= total length of edge (or perimeter) of class i in terms of number
of cell surfaces; includes all landscape boundary and background 0 ≤ NLSI ≤ 1
edge segments involving class i. min ei = minimum total length of
edge. Maximum ei = maximum total length of edge.

3 Clumpiness- Gi−Pi
for Gi<Pi and Pi<5; else
CLUMPY Pi
Gi − Pi
1 − Pi −1≤CLUMPY≤1
gii
where, Gi =
(∑𝑚
𝑘=1 𝑔𝑖𝑘 )−min 𝑒𝑖
gii= number of like adjacencies between pixels of patch type (class)
i based on double count method.
gik=number of like adjacencies between pixels of patch type (class)
i and k based on double count method. min ei=minimum perimeter
of patch type (class) i for maximally clumped class. Pi= proportion
of the landscape occupied by patch type (class) i.

4 Aggregation index –
AI
0 ≤ AI ≤ 100
gii=number of like adjacencies between pixels of patch type
Pi=proportion of landscape comprised of patch type

5 Largest patch index – LPI =

max 𝑎𝑖𝑗
* (100)
LPI 𝐴 0 < LPI ≤ 100
Where,
aij=area (m2) of patch ij.
A=total landscape area (m2).

6 Interspersion and
Juxtaposition – IJI

0≤ IJI ≤100
eik=total length (m) of edge in landscape between patch types
(classes) i and k.
E: total length (m) of edge in landscape, excluding background .
m=number of patch types (classes) present in the landscape,
including the landscape border.

Table 5: Agents and constraints considered for modelling

Agents Industries, proximities to roads, railway stations, metro stations, educational

institutes, religious places etc.
Constraints Drainage lines, slope, water bodies, regulated regions for non-development,
Protected areas, coastal regulated areas, sensitive regions in a catchment, etc.
 156

Journal of Geomatics Vol 11 No. 2 October 2017

5. Results and discussion during 1989-2014, highlight the grave situation in the
region and the need to restore and rejuvenate water
Land cover analysis: This analysis helps in bodies which aid as a lifeline of the society. Table 7
delineating regions under vegetation and non- summarizes the land use details for Chennai and
vegetation. Vegetation cover analysis was done Hyderabad respectively.
through NDVI. Figure 3, 4 and table 6 indicates the
land cover changes of different time periods for
Chennai and Hyderabad regions respectively. In
Chennai, vegetation cover has dramatically decreased
from 70.47% (1991), to 35.53% in 2013, whereas the
non- vegetation i.e. built up, paved areas, bare soil etc.
have increased 29.53% in 1991 to 64.47% in 2013.
Hyderabad also shows similar trend with decrease in
vegetation from 95.64% (1989), 93.28% (1999),
82.67% (2009) and 61.15% (2014). Land use analyses
was performed to understand the transitions across
land use categories like built up, forests, water bodies,
etc., Vegetation cover and water bodies aids in
moderating local climate and also help in mitigating
floods, etc.

Land use analysis and Accuracy assessment: Figure

Figure 3: Land cover changes 1991-2013, Chennai
5 and 6 represents land use dynamics for the study
region
regions during the past four decades. Results revealed
the steep increase of 72% in built up areas in Chennai
at Ponneri, Pattabiram, Sriperumbudur, Tambaram,
etc. during 1991-2000 and 646% during 2000-2013.
Areas such as Malakpet, Madapur, Bollaram,
Kukkatpally, Cherlapally, etc. Hyderabad showed an
increase in built-up area by 93% during1989-1999,
319% (during 1999-2009) and 56% (2009-2014). It is
important to notice that both the study regions show
significant increase during the years 2000-2010 with
emergence of various industrial sectors such as
automobile, hardware manufacturing as well as
information technology parks. Others category
(mainly open spaces) has consistently reduced from
69.5% to 50.5% (1991-2013) in Chennai and 90.5% to
72.6% (1989-2014) in Hyderabad region indicating a
large scale conversion to urban land use type. Water
Figure 4: Land cover changes during 1989-2014 in
bodies of Hyderabad shows a very critical decrease
Hyderabad region
which indicates either these land uses are converted or
they have been dried up. Decline from 3.75% to 1.84%

Table 6: Temporal land cover details for Chennai and Hyderabad

CHENNAI HYDERABAD
Non- Non-
Year Vegetation (%) Year Vegetation (%)
Vegetation (%) Vegetation (%)
1991 70.47 29.5 1989 95.64 4.36
2000 56.7 43.27 1999 93.28 6.72
2012 48.18 51.85 2009 82.67 17.4
2013 35.53 64.47 2014 61.15 38.85
 157

Journal of Geomatics Vol 11 No. 2 October 2017

Figure 5: Land use dynamics during 1991 to 2013 in Chennai metropolitan area

Table 7: Land use dynamics in Chennai and Hyderabad

Chennai Hyderabad
Year 1991 2000 2012 2013 2016 1989 1999 2009 2014 2016
Urban 1.46 2.52 18.55 18.81 22 1.75 3.39 14.21 22.19 24.18
Vegetation 1.38 0.8 1.51 2.76 1.83 4 3.53 3.83 3.38 2.43
Water 27.64 27.25 28.15 27.92 28.34 3.75 2.89 2.46 1.84 0.64
Others 69.52 68.35 51.38 50.51 47.83 90.5 90.19 79.5 72.59 72.76

Figure 6: Land use dynamics in Hyderabad

 158

Journal of Geomatics Vol 11 No. 2 October 2017

Table 8 lists overall accuracy and Kappa statistic for in Chennai and circles 1-11 in Hyderabad), each
land use classified information for Chennai and patch has agglomerated into a single large urban
Hyderabad. Overall accuracy for Chennai varied from patch i.e. there is a saturated urban landscape with
86% to 97% and for Hyderabad 87% to 94% highlights no other land uses (Egmore, Nugambakkam in
the agreement of classified information with the field Chennai and Abids, Secunderabad,
data. Narayanaguda, Somajiguda, etc. in Hyderabad).
Sprawl is evident with higher number of patches
Table 8: Accuracy assessment of Chennai and in NW, SW directions (Chennai) and NE, SE, SW
Hyderabad regions directions (Hyderabad).

CHENNAI HYDERABAD  Normalized landscape shape index (NLSI): This

Year Overall Kappa Year Overall Kappa metrics provides measure of class aggregation.
Accuracy Accuracy All four zones show lesser value of NLSI in 2013
(%) (%) and 2014 compared to 1991 and 1989 (reflected
1991 92 0.92 1989 94 0.73
in figures 9 and 10 for Chennai and Hyderabad).
2000 91 0.9 1999 87 0.85
The minimum values (NLSI < 0.5) especially in
2012 97 0.93 2009 90 0.90
2013 86 0.78 2014 91 0.76
CBD areas (such as Ambattur, Nungambakkam,
Sowcarpet, Egmore, in Chennai region and also
Nampally, Secunderabad, Medhipatnam etc. in
Urban sprawl analysis: Shannon’s entropy is
Hyderabad region) indicates that the landscape
calculated direction wise considering the proportion of
consists of a single square urban patch or it is
built-up/paved urban area in the gradient and results
maximally compact (i.e., almost square) in
are listed in table 9. Shannon’s entropy values ranges
contrast with the higher values in 1991 and 1989
from 0 (concentrated growth) and log n (dispersed
(NLSI ≈ 1, specifying maximally disaggregated
growth or sprawl). ‘n’ indicates the number of circles
urban patches with complex shapes.
/gradients in the respective direction. Analysis
highlights the tendency of urban sprawl during 2000
and 2012 for Chennai and 2009 and 2014 for  Clumpiness: This metric indicates the
Hyderabad. Higher entropy values of 0.444 (NE), aggregation and disaggregation for adjacent
0.396 (NW), 0.409 (SE) for Chennai and 0.442 (SW); urban patches. Figures 11 and 12 show the values
0.352(NE), 0.422 (NW), 0.444 (SE) and 0.355 (SW) closer to 0 for the regions corresponding to
in for Hyderabad shows of dispersed growth. The circles 25-35, in NE direction (Manali new town,
sprawl phenomenon is evident in figures 5 and 7. Ennore and SW direction includes
Irungattukottai, Kondavakkam) of Chennai and
Quantifying spatial patterns of urbanization the regions corresponding to circles 23-31, in
through metrics: Six metrics were computed using NE, SE and SW directions (Keesara Mandal,
FRAGSTATS for each gradient, zone wise to Rampally, Manneguda, Shankarpally) in
understand the spatial patterns of urban growth. Hyderabad, highlighting less compact growth or
maximal disaggregation. In 2012 and 2014
values, approaching +1 in core city areas (circles
 Number of patch (NP): Figures 7 and 8 gives NP
1-15) of all directions indicate of very complex
for Chennai and Hyderabad. Patches shows rapid
growth with all maximally aggregated patches
growth in all the directions pointing out
forming large urban monotype patch.
fragmentation during 2009, 2012, 2013 and 2014.
During 2013 and 2014, core city area (circles 1-9

Table 9: Year wise Shannon’s entropy values for the two cities

CHENNAI HYDERABAD
Year/Direction NE NW SE SW Year/Direction NE NW SE SW
1991 0.052 0.041 0.078 0.048 1989 0.029 0.046 0.081 0.055
2000 0.116 0.108 0.107 0.118 1999 0.034 0.052 0.106 0.096
2012 0.423 0.468 0.416 0.473 2009 0.249 0.326 0.354 0.321
2013 0.444 0.396 0.409 0.442 2014 0.352 0.422 0.444 0.355
Threshold limit = log 37 = 1.568 Threshold limit = log 33 = 1.518
 159

Journal of Geomatics Vol 11 No. 2 October 2017

Figure 7: Number of patch metrics for Chennai

Figure 8: Number of patch metrics for Hyderabad

Figure 9: NLSI metrics for Chennai

 160

Journal of Geomatics Vol 11 No. 2 October 2017

Figure 10: NLSI metrics for Hyderabad

Figure 11: Clumpiness metrics for Chennai

Figure 12: Clumpiness metrics for Hyderabad

 Interspersion and juxtaposition index (IJI): This associated only with one other adjacent patch
metric show how well an urban patch is type. This phenomenon does not hold well at the
associated or interspersed with other adjacent outskirts since urban patch is equally adjacent to
patch types. Lower values as observed (figures all other patch types (i.e., maximally interspersed
13 and 14) in 1990’s indicates an urban patch is and juxtaposed to other patch types) showing
sprawl in these areas.
 161

Journal of Geomatics Vol 11 No. 2 October 2017

Figure 13: Interspersion and juxtaposition metrics for Chennai

Figure 14: Interspersion and juxtaposition metrics for Hyderabad

Urban growth modelling: Land uses in 2025 were constraint of CDP wherein water bodies, forest areas,
predicted considering various agents (amenities, road catchment areas and coastal regulation areas (only in
and railway network) and constraints (protected areas, Chennai region) as no development areas. and (ii)
drainage lines and slope) for Chennai and Hyderabad without considering constraint of CDP. Table 11 lists
regions to visualize and understand the likely urban percentage changes in land use categories, especially
growth i. Pair wise comparison between two factors two fold increase in built-up areas with the decrease in
were done to obtain weights for these factors using vegetation and other categories. Two scenarios i.e.
AHP. Consistency ratio of 0.05 and 0.07 were with CDP and without CDP showed similar statistics,
achieved for Chennai and Hyderabad respectively, but it is very essential to note that with the constraint
which is considered satisfactory to continue with of implementation of CDP, urban growth would be at
further analysis. Land use changes for the year 2013 the outskirts or at the periphery of the city boundary.
for Chennai region and year 2014 for Hyderabad However, in absence of CDP, distressing trend of large
region were simulated. This helped in validation for scale land use changes in areas within the CMDA
comparing simulated land use with the actual land use boundary such as Korathur and Cholavaram lake bed,
based on the classification of respective remote Redhills catchment area, Perungalathur forest area,
sensing data. Satisfactory kappa values with greater Sholinganallur wetland area etc. which will either be
accuracy achieved, indicate of higher agreement encroached or completely occupied by built-up
between the actual and predicted land uses (table 10). category (figure 15). A similar trend is observed in
Hyderabad with violations in CDP, vulnerable
Prediction for the year 2025 was performed using ecologically sensitive areas such as Musi river bed
Markovian transition estimator tool considering (i) (Malakpet), Mir Alam and Madeena guda lake bed,
 162

Journal of Geomatics Vol 11 No. 2 October 2017

Kanchan bagh, Alwal wetland area and Janakinagar for Hyderabad region. This decreasing trend in urban
wetland gets changed into built-up categories (figure landscape shape index further confirm of landscape
16). attaining a standard or regular shape with the decrease
in length of the edges. Largest urban patches were
Zone and gradient wise spatial metrics were computed observed in circles 7-15, NE, SE (Kolathur and
with 2025 predicted images to understand the spatial Vadapalani) and 7-11 SW (Poonamalle) of Chennai
patterns of urban growth and sprawl. Figures 9 - 16 and circles 9-13 NW (Kukkatpally and Jeedimetla),
depict the metric wise spatial patterns of urbanization. 11-15, 23-29 SE (Secunderabad, Ghatkesar and
Number of patches and patch density in the core city Cherlapally IDA) and 9-11 SW (Manikonda and
area (circles 1-9, Chennai and circles 1-12, Hitech city) of Hyderabad. These metrics clearly
Hyderabad) in all directions showed almost zero indicate of intensified and concentrated urban growth
values implies that the entire landscape is completely in the core city and fragmented or dispersed growth in
dominated and saturated by only one single urban peri-urban regions.
patch. For both the regions, in all directions (except
Hyderabad, SE, circles 23-35) NLSI values were Table 10: Validation statistics (Simulated and
observed to be lesser than 0.2 indicating the urban classified image)
patches are more compact, dense and has attained a
standard shape. Clumpiness values almost reaching +1 City Chennai Hyderabad
as well as aggregation index values to 100, showed KLocation 0.9058 0.8829
urban landscape maximally aggregated in both KStandard 0.8229 0.8624
regions. Urban shape index values for 2025 are less Overall Accuracy 92% 93.8%
compared to 2012/2013 for Chennai region and 2014

Figure 15: Predicted land use categories for the year 2025 – Chennai region

Figure 16: Predicted land use categories for the year 2025 – Hyderabad region
 163

Journal of Geomatics Vol 11 No. 2 October 2017

Table 11: Predicted land use statistics for the year 2025 for Chennai and Hyderabad regions
Chennai region Hyderabad region
Categories / Predicted 2025 Predicted 2025 Predicted 2025 Predicted 2025
Year with CDP without CDP with CDP without CDP
% land use % land use
Builtup 36.6 36.5 51.01 51.02
Vegetation 2.4 2.4 2.98 2.97
Water 27.9 27.8 1.98 1.98
Others 33.1 33.3 44.03 44.03

Conclusion and discussion natural resources and agrarian lands. There has been a
spurt in population and increased population density in
Poor environment, infrastructure and living conditions the urban core; this would put lot of pressure on
due to unplanned urbanization has been a major improving the accessibility of basic amenities to
concern in metropolitan cities of India. Understanding citizens both in Chennai and in Hyderabad. Social
spatio-temporal patterns of urban growth and its factor (S) and economy are two major factors that play
impacts on environment is possible with the a vital role in managing urban strata in a city and urban
availability of remote sensing data acquired through space. Social amenities as considered in the study in
space borne sensors. In the current study, land cover modelling the land use change are more concentrated
dynamics during 1991-2013 and 1989-2014 were in the city in Hyderabad, pushing growth around the
assessed through vegetation index. Chennai which had region. Whereas in Chennai social amenities are
a lush green cover of around 70.47% in 1991 present in both core city and outskirts in large number
consistently declined accounting to about 50% in have fueled the growth and would be main factors that
2013. Land use dynamics analyses during four would allow growth in coming years. Chennai being a
decades show a drastic increase of urban area by more hub of industry and Hyderabad being a hub of
than 20 times with the conversion of grazing, information sector units would provide a huge push of
agricultural and open areas. Urban area was found to economic growth near outskirts of the city and in the
be spread over 77104 ha (2013) in Chennai region and buffer zones thus all these factors showed a great
75768 ha (2014) in Hyderabad region. This influence in rampant urban sprawl and urbanization in
tremendous growth may be clearly visualized in both study regions.
industry oriented land fragments like Ponneri, Avadi,
Sriperumbudur in Chennai region and Malakpet, The simulation and prediction of land uses with
Madapur, Bollaram, Kukkatpally etc. in Hyderabad violations of CDP show of intensified urban growth
region. Higher overall accuracy values ranging from within CMDA and HMDA boundary limits.
86% - 97% (Chennai) and 87% - 94% (Hyderabad) Compared to this, with constraints of CDP
proves the consistency of land use classifications. implementation, indicate of fairly distributed built-up
Shannon’s entropy values indicate of sprawl or along highways such as Avadi, Triuchinapalli, Ponneri
dispersed growth in recent years. Spatial metrics were (NH- 4, 45 and 716 roads) in Chennai region and
used considering the area, shape and contagion Kushaiguda, Safilguda, Uppal, Ghatkesar, Katedan,
obtained through the moving window method to Serilingampally, Patancheru (NH – 5,7,9 and 202
quantify the urban built up land density. The analysis roads) in Hyderabad region on the peripheries of
also revealed that the process of densification at the metropolitan boundary zones. Urban areas are
city center (CBD: Central business district) with the observed to be increased from 77104 ha (2013) to
initiation of the process of aggregations during 2010’s. 151428 ha (2025) in Chennai and 75768 ha (2014) to
Predication of urban growth in 2025 of complex urban 175009 ha (2025) in Hyderabad. These findings aid
landscape systems was done with integrated fuzzy- policy makers in provisioning basic amenities and
AHP, cellular automation and Markov chain adequate infrastructure in rapidly urbanizing
techniques. The predicted spatial patterns of 2013 landscapes. Decline of vegetation and wetlands in the
(Chennai) and 2014 (Hyderabad) were validated by landscape will lead to instance of frequent flooding,
comparing with the actual land use show conformity traffic congestions, higher level of pollutants, water
with higher accuracies and kappa statistics. The spatial scarcity, etc. which necessitates sustainable
analyses helped in visualizing and identification of management of natural resources.
urban growth regions and assessment of impacts on
 164

Journal of Geomatics Vol 11 No. 2 October 2017

Acknowledgement July 13th – July 19th 2014, Quebec City convention

centre, Quebec, Canada.
We are grateful to (i) ISRO-IISc Space Technology
Cell, Indian Institute of Science, (ii) SERB division, Brueckner, J.K. and H. Kim (2003). Urban sprawl and
Ministry of Science and Technology, DST, the property tax. International Tax and Public Finance,
Government of India and (iii) Asia Pacific Network 10: 5–23.
(APN) (iv) Indian Institute of Technology Kharagpur
for financial and infrastructural support. We thank Census of India (2011). Government of India,
USGS Earth Resources Observation and Science http://censusindia.gov.in/
(EROS) centre for providing the environmental layers
and Global Land Cover Facility (GLCF) for providing Chang, K.T. (2006). Introduction to geographic
Landsat data. information systems. 4th edition, McGraw-Hill
Higher Education publication.
References
Cheng, J. and I. Masser (2004). Understanding spatial
Alberti, M. (2005). The effects of urban patterns on and temporal processes of urban growth: Cellular
ecosystem function. International regional Science automata modelling. Environment and Planning B:
Review, 28 (2): 168-192. Planning and Design, 31: 167–194.

Almeida, C.M., A.M.V. Monteiro, G. Camara, B.S. Deng, J.S., Ke Wang, Yang Hong and Jia G. Qi,
Soares-Filho, G.C. Cerqueira, C.L. Pennachin and M. (2009). Spatiotemporal dynamics and evolution of
Batty (2005). GIS and remote sensing as tools for the landuse change and landscape pattern in response to
simulation of urban land-use change. Int. J. Remote rapid urbanization. Landscape and Urban Planning,
Sens. 26 (4): 759–774. 92, 3–4: 187-198.

Al-sharif, A.A. and B. Pradhan (2013). Monitoring Duda, R.O., P.E. Hart and D.G. Stork (2000). Pattern
and predicting land use change in Tripoli Metropolitan classification (2nd edition). Wiley- Interscience
City using an integrated Markov chain and cellular Publication, ISBN 978-81-265-1116-7.
automata models in GIS. Arabian Journal of
Geosciences, 7 (10): 4291-4301. Dupont, V. (2005). Peri urban dynamics: Population,
habitat and environment on the peripheries of large
Barnes, K.B., J.M. Morgan, M.C. Roberge and S. Indian metropolises - A review of concepts and
Lowe (2001). Sprawl development: Its patterns, general issues. French Research Institutes in India
consequences, and measurement. A white paper, Publication.
Towson University.
Ewing, R., R. Pendall and D. Chen (2002). Measuring
Batty, M. (1998). Urban evolution on the desktop: sprawl and its impacts. Washington, DC: Smart
Simulation with the use of extended cellular automata. Growth America.
Environment and Planning, 30: 1943-1967.
Furberg, D. and Y. Ban (2012). Satellite monitoring of
Batty, M. and Y. Xie (1997). Possible urban automata. urban sprawl and assessment of its potential
Environment and Planning B, 24: 175-192. environmental impact in the greater Toronto area
between 1985 and 2005. Environmental Management,
Bharath, H.A., S. Bharath, S. Sreekantha, D.S. 50: 1068-1088.
Durgappa and T.V. Ramachandra (2012). Spatial
patterns of urbanization in Mysore: Emerging tier II He, C., N. Okada, Q. Zhang, P. Shi and J. Zhang
city in Karnataka. Proceedings of NRSC UIM 2012, (2006). Modeling urban expansion scenarios by
Hyderabad. coupling cellular automata model and system dynamic
model in Beijing, China. Appl. Geogr. 26: 323–345.
Bharath, H.A., S. Vinay and T.V. Ramachandra
(2014). Landscape dynamics modelling through Hepinstall-Cymerman, J., C. Stephan and L.R. Hutyra
integrated Markov, fuzzy-AHP and cellular automata. (2013). Urban growth patterns and growth
in the proceeding of International Geoscience and management boundaries in the Central Puget Sound,
Remote Sensing Symposium (IEEE IGARSS 2014), Washington, 1986–2007. Urban Ecosystems, 16:
109–129.
 165

Journal of Geomatics Vol 11 No. 2 October 2017

Herold, M., H. Couclelis and K.C. Clarke (2005). The Lillesand, T.M., R.W. Kiefer and J.W. Chipman
role of spatial metrics in the analysis and modeling of (2012). Remote sensing and image interpretation. 6th
urban change. Computers, Environment, and Urban edition, John Wiley and Sons publication.
Systems, 29: 339–369.
Loibl, W. and T. Toetzer (2003). Modeling growth and
Hossein, S.M. and H. Marco (2013). Spatiotemporal densification processes in suburban regions-
urbanization processes in the megacity of Mumbai, simulation of landscape transition with spatial agents.
India: A Markov chains-cellular automata urban Environmental Modelling & Software, 18: 553–563
growth model. Applied Geography 40: 140-149
McCuen, R.H. and R.E. Beighley (2003). Seasonal
Hotton, D. (2001). Deer. In Maryland Game Program flow frequency analysis. Journal of Hydrology, 279:
Annual Report 2000-2001. Annapolis, MD: Maryland 43–56.
Department of Natural Resources: 11-29.
McGarigal, K. and B. Marks (1995). Fragstats –
Irwin, E. and N. Bockstael (2007). The evolution of Spatial pattern analysis program for quantifying
urban sprawl: Evidence of spatial heterogeneity and landscape structure. Forest Science Department,
increasing land fragmentation. Proceedings of the Oregon State University.
National Academy of Sciences of the USA, 104(52):
20672-20677. Mukunda, D.B., N.B. Santosh, N.P. Sudhindra, R.B.
Priti and S.R. Partha (2012). Modelling and analyzing
Jat, M.K., P.K. Garg and D. Khare (2008). Monitoring the watershed dynamics using Cellular Automata
and modelling of urban sprawl using remote sensing (CA)–Markov model – A geo-information based
and GIS techniques. International Journal of Applied approach. J. Earth Syst. Sci. 121: 1011–1024
Earth Observation and Geoinformation, 10: 26-43.
O’Neill, R.V., J.R. Krummel, R.H. Gardner, G.
Khwanruthai, B. and M. Yuji (2011). Site suitability Sugihara, B. Jackson, D.L. DeAngelis, B.T. Milne,
evaluation for ecotourism using GIS and AHP: A case M.G. Turner, B. Zygmunt, S.W. Christensen, V.H.
study of Surat Thani province, Thailand. Procedia Dale and R.L. Graham (1988). Indices of landscape
Social and Behavioral Sciences 21: 269–278. pattern. Landscape Ecology 1: 153–162

Kong, F. and N. Nakagoshi (2006). Spatial-temporal Ottensmann, J.R. (1977). Urban Sprawl, Land Values
gradient analysis of urban green spaces in Jinan, and the Density of Development. Land Economics, 53
China. Landscape and Urban Planning 78: 147–164. (4): 389-400.

Kowarik, I. (1990). Some responses of flora and Praveen, S., S. Kabiraj and T. Bina (2013).
vegetation to urbanization in central Europe. In: Application of a hybrid Cellular Automaton – Markov
Sukopp, H., Hejny, S., Kowarik, I. (Eds.), Urban (CA-Markov) model in land-use change prediction: A
Ecology: Plants and Plant Communities in Urban case study of Saddle creek drainage basin, Florida.
Environments. SPB Academic Publishing B.V., The Applied Ecology and Environmental Sciences 1, no. 6:
Hague, The Netherlands: 45–74. 126-132.

Lau, K.H. and B.H. Kam (2005). A cellular automata Pryor, R.J. (1968). Defining the rural-urban fringe.
model for urban land-use simulation. Environment and Social Forces, University of North Carolina Press, 47:
Planning B: Planning and Design 32: 247–263. 202–215.

Li, X., Q. Yang and X. Liu (2008). Discovering and Ramachandra, T.V., H.A. Bharath and D.S. Durgappa
evaluating urban signatures for simulating compact (2012). Insights to urban dynamics through landscape
development using cellular automata. Landsc. Urban spatial pattern analysis. International Journal of
Plann. 86: 177–186. Applied Earth Observation and Geoinformation 18:
329-343.
Li, Y., X. Zhu, X. Sun and F. Wang (2010). Landscape
effects of environmental impact on bay-area wetlands Ramachandra, T.V., H.A. Bharath and S. Vinay
under rapid urban expansion and development policy: (2013). Land use land cover dynamics in a rapidly
A case study of Lianyungang, China. Landscape and urbanising landscape. SCIT Journal, 13: 1-12.
Urban Planning 94(3): 218-227
 166

Journal of Geomatics Vol 11 No. 2 October 2017

Ramachandra, T.V. and U. Kumar (2009). Land United Nations (UN), (2012). Millennium
surface temperature with land cover dynamics: Multi development goals indicators, database.
resolution, spatio-temporal data analysis of Greater
Bangalore. International Journal of Geoinformatics, Vishwanatha, Bhat, H.A. Bharath and T.V.
5(3), 64-75. Ramachandra (2015). Spatial patterns of urban growth
with globalization in India’s Silicon Valley.
Ramachandra, T.V. and U. Kumar (2010). Greater Proceedings of National Conference on Open Source
Bangalore: Emerging heat island. GIS for GIS: Opportunities and Challenges, IIT (BHU),
Development, 14(1): 86-104. Varanasi. October 9-10, 2015.

Ramachandra, T.V., H.A. Bharath and B. Barik Wagner, D.F. (1997). Cellular automata and
(2014a). Urbanisation pattern of incipient mega region geographic information systems. Environment and
in India. Tema. Journal of Land Use, Mobility and Planning B: Planning and Design 2: 219–234.
Environment, 7(1): 83-100.
White, R. and G. Engelen (2000). High resolution
Ramachandra, T.V., H.A. Bharath and M.V. modelling of the spatial dynamics of urban and
Sowmyashree (2014). Urban structure in Kolkata: regional systems. Comput. Environ. Urban Syst. 24:
Metrics and modeling through geo-informatics. 383–400.
Applied Geomatics, 6(4): 229-244.
Whittaker, R.H. (1975). Communities and
Ramachandra, T.V., H.A. Bharath and M.V. ecosystems. MacMillan, New York.
Sowmyashree (2015). Monitoring urbanization and its
implications in a mega city from space: Wolfram, S. (2002). A new kind of science. Wolfram
Spatiotemporal patterns and its indicators. Journal of Media, Canada.
Environmental Management, 148, 67-91.
World Urban Prospects - Highlights, 2014 revision.
Saaty, T.L. (1980). The analytical hierarchy process: Department of Economic and Social Affairs. United
Planning, priority setting, resource allocation. Nations, New York-2014.
McGraw-Hill publication, New York.
Wu, J. (2006). Environmental amenities, urban sprawl,
Sadhana, J., K. Divyani, R. Ram Mohan and B. and community characteristics. Journal of
Wietske (2011). Spatial metrics to analyze the impact Environmental Economics and Management, 52(2):
of regional factors on pattern of urbanisation in 527-547.
Gurgaon, India. Journal of Indian Soc. Remote Sens.,
39(2): 203–212. Xian, G., M. Crane and J. Su (2007). An analysis of
urban development and its environmental impact on
Smart Cities - Mission Statement & Guidelines, 2015. the Tampa Bay watershed. Journal of Environmental
Ministry of Urban Development, Government of Management 85(4): 965-976.
India. http://smartcities.gov.in/ accessed on
07.04.2015 Yang, X. and C.P. Lo (2003). Modelling urban growth
and landscape changes in the Atlanta metropolitan
Stevens, D. and S. Dragicevic (2007). A GIS-based area. International Journal of Geographical
irregular cellular automata model of land-use change. Information Science, 17: 463–488.
Environment and Planning B: Planning and Design 34:
708–724. Yuzer, M.A. (2004). Growth estimations in settlement
planning using a Land Use Cellular Automata Model
Sukopp, H. (1998). Urban ecology—Scientific and (LUCAM). Eur. Plann. Stud. 12 (4): 551–561.
practical aspects. In: Breuste, J., Feldmann, H.,
Uhlmann, O. (Eds.), Urban Ecology. Springer, Berlin, Zhu, W. and M.M. Carreiro (1999). Chemoautotrophic
3–16. nitrification in acidic forest soils along an urban-to-
rural transect. Soil Biol. Biochem., 129: 1091–1100.
Taubenbock, H., M. Wegmann, A. Roth, H. Mehl and
S. Dech (2009). Urbanization in India-Spatiotemporal
analysis using remote sensing data. Computers,
Environment and Urban Systems, Volume 33, Issue 3:
179–188.

View publication stats