Home Articles Measuring urban sprawl: A case study of Hyderabad

Measuring urban sprawl: A case study of Hyderabad

K. Madhavi Lata, Dr. V. Krishna Prasad, Dr. K. V. S. Badarinath, Dr. V. Raghavaswamy
National Remote Sensing Agency (NRSA), Department of Space
Government of India , Hyderabad
[email protected]

C. H. Sankar Rao
Dept. of Geo-engineering and Resource Development Technology
College of Engineering, Andhra University, Visakhapatnam

Rapid urban development and increasing land use changes due to increasing population and economic growth in selected landscapes is being witnessed of late in India and other developing countries. The measurement and monitoring of these land use changes are crucial to understand land use cover dynamics over different spatial and temporal time scales for effective land management. Today, with rapid urbanization and industrialization, there is increasing pressure on land, water and environment, particularly in the big metropolitan cities. Urban sprawl may be defined as the scattering of new development on isolated tracts, separated from other areas by vacant land (Ottensmann,1977). It is also often described as leapfrog development (Gordon and Richardson, 1977) as observed in all the major cities across the world. Urban sprawl has been criticised for inefficient use of land resources and energy and large scale encroachment onto the agricultural lands. There are many problems associated with fragmented conversion of agricultural land into urban use. The cities are expanding in all directions resulting in large-scale urban sprawl and changes in urban land use. The spatial pattern of such changes is clearly noticed on the urban fringes or city peripheral rural areas, than in the city centre. Inadvertently this is resulting in increase in the built up area and associated changes in the spatial urban land use patterns causing loss of productive agricultural lands, forest cover, other forms of greenery, loss in surface water bodies, depletion in ground water aquifers and increasing levels of air and water pollution. Further, it is widely agreed that fragmentation of land use is also harmful to biological conservation. There have been lot of debates on how to confine urban sprawl and conserve agricultural land resources (Bryant et al., 1982;Ewing, 1997;Daniels, 1997). There is a demand to constantly monitor such changes and understand the processes for taking effective and corrective measures towards a planned and healthy development of urban areas. In the recent times, Remote sensing data is being widely used for mapping and monitoring of urban sprawl of cities. The spatial patterns of urban sprawl over different time periods, can be systematically mapped, monitored and accurately assessed from satellite data along with conventional ground data. In the present study ‘Entropy Approach’ for studying the urban sprawl patterns of Hyderabad over different time scales has been attempted in the present study. Further, the use the GIS for quantifying the urban sprawl trends at various land use sites, viz., commercial, industrial, residential sensitive and mixed zones is also attempted.

Study Area
The study area of Hyderabad city and environs extend from 17010/-17050/N and 78010/-78050/ E. The Hyderabad Urban Development Area (HUDA) is around 1907 sq.km. The HUDA area is divided into 29 planning zones (11 zones inside municipal limits and 18 zones in the non-municipal limits or peripheral areas). The city is located around 580m above Mean Sea Level (MSL). It experiences a minimum temperature of 11.60C and a maximum of 40.50C with an average annual rainfall of 73.55 cms. The city is situated centrally between the other metropolises of Mumbai, Chennai and Bangalore and is well connected by road, rail and air.

Datasets and Methodology
In the present study, IRS-1C (LISS-III + PAN) merged data of 1999, Hyderabad area is used for studying the entropy characteristics of urban sprawl patterns and their areal estimates are derived using satellite and GIS techniques.

The satellite characteristics of IRS-1C (LISS-III +PAN) data is given in table1. Land Use/Land Cover estimates of the previous years obtained from the local district records for the periods of 1980 and 1992 (Mary and Raghavaswamy, 2000) were also referred. The zones selected correspond to different locations in the Hyderabad city dominated by Residential (Banjara Hills), Industrial (Balanagar), Commercial (Paradise, Panjagutta, Abids and Charminar) Sensitive (Zoo park) and Mixed zone (Uppal). In the present study, we use the methodology of Yeh and Li (2001) is adopted for studying the urban sprawl characteristics through Entropy approach.

Shannon’s entropy (Hn) is used to measure the degree of spatial concentration or dispersion of geophysical variable (Xi) among n zones (Theil, 1967; Thomas, 1981). It is calculated by

Where Pi is the probability or proportion of a phenomenon (variable) occurring in the ith zone , and xi is the observed value of phenomenon in the ith zone, and n is the total number of zones. The value of entropy ranges from zero to log (n). If the distribution is maximally concentrated in one zone, the lowest value, zero will be obtained. Conversely, an evenly dispersed distribution among the zones will give a maximum value of log (n). Relative entropy can be used to scale the entropy value into a value that ranges from 0 to 1.Relative entropy Hn’ is (Thomas, 1981)

Entropy can be used to indicate the degree of urban sprawl by examining whether land development in a city is dispersed or compact. If it has a large value, then it indicates occurrence of urban sprawl. The buffer function of a GIS can be used to define buffers of zone from city/town centers of roads and thus the density of land development in each of these buffer zones can be used to calculate the entropy.
Further, measurement of the difference of entropy between time t and t+1 can be used to indicate the magnitude of change of urban sprawl, i.e.,

Results and Discussion
Remote sensing data is capable of detecting and measuring a variety of elements relating to the morphology of cities, such as the amount, shape, density, textural form and spread of urban areas (Webster, 1995;Mesev etal.1995, Yeh and Li, 2001). Hyderabad occupies fifth position in terms of area and population in the country. The city has been witnessing rapid growth in urban population between 1981 and 1999 (Mary and Raghavaswamy, 2000). The urban population of the city has increased by 41.57% as against 43% of the total Andhra Pradesh state and 36% of total country. In such a scenario, studies on land use cover dynamics over the Hyderabad and its environs gain importance. Land use / cover analysis from the Remote sensing data suggested different land use cover classes viz., residential, industrial, public, semi-public, water bodies and forest (Fig. 1). The detailed land use cover estimates obtained from IRS-1C (LISS III + PAN) data for the years, 1980, 1992 and 1999 (Mary and Raghavaswamy, 2000) are given in Table 2. Analysis of the results suggests a clear increase in residential, commercial, industrial and transportation in the urban area. In the non-urban area, there is a clear reduction in agriculture area and also in vacant land suggesting the increased intensity with urbanization activities. In the present study, the areal estimates of 1999 of Hyderabad and its environs have been used for studying the urban sprawl intensity at different zones, viz., Residential (Banjara Hills), Industrial (Balanagar), Commercial (Paradise, Panjagutta, Abids, Charminar) Sensitive (Zoo park) and Mixed zone (Uppal). The classified data obtained from remote sensing has been transferred to GIS domain for performing the spatial operations. Relative entropy of two types of buffer zones viz., based on the site (site buffer) and road (road buffer) respectively, for each site has been calculated to measure the degree of urban sprawl in each of the buffer zone (Fig.2). Density of land development (%) defined as the amount of land developed divided by the land area in each of the buffer zone has been calculated following Yeh and Li approach (2001). Apart from the above analysis for the year 1999, the entropy values have been calculated for the years of 1980 and 1992 also. The results suggest that there is substantial variation in the patterns of urban sprawl among the different zones of the study area corresponding to residential, industrial, commercial, sensitive and mixed zones. The pattern of land development away from city center is slightly different from that along the roads. The detailed analysis has been carried with respect to density of land development and the road distance for each of the sites. Analysis of the results suggest that in case of residential areas, as the road distance increased, the density of land development also increased and vice versa. This relationship has been found to be high for site of Banjara hills and lowest for Paradise. The density of land development (%) declined rapidly as the distance from road increased for Zoopark and Charminar in residential sites (Fig.3). When compared to residential sites which showed a positive correlation of r2 = 0.72 with respect to density of land development, negative correlation has been noted for remaining industrial, public and forest areas. This suggests that, public amenities such as colleges, university and industrial and forest areas are not in proportion i.e., as the road distance from the city center increased, the above amenities also decreased considerably, suggesting the aggregation of the above amenities at some localities. The temporal change of spatial patterns of urban development can be easily measured from the change of entropy equation. Urban sprawl as reflected from the entropy values in different sites for Hyderabad and its environs suggest that the buffer zones 1(10m) higher entropy at all sites compared to buffer zone2 and buffer zone 3. This indicates that around the center of city, the entropy values are high suggesting highly dispersed nature of the environs. It is found that the average increase in entropy from city center is around 0.55 during the year, 1999. Results of the analysis suggested that in the Hyderabad city, higher entropy is noted for Paradise area followed by Banjara hills and Balanagar (Fig.4). Using the areal estimates of the years 1980-1999, entropy values have been calculated for urban and sub-urban areas. The entropy for urban areas increased from 0.75 in 1980 to 0.80 in 1999 compared to increase of entropy in non-urban areas from 0.55 in 1980 to 0.69 in 1999 (Fig 5). The increase in the values of entropy indicates that there is an increase in urban sprawl and development tend to be more dispersed over a period of time. This indicates rapid increase of urban sprawl. The entropy values in urban areas are much higher than sub-urban areas indicating rapid urbanization process. Overall analysis suggests that among the different zones classified as Industrial, Commercial, Sensitive and Residential, the entropy values are considerably high in the Residential zones suggesting high rate of urban sprawl over a period of time.

Table 1. Sensor
Sensor Spectral Bands Ground resolution (m) Swath (Km)
IRS-1C LISS III 0.52-0.59 23 (VNIR) 140
0.62-0.68 70 (MIR)  
IRS-1C (PAN) 0.5-0.75 5.8 70

The present study has demonstrated the utility of entropy approach to identify, measure and monitor spatio-temporal patterns of urban sprawl in Hyderabad city and its environs, by integrating with remote sensing and GIS techniques. The entropy method can be easily implemented within GIS to facilitate the measurement of urban sprawl. The study suggests that entropy is a good indicator for identifying the spatial processes in land development.

Table 2. Land Use / Land Cover estimates of Hyderabad and its environs – 1980-99 (Mary and Raghavaswamy, 2000)
Landuse/Landcover 1980 1992 1999
Area(Sq.km) % Area (Sq.km) % Area(Sq.km) %
Residential 29.38 1.65 90.01 5.32 151.99 8.98
Commertial 0.53 0.03 2.04 0.12 2.04 0.12
Industrial 35.79 2.01 41.15 2.43 66.38 3.92
Public/Semi-Public 82.27 4.62 92.85 5.49 93.00 5.49
Public Utility NA 0.17 0.01 1.49 0.09
Recreation NA NA 0.87 0.05
Transportation 14.07 0.79 18.63 1.10 16.92 1.00
Layouts/Plotted NA 56.93 3.36 71.84 4.24
TOTAL 162.04 9.10 301.87 17.83 404.53 23.89
Agriculture 749.83 42.11 556.04 32.86 524.60 31.00
Reserved forest NA 82.58 4.88 82.58 4.88
Hillock/Rocky Area NA 116.88 6.91 109.98 6.50
Water bodies 87.79 4.93 98.65 5.83 84.72 5.01
Vacant Land 780.99 43.86 536.25 31.69 485.86 28.71
TOTAL 1780.6 100.0 1692.27 100.0 1692.27 100.0

Authors are grateful to Dr. R. R. Navalgund, Director, NRSA and Prof.S.K.Bhan, Associate Director, NRSA (Applications) for the guidance and encouragement.V.Krishna Prasad and and M.Lata thank ISRO-GBP for providing the fellowship.


  • Bryant,C.R., Russwarm, L.H. and McLellan, A.G. 1982. The city’s Countryside:Land and Its Management in the Rural-urban Fringe,Longman Group Ltd,New York,N.Y.,249p.
  • Daniels,T.L.,1997.Where Does Cluster Zoning Fit in Farmland Protection?Journal of the American Planning Association? 63(1):129-137.
  • Ewing,R.,1997.Is Los Angles-Style sprawl Desirable? Journal of the American Planning Association, 63(1):107-126.
  • Gordon,P., and H. W. Richardson. 1997. Are compact Cities a Desirable Planning goal? Journal of the American planning Association,63 (1):95-106
  • Mc.Arrthur,R.H.,and e.O. Wilson, 1967.The Theory of Island Biogeogrophy,princetion University press,princetion,N.J.,203 p.
  • Mary Ashalata and Raghavaswamy. 2000. Remote sensing and GIS based study on ‘Air Quality and Land Use / Cover Hotspot Characterization in Hyderabad City, Andhra Pradesh, India. NNRMS bulletin (B)-25. Bangalore. Pp.30-36.
  • Mesev, T. V., P. A. Longley, Batty, M., and Y. Sie, 1995. Morphology from Imagery:Detecting and Measuring the Density of Urban Land Use,Environment and planning A,27:759-780.
  • MSS, Photogrammetric Engineering & Remote Sensing,49 (9):1303-1314
  • O’Connor, K. F.,F. B. Overmars and M. M. Ralston ,1990. Land Evolution for Nature Conservation,Caxton press Ltd, Wellington, New Zealand, 328 p.
  • Ottensmann,J.R.,1977,Urban Sprawl, Land Values and the Density of Development,Land Economics, 53(4):389-400.
  • Thomas, R.W., 1981.Information Statistics in Geography, GeoAbstracts, University if East Anglia, Norwich, United Kingdom 42p.
  • Webster,C.J.1995,Urban Morphological Fingerprints, Environment and planning B,22:279-297.
  • Yeh,A.G., and X.Li, Photogrammetric Engineering & Remote sensing vol.67, No.1, January 2001,pp.88-90.