GISdevelopment.net ---> AARS ---> ACRS 1989 ---> Poster Session 2

Monitoring urban Sprawl in Kanpur Metropolis, India, using Multidate Satellite data

S.K. Shukla & Ms Suchandra Ray
Remote Sensing Application Centre,
Uttar Pradesh, Sector-G,
Janakipuram, Kursi Road,
Lucknow-226007, India.



Abstract
During the last two decades, rapid growth of certain large cities in India has led to extensive urban sprawl. For the management of these urban and periurban areas, accurate and updated Landuse maps are required. Urban Landuse/Lanedcover categories and their spatial extensions can easily be temporally monitored using multidate satellite data.

Kanpur metropolis, the industrial hub of the state of Uttar Pradesh with a population of 1.68 million and an urban area of approx. 300 sq km is the ninth largest city in the country,. In the decennium 1971-81, a population growth of about 3.2% per annum has been registered. The paper records the result of the study undertaken to monitor the spatial extension of urban growth of Kanpur city during fourteen years suing multidate satellite images.

The urban landuse/landcover status was derived from the Landsat MSS image of December 12,1975 and the information was transferred on Survey of India topographical maps to produce the threshold for comparison with subsequent date satellite images. The urban sprawl was thereafter monitored from Landsat TM image of March 2, 1986 and IRS-1A LISS-2 image of January 15, 1989 in order to assess the extent of increase in the growth of city. The comparison revealed that between 1975-89 the growth has taken place @ 6.4% per annum. Though the increase in urban sprawl has been noted to wardsa east and west of the metropolis also, but due to rapid expansion of industries in southern part of city, greater urban growth has taken place there.

Introduction
The inexorable growth of urban population in the coming decades presents a rare opportunity o assure an acceptable standard of living for the urban people (DdijKgraaf, 1987). Despite extensive efforts to improve the living conditions and basic amenities of the urban population, the number of badly sheltered lives in villages, still the urban growth has reached phenomenal proportions and continued to present almost insuramountable problems for the urban planners (Mishra, 1987).

The city of Kanpur is known to have undergone extremely fast areal expansion in recent years due to further industrialisation resulting in an unprecedented population growth @ 3.2% per annum during the last decade (Census of India, 1981) The rapid changes in the urban landuse and their expansion need to be monitored frequently for effective physical planning of the urban sprawl and to check uncontrolled growth of the city (Kalubnarme, et al., 1985). The present investigations were undertaken with the obejctive of mapping the urban landuse and monitoring the urban sprawl during the period of 1975-89 alongwith the dominant growth pattern of Kanpur metropolis using satellite remote sensing techniques.

Study Area
Kanpur metropolis, with a population of 1.68 million having an urban area of about 300 sq km is the ninth largest city in the country. It is the industrial hub of the state of Uttar Pradesh an is situated on the bank o river Ganga. It sprawls between latitude 26020’N and 80010’ to 80030’E (Fig.1)


Fig .1

Data Used & Methodology
The following data were used in the assessment of landuse/landcover categories and in monitoring changes in the urban sprawl:
  • Survey of India topographical maps on 1:50,000 scale (published 1977, surveyed 1974-75).
  • Landsate MSS FCC Bands 4,5,7, (Path & Row No. 155-042) of December 12, 1975 on 1:250,000 scale.
  • Enlarged Landsat TM FCC Bands 2,3,4 (Path & Row No. 144-042) of March 2, 1986 on 1:62,500 scale.
  • IRS-1A LISSs-2 FCC bands 2,3,4 (Path & Row No. 26-49) of January 15, 1989 on 1:125,000 scale.
The Survey of India topographical maps on 1:50,000 scale of kanpur City were used as the threshold base maps (Fig.2) from which the relevant ground truth information was extracted for interpretation of Landsat MSS FCC of December 2, 1975 since there was correspondence of the two data sets. The visual interpretation of Landsat MSS FCC was carried out with a view to identifying the different mappable categories of urban landuse/landcover following the usual elements of interpretation. Using this as a data base , the information on landuse/landcover categories extracted by the visual interpretation of Landsat TM FCC of March 2,1986 and IRS-1A FCC of January 15, 1989 were transferred on the base map. A series of visits were also made in the city area for verification and collection of ground truth. The information on the urban landuse patterns was integrated and urban landuse/landcover maps were prepared for the different dates. The spatial extent of the city growth was thereafter computed and growth pattern analysed for town planning purposes.


Fig .2

Results & Discussions

1. Urban Landuse Categorisation:
Based on the visual interpretation of Landsat MSS FCC of December 12, 1975, Enlarged TM Data of March 2, 1986 and IRS-1A Data of January 15, 1989, it has been possible to compare the extent of extraction of information related to urban landuse categories. The following table gives the comparison of the information interpreted from the different satellite data used in the present study. It is evident from the Table-1 that only five categories of urban landuse could be delineated from Lnadsat MSS data of 1975. While 28 urban landuse classes could be extracted from the Landsat TM, and 26 categories were identified from IRS-1A data.

From the Landsat MSS imagery, discrimination between the different canal, and categories like water body, zoological garden comprising dense forest, and the cantonment area (Fig.3). On the TM and IRS imagery, residential and commercial areas could not be discretely separated out because in Kanpur metropolis, like many other metropolis in developing countries, the commercial areas are inter-mixed with residential ones (Fig. 4&5). However, on the basis of the compactness of built-up areas, three categories of residential/commercial areas viz., high medium and low density could be identified. Among the industrial areas, only a few major industrial complexes could be identified because of their typical features like ash pond, cooling thank and characteristic pattern of the built-up area. Small industrial units could not be separated. Among the institute of Technology, Medical College could be located and mapped. Among recreational areas, most of the large Parks, play grounds, stadium and zoological garden could be identified. The urban infrastructure like roads, railway lines, airports were easily identified and delineated. The central bus depot due to its large size and better contrast was also identified. The extensive farm/cropland area, a few or-chards, and forest areas were also delineated. Water bodies like river, canal, ponds were, however, very easily identified. It was possible to identify and delineate two categories of wastelands viz., waterlogged and salt affected at the fringe of the built up area of the metropolis. Because of their conspicuous location, contrast and association, one major religious place of worship and one hospital could be identified. It was not possible to identify and delineate many smaller hospitals and religious places. The military cantonment area, vacant land and garbage dumping ground could easily be demarcated. The graveyard and jail could be additionally identified on Landsat TM imagery by intergrating ground truth information.


Fig .3


Table 1: Landuse categories interpreted from different satellite data:
Landsat MSS Enlarged Landsat TM IRS-1A LISS-2
Rail/Road/Canal RESI/COMMERCIAL ESI/COMMERCIAL
01. High Density 01. High Density
Zoo/Museum 02. Medium Density 02. Medium Density
03. Low Density 03. Low Density
Cantonment INDUSTRIAL ( Major) NDUSTRIAL ( Major)
(Vegetative RECREATIONAL RECREATIONAL
dominant) 01. Parks 01. Parks
02. Playground /Stadium 02. Playground/Stadium
Water bodies 03. Zoo/Meseum 03. Zoo/Museum
Urban Sprawl  
URBAN INFRASTRUCTURE URBAN INFRASTRUTURE
01. Road (major) 01. Road (major)
02. Railway line/yard 02. Railway line/yard
03. Airport/Air Strip 03. Airport/Air Strip
04. Bus Depot (major) 04. Bus Depot. ( major)
AGREECULTURE LAND AGREECULTURE LAND
01. Cropland 01. Cropland
02. Orchards 02. Orchards
FOREST/VEGETATION FOREST/VEGETATION
WATER BODIES WATER BODIES
01. River/sand 01. River/sand
02. Canal 02. Canal
03. Pond/tank 03. Pond/tank
WASTELANDS WASTELANDS
01. Waterlogged 01. Waterlogged
02. Salt affected 02. Salt affected
MISCELLANEOUS MISCELLANEOUS
01. Religious place 01. Religious place
02. Cantonment Area 02. Cantonment Area
03. Vacant land 03. Vacant land
04. Dumping ground 04. Dumping ground
05. Ash pond 05. Ash pond
06. Hospital 06. Hospital
07. Graveyard 07. Smoke plume
08. Jail

2. Analysis of the Dynamics of Urban Sprawl:
For monitoring the growth of the Kanpur metropolis the Survey of India topographical maps have been used as the data base. The urban extent of kanpur City in 1974-75, as computed from the Survey of India maps was 115.28 sq km (Table 2.) The urban, as computed from Landsat MSS FCC of 1975 is 114.22 sq km which compares favorably with the Survey of India maps. It shows that Landsat MSS imagery provides an accurate boundary of urban sprawl.

Table 2: Urban growth of Kanpur in different years:
S. No. Year Data Base Urban area (sq km) Urban growth (sq km) %of urban growth per (base 1974-75) % growth annum
1. 1974-75 Topographical 115.28 - -  
2. 1975 Landsat MSS 114.22 - -  
3. 1986 Landsat TM 170.59 55.31 47.97 4.3
4. 1989 IRS-IA 211.40 40.81 23.92 7.9
5. Growth between 1986-89   40.81 23.92 7.9

The urban extent of the metropolis as interpreted from Landsat TM imagery of 1986 is 170.59 sq km, while from IRS-1A data of 1989, it is 211, 40 sq km. Thus, the urban growth recorded during the 11 years period of 1975-86 is 55-31 sq km which amounts to an increase of built-up area by 47.3%. The average annual urban growth rate during the above period is about 4.3%. The rate of growth between 1986 to 1989 is 40.81 sq km which amounts to an increase in growth by 23.92% within three years giving an average annual growth rate of 7.9% per annum.

the average annual growth rate of the metropolis is found to be 4.3%. during 1975-86, but has shown an increase of 7.9% during 1986-89. The increase in the built-up area has obviously led to adverse effects on other landuse categories. The co-relation with the increase in population , as evidenced from the decennial census report, shall be attempted subsequently and the annual urban growth shall be monitored using satellite data along with assessment of the consequent damage to other landuses. It has been noticed that the urban sprawl has extended along the east-west transportation network. Furthermore, specific growth has been identified in the south of city.



Fig .4



Fig .5

Conclusions
The study of Landsat MSS, Landsat TM and IRS-1A data of Kanpur metropolis of different dates has shown that Landsat MSS data provided rudimentary information of the urban landuse categories, the TM and IRS-1A FCC images provided information on the broad urban landuse categories which are of relevance in urban planning. thus, a quick inventory of broad urban landuses can be obtained very easily for the desired specific period using either landsat TM or IRS-1A, LISs-2, images in visual interpretation mode. For detailed urban landuse classes, it would be desirable to use SPOT data.

Satellite remote sensing technique provides an effective system of temporal monitoring of urbanisation with consequent depletion of other natural resources in the immediate environs of big cities and detropolis. It will provide inputs in the urban landuse planning of large cities and metropolis with a futuristic trend based on he past dynamic observations.

Acknowledgements
The author would like to thank the Director, Remote Sensing Applications Centre, Uttar Pradesh, Lucknow, India, for permitting to submit the paper at the 10th Asian Conference of Remote Sensing.

References
  • Census of India, 181. Registrar Gen. and Census Commr., New Delhi, India.
  • Cor DijKgraaf (1987), The Urban Planet, ITC J.3: pp. 194-96.
  • Kalubarme, M.H., Baldev Sahai and SU Avarani (1985), Mapping and change detection in urban landuse of Surat City, Proceedings of Vi ACRS, pp. 103-8.
  • Mishra, (1987), Information and Manpower requirements for urban and Regional Planning in India, ITC J.3: pp. 206-10.