Overview |
Urban Sprawl |
Fringe Area Development |
Urban Agglomeration |
Emerging Technologies |
Relevant Links
Metropolitan Spatial Data Infrastructure Empowering Government Authorities Through Modern Spatial Tools and Techniques
2.3. Satellite imagery preparation
Remotely sensed images for the project came from three sources:
- SPOT 5: The image from this French satellite represented one scene covering an area of 3600 km2 with a resolution of 5 m acquired on the 6th of March 2003. It was generated by merging a multispectral and a panchromatic SPOT 5 image taken on the same date. This image was instrumental in giving the urban planners an immediate, global view of the city. It gave them a preliminary understanding of some of the stakes involved at the level of the city. The SPOT 5 satellite image was used to create the Land Occupation Mode following the CORINE Landcover nomenclature.
- QuickBird: A series of 19 high-resolution images acquired by this satellite between February and March of 2002 were mosaiced together to cover 1500 km2 with a resolution of 60 cm. These images were orthorectified with 497 Survey of India ground control points. The output of this process was a single image which became the georeferenced raster image base for the project.
- ASTER: The Advanced Spaceborne Thermal Emission and Reflection Radiometer satellite provided a pair of 30 m resolution, covering an area of 12600 km2, which were used to generate a digital elevation model for Bangalore.
2.4. Ground control points of Survey of India
The Survey of India has covered the city of Bangalore with 600 survey points. The latitude and longitude of each of these points is accurately known in the UTM zone 43, WGS 84 coordinate system.
After verification, only 508 points were found on the ground. A few discrepancies were found between lat/long values. Only those with an offset less than 23 cm were made use of in the project. 497 points fulfilled this condition. These points were the official georeferenced vector base for the project.
These points were essential in orthorectifying the QuickBird satellite image as well as in calculating the shift on each of the NRSA tiles mentioned further on in this paper.
2.5. Digital base map creation
The underlying objective was to create a base map that would be made available to the city’s stakeholders like the water supply and sewerage board (BWSSB), the electricity board (BESCOM). They can then add layers that are relevant to their particular requirements. Basically to create a ‘pizza base’ over which each stakeholder can add its own ‘toppings’ according to its needs. Before the advent of the MSDI project, the city of Bangalore did not possess a standardized GIS enabled base map. Most maps present with agencies like the BDA were in the form of engineering drawings of proposed or recently formed layouts often on paper or in a digital CAD format.
In 1998, a decision was taken by a number of stakeholders, the BDA included, to carry out a large scale aerial photo and mapping campaign. The National Remote Sensing Agency was designated to carry out this mission which it began towards the end of the same year.
The photographs arising from this campaign were distributed to 12 companies for digitization at 1:2000 scale.
The end result was a series of 2 km2 tiles which when put together covered an area of 1490 km2 represented in 514 layers.

2. Themes and their names
Theoretically this work could have constituted an excellent base map. Unfortunately, due to a number of reasons, this was not the case. The resulting CAD tiles were of inconsistent quality. A few of the major issues were:
- The tiles were not edge matched
- No standard nomenclatures for layers
- Buildings often represented as unclosed polylines.
Though the aerial photo campaign itself was completed quickly, the time taken to finish the digitization took extremely long. The last of the tiles arrived close to 5 years after the campaign was carried out. Rather than waste the data and start from scratch, it was decided to clean up the digitized data.
The first step was to reduce the number of layers. Since each company involved in the digitization process had its own specifications, methodologies and naming conventions, when the tiles were put together, a particular theme could be represented under numerous names. According to the original NRSA specifications, only 138 layers were supposed to be present in the output DWG files.
From the initial 514 layers spread over the 745 tiles, duplicated layers with different names were standardized.
The next stage involved the edge matching of data within each tile as well as between tiles. Since layers like buildings, which are normally represented as polygons, were digitized as polylines which were not always closed, this task represented a time consuming challenge.
For the cleaning up of the tiles, Survey of India’s (SOI) ground control points for the city as well as the QuickBird image were used. Out of 600 ground control points, 106 were checked on the ground, and then digitized on the NRSA digital data according to their exact location, taking into consideration their immediate neighbourhood. In this way, it was possible to calculate the offset between the NRSA coordinates and the coordinates provided by the SOI.
The main results were:
- X offsets range from -26 m to +14.7 m, while Y offsets range from -21 m to +19.2 m.
- Offset distance range from 0 to 29 meters, averaging 9 meters
- These distortions are not homogeneous throughout the Bangalore area
An in-house program was then run to move each tile according to the shift calculations made earlier.
Once this was done, with the QuickBird in the background, the edges of each of the tiles were visually matched by digitisers. Buildings were converted from polylines that were often open to closed polygons. Similarly, road blocks were also created as polygons.
The creation of the base map for Bangalore followed the simple rules listed below:
- A building must be represented by a closed polygon
- A road block is a polygon that is bounded on all sides either by roads, rivers/streams or railway lines.
- Every building must be represented within a road block
- Topology rules were set up to prevent a building, road centerline, stream or railway line from crossing a road block
It must be remembered that the data was five years old when it was finally converted into a seamless GIS-enabled base map. During this time, the city had undergone a lot of development especially around the outskirts of the city. These changes were digitized directly from the QuickBird image.
The task of creating the GIS-enabled base map took approximately 8 months.
2.6. Creation of the Land Occupation Mode (LOM)
This theme was developed through a visual interpretation of the SPOT 5 satellite image for 1300 km2. The LOM, at a scale of 1:25000, comprises 34 categories developed on the CORINE (COordination of INformation on Environment) Land Cover (CLC) nomenclature. The minimum polygon size for this task was fixed at 5000 m2. The CORINE Land Cover nomenclature, initiated by the European Environment Agency in 1985 is officially divided into 3 levels. For the context of Bangalore, a 4th sub level was added. The categories used for the creation of the LOM are given in Annexure I.

3. Land Occupation Mode for Bangalore
The interpretation began with a series of ground truthing surveys based on pre-selected areas of doubt. Once these zones were clarified and the CLC categories finalised, the actual process of interpretation started.
After the interpretation phase, a second round of ground truthing was carried out on randomly selected sites scattered over the project area. This was done to crosscheck the image classification.
The LOM provided the urban planners and the BDA with a quick overview of the existing situation on the ground. It was the first time that an individual with no remote sensing skills could grab a quick understanding of the city. The entire process took 6 months to complete.

4. Sample A3 printout of the base map after survey
2.7. Surveys
One of the most important and challenging stages of the creation of the DUSR was the Survey phase. When data was questionable, or not available from a stakeholder, it was imperative that it be collected off the ground.
A good example of a survey carried out for the project would be the land use survey. With the base map ready, it was necessary to map out the land use of Bangalore at the sub-road block level. Over 55% (690 km2) of the project area covering the entire central zones of the city was put under survey. Field personnel, armed with portions of the base map and QuickBird on A3 sheets, would divide road blocks into categories defined under the Karnataka Town and Country Planning Act (KTCP Act).
After the surveyors’ work, the A3 sheets were then used to update the database by subdividing the roadblocks into land use blocks that conformed to the Government of Karnataka’s requirement.
The land use for the remaining areas around the edges of the city were gathered from the LOM since these zones had not undergone much development and the classifications were fairly stable. This stage took 5 months to complete.

5. Portion of the core area after the land use survey
At the end of this phase, it was possible to make a comparison between the existing Comprehensive Development Plan (CDP) for 1995 document present with the BDA and the existing situation. It allowed the urban planners to gauge the extent at which the previous CDP had been followed as well as understand the possible pitfalls and challenges that they might meet when designing the new one.
Some of the surveys carried out are listed in the table below
| Survey | Area/distance covered | Manpower | Duration (months) |
| Railway right of way | 230 kms | 2 | 1 |
| Landmarks | 700 km2 | 3 | 3 |
| Photographs | 700 km2 | 3 | 3 |
| High Tension power lines | 450 kms | 2 | 3 |
| Drains | 815 kms | 2 | 1.5 |
Not all surveys involved setting out on field. Tasks such as gathering educational information on the number of teachers per school, as well as information related to the medical situation like the number of beds per hospital were gathered through telephonic interviews.
The Digital Urban Spatial Repository (DUSR) at a glance:
A set of 12 geo-databases that comprise:
- 1:2000 scale base map covering 1307 km2
- 553 villages
- 55,000 parcels
- 650 000 buildings
- 15 500 km of roads
- 230 km of railway
- 2 546 places of worship
- 100 691 manholes
- 330 903 water supply consumer connections
- 4 010 km of water pipelines
- 3 245 km of sewage lines
- 815 km of drains
- 450 km of High tension power lines
- 90 000 parcels of existing land use
- 15 000 landmarks
- 400 layouts
- 1991 and 2001 Census information
- Low (30 m), mid (5m) and high resolution (0.60m) satellite images
- Historical maps of the city
- Attribute data from over 30 stakeholders attached to various geographical objects
The DUSR provides the main public stakeholders with an up-to-date urban large scale digital map (1: 2000) that is fully GIS enabled. It acts as the unique spatial platform which displays and consolidates all ground and underground urban information. As such, it plays a key role in spatially coordinating the various urban projects and in making them more coherent: for example, the impact of the metro on land use, the implementation of a water network consistent with road maintenance work, monitoring of layout development with respect to the approved plan, tracking of diseases in relation with water quality etc.
3. Revision of the Master Plan (MP)
The Bangalore Development Authority has been recognised as the Local Planning Authority for the Bangalore Metropolitan Area (BMA) covering 1307km2. Under the purview of its tasks, it was necessary for the BDA carry out a revision of the MP for the year 2015. One of the main goals of the MSDI project is to achieve this task by making use of state-of-the-art urban planning and spatial decision support tools.
The revision of the Master Plan (MP) followed an integrated approach which comprised of the following stages:
- City Diagnosis
- Urban Strategies Design
- Design of the Master Plan and the Proposed Land Use Map

6. Administrative boundaries

7. Urban Sprawl through the ages

8. Another example of a thematic map generated for the project
3.1 City Diagnosis
Prior to any planning process, a thorough understanding of the current situation of the city and the stakes involved was necessary. The fact that Bangalore is a bustling city with a rapidly growing population that aims at becoming a world class metropolis is well known. In view of this, it was vital to identify and, if possible, quantify the existing and future pressure points that could hamper its growth and evolution.
The DUSR played a crucial role at this stage. Not only did it form the basis for the visualisation of raw data like satellite images and information from the various stakeholders; its true power was brought out through the analysis and thematic interpretation of this data. Basic maps like those showing the urban sprawl, populations density etc, as well as more complex ones like the Synthetic Social Index maps gave the urban planners a method to gauge the health of the city.