Determination of the build up area development in the Greater Municipality of Istanbul by space images

The population growth in the Greater Istanbul Municipal area is causing several changes of the build up parts. Often these changes are not legal. Especially in the protected water catchments areas illegal housing is critical and are not accepted by the municipality. Due to the legal situation a close to on-line detection of new buildings is necessary. In addition information for urban planning is required. For these reasons the municipality area of 5378 km² is investigated regularly by means of satellite imagery. In addition also the surrounding Marmara region is inspected because of its influence to the development of the Greater Municipality area.
The analysis of the changes is based on a time serious of Landsat images, giving an overview of the land classes, supported by a complete coverage of the Marmara region by SPOT 5 supermode pansharpened images and IKONOS scenes covering the municipality area, taken every 3 months.
The Landsat images are available for 5 year time interval; they are used for a classification of the main feature classes like forest, agriculture and build up areas. This allows an overview of the urbanisation of the Marmara region, not only limited to Istanbul. In general build up areas are growing, while forest and agriculture parts are shrinking. The location of the changes is important information for urban planning applications. The changes have to be determined by means of remote sensing because a ground survey is too expensive and in most cases the changes are not based on allowances by the municipalities.
The complete coverage of the region by pansharpened SPOT 5 supermode images, with 2.5m GSD, was used for updating the existing topographic maps as required base for the urban planning. The automatic classification of the SPOT 5 images gave more detailed information about forest and agriculture areas, while a manual update could not be avoided in the build up parts.
Based on pansharpened IKONOS images, shape files including all buildings of the Greater Istanbul Municipality area have been generated. For controlling illegal housing and as base for urban planning a time serious of IKONOS scenes with 3 months repetition rate has been used starting in June 2005. With approximately 10 scenes the whole area of Greater Istanbul Municipality is covered. In the case of partial cloud coverage additional images are ordered. The ground resolution of 1m did not allow an automatic identification of new buildings. Often the buildings are too small and partially they are under construction or in preparation. Even not solid buildings, like green houses with plastic covers, had to be mapped. Over 3 months time interval the radiometric changes caused by different sun elevation, changes of the vegetation and the atmosphere are fundamental so that automatic change detection is not possible with very high resolution images. Especially, due to the different length of shadows a reliable result can not be achieved by the automatic mode.
As base for the manual detection of building changes IKONOS ortho-image mosaics are used. They are overlaid by shape files of the update period before. Any detectable change is checked against the images of preceding period to be sure about the start of the construction. Important changes are checked also by a second operator. Not only new buildings, also the preparations and obvious changes of the building heights are recorded. In the critical water catchments areas a field check is made for verification.
The manual periodic update of the shape files takes approximately 30 operator days for the area of Greater Istanbul Municipality, which is distributed to a group of ten operators. Thus, a fast reaction to not acceptable illegal constructions especially in the protected areas can be guaranteed. Around the water basins for the drinking water supply in the first zone no buildings are allowed, while in the neighboured zones several restrictions for guaranteeing the water quality exist.
The whole procedure of information acquisition by automatic classification and manual inspection, and periodic update is operational and only possible in using high resolution space images.