Multitemporal analysis of rapid urban growth in Istanbul using remotely sensed data

The population of Turkey is increasing very rapidly. The population and urbanization in the large cities have increased due to a heavy immigration rate from rural areas to access job opportunities, better education, and better health facilities in the cities. While the population rate in the cities had been 57.5% in 1985, it has reached 64.9% in 2000. The European side of Istanbul has the highest immigration rate of Turkey and was selected as the study area. While the average population increase rate is 2.2% in all of Turkey, it is 3.8% in Istanbul. This study area has the most affected land use changes in Turkey because of the immigration. Remotely sensed data provide reliable information on current and temporal urban land use/cover characteristics and changes. In this study, Landsat 5 TM (1987) and Landsat 7 ETM+ (2001) satellite sensor data were used. After the classification of these images separately, eight main land use/cover classes were obtained and then reduced to five in processing. It has been shown that the urban areas increased 1000 ha yr^-1 and, by contrast, agricultural, bare soil, and forest areas have decreased. The overall classification accuracy was 83 and 80% for the years of 1987 and 2001, respectively. In this study, spatial growth of the urban site was compared with the population growth.