Dynamic spatial analysis of urban sprawl through fractal geometry: the case of Istanbul

During the last decade, there has been a rapid increase in the research literature dealing with the measurement of urban sprawl. Density gradients, sprawl indexes, and certain simulation techniques are some of the quantitative measures that have been used in previous studies. Fractal analysis has also been applied in describing urban areas and a fractal theory of cities has been proposed. This study attempts to measure urban sprawl using an index relating density and proximity and relating this to the fractal analysis of urban sprawl using the example of Istanbul, a large city that hitherto has not been measured or characterized using such techniques. Measures of sprawl were calculated at each neighbourhood level and then integrated within an index using 'density' and 'proximity' factors. This identifies the pattern of urban sprawl during six time slices defining five periods from 1975 to 2005. The urban form of Istanbul is then quantified through fractal analysis in the given years in the context of the dynamics of urban growth. Our findings suggest that the fractal dimension of urban form is positively correlated with the urban sprawl index when urban growth is more likely to be 'concentrated'. This is in contrast to a negative relationship which is observed between the fractal dimension and the sprawl index when the urban growth pattern has a more dispersed, semilinear form.