46th Congress of the European Regional Science Association: "Enlargement, Southern Europe and the Mediterranean", August 30th - September 3rd, 2006, Volos, Greece
The application of fractal geometry at the analysis of the urban development patterns has been widely investigated during the last two decades, providing further evidence that cities are complex and emergent structures . The fractal dimension of an urban area is an extremely useful indicator of the urban spatial structure and it's transformation through time. Various models and algorithms are used to calculate the fractal dimension, such as the Box-Counting Method and the Radial Analysis . The paper concentrates on the urbanization at the edges of the city, the outskirts of the metropolitan areas, which can be considered the examples par excellence of complex, fractal-like urban structures, revealing at the same time dynamic processes of growth and transformation. The basic models of fractal analysis are presented, with the focus on the interpretation of the calculated values and the evaluation of the observed urban patterns: Fractal dimensions can be used as indicators of urban sprawl, of the degree of fragmentation of the urban landscape and of the transition from monocentric to polycentric structures. A further exploration of the above notions is based on the application of the models on an area at the outskirts of Thessaloniki, Greece. The calculation of the fractal dimensions is based on the logarithmic expression of the relationships N(l) = al-D, Í(R) = bRD, where N presents the number of cells that are developed, l the size of the grid that is used for the calculation, D the fractal dimension, R the distance from and appropriately defined center and a,b are constants. The results are displayed by the presentation of the scatter diagrams and the linear regression between the calculated values. The important changes observed during the last decades at the urban spatial structure of the area under investigation are being evaluated by the quantitative methods provided by the fractal analysis. Attention is given to the divergence of the values across space and scale and through time. As a conclusion the need for further investigation of the models is being noted. The correlation of the fractal analysis with other methods and indices used to quantify the configuration and composition of the urban areas, and also with more general socioeconomic data, should be regarded as a promising field of further research.