Towards an understanding of morphogenesis in metropolitan street-networks

We propose a GIS-based method to enable the understanding of how global street-network properties emerge from the temporal accumulation of individual street-network increments. The method entails the adoption of quantitative descriptions of individual street-patterns and of classification algorithms, in order to obtain numerically defined typomorphologies, which may then be statistically associated with the numerical outputs of street-network analysis. We apply the method to the case of Oporto Metropolitan Area, whose development we observed over 60 years. We isolate each increment of development entailing the creation of new streets (4208 objects), we quantify the morphology of their street-layouts, and we classify them into typomorphologies with clustering techniques. Through the investigation of the temporal and spatial frequencies of those typomorphologies, we assess their impacts on the street-networks of a set of selected civil-parishes of the metropolitan region, demonstrating that different typomorphological frequencies result in also different global street-network properties. We conclude by summarising the advantages of the method to generic urban morphological research and by suggesting that it may also contribute to inform bottom-up metropolitan spatial planning.