Shpuza; Urban Shapes and Urban Grids: A Comparative Study of Adriatic and Ionian Coastal Cities
Connectivity Skewness; hence it is possible to speculate that the affinity between the bias of shape compactness and grid integration is a result of the lack of connectivity bias in urban grids. Second, the asymmetry of integration is affected by both the symmetry of shape fragmentation and by the symmetry of shape compactness, as shown by the negative correlations Drelative Skewness vs. Integration Skewness and Fvisual Skewness vs. Integration Skewness. Hence, as far as asymmetry of integration is concerned, unbiased-dense urban grids are sensitive to both the fragmentation and compactness of the containing urban shapes. Third, unbiased-dense urban grids display the best correlations between Integration and shape measures in comparison to the other two types. There exist strong and significant negative correlations Drelative vs. Integration (figure 8f) and Drelative vs. Mean Depth, which improve for shape hulls Drelative() vs. Integration. Therefore, unlike the biased type, the effect of urban shape on the integration of unbiased-dense cities is exerted according to the compactness of urban shape; this is further improved when shape hulls are considered. As a conclusion, cities with dense and undifferentiated grids located inside compact urban areas tend to be more integrated than those located in elongated areas.
In the study of offices, the theoretical fishbone in which secondary lines intersect the main double-lined spine, was defined as the ideal type representing the underlying structure of biased office layouts, whereas the theoretical grid evenly extended in both directions was defined as the ideal type representing the underlying structure of unbiased-dense office layouts. Fishbones consistently generated into a sample of 25 actual office floorplates showed strong and significant positive correlations between Integration and shape compactness and fragmentation (figure 8i and 8j). Grids showed strong and significant
negative correlations between measures (figure 8k and 8l).
and the two shape
At a glance, the findings about urban grids seem to partially contradict and partially reinforce the results of theoretical experiments with office floorplates. Correlations for fishbones are compared to the respective correlations for biased urban grids, while those for theoretical grids are compared to unbiased-dense urban grids. For fishbones, while the best match was the positive correlation Drelative vs. Integration (figure 8i), the respective correlation for biased urban grids is nonexistent (figure 8c). In addition, the correlation for fishbones Fconvex vs. Integration (figure 8j) entirely contradicts the respective one for biased urban grids (figure 8d).
For unbiased-dense urban grids, the strong and significant negative correlation between shape compactness and Integration (figure 8f) perfectly resembles the respective one for theoretical grids (figure 8k). The strong correlation between fragmentation and Integration for theoretical grids (figure 8l) is not matched by the poor correlation for unbiased-dense urban grids (figure 8g). However, when shape hulls are considered, a good and significant correlation is found (figure 8h).
Leaving the fishbones aside, biased urban grids resemble theoretical grids from the point of view of the affinity between shape fragmentation and grid integration (figure 8d and 8l). A closer examination of twelve urban grids which belong to the biased type reveals that their linear maps do have characteristics of grids where axial lines run mostly in two directions and cross each other to form rectangular city blocks. The bias in these grids is due to the existence of a few long connecting boulevards. These cities are thus termed “boulevard grids”. The extension of long boulevards in these cities takes advantage of long corridor strips in urban shapes that stretch along flat land or extend across bridges (e.g. Messina, Taranto,
Proceedings, 6th International Space Syntax Symposium, stanbul, 2007