Robustness Analysis of City Road Network at Different Granularities - Space-Time Integration in Geography and GIScience - page 129

Fig. 8.3 Community-based
urban road network

imply that community structure in urban road networks can provide novel perspec-
tive for network robustness analysis. Different from segment-based and stroke-based
urban road networks, if a blue segment fails, we consider all blue segments in
community C 1 is affected. Thus we can see that destroy of green community can
be more harmful than blue community to the entire network since the malfunction
of green community will isolate the blue and the red community. Furthermore, the
community perspective can help improving road network robustness. As we can
see in Fig. 8.3 , if another community is built between blue and red community and
provides more connections, the damage of green community can be less harmful to
the entire network. From Fig. 8.3 , we can see that community can provide a novel
and important perspective on urban road network's structure and help enhance its
robustness analysis.
To evaluate the feasibility of city road network robustness analysis at multiple
granularities, we provide a series experiments with real world urban road network.

8.3 Data and Methodology

8.3.1 Experimental Data

In this article, we use the Beijing road network to test our method. This network
consists of 16,491 road segments which cover the downtown of Beijing. Figure 8.4
shows the initial road segments.
Based on the initial road segments, we generate three types of Beijing road
network including segment-based, stroke-based and community-based dual graph
using the methods described in Sect. 8.2 . To display the differences in three dual
graphs, we use open source software GUESS (Adar and Kim 2007 ) to visualize the
dual graphs in topological spaces in Fig. 8.5 .