Page 380 - 360.revista de Alta Velocidad - Nº 6

Yang, Haoran. Dobruszkes, Frédéric. Wang, Jiao’e. Dijst, Martin.

Figure 7. The spatial distribution of communities

In the HSR network, the modularity value is 0.58, and there exist eight subgroups with
typical characteristics of geographical distributions in China, reflecting that the geographical
proximity matters most in intercity relationships. For instance, in the YRD region, Shanghai
services as the core and form the strongest bonds with Hangzhou along the Jinghu and
the Shanghai-Hangzhou HSR routes. In the north-east with Liaoning, Jilin and Heilongjiang
provinces, Jilin-Changchun and Dalian-Shenyang form strong bonds along the Jingha HSR
routes covering the north-east regions. Besides, in Jiangxi and Fujian provinces, there are
strong agglomeration effects within each province where Nanchang-Jiujiang and Fuzhou-
Xiamen form the strongest bond, and then other peripheral prefecture-level cities are
connected by parts of the Hukun HSR route and the southeast coastal HSR route within each
province, respectively. Along the Wuhan-Guangzhou route connecting the middle and the
southeast, cities are also clustered. Furthermore, two clusters are formed mainly due to the
geographical isolation and less developed HSR network connections: Sanya and Haikou as
mentioned before are isolated from the mainland of China. There is no direct HSR channel
connecting the Hainan island to the mainland of China, which forces them to be a cluster
connected by the Hainan HSR route. There exists another cluster in the southwest where
Chengdu and Chongqing serve as the cores with limited connections with adjacent cities by
parts of the Huhangrong HSR route due to a rather slow development of HSR networks up to
2013 in the southwest. In sum, HSR cities in the specific regions tend to be clustered with
adjacent cities along the HSR routes.
In contrast, in the airline network, the modularity value is only 0.12 and accordingly there are
four subgroups in the airline network without the obvious characteristics of the geographical
cluster distribution in specific regions. For instance, in the airline transportation network, Sha-
nghai and Shenzhen in the middle-east and the south-east forms the strongest bond and starts
the first subgroup, and this subgroup grows with the addition of Xiamen and Tianjin. Then loose-
ly connected cities in different parts of China then are added to enlarge the original group. Des-
pite four subgroups in the dendrogram, they have not formed any major geographical clusters
separately in specific regions due to the overlapping wide influence scopes of aviation centers.

378 360.revista de alta velocidad