Network based definition of functional regions: A graph theory approach for spatial distribution of traffic flows

Functional regions are autonomous (internally coherent and externally self-contained) spatial structures based on vector data, so-called spatial interactions. Typically, travel-to-work, travel-to-school flows and migrations are analysed by various methods of functional regional taxonomy in order to define functional regions. There is still another type of statistically recorded vector data which has, up to now, rarely been used for this purpose: traffic flows. However, these data differ distinctly from the above mentioned flows. In this paper we pursue two objectives: (i) to define functional transport regions based on a graph theoretic analysis of individual traffic flows, and (ii) to add knowledge to the issue of the self-containment of functional transport regions. The specific nature of transport data compared to the above-mentioned spatial interactions requires a specific methodological approach, which is presented in the paper. The existing graph theoretic procedures do not seem suitable for the definition of functional transport regions due to data specifics. Therefore our analysis is based on a rough analogy to the minimum cut method – we identify minimum flows in a graph representing a transport network. The territory of the Czech Republic is used as the example. Two regional systems are defined (based on 2000 and 2016 data) and compared in time. The paper achieves two main findings. First, the proposed methodological approach allows us to define autonomous functional transport regions, and the means to calculate their self-containment is discussed. Second, functional transport regions in the Czech Republic show unexpected stability over time compared to functional regions based on such spatial interactions as commuting flows.