Delimiting boundaries of market areas of central places using the density of retail facilities in an urban space

Abstract

When Christaller's central place theory is applied to an anisotropic urban space, the hierarchical structure of the central places in hexagon shapes does not appear. Thus, we have difficulties delimiting boundaries of market areas of central places. In order to overcome this limitation, we suggest the geographic information system (GIS) method based on a hydrological modelling which uses raster density and apply the method to the retail facilities of Seoul, Korea. Kernel density function is used to measure the density of retail facilities, and search radii are set up as 500, 1000, and 5000 m by considering the size of the neighbourhood unit, the neighbourhood district, and the sub-centre region. As a result, we confirm market areas of central places with irregular shapes, which become nested in those of higher order central places. This means that central places in an urban space have the hierarchy, which is the core concept of Christaller's theory. In order to evaluate the result of delimiting market areas, we suggest the normalized internal traffic volume ratio index (NTI). The average NTIs of the market areas of central places are higher than those of administrative units, and thus, we conclude that market areas of central places are effectively delimited and the GIS method based on a hydrological modelling is useful.