Node geography and population: an alternative spatial model for estimating population distribution

Abstract

The paper presents an alternative spatial model based on data from OpenStreetMap and explores its capabilities and its relationship to population estimation and its distribution. By using OSM g eodata, a nd the level o f r oads, a spatial model i s d one having s patial network structure with nodes and links, as the main structural entities. Then, as a spatial model we use only the level of nodes and with spatial statistics and analysis techniques offered by the GIS, it is compared with the distribution of the population and the relation between them is examined. The systematic examination of the distribution of nodes leads to interesting conclusions and possibilities, such as the strong statistical correlation with the population distribution. Based on this finding, an alternative way of spatial representation is created / examined: the Node Geography. For further confirmation, the model was applied f or Greece and for selected regions of  rance. It a ppears thatt he correlation has no local characteristics, at least in Europe, and the implementation of the proposed model offers an alternative reliable solution, in cases where the estimation of population distribution is not easily feasible, and mainly an alternative spatial model with special dynamics and plenty of applications.