A model of urban growth based on the empirical spatial distribution of establishments in Metro Manila, Philippines

Abstract

Metro Manila represents a self-organized growth of structures with minimal global coordinated planning. Apart from the fact that establishments are placed along major thoroughfares, our empirical analyses based on $k$-nearest-neighbor analyses of GIS data from OpenStreet Maps (OSM) show that administrative buildings tend to spatially scatter as evidenced by longer characteristic separation distances, while economic facilities tend to cluster with very short spatial separations. Inspired by these observations, we present a model of the growth of urban centers where the clustering and scattering tendencies are parameterized by effective "charges" representing the influence of the establishment at a given point in the surrounding space. Scanning for different values of the parameters result in realistic-looking spatial scatters and distributions.