Spatial economic self-organization with periodic and quasiperiodic dynamics

Abstract

The formation of regions through spatial self-organization is a key issue in spatial economics. The standard approach to the modelling of space in regional science has been to assume that space can be modelled as a one-dimensional system, often locations arrayed on a circle. This paper studies spatial self-organization in an economic geography model defined on a two-dimensional surface, assuming periodic or quasiperiodic dynamics. A discrete time model with discretized locations is applied, motivating a cellular automata approach. Numerical simulations suggest that the number of regions emerging on a two-dimensional model is different than that on a one-dimensional model, keeping other parameters the same. The spatial distribution of economic activity on a two-dimensional space (torus) appears less stable than one on a circle. A two-dimensional economic model can be extended to capture nonuniformity in the landscape by using a quasiperiodic dynamical system.