WorksPhysics and Astronomy
Modelling population distribution across cities on spatial networks using the kinetic theory of multi-agent systems
Abstract
We investigate the urban population distribution of the Italian region of Sardinia through the kinetic theory of multi-agent systems on spatial networks, interpreting cities both as agents and vertices of a graph, and representing interurban links as edges. We consider various emigration rules to model the interactions between agents, obtaining two main results: the bimodal lognormal fit of the total population distribution among cities and the Pareto fit of the tail. Overall, our model provides results consistent with the actual distribution derived from ISTAT data of Sardinia; furthermore, the bimodal lognormal fit generally follows the distribution tail more accurately than the Pareto one, although the two coincide in the first half.