A Spatial and Temporal Autocorrelated Growth Model for City Rank—Size Distribution

Abstract

Rank—size distribution has been an important tool in characterising and analysing city size distributions across spatial and temporal scales. Zipf’s law in city rank—size distribution has been observed in many analyses and is considered an important empirical regularity describing the organisation of cities. Based on analyses of the evolution of cities in Texas from 1850 to 2000, this paper documents spatial and temporal autocorrelation in city population growth rates. A modelling strategy has been developed that accounts for the spatial and temporal autocorrelated growth in Texas cities and is effective in reconstructing the empirical rank—size distribution. This study shows that it is necessary to take into account the interdependence among cities in simulating the city size distribution.