Urban expansion simulation of Southeast England using population surface modelling and cellular automata

The question of where to accommodate future urban expansion has become a politically sensitive issue in many regions. Against the backdrop of 'urban compaction' policy, this study uses population surface modelling and cellular automata (CA) to conduct an empirical urban growth simulation for Southeast England. This implementation leads to a consideration of the proper balance between the theoretical abstraction of self-organised growth and empirical constraints to land development. Specifically, we use 1991 and 1997 postcode directories to construct population surfaces. From these, the distributions of developed and vacant (rural) land are derived. Development potential is assessed through accessibility surfaces, which are constructed from the travel/commuting time to major London rail termini, to motorway junctions, and to principal settlements. Through investigating the frequencies of land development in relation to the accessibility surfaces, we can begin to understand the distribution of land development in this region. Based on this empirical relationship, the transition rules of a CA simulation of future urban expansion are constructed. In addition, government population projections at the county level are used to constrain simulation to the year 2020. The study demonstrates the utility of empirical CA in urban growth modelling; in particular the importance of empirically informed CA simulation rules in characterising the distribution of land development.