Integrated modelling of population, employment and land-use change with a multiple activity-based variable grid cellular automaton

A constrained cellular automaton (CA) framework is used to model both land use and, at the same resolution, densities of population and economic activity. The multi-scale processes determining the location of population, economic activity and land use are captured in a variable grid CA, in which the neighbourhood of each cell includes the entire modelled area. The transition rules generating the spatial dynamics incorporate both the land use and the activities, and because they cover the entire modelled area, they represent interaction effects at all spatial scales; effectively, they are distance decay functions. In general, any particular cell hosts a number of activities (population, employment in various sectors) regardless of its land use, so in effect multiple land uses are modelled as multiple activities, although activity levels are normally highest on cells of the corresponding land use. The model is applied to both the urban-centred Greater Dublin Region and the country of Belgium, which consists of the entire polycentric urban system and its rural matrix. Results for both applications are good, as measured by the errors of both predicted populations and fractal dimensions, and the model outperforms the best existing models by these measures.