Cellular automata (CA) models are used for simulating land use change for more than two decades. These models have a simple structure based on a cellular partition of space, considering a finite set of cell states (or land uses) and their interaction within a given neighborhood area, changing throughout time under a set of transition rules. Transition rules are based on more or less sophisticated measures of state transition: they can be more complex rules that try to incorporate the different drivers of land use change or they can be purely probabilistic rules that take into account only the states of neighboring cells. This last approach is often based on a measure of a transition potential that establishes a rank for state transition for every cell. There are many drivers of land use change and accessibility is acknowledged as being one of the most important ones. At the same time, transportation systems (thus accessibility) are also influenced by land use change. This suggests that CA models are potentially good tools to simulate these phenomena by considering the cross-interdependences between both. In this presentation, we make a reflection on how accessibility can be measured, incorporated, and used to improve CA transition rules based on transition potentials towards more representative models of land use change. We address not only modeling requirements but also the potential of using CA models to evaluate both the impacts of transportation policies in land use change, and vice versa.

Peer Reviewed