An Object-oriented Geographical Cellular Automata

Xia LI,Shao-kun WU
DOI: https://doi.org/10.3321/j.issn:0529-6579.2006.03.023
2006-01-01
Abstract:Recently, there are growing studies on applying cellular automata (CA) techniques to the simulation of complex geographical phenomena, such as land use dynamics. There are problems on how to define transition rules and provide user-friendly interface for using cellular automata. An object-oriented CA is proposed, which has a number of advantages in simulating complex geographical phenomena. This architecture allows a better opportunity for model extension and model integration. In this paper, MCE (Multicriteria Evaluation) techniques are mainly used to recover the transition rules of GeoCA, which take into account a number of spatial variables. A stochastic variable is embedded to address the uncertainties in urban simulation. A universal GeoCA model is designed based on the objected-oriented architecture with the capabilities of highly extendable, user-friendly through the Visual basic language and the ArcObject components of ARCGIS. This GeoCA model is tested in the simulation of fast land use changes in the Dongguan city, the Pearl River Delta of China. Plausible simulation results have been obtained by using this object-oriented simulation model.
What problem does this paper attempt to address?