An object-based cellular automata model to mitigate scale dependency

摘要: Cellular automata (CA) are individual-based spatial models increasingly used to simulate the dynamics of natural and human systems forecast their evolution. Despite simplicity, they can exhibit extraordinary rich behavior remarkably effective at generating realistic simulations land-use patterns other structures. However, recent studies have demonstrated that standard raster-based CA sensitive scale, more specifically cell size neighborhood configuration for simulation. To mitigate dependency, a novel object-based model has been developed where space is represented using vector structure in which polygons correspond meaningful geographical entities composing landscape under study. In addition, proposed allows geometric transformation each polygon, expressed as change state part or totality its surface, based on influence respective neighbors. The implementation testing this real data reveals it generates configurations patches than conventional model.