Geospatial Modeling of Urban Landscape Changes through an Agent-based Approach

Over the past years, the agent-based modeling approach has emerged as a promising geospatial technique for urban change simulation. In this paper, we review the status of agent-based models for urban land change research, with the emphasis on the issues of how far agent-based models have been implemented to model the real world complexity. For such a purpose, we have identified four features of agent-based land use models: agent heterogeneity, driving forces, system feedbacks, and spatial and temporal scales. We focus on how these issues have been addressed in existing models with respect to urban land changes. Along with these discussions, we have also identified some challenges for model implementation as well as some issues that need further research. Finally, we have implemented an agent-based model to explore the impacts of cross-scale feedbacks on land use patterns in an urban area. The results have demonstrated the influences of cross-scale feedbacks on land fragmentation, compactness and dispersion.