Urban Growth Simulation Through Agent-Integrated Irregular Automata (AIIA)
MetadataShow full metadata
The research goal of this dissertation was to build a model for simulating urban growth and producing different future scenarios. This study proposed the Agent-Integrated Irregular Automata (AIIA) model – a hybrid between the cellular automata and agent-based modeling. The model uses irregular geometries (i.e. vector data format) as the unit of operation.
This dissertation is comprised of three interrelated research projects that are summarized into distinct chapters. The first project focuses on the development and implementation of the AIIA model. The model was deployed to model the urban growth of San Marcos, Texas. By validating against empirical data, the results demonstrated that the AIIA model produces accurate future growth scenarios. The model contributes to the advancement of existing methodologies on urban modeling and its simulations provide useful insights for urban planning and policy making.
The second project delves into the neighborhood subdivision component of the AIIA model. It is important to focus on automation of subdivision of developable lands into residential parcels because meaningful execution of behavioral rules in the AIIA demands interaction at the household level. The Parcel-Divider, a GIS toolset was created to automate parcel subdivision and generation of urban layouts. By comparing with many real-world subdivisions, the resulting toolset generates uniform and regularly-shaped lots while maintaining egress and continuity of road networks to the newly developed areas.
The third study examines the impact of neighborhood configuration to the proposed AIIA model. It concludes that the AIIA simulations are highly sensitive to the type and size of the neighborhood and recommends that sensitivity analysis should be an integral part of calibration to urban growth simulation. This dissertation research contributes to the advancement of urban growth modeling.