Authors: Heeseo Rain Kwon*, University of Cambridge, Elisabete A. Silva, University of Cambridge
Topics: Spatial Analysis & Modeling, Land Use and Land Cover Change, Geographic Information Science and Systems
Keywords: SLEUTH, behavioral theories, agent-based modelling, cellular automata, data science, complexity theory, geographic information systems, planning support systems
Session Type: Paper
Presentation File: No File Uploaded
Understanding and predicting spatial patterns of urban change is greatly useful for planners to deliver evidence-based and adaptive policies to address current problems and future sustainability. While approaches like agent-based modelling (ABM) and Cellular Automata (CA) have been applied to generate dynamic simulation, the current approach is limited in taking multiple behavioral rules into complex agent behavior. This paper extends the existing CA-based SLEUTH model into a CA/ABM-based land use-transport interaction model on NetLogo using data from Sejong, Korea, and attempts to increase the model efficiency by using two approaches. First, we intervened in the CA and tried to improve the calibration of the five coefficients of SLEUTH by applying building height data into the urban layer and traffic volume big data into the transport layer. Second, we tried to embed ABM functions into the CA-based model by applying metrics for travel behavior (car to non-car mode-switch behavior) linked with variables based on behavioral theories from disciplines like psychology, sociology, and economics. For this, we added household socio-economic survey data and applied language-based coding based on three most applicable behavioral theories with a hypothesis that a lower percent share of car use will lower the coefficients of road gravity, dispersion, and breed, therefore lower the urban growth in the rural areas of Sejong especially along the road network. We expect around 10% increase in efficiency measured in Overall SLEUTH Metric, error margin and land use allocation and expect to draw discussions pointing to data science, complexity and behavioral theories.
To access contact information login