LandSys: an agent-based Cellular Automata model of land use change developed for transportation analysis

This paper reports on LandSys, a land use model that integrates a Cellular Automata (CA) model and agent-based models to facilitate transportation demand modeling and analysis. The LandSys model simulates the spatial suitability of land use change over time based on both the land use suitability and the impacts of neighboring land use types in a CA model framework, coupled with individual decision-makers’ behaviors with agent-based models (e.g., household, employment, and developer agents). To generate inputs for transportation models, two equilibriums in the land use market are considered: land development equilibrium simulates land use change at a manageable cell level (50 m × 50 m), while land supply–demand equilibrium allocates firms and agents based on the bid-rent theory. Data from Orange County, FL, are used in a case study for model estimation and validation. A comparison of model results with actual data from 2000 shows that the proposed LandSys model is effective in estimating land use change, with an accuracy of 85.4%. The model is also responsive to land use policies and could be used for what–if analysis.

► LandSys integrates CA and agent-based models to facilitate transportation analysis.
► Land development equilibrium simulates land use change at a cell level.
► Land supply–demand equilibrium allocates agents based on the bid-rent theory.
► LandSys accurately predicted the land use change at 85.7%.
► The household and employment change was predicted at 75.7% and 69.9%.