Developing a web-based cellular automata model for urban growth simulation

Cellular automata as an emerging technology have been adapted increasingly by geographers and planners to simulate the spatial and temporal processes of urban growth. While the literature reports many applications of cellular automata models for urban studies, in practice, the operation of the models as well as the configuration and calibration of relevant parameters used in the models were only known to the model builders. This is largely due to the constraint that most cellular automata models were developed based on desktop computer programs, either by incorporating the model within a desktop GIS environment, or developing the model independent of a desktop GIS. Consequently, there is little input from the user to test or visualise the actual operation or evaluate the applicability of the model under different conditions. This paper presents a methodology to implement a fuzzy constrained cellular automata model of urban growth within a web-based GIS environment, using the actual urban growth of Metropolitan Sydney, Australia from 1976 to 2006 as a case study. With the web-based cellular automata model, users can visualise and test the operation of the model; they can also modify or calibrate the model's parameters to evaluate its simulation accuracies, or even feed the model with various 'what-if' conditions to generate alterative outcomes. Such a web-based modelling platform provides a useful and effective channel for government authority and stakeholders to evaluate different urban growth scenarios. It also provides an interactive environment that can foster public participation in urban planning and management.