Building an Environment for Spatial Modelling of Canadian Cities Using Open Data: A Replicable Workflow Applied to Winnipeg

Microscopic approaches, such as agent-based modelling, are increasingly used by urban planners to model cities. The development of accurate microscopic models is facilitated by the increasing availability of real-world open data, essential for agent-based models to be widely used and replicated. Whereas open data is gaining popularity in Canada, only few agent-based models have been built for urban planning using open data. This paper presents a workflow for building an environment for agent-based modelling of Canadian cities based exclusively on open data and tools. Combining data from Census, OpenStreetMap, and local portals, we propose a synthetic population and an environment model that form a basis for agent-based modelling of any Canadian city. The approach is implemented and validated on Winnipeg. Results indicate that the model offers a valuable basis for urban analysis. The proposed workflow has the potential to support human-centred urban planning and to facilitate replicable modelling in Canada.