Open data-based urban models: an assessment for Canadian cities
Matthew Glowacki, Yuhao Lu, Zach Pankratz, Reece Cullen, Benjamin Boswick, Kyra Kwiatkowski
Radar charts of the average score across all categories for each of the 11 cities analyzed: a) Zurich-2.6; b) Helsinki-2.2; c) Singapore-1.1; d) Chicago-1.7; e) New York-2.4; f) Montreal-2.1; g) Ottawa-2.0; h) Toronto-2.7; i) Winnipeg-1.6; j) Edmonton-2.3; and k) Vancouver-1.8.
Project Overview: Digital urban models, such as Urban Digital Twins (UDTs), are rapidly advancing tools for urban design, simulation, and policymaking. However, the high costs of proprietary data and software often restrict their adoption. To democratize access to these advanced planning tools, this research investigates the feasibility of utilizing Open Data-Based Urban Models (ODUMs) in Canada. This study provides a comprehensive evaluation of municipal open data infrastructure across Canada and demonstrates its practical application at the City of Winnipeg.
Methodology + Key Findings: We developed an assessment framework to compare the open data portals of 11 North American cities, including Toronto, Montreal, Vancouver, and Winnipeg, and 3 international benchmark cities, namely Helsinki, Singapore, and Zurich. The open datasets were evaluated for timeliness, accessibility, and metadata quality to assess their readiness for GIS-based modelling. The assessment revealed significant cross-city variability and highlighted distinct gaps in Canada’s open data infrastructure. While vegetation and transportation datasets are consistently available and high-scoring across Canadian portals, critical layers such as buildings, energy, and socio-economic data are frequently absent or unevenly published.
The Winnipeg Case Study: To demonstrate that meaningful urban models can still be developed in cities using only open data, we constructed a stage 3 UDT for Winnipeg entirely from open-source datasets and software (QGIS and R). Despite the need for substantial data cleaning and preprocessing, the model successfully integrated nine open datasets, resulting in a comprehensive city replica. The final model incorporates a highly detailed street network encompassing over 4,600 km of roadways, 284,621 geometrically validated building footprints, and 1.48 million individually delineated tree canopies extracted from aerial laser scanning (ALS) data.
Contribution: This research establishes a replicable methodology for resource-constrained municipalities to build their own urban digital twins. Ultimately, we advocate establishing a national open data standard in Canada to improve data accessibility, enhance interoperability among cities, and further empower data-driven urban planning and community engagement.
Acknowledgement
We thank the University of Manitoba’s University Research Grants Program (Early Career Stream) and the Faculty of Architecture for funding (research start-up funding for Y.L.) for this research.
We also appreciate the valuable comments and suggestions from the editors and two anonymous reviewers.

