Researchers from ETH Zurich and the University of Wisconsin-Madison have correlated geoinformation data from 30 major cities worldwide with traffic congestion data. For the first time, they were able to demonstrate not just correlations, but genuine cause-and-effect relationships between urban changes and traffic flow. The study was published in April 2026 in the journal “Nature Communications”.
Three factors, one traffic jam
The research team led by first author Yatao Zhang distinguished three dimensions. The structure of the road network, the spatial form of the city and the function of individual areas, i.e. whether people live, shop or work there. Surprisingly, it is not only the road network that determines the flow of traffic. An urban sprawl structurally generates more traffic. The concentration of leisure activities in a neighborhood drives up weekend traffic. Mixed-use developments, on the other hand, bring living and working close together, shorten commuting distances and reduce the volume of traffic. “Traffic is created by what people do, not just by the existence of roads,” Zhang sums it up.
Singapore versus Zurich
An international comparison shows major differences. In Singapore, residential areas are clearly separated from the service center and structural changes have a direct impact on commuter flows. In Zurich, this link is much weaker because apartments are spread across the entire city and commutes are shorter and more diverse. Such differences can now be systematically measured and compared for the first time.
What this means for planning
ETH Professor Martin Raubal, who supervised the study, sees great potential for urban and transportation planning. The new method makes it possible to forecast how an intervention, such as the construction of a large shopping center, will affect traffic in the medium term. Cities could use it to better simulate measures before they are implemented. However, more in-depth detailed analyses are still needed before concrete recommendations can be made in Zurich or other cities.
Data from open sources
Open Street Map was the main source of data, supplemented by traffic congestion data from Here Technologies, which is updated every five minutes worldwide. For Los Angeles alone, the congestion values of over 18,000 road sections were included in the analysis. The fact that such a study is based on publicly accessible geodata makes the approach reproducible and scalable.
