Skip to main content

Data for: Urban form and its impacts on air pollution and access to green space: A global analysis of 462 cities

Cite this dataset

Rezaei, Nazanin; Millard-Ball, Adam (2022). Data for: Urban form and its impacts on air pollution and access to green space: A global analysis of 462 cities [Dataset]. Dryad.


A better understanding of urban form metrics and their environmental outcomes can help urban policymakers determine which policies will lead to more sustainable growth. In this study, we have examined five urban form metrics – weighted density, density gradient slope, density gradient intercept, compactness, and street connectivity – for 462 metropolitan areas worldwide. We compared urban form metrics and examined their correlations with each other across geographic regions and socioeconomic characteristics such as income. Using the K-Means clustering algorithm, we then developed a typology of urban forms worldwide. Furthermore, we assessed the associations between urban form metrics and two important environmental outcomes: green space access and air pollution.

Our results demonstrate that while higher density is often emphasized as the way to reduce driving and thus PM2.5 emissions, it comes with a downside – less green space access and more exposure to PM2.5. Moreover, street connectivity has a stronger association with reduced PM2.5 emissions from the transportation sector. We further show that it is not appropriate to generalize urban form characteristics and impacts from one income group or geographical region to another, since the correlations between urban form metrics are context specific. Our conclusions indicate that density is not the only proxy for different aspects of urban form and multiple indicators such as street connectivity are needed. Our findings provide the foundation for future work to understand urban processes and identify effective policy responses.


The dataset includes the raw data required to calculate the urban form metrics used in the article including weighted density, street connectivity, density gradient slope, density gradient intercept, and compactness. We merged it with the GHSL statistics data (open-source data available here) for the analysis.

The street connectivity metric was calculated using the approach proposed by Barrington-Leigh and Millard-Ball (2020). The metric was calculated based on OpenStreetMap.

Other metrics were calculated using the expressions and methods explained in the paper. The data source for the process was the Global Human Settlement Layer (GHSL) 2015.

Please see the README document for more information. 

Usage notes

The data is in the .csv format which can be opened and read with Google Sheets, Microsoft Excel, R, Python, and several other programs.