Street-network sprawl map site

Below is information related to the interactive map site, ( which presents data from our research on global street-network sprawl.

Is there peer-reviewed research behind this site?

The sprawl map is an interface to data published in 2020 in two papers: (For now, preprints are available here:

What are the policy implications of the findings?

We discuss policy implications in the conclusions of our papers, above. More discussion will appear here shortly.

Where does my country / city rank?

You can explore SNDi listings of 200 cities and most countries in sortable tables, here: You can also download the data using the link at the top of each table.

Is this work reproducible? Can I get the code?

Our work was carried out using entirely open-source data, and all of the software we wrote to do the analysis is available under an open source license for anyone to use and modify (as long as all changes remain open source). See the research papers, above, for more detail about our underlying data sources and our algorithms. Briefly, we rely on data from OpenStreetMap and the Global Human Settlement Layer (and, for our research papers, the Atlas of Urban Expansion as a secondary source of information about time of development).

Our analysis was carried out with nearly 15,000 lines of custom Python and Postgresql code written by the two authors of the research, over 5 years, leveraging existing amazing open source tools like Networkx and Pandas in Python and the PostGIS layer on postgresql. Not only is our code open sourced, but all of our development was carried out using Ubuntu GNU/Linux (open source operating system) and exclusively open source software. Similarly, our research papers were written entirely using open source software. We made use of a powerful computation server (56 cores and 756G of RAM) but our code is scalable (and mostly parallelized).

After our research was nearly complete, Chris Barrington-Leigh and Muhammad Sumbal, with some early contributions from Sam Lumley, developed the interface to part of our data, also using entirely open source tools and a lot of custom code. That code, too, is available with an open-source license if you wish to use it.

Help! I'm having trouble with the web site,

Visit and join our discussion group to ask other users for help, leave comments, or report an issue. You can also discuss interesting features in the maps there. You can also directly contact Professor Barrington-Leigh at

How did you make this interactive map web site?

Even the code we used to build this site is open source. Thanks to the expert collaboration and initiative of Muhammad Abdullah Sumbal, with earlier help from Sam Lumley, the code for our entire site including the baking of tiles, the tile server, and the user interface are all open source, and were built on open source tools using open source operating systems (GNU/Linux). Our codebase is available on GitLab. We used open data (OpenMapTiles) for our basemaps. And our servers are of course running on open source operating systems too.