Author: Gareth Simons, Research Fellow, Bartlett School of Architecture, University College London.

Evidence Based methods for Urban Design and Planning (EBDP) require spatial analytic workflows to derive insights. For example, the street network structure is analysed to understand which streets are more prominently located; proximities are used to compare access to land uses such as schools, retail, and green spaces; and census statistics provide information about population densities or employment levels for a given area.

Creating these forms of spatial analytic models for a given town or city can be time and resource intensive while also being data-source specific. This limits the applicability of the methods and generally precludes broader application across different contexts. There has resultantly been growing interest and application of wider-scale automated models run against data sources with more extensive coverage. Examples of this approach include the analysis of 27,000 street networks across the US (Boeing, 2018), analysis of street networks, land uses, and population densities for 931 towns and cities in the UK (Simons, 2021a), and a dataset encompassing networks, land uses, and building morphology for 50 cities spanning 29 countries (Yap & Biljecki, 2023). In each of these cases, the use of datasets with extensive geographic coverage allows for a more generalisable approach.

In the case of OpenStreetMap derived models, the workflows can theoretically be scaled globally, though this entails a drawback in that the quality of the data can be highly variable across locations (particularly for Points of Interest land-use data) and global sources of census data are limited in detail and resolution. An alternative is to use national-scale datasets which offer greater consistency and richer information on land uses and census statistics (as in Simons, 2021a), though these limit the application to a single country and can have more restrictive data licensing terms.

The T2E consortium explores this topic within the context of the EU. The interest in extents greater than a single country means that we are still interested in datasets with global coverage (Overture Maps, in our case), but by limiting the area of extents to the EU we are able to leverage EU specific datasets where available. This allows us to make use of:

• Eurostat 2021 Urban Centres / High Density Clusters data to rigorously define urban extents, from which we’ve extracted 699 towns and cities.
• Eurostat 2021 homogenised 1x1km census statistics for information on population densities, employment levels, place of birth, and population change.
• Copernicus Urban Atlas data from which we derive Urban Blocks, building heights, and tree canopies data.

We’ve opted to use Overture Maps instead of OpenStreetMap (OSM) for information on street networks, infrastructure (transport stops), Points of Interest (land-use information), and building footprints. Overture Maps is a relatively new alternative which includes a more formalised data release cycle, improved data validation in respect to Points of Interest, and expanded coverage for building footprints through supplementing OSM building footprints data with open-source data from Google and Microsoft. The data ingestion workflows are designed to work with Overture global data releases, a process which is facilitated by the overturemaps-py Python package, allowing for direct extraction of Overture information for each urban boundary.

Network data processing has also been enhanced through the use of the latest available automated street network cleaning workflows through the open-source cityseer Python package (Simons 2021b). These include level-aware cleaning to prevent incorrect merging at bridge locations and targeted merging based on highway classifications. Within the context of Nicosia, Cyprus, we are looking at how these automated techniques compare to models that have been manually created so that we can better understand the opportunities and limitations of automated procedures.

The development of the SOAR data model helps to assess the potential of scalable and automated methods for comparing cities across the EU. A core principle of SOAR is its commitment to openness, reproducibility, and collaboration across research institutions. By leveraging open data and by providing a replicable methodology, SOAR aims to provide researchers and planners with actionable insights to support evidence-based decision-making in planning and policy across the EU.

References:

Boeing, G., 2018. A multi-scale analysis of 27,000 urban street networks: Every US city, town, urbanized area, and Zillow neighborhood. Environment and Planning B: Urban Analytics and City Science. Available at: https://doi.org/10.1177/2399808318784595.

Simons, G.D., 2021a. Detection and prediction of urban archetypes at the pedestrian scale: computational toolsets, morphological metrics, and machine learning methods. Ph.D. thesis, UCL (University College London). Available at: https://discovery.ucl.ac.uk/id/eprint/10134012/

Simons, G., 2021b. The cityseer Python package for pedestrian-scale network-based urban analysis. Environment and Planning B: Urban Analytics and City Science. Available at: https://doi.org/10.1177/23998083221133827.

Yap, W. and Biljecki, F., 2023. A global feature-rich network dataset of cities and dashboard for comprehensive urban analyses. Scientific Data, 10, p.667. Available at: https://doi.org/10.1038/s41597-023-02578-1.

Disclaimer: The TWIN2EXPAND Project is funded by the European Union under grant agreement 101078890 and by the UKRI under grant numbers 10052856 and 10050784. Views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Executive Agency (REA). Neither the European Union nor the granting authority can be held responsible for them.