How others have defined cities and urban boundaries

[carlhiggs]

We have been considering again the question of 'what is a city' -- and how to define the boundaries and population for them again. @eugenrb mentioned she would look into the city boundary question, and to help this I thought I'd share some links to some resources which I located recently when re-considering this ¹.

• Our project drew upon the European Commision's Global Human Settlements Layer Urban Centres Database (r2019a) ---- which is founded on their analysis of 'degree of urbanisation' https://ghsl.jrc.ec.europa.eu/degurba.php ---- with modelled temporal coverage for urban centre statistics (including areas and population) for 1975, 1990, 2000, and 2015

• The UN population division has

  • a glossary of their terms describing meaning of city area description: https://population.un.org/wup/General/GlossaryDemographicTerms.aspx
  • a detailed discussion of 'city boundaries and populations' https://population.un.org/cityboundaries/resources/About_the_United_Nations_Database_of_City_Boundaries_and_Populations.pdf
  • a database of city boundaries https://population.un.org/cityboundaries/index.html
    • this is great --- and highlights the problem, e.g. Perth in Western Australia has 8 boundaries --- which is the right one? depends who you ask / what they're interested in...
    • notwithstanding the challenges, this data was used to produce the 2018 'the worlds cities' report: https://www.un.org/en/development/desa/population/publications/pdf/urbanization/the_worlds_cities_in_2018_data_booklet.pdf

• UN Habitat has a 2020 paper on 'what is a city' (great up to date summary of the issues, and includes a proposal for a 'unified city definition'; as far as I can tell, that hasn't been operationalised anywhere yet!) https://unhabitat.org/sites/default/files/2020/06/city_definition_what_is_a_city.pdf

• I just found some metadata at World Bank which also discusses how they defined a city and a metropolitan area (the latter is basically an urban agglomeration, or functional urban area --- GHSL also has a definition and database of FUAs incidentally): https://databank.worldbank.org/reports.aspx?source=2&type=metadata&series=EN.URB.LCTY#

  • the dataset itself is problematic (e.g. doesn't specify which is the largest city identified at each time point), but the metadata contains a great discussion of the challenges e.g. with inconsistent urban definitions throughout the world, and how estimates of urban populations are very sensitive to these. Points to other resources, which is useful!

Hope this is useful --- so the above isn't lost in an e-mail thread, I'll also add it as a discussion on the GitHub site!

Footnotes

1. Deepti and Jim were wanting to relate urban populations to Covid mortality -- and that has the same problem, how to consistently/comparably/meaningfully define urban boundaries ↩