The Degree of Urbanisation Classification
for Postcode Areas

This repository contains R code to implement level 1 and 2 of the degree of urbanisation (DEGURBA) classification (see Dijkstra et al. 2020 and Eurostat 2021) and to superimpose the grid cell classifications on postcode areas. The method is applied for five countries: France, Germany, Spain, Switzerland, and the United Kingdom.

Data sources

The classification is based on the following data sources. Data can be downloaded with the 01-data.R script.

• 2018 GEOSTAT 1sqkm population grids
• 2021 NUTS0 country border polygons
• 2018 Urban audit data including functional urban areas
• Postcode boundaries for:
  • France (2014 5-digit postcodes)
  • Germany (2020 5-digit postcodes)
  • Spain (2017 5-digit postcodes)
  • Switzerland (2021 4-digit postcodes + supplementary digit)
  • United Kingdom (2015 postcode sectors)

Classification procedure

The degree of urbanisation offers a common classification scheme of the urban-rural continuum that facilitates cross-national comparability. It was developed by the European Commission, the Food and Agriculture Organization of the United Nations, the United Nations Human Settlements Programme, the International Labour Organisation, the Organisation for Economic Co-operation and Development, and The World Bank. The methodology is described in detail in Eurostat (2021).

The classification is applied in two steps. First, 1 sq km grid cells are classified based on population density and contiguity. Second, spatial units are classified based on the share of their population in classified grid cells. Two classification levels allow for varying granularity of the urban-rural continuum. Here is an overview of category definitions for both levels.

Grid cell classification:

Level 1

1. urban centre: contiguous (Rook's case) grid cells with population >= 1,500 inhabitants and collectively a population >= 50,000 inhabitants. Gaps are iteratively filled afterwards, see majority_rule() in functions.R script.
2. urban cluster: contiguous (Queen's case) grid cells with population >= 300 inhabitants and collectively a population >= 5,000 inhabitants. Urban centres are removed from urban clusters afterwards.
3. rural cells: grid cells that are neither urban centres nor urban clusters.

Level 2

1. urban centre: same as for level 1.
2. dense urban cluster: urban cluster with contiguous (Rook's case) grid cells with population >= 1,500 inhabitants and collectively a population >= 5,000 and < 50,000 inhabitants.
3. semi-dense urban cluster: urban cluster with contiguous (Queen's case) grid cells with population >= 300 inhabitants, collectively a population >= 5,000 inhabitants, and outside a buffer of three grid cells of an urban centre or a dense urban cluster.
4. suburban cells: urban cluster cells that are neither part of dense or semi-dense urban clusters.
5. rural cluster: rural cells with contiguous (Queen's case) grid cells with population >= 300 inhabitants and collectively a population >= 500 and < 5,000 inhabitants.
6. low-density rural cells: rural cells with population >= 50 inhabitants and not part of a rural cluster.
7. very low-density rural cells: rural cells with population < 50 inhabitants.

Spatial unit classification

Level 1

1. cities: >= 50% of spatial unit population in urban centres.
2. towns and semi-dense areas: < 50% of spatial unit population in urban centres and <= 50% of spatial unit population in rural grid cells.
3. rural areas: > 50% of spatial unit population in rural grid cells.

Level 2

1. cities: same as for level 1.
2. dense towns: towns and semi-dense areas with larger population share in dense than semi-dense urban clusters and larger population share in dense + semi-dense urban clusters than suburban cells.
3. semi-dense towns: towns and semi-dense areas with larger population share in semi-dense than dense urban clusters and larger population share in dense + semi-dense urban clusters than suburban cells.
4. suburban areas: towns and semi-dense areas with larger population share in suburban cells than dense + semi-dense urban clusters.
5. villages: rural areas with largest population share in rural clusters.
6. dispersed rural areas: rural areas with largest population share in low-density rural cells.
7. mostly uninhabited rural areas: rural areas with largest population share in very low-density rural cells.

Note that when superimposing the grid cell classifications on spatial units, the population of a grid cell is here always weighted by the fraction of the cell that is actually covered by the spatial unit/polygon.

In addition, functional urban areas can be used to take commuting flows into cities into account. Here, NUTS3 functional urban area polygons from urban audit data are rasterized to make them available at the level of grid cells. Spatial units are classified as functional urban areas if >= 50% of the spatial unit population reside in functional urban area grid cells.

Example: Part of the Rhine-Neckar Metropolitan Region

Population Grid

DEGURBA Level 1 - Grid

DEGURBA Level 2 - Grid

DEGURBA Level 1 - Postcode Areas

DEGURBA Level 2 - Postcode Areas

Country-specific Postcode Area Classification

DEGURBA Level 1

France

Germany

Spain

Switzerland

United Kingdom
(Excluding Northern Ireland)

DEGURBA Level 2

France

Germany

Spain

Switzerland

United Kingdom
(Excluding Northern Ireland)