Settlement Grid - JRC

The GHS-SMOD is the rural-urban Settlement classification MODel adopted by the GHSL. It is the representation of the degree of urbanization (DEGURBA)[1] concept into the GHSL data scenario. Each grid in the GHS-SMOD has been generated by integrating the GHSL built-up areas and GHSL population grids data for reference epochs: 1975, 1990, 2000, 2015.

[1] The DEGURBA classification schema is a people-based definition of cities and settlements: it operates using as main input a 1 km² grid cell accounting for population at a given point in time. The DEGURBA discriminates the population grid cells in three main classes: 'urban centers' (cities), 'urban clusters' (towns and suburbs), and 'rural grid cells'. (base). These class abstractions translate to 'high density clusters (HDC)', 'low density clusters (LDC)', and 'rural grid cells (RUR)', respectively, in the GHS-SMOD implementation. The 'HDC' differ from the DEGURBA 'urban centers' in that they account for the over-fragmentation of cities in regions with large low-density residential development by integrating the built-up layer. In the GHS-SMOD representation, the 'HDC' are the spatial generalization of contiguous population grid cells (4-connectivity, gap-filling) with a density of at least 1,500 inhabitants per km² or a density of built-up surface > 50%, and a minimum total resident population of 50,000. The 'LDC' are continuous grid cells with a density of at least 300 inhabitants per km² and a minimum total population of 5,000. The 'RUR' are grid cells outside 'HDC' and 'LDC' with population > 0 and < 300. Everything else is classified as inhabited areas where population = 0. This dataset was produced in the World Mollweide projection (EPSG:54009).