Geovisualization of artificial land use in European cities in 2006, with urban scaling laws
Résumé
Almost all cities, even those with a decreasing population, continue to spread out with a growing urbanization. Current urbanization tends to consume a lot of space (suburban areas, business parks), combined with strong de-densification in rural areas (less attractive), which leads to a significant increase in soil artificialisation on the outskirts of large cities. The constant growth of cities reveals sustainability issues in the housing and transport sectors by challenging the spatial organization of cities, for example by increasing travel times and car use. Cities are major sources of pollution, even more in the context of climate change and the health of city residents. They are also responsible for the depletion of many energy resources. Moreover, heavy urbanisation and the constant increase in artificial land use create profound environmental consequences (impacts on fauna and flora, urban heat island...). And urban sprawl is often an irreversible process.
The desire to have a more sustainable management of urban space, for example by reducing the artificial land use, have been at the heart of urban development policies for decades now.
This web application presented here focuses on artificial land use in 2006 with urban scaling laws, in 305 European urban areas with more than 100,000 inhabitants, described by harmonized land use databases. We use the Copernicus Urban Atlas 2006 land use database at 5 m resolution and the Geostat population grid at 1 km resolution. Artificial land use refers to all land use categories in the Urban Atlas except water areas, agricultural spaces, urban green spaces and forests.
The objective is to analyze how the urban structure (more specifically land uses) changes with distance from the main center.
Urban scaling laws (which allow to transform a set of objects from one spatial scale to another without changing their structure) prove themselves to be a great tool to compare cities (Batty, 2015). Indeed, after rescaling distances from City Hall, European cities are positioned at similar levels on the issue of land use. The large corpus of cities to be treated and the objective of comparison on a European scale reinforces the relevance of the use of urban scaling laws. Thanks to rescaling, comparison becomes possible for cities of different sizes and geography makes it possible to provide spatial comparability of territories.
The methodological choice behind this work is the use of radial analysis to describe the cities. The radial analysis allows the complete distribution of land and people in the city. This is based on Alonso's model, the city is considered as circular and monocentric. The reference center for the city corresponds to the city hall in each case.
The city does not only represent the urban core, but the peripheral spaces must also be considered in the analysis of urban scaling laws (and thus the whole urban environment is taken into account). The city is studied morphologically, focusing on the organization of the succession of land uses. Using a distance-based approach allows us to better grasp the phenomenon of urban sprawl and how to better control it.
The main result of this work is that the organization of land use successions occurs at the same rate if we cancel out the size effect of urban areas (Lemoy, 2020). Larger cities are larger in terms of area but are not more parsimonious in their land use per capita.
Nevertheless, urban scaling laws can't explain everything either: there are some cities that stand out, and some variations that appear at the margin. In particular, two large countries stand out compared to the average European city: French cities are more artificial than the average while Spanish cities are less, more specifically in the first 10-30 kilometers (rescaled distance with London as reference). It is therefore interesting to study the differences in artificial land uses and to analyse which types of artificial spaces can be found in excess in France.