Comparing WUDAPT Level 0 Cartography with a More Detailed Urban Database. Some Examples for French Cities Using the MAPUCE Database

The World Urban Database and Portal Tool (WUDAPT) project is an international collaborative project started in 2014. It aims to acquire, to store and to disseminate relevant data on the physical geographies of cities worldwide for climate modelling purposes. The step of urban data production aims to acquire information on the form (surface cover, construction materials and urban geometry) and functions (activities that drive the urban metabolism) of cities. In WUDAPT data gathering is categorized in three levels of detail. The first of them, WUDAPT level 0 (WL0), is based on the Local Climate Zone scheme and uses Landsat images to classify the urban surface at 100m of resolution. The methodology to produce WL0 data is now stabilized and LCZ cartography has already been collaboratively obtained for several cities around the world(a) including three French cities (Paris, Toulouse and Nantes).

In this contribution, WL0 cartography for those French cities is compared to a more detailed database, the MAPUCE urban database, recently obtained in the framework of a national research project(b). In MAPUCE, an automatic platform under a Geographical Information System is built in order to derive 80 urban and social indicators at building and islet (street blocks) scale for any city in France. This French urban database is based on national geographical and socio-economical databases at several spatial resolutions ranging from the building to the neighborhood, by way of the parcel scales. Based on those topographic sources, morphological and architectural indicators are computed over new zoning based on contiguous street-blocks, the islet scale. Data will be publicly available at this spatial scale at the end of the project for so it’s the reference spatial scale used in this WL0 vs MAPUCE comparison.

WUDAPT L0 maps obtained for Paris, Toulouse and Nantes captures well the morphological diversity within the city and the land cover differences between those cities. To compare WL0 cartography with the MAPUCE one, a methodology to classify each MAPUCE islet on LCZs combining the building typology and the land cover was developed. This methodology was initially developed using the Toulouse study case and tested over Paris and Nantes. Thresholds chosen for this classification sometimes differs from those provided by Stewart and Oke’s study and will be discussed. Parameters ranges characterizing the existent LCZs for those cities are calculated and how the spatial distribution obtained and the standard LCZ parameters proposed by the WL0 framework discussed.

(a)    http://www.wudapt.org/cities/

(b)   http://www.cnrm-game-meteo.fr/ville.climat/spip.php?rubrique120&lang=en