Simple download service (Atom) of the dataset: Database Occupation of the land at the municipal level (OSCOM) in 2016 in Indre-et-Loire

The OSCOM database is produced from various layers of so-called large-scale data (BD-TOPO®, BD-FORET®, MAJIC, etc.), so OSCOM can be considered as a large-scale land cover database, unlike CORINE LAND COVER for example. While the CORINE LAND COVER work scale thus falls within 1/100 000, the OSCOM product’s work scale is 1/10 000. This scale makes it possible to envisage uses of the database on an inter-communal scale, but also communal or infra-communal. This scale is well adapted to local environmental monitoring and management or spatial planning needs. In particular, it makes it possible to envisage a precise monitoring of the consumption of space at the municipal level. I — The Minimal Collection Unit The layers used to produce the database are vector data layers that have different validity scales depending on the source. Therefore, it is not strictly possible to determine a homogeneous Minimal Collection Unit (MCU) for the entire land cover database. Despite this heterogeneity, the accuracy of the different data sources is either directly derived from the IGN Large Scale Reference (GSR) (BD-TOPO®) or from a scale level compatible with it. This is notably the case of the Graphical Parcellar Register (GPR) of the ASP and the MAJIC database of the DGFiP, the accuracy of which is provided to the cadastral parcel. Within the framework of OSCOM, it was not envisaged to generalise topography in order to create an artificial UMC, as this would have led to a deterioration in the quality of the data sources used to form the database. II — The temporality and concept of vintage As with the Minimal Collection Unit, the layers used to produce the database have different temporalities, which poses problems in determining a homogeneous vintage for the soil occupation database. Although this bias is one of the assumed limitations of the method, the scope of the bias should be relativised in the light of the following factors. Indeed, among the databases mobilised, the RPG and MAJIC are vintage databases which therefore do not pose any problems as such. As part of the OSCOM version in 2016, all the data layers used have a vintage oscillating between 2012 and 2014, which makes it possible to use an indicative final 2013 vintage for this database. Although the timeliness of BD-TOPO® data now differs according to the classes of objects, which poses problems in determining a vintage, the most important classes of objects for the creation of the database, in particular the “BATI” class, follow the cycle of updating cadastral boards by the DGI. For other items, the IGN guarantees a maximum one year lag on the information provided. The timeliness of the data according to the different deliveries can therefore be regarded as relatively reliable. For the BD-FORET®, it is available in its version 2 at the level of the Auvergne-Rhônes-Alpes region only relatively recently. The observed deviations therefore remain acceptable in view of the unscalable nature of the data. In conclusion, the vast majority of geographic objects used to build the database respond to an update timeline compatible with the concept of vintage. However, special attention should be paid to certain classes of BD-TOPO® (business areas, in particular, etc.) whose cycle of updates (even if it is less than one year) may occasionally show greater deviations from the selected vintage.