.pdf
Database management
The three databases describes in the previous section have a size of 3TB for the France. Even though we restricted the zone to three departents (Bas-Rhin, Haut-Rhin and Moselle), the size of the database is 300GB. Except for the size managing these data comports a another challenge: their nature. Indeed, this data are geospatial data, which means that a traditional relational database will not be able to manage them efficiently because it will require to test every single row of the database to know if it is in the zone of interest.
Geospatial database
To tackle this issue, I used a geospatial database called PostGIS. This database was chosen because it is the most popular geospatial database and because BD Topo is available in the PostGIS format.
PostGIS is an extension of PostgreSQL, which means that it is possible to use PostgreSQL to manage the data. The main advantage of PostGIS is that it provides a set of functions to manage geospatial data. For example, it is possible to query the database to know if a point is in a polygon. Moreover, the data are stored in a Octree structure, which means that the database will not have to test every single row to know if it is in the zone of interest.