Integrating geospatial information in a PostGIS database

ongoing

Statistics Denmark receives and stores most geospatial data in compressed files. Currently we have only addresses and cadastres available with spatial information in an ORACLE database. This makes searching, logging and accessing the data difficult, if other geospatial data than these are to be used. It creates a barrier for statisticians to publish data with geographical distributions beyond municipal level (NUTS3/LAU), while also adding extra costs. There is a wealth of administrative data containing geospatial information, such as agricultural land parcels, buildings, road and rail networks, etc. Easier access to these will open new opportunities for producing statistics. A better utilisation of registry data can potentially make existing production cheaper and reduce reporting burden. The project's aim is to generate a model for loading geospatial data into a PostGIS database, which is easy and fast to connect with other statistical data. It makes the data searchable and standardized and allow us to communicate more proactively about integrating geospatial data in statistical production. On a long term, it will extend the range of possible map visualizations in our Statbank while expanding the competencies of both the statistical community and IT staff. For the duration of the project, four data sets will be chosen as templates for each data type: one for raster and three for vector (one for each geometry). Integrating raster data into a database will also support new projects involving Earth Observations. In summary, the projects main goal is to automate the load of geospatial data to a PostGIS database connecting them to existing statistical data.