A database oriented approach to GIS designing

In the framework of Land Planning and Management activities, a database including a large variety of datasets of heterogeneous typology and acquisition is fundamental in order to allow for diversified queries and efficient selection of data. Local municipalities (and unions or consortia of municipalities) in Regione Lombardia are encouraged to create their own geographical information systems (GIS) for territorial data management and to incorporate in the database the results of the nowadays on-going photogrammetric surveys which will produce the updated cartography of the region in the form of a database with geo-referenced and topologically structured data (the Italian definition is Database Topografico, and the acronym is DBT). Usually, the towns’ technical departments have cartographic data available in digital format, which means that at least a part of their digital information archives are geographically referenced. However, this does not imply that all the data are already organized as a DB. In general, when newly surveyed data have to be integrated with already existing digital archives containing spatial and thematic information, the challenge is twofold: (a) to structure the existing data archives according to the rules of relational DB design (in the case that it is not already so), in order to obtain a model which univocally organizes and describes the information content of the GIS (thus allowing also for data interoperability); (b) to incorporate the new updated cartographic data (DBT) in the “old” database, without compromising the newly created structure. Both aspects can be studied and solved in the frame of conceptual modeling, which represents the basis for structuring large sets of data enabling to manage a variety of applications for a wide spread of potential users. In this paper, a conceptual model is presented for a relational database for a GIS application to be the main tool supporting everyday work of a Consortium of Local Municipalities in Regione Lombardia. Solutions are presented for both problems described above. After logical modeling, the system was implemented and tested at a prototype level, showing that several data sets (many of them with spatial reference) could be structured in a unique database, producing analysis and elaborations both in an automatic and efficient way.