Resumen
The demand for easily available geographic information is increasing in society. Moreover, knowledge of spatial data infrastructures (SDIs) has increased in many European governmental agencies, in large part because of the implementation of the INSPIRE directive. Many countries, thus, recognise the need to provide more detailed geographic information as network services at the national level. One means of realising this goal is to create INSPIRE extensions, i.e., to extend the INSPIRE data specifications with more detailed and specific national information. This paper describes a study where a complex INSPIRE extension has been created to describe the national need of hydrography information in Sweden, based on the Swedish water system standard (SWSS). The study includes the creation of a UML application schema that extends the INSPIRE Hydrography (HY) theme, the transform from UML to an XSD schema, the creation of GML files, and finally, testing and evaluating the approach of using INSPIRE extensions. When evaluating the results, the consequences of replacing existing dataset/download services with one extended INSPIRE HY dataset/download service are evaluated from the perspectives of both users and data providers. The evaluation is carried out as quantitative tests of the resulting GML files, in a user-centric test where a user tests the applicability of the GML files in hydrological analyses, and by telephone interviews with personnel from Lantmäteriet, the Swedish mapping, cadastral and land registration authority. Beside these evaluations, the possible effects on the information modelling process when creating an INSPIRE extension are also examined. The study shows that it is possible to create complex INSPIRE extensions that include many object types, attributes and relations. From a user perspective, extended INSPIRE HY files do not differ substantially from SWSS files, and can be used in hydrological analyses. Data providers can relatively simply replace their current download services with one for the extended INSPIRE HY, but the specific economic consequences for this could not be drawn. It could be expected, though, that there can be both economic, administrative and maintenance advantages if today?s separate INSPIRE and national download services are replaced with services exposing datasets based on an extended INSPIRE data model for all adequate themes.