Ongoing efforts and high investments have been made in recent years to develop models and techniques for the acquisition, maintenance and analysis of Geo data. The increasing use and development of modeling approaches in Geo sciences has caused a demand for, and the production of a vast amount of data. Additionally, many new case studies produce a considerable variety of field data. A great diversity of methods such as hydrological modeling, analyses of sediment covers, geomorphological mapping or dating methods provide heterogeneous data that are stored in diverse formats and in different databases.

The development of centralized data stores and respective query systems aims at providing large data sets, that are maintained centrally, widely available, and that have a well known homogeneous format. These systems are an important step to getting rid of the so called "data cemeteries". A major disadvantage of this approach, however, is the given structure of the central database. As a consequence, it is an additional expenditure for the owner of the data to restructure his data according to the demands of the central schema and, even worse, this schema will more or less not fit his special needs.

The progress in Geo sciences will continue to create specialized data types to support new applications and methods. Nowadays, we face, for instance, an increasing development and use of three and four dimensional models based on physical methods. But there are still no satisfactory solutions for handling spatio-temporal data within standard GIS.

The scenario described above obviates the need for open and extensible information systems supporting the integration and administration of heterogeneous Geo data. These systems should not only provide facilities for integrated and uniform access to various data sources with respect to both data format and employed source system, but also be able to act as a mediator between data sources and existing highly developed end user tools.