In 1983 I left the International Institute for Aerospace Survey and Earth Sciences (ITC) and went to Wageningen Agricultural University, where I took the chair of “Land Surveying and Teledetection”. Up to that time I had been active mainly in geodesy and photogrammetry, but at Wageningen it soon became clear to me that the students there were more interested in the use of spatial data, rather than in its acquisition. They were students in disciplines such as Soil Science, Forestry, Soil and Water Conservation, Hydrology, Physical Planning, and Land Use Management. They needed spatial data for inventories, for problem analysis and for project planning and design. It was also evident that they needed the thematic data provided by remote sensing rather than the exact large-scale data produced by land surveying and photogrammetry. Therefore the emphasis of the activities of my department shifted to remote sensing and spatial information processing. In the introductory courses we taught the students the elementary concepts of these disciplines and showed them how to handle the technology of image processing and GIS.

The spatial problems that the other disciplines deal with appear to have a high level of complexity so that a straightforward application of RS and GIS technology is not sufficient, and is often not even possible. The application of modern tools requires a thorough understanding of the spatial descriptions used in these disciplines. They all deal with dynamic spatial complexes consisting of geographic objects that are spatially related and change through time due to processes acting on them. Each discipline is interested in some particular types of processes, so that the semantic aspects of the spatial descriptions differ from discipline to discipline—that is, the semantics depend on their scientific context. The structural aspects of these descriptions show strong similarities, however, and this has made it possible to develop a structural approach for spatial modelling that is relevant for a large variety of application fields.

Over the years in Wageningen a theoretical framework developed encompassing different aspects of spatial modelling. This framework has been used to structure the series of advanced lectures for postgraduate and PhD students that have been given over many years under the title of “Geo-information Theory”. Many of these students contributed to the development of these ideas through discussions and through their thesis work, and these contributions have been of great value to me. In the thirteen years that I worked at Wageningen Agricultural University I also had many discussions with colleagues from other departments.