Models, which represent in 3 dimensions the geometric elements of a city, are called 3D city models. Those models are used for an intended wide range of applications beyond mere visualization. Such uses are made possible by enriching the geometrical aspects with urban or environmental data that cannot be perceptible to a user without a visual abstraction. Using such models implies interacting with them in order to perform specific tasks related to the visualized data. Those data are potentially heterogeneous in terms of types, scales, issues, etc. When creating an enriched 3D city model, the designer faces the problem of choosing the most relevant visualization techniques for those heterogeneous data that have to be displayed within the same scene. In this paper we address the problem related to the visualization of enriched 3D city models by proposing a formal representation of existing visualization techniques in the form of an ontology of 3D visualization techniques. We also show how this ontology can be used for computational reasoning and help for the selection of visualization techniques by checking if a technique is compatible with a dataset to display or to a technique already selected and used in the same scene.