Mapping applications such as Google Maps or the French Géoportail provide a wide range of users with on-demand mapping. But combining data sources, merging different representations, and in some cases navigating maps remains tedious and cognitively demanding. For instance, users are unable to easily get details about a point of interest while preserving a general view of their itinerary. Advanced scenarios involving subject-matter experts who must perform more complex tasks in possibly critical situations, reveal even more issues. In the case of, e.g., a natural disaster crisis management such as a tsunami, exposed areas have to be identified quickly, and emergency access to secured areas communicated to evacuation teams on the ground. This thesis will explore novel ways of combining different types of maps and navigating from one representation to another.