Urban infrastructures, such as electricity and communication networks, laid and maintained by different companies, can be affected from each other's operations. Thus, there is a need to share the geospatial data and be able to infer the ramifications of a particular operation, which requires the effective interoperability of geographic information systems (GISs). This paper describes a solution taking into account the dynamic context in which essential services exist. The focus is on the semantic and behavioral changes of a GIS element when the context is changed. To explore this issue, an experimental system was implemented involving the GIS of two infrastructure companies. The models of both infrastructures were constructed as application ontologies that import upper-level ontologies. Context ontologies were used to model the maintenance and emergency situations. Guided by a rule base, responses to context changes were the key elements for the dynamic interoperation of GISs.