Semantic mediation is increasingly at the heart of the design of emerging web based information systems, particularly spatial information systems that require the integration or interoperation of a collection of heterogeneous data sources. Ontologies are increasingly used to represent agreed semantic of applications by providing formal description models and related tools to explicitly specify the conceptual entities of an application. Past research in traditional database integration has identified key issues in the resolution of various structural and semantic conflicts. Recent research effort has extended this early research to address the interoperability of spatial information systems. In this paper, we present an ontology based semantic reconciliation and interoperability methodology of heterogeneous spatial information systems based on three types of ontological components: 1) global mediation entities (GME) derived from a shared ontology, 2) cooperative semantics entities (CSE) which are used for inter-context mediation, and 3) local ontologies representing entities of the local data sources. The shared ontology provides the generic ontological terms that can be refined by defining ontological agreements to adapt the structure and the intended meaning of the components of the shared ontology to the semantic characteristics of a cooperative data source. Context transformation functions are used to achieve semantic information integration by mapping cooperative semantic entities to the local entities or to the entities of another cooperation context. We use OWL ontology description language to describe the structure and the logical domain properties of the global and cooperative semantic entities. The reasoning and inference capabilities of OWL are used to allow automatic or semi-automatic processing of the mediation approach.