The search of semantic understanding in information systems has led us to take advantage from fields like Natural Language processing from the Artificial Intelligence. This is the case of ontologies which have been used to solve many problems such as interoperability and integration. This paper presents a point of view over specific-domain ontologies to be used in geospatial activities which are named spatial ontologies. An overview of ontologies which could cover this need is presented, focusing on their strengths and weaknesses. In the last part we present the ontology to be used for the semantic integration of spatial schemas in the SIT-SD (Semantic Integration Tool for Spatial Data) prototype. This Variable-Level Spatial Ontology is being developed as part of the SIT-SD. In this ontology we have defined the main spatial characteristics to be used in the inference with spatial information.