This paper presents the SEMAP framework designed to maintain and analyze the spatial data of a multi-modal environment model. SEMAP uses a spatial database at its core to store metric data and link it to semantic descriptions via semantic annotation. Through in-built and custom-made spatial operators of a PostGIS database, we enable the spatial analysis of quantitative metric data, which we then use in the context of semantic mapping. We use SEMAP to query for task-specific sets of spatial and semantic data, to create semantically augmented metric navigation maps, and to extract implicit topological information from geometric data.