Vagueness is prevalent within the geographical domain, yet it is handled poorly in existing ontology approaches. A proposed way to rectify this is to ground the ontology upon the data. By grounding the ontology, we make an explicit link between the ontology and the data, and thus allow reasoning to be made within the context of the particular data. In order to ground the ontology upon the data, we must first decide how to represent the data and how to handle the vagueness with reasoning. This paper illustrates the stages required to prepare geographical data for an ontology to be grounded upon, including considering how to reason about the vagueness, how to represent the data in a more efficient manner and how to reason about relations within the data to extract attributes that would be used within an ontology.