This paper presents a statistical phrase-based machine translation system which is enriched with semantic data coming from a spatial ontology. Paper presents the spatial ontology, how it is integrated in statistical machine translation system using factored models and how it is being evaluated using both automatic and human evaluation. Spatial information is added as a factor in both translation and language models. SOLIM spatial ontology language is used to implement ontology and to infer necessary knowledge for training statistical machine translation system. The machine translation system is based on Moses toolkit.