Abstract: Knowledge graphs express the spatial relationships of water conservancy objects through node attributes in triplets. However, when visualized, they fail to reflect the relative positional relationships among water conservancy objects, resulting in unintuitive spatial semantic representation and inconsistency with the reading habits prevalent in the water conservancy industry. To address this issue, this paper proposes a water conservancy knowledge graph construction method that integrates spatial location information of water conservancy objects, offering a complete solution from knowledge graph construction to visualization. In the key stages of knowledge graph construction—including ontology modeling, schema design, spatial relationship computation, knowledge extraction, and storage—the spatial topological relationships of water conservancy objects are incorporated. Coordinate information is utilized to control the visualized representation of the knowledge graph, thereby achieving a faithful mapping of the actual spatial relationships of water conservancy objects. This method represents a tailored adaptation of traditional knowledge graph construction techniques to the specific characteristics of the water conservancy industry, and can more effectively support the development of water conservancy knowledge platforms.