Smart cities effectively integrate human, physical, and digital systems operating in the built environment to provide automatic and efficient applications. While 3D city models, Internet of Things (IoT), and domain models are essential components of smart cities, the integration of IoT resources and 3D city models is a central information backbone for smart city cyber-infrastructures. However, we argue that most of the existing solutions integrating 3D city models and IoT resources are usually customized according to applications and lack of interoperability. To improve the interoperability between smart city modules, this study proposes a semantic-based methodology to integrate OGC CityGML and SensorThings API standards. Based on the data models from these two standards, this study proposes an integration ontology to connect information from these two standards. Due to the flexible definition of Thing in the IoT, the proposed ontology also considers multiple views of Thing. As a result, information from the CityGML and SensorThings API can be connected and queried via SPARQL queries. In general, this proposed integration ontology can facilitate the integration of IoT and 3D city models to achieve interoperable smart city.