The Sensor Web provides wider access to sensors and their observations via the Web. A key challenge is to infer information about geographic events from these observations. A systematic approach to the representation of domain knowledge is vital when reasoning about events due to heterogeneous observational sources. This article delivers a formal model capturing the relations between observations and events. The model is exploited with a rule-based mechanism to infer information about events from in-situ observations. The article also describes how the model’s vocabularies are used to formulate spatiotemporal queries. A use case for reasoning about blizzard events based on real time series illustrates the formal model.