The nature of today's geographically and managerially distributed geospatial information sources makes the interoperability across different sources from different organizations difficult. The integration of real-time water quality assurance data with geographic data is even more challenging. In this paper, we present a context-aware geospatial data and service integration framework that is based on the combination of a syntactic model, a semantic model and a pragmatic model using Semantic Web technologies. This model is context-aware, with the ability to analyze existing dependencies, predict causes and effects and provide context-aware services (which information services are relevant, how to perform the services, how often they are needed, etc). As a proof of concept, we demonstrate the HIS-KCWater system for supporting users in analyzing real time watershed data, predicting the water quality using hydrologic model simulators, interpreting the results, dynamically forecasting problems and generating alerts about water quality issues.