Scientific models are increasingly dependent on processing large volumes of streamed sensing data from a wide range of sensors, from ground based to satellite embarked infrareds. The proliferation, variety and ubiquity of those devices have added new dimensions to the problem of data handling in computational models. This raises several issues, one of which -- providing means to access and process these data -- is tackled by this paper. Our solution involves the design and implementation of a framework for sensor data management, which relies on a specific component technology -- Digital Content Component (DCC). DCCs homogeneously encapsulate individual sensors, sensor networks and sensor data archival files. They also implement facilities for controlling data production, integration and publication. As a result, developers need not concern themselves with sensor particularities, dealing instead with uniform interfaces to access data, regardless of the nature of the data providers.