If you have any problems related to the accessibility of any content (or if you want to request that a specific publication be accessible), please contact us at firstname.lastname@example.org.
SUNPRISM: A Software Framework for Climate Change Research
AdvisorDascalu, Sergiu M.
Computer Science and Engineering
AltmetricsView Usage Statistics
This dissertation presents the SUNPRISM approach and tools aimed at supporting scientific investigation via new capabilities for combining data transformations, model simulations, and output visualizations in application scenarios developed for climate change research. Consisting of two specialized software tools, SUNPRISM Scenario Manager and SUNPRISM Visualizer, the SUNPRISM framework's defining characteristics are a visual, user-friendly object-based interface for scenario configuration; a workflow-based environment that allows code generation and dataflow scenario execution; and data visualization capabilities for 3D environments, including for immersive virtual environments such as CAVE. An application scenario consists of a combination of data resource objects and computational activity objects and, procedurally, there are five modes of operation in SUNPRISM: resource registration, scenario configuration, automated scenario code generation, scenario execution, and data examination.The proposed framework addresses known issues pertaining to data and model interoperability, in particular inadequate tool support, and provides new software solutions for defining and running scenarios pertaining to climate change research. The dissertation presents details of SUNPRISM procedural steps, describes the requirements and architecture of its associated software tools, and illustrates the framework's capabilities through a detailed description of its components and step-by-step presentation of two application scenarios. The first such scenario accesses a locally stored file set that contains the global summary of day (GSOD) data that consists of historical measurements from over 11,000 worldwide stations. The second scenario retrieves forecast data from the National Digital Forecast Database (NDFD) through their provided web service. Both scenarios have been developed to produce 3D visualizations of specific climate data superimposed over the world map.Although the framework has been designed for use in climate change research it can be easily adapted to other areas of research in which model simulation and data processing activities play a significant role.