| The graphical depiction of uncertainty information is emerging as a problem of great importance in the field of visualization. Scientific data sets are not considered complete without indications of error, accuracy, or levels of confidence, and this information is often presented as charts and tables alongside visual representations of the data. Uncertainty measures are often excluded from explicit representation within data visualizations because the increased visual complexity incurred can cause clutter, obscure the data display, and may lead to erroneous conclusions or false predictions. However, uncertainty is an essential component of the data, and its display must be integrated in order for a visualization to be considered a true representation of the data. The growing need for the addition of qualitative information into the visual representation of data, and the challenges associated with that need, command fundamental research on the visualization of uncertainty.;This dissertation seeks to advance approaches for uncertainty visualization by exploring techniques from scientific and information visualization, creating new visual devices to handle the complexities of uncertainty data, and combining the most effective display methods into the Ensemble-Vis framework for visual data analysis. Many techniques exist for graphical data display. However, their usage on data with uncertainty information is not straightforward. This work begins by first exploring existing methods for data visualization and assessing their application to uncertainty. New visual metaphors are then presented for the depiction of salient features of data distributions, including indications of uncertainty. These new methods are inspired by proven visual data analysis techniques, but account for the requirements of large, complex data sets. Finally, Ensemble-Vis is presented, which combines effective uncertainty visualization techniques with interactive selection, linking, and querying to provide a user-driven, component-based framework for data investigation, exploration, and analysis. |
