The primary objective of this research is to design improved methods of visualizing uncertainty in scientific displays. A secondary objective is to identify and calculate uncertainty from data and algorithms. Almost all scientific data contain uncertainty, which are as important to the proper interpretation of the results as the original data. However, very few visualization techniques depict uncertainty. This research will take two approaches: mapping uncertainty information as an additional piece of data and creating new visualization primitives and abstractions. The researchers will use experimentation, metrics, and application expert(s evaluations to find the right combination of techniques. Visualization is a valuable tool in understanding large amounts of data and the phenomenon represented by these data. Complete specification of data should include uncertainty. The resulting techniques should significantly improve all visualization and graphics applications where data uncertainty is a concern.
