Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


School of Computing

Committee Member

Dr. Donald H. House, Committee Chair

Committee Member

Dr. Joshua A. Levine

Committee Member

Dr. Brian C. Dean

Committee Member

Dr. Amy W. Apon


The track forecast cone developed by the U.S. National Hurricane Center is the one most universally adopted by the general public, the news media, and governmental officials to enhance viewers' understanding of the forecasts and their underlying uncertainties. However, current research has experimentally shown that it has limitations that result in misconceptions of the uncertainty included. Most importantly, the area covered by the cone tends to be misinterpreted as the region affected by the hurricane. In addition, the cone summarizes forecasts for the next three days into a single representation and, thus, makes it difficult for viewers to accurately determine crucial time-specific information. To address these limitations, this research develops novel alternative visualizations. It begins by developing a technique that generates and smoothly interpolates robust statistics from ensembles of hurricane predictions, thus creating visualizations that inherently include the spatial uncertainty by displaying three levels of positional storm strike risk at a specific point in time. To address the misconception of the area covered by the cone, this research develops time-specific visualizations depicting spatial information based on a sampling technique that selects a small, representative subset from an ensemble of points. It also allows depictions of such important storm characteristics as size and intensity. Further, this research generalizes the representative sampling framework to process ensembles of forecast tracks, selecting a subset of tracks accurately preserving the original distributions of available storm characteristics and keeping appropriately defined spatial separations. This framework supports an additional hurricane visualization portraying prediction uncertainties implicitly by directly showing the members of the subset without the visual clutter. We collaborated on cognitive studies that suggest that these visualizations enhance viewers' ability to understand the forecasts because they are potentially interpreted more like uncertainty distributions. In addition to benefiting the field of hurricane forecasting, this research potentially enhances the visualization community more generally. For instance, the representative sampling framework for processing 2D points developed here can be applied to enhancing the standard scatter plots and density plots by reducing sizes of data sets. Further, as the idea of direct ensemble displays can possibly be extended to more general numerical simulations, it, thus, has potential impacts on a wide range of ensemble visualizations.