Toward visual analysis of ensemble data sets

The rapid and continuing increase in available high-performance computing resources has driven simulation-based science in two directions. First, the simulations themselves are growing more complex, whether in the fidelity of the models, spatiotemporal resolution or (more frequently) both. Second, multiple instances of a simulation can be run to sample the results of parameters within a given space instead of at a single point. We name the results of such a family of runs an ensemble data set. In this paper we discuss the properties of ensemble data sets, consider their implications for analysis and visualization algorithms, and present a few insights into promising avenues of investigation.