Spot the Difference: Tornado Visualizations

Since tornado prediction is a critically vital task, the need to understand this complex weather phenomena drives a wide and fascinating range of research. Amy McGovern and her team at the University of Oklahoma are applying data mining techniques to hundreds of simulated storms to identify tornado precursors that could increase tornado warning lead time and prediction accuracy. The visualizations created in this project supported by the Extreme Science and Engineering Discovery Environment, via its Extended Collaborative Support Service (XSEDE ECSS) give the group a novel view of their data, helping them to refine the objects they use for the machine learning and data mining, and letting the scientists visually experience all the storms they want to as well as enabling them to see features that are not visible to the naked eye in nature.