Vortex Analysis in Uncertain Vector Fields

Abstract

We present an approach to extract and visualize vortex structures in uncertain vector fields. For this, we generalize the concepts of the most common vortex detectors to uncertain vector fields, namely the λ <sub>2</sub>-criterion, Q-criterion, and the concept of parallel vectors at the example of the method by Sujudi and Haimes. All these methods base on the computation of derivatives of the uncertain vector field which are uncertain fields as well. Since they generally cannot be computed in a closed form, we provide a Monte Carlo algorithm to compute the respective probability distributions. Based on this, uncertain versions of vortex regions and core structures are introduced. We present results of our approach on three real world data sets in order to give a proof of concept.