Topological Analysis of the 2D von Kármán Street

Topology based analysis and feature tracking is a well studied area. In this work, we focus exclusively on a dataset called the von Karman street, and apply topology-based methods to understand its vortices. For this analysis, we adapt the recently proposed edit distance between merge trees. We discern several interesting results. One, we observe spatial periodicity between the vortices, alternating every half-cycle. Two, we observe a distinct difference in spatial probability of vortex regions during a half-cycle. Further, we compare the accuracy of our spatial probability with an off-the-shelf machine learning approach.