Hurricanes are the most destructive storms on Earth. They pose the greatest threat to life and property in many coastal areas. It is crucial to understand the triggering mechanism for rapid intensification of tropical storms.

Tropical storms are complex systems, characterised by intense winds, a low-pressure center, a closed atmospheric circulation and a spiral arrangement of rainbands. The study is based on the asymptotic analysis of nearly axisymmetric vortices with large tilt in dry air by Päschke et al., J. Fluid Mech. 701, 137–170, (2012). According to this theory, certain arrangements of non-axisymmetric heating and vortex tilt can induce vortex amplification. The main goal is to verify the asymptotic model by finding the theoretically predicted phenomena/structures in observational data and numerical simulations. For this purpose, advanced data analysis and visualization techniques for extraction and tracking of spatiotemporal characteristic features shall be applied and extended. Additionally, methods for visual comparison of multiple time-dependent fields shall be developed to reveal similarities and differences between results from asymptotic analysis, high-resolution simulations, and meteorological observations.

Feature extraction and tracking

According to the asymptotic theory, certain arrangements of non-axisymmetric heating and vortex tilt can induce vortex amplification. The video below shows the correspondent arrangement of the vortex core line and vertical velocity dipoles based on the idealized simulation conducted with EULAG  (Eulerian/semi-Lagrangian fluid solver).

In order to find a similar organisational pattern for real-case simulations first the vortex region should be identified. There are plenty of methods for vortex extraction which do not give the same result. Additionally none of them is parameter free. The choice of the method and the parameter values is critical. We study this definitional uncertainty by investigating the stability of various extraction methods.