Visualization of 3D extreme events in extreme-scale turbulence datasets

Fluid dynamics is an inherently visual subject, and turbulence is well known to possess intriguing yet hard-to-define structural features in space.  The availability of massive datasets from $32768^3$ direct numerical simulations (Yeung etal. J. Fluid Mech.  Vol. 1019 (2025)) is providing unique opportunities for visualizing intricate flow structures but effective imaging  at this scale is also a daunting task.  We have implemented a tractable technique for extreme-scale volume rendering on distributed GPU systems, consisting of mapping a number of planes of data to an appropriate color and opacity that best highlights the structures of interest, then stacking such planar images on top of each other with an offset and compositing them together into a final 3D render.  This technique has been applied to the study of couplings between pressure and vorticity fields focused on 3D sub-domains that possess observed extreme activity.  Apparently, low-pressure structures can be matched with vorticity structures if the vorticity field is spatially filtered, which potentially inspires new ideas about the formation and evolution of extreme events in isotropic turbulence.  Extensions to animation of Lagrangian pressure and enstrophy events are also briefly addressed.