3D modeling of horseshoe clouds with VGSTUDIO and COMSOL

Published on May 24, 2026 | Translated from Spanish

Horseshoe clouds are atmospheric formations as ephemeral as they are fascinating. Their inverted U-shaped structure, generated by a horizontal air vortex that curves upon encountering ascending thermal currents, disappears within minutes. To study them, scientific visualization offers tools capable of capturing their dynamics and geometry in three-dimensional environments, allowing analysis of phenomena that direct observation cannot retain.

3D modeling of horseshoe clouds with VGSTUDIO and COMSOL for atmospheric scientific visualization

Simulation of the horizontal vortex and ascending thermals 🌪️

The process begins in COMSOL Multiphysics, where a bio-electromagnetism model adapted to fluid dynamics is configured. An initial velocity field simulating a rotating horizontal vortex is defined, with variable pressure and temperature parameters. Ascending thermal currents are modeled as heat gradients that deform the vortex, curving it upward. The resulting data is exported to Volume Graphics VGSTUDIO MAX, where the cloud is reconstructed as a 3D volume. Here, density and opacity thresholds are applied to visualize the horseshoe shape, adjusting lighting to highlight internal turbulence. Finally, Materialise Mimics allows segmentation of areas with higher vorticity, creating meshes that facilitate structural analysis.

Capturing the ephemeral to understand the eternal ⏳

These clouds are a reminder that nature operates on scales that escape our senses. By modeling them in 3D, we not only document a rare event but unlock its hidden mechanics. The combination of simulation and visualization transforms a fleeting instant into a tangible object of study, demonstrating that technology not only records but reveals the underlying logic of the most elusive phenomena.

What methodology was employed in VGSTUDIO and COMSOL to validate the fluid dynamics of horseshoe clouds, and how was it compared with real atmospheric data given their ephemeral nature?

(PS: modeling manta rays is easy, the hard part is making them not look like floating plastic bags)