Christopher Healey and his research team have developed a painterly approach to data visualization. His software correlates dimensions of multivariate data to dimensions of visual sensation. The results are stunning and have both artistic and scientific merit.
In the images shown here, Healey’s software has been turned to the task of visualizing data from the explosion of a dying star. His colleagues at NCSU record three-dimensional timesnaps of the supernova, then investigate that data by looking at two-dimensional slices. These slices, at a 500×500 resolution, carry information across the variables of velocity, magnitude, pressure, and density.
The following mappings are in use:
- magnitude → color
- density → stroke size
- Δx and Δy → stroke orientation
Continuity of strokes (image B) indicate a continuous flow, but sharp contrasts in color or orientation (note the right side of image A, at the edge of the shock wave) suggest discontinuties and a swarm of vortices.
Healey and his team have been working on painterly visualizations for several years. His website is flush with papers and examples of how this approach can often yield useful and beautiful results. In particular: Engaging Viewers Through Nonphotorealistic Representations (PDF).