Scientific Visualization in Astronomy: Towards the Petascale Astronomy Era

This week’s post is about a fascinating paper by Amr Hassan and Christopher J. Fluke of Swinburne University of Technology in Australia on “Scientific Visualization in Astronomy: Towards the Petascale Astronomy Era”

Four common visualization methods applied to a cosmological N-body dataset. Scattered point data (top left); Gaussian “splats”, using transparent-mode texture blending (top right); three represen- tative isosurface levels with density increasing from red-orange-yellow (bottom left); and texture-based volume rendering with “heat” colour map increasing from black through red and yellow to white (bot- tom right). Data courtesy Madhura Killedar (University of Sydney).  From Hassan and Fluke

Four common visualization methods applied to a cosmological N-body dataset. Scattered point data (top left); Gaussian “splats”, using transparent-mode texture blending (top right); three representative isosurface levels with density increasing from red-orange-yellow (bottom left); and texture-based volume rendering with “heat” colour map increasing from black through red and yellow to white (bot- tom right). Data courtesy Madhura Killedar (University of Sydney). From Hassan and Fluke

They provide a thorough review of visualization applications in astronomy, and identify six grand challenges that they think must be met to meet astronomy visualization needs in the era of Petabyte scale data driven science. They are:

  • Support Quantitative Visualization.
  • Effective Handling of Large Data Sizes.
  • Discovery in Low Signal-to-noise data.
  • Better Human-Computer Interaction and Ubiquitous computing.
  • Encourage adoption of 3D scientific visualization techniques.
  • Better workflow Integration.

I strongly recommend this paper to anyone interested in the future of visualization in astronomy.

Advertisements
This entry was posted in astroinformatics, Astronomy, cyberinfrastructure, Data Management, High performance computing, Parallelization, visualization and tagged , , , , , , , . Bookmark the permalink.

6 Responses to Scientific Visualization in Astronomy: Towards the Petascale Astronomy Era

  1. Stefan, says:

    They apparently not well aware about available VR solutions; it has been out for a while – with the latest Intel/AMD CPUs 4K cube may be interactively rendered at very high quality and resolution on $5K machine…

  2. Stefan says:

    >do have any good references on this topic?

    Some image samples:
    http://www.fovia.com/gallery.php?view=Mummy

    In fact any 3D or 4D scalar fields are fine. The technique is adaptive – means it keeps sampling densities according to their contribution to final image as result the time complexity is a view-scene dependent with logarithmic dependency on size of volumetric data resulting exceptionally high interactive rendering quality. 4k cube of scalars may be rendered interactively on system with 180GB available RAM; apparently top CPUs setup is a requirement for such data sizes. It is commercial but they offer the research license as well.

    To apply for research license:
    http://forms.fovia.com/forms/register.php?t=research

    Stefan
    Disclosure; I associate with them.

  3. astrocompute says:

    Terrific – thank you!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s