3D Visualization in Astronomy

The Need for 3D Tools in Astronomy

Viewing your data in 3D allows for a deeper understanding of data rich in structure. Based on "The Need for 3D Visualization in Astronomy" (presented as part of our recent 3D Visualization Tutorial at ADASS, London), we explain why 3D visualization can be crucial when exploring observational astronomy data. It indeed can be used in a vast range of applications far beyond outreach and education.

L1448

Even in data where not all axes are
spatial, 3D visualization helps to discover
and memorize features not accessible by
other means.

Why do we actually do 3D visualization in observational astronomy? The purpose of displaying data in 3D is pretty clear when dealing with data having three actual spatial dimensions. However, only a small fraction of observational astronomical data contains, e.g., information on the "depth" of the objects studied. Instead, in our work we often use two axes to express spatial information (such as right ascension and declination) while the third axis is used to express non-spatial information, such as velocity.

In the particular case of radio astronomy, spectroscopic observations yield the intensity as a function of velocity along the line of sight, I(v), towards various positions in the sky, (α, δ). (Observed frequencies of molecular lines can be related to velocities using the Doppler effect.) Formally, this can be expressed as I(v, α, δ), i.e., the intensity as a 3D function. Correspondingly, this function can - simply from a formal standpoint - be visualized in 3D.

The critical point is that the use of a third dimension in data display enables to visualize at least one more dimension of the data that otherwise has to be folded away. In the above example, the third dimension makes it, e.g., possible to represent the change of the intensity with velocity at all positions in an observed molecular cloud, where one usually only would display the spatial intensity distribution at a given velocity. This enables the easy discovery of 3D features otherwise barely extractable from series of 2D snapshots.

Beyond visualization, 3D representation of non-spatial data also helps the investigator to develop an intuitive understanding of the data. This relies on the human recognition, which inherently works in three dimensions. To give an example, it is much simpler to memorize a 3D model of the velocity structure of a molecular cloud than to memorize a series of several hundred 2D images of the cloud at increasing velocities.

Our 3D visualization research is driven by recent surveys that are too large to be explored using traditional methods. To give examples, some of the surveys of the COMPLETE project have several 105 pixel and have several hundreds of spectral channels per pixel. Therefore, tools frequently in use in the past decades (e.g., "channel-maps") cannot be used anymore because of excessive data sizes.

Summary handouts: "The Astronomical Medicine Project" & "Getting Started with 3D Software"

3D Visualization Tutorial (co-organized by AM)

This tutorial was organized by ourselves, Nick Holliman (eScience Research Institute at Durham University), Ugo Becciani (Ist. Naz. di Astrofisica – Osservat. Astrofisico di Catania), and Claudio Gheller (Consorzio Interuniversitario di Calcolo).

Jens Kauffmann (IIC & CfA): Tutorial Introduction (PDF)

Jens Kauffmann (IIC & CfA): The Need for 3D Visualization in Astronomy (PDF)

Michael Halle (IIC & BWH): Commonalities between Medical and Astronomical 3D Imaging (PDF, PPT)

Nick Holliman (Durham Univ.): Binocular Imaging (PDF)

Michelle Borkin (IIC) & Jens Kauffmann (IIC & CfA): 3D Data Exploration Examples (see our 3D Slicer and OsiriX pages)

Claudio Gheller (CINECA): A Glimpse on VisIVO (PDF; also see http://visivo.cineca.it/)

Chris Fluke (Swinburne Univ.): A Glimpse on S2PLOT (PDF, PPT)

Please also refer to our Research Projects page and our list of 3D software compiled during the ADASS meeting.

Invited and Contributed Talks

Alyssa Goodman (invited; IIC & CfA): Astronomy + Medicine = Understanding (PDF)

Michelle Borkin (IIC): 3D Visualization and Detection of Outflows From Young Stars (PDF, PPT)

ResearchBackground (last edited 2011-02-18 17:26:13 by borkin)