Overview

 

Why do we need Virtual Observatories?

Imagine being able to browse a region of sky, imaged in different colours and at several epochs, looking for the rarest of celestial objects. Or suppose you could instruct a software agent to coordinate the processing and statistical comparison of half a dozen remote image archives looking for hitherto unknown patterns in the cosmos, all from the comfort of your desktop workstation.

Sound far-fetched? Australian and international astronomers are working together to build systems which will do just this, systems known as virtual observatories aim to federate the world’s astronomy archives and to provide a grid of computing and storage resources for astronomers. A rich suite of analysis and visualisation tools will exploit the unified archives to enable ground-breaking discoveries – discoveries that could otherwise take years of tedious manual work, or might not be possible at all.

In 2003, researchers from the University of Melbourne, the Anglo-Australian Observatory, the CSIRO’s Australia Telescope National Facility and the University of Sydney launched a collaborative venture called the Australian Virtual Observatory (Aus-VO), partly funded by the Australian Research Council. This followed the establishment of the Australian Virtual Observatory Working Group in December 2000.

Funding has continued with a renewed ARC grant in 2004, involving ten partners. We are also a part of the Australian Partnership for Advanced Computation's National Grid Project. The AustralianVirtual Observatory is one of the key science application projects– driving development of the National Grid.

Our project aims to build a virtual observatory environment linking the archives of all Australian telescopes and which will interface seamlessly with other international virtual observatories – a concept known as interoperability. Our group is an active and founding member of the International Virtual Observatory Alliance (IVOA) and we are establishing active collaborative links with several alliance members.

Virtual observatory standards are the key to interoperability.

While observatories can store telescope data in any format, virtual observatory compliance requires that they be able to describe and provide the data in a common format. In April 2002, the VOTable standard was adopted by the IVOA for the description and transport of a wide range of astronomical data. Australia will contribute World Coordinate System definitions and tools for the next VOTable version and in particular will focus on support for spectral line data.

Observational data and the results of large-scale theoretical simulations are key ingredients in the virtual observatory.

In 2006-07, our group will publish four of Australia’s premiere data archives using virtual observatory techniques:

  • A catalogue of 4,329 galaxies in the southern sky, detected in neutral Hydrogen emission by the Parkes radio telescope.
  • A catalogue of some 22,000 quasars identified spectroscopically and their associated optical light spectra, collected by the 2dF instrument on the Anglo-Australian Telescope.
  • The complete collection of observations made over 13 years by the Australia Telescope Compact Array.
  • A catalogue of more than 100,000 sources from a deep survey of the southern radio continuum sky at 843 MHz made with the Molonglo Observatory Synthesis Telescope.

Storage, access and manipulation of data published to virtual observatories will require access to a grid of high performance computational resources and mass storage systems

We will promote the development of Australia’s computing and network infrastructure so that we can address virtual observatory challenges using grid computing. This paradigm is based on the regionalised distribution of data, storage facilities and computing nodes, expertise, and demand for the data or processed products. Grid computing offers a way to share remote but powerful resources amongst a diverse set of users, with the convenience of single sign-on authentication. We plan to have a small but functional Australian Astronomy Grid by late 2006. Analysis tools and a sophisticated user interface will enable astronomers to use the resources of virtual observatories. Australia has significant experience in data visualisation and the virtual observatory will provide a new home for our tools. Data mining projects are also underway in the Australian community, including work on topics such as machine learning and pattern searching.

Media, sponsors and other interested parties may contact the Project Scientist, David Barnes, at barnesd@unimelb.edu.au or the Lead Investigator, Rachel Webster, at rwebster@physics.unimelb.edu.au.

Below is a brochure on the Australian Virtual Observatory which was presented at the IAU General Assembly 2003.

Page 1 and Page 2