PITTSBURGH- Experts in computer science and astrophysics from Carnegie Mellon University and the University of Pittsburgh are among researchers from 17 institutions nationwide who will share a $10-million, five-year Information Technology Research (ITR) grant from the National Science Foundation (NSF) to help "put the universe on line" via a National Virtual Observatory (NVO)- http://us-vo.org.
The NVO is headed by astronomer Alex Szalay of The Johns Hopkins University and computer scientist Paul Messina of The California Institute of Technology. They and their colleagues will use the latest computer technology and data storage and analysis techniques to unite the astronomical databases of many earthbound and orbital observatories. They hope to gain new scientific insights from the data by making them available in an accessible, seamlessly unified form. The information will not only be available to professional researchers, but to amateur astronomers, students and the general public as well.
At this point, all of the major archives of astronomy data in the United States are signed up to participate in the NVO and links are being created to similar initiatives in Europe and Asia.
"When it comes to exploring the universe, the National Virtual Observatory will be as important as spacecraft," said Andrew W. Moore, Carnegie Mellon's A. Nico Haberman associate professor of robotics and computer science. "The biggest problem in physics today is that we're drowning in data, which is gathering dust in warehouses. The NVO will enable researchers and the public to ask questions of all the data that's been collected."
Moore along with Carnegie Mellon Assistant Professor of Physics Robert C. Nichol and Andrew Connolly, assistant professor of astrophysics at Pitt, comprise the Pittsburgh team. The three have been collaborating since 1998 when they received a $1.6 million grant from NSF to develop algorithms to automate the process of finding information in astrophysical databases like the Sloan Digital Sky Survey.
Just recently, the trio collaborated with four other experts in statistics and robotics from Carnegie Mellon and Pitt to win a $3.4 million ITR grant from NSF to develop new statistical and computational algorithms that will enable researchers to ask more interesting questions of the databases.
"Our connection to the NVO is to put the algorithms we develop into the public domain," Moore said.
"The interesting thing from a computer science point of view is that the question has to be distributed to a bunch of databases around the country. By the time the question is answered, you've touched a trillion points of data. "Our group at Carnegie Mellon specializes in making the statistical kinds of questions go magnitudes faster," he explained. We currently have data on 50 million galaxies. There are some kinds of queries which would previously have taken five years to answer even if we only had 50,000 galaxies to deal with. But now, we've got it down to an hour."
According to Nichol, the Carnegie Mellon and Pitt team members' expertise in the creation of scientific algorithms will enable scientists to more efficiently use the huge infrared, optical, x-ray and microwave astronomical databases that are now available.
"A large part of our contributions are about helping to set scientific standards - providing fast and efficient statistical analysis tools for studying the wealth of information that is now available to astronomers," Nichol said. "There are many versions of the sky out there and we need to have programs that enable us to share this information not just at the terabyte scale but also at the petabyte scale of information."
"In itself, the NVO will be a much richer resource to astronomers than the sum of the individual astronomical databases," said Connolly. "It will provide the first panchromatic view of the night sky that will help astronomers to solve the questions of the origin and fate of our universe." The National Virtual Observatory was inspired by the Sloan Digital Sky Survey, which made data from four different astronomy databases available through one seamless web portal. The NVO challenge is to simultaneously analyze information from several of the dozens of astronomical databases in existence today. Each is organized differently, which makes it difficult to perform analyses concurrently.
The researchers will do their work through a technique called grid computing, which allows scientists in multiple institutions to easily and rapidly share data in the way that power grids around the country seamlessly share their stocks of electricity.
Alex Szalay said the National Virtual Observatory project represents a significant step toward formalizing a third major approach to scientific research that's been growing in popularity and usefulness in recent years. In addition to theoretical models and experiments, there is now scientific exploration through computational methods, which has come to the fore in response to the tremendous volumes of data being gathered in many of the sciences.
"The National Virtual Observatory shows us that the whole paradigm for doing astronomy has changed," said Nichol. "An astronomer like me can sit at his desk a Carnegie Mellon and, via a PC, have the whole universe in front of him."