The Johns Hopkins Gazette: February 9, 1998
Feb. 9, 1998
VOL. 27, NO. 21


A Million Galaxies to Be Tracked

Mapping: Sloan Digital Sky Survey will provide astronomical amounts of critical new data

Emil Venere
News and Information

Johns Hopkins Gazette Online Edition

Scientists are completing work this month on the final pieces of an advanced telescope that will give astronomers their first clear view of the large-scale structure of the universe.

One of its chief goals will be to probe the overall framework of the cosmos, a curious pattern in which galaxies seem to be arranged only on the outskirts of huge "bubbles" that contain vast voids of empty space.

Astronomers from around the world will use the telescope to produce the Sloan Digital Sky Survey, an unprecedented catalog of galaxies and other objects that will contain at least 10 times more data than any previous sky catalog. Whereas scientists presently have accurate distance measurements for only a few tens of thousands of galaxies, the Sloan survey will measure the distances of a million galaxies.

The telescope's primary mirror was installed in January. And the final pieces, two spectrographs from Johns Hopkins, will be ready this month for delivery to the observatory, located at Apache Point, N.M., said Alan Uomoto, a Hopkins astrophysicist heading up the team of scientists who designed and built the spectrographic system.

The instruments separate radiation into a spectrum of individual bands, or colors, which reveal important details about the nature of the objects emitting light. Within Sloan's planned five-year lifetime, the survey will contain more spectra than have been published in the history of science.

Astronomers will take their first peek through the telescope in May, and they expect to be collecting useful data in about a year, said James Crocker, director of program management for the survey and associate research director of the Center for Astrophysical Sciences at Hopkins.

Spectra and images from the telescope will reveal the distances and locations of galaxies, enabling astronomers to build the largest three-dimensional map of the cosmos ever created. Such a map will provide important clues to the overall structure and evolution of the universe. For example, scientists are puzzled about how the cosmos developed into its present, clumpy structure from the evenly distributed mixture of gas and dust that existed originally, following the Big Bang.

"There are huge, incredible volumes of space that are empty," Crocker said. "Galaxies seem to be grouped on the outside edges of the voids."

The $1.1 million spectrographic system took about three years to develop. It consists of two identical 650-pound spectrographs, each about 8.5-feet long. They will be mounted at the back of the telescope. For each observation, a round aluminum plate larger than a pizza will be placed between the spectrographs and the telescope. The plates are riddled with tiny holes that line up precisely with the positions of galaxies and other objects being observed. As light comes through each hole, it will be picked up by an individual optical fiber and carried to the spectrographs, which have a combined 640 fibers. The spectra will then be recorded by light-sensor chips called charge-coupled devices, or CCDs, so that they can be stored on a computer.

Astronomers expect to survey about 100 million objects overall, collecting 20 to 40 trillion bytes, or terabytes, enough data to fill 20 million floppy disks.

But the huge volume of astronomical information in the Sloan and other future surveys of similar magnitude poses a scientific dilemma. They will contain so much information that there is no way to retrieve the data quickly. Simple cross referencing could be prohibitively time consuming; astronomers would be unable to readily compare specific galaxies or stars as they appear in different catalogs, which is essential to learn as much as possible about objects in space.

However, astrophysicist Alex Szalay, working with other Hopkins astronomers, has solved the problem by devising a standardized method--a computer program--for breaking the sky into an infinite number of triangular sections. Large triangles contain a nest of smaller and smaller triangles, ad infinitum, enabling astronomers to match up the same regions in different surveys. Each triangle will be labeled for easy cross-referencing.

"Everyone can take as fine a subdivision as they want," Szalay said. "If they are working within one region, it is always part of a bigger triangle."

Catalogs covering the same regions of space often are taken in different wavelengths. Because certain wavelengths reveal different characteristics about an object, astronomers need to compare the same object in various catalogs to get the whole picture.

A universal "sky partitioning" system will become increasingly necessary as more massive sky surveys are produced, said Szalay, who is building the Sloan Survey's database engine--which includes the computer hardware and software--in the university's Department of Physics and Astronomy. The nucleus of the system is a bank of 24 computers provided as part of a three-year grant from Intel Corp.

"This is a new paradigm, a new way of doing science," Crocker said. He likened the advancement to the giant step of going from naked-eye astronomy to the first telescopes.

Such surveys will speed the pace of research by providing fast access to images and data. Ordinarily, astronomers apply for observation time on a telescope, sometimes waiting months for their turn to come and often enduring weather-related setbacks. Then, after observing an object in one range of the spectrum, an astronomer may find it necessary to observe the same object in other wavelengths, requiring another request for observation time and more waiting.

But future sky surveys would enable the astronomer to gain access to a huge catalog of objects almost instantly.

"By looking at data in this new way, we will see new relationships and new things that we didn't see before," Crocker said. "What we are talking about is a production system to do astronomy.

"To answer questions you go into the data base, and you 'data mine.' So you are looking at this data set rather than looking at the sky."

The Sloan project is being managed by the non-profit Astrophysical Research Consortium, made up of the University of Chicago, Johns Hopkins, Princeton University and the University of Washington. Also involved are Japanese astronomers, the Institute for Advanced Study, Fermi National Accelerator Laboratory and the Naval Research Observatory. Funding for the project has been provided by the Alfred P. Sloan Foundation, members of ARC, the National Science Foundation and other sources.