The Universe en Rose

Science News, March 13, 1999 by Ron Cowen

The view through a better infrared camera

On a clear night, most astronomers would consider themselves lucky to be on Hawaii's Mauna Kea, in the control room of the world's biggest optical telescope. What could be more exhilarating than to know that 3.6 kilometers away, on the summit of this extinct volcano, the giant mirror of the Keck I Telescope is at your command? By all rights, astrophysicist Richard G. McMahon should have been sitting pretty. But on this night, he was frustrated.

It wasn't the weather, and it certainly wasn't the telescope. Keck I had followed McMahon's orders to the letter, recording precious light from a particular patch of the distant cosmos. The problem was the telescope's near-infrared detector. Like all such devices used in astronomy, it was too small to do the job efficiently.

McMahon, based at the University of Cambridge in England, was at Keck I to examine two extremely distant galaxies that lie close together in the sky. Even though the starlit bodies are separated by an angular distance of just 70 arcseconds--a minuscule fraction of the width of the full moon as seen from Earth--the near-infrared sensor was so tiny that it could only record the infrared light from one galaxy at a time.

It was sheer tedium, not to mention a waste of valuable telescope time. McMahon and his colleague Esther M. Hu of the University of Hawaii in Honolulu would have to repeat their observations, slewing the telescope ever so slightly from one galaxy to the other, in order to obtain near-infrared images of both.

On that October night 18 months ago, McMahon took comfort in one thought. He and his colleagues had nearly completed construction of a bigger and more powerful near-infrared camera.

That camera, installed in late 1997 on the 2.5-meter Isaac Newton Telescope on La Palma in the Canary Islands, is the largest infrared imaging device ever assembled. Because of its large size, the camera can detect near-infrared radiation from a given region of the sky more than 50 times faster than the detector that McMahon and Hu used at Keck I.

The Keck I device can only image a region 0.05 percent the area of the full moon. On the Newton Telescope, the new camera, known as the Cambridge Infrared Survey Instrument (CIRSI), views a patch of sky 25 percent the area of the moon. That's more than big enough to encompass both of the distant galaxies that McMahon and Hu had to observe separately.

The difference results primarily from the size of the infrared detector, but there's a second factor at play. Keck I, a bigger telescope, acts as a telephoto lens, magnifying details more than the Newton Telescope does but covering a smaller patch of sky. The large-format CIRSI thus provides a faster way to survey large regions of sky, McMahon says.

"This camera has been designed to find objects that can then be examined in detail on a large telescope," McMahon notes.

The precision optics of Keck I, the Hubble Space Telescope, or the Very Large Telescope, a quartet of 8-m telescopes now under construction on Cerro Paranal in Chile, can do follow-up studies.

CIRSI consists of four sensitive infrared arrays--electronic detectors that are now standard for recording near-infrared light in the laboratory. Both visible-light and near-infrared detectors rely on semiconductors, which convert tiny light signals into electrical currents. Visible-light detectors, known as charge-couple devices (CCDs), consist of layers of silicon. Highly developed because of their widespread use in digital cameras and computer circuitry, CCDs can be made relatively easily and in large sizes.

An infrared array is considerably more complex. One layer of semiconductors--a mixture of mercury, cadmium, and telluride--records the near-infrared radiation, while a layer of silicon bonded to this material reads out the electronic signals. In contrast to a CCD, the signal from each light-sensing picture element, or pixel, must be read separately.

"The electronics are as complex as those in a Pentium chip," says McMahon. Moreover, because the detectors have limited use outside astronomy, "the infrared technology is lagging a decade behind CCDs," he adds.

Each of the CIRSI arrays measures 19 millimeters on a side and contains 1 million individual pixels. The Keck I device consists of a single array of 65,500 pixels.

"The combination of many more pixels and the fact that the scale of the telescope we're using is well matched [to the camera] gives us a giant advantage over Keck," says McMahon.

Although the Rockwell International Science Center in Palo Alto, Calif., supplied the arrays, McMahon and a team of Cambridge scientists designed and put together the trio of computers and the software to operate the camera, a project that took months to complete. In a single night of observing, the camera's ultrasharp scans can produce 30 gigabytes of data, enough to fill 20,000 floppy disks or 50 CD-ROMS.

"We have some spectacular images--the biggest images you can make with a single [near-infrared] camera," says McMahon.