Astronomers have found two new planets circling other stars, thanks in part to a spectrograph that is the most sensitive of its kind. Geoffrey Marcy of San Francisco State University (San Francisco, CA) and Paul Butler of the Anglo-Australian Observatory (Epping, NSW, Australia), who in 1995 discovered several new planets, announced in September that they had increased the number of known worlds outside the solar system to 12.
One of the planets, a Jupiter-sized body, orbits in about three days at a distance 25 times closer to its star than Earth is to the Sun. That star, HD 187123, is 154 light-years away in the direction of the constellation Cygnus. The other, also about the size of Jupiter, is in a more Earth-like orbit and takes 437 days to go around its star, HD 210277. That star is 68 light-years away, toward the constellation Aquarius.
Both planets were found with the High-Resolution Echelle Spectrograph (HIRES) attached to the right Nasmyth platform of the Keck 10-m telescope on Mauna Kea, HI. Marcy and Butler find planets by studying the Doppler shift in the light coming from a star and examining changes in the shift caused by the wobble in a star that results from an orbiting planet. The HIRES allows the scientists to detect a wobble as small as about 2 m/s.
"It's the most powerful of its type in the world," said Steven Vogt, professor of astronomy and astrophysics at the University of California, Santa Cruz, and an astronomer at the Lick Observatory (Santa Cruz, CA).
Vogt designed and built HIRES, as well as the Hamilton Spectrograph at Lick, which Marcy—formerly one of Vogt's graduate students—and Butler used for their last batch of planet discoveries. The HIRES is based on an echelle (French for "ladder") spectrum, after the stair-step-shaped spectrum it produces. Instead of a prism, the light passes through a grating, correction lenses, a mirror, and a field flattener to a charge-coupled-device (CCD) camera. It hits the CCD as a rough square, rather than the long rectangle of a typical spectrum.
The concept is not particularly new, but this instrument uses three 12 × 16-in. gratings strung end-to-end into a single 12 × 48-in. grating, three times larger than any that exists. Vogt said it was very difficult to make this size grating, which had to be lined up within 0.5 arcsec and had to have very high flatness across its entire length.
In addition, the spectrograph uses an extremely achromatic, extremely fast camera that can detect light from 300 nm to 1 µm without refocusing. The camera includes lenses 30 in. in diameter and a mirror 45 in. across. "It's like a telescope in its own right," Vogt said. "It took us several years to make this thing. We didn't think it could be done."
Because of the size of the optics and the sensitivity needed, the camera could use no aspheres or any exotic materials for the glass, Vogt said. The lenses are fused silica. The system is attached to a 2048 × 2048 CCD, with low noise, low dark current, and high linearity. "These are big pieces of optics pieced together," Vogt said.
Improvements needed
But, for all its size, "to find planets, even a spectrograph like this is not good enough," he said. To make it good enough, Marcy and Butler add an iodine-absorption cell near the focus of the telescope. The cell consists of a glass bottle about the size of a beer can containing a vapor of iodine molecules. Starlight passing through the vapor will show an absorption spectrum, which remains steady. The Doppler shift of the starlight can then be seen against the unchanging background of the iodine-absorption spectrum. The technique is sensitive enough to detect a shift in wavelength as small as one part in 100 million.
"You're measuring a star with the accuracy of the speed someone walks in the park," Vogt said. What`s remarkable about the iodine technique, he said, is that it allows astronomers to make measurements below the mechanical limits of the spectrograph itself. Just the weight of the device, or the difference in thermal effects between summer and winter, causes distortions greater than what the planet hunters are trying to measure. And because a planet must orbit its primary several times for the pattern of the wobble to be apparent, the measurements must stay steady over several years.
The Hamilton spectrograph, which found earlier planets, contains only an 8 × 16-in. echelle and has prisms instead of a grating. The attached camera, only one-third the size of the HIRES camera, contains aspheres and is hooked up to a smaller CCD.
The HIRES is not sensitive enough to find an Earth-sized planet. Earth causes the Sun to wobble only at a rate of about 9 cm/s. Even if there were a device that sensitive, currents and movements in the Sun`s atmosphere would mask any movement smaller than about 25 cm/s, Vogt said.
The star survey, which has been underway for about two years, covers 430 stars similar in size to our own Sun, each at least 2 billion years old. Those stars, said Vogt, should be stable enough for the HIRES to find any planets. Vogt said about 5%-6% of the stars the team looks at show planets that are easily spotted, so the search should turn up two or three dozen bodies. In fact, searchers have already identified several candidates but need more measurements to confirm their findings.
"There are a lot of cases where we know there's a planet but we don't yet have enough data to publish," Vogt said.