Dusty disks may reveal hidden worlds; on the trail of extrasolar planets.

Like a crab scuttling through sand, an orbiting planet leaves a telltale trail in the dust surrounding its parent star. Astronomers began scouring nearby stars for such trails nearly 2 decades ago, but telescopes provided only fuzzy images. With the keen-sighted instruments available today, however, the dust trails are coming into sharper focus, opening the way to finding and characterizing the properties of hundreds of planets beyond the solar system.

Doughnut-shape patterns of dust, or debris disks, are much easier to detect than planets themselves because the disks have a much larger surface area. Recent pictures of debris disks around 10 or so nearby stars show gaps, arcs, rings, warps, clumps, and bright patches.

Some of these structures are serving as guideposts, helping astronomers home in on planets they hope to directly image. Others suggest the presence of planets that can't currently be detected by other means.

Although the most successful planet-detection protocol so far has been the measurement of the slight wobble that extrasolar planets induce in the motion of their parent stars, the technique has limitations. It fails for stars that rotate rapidly, have insufficient mass, or lie at great distances from their parent star.

"Debris disks can definitely point you to the existence of a planet that you might not be able to detect in any other way," notes Alycia J. Weinberger of the Carnegie Institution of Washington (D.C.). Although she cautions that astronomers haven't proved that the features they see have been sculpted by planets, theorists are using the newest data to refine estimates of the mass of proposed planets, their distance from their parent stars, and the shapes and tilts of their orbits.

Trying to discern the properties of planets through the patterns they generate within debris disks is an arduous task, says Charles A. Beichman of NASA Jet Propulsion Laboratory in Pasadena, Calif. But the increasing resolution of images taken at near-infrared and longer wavelengths is making the job easier.

DUSTY DOINGS Astronomers have known since 1984 that searching for planets can be a dusty business. That's when they began analyzing data from the first spacecraft to survey the entire sky at mid- and far-infrared wavelengths. Launched in 1983, the Infrared Astronomical Satellite (IRAS) detected much more infrared light around the brilliant star Vega than could be accounted for by the star's own radiation. About 15 percent of the stars surveyed by IRAS showed a similar excess of infrared radiation.

Researchers deduced that dust must encircle such stars, soaking up the stars' ultraviolet and visible light and reradiating the energy at longer, infrared wavelengths. A swirling distribution of dust around a star flattens into a debris disk. These disks are considered the remains of much denser, gas-rich rings of dust, known as protoplanetary disks, that swaddle infant stars. During the first few million years of the life of some stars, the gas, dust, and ice within a protoplanetary disk clump together to form planets and smaller bodies, such as asteroids and comets. These processes drastically thin out the material in the disk.

Calculations indicate that most of the grains in a debris disk quickly spiral toward the star or get ejected from the system. That means that for the ring to survive, the dust must continuously be replenished. Beichman cites three possible sources of this dust: asteroids smashing into each other, comets evaporating, and material drifting in from the outer edge of the original protoplanetary disk.

Although IRAS lacked the resolution to image debris disks, the excess infrared radiation it detected sparked a hunt for what came to be called Vegalike stars. Astronomers then used larger telescopes to examine these stars for evidence of planet-bearing debris disks.

In 1984, researchers obtained their first image of a debris disk. It surrounds Beta Pictoris, a star just 63 light-years from our solar system. Spectroscopy later revealed that the grains of debris are about the size and composition that scientists predict would be produced by colliding asteroids or comets.

During the 1990s, images taken by ground-based telescopes and the Hubble Space Telescope revealed a 3 [degrees] tilt, or warp, in a section of the disk. Some astronomers suggested that the warp was the handiwork of an unseen planet (SN: 2/3/96, p. 77), but others calculated that the gravity of a passing star could have just as easily created the tilt, which lies about 70 astronomical units (AU) from the star. One AU is the distance between Earth and the sun.

DOUBLE WARP Characteristics of new features detected in debris disks surrounding Beta Pictoris and Vega, as well as several other stars, make a planetary origin for the features seem likely. Several groups presented their initial findings in January at a meeting of the American Astronomical Society in Washington, D.C. Some of the researchers reported additional findings last month in Tucson at a symposium on debris disks and the formation of planets.

At the January meeting, two groups presented new infrared observations of Beta Pictoris' debris disk. Weinberger, along with Eric E. Becklin and Ben Zuckerman of the University of California, Los Angeles, used the Keck I telescope atop Mauna Kea. Zahed Wahhaj and David W. Koerner of the University of Pennsylvania, along with Koerner and their colleagues, used Keck II.

Both teams saw evidence within the disk of an additional warp, this one located much closer to the star. The warp lies between 5 AU--Jupiter's distant from the sun--and 30 AU, roughly Uranus' distance. Moreover, it tilts about 14 [degrees] out of the plane of the disk, in the opposite direction of the warp that had been previously seen. Two warps tilted in opposite directions make for a more compelling case that the gravitational tug of one or possibly two planets is responsible for the features, says Weinberger.

"We've seen disk features before that could be due to planets ... but most of these were discovered far outside the region where planets reside in our own solar system," notes Koerner. In contrast, the new images reveal inclinations in the orbits of dust grains that resemble the orbits of planets in our own solar system, he notes. The Keck images may therefore provide "circumstantial evidence for a similarly organized planetary system," Koerner suggests.

However, says Weinberger, "some clever person ought to think about how to do this without a planet before we rush to judgment that it is a planet."

Both teams also see evidence of bands, or rings, of more concentrated dust embedded within the inner part of Beta Pictoris' disk. Other researchers, observing the star with the Hubble Space Telescope, had previously reported that the outer part of the disk is composed of a sequence of dust rings.

Rings of dust are intriguing because they might constitute a larger-scale replica of Saturn's rings. Just as Saturn's rings are corralled by the gravity of closely orbiting moons, the rings within Beta Pictoris' debris disk may owe their slender shape to the shepherding action of one or more planets. A single, narrow dust ring that surrounds another star, HR 4796A, may also have a planetary origin (SN: 1/9/99, p. 20).

Pawel Artymowicz of Stockholm Observatory notes that dust rings don't require the presence of a planet--especially if the dust is mixed with a substantial amount of gas. A balance between two opposing forces--the outward push exerted on dust by the star's radiation and the drag exerted on the dust by neighboring gas particles -- could in theory maintain a ring of dust at a fixed location.

At the April meeting, astronomers debated whether enough gas resides in Beta Pictoris' disk to have such an influence. Other debris disks show no evidence they contain substantial amounts of gas, says Weinberger.

Irregular features that have been observed within debris disks, such as rings off-center from the star or arcs that don't form a complete ring, are harder to explain without invoking a planet, Weinberger and Artymowicz agree. At the meeting last month, Weinberger showed new evidence that one side of Beta Pictoris' disk is brighter than the other. Spectra taken by her team suggest that the difference in brightness, which is strongest at the shortest infrared wavelength that Weinberger and her colleagues observed, can't be explained by differences in the composition or amount of dust.

Hot dust, which tends to be composed of the smallest particles, glows more brightly at shorter wavelengths than at longer ones. Weinberger therefore speculates that dust on the more luminous side of Beta Pictoris contains a higher abundance of fine particles. A lurking planet, she proposes, could generate such an abundance of small grains by generating collisions between rocky bodies in the neighborhood.

DUST CLUMPS Theorists haven't yet deduced the exact mass or orientation of any planet that may orbit Beta Pictoris, but new images of the debris disk surrounding Vega have prompted some astronomers to calculate several properties of a proposed planet circling that star. The first hint that Vega's disk might contain clumps of dust surfaced 4 years ago, when astronomers observed the disk at submillimeter wavelengths using the James Clerk Maxwell Telescope atop Mauna Kea (SN: 4/25/98, p. 260). Submillimeter observations make it easier to image disks because stars radiate only dimly at these wavelengths.

Having observed Vega with higher-resolution telescopes, two teams of astronomers reported at the January meeting that they had confirmed that the star's debris disk contains prominent clumps of dust. A group led by Koerner used the Owens Valley Radio Observatory near Bishop, Calif. Another team, which includes David J. Wilner, Matthew J. Holman, and their colleagues, all of the Harvard-Smithsonian Center for Astrophysics in Cambridge, Mass., relied on the Plateau de Bure Interferometer in the French Alps.

Wilner's team found one clump located 60 AU to the southwest of the star, and another 75 AU to the northeast. Computer simulations suggest that both features could result from a single planet, several times as massive as Jupiter, in an elongated orbit. On average, the planet would reside about 30 AU from the star, Holman says. That's far enough away from Vega that planet hunters have a chance of directly imaging the planetary body.

Holman notes that a gravitational resonance with the proposed planet would trap the dust in the observed locations. In this model, every time the orbiting clumps complete one pass around Vega, the planet would complete an integer number of passes.

Many of the planets detected by the wobble method also have elongated orbits, but they all lie close in, less than 3 AU--roughly the distance between the sun and Mars--from their parent star. The clumps in Vega's disk could be signs that a planet with a highly elongated path resides at a distance from its parent similar to that of the solar system's outer planets, Holman and his colleagues suggest in an upcoming Astrophysical Journal Letters.

Some theorists have proposed that Neptune once had an equally elongated orbit that became more circular after gravitational interactions with the Kuiper belt, an icy reservoir of comets that lies beyond Pluto. "Perhaps we are witnessing a similar phase in the evolution of the Vega system," Holman and his colleagues note.

For all their efforts, astronomers have found fewer than a dozen debris disks. That's why planet hunters are eagerly awaiting the launch next year of NASA's Space Infrared Telescope Facility. Although the mirror of this orbiting telescope is too small to image debris disks, it will have 10 times the sensitivity of IRAS to excess infrared radiation, an indication of debris disks.

The observatory is expected to generate a catalog of hundreds of stars, some of them no brighter than the sun, likely to have debris disks. In just a few years, as astronomers train their high-resolution telescopes on these disk-enshrouded stars, they may discover dozens of dusty trails left by planets.