Astronomers have found a wild bunch of planets outside our solar system. Now, scientists are asking what these strange giants can tell us about how planets form.

On the isolated peak of Mt. Hamilton, 20 miles up a tortuous road from San Jose, California, the white domes of Lick Observatory stake out the sky. On most fair-weather nights, the observatory buildings house an astronomer searching starlight for clues to other worlds.

At 4 p.m. on an early February day, 2001, the center slice of the largest dome slides back. In the submarine-like spaces around the 3-meter reflector telescope, Debra Fischer, a member of the world’s most successful planet-hunting team, hurries to prep the equipment for a 12-hour observing stint. She races beneath the massive yellow telescope, pausing briefly to examine the exposed slice of now-violet sky. Light, high clouds—no problem, the telescope can see right through them. Lithe and purposeful, she darts through gray metal doors, down a narrow stairwell, around the corner, farther underground, ducking under instrument arms as she pushes and pulls equipment into place.

Blonde hair tucked behind her ears, Fischer pours liquid nitrogen into the body of the light detector, cooling the key machine to operating temperature. “This is why I became a scientist,” she says, smiling, “to pour steaming, smoking stuff from one vial to another.”

At 6:15, she and the telescope operator begin a meticulous hunt. They point the telescope at one star after another and take a digital photograph of each, snapshots that will tell her whether or not each star is moving. Though she examines the stars, she’s actually searching for planets. The planets themselves are invisible, so Fischer must infer their presence by checking for planets’ effect on their stars. A star without any orbiting planets appears at rest, or randomly fidgeting. A star with planets, however, moves to a beat; it waltzes with each planet. One photo reveals a single sidestep or half turn. Fischer must compile a sequence of photographs, taken over weeks, months, or even years, to be certain the star is circling through an entire box step. Once she’s sure the star is waltzing, Fischer looks to see how big an area its covering and how evenly. The grace and breadth of the star-dance tells her something about the heft and style of the star’s planetary partner. Finally, she and her colleagues proclaim a new planet, called an extra-solar planet because it is in orbit around a distant star rather than our own sun.

The first extra-solar planet, found in 1995, sparked a barrage of discoveries that has shot holes in traditional explanations of how planets form. Before 1995, theorists were forced to make guesses at the processes that shape planets based solely on evidence from our own solar system, with no idea whether our system is rare or common. Extra-solar planets offer a wealth of new data, revealing the results of some of the universe’s other planet-making experiments. Those results hint at a laboratory far different than most scientists imagined. The credos of planet making have fallen by the wayside while proposals that once seemed flagrant speculation are winning the race.

Planet-hunters have discovered a troupe of behemoths that travel wilder paths and dwell far closer to their stars than anyone thought possible. Old-hat theories of planet formation maintained that planets Jupiter’s size form farther from their star, leaving smaller Earth-like planets to reside peacefully in the warm embrace of the Sun. But from the get go planet-hunters found gas giants circling just a hairs breadth from their star’s scalding surface. Sticking to the logic of our solar system, theorists declared that planets could grow no larger than Jupiter. Among extra-solar planets, however, planets two, three, even four times Jupiter’s mass are not obese. And many of the big guys wend long squashed orbits, though theorists once thought the nature of a planet required it to travel a circular orbit.

Trying to figure out how planets form is like reading a mystery novel, said Douglas Lin, an astrophysicist at the University of California, Santa Cruz. “If you read Agatha Christie for the first time, very often you jump to conclusions based on a limited number of facts. That is sort of the way we are attacking nature.” Lin smiles. “Nature is much more imaginative than we are.”

Lin was one of the few theorists to challenge prevailing theories before astronomers found any extra-solar planets. Now, there are enough clues to give astronomers a better idea of what lurks around some five percent of stars—and it appears that the clues bear out many of Lin’s seemingly preposterous ideas.

The discovered extra-solar planets have jumpstarted the imaginations of theorists everywhere. “It’s the golden age of theorists. They can let their theories be fruitful and multiply,” said Alan Boss of the Carnegie Institution in Washington, DC. Boss is another astrophysical pioneer, holding a few theories of his own to explain the wilder residents of our universe.

While theorists struggle to explain the strange qualities of extra-solar planets, planet hunters like Fischer are searching for more. It’s easiest to spot planets in small orbits; the closer a planet is to a star, the harder it’s gravity yanks. Planet-hunters are limited to what they can tell from changes in starlight, so the bigger the motion the better.

To see the star’s movement, Fischer takes advantage of a phenomenon called Doppler shift—the same reason the siren of an approaching fire-truck drops in pitch once it speeds past you. The phenomenon works for both light and sound because both travel in waves. Since a star, like the fire truck’s speaker, is the source of the waves, its motion changes how the light appears to us, as distant observers. Visible light is made up of a rainbow of composite colors. Each color is represents a different wavelength. When the star waltzes toward Earth, it compresses the wavelengths of its light, just as the approaching fire truck compresses the wavelengths of the siren. The shortened wavelengths make the siren sound higher in pitch, but they lend the starlight a blue cast, because blue light is at the shorter end of the wavelength scale. If the star moves away from us, it stretches the wavelengths of the light it shines. For the receding fire truck, stretched wavelengths sound like a lower pitch. For a star, we see a reddish hue because red light has longer wavelengths. The faster the fire truck or star travels toward or away from us, the more dramatic the shift. A star moving sideways relative to the Earth, however, won’t shift in color because it doesn’t compress or stretch the wavelengths of its light. Since the shift in wavelength is our only proof that a star is moving, we can only detect the star’s motion toward and away from Earth, even though a star with an orbiting planet is moving in a circle.

click to see flash animation (89 Kb)

Regular shifts between blue and red light suggest that an orbiting planet is tugging on the star.

Fischer can, however, tell that the star is moving in a circle because its speed toward or away from us changes as it travels the circle. When Fischer maps the star’s color shifts over time, she gleans a plot of the star’s changing speed. If the plot undulates like a sine wave, a line that curves up and down as the star moves toward us and then away, she knows she’s found a planet. No other phenomenon could create such a regular pattern, astronomers believe.

If a planet is too small or travels too large an orbit, its pull is imperceptible. Current instruments are sensitive enough to detect Saturn-mass planets, but only if they are in Mercury-size orbits. Plus, to determine the size of the planet and its orbit, observers must watch an entire rotation. If we were looking at the Sun from space, we wouldn’t have seen Jupiter yet, Fischer said. That would take 12 years, but we’ve only been looking for about eight.

The time restriction explains why researchers are only beginning to identify systems of several planets orbiting the same star. Her team has found “four completely-done, fully-cooked multiple systems,” Fischer said. There are many others where they see the beginnings of a pattern, but haven’t watched long enough to see each planet’s full orbit. “The systems we’re finding appear to be quite full,” she said. “If you want to find another planet, look where there’s already one.”

The first planet found orbiting a distant star is an exception: it lives alone. This planet also shot down the theory that large gaseous planets form far from a star and never venture into its warm inner circle. This planet, named 51 Pegasi B, is a giant the mass of Jupiter. Living in entirely different habitat, 51 Peg skims the surface of its star, traveling once around in just four days compared to Jupiter’s leisurely 12 years.

Jupiter’s home was long thought prime real estate for the making of giant planets. In the planetary nursery— vast spinning disks of gas and dust that swathe young stars—the most raw material sits halfway between the star and the outer edge of the disk. That’s partly because the climate there is cool enough for ice to form, tripling the planet-making ingredients. Astronomers are still relatively certain that big planets cannot form in the veritable inferno of a star’s immediate surroundings. With the discovery of 51 Peg, though, they are pressed to explain how a planet might get there.

Before 51 Peg’s discovery, Lin and others argued that a gas giant might form at a distance and then spiral slowly toward its star. But this theory of migration was only a worry to a “handful of theorists” who thought that if a planet did migrate, it would keep on going, eventually getting engulfed by its star, said Alan Boss. The worry originated in the 1970s when researchers from the California Institute of Technology observed the rings of Saturn dragging on inner moons, slowing them down and causing them to spiral toward Saturn. They proposed that the disk might do the same to young planets. For Lin, who expanded the theory, the idea that disks drag planets became a sort of battle cry, proclaimed at a 1993 meeting in Hawaii.

“Everybody was getting up and saying, ‘There are no planets, there are no planets, there are no planets out there. It’s very difficult to form planets and we’re unique, unique, unique.’ So I said, ‘Maybe we’re not so unique. It’s just that [planets] form. They move. Then they move inside the star—they get gobbled up,” Lin said.

As planets grow, Lin says, they sweep up the gas and dust in their path, clearing a gap that splits the disk into two regions. The inner dust and gas dashes around its star, traveling a tighter circle. This shorter orbit means that the inner disk has more of an energy currency, called “angular momentum,” than the planet. The planet, in turn, has more angular momentum than the outer ring. The greedy outer ring steals energy from the planet. To recoup, the planet steals energy from the inner disk. But as the inner disk loses energy, it sinks toward the star. The outer ring continues to steal from the planet, but now the planet can’t reach the inner disk. The planet, too, spirals slowly inward.

Click to see Flash animation (381 Kb)

This theory of migration is helpful, but it doesn’t explain why the gas giant stopped short of certain annihilation. A few days after the discovery was confirmed, Lin and his colleagues proposed two explanations. Perhaps the star’s magnetic field swept away the dust and gas closest to it, getting rid of the disk’s friction-tug. Once 51 Peg made it into the open, it could safely circle in a drag-free orbit, they proposed.Or maybe once the planet drifted close enough, it would excite tides in its molten star, like the moon does to liquid on Earth, they proposed. The star spins just faster than the planet orbits, so the tides would keep just ahead of the planet. The extra gravitational pull from the bulges would tug the planet forward, they said, fighting the disk’s drag, and keeping the planet from sinking into the star.

The disk-drag migration scenario was bolstered by the discovery of an unusually harmonious two-planet system early this year. In the system, announced by the leaders of Fischer’s team, Geoffrey Marcy at UC Berkeley and Paul Butler of the Carnegie Institution, two planets circle a star 15 light-years away in the constellation Aquarius. The smaller planet, at half Jupiter’s mass, circles the star twice as fast as the larger, which is nearly twice Jupiter’s mass. Lin and his colleagues predicted that planets of that size could exist in just such a partnership, called resonance.

Lin’s team thinks the two planets formed in more spacious orbits, farther from their star and from each other. As the two grew up, they gobbled up new material and picked their orbital paths clean. That process split the disk into three parts, with a disk-ring between the planets. Like all full-fledged planets, each created ripples in this middle disk ring as they circled. If only one planet were present, the ripples would have fizzled out like the waves of a pebble dropped into a pond. In between the two planets, however, the waves crashed into each other, quickly scattering the dust and gas. The middle ring disappeared.

The middle ring now gone, the outer ring stole energy from the large planet, pushing it in. The smaller, inner planet stole from the inner disk and used its new energy to bound outward. Finally, the two came close enough that the smaller planet circled exactly twice as fast as the larger planet. In this configuration their gravity tugged at each other in the same spots in each orbit, and this tugging locked them together. The inner planet circled twice for every orbit of the outer planet. Once together, they slowly migrated inward, like a single planet, to their current spots.

The new discovery sets Lin’s theory ahead of others. Some have suggested, less successfully, that large planets might have formed close to their stars, but it’s unlikely that these two planets could have formed in resonance. Others suggest that certain forces could sling large planets quickly in, but this also would disrupt any resonance.

Still, Lin hasn’t always been right. In the 1970’s, when his current theories were just wild ideas, Lin argued that planets could grow no bigger than Jupiter. Many astronomers agreed, but nature had a surprise in store.

“Many of these objects are so darn massive,” said Carnegie Institution astronomer Alan Boss. Astronomers have found some planets more than 5 times Jupiter’s mass. Two major theories have been proposed to explain how these giants might form. The older theory suggests that bigger planets could form just like their smaller cohorts, building up slowly as dust and ice collides and sticks. The planet grows fatter as it plows through the material in its orbital path, much as a snowball rolling down hill picks up snow and ice. Once this core of matter gets large enough, its gravity sucks in gas. This process, called core accretion, could take more than a million years. By then the disk gas would have slipped from the planet’s grasp and drifted into the star or out to space.

Particles of ice and dust collide and stick together, slowly building up a planet.

Boss thinks monster planets might form differently. The disk surrounding a star can be clumpy, with the dust, ice and gas distributed unevenly. Boss predicts that the gravity of a large enough clump would compress the matter into a planet. The gravity acts much like a child’s hands taking two handfuls of fluffy snowflakes and packing them into a dense snowball. When gravity packs the dust and gas into a planet, a process called disk instability, gas giants Jupiter’s mass or larger might form in less than half the time predicted by the theory of core accretion.

Gravity quickly compresses a particularly dense clump of dust, ice and gas into a planet.

It may be that different sized gas giants form differently. It is difficult to follow the rules of core accretion and come up with a planet even a large as Jupiter, Boss said. On the other hand, predictions show that disk instability best creates planets larger than Jupiter.

Boss predicted disk instability could spawn a system of lumbering giants almost identical to another new system two-planet system announced early this year by Marcy and Butler. The smaller of the two is a whopping seven times Jupiter’s mass. It orbits its sun-like star at one-third the Earth’s distance from the sun. The second and larger planet is between 17 and 40 times Jupiter’s mass, traveling at a distance that is three times the space between the Earth and the sun.

Differences in birth may also explain why many gas giants travel in stretched-out orbits. “We sort of looked in the back of the book to see what the answers were. Saturn has a circular orbit. Jupiter has a circular orbit. Planets must have circular orbits,” Boss said. Actually, of the discovered planets, those farther than a stone’s throw from their star travel orbits that are distortedly elliptical, coming very close to their star and then shooting off into the distance before returning.

Theorists scratched their heads and wondered if something had forced the planets out of a comfortable circular orbit, or if they were born eccentric. Disk instability could give birth to planets in eccentric orbits, Boss said. Core accretion would produce planets in circular orbits. A core-accreted planet, however, could be pushed and pulled into eccentric orbit if it were born in a system with other large planets, Lin says.

The difference between a one-planet system and a multiple-planet system is “like having a one child family versus a three child family,” Lin said. “The three perturb each other constantly.” After a billion-year tug of war, the planets barrel around their star in erratic orbits. If their orbits cross, they may collide to form a doubly massive planet with a looping orbit. Otherwise, their orbits become progressively wilder until one is shot into the sun or out to interstellar space, leaving the remaining sibling to orbit in peace. If gas giants are such grumpy bedfellows, there should be planets floating through space, free from any star. Lin is certain astronomers will find free-floaters.

Maybe they already have. Members of a research team in Tenerife, Spain, discovered free-floating objects in the Orion Nebula. Their possible masses—5 to 15 times Jupiter’s—are low enough that the researchers called them “isolated giant planets” in a paper published on October 6, 2000, in the journal Science. Most scientists think they are too large to have been ejected—usually it’s the smaller, weaker planet that gets kicked out.

The debate circles back to a question that has astronomers quarrelling: what is a planet? Simple, you might say, a planet orbits a star. Unfortunately, some very big things orbit stars, while planets might have been kicked out. The largest objects might be failed stars called brown dwarfs, but it’s tough to tell. Brown dwarfs form like a star—from a collapsing cloud of dust and gas—but never grow massive enough to burn hydrogen. Only when it hits 80 Jupiter masses is a star’s core hot enough, and pressurized enough, to spark major nuclear fusion. Over 13 Jupiter masses, though, a brown dwarf will burn a heavier form of hydrogen before fading into oblivion. Since it’s impossible to look at an object and tell how it formed, some astronomers argue that anything smaller than 13 Jupiter masses should be called a planet.

A division based purely on size might be convenient, Alan Boss said, but it’s not entirely accurate. He has calculated the same core collapse that gives birth to a star could create objects only a few times Jupiter’s mass. So, he says, there are objects a good deal smaller than 13 Jupiter masses that should probably be termed brown dwarfs.

Alan Boss heads a consortium of scientists that are trying to settle on a definition. Personally, Boss said, contemplating definition of a planet has often kept him up late. “I view a star as something that forms through gravitational collapse,” Boss said. “Planets are the epilogue, after the final act of star formation—leftovers that manage to clump together and form a planet by any mechanism. That’s what helps me sleep at night.”

Lin is not as interested in the debate. “Maybe the best way to refer to these objects is not to give them a name, or not to worry so much about a name—like calling so and so a Democrat or a Republican. It’s almost irrelevant these days. Some Democrats sound awfully like Republicans and some Republicans, well, may have certain shades of Democrat in them. Perhaps it’s better just to ask what their properties are.”

He did agree with his UCSC colleague Peter Bodenheimer, who defined a planet as something, below some maximum mass, that was borne from the disk around a star. That doesn’t mean the planet is still circling. They probably won’t settle the debate until astronomers can gaze directly at planets and piece together a planetary life cycle by comparing systems in all stages of planet formation.

Then, they’ll also detect earth-like planets. But do any rocky earth-like planets hide in the neighborhoods we’ve discovered so far? Most of the planets found so far are “gravitational bullies” that would have either stopped an Earth-like planet from forming or would have kicked it out long ago. Barrie Jones of The Open University in Great Britain has pinpointed at least two planetary systems where a rocky planet might have been able to form in the so-called habitable zone of its star—where it is warm enough for water to exist as a liquid. The two planetary systems have gas giants either well inside this habitable zone, or well outside it. There’s a fifty-fifty chance that life might have found a fertile foothold in either system, Jones believes.

Planetary systems like our own cozy solar system, with its circular orbits and svelte planets, could crop up within the next ten years. Will the planet hunters find a plethora, or are we one of a lucky few?

“It’s premature to say that we’re unique, or even odd,” Lin said. So far astronomers have found large, close planets orbiting five to eight percent of stars. “All we can really say is that the frequency of solar systems could be somewhere between zero and 90 percent,” Alan Boss said. “My guess is it’s somewhere toward the middle.”

Two NASA programs seem likely to help astronomers find solar systems like our own. The first is the Space Interferometry Mission, scheduled to launch in 2006. Outside the distorting effects of the Earth’s turbulent atmosphere, a machine called an interferometer will manipulate light waves to reveal a star’s motion, not just a color shift. The mission will be able to see planets five Earth-masses around stars closer than 35 light-years, and it will be able to find them around younger stars whose surfaces are too tumultuous for current methods. Some astronomers think they can do the same thing from the ground by canceling out atmospheric interference.

The grander Terrestrial Planet Finder is still on the drawing board. This project will look specifically for Earth-like planets. It will use a larger interferometer to dim the image of a star, allowing researchers to actually see planetary systems. It is expected to take more than a decade to come to fruition. Either way, theorists must wait to have their hypotheses confirmed or discounted.

Theorists don’t seem frustrated, however. “I like to have a little room to lead the observational effort a little bit—to make some predictions that people can actually test rather than always following what people already found,” Lin said. “You always feel enormously rewarded when even the smallest amount of ‘prediction’ you make is verified, because you thought of it before people revealed it.” Boss is more expansive. “This is the most thrilling part of my life. I’ve been working on this stuff since 1975. For years we had only our Solar System, only one answer. Now we are in a phase where there are new answers. It’s exhilarating.”