GALILEO would have loved it: for the first time, astronomers can study a variety of solar systems. And this growing collection of planets is full of surprises. Gas giants larger than Jupiter whip round their stars in tiny orbits. Others trace eccentric paths, looping about like loose cannonballs. Only a few planets circle within their stars鈥� habitable zones-where water and perhaps life could exist. Planets probably orbit more than one in five Sun-like stars, but cosy, stable solar systems like ours may be rare.
As the planetary zoo fills in, theorists are competing to answer the essential questions. How do planetary systems form? How long does it take? How many stars grow planets? How many systems survive their wild adolescence and reach stable maturity? Most importantly, is Earth unique, or do many planets exist where life could evolve?
Stars and planets are born simultaneously, as the German philosopher Immanuel Kant first outlined in 1755. The story starts with a vast, slowly rotating molecular cloud-cold gas and dust-within a galaxy. As gravity pulls a denser region together, the gas rotates faster. Then, rather like a spinning lump of pizza dough, it spreads out and thins. The result is a primordial solar nebula-a rotating disc that鈥檚 99 per cent gas, peppered with motes of dust.
Advertisement
RECIPE FOR PLANETS
Take a cloud and stir
For several million years, material rains into the disc and spirals inward. We now know that if the core gobbles up enough mass, fusion begins, releasing huge amounts of energy. A star is born.
Discs with less mass may fail to form stars. They can produce brown dwarfs, which weigh from 13 to 75 times as much as Jupiter, the most massive planet in our Solar System. Normal stars may shine for billions of years, powered by hydrogen fusion, but brown dwarfs flare briefly by fusing deuterium, a heavy form of hydrogen. Then, over 100 million years or so, they cool and fade from view.
While a star is forming, grains of dust and ice start to clump together, or accrete, in a thin layer within the disc. The grains gradually accumulate more material, forming chunks. Some of these, called planetesimals, eventually reach a kilometre in diameter. This is vital: their gravitational grip is now strong enough to pull in nearby material.
From the start, planet formation is a winner-take-all race. Collisions sometimes destroy growing planetesimals but, on the whole, larger bits grow faster than smaller ones. And, like snowballs rolling downhill, the bigger they get, the faster they grow. Within a million years or so, the winners gobble up all the solid material within their gravitational reach. Neighbours orbit far enough apart for collisions to become rare. These isolated protoplanets may become massive enough-10 to 20 times as heavy as Earth-to attract and devour huge quantities of gas, carving wide gaps in the disc. They quickly balloon into gas giants like Jupiter and Saturn (see Figure 1).FIG-mg22566801.JPG
This model seems to explain our Solar System very well (see Inside Science No. 30). Near the Sun it was too hot for ice to form. Rocky planets accreted slowly, and remained lightweights, like Earth. Out beyond the so-called 鈥�snow line,鈥� the solid cores of Jupiter and Saturn feasted on dust plus ice, and so grew faster and bigger. Ice giants Uranus and Neptune, which lack massive atmospheres, may have formed after Jupiter and Saturn hoovered up the giant鈥檚 share of available gas. Tiny Pluto formed farther out, where material was scarce.
Astronomers now know our Solar System extends far beyond Pluto. They count Pluto as just one of perhaps 100 000 mini-planets in a distant ring around the Sun called the Kuiper Belt. And even farther out a shell of frozen debris, the Oort Cloud, is thought to surround the Solar System, revealed by the newborn comets it occasionally sends our way.
It鈥檚 a tidy explanation. But it fails badly when it comes to the planets that orbit other stars. On 6 October 1995 Swiss astronomers Michel Mayor and Didier Queloz discovered the first planet outside our Solar System, a find confirmed four days later by planet hunters Geoff Marcy and Paul Butler. The planet circles 51 Pegasi, a Sun-like star 50 light-years from Earth, in the constellation Pegasus.
This epoch-making discovery immediately raised new questions. No one, however, was fazed by the planet鈥檚 great size. The minute wobbling of the star, caused by the planet鈥檚 gravitational tug, gave its presence away. Only a giant planet could jiggle its star enough to be detected from Earth (see 鈥淐elestial sleuthing鈥�). The surprise was the planet鈥檚 orbit. It circled 51 Pegasi every four days in an orbit just one-twentieth the size of Earth鈥檚.
In fact, more than a third of the planets discovered so far are like 51 Pegasi鈥檚-and unlike anything in our Solar System. Dubbed 鈥�hot Jupiters鈥�, they are gas giants, but they zip around in orbits far smaller than Mercury鈥檚. Most theorists do not believe they could form so close to their stars, as intense radiation and blasts of charged particles would blow away the raw materials they need. But if they grew farther out, why are they now in star-skimming orbits?
Even more planetary systems show signs of a chaotic adolescence. Most extrasolar planets trace highly eccentric, or oval-shaped, orbits. But almost all the planets in our Solar System follow circular paths. That鈥檚 how a planet growing in a rotating disc should behave. So what鈥檚 creating all those planetary oddballs?
Because these giant planets migrate or orbit through the regions near their stars, it鈥檚 likely that Earthlike planets would be swallowed or elbowed out of the system. That鈥檚 made some astronomers think that even if planets abound, systems like ours are rare.
As more new planets have been discovered, two competing theories-championed by Douglas Lin at the University of California, Santa Cruz, and Alan Boss at the Carnegie Institution of Washington-have emerged to explain how giant planets form and survive.
Lin was one of the few astronomers not astonished by the planet circling 51 Pegasi. Fourteen years earlier, his mathematical models had led him to predict that giant planets would form readily. But the models also revealed tide-like forces that inevitably shunted giant planets inward toward their stars. Until 51 Pegasi, nobody believed him. Now theorists are trying to specify conditions that would let planetesimals and protoplanets stay in orbit and not spiral into their stars.
Building on work by the Russian theorist V. S. Safronov, Lin believes dust and ice must accrete into massive planetary embryos before burgeoning into gas giants. His calculations show that a core about ten times as massive as Earth turns into a ravenous gravitational predator. It quickly devours all the gas within reach, and would consume the whole disc if it could. Once the core reaches that criical mass, it can wrap itself in a massive atmosphere of hydrogen and helium in a mere 10 000 years.
Lin鈥檚 models also neatly explain why giant planets stop growing. As they develop, they sweep out a gap in the disc. They do this by draining energy from faster-moving material inside their orbits, so that material falls inward. They also transfer energy to the slower-moving gas outside their orbit, pushing it outward. The Hubble Space Telescope (HST) has snapped pictures of just such a gap in the disc round a star in the constellation Libra.
Recent computer studies show that even after carving a gap, a planet can still grow, fed by a stream of gas from the outer disc. But the more massive the planet, the wider the gap. The ravenous giant eventually grows so large, it cuts off its own food supply. This may explain why planets many times larger than Jupiter are rare (see Figure 2).FIG-mg22566802.JPG
There鈥檚 a problem with Lin鈥檚 core-accretion model: the two-step process may take too long. By the time a sufficiently massive core has formed-after a million years or more-the disc may have lost most of its gas. However, Boss believes giant planets don鈥檛 need those slow-growing solid cores at all. His models show that if a disc is cool and massive enough, it will break up spontaneously into blobs that contain enough matter to be bound by gravity. Mutual attraction quickly compresses the gas, creating a planet like Jupiter in less than 100 000 years.
It now appears that in a cool, massive disc, the first giant planet may spawn others. Detailed computer simulations by Philip Armitage of the Canadian Institute for Theoretical Astrophysics show a giant planet spinning off twin wakes of high-density gas. These spiral inward and outward through the disc as shown in Figure 1. The bands are unstable and break up into segments massive enough to coalesce into planets. In Armitage鈥檚 models, several gas giants form at nearly the same time, at 5 to 10 astronomical units from the star. (An astronomical unit is Earth鈥檚 average distance from the Sun-about 150 million kilometres.) If Armitage is right, Jupiter was the kingpin of our Solar System, and Saturn, Uranus and Neptune coalesced from its wake.
Theoretical pitfalls
Ingenious solutions
Competition aside, all theorists face a number of roadblocks when they try to explain how planets form and survive. For one, tiny grains of ice and dust have to stick together to grow into planetesimals. But their feeble gravity is far too weak do this. Worse, at the temperatures out where giant planets probably form-colder than -175掳C-chunks of ice are bouncier than rubber balls, and even dust motes won鈥檛 stick together.
One team of researchers think they can dodge this obstacle. They discovered that when ice picks up a frosty coating, especially one of organic molecules such as methanol, it becomes 100 times stickier. Since such chemicals abound in molecular clouds and protoplanetary discs, there should be enough of this organic Velcro to help tiny grains accrete into planetesimals a kilometre across.
Still, a growing planet鈥檚 troubles are far from over. Bodies a kilometre or more in size don鈥檛 gently drift together. They slam into each other, often with enough energy to shatter to bits or to blast huge craters or. Modellers wonder how any promising planets avoid being destroyed.
It could be that planetesimals have an unusual consistency. Working from close-up images of asteroids, researchers now think embryonic planets are porous-piles of gravel rather than monoliths. A collision that would shatter a rigid body can use up all its energy just rearranging the rubble. So planetesimals may grow past the magic size of 1 kilometre because they鈥檙e just a collection of fragments: they can be shattered, but have enough gravitational 鈥済lue鈥� to patch themselves back together again.
What of the final phase of planet formation? This is an even bigger drama of gravitational jousting and titanic collisions. Computer simulations show dozens or hundreds of planet-sized competitors disrupting each other鈥檚 orbits or smashing together. Some collisions are so energetic they melt entire planets. When the free-for-all ends, radically different planetary families can result. Some contain many small planets, others just one lonely giant. Almost all such simulated planets start life in eccentric orbits.
A similar smash-up probably created the Moon. Astronomers believe a planet at least as massive as Mars rammed Earth 4.5 billion years ago. The Moon condensed from a ring of debris blasted into orbit by this spectacular crash.
Like Armitage, Fred Rasio of the Massachusetts Institute of Technology thinks it is common for several giant planets to form at the same time in closely spaced orbits. He believes this causes a life-and-death game of gravitational bumper cars, flinging the losers into space within 100 000 years or so. Only a few planets survive. Their wobbly orbits reflect the battles they鈥檝e been through.
Theorists now suspect that in our Solar System, Jupiter and Saturn may have flung their smaller siblings Uranus and Neptune into tilted, eccentric orbits. Over time, leftover material in the outer part of the protoplanetary disc could have eased them into their present paths.
While theorists refine their models, observers are filling in the details of many of the stages of star and planet formation.The HST has captured dramatic images of huge pillars of gas and dust in the constellations Hercules and Orion. Dozens of newborn stars called proplyds are emerging from the clouds, wrapped in dense protoplanetary discs. They鈥檙e ready to run the planet-building race.
Observers first detected discs in the early 1980s from faint infrared radiation emitted by dust within them. Currently, infrared surveys indicate that 60 per cent of young stars have discs and could grow planetary families.The HST recently sent back images of dusty rings round several nearby stars. The smallest pack perhaps 100 Earth masses of dust and ice into a disc smaller than Saturn鈥檚 orbit. The largest contain enough material to make thousands of Earths, and they measure 25 times larger than our Solar System. To compare, our family of planets contains about 500 Earth masses of material.
Planet-forming discs have now been observed at several stages of evolution. Young discs are massive and symmetrical. After a few hundred thousand years, many go doughnut-shaped, their centres probably swept clear by developing planets. A few discs, such as one round the star Beta Pictoris, appear warped, perhaps by the gravitational pull of one or more planets. Others, including the disc girdling the Sun-like star Epsilon Eridani,have gaps and bright spots, possibly from dust surrounding new planets.
Discs spawn planets surprisingly quickly. Studies of young, low-mass T Tauri stars show that once planets start to form, the discs have all but vanished after a few million years. Our Solar System probably developed quickly too, about 4.6 billion years ago.
Discs thin with age, but may not vanish altogether. Stars older than a few hundred million years have faint, distant rings. Like the Sun鈥檚 Kuiper Belt, these probably store rubble left over after planets form.
Earthlike planets
Signs of life
All the extrasolar planets found so far are gas giants. Like Jupiter and Saturn, they have no surface where life could evolve. Many bake in tight orbits around their suns. More than half follow eccentric paths that would obliterate any smaller terrestrial planets orbiting near their stars. A very few stars-such as HR 810 and Upsilon Andromedae-harbour planets within their habitable zones (see Figure 3). Upsilon Andromedae is the first of two stars, other than the Sun, found to possess fully fledged planetary families (see Figure 4). Observers now believe that when there鈥檚 one planet, they鈥檒l likely find more.FIG-mg22566804.JPG
Astronomers disagree about how common Earth-like planets may be. Some, including the leading planet-finders Marcy and Butler, are optimistic. They believe that many stars lacking hot Jupiters harbour smaller planets. They point out that current techniques for spotting planets work best for giants close to their suns, and may miss systems like ours.
Lin thinks that after the last migrating giant has fallen into a star, a disc can still gather enough material to form a generation of terrestrial planets safe from marauding giants. But Armitage does not believe many terrestrial planets survive. In his models, planet formation is a violent, chaotic process. He thinks gravitational battles between competing giants likely destroy small planets like Earth.
Jack Lissauer of NASA has tried to weigh up the factors in favour of, or against, the formation of habitable planets. He points out that massive stars burn out too fast for life to develop. Low-mass stars shine for many billions of years, but are prone to potentially life-destroying stellar storms. Even around stable, Sun-sized stars, Lissauer notes, planetary systems may form, only to fly apart in the throes of gravitational battles. And in systems where no giant planets form, Earth-sized planets would suffer a continuous rain of destructive collisions. A giant planet mops up debris, just as Jupiter did with Comet Shoemaker-Levy 9.
Despite the many factors that could keep habitable planets rare, Lissauer still believes that planets like Earth are numerous, given the billions of Sun-like stars in the Milky Way. These include rocky planets with stable orbits and cosy temperatures, where oceans form and life could emerge.
The argument will continue until observers can detect Earth-sized planets and study their atmospheres for chemical signs of life. Happily, space-based instruments able to do just that may be sending images and spectra back to Earth within the next twenty years (see 鈥淰isions of the future鈥�).
Four hundred years ago, the Italian Renaissance philosopher Giordano Bruno was burned at the stake in Rome, in part because he argued that Earth was not unique. In 1584 he had written:
鈥淚nnumerable suns exist; innumerable earths revolve around these suns in a manner similar to the way the seven planets revolve around our sun. Living beings inhabit these worlds.鈥�
Astronomers have now shown that Bruno was right when he intuited the existence of innumerable planets. In the very near future we will learn if he was also right about other Earths-and about life on those distant worlds.