Almost 5 billion years ago, an insignificant cloud of gas on the fringe of an ordinary spiral galaxy began to collapse. Its centre started to glow as a star formed; the disc of gas and dust around it coagulated into smaller bodies in orbit. Such a scene had been played out billions of times before in this and countless other galaxies. But this was a special occasion in one respect: intelligent beings would one day emerge on one of these smaller planets. They would call their home Earth, and would wonder how their star - the Sun - and its family of planets had come into being.
The creation myths of the world's religions naturally linked the creation of the Earth with the birth of the Universe as a whole. Since the 17th century, astronomers have tried to take a more rational stance, but they have faced one serious problem: there is only one Solar System. While other scientists can compare and contrast different specimens to find what is normal and what is a deviation, we do not even know if planetary systems like ours are common in the Galaxy. If many stars have planets, then the birth of planets is a natural by-product of the formation of a star and a theory of planet formation should follow on from theories of star birth. At the other extreme, our Solar System might be unique, created by some freak event.
At the end of the 19th century, astronomers thought that the planets owed their existence to a chance brush between the Sun and another star. The close encounter drew a streamer of gas from the Sun and this matter condensed into the planets. Michael Woolfson, of the University of York, has developed this idea and proposes that the gas streamer came from a distended star near the young Sun. But the pendulum of opinion has now swung the other way; most researchers believe that the planets were born from the same cloud of gas as the Sun, in a process that must accompany the birth of many stars.
In its simplest form, this theory dates back to the middle of the 18th century. The philosopher Immanuel Kant envisaged that the planets con-densed from a rotating disc of gas around the Sun. The French physicist Pierre Simon de Laplace fleshed out the theory math-ematically. According to Laplace, as the disc contracted it threw off denser rings of matter that eventually condensed into the planets.
Physicists have had to discard Laplace's original theory, because it cannot explain why the Sun has only a tiny proportion of the angular momentum in the Solar System. The Sun collected up most of the mass when the cloud of gas collapsed and began to spin more quickly, but it should spin faster than it does to fit the original form of the theory. Andrew Prentice at Monash University in Melbourne has however resurrected Laplace's idea in a new guise. He invokes supersonic movement of gas to spread angular momentum outwards from the Sun. Prentice extends the theory to explain not only the form of the Solar System but also the families of satellites, which he sees as 'miniature Solar Systems' around the giant planets such as Jupiter.
While retaining Laplace's concept of a disc around the early Sun, other researchers have generally dispensed with the idea that it threw off rings. Instead, they believe that the planets condensed within the disc of gas and dust. According to these theories, the Sun lost its angular momentum as some of its original matter streamed out to space, a version of the solar wind that exists today.
Variations on disc theories now hold centre stage for one strong reason - astronomers have found discs around other stars that appear to be young. Radio telescopes show stars being born inside dense black clouds of dust and discs of gas surround many of these embryonic stars. Dust, mixed with the gas, shines at infrared wavelengths, and infrared telescopes have found signs of discs surrounding many young stars in the constellation Taurus - including a star called T Tauri that astronomers believe resembles the Sun when our star was only a million years old.
Most convincing of all has been evidence from the Infrared Astronomical Satellite, which made its pioneering survey of the sky in 1983. The satellite found that many stars are producing more infrared radiation than their temperature and composition suggests, indicating dust that is much cooler than the stars themselves. The list includes several bright, nearby stars, such as Vega and Fomalhaut. Most likely, the dust in these systems lies in a disc around the star itself - an interpretation that was dramatically proved in the case of the star Beta Pictoris. American astronomers Brad Smith and Rich Terrile took a long-exposure image with an electronic light detector to reveal that this star was surrounded by a disc of matter twice the size of the Solar System.
All these stars and their discs are giving us glimpses of how the young Sun must have looked but we can observe none of them in enough detail to see how the material in the disc turns into planets. Astronomers have yet to find proof of other planetary systems: the Hubble Space Telescope should have had a sporting chance of finding planets in orbit about some nearby stars, but the telescope's flawed mirror has ruled this out. We must turn back to theory and to the little evidence that has survived from the earliest days of our own system.
We live in the wrong place to find traces of the early Solar System: mountain building, the recycling of ocean floors and erosion have altered the Earth so radically that no part of our planet's surface dates back more than 3800 million years, and most is far younger. A few clues come from the Earth's composition, especially the distribution of rare elements and isotopes in the deeper, less reworked rocks, but for more information we need to look farther afield than our own familiar planet.
Fortunately, the most valuable type of information comes to us free and unbidden, in the form of meteorites, the bewildering variety of lumps of rock that fall to Earth from space. Most meteorites are made of pale-coloured rock, but a few per cent are composed of metal, usually iron or nickel and alloys, while others are a dark, powdery rock containing large amounts of carbon. The metal and rocky meteorites have broken free from the interiors of asteroids in collisions: the high temperatures within the asteroids erased most of the information about their origins. But the dark, carbon-bearing meteorites - carbonaceous chondrites - once formed near the surfaces of asteroids, and they have suffered comparatively little alteration since the birth of the Solar System.
The carbonaceous chondrites are treasure trove for researchers. Many of the crystals that compose them could have formed on Earth; they even contain microscopic diamonds. But among the jumble of crystals are unaltered pieces of interstellar dust that date from before the formation of the Solar System (as described by Monica Grady and Ian Wilson in 'A cosmic cake mix', New Scientist, 15 September 1990). But the special feature of the chondrites are the chondrules, rounded clumps of crystals that look like droplets. The crystals within them have a characteristic pattern, recognisable from rocks on Earth that have melted and cooled instantaneously. These are drops of rock that melted during the birth of the Solar System. They show that temperatures in the disc reached around 1500 oC.
Top-down or bottom-up?
This ties in well with theoretical calculations of what happens during the birth of stars. Alan Boss of the Carnegie Institution of Washington, for example, calculated that the original material collapsed under its own gravity to form the Sun and the surrounding disc in only 100 000 years - just an instant on the astronomical timescale. At this moment, marking the birth of the Solar System, Boss found that compression would have heated the gas in the disc to around 1200 °C, at the distance from the Sun where the asteroids formed.
This high temperature melted or vaporised most of the material that went into the disc. Some of the minerals in the carbonaceous chondrites have not been altered by heat or pressure since then. Researchers can use the proportions of radioactive elements in such crystals to measure the time that has elapsed since the crystal solidified, which for all practical purposes is equal to the age of the Solar System. A group at the Paris Institute of Earth Physics has made the most precise measurements, looking at the relative proportions of radio-active uranium-238 and its 'daughter' element lead-206 in minerals such as merrilite and apatite, which are phosphates.
These crystals are naturally high in uranium and almost free of lead, so the isotope of lead present now must have come from the decay of uranium. As a result, the French group can date the formation of a crystal with formidable accuracy - to within a million years. They find that the Solar System condensed from a ball of gas close to 4560 million years ago.
When it comes to turning the material of the disc into planets, there are two possibilities: 'top-down' or 'bottom-up' theories. In the top-down theory, clumps of gas in the disc became unstable and collapsed under their own gravitational attraction to form planets, like miniature versions of the collapsing Sun. According to the bottom-up idea, specks of dust accumulated to make bigger and bigger solid bodies that grew into the present-day planets and asteroids.
The top-down idea has been championed, in different forms, by Prentice, Woolfson and Alistair Cameron of Harvard College Observatory. It provides an excellent means of creating the giant gas planets such as Saturn and Jupiter and their families of moons that resemble miniature Solar Systems. The advantage of the top-down process is its speed: a gas cloud can collapse under its own gravity in less than a million years. Jupiter, in particular, must have formed this quickly. This mammoth planet grabbed most of its gaseous bulk - hydrogen and helium - from the disc of gas around the early Sun, and this disc dissipated within a million years.
In addition, Jupiter reached its present mass quickly enough for its gravity to prevent the asteroids, orbiting between Jupiter and Mars, from coalescing into a single planet. And the asteroids themselves formed in a few million years, at most, according to Robert Hutchison of the Natural History Museum in London, and Ian Hutcheon of the California Institute of Technology. They have investigated a portion of a meteorite with a composition that indicates it once formed part of an asteroid-sized body. This small piece of rock contains a rare isotope of magnesium, one that is the daughter product of the isotope aluminium-26, which has a half life of only 720 000 years. The radioactive aluminium must have been created shortly before the Solar System was born, probably in the explosion of a nearby supernova. It had to be incorporated into the parent body of this meteorite before a significant amount of the aluminium had decayed - within a million years at most.
Despite its successes in explaining the giant planets, the top-down theory is no longer very fashionable. It has lost favour largely because the astronomers' model of the Solar System has changed. We used to regard the nine planets as the only part of the system worth considering, and a theory was satisfactory if it succeeded in explaining just the birth of the planets. Now astronomers give as much weight to the many smaller bodies of the Solar System: despite their small physical size, a successful theory must explain them, too.
The small worlds include Pluto and its moon Charon, which together make up only 0.0002 of the mass of the Earth, and the thousands of asteroids. There are also billions of comet nuclei - lumps of ice several kilometres in diameter - that live beyond the orbit of Pluto. Alan Stern of the University of Colorado at Boulder has recently proposed that Pluto and Neptune's moon Triton (which orbits the planet backwards) are the few survivors of thousands of icy worlds between 1000 and 2000 kilometres across, which originally formed near Uranus and Neptune. These were the big brothers of the comet nuclei, and they have all been flung out to a region beyond Pluto's orbit. One that has made its way back is the small world Chiron, which is now in an unstable orbit between Saturn and Uranus.
The top-down theory accounts for the creation of the large planets very easily, but its large gaseous swirls would not naturally form tiny solid worlds such as the asteroids. The bottom-up theory, on the other hand, leads quite naturally to the existence of solid bodies of all sizes - from microscopic grains through comet nuclei and Stern's 'ice dwarfs' up to fully fledged planets.
The bottom-up theory is flawed because it does not explain the formation of the four largest planets, the 'gas giants'. But this has not stopped many of today's leading researchers - intent on forming all the planets by the same mechanism - from trying to develop a workable version. They believe that the icy material in the outer reaches of the Solar System built up into planets around ten times heavier than the Earth, at which point the worlds had enough gravity to attract gases from the surrounding disc and grow to their present enormous sizes. The problem is that the growth of solid bodies from the bottom up is slower the farther they are from the Sun. Jupiter's core should have taken ten million years to accumulate in this way, by which time the surrounding gases would long since have dissipated. Even worse, the time taken to make Uranus and Neptune would be more than the age of the Solar System, so these planets should not yet exist!
Apart from the problems with the giant planets, however, there is now little doubt that the rocky planets - Mercury, Venus, Earth and Mars - and the smaller solid worlds were built up from microscopic solid grains. The dust in interstellar space consists of fragments of rock, carbon and ices, generally less than a micrometre across. In the inner region of the disc around the young Sun - out to the present orbit of Jupiter - the gas became hot enough to boil away ice. As a result, the planets that formed closest to the Sun are mainly rock, while the bodies in the outer part of the Solar System are largely ice or water (along with hydrogen and helium captured from the gaseous disc).
In the course of a few thousand years, the dust particles in the disc settled into a thin layer, like an enormous version of the rings circling Saturn today. Saturn's rings are so close to the planet that its gravity prevents the particles from banding together. The particles surrounding the Sun were much further out, proportionately, and had no such problem. According to proponents of the bottom-up theory, the mutual gravitational pull of the tiny dust particles - possibly aided by their natural stickiness - clumped them into bodies that grew to about a kilometre across.
These small, loose aggregates of dust, known as planetesimals, are a far cry from fully-fledged planets. There were so many of them that they must have run into one another, building up larger bodies. But there is the danger that a collision might instead smash the planetesimals apart. The bottom-up theory has become popular in recent years largely because calculations of the likely effects of collisions show that the aggregation of planetesimals will win out over their destruction.
The basis of the theory was set out in the 1960s by Soviet researchers led by Victor Safronov and his wife Evgenia Ruskol. A planetesimal either grows or splits in a collision depending on whether the closing speed is higher or lower than the planetesimal's escape velocity - a measure of its gravitational pull. If two planetesimals collide more slowly than their combined escape velocity, the fragments from the collision move off at a low speed, and their gravity can pull them back together into a single body. If the two bodies smash at a speed higher than the escape velocity, the fragments shoot off too fast for gravity to drag them back.
Safronov and Ruskol realised that planetesimals themselves control their relative speeds as they all whirl around the Sun. The speed of a planetesimal alters all the time through close encounters with other planetesimals, either a collision or a gravitational tug as two pass nearby. The calculations showed that this process always gives speeds that are less than the escape velocity from the largest planetesimal involved. Collisions between smaller planetesimals may lead to either growth or disruption, depending on the sizes and velocities. But the largest planetesimals can always hold onto any smaller body that collides with them: the law of the jungle applies, with the fattest cats growing fatter at the expense of the others.
Other researchers have now used computers to simulate the growth of planetesimals. George Wetherill, of the Carnegie Institution of Washington, has looked in detail at the growth of the four innermost planets. In a typical calculation, he takes 500 planetesimals with masses a bit less than that of our Moon, circling the Sun roughly between the present orbits of Venus and the Earth. In dozens of trials with slightly different starting conditions, Wetherill finds consistent trends that lead repeatedly to planets similar in sizes and arrangements to those that exist today.
In one typical simulation, the largest of these small worlds grow quickly, with nine bodies as massive as Mars appearing within half a million years. After two million years, there are more than a dozen planetesimals as massive as Mars and two that are more than twice as heavy as the others: these two grow to become the largest of the inner planets, Venus and the Earth. Over the next 30 million years, all except one of the Mars-sized bodies is swept up by the embryonic Venus or Earth. The Earth reaches its present size and mass some 60 million years after the birth of the Solar System.
Wetherill's latest simulations (New Scientist, Science, 10 August) have included the effects of Jupiter's gravity, which stirred up the planetesimals. Over dozens of simulations, he finds some common themes. Jupiter flings most of the planetesimals away from a region that corresponds to the asteroid belt, while material closer to the Sun generally forms four planets. The heaviest of these rocky planets is very similar to the Earth in size and distance from the Sun. There is usually a second large planet, the size of Venus or slightly smaller, which lies either nearer to the Sun or farther away. The other two are typically much smaller and lie at the inner and outer edges of this four planet family - the very positions of the small planets Mercury and Mars. Although Jupiter's gravity is needed to produce this suite of four inner planets, Wetherill finds that an 'Earth' forms at about our distance from the Sun in about three-quarters of the simulations, which is good news for those looking for life around other stars.
In these calculations, it turns out that the collisions move the growing planets around, so that the original 'planet-embryos' did not lie at exactly the same distances from the Sun as we find the planets today. The most extreme case is Mercury. This small planet is a lucky survivor: literally dozens of planetesimals this size and larger ended up on collision course with Venus, Earth or Mars. In one of Wetherill's simulations, Mercury began life beyond these planets, and a narrow brush with Mars swung it into its present orbit near the Sun.
The overall message from these simulations is startling. The final stages in the growth of the Earth and the other inner planets was not the gentle sweeping up of small meteorites. Instead, the scene was dominated by high-speed collisions between fully grown planets. During the formation of the Earth, for example, Wetherill finds that our planet was hit by about three planetesimals the size of Mercury and at least one bigger than Mars. As a result, astronomers now accept that the early history of the Earth and its neighbours must have involved catastrophic collisions. A driving force behind the acceptance is that this idea can explain several of the puzzling features of the inner planets.
One long-standing problem has been Mercury's high density. This small planet has a dense core of iron that is huge compared with the other rocky planets, as described by Ken Croswell in 'Mercury, the impossible planet' (New Scientist, 1 June). Alistair Cameron, with colleagues from the Harvard-Smithsonian Center for Astrophysics and the Los Alamos National Laboratory, has found that a collision can easily account for Mercury's present state. In their computer model, they start with a 'proto-Mercury' twice as big as the planet today. A smaller world runs into it at 20 kilometres per second, vaporising much of the planet's rocky outer layers without disrupting its interior, to leave a world with a disproportionately large metallic core.
Venus has long puzzled astronomers by spinning backwards relative to the rotation of the other planets. Again a giant collision could be the answer. If a world the size of Mars ploughed into Venus in a direction opposite to the planet's original spin, it would have delivered a sufficiently large kick to make the planet turn the other way. A similar collision may have tipped up the planet Uranus, which orbits the Sun while spinning on its side.
Most exciting of all, this cosmic snooker in the early days of the Solar System provides a way of making our Moon. Familiar as it is to everyone, the very existence of the Moon has always been a headache for astronomers. In many ways, it is the oddest moon in the Solar System, forming with the Earth a system that is more like a double planet that the usual pattern of big planet and tiny satellite. The Earth is the only one of the inner planets to have a sizeable moon: Mars has two small moons, but they are almost certainly asteroids captured by the planet. Our Moon is larger in relation to its parent planet than any other satellite, except for the moon of tiny Pluto.
One early theory said that the Earth and Moon were once a single planet, but one that rotated so quickly that it split in two. Among the problems of this theory is the fact that the Moon ought then to orbit above the equator of the spinning Earth. But our Moon is the only major satellite in the Solar System that does not orbit above the equator of its parent planet. Its orbit is tipped over so that it lies more or less in the plane in which the planets orbit the Sun. In that case, perhaps the Moon formed as a separate planet? It might have been born alongside the Earth, or it might have formed in a different part of the Solar System and was later captured by the Earth's gravity.
The rocks brought back from the Moon by the Apollo astronauts in the late 1960s and early 1970s bore out none of these theories. They were too different in composition for the Moon to have been part of the Earth or to have condensed as a neighbour in the same cloud of gas and dust.
In fact, the Moon's composition is so strange that astronomers could not think where in the Solar System it could have formed. Like Sherlock Holmes' dog that did not bark, the most important clue was not what the Moon contains, but what it lacks. First, the Moon contains very little iron. Secondly, it lacks the most volatile substances - such as chlorine, potassium and water - indicating that the Moon must once have been heated to incandescence.
Faced with these facts, American astronomers William Hartmann and Don Davis of the Planetary Science Institute at Tucson, Arizona, devised an alternative theory for the birth of the Moon: the 'big splash'. When first proposed in 1975, the theory seemed highly unlikely, because it required a wayward planet even larger than the Moon colliding with the Earth. But the recent simulations of the birth of the planets through collisions have brought the big splash to the fore as the leading contender to explain the origin of the Moon.
According to the Big Splash theory, a world as large as Mars - fully half the diameter of our planet - smashed into the early Earth at 10 times the speed of a rifle bullet. The impact smashed whatever crust the Earth had, melting rock to a depth of hundreds of kilometres. The other world was completely destroyed. The iron that had formed its core sank through the ocean of molten rock into the Earth's own core. Its surface rocks exploded into space in an incandescent plume, to form a fiery ring around the Earth.
Within a day, these drops of molten rock came together to form a new world in orbit around the Earth. This body had very little iron, because the iron in the impacting world had ended up in the Earth's core. It was also bone-dry, with little left of the water or volatile elements that had boiled away from the fiery plume. The Earth had acquired its Moon.
As these loose ends begin to fit into a single broad-brush picture, many astronomers are confident that we now have - in outline at least - the correct theory for the origin of the planets. Wetherill sees the present growing consensus as an advance on the more idiosyncratic and individualistic ideas of a few years ago, and refers to the bottom-up theory as the 'standard model'.
Hutchison, however, sounds a more cautious note. For all the computer modelling on the accumulation of dust into planets, he believes that the available evidence does not rule out the competing theories, such as Woolfson's idea of the planets forming from a gas streamer in a dense cluster of young stars. More evidence, both from planets and meteorites, is required. 'It's still a great area for speculation,' Hutchison concludes. 'That's what makes it so exciting.'
* * *
A plan for the planets
Any successful recipe for the Solar System must produce the following combination of planetary properties:
1. Nine planets orbiting a star in the same direction, in orbits that are almost circular and lie roughly in the same plane.
2. Each planet's orbit to be about 73 per cent larger than that of the next planet closer to the Sun.
3. Most of the angular momentum in the system to lie in the planets rather than in the rotation of the Sun - in other words, the Sun must rotate surprisingly slowly.
4. A distinct difference in the type of planet, depending on the distance from the Sun: small rocky worlds are close to the Sun, the gas giants are farther out, with a small icy world, Pluto, at the outer edge.
5. Most planets must themselves rotate in the same sense in which they orbit the Sun. The exceptions are Venus, Uranus and Pluto.
6. Other individual quirks: Mercury's large core, Venus's slow, backward rotation, and the sideways tilt of Uranus, which points its pole of rotation at the Sun.
7. The many different types of rubble: the asteroids between Mars and Jupiter, the odd icy worlds farther out, and the millions of tiny icy bodies, the nuclei of comets, beyond the orbit of Pluto.
8. Moons of different kinds: the giant planets have large regular families, Neptune has a large moon orbiting in the opposite direction to the planet's own rotation and Earth and Pluto are double planets.
By Bernard Pelletier
Thu Mar 27 14:08:27 GMT 2008
There are several problems with the solar nebula model: it does not well explain angular momentum distribution, nor accretion stickiness, nor planetismal collisions, and has no explanation at all for the twinning of planets. Solar fission (Van Flandern) accounts for the twinning of planets and angular momentum distribution, and does not require accretion stickiness nor planetismal collisions.All comments should respect the New Scientist House Rules. If you think a particular comment breaks these rules then please use the "Report" link in that comment to report it to us.
If you are having a technical problem posting a comment, please contact technical support.
10:45 10 October 2008
