Image
© Andy MartinDark forces at work
Where will astronomers stop in their love affair with the enigmatic substance called dark matter? First we were told it was essential to allow a galaxy to spin without falling apart. Then it was the glue that held clusters of galaxies together. Later it was said to have catalysed the formation of the galaxies in the first place. Now, surely, they have gone too far. If the latest theories pan out, dark matter has also given us some of the world's most enduring astrological myths.

In this rational age, we have come to recognise constellations as chance alignments of groups of stars. No longer do we think of Orion as a mighty hunter, brought down by the sting of a fateful encounter with the scorpion Scorpius. Yet it seems that an unseen hand may after all have been responsible for placing these stars in the sky.

Hints are emerging that around 30 million years ago, a giant clump of dark matter struck our part of the Milky Way, creating a rippling disc of star formation that eventually produced Orion's belt, the bright ruby jewel of Antares in Scorpius, and many more of the sky's most notable stars. If the scenario is correct, it could guide us in the search for a solution to one of the abiding mysteries of physics: what exactly is dark matter made of?

In the middle of the 19th century, the English astronomer John Herschel noticed that we are surrounded by a ring of bright stars. But it was Boston-born Benjamin Gould who brought this to wider attention in 1874. Gould's belt, as it is now known, supplies bright stars for many famous constellations including Orion, Scorpius and Crux, the Southern Cross, which appears on the official flags of five countries and several territories. Perseus and Canis Major in the north, along with Vela and Centaurus in the south, also contain stars in Gould's belt.

It is a sizeable structure, some 3000 light years across, and can be traced as a bright band of stars tilted at about 20 degrees to the Milky Way. Within it are several thousand high-mass stars as well as up to a million low-mass ones. Most importantly, these stars appear to have formed separately from the rest of the stars in the galaxy - and that's what makes them so interesting.

Stars do not form randomly in the Milky Way. Instead they are confined to the arms that spiral around its nucleus. Giant clouds of gas molecules form in these denser regions and are eventually compressed to form the stars that shine out in the orderly spiral pattern that characterises our galaxy and many others like it.

"Gould's belt is not part of the spiral structure," says Fernando Comerón of the European Southern Observatory in Garching, Germany, who has been studying the belt for more than 30 years. "It must have been triggered by some local, violent event." But what caused that event?

The most obvious culprit for violence on a galactic scale is an exploding star. Such celestial cataclysms temporarily give out more light than a hundred billion stars combined, and are a great way of pumping energy into space and sending shock waves through gas clouds that could eventually trigger the formation of stars. That won't do as an explanation for Gould's belt, though. Exploding stars shoot out their energy in all directions at once, and it is hard to imagine how this could have tipped the belt so far out of the plane of the galaxy. "It is very difficult to knock a giant molecular cloud out of the galactic disc," says Comerón.

Fifteen years ago, however, he thought he had the answer. Comerón suggested that an intergalactic gas cloud falling onto the Milky Way might have struck a giant molecular cloud, generating shock waves that set the formation of Gould's belt into motion.

The idea had a lot going for it, yet it also had a fatal flaw. On one hand, Comerón showed that an intergalactic cloud coming in at a suitable angle could produce the characteristic tilt of Gould's belt. The trouble is, such clouds are extremely tenuous, making it hard to understand how they could have provided the oomph needed to bring about the birth of the hundreds of thousands of stars in Gould's belt. So the search for the culprit continued - but to no avail.

That lack of success set astrophysicists thinking. What is the one thing that astronomers cannot see but are nevertheless convinced exists in vast quantities? Dark matter, of course.

Every galaxy is thought to be surrounded by a giant sphere of dark matter more than 10 times the width of the visible portion of the galaxy, and containing at least 10 times the mass of the visible stars. The dark matter "halo" is clumpy rather than smooth, because it contains residues of smaller halos that the galaxy has swallowed to grow to its present size. These clumps are referred to as the "halo substructure", and there are usually plenty of them.

"Count them all and you get a huge number," says Jürg Diemand who runs high-resolution supercomputer simulations of the Milky Way's halo substructure at the University of California, Santa Cruz. He estimates that in the halo surrounding the Milky Way - which is a pretty typical sort of galaxy - there could be a million billion (1015) such clumps of dark matter.

Crucially, this substructure is remarkably durable. Inside the halo, clumps of ordinary matter tend to radiate energy away, and this allows them to settle onto the disc of the galaxy. Dark matter clumps do not normally emit radiation, and so continue to circulate in the same orbits into which they were born - and this means that from time to time a clump of dark matter will come crashing through the disc of the Milky Way.

Dark collisions

For Kenji Bekki of the University of New South Wales in Sydney, Australia, this line of thinking was encouragement enough to set to work. He cranked up a computer model and began simulating what happened as clump after clump of dark matter smashed into one giant molecular cloud after another. Sure enough, he found that it was child's play to generate something resembling a Gould's belt. "The tilt is easy to reproduce," he says. "These collisions may be entirely natural.

Bekki's simulation shows that as a clump of dark matter with a mass of about 10 million suns passes through a giant molecular cloud, it pulls gas in towards the centre of the collisions and triggers a wave of star formation. As it continues through to the other side, it tilts the cloud and sets it spinning, as observed in Gould's belt (Monthly Notices of the Royal Astronomical Society: Letters, vol 398, p L36).

Image
© New Scientist
Such collisions may not just be happening to the Milky Way. Our nearest satellite galaxy also shows evidence consistent with it having been pummelled in a similar way. The Large Magellanic Cloud (LMC) sits just 186,000 light years away, and at first sight it appears to be a mess of stars with no obvious structure. Closer inspection, however, has revealed a bar-like structure of older stars buried near its heart. This could have happened if the LMC was once a "barred spiral galaxy" - one in which looping spiral arms are joined to the central nucleus by a bar of stars. What has so far defied explanation, however, is why the bar of stars is significantly off-centre inside the galaxy.

One suggestion was that the Milky Way's gravity has gradually been pulling the LMC out of shape, but this has come to nothing as simulations were unable to show how this gentle distortion could have displaced the bar.

Bekki wondered whether a collision with a piece of halo substructure could be the cause - and in his simulations, it could. He found that a clump of dark matter with 100 million times the mass of the sun - which equates to about 1 per cent of the mass of the LMC itself - could reproduce the offset bar. It would do this not by moving the bar itself, but by rearranging the surrounding stars so that the bar appears to be off-centre. "The results strongly suggest that apparently isolated dwarf galaxies develop off-centre bars from interactions with dark matter clumps," he says.

Evidence for similar collisions is starting to turn up even further afield. Bekki is casting a critical eye at dwarf galaxy NGC 6822, which has a noticeable hole about 5000 light years across. "It looks like a collision to me," he says.

Meanwhile, Comerón has also found something that he believes to be a Gould's belt in another galaxy. Down in the south-east quadrant of the magnificent spiral galaxy M83 is a bright burst of stars, sitting in an otherwise dark lane between spiral arms. It is about 1500 light years across. "The complex stands out clearly from the spiral pattern of the galaxy, and its size and age make it look like a Gould's belt," he says.

Bekki estimates that a collision with a chunk of dark matter happens only rarely - about once every 300 million years somewhere in the Milky Way - and that the resulting Gould's belt will typically fade away in a few tens of millions of years. That makes us very lucky indeed to be sitting almost at the bull's eye of the collision, not least because it may give us a remarkable chance of getting to grips with dark matter.

Though there is no doubt in most astronomers' minds that dark matter exists, the evidence remains circumstantial. No one has directly detected a single particle of the stuff, and we are not even sure what it is made of. According to the leading theory, it is composed of particles called neutralinos. These hypothetical particles do not interact with passing light rays, but they are expected to unleash gamma rays if two happen to collide with each other. Evidence for their existence may be available soon, thanks to NASA's latest space telescope. Called Fermi, it was launched in June 2008, and part of its mission is to search for the characteristic gamma rays from dark matter annihilations. "Fermi represents the first time we can probe realistic models of dark matter," says Tesla Jeltema at the University of California's Lick Observatory in San Jose and part of the team searching for the telltale signals.

And what better place is there to look for these signals than in the galaxy's halo substructure? There are two possible approaches: look at known satellite galaxies, as these should be sitting in the largest clumps of dark matter, or go fishing across the whole sky for smaller clumps that may be considerably closer. According to Bekki's model, the clump that formed Gould's belt should now be about 9000 light years away, somewhere in the direction of Scorpius. Being 20 times closer than the LMC makes it the best target for these searches by a long way.

"That's very close," says Jeltema. "It should be much easier to detect than the dwarf galaxies." She goes on to suggest that the gamma-ray signal could be hundreds of times stronger than from a typical dwarf galaxy if - and it's a big if - the clump has remained intact during its passage through the Milky Way. Diemand is not optimistic. He reckons that the gravity of the Milky Way will have pulled it to pieces as it passed through the plane of the galaxy. "The clump will suffer significant mass loss or be completely destroyed."

Whether or not the clump responsible for producing Gould's belt is found, Bekki is certain of one thing. "Gould's belt is not just a magnificent stripe of stars in the night sky; it is also a fossil record of a dramatic event that happened in our solar neighbourhood." It's a story written in the stars, and it promises to tell us a lot more than the one about the hunter and the scorpion.

Bibliography
1.Stuart Clark's latest book is Galaxy, published by Quercus. His personal blog is at stuartclark.com