Two billion years from now, our galaxy is in for a shock. With every hour that passes, the Milky Way gets half a million kilometres closer to another large spiral galaxy called Andromeda, and it is only a matter of time before we collide. Over the course of 3 billion years or so, the galaxies will swing through each other, stretching out long, wispy streams of stars before settling down and merging into one. That much we know.
Yet that picture is far from complete. Lying at the centre of our galaxy is a giant black hole more than 3 million times as massive as the sun. The black hole at the heart of Andromeda is believed to be 10 times the size. What will happen to these supermassive black holes during the encounter is anyone鈥檚 guess.
Astronomers have recently started to find some clues, though. Most, if not all, galaxies have a supermassive black hole at their centres. Everyone thought that these hungry behemoths simply sat at the heart of their parent galaxies, vacuuming up gas clouds and ripped-apart stars. Now it seems they can go off sightseeing. A black hole can cut loose when two galaxies collide: their central supermassive black holes coalesce into a single object, and this can receive a tremendous kick in the process. Some supermassive black holes travel to the outskirts of their galaxy before returning home, others go into exile for good, catapulted unceremoniously into the lonely deep freeze of intergalactic space.
Advertisement
These new insights could explain some of the most puzzling observations chalked up in outer space. It鈥檚 all thanks to new ways of modelling the complex distortions of space-time wrought by black holes鈥� awesome gravitational power. Theorists have finally learned how to simulate the merger of two black holes ( of two galaxies merging), and the discoveries are coming thick and fast.
鈥淔or 30 years, people had been struggling to solve this problem,鈥� says Manuela Campanelli of the University of Texas at Brownsville and the Rochester Institute of Technology, New York. 鈥淪uddenly, there has been a spectacular breakthrough.鈥�
What had stymied progress for so long was the horrendous complexity of applying Einstein鈥檚 general theory of relativity to pairs of black holes. General relativity describes gravity as warps in space-time, and the tremendous gravity of a black hole pulls those warps to extremes as it sucks in everything around it. Two black holes rushing towards each other change the shape of nearby space-time dramatically, which in turn affects their motion. This feedback process makes the equations of general relativity around pairs of black holes fiendishly difficult to handle: when researchers tried to simulate black hole mergers, the complexity of the calculations skyrocketed and caused their supercomputers to crash. Some astronomers became convinced that we would never be able to understand how black hole collisions work.
Enter Frans Pretorius, a physicist at Princeton University. In 2005, he surprised everyone by finding an answer. His remedy exploits the fact that Einstein鈥檚 equations hold good no matter how you describe your location in space-time. This had always meant that, tantalisingly, if we just used a different way of specifying the position of two black holes, the mathematics might become simple enough to allow theorists to calculate what happens when the black holes collide.
Pretorius鈥檚 breakthrough was to find the right sort of coordinates. He then proceeded to model the coming together of two black holes with equal masses. 鈥淚t totally galvanised the field,鈥� says Campanelli. 鈥淧retorius smashed through a mental barrier. Everyone went back to their own simulations with the priceless knowledge that the problem was soluble.鈥�
Within a month, Campanelli鈥檚 group had adapted their own, quite different, method 鈥� and got it to work. They presented their results at a conference at NASA鈥檚 Goddard Space Flight Center in Greenbelt, Maryland. Coincidentally, Campanelli attended a talk at which the local NASA researchers led by Joan Centrella presented an almost identical method. 鈥淚t鈥檚 amazing,鈥� says David Merritt, a colleague of Campanelli at Rochester. 鈥淭hree decades of total frustration and then three groups solve the problem.鈥�
Other physicists have reason to welcome the breakthrough. Several groups around the world are hunting for gravitational waves 鈥� ripples in space-time radiated by massive moving objects such as black holes and neutron stars. When, in the next few years, our detectors become capable of picking up a gravitational wave, we will want to work out what is producing it. That means knowing how to identify the gravitational wave signatures of different kinds of violent cosmic events, including mergers of black holes.
Two become one
Today鈥檚 gravitational wave detectors, such as LIGO in the US, Virgo in Italy and Geo600 in Germany, should soon be picking up the merger of much smaller black holes, created by the collapse of individual stars. To hear the vibrations caused by mergers of the supermassive kind, however, detectors need to escape seismic 鈥渘oise鈥� and get much bigger so that they can tune into a wider range of wavelengths. That鈥檚 why a space-borne gravitational wave observatory called LISA is due to launch some time after 2015.
Astrophysicists are eagerly awaiting LISA because they think that such mergers played a major role in the creation of today鈥檚 galaxies. 鈥淚t鈥檚 very likely that every galaxy, including the Milky Way, has undergone at least one merger in its history,鈥� Merritt says.
When two galaxies collide, the black holes at their cores are thought to go into orbit around each other. Gravity pulls them ever closer, so the black holes spiral together until they coalesce, releasing gravitational waves all the while. In the past researchers have been able to work out using general relativity what the waves should look like as the black holes begin stalking each other, and they can also predict what sort of waves the single, merged black hole should give off. But the final moments before two black holes collide, when gravity is at its strongest, have remained obscure until now. 鈥淭his is where all the interesting stuff happens,鈥� Campanelli says.
The latest simulations by Campanelli and others have thrown up some surprises. Astronomers believe that most black holes rotate, and it turns out that this spin plays a far greater part than the behemoth鈥檚 mass in determining what happens during a black hole merger (see Illustration). In the simulations, the resulting black hole usually spins about a different axis to its two precursors. What鈥檚 more, this change in spin occurs very suddenly 鈥� just as the black holes are about to merge.
The researchers didn鈥檛 have to think up a name for this phenomenon because it already had one: 鈥渟pin-flip鈥�. Along with his colleague Ron Ekers of the Australia Telescope National Facility in Sydney, Merritt had predicted it in 2001 while pondering the nature of a class of active spiral galaxies known as Seyfert galaxies. Theory says that a galaxy and its central black hole should spin around the same axis, since the two were presumably born together. However, many Seyfert galaxies have a black hole whose spin direction is different to that of the rest of the galaxy 鈥� something revealed by the narrow jets of energetic particles that shoot out from the poles of supermassive black holes in active galaxies. 鈥淚n 2001, I speculated that this is because such galaxies have undergone multiple mergers so that the spin-flips have randomised the orientation of their black holes,鈥� says Merritt.
Now that it has been confirmed by the new simulations, spin-flip may explain a curious class of active galaxies that appear X-shaped when viewed with radio telescopes. These ultra-bright galaxies owe their brilliance to a central, spinning supermassive black hole: as its jets slam into the ionised gas that makes up the intergalactic medium, they emit radio waves that form 鈥渓obes鈥� stretching millions of light years either side of the galaxy.
Astronomers have noticed that several X-shaped galaxies have much larger lobes than the others. One is called NGC 326, and detailed observations show that in fact it has two sets of lobes superimposed. 鈥淭he simplest explanation is that the central galaxy had a recent merger, causing the spin of its black hole to flip and the jets to change direction,鈥� says Merritt.
The latest simulations have done far more than confirm Merritt and Ekers鈥檚 ideas. They have allowed astronomers to study mergers between black holes with different masses and different spins. This way Campanelli鈥檚 group and others have uncovered the brightest sources of gravitational waves in the universe.
In their simulations, they studied two black holes rotating in the same direction as they orbit each other. Campanelli鈥檚 team found that such black holes take up to three times longer to merge than ones whose spins are different, because they experience a large centrifugal force that fights to keep them apart. This 鈥渉ang-up鈥� has important consequences. 鈥淎 slower merger means the two black holes experience each other鈥檚 intense gravity for longer and therefore radiate more intense gravitational waves,鈥� says Campanelli. 鈥淭hese kinds of systems will be the brightest merger events in the history of the universe for gravitational wave detectors.鈥�
By far the most exciting consequence of a black hole merger, though, is the 鈥渒ick鈥� the merged object can receive. The size of the kick depends crucially on spin, because unequal spins make the merger asymmetric, and that produces asymmetric gravitational waves. 鈥淭hese act like a rocket exhaust, pushing the black hole in the opposite direction,鈥� says Merritt. 鈥淚 like to think of it as space itself pushing the black hole.鈥�
That kick has all kinds of novel astrophysical consequences. For instance, in the merger of galaxies with relatively small black holes, the kick may be only a few hundred kilometres per second, so the merged object may be booted only as far as the outer regions of its parent galaxy before falling back to the centre. Since the merged object may very well take its super-hot disc of swirling matter and jets with it, it will appear as a very bright, compact object called a quasar, displaced from the centre of the galaxy.
Calculations show that in a small galaxy, a kicked supermassive black hole may oscillate back and forth, crossing the galaxy perhaps 10 times before coming to rest in the core again. A bigger black hole in a larger galaxy, on the other hand, might make only one excursion before returning home.
According to Merritt, such richocheting black holes may already have been observed. They are known as ultra-luminous X-ray sources, or ULXs, and astrophysicists have previously postulated the existence of middleweight black holes to explain them. It isn鈥檛 clear how such black holes would form, however, and their existence remains controversial.
Absent without leave
Merritt wonders whether there is any need to imagine such objects at all. 鈥淎 ULX could simply be a supermassive black hole that has been kicked out of a galactic core,鈥� he says. Telling a wayward giant apart from an intermediate-mass black hole won鈥檛 be easy, though. Without following the motion of the intense swirl of stars and hot gas at the centre of a galaxy, it is difficult to pin down its mass accurately.
It is not just the kicked black hole that is interesting. According to Merritt, a supermassive black hole gone AWOL profoundly affects the galactic core it leaves behind. For one thing, the stars of the core are no longer held in place by the strong gravity of the supermassive black hole, so they will spiral outwards in their orbits. When the prodigal black hole returns, the energy it dumps onto the gas and stars as it slows down will cause them to move even further out 鈥� a kind of double whammy. 鈥淪ure enough, people have long noticed that the cores of luminous elliptical galaxies are often twice as big as in other galaxies,鈥� Merritt says. 鈥淭ravelling black holes may be the explanation.鈥�
Kicked black holes may do more than just bounce around within galaxies. Campanelli鈥檚 simulations have shown that the merged black hole can gain a kick of thousands of kilometres per second. 鈥淭here are some indications that the kick could be as high as a million kilometres per hour,鈥� Merritt says. 鈥淲hat鈥檚 exciting here is this is easily enough to eject a supermassive black hole from even the most massive galaxy.鈥�
Even at those speeds, such a hole would still take about 100 million years to reach the outskirts of its host galaxy. During that time, it would appear as a quasar displaced from the centre of the galaxy. Then it would be gone, fading into the darkness between galaxies. 鈥淚ntergalactic space might be criss-crossed by high-speed, runaway supermassive black holes,鈥� Merritt says.
鈥淚ntergalactic space might be criss-crossed by high-speed, runaway supermassive black holes鈥�
What would they look like? 鈥淭his is the key question,鈥� Merritt says. By definition, we can鈥檛 see black holes; they give themselves away by sucking in matter that gives off radiation before it vanishes down the gravity plughole. 鈥淲ith very little gas floating about in intergalactic space to feed such holes, they are likely to be very faint indeed,鈥� Merritt says. Abraham Loeb of the Harvard-Smithsonian Center for Astrophysics in Cambridge, Massachusetts, who has studied galaxy mergers in detail, says that under some circumstances ejected black holes might shine brilliantly ().
Remarkably, it is possible that one has already been seen. Discovered in 2005 by a team led by Pierre Magain of the University of Li猫ge, Belgium, HE0450-2958 is a faint quasar near a galaxy that has recently undergone a merger. 鈥淪ome say that the quasar looks like what you would expect if the black hole powering it had been kicked from the nearby galaxy,鈥� says Merritt. 鈥淗owever, I think the jury is still out.鈥�
Merritt thinks the idea of rogue supermassive black holes flying through space could be a hugely fruitful new astrophysical paradigm. The quasar-like activity we see in intergalactic space, in the outer regions of a galaxy and in galactic cores may all be examples of the same phenomenon. 鈥淚t could be a unifying idea,鈥� he says.
The findings have relativists writing papers on black hole mergers by the score. 鈥淣ow that we can predict the consequences of black hole mergers, there is so much we can do,鈥� says Merritt. He wonders what will happen when a galaxy with a pair of coalescing black holes merges with another galaxy with a single hole or even another binary. 鈥淭hat鈥檚 three or four black holes merging together,鈥� he says. 鈥淣ow that the mental barrier has been smashed, the possibilities in this field are endless.鈥�
So what鈥檚 the fate of our own black hole? Most astronomers agree that it will coalesce with Andromeda鈥檚. Beyond that, no one has yet worked out the details. Our new black hole could be a stay-at-home giant. Then again, it could suffer a serious case of intergalactic wanderlust, leaving what鈥檚 left of our solar system to drift through space. So if we want to study our black hole up close, we鈥檇 better develop a star-faring craft soon. Otherwise we鈥檒l have to chase it across the universe.
