快猫短视频

GALAXY clusters are the largest objects in the Universe, so huge that
individual galaxies are only specks of light within them. You wouldn鈥檛 want our
Galaxy to bump into one on a dark night. But that, it seems, is what fate has in
store for us.

Images from two new telescopes show that clusters are colliding, and in the
process they are avidly consuming each other and all the matter around them.
Once they have finished, astronomers believe, we will be in a very
different-looking Universe from the one we see now鈥攁 Universe inhabited by
a new race of giants called superclusters.

That is a profound change from the status quo. For most of the 15 billion
years since the big bang, the basic cosmic building blocks have been galaxies.
These huge aggregations of dust and gas today hold anything from a million to
more than a trillion stars. Our own Galaxy is a light heavyweight, holding
around a hundred billion stars in a many-armed spiral about 50,000 light years
across.

As the Universe has evolved, some galaxies have joined forces to form galaxy
clusters, in which hundreds or thousands of galaxies swarm within a shared
atmosphere of gas. But clusters are a rarity鈥攐nly about 10 per cent of
galaxies are part of one. 鈥淭he average structure that is expected today is a
group of four or five galaxies,鈥� says Frazer Pearce at the University of
Nottingham. That鈥檚 much like the loose association of galaxies we belong to, the
Local Group, dominated by our Milky Way and the Andromeda galaxy.

Is this to be the Universe鈥檚 final form? Or could larger structures come to
dominate? A few years ago, the picture was unclear. Some astronomers thought
clusters might be merging into larger objects, but the observations were
inconclusive. Most clusters appeared to have settled down, having reached the
end of their growth.

Computer models are one way to find out why the Universe looks the way it
does, and what it might look like in future. Theoreticians start with the hot
gas expanding out of the big bang. If this gas had been evenly spread, it would
have stayed even while becoming more and more tenuous. That鈥檚 a pretty dull
outcome, and nothing like the structured Universe we see. So there must have
been some patches that started out a bit denser than the rest. Then gravity
would have pulled in the surrounding gas, eventually forming a much denser chunk
of matter.

What kinds of structure you get depends on the tug of war between gravity and
the expansion of the Universe. A small, dense knot of gas will rapidly collapse
under its own gravity. A larger, not-so-dense bump will probably continue
expanding. So to simulate the process, you need to know how much matter is
around, how much its density fluctuates over different distances and how fast
the Universe expands.

Cosmologists think they have a good handle on the average density of matter.
But when it came to expansion, they got a surprise. In 1997, measurements of the
motion of distant supernova explosions showed that the Universe鈥檚 expansion was
accelerating, not decelerating as had been thought. The cause of this is still a
mystery, but the acceleration seems to stand up to observation, so it has become
part of the computer models.

The final ingredient is variation in the density of matter just after the big
bang. This can be gleaned from the cosmic microwave background radiation,
emitted just a few hundred thousand years after the beginning of the Universe.
Brighter and darker spots in the microwave sky reveal where the density was
higher and lower.

Models are pretty good at reproducing the kinds of structure we see in the
Universe. For example, modelled galaxies don鈥檛 end up scattered uniformly
throughout space. Instead, they are strung out along a complex web of filaments
that stretches across the Universe. Maps of real galaxies confirm the existence
of this cosmic lattice.

Pearce thinks that clusters form where the filaments cross. 鈥淚n the
simulations, you can clearly see little knots of matter, anywhere from the size
of galaxies to small groups of them, draining away into these clusters,鈥� he
says. 鈥淎nd that is what we think is happening in the real Universe.鈥� Pearce鈥檚
simulations imply that clusters are still growing by swallowing nearby
galaxies.

But confirming this idea has been difficult. With an ordinary telescope,
which gathers visible wavelengths of light, all you can see are the galaxies
themselves. Two swarms of galaxies colliding don鈥檛 look much different from a
single swarm just sitting there.

Instead, you need to look at the gas that surrounds the galaxies, outweighing
them by 10 to 1. This gas is so hot鈥攂etween 50 million and 100 million
kelvin鈥攖hat it emits X-rays. And to catch the X-rays you need to put a
detector on a satellite.

Our nearest big clusters, called Virgo and Coma after the constellations they
appear in, were mapped with the earliest X-ray telescopes in the 1960s. Their
gas seemed to have a uniform temperature, which led astronomers to believe that
clusters are settled, mature objects. But during the 1990s, the satellites ROSAT
and ASCA found hints that clusters were not simple clouds with all the gas at
one temperature. Now, two new missions have returned X-ray images of
unprecedented clarity.

NASA鈥檚 Chandra, launched in July 1999, and ESA鈥檚 XMM, launched six months
later, both take the temperature of the gas by looking at the frequencies of
X-rays it emits鈥攈otter gas emits higher-frequency X-rays. And they can do
it in extraordinary detail, so astronomers end up with a map of hot and cool
spots, rather than the blurry averages they got with older instruments.

Chandra and XMM show that clusters are in turmoil. 鈥淭he maps appear much more
chaotic than expected,鈥� says Benjamin Mathiesen of Stanford University in
California. This chaos can only mean that clusters are colliding.

Maxim Markevitch of the Harvard-Smithsonian Center for Astrophysics has used
Chandra to map several clusters, all of which seem to be scarred by violent
mergers. Two clusters in particular, Abell 2142 and Abell 3667, were
intriguing.

Markevitch was looking for shock waves in the gas, a sign of cluster
collisions. Shock waves would heat the gas suddenly, so he was expecting to find
rapid temperature hikes. But instead, he found a place where the cluster
temperature fell. 鈥淚t turns out that the temperature jumps in the opposite
direction,鈥� he says. Although initially disappointed, he soon realised that he
had discovered something important.

What he had actually detected was the undigested remains of the cluster鈥檚
last meal鈥攖he surviving core of a smaller cluster that had been consumed.
As the clusters collided, the smaller one had its outer, hotter layers stripped
away, leaving only a dense core of relatively cool gas. Markevitch and his team
named these areas 鈥渋ntergalactic cold fronts鈥�. He has also found the hot fronts
he was originally looking for. 鈥淚n at least two clusters with cold fronts, at
some distance ahead we also see weaker edges that must be the genuine shock
蹿谤辞苍迟蝉.鈥�

The level of violence in the mergers has taken many astronomers by surprise.
About half the nearby clusters studied have been caught eating something
substantial, perhaps a group of a few dozen galaxies. 鈥淲e now know that clusters
are accreting smaller blocks and bits all the time,鈥� says Andrew Fabian of the
Institute of Astronomy in Cambridge.

About 10 per cent of clusters are actually undergoing mergers right now.
Compared with the sudden release of fury that is a supernova or a gamma-ray
burst, these collisions are slow, insidious catastrophes that take millions of
years to complete. But in the process, they release huge quantities of
energy鈥攁bout 10⁵⁷ joules. It would take a trillion years for the stars of
the Milky Way to emit that much.

These grand conflicts are changing the face of the Universe. The sea of small
groups of galaxies won鈥檛 last, as most of them will end up being eaten by
clusters. Astronomers are now witnessing the age of the cluster.

That鈥檚 not the end of the story. In perhaps a few tens of billions of years,
there will be an age of even more massive structures. 鈥淭he next stage will be
superclusters,鈥� says Mathiesen. Astronomers already use the term supercluster to
refer to some collections of clusters. The Shapley supercluster, for example, is
an enormous gathering of about 20 clusters visible in the southern sky. But
until now, astronomers haven鈥檛 been sure whether Shapley and other so-called
superclusters are really just patterns in space, not bound together by gravity,
or whether the clusters within them will eventually fall together into a single
unit.

However, the high level of merging activity convinces many astronomers that
gravity will eventually pull most of these loose alliances of clusters together
into giant hordes of up to 100,000 individual galaxies. At that point, structure
formation will almost certainly have run its course. In that far future, the
expansion of the Universe will have driven the individual superclusters so far
away from one another that they will be effectively out of reach of each other鈥檚
gravitational fields.

Our Local Group of galaxies is even now being pulled towards the Virgo
cluster, and we will merge with it in perhaps a few tens of billions of years.
Eventually we will join the local supercluster.

So what will it be like? Galaxy clusters are already dangerous places for
their more delicate members鈥攕piral galaxies like ours. The spiral arms of
galaxies are where new stars form out of old gas. But when two galaxies clash,
their delicate spiral structure is destroyed, and they become dull elliptical
galaxies. In the future, as clusters pack in more and more galaxies, those
galaxies will be far more likely to smash into each other. The superclusters
will gradually turn all their spirals into ellipticals.

Often there is one elliptical galaxy in the centre of today鈥檚 clusters, a
monster that has grown by swallowing its neighbours. The nearest example is the
galaxy M87 in Virgo. Eventually, these central galaxies may consume so much that
they come to dominate the whole cluster. In the process, though, there will be
some escapees鈥攇alaxies and stars flung out of the cluster after a close
encounter with the central elliptical.

And there鈥檚 a monster within the monster. M87 contains a whopping black hole,
3 billion times as massive as the Sun, and it is believed that all other central
galaxies do too. In the supercluster age, such black holes might reach
unheard-of sizes, fed by stars and by a river of gas flowing into the centre of
the cluster鈥攁 so-called cooling flow. 鈥淭hat would be a neat way of
building a really massive black hole,鈥� says Fabian.

So in 100 or 1000 billion years, our Milky Way will have had its spiral arms
wrenched off. A larger elliptical galaxy, perhaps the giant elliptical at the
core of the supercluster, will have eaten it and much of it might even have
fallen into the cluster鈥檚 central black hole. What a way to go.