¿ìè¶ÌÊÓƵ

THROUGHOUT the long history of astronomy, researchers have had one main
way to understand the Universe: from the message carried by electromagnetic
radiation. Although this type of radiation covers a wide range in frequency
– from gamma-rays to radio waves – produced by many different sources in
the Universe, electromagnetic waves are caused by the same basic physical
process, the acceleration of electric charges.

Astronomers could add to this knowledge slightly by studying some particles
that reach us from space: the ‘cosmic rays’ that are in fact high-speed
atomic nuclei, and the highly-penetrating neutrinos from the Sun and from
the nearby supernova which exploded in 1987.

But now astronomers are poised to make a breakthrough in observing the
cosmos. Within the next few years, four or five new ‘telescopes’ around
the world will begin to pick up an entirely new kind of signal from space.

These gravitational waves are shudders in the fabric of space-time itself.
The ‘gravitational wave telescopes’ should detect radiation from known sources
of cosmic violence, such as nearby supernovae, and from events such as the
crashing together of two neutron stars as they spiral in towards one another.
More exciting is the chance of making new discoveries about the Universe
that are not predicted by our current knowledge and theories.

Gravitational radiation has been around, in the minds of theorists at
least, for many decades. Albert Einstein deduced that it must exist, as
a natural corollary of his theory of gravity, the general theory of relativity.
One simple way to visualise general relativity is to think of space-time
(reduced to two dimensions) as a sheet of rubber. A mass, such as a star,
placed on the rubber forms a depression in the sheet. If the mass wiggles
about, ripples – waves of gravitational radiation – will spread out in the
sheet.

This is similar to the way in which charged particles, such as an electron,
produce electromagnetic waves. But there is one fundamental difference.
A single electron may oscillate without disturbing any other electric charges
causing what is called a dipole. The result is electromagnetic ‘dipole’
radiation which is relatively intense.

For a star to produce gravitational radiation, however, then a force
must be applied that reacts on another mass. According to Einstein as well
as Newton, this reaction must be ‘equal and opposite’. So, the second mass
wiggles too, and the gravitational waves from it more or less cancel out
those from the first mass. In fact, a detailed calculation shows that the
motion of the second mass always wipes out the dipole radiation of the first
mass completely. All that is left is a much weaker ‘quadrupole’ radiation.

This means that accelerating masses do not always produce gravitational
waves. One example would be a star that explodes as a perfectly symmetrical
sphere of gas. In such a system, the quadrupole radiation is cancelled as
well. So an absolutely spherical supernova would generate no gravitational
waves however violently it exploded. In practice, however, astronomers think
that a supernova is rather messier than this and is sure to generate some
gravitational radiation.

Astronomers have thought of at least two other kinds of star systems
that should produce a detectable amount of gravitational radiation. A neutron
star is a very compact object, made almost entirely of neutrons packed cheek
to jowl, so it would have an extremely high gravitational field. Many neutron
stars spin rapidly, producing pulses of electromagnetic radiation. If a
spinning neutron star has a small ‘mountain’ on its surface, it becomes
an excellent generator of gravitational waves.

Any two objects orbiting each other also generate gravitational waves.
In the case of the Sun and the Earth, these waves are too weak to be measured.
But if we have two neutron stars very close together, orbiting at very high
speed, then they will pump out a significant amount of gravitational radiation.
In 1974, radio astronomers found such a pair of neutron stars, swinging
around each other in just under eight hours. Over the years, their mutual
orbit has gradually shrunk, implying that the system is losing energy –
and at exactly the same rate as theory predicts for the energy carried away
by gravitational waves. There is little doubt that this pair of neutron
stars is generating gravitational radiation, although it has not been detected
directly.

Bernard Schutz of the University of Wales in Cardiff pointed out that
a pair of stars, such as this one, will eventually spiral together and merge
into one body. In its final few seconds, the star system should emit a crescendo
of gravitational waves: a loud and characteristic burst of radiation that
would sound like a ‘chirrup’ if we hear it.

The gravitational radiation from these sources can carry a lot of energy.
The burst of gravitational waves from two stars coalescing in a galaxy 100
million light years away would arrive at the Earth with a power – if we
converted it into terms of optical light – equal to the brightness of the
full Moon.

As this ripple in space-time spreads through the Universe, it distorts
the shape of any object it passes through. This gives us the means, in theory
at least, to detect gravitational waves. It will squeeze a piece of matter
in one direction and stretch it in the perpendicular direction. The shape
oscillates in time with the frequency of the vibration of the wave as it
passes through.

But in reality, the change in size of the object is absolutely minuscule.
The gravitational waves from a typical astronomical source would deform
a piece of matter by about one part in 10**2 of matter a metre across would
change in size by much less than the diameter of an atomic nucleus.

Undeterred by such predictions, an American physicist, Joe Weber, decided
in the 1960s to try to detect these waves. His ‘lump of matter’ was a cylindrical
bar, weighing a tonne and made of aluminium. Over the intervening years,
many other groups around the world have made more and more sophisticated
detectors of this type, usually cooled to a few degrees above absolute zero
to improve their sensitivity .

Despite all this progress, most physicists believe that bar detectors
have reached the limit. One limitation is that they can detect gravitational
waves with only a limited range of frequency, near to the tone at which
the bar naturally vibrates. More fundamentally, the experimenters are now
looking for such tiny movements that they are running into Heisenberg’s
uncertainty principle, whereby the very act of measuring the changing size
of the bar affects the dimensions of the bar by a comparable amount. Along
with more recent and pessimistic estimates of the amount of change a gravitational
wave would produce, this limitation means that even the most sophisticated
bar detectors will be able to detect only the most spectacular – and hence
comparatively rare – outbursts of gravitational waves.

A novel idea for detecting gravitational waves has, however, led to
a huge resurgence of interest in gravitational wave astronomy. It promises
a more sensitive kind of detector that can pick up much weaker signals,
much more often.

After Weber claimed he had detected these waves, Bob Forward, an American
scientist then working for Hughes Aircraft, suggested an entirely different
kind of device.

Forward argued that a gravitational wave changes the size of a detector
by the same tiny fraction, regardless of the detector’s size. If we use
a detector that is as large as possible, then we would have a change in
size that is larger, and so is easier to measure. Forward’s scheme was simply
to measure the change in distance between two separate masses that are a
long distance apart using a laser. In 1973, Forward built the first prototype
of this type of detector, with masses 10 metres apart. His experiments led
to improvements in laser technology that Hughes incorporated in displays
for aircraft cockpits, but Forward could not carry on to build larger detectors
for pure astronomical research.

A few years later, Rai Weiss, of the Massachusetts Institute of Technology
developed the theory further, while experimenters at Munich in Germany and
the University of Glasgow in Scotland began to build prototypes that were
30 to 40 metres long. The leader of the Scottish group, Ron Drever, was
later scooped up by Caltech in California to research gravitational waves
there. As interest has grown, physicists in France, Italy and Japan have
also started to test prototype detectors. These groups are now confident
that they can construct full-scale gravitational wave telescopes – several
kilometres in size – that are capable of detecting gravitational radiation
if our present theories are anywhere near correct.

All the new detectors work in the same simple way: a laser beam is used
to measure the distance between two mirrors, fixed to separate masses. When
a gravitational wave passes through the detector, the distance between the
masses alters, leading to a change in the time it takes for the laser beam
to return. The gravitational wave also affects the propagation of the light,
but the two effects do not cancel each other out.

In practice, it is easier to measure the difference in the time taken
for the light to travel in two different directions, at right angles: the
gravitational wave makes one of these distances increase and the other diminish.
Even more important, you do not actually have to time the light. You let
the beams interfere with one another, and the changing lengths of the arms
appears as a change in the interference fringes. This is, in essence, a
Michelson interferometer, of the type that undergraduate physics students
quickly come to swear at.

By making the arms of the interferometer as long as possible, you can
make it more and more sensitive to gravitational waves. But you also run
into practical problems. First, the laser beams cannot simply travel through
air. Changes in the refractive index of air would easily mask the effect
being sought. So the path of light, along both arms, must be enclosed in
long metal tubes that are pumped to a very high vacuum.

According to the theory, we need arms that are hundreds of kilometres
long if we hope to detect a reasonable number of weak bursts of radiation.
Clearly this is not feasible, especially when the beams have to travel in
evacuated tubes – even for arms a few kilometres long, these tubes and the
powerful vacuum pumps account for most of the cost of the instrument. The
answer is to bounce the light back and forth within each tube, up to a hundred
times. In some of the prototypes, a narrow laser beam reflects off different
parts of the mirrors each time, until it escapes back through a small hole
in the mirror at the centre of the device. The groups at Glasgow and Caltech
have taken a slightly different approach. Their mirrors form a Fabry-Perot
etalon: the distance between the mirrors is an exact number of wavelengths,
so the reflected light sets up a pattern of standing waves.

Either way, the researchers depend on highly reflective mirrors to reflect
the laser beam up to a hundred times without it becoming significantly dimmer.
These instruments could not have been built at all without the development
of mirrors that absorb less than one ten-thousandth of the light falling
on them. These mirrors are a spin-off from the development of laser-ring
gyroscopes, used in the navigation systems of military aircraft and cruise
missiles, where a beam of light reflected round and round acts as a steady
reference against which the electronics can measure the motion of the craft.

Another important component is the laser itself. It must produce light
that is always precisely the same wavelength. Drever, at Glasgow and then
Caltech, has pioneered ways of controlling the laser so that the wavelength
it produces is extremely stable. The laser must also be bright enough for
its light to undergo many reflections and still be detectable. The prototype
instruments generally use argon lasers, with an output of a few watts, but
the groups are now working on much more powerful neodymium-yttrium-garnet
lasers, which should produce about 100 watts.

The research groups that are about to build full-scale gravitational
wave detectors are not so much in competition as in a loose collaboration.
Although there will be a certain kudos attached to the first definite detection
of gravitational waves, it is an honour that no group expects to have to
itself. The history of the subject has taught the experimenters that other
scientists will be sceptical of a claimed detection unless it is backed
up by results from a completely independent instrument.

The other advantage of working together is that a single gravitational
wave detector can tell astronomers very little about the direction from
which the radiation has come. As the wave sweeps through the Earth, at the
speed of light, it will, however, pass different detectors at different
times. If the experimenters compare notes of the exact times when the waves
are passed through each detector, they can calculate the direction in which
it was travelling.

The more spread out the detectors are, the more accurately the researchers
can determine the direction of the source of the gravitational waves. With
a set of detectors spanning the Earth, they should be able to pin down the
direction of the source to within a few arcminutes (about a quarter the
apparent size of the Moon). For sources beyond our own Galaxy, this should
be good enough for the astronomers to identify the galaxy that produced
the radiation. This accuracy will also allow astronomers to check if there
was a burst of electromagnetic energy at the same time – as would be the
case with a supernova.

For direction-finding, the best way of locating the source of the radiation
is to use four detectors at the corners of a tetrahedron. More by luck than
anything else, the research groups interested in building the detectors
are spread out in roughly that way: on the east and west coasts of the US,
in Europe and in Australia.

The American groups at Caltech and MIT have formed a consortium, the
Laser Interferometer Gravitational Wave Observatory, to construct a pair
of identical instruments, one on each coast. The western detector may be
at the Edwards Air Force Base in California (where the space shuttle lands)
and the eastern one in Maine or Louisiana. Each will have arms that are
four kilometres long, with their light paths side by side, so that several
different gravitational wave detectors can work simultaneously. These could
be of slightly different design, or respond to different frequencies. Both
the National Science Foundation and President Bush have backed the proposal,
worth $192 million (about Pounds sterling 118 million) and it is awaiting
approval from Congress later this year.

All being well, Europe will also have two gravitational wave detectors,
though these are being built by separate teams. A French-Italian consortium
is planning a gravitational wave detector with arms three kilometres long,
that will be sited near Pisa. Researchers at the University of Glasgow and
the Max Planck Institute for Quantum Optics at Garching are also designing
a detector with three-kilometre arms. Jim Hough, at Glasgow, has already
located a suitably flat and accessible piece of land in Tweedsmuir Forest,
near St Andrews, and the instrument has been granted planning permission.
But the location will almost certainly be decided by the relative financial
contributions from the two countries. In July, the British Science and Engineering
Research Council allocated Pounds sterling 5.5 million to the telescope.
Researchers expect that the Germans will soon commit themselves to providing
the rest of the Pounds sterling 30 million required. The contributions will
be split between the Max-Planck Institute, the Federal Ministry for Research
and Technology (BMFT) and the local government of Lower Saxony.

All these detectors are in the northern hemisphere, and at much the
same latitude. In order to tell accurately how far north or south a source
lies in the sky, astronomers must also use a detector in the southern hemisphere.
To complete the ‘tetrahedron’ at its southern corner Australian and Japanese
scientists are planning to build an instrument near Perth. The Japanese
research group at ISAS, near Tokyo, has already built a prototype with arms
10 metres long. They will work with the Perth group that already operates
a cooled bar detector and with optical experts from the Australian National
University. Although the Australian federal government has yet to approve
the project, the state government of Western Australia has already promised
A$5 million (about Pounds sterling 2 million) of support, in the form of
land, roads and electricity.

All the groups are confident that their experience with the small prototypes
has ironed out all the major problems. It will take about three years to
build a full-scale instrument, and another two or more years to test it
out and get it running smoothly.

By the late 1990s, then, astronomers should have ploughed well into
the new field of gravitational wave astronomy.

Every year, the detectors should be picking up several bursts of radiation
from supernovae in other galaxies, out to a distance of perhaps 100 million
light years. There should be dozens of ‘chirrups’ per year from neutron
stars that are running together and coalescing in galaxies up to 300 million
light years from us. Bernard Schutz has calculated that by measuring the
gravitational waves alone we can we can work out how far it is to such a
coalescing star pair: by identifying the parent galaxy and measuring its
speed away from us, astronomers will be able to calculate the rate at which
the Universe is expanding (the Hubble constant). This will help astronomers
locate more precisely more remote galaxies and quasars, and provide new
evidence on the age of the Universe.

Still in the realm of cosmology, some theorists have predicted that
the big bang produced ‘cosmic strings’ – discontinuities in space-time that
act like long, thin strings with a strong gravitational pull. Loops of cosmic
string could have acted as seeds around which the galaxies formed. If this
were the case, the loops of string would get smaller and eventually disappear,
producing a burst of gravitational waves in the process. The new detectors
may well find this radiation forming a background to their other measurements.

And, at the most basic level, detecting gravitational waves at all is
an important test of general relativity. The experimenters will also be
able to test if the polarisation of the wave matches up with the predictions
of relativity.

But the most exciting prospect is that a new window on the Universe
opens up the chance of unexpected discoveries. Radio astronomers stumbled
upon quasars, and X-ray astronomers were unprepared for black holes. Who
knows what objects may exist in the Universe that choose to reveal themselves
only by their output of gravitational radiation.

* * *

1: HOW TO OBTAIN HOT RESULTS FROM COLD METAL

THE HEART of the first gravitational wave telescope was a bar of aluminium,
1.5 metres long. A passing gravitational wave would alter its length by
less than the diameter of atomic nucleus. Its designer, Joe Weber of the
University of Maryland, fitted very sensitive electronic transducers to
the ends of the bar, to record changes in its length. He knew that the transducers
would show that the bar was continuously shrinking and expanding by small
amounts, just because of the randomly changing vibrations of the atoms.
Normally, such vibrations would have swamped any disturbance from a gravitational
wave.

This was where the choice of aluminium came in. If you hit a suitably
supported bar of aluminium, it will keep on ringing for a very long time,
perhaps for hours. Each random vibration of an atom acts like a minute hammer
blow, and sets the bar ringing, so the changes in length from these vibrations
is averaged out over a period of hours. A gravitational wave produces a
much smaller signal in the detector, but because it changes the length of
the bar in only a fraction of a second this signal will stand out from the
much larger but very gradual changes caused by the vibrating atoms.

In 1969, Weber confidently announced the discovery of gravitational
radiation: his bar was detecting one gravitational wave burst per year.
This result was astonishing. The strength of the signal Weber claimed to
have measured meant that our Galaxy was destroying a star like the Sun every
year. At that rate, one-tenth of its matter would have been destroyed since
the Galaxy was born.

This extraordinary claim spurred other scientists to build similar detectors,
to see if they could duplicate Weber’s result. In the end, no one could.
To this day, it is not clear what Weber was detecting – most probably some
kind of noise in the apparatus. But the episode had important consequences.
It opened researchers’ eyes to the fact that a gravitational wave detector
could be built. But they would need simultaneous detections from at least
two different detectors before other scientists would be convinced of the
discovery of a burst of gravitational waves. Weber’s work also prompted
theorists to take a closer look at the generation of gravitational waves.
They concluded that the waves would be even weaker than first believed.

Clearly, the experimenters needed to improve the sensitivity of their
detectors.

Since the main problem is noise from the natural vibration of the atoms,
it helps to cool the bar down to just a few degrees above absolute zero.
Several groups around the world have now built such ‘cooled bar’ detectors.
There is an Italian detector at the European particle physics laboratory
near Geneva (CERN) and there are detectors in Japan and in Louisiana in
the US. Physicists at Stanford in California have been operating the largest
such ‘telescope’, an aluminium bar weighing five tonnes: the more massive
the bar, the better it responds to both gravity and gravitational waves.
This instrument was damaged in the Californian earthquake last year, focusing
the group’s work on a new and even more sensitive aluminium bar detector.

At the University of Western Australia in Perth, David Blair has gone
for quality rather than quantity. Instead of using larger amounts of aluminium,
he has investigated which materials will ‘ring’ for longest when struck.
The best is sapphire, but it would be difficult to obtain a lump of pure
sapphire weighing a tonne or more. So Blair has opted for a 1.5-tonne bar
of niobium, the metal that will ring for longest. Because the metal is not
consumed in the experiment, the bar worth A$500 000 (about Pounds sterling
200 000) is owned by the university’s investments board. Blair is still
sorting out problems in preventing vibrations from reaching the bar from
the outside world; later this year he expects to reach the same sensitivity
as the biggest aluminium bars.

Japanese researchers have taken a different tack. Their detectors are
tuned to the frequencies of some nearby rotating neutron stars that are
powerful emitters of electromagnetic radiation, such as the neutron star
in the centre of the Crab Nebula. By looking for just one frequency that
is known in advance, they can filter out most of the noise.

* * *

2: THE SUPERNOVA AND THE WARM BAR

AT THE MOMENT, the most sensitive detectors are the big cooled bars:
they are about 10 times better than the prototype laser interferometers.
The cooled bars should certainly be able to detect cataclysmic events in
our Galaxy or our closest neighbours, although such events are rare.

By a strange irony, one such event did occur recently: supernova 1987A
in the Large Magellanic Cloud. At the time of the outburst, however, none
of the cooled bar detectors was running. They operate for only a few months
a year, with the rest of the time being spent in improvements, or simply
in cooling down and warming up the tonnes of metal.

But the supernova saga has a strange footnote. There are a few bar detectors
that are not cooled at all. These are much less sensitive than the cooled
bars, and few scientists expect them to detect anything astronomical. But
two ‘room temperature bars’, in Maryland and Rome, were operating when the
supernova exploded, and seemed to have registered a signal.

If this signal was real, the wave was carrying the energy equivalent
of thousands of Suns, while the star that exploded was only as heavy as
20 Suns. The Rome group think their ‘signal’ was probably just noise that
happened to occur at the right time. Weber, in Maryland, believes that it
was a real signal, and that the theory of the detection of gravitational
waves is wrong: he claims that bar detectors are thousands of times more
sensitive than any one had previously thought.