Among the biggest puzzles for early radio astronomers were the bright
‘radio stars’ that turned up on sky surveys but seemed to have no relation
to known astronomical objects. As the resolving power of radio telescopes
improved, it soon became clear that these objects were not stars at all,
but galaxies radiating enormous power from regions far outside their visible
confines. In the past three decades, the diversity of these objects has
multiplied, and we now have, in addition to the classical ‘radio galaxies’,
a zoo of exotic species, which are powerful emitters of cosmic energy, bearing
such names as Seyfert galaxies, N galaxies, BL Lac objects, blazars, QSOs
and quasars.
Astronomers now believe that all these objects are essentially similar,
being otherwise normal galaxies dominated by an extremely powerful source
of energy within their cores. They are known collectively as ‘active galaxies’.
One of the most intriguing and important phenomena shown by active galaxies
are cosmic jets.
High-resolution radio observations made with such telescopes as the
Very Large Array in New Mexico and MERLIN in Britain show that most active
galaxies have some kind of ‘jet’-a narrow feature associated with the nucleus
of the galaxy. Radio astronomers have recorded several hundred jets, ranging
in length from a few parsecs up to many hundreds of kiloparsecs, and in
shape from pencil-thin beams to billowing gushers. Despite the diversity
of jet sources, there does seem to be some order underlying their complex
structure. They fall neatly into two groups according to their radio brightness,
a classification first noted in 1974 by Bernard Fanaroff and Julia Riley
at the University of Cambridge.
Advertisement
The weaker, Type 1, sources usually have twin jets emerging from either
side of the central core. The jets broaden rapidly, each merging into a
diffuse cloud called a ‘lobe’. A typical example is the radio galaxy 3C449,
though the best studied jets in this class are those of the galaxies M87
(Virgo A), the most powerful galaxy in the Virgo cluster, and 3C120
both of which are untypically one-sided. The more powerful, Type 2, sources
tend to have only one, narrow jet which terminates in a sharp-edged lobe
with local brightenings called ‘hot spots’. A classic and much studied example
is Cygnus A, which was the first radio source to be identified with a galaxy.
Looking at pictures of these jets, it is hard to avoid the conclusion
that we are witnessing vast amounts of matter gushing from the nucleus of
the galaxies and streaming out into space. Although we must be careful not
to go by appearances alone, astronomers now believe that is precisely what
is happening. The jets are channels transporting matter and energy from
the centre of the galaxy and depositing it in the lobes. Although no two
jets look the same, they all seem to originate from a point-like core that
coincides with the nucleus of the galaxy. It is to this nucleus that astronomers
look for the energy source that powers the jets.
Powered by a black plug hole
Whatever the nature of this ‘central engine’, it not only produces up
to 1000 times the power of a whole galaxy from a volume not much bigger
than the Solar System, but it also maintains that output for millions of
years. No telescope is powerful enough to show us the engine itself, and
we have to guess its nature from the effects it has on its surroundings.
A consensus is emerging that the energy source must be gravitational, and
the most likely candidate is a massive black hole with a mass anywhere between
a million and a billion times that of the Sun.
In this picture, a spinning black hole is fuelled by interstellar gas
which is drawn in by the hole’s intense gravitational field, rather like
bath water going down a plug hole. The matter falling inwards crashes into
a fat, swirling ‘accretion disc’ that envelops the hole except for two funnel-shaped
openings above the poles. By means not altogether clear, material is accelerated
out of the funnels to create two opposed jets.
Depending on the mass of the hole, the rate with which it is supplied
with fuel, the nature of the surrounding medium and the orientation of the
funnels to our line of sight, we can explain many of the diverse properties
of active galaxies. This ‘twin-beam’ model is the starting point for trying
to understand the physics of jets, but as we shall see there are still many
questions left unanswered.
Not the least of these is: ‘How are the jets generated?’ Although we
cannot probe the region in which the jets originate, techniques based on
very long baseline interferometry, (VLBI-see Box) allow astronomers to pick
up the jets as close as a few light weeks from the nucleus in the nearest
sources. By making observations over several years, we can follow the motions
of blobs in the jets and, knowing the distance, work out the velocity of
expansion.
In 1966, Martin Rees at the University of Cambridge predicted that material
ejected from a quasar at relativistic speeds (close to that of light) would,
if directed towards the observer, appear to have a velocity much greater
than the speed of light. This paradoxical ‘superluminal motion’ remained
a theoretical oddity until 1970, when a team of American and Australian
astronomers announced that they had detected superluminal motion in the
powerful quasar 3C279. About 30 sources are now known to be superluminal,
including galaxies, quasars, BL Lacs and blazars.
There are grave difficulties in working out what kind of mechanism could
draw energy from a black hole that would accelerate material up to the highly
relativistic speeds revealed by superluminal motion. Some astronomers wonder
whether the source of energy may be something other than a supermassive
object, but the black hole remains the best working hypothesis.
Astronomers have made good progress in understanding how the jets carry
energy from the core and deposit it in the lobes. VLBI studies reveal that
jets emerge from the nucleus as a continuous fluid flow, often tracing curved
paths, and it is clear that we are dealing with a complex problem in fluid
dynamics. The emission from jets is of a form familiar to radio astronomers
and known as synchrotron radiation. This tells us that the jets contain
charged particles, probably electrons, moving in a magnetic field at close
to the speed of light. The field causes the electrons to wind along spiral
paths and, as they do so, their kinetic energy radiates away as radio waves.
By studying the polarisation of the waves, the direction in which they vibrate,
we can map the magnetic field.
Apart from that, we know very little about the material making up the
jets. Presumably the electrons are accompanied by positively charged particles,
but given our ignorance of how jets are created, it would be a mistake to
assume that they must be protons. Indeed, according to one theory the jet
is made of a mixture of electrons and positrons. For synchrotron radiation
to be effective over a period of thousands or millions of years, the electrons
have to be kept supplied with energy. This energy is derived from the bulk
kinetic energy of the jet itself and is transferred to the electrons as
the jet is slowed down by the resistance of the intergalactic gas.
Although superluminal motion seems to be very common deep in the cores
of jet sources, there is a surprising dearth of hard evidence that there
is any motion along the outer, large-scale jets that make these objects
so distinctive. Astronomers have measured the velocities of jets only in
galaxies M87 and 3C120. These, too, reveal flows close to the
speed of light.
Most of the progress in understanding how jets propagate has been made
by computer simulation. For example, it seems that jets emerging from the
nucleus are lighter than the tenuous interstellar gas through which they
pass. Jets from the weaker sources rapidly slow down as they plough through
this gas, while the stronger jets remain what is called ‘supersonic’, that
is, their forward velocity is greater than the random speed of the particles
within them.
A central problem of understanding jets is the one of collimation and
confinement, that is, how they are focused into tight beams and remain jet-like
over distances up to hundreds of kiloparsecs and for periods up to a million
years. If they contain energetic particles, why don’t they splay out as
soon they leave the nucleus?
A jet moving at supersonic speeds, even if allowed to expand freely,
will have sufficient forward velocity for it to remain quite narrow. In
general though, the physics is complicated and depends on the speed and
density of the jet and the detailed properties of the surrounding gas and
magnetic fields.
Until recently it was thought that the pressure of the intergalactic
medium, though very low, was enough to confine a jet. Estimates of the pressures
within the jets were comparable to those outside. But it now seems that
this explanation will not do for all jets. Recent observations of the jets
in Cygnus A and M87 show that their internal pressure is actually well above
that of their surroundings. In such cases, it is likely that the magnetic
fields within the jets play a part in keeping them confined.
The bright lobes seen at the ends of jets, which also emit synchrotron
radiation, are believed to be cavities swept out of the intergalactic medium
and filled with material conveyed along the jets. In the stronger sources,
the end point of the jet corresponds to a local brightening or hot spot,
where the stream is strongly interacting with the intergalactic medium,
accelerating particles to relativistic energies. The lobe itself consists
of material flowing away from the advancing hot spot back towards the nucleus.
The closer we look at jets the more we find they are not the smooth,
straight tubes of the simpler computer models. Almost every jet shows some
kind of irregularity, such as small bright spots (‘knots’), wiggles, kinks
and even sharp bends. Some of these may be related to irregular outflow
from the central engine, carried outwards like ridges on a fingernail. Some
jets, like the one in M87, have a wavy structure as if material is issuing
in a corkscrew fashion. This may be evidence that the black hole in its
core is precessing, wobbling on its axis, somewhat like a rotating lawn
sprinkler. For this to happen requires the presence of a nearby massive
object, such as a second black hole or a companion galaxy.
Some theorists argue that a double black hole may be a more plausible
energy source than a single hole, and could arise from the collision and
merger of two massive elliptical galaxies. Indeed, some sources, such as
galaxy 3C75, clearly have two nuclei, each with their own twin jets.
Waves and other irregularities may also arise inside the jet itself.
Heavy jets are prone to internal instabilities that could show up as local
brightenings. The state of the surrounding medium is also likely to play
a part. If it is clumpy, then even the light, supersonic jets are likely
to be deflected giving rise to some of the bends and knots.
Blowing in the intergalactic wind
In many sources, we see jets that seem to have been distorted or disrupted
on a vastly larger scale than can be explained by processes within the jets
themselves. Curved jets are common, and may often be swept back like plumes
of smoke, as if they are being blown by some intergalactic wind. In others,
the jet suddenly splashes out as if it has collided with an object far outside
the core.
As noted earlier, the more powerful sources usually show only one jet
where the twin-beam model would predict two. Quasars, interestingly, only
ever show a single jet, such as the spectacular one in 0800+608, though
they usually have two lobes. The most likely explanation is that there are
really two jets, but one is too faint to be seen.
Why should one jet be brighter than its companion? If the jets are flowing
at speeds close to the velocity of light, an effect called relativistic
beaming will concentrate their emission in the forward direction. A jet
coming towards an observer will appear much brighter than an identical jet
moving away. For this to work, the jet axis must be aligned roughly towards
us; the faster the jet the tighter this constraint and the greater the contrast
in brightness.
Support for this explanation comes from recent studies of a phenomenon
known as ‘depolarisation’. The jet side of the source is always more strongly
polarised than the opposite side. This is to be expected if the jet side
is nearer to us and its emission is less affected by passage through intergalactic
gas.
Some astronomers argue that this apparent one-sidedness is not a relativistic
effect but intrinsic to the source. For example, in the so-called ‘flip-flop’
model the outflow alternates between one side of the nucleus and the other,
but most astronomers are reluctant to accept this theory so long as relativistic
effects provide an adequate explanation of one-sidedness. If we accept the
relativistic beaming explanationfor one-sided jets, then we are left with
another awkward problem. All quasar jets are one-sided, which implies that
they are all pointing towards us. Where are the quasars that are not pointing
towards us? The answer may lie in the suggestion that the quasars are classical
radio galaxies, with two jets like Cygnus A, but are seen as quasars only
when their axes are pointed at us.
Jets are among the most fascinating and intriguing phenomena of modern
astrophysics. But despite the wealth of high quality radio images, we are
far from a complete understanding of how jets originate and how they propagate.
We could do with equally sensitive and detailed images at other wavelengths,
but these will not be forthcoming in the foreseeable future. We also need
more sophisticated and realistic computer models than those available at
present.
In the next few years, as even more powerful radio telescopes become
available , we can expect a deluge of new and spectacular images of active
galaxies and their jets. But it remains to be seen whether our capacity
for understanding these images can keep pace with our virtuosity in creating
them.
Chidi Akujor is a senior lecturer in the Department of Physics and Astronomy
at the University of Nigeria at Nsukka and visiting research fellow at Jodrell
Bank. Tony Jones is an astronomer and science writer.
* * *
Many radio eyes are better than one
High-Resolution radio images of galaxies and quasars have been made
possible by the technique of ‘aperture synthesis’ developed by Martin Ryle
and his colleagues at the University of Cambridge in the 1950s. The angular
resolution of a radio telescope, a measure of the fineness of detail it
can see, depends on the diameter of its reflecting dish or ‘aperture’. In
aperture synthesis, several small telescopes are linked by cables to operate
as a single instrument. Their combined signals are processed by computer
to make an image. The angular resolution then depends not on the size of
the individual telescopes, but on their maximum separation or ‘baseline’.
The longer the baseline, and the shorter the wavelength, the more detail
can be seen.
The first successful aperture-synthesis telescope, the Cambridge One-Mile
Telescope, began work in 1964 and has since been followed by several others,
notably the 2.8-kilometre array at Westerbork in the Netherlands and the
4.6-kilometre Ryle Telescope at Cambridge. Arrays like these can achieve
typical resolutions of a few arcseconds.
Most of the pictures accompanying this article come from the most celebrated
of all synthesis telescopes, the Very Large Array of 27 dishes in the New
Mexico desert. With a maximum baseline of 36 kilometres, the VLA can achieve
a resolution of less than 1 arcsecond and is responsible for some of the
finest images in radio astronomy.
Still larger arrays can be built by linking telescopes by radio beams
rather than cables. In Britain, Jodrell Bank’s MERLIN array of eight telescopes
now stretches 220 kilometres across central England, achieving a resolution
of 0.05 arcseconds at short wavelengths.
New arrays include the 6-kilometre Compact Array of the Australia Telescope
in Narrabri, which should be completed this year, and the Giant Metrewave
Radio Telescope of 30 dishes spanning a baseline of 25 kilometres, under
construction at Poona in India.
Even higher resolutions can be achieved by very long baseline interferometry
(VLBI) in which independent observatories agree to observe the same objects
at the same time and record the radio signals on magnetic tape. The tapes
are later replayed together and the signals combined as if coming live from
the telescopes. Observatories all around the world now routinely take part
in VLBI sessions, and the very long baselines-up to the diameter of the
Earth-allow astronomers to achieve resolutions as fine as 0.001 arcseconds.
Both the US and Australia are creating dedicated VLBI arrays over baselines
the size of a continent.
We should be able to do even better with the proposed International
VLBI Satellite. This will be a radio telescope in an elliptical orbit around
the Earth, operating with an array of telescopes on the ground. Baselines
will be at least 10 times longer than Earth-bound VLBI, with correspondingly
better resolution.
Such a resolution of 0.0001 arcseconds corresponds to distances from
about 0.01 parsec in nearby radio sources to about 0.5 parsec in the most
distant galaxies that can readily be observed with VLBI. (Interstellar distances
are typically a few parsecs.) The radius of a black hole with a mass of
a billion times that of the Sun is only 10-4 parsecs (3 billion kilometres).
So even with these powerful techniques we are still far from penetrating
the innermost parts of active galaxies.