THE ancients had no problems distinguishing stars and planets: stars are
fixed in the firmament, planets move relative to them. Ours is a more
sophisticated age. We don’t know what the difference is.
To find the answer, we must enter the twilight world of the brown dwarf.
Brown dwarfs are hardly stars at all. Too lightweight to trigger hydrogen fusion
in their cores, they glow so feebly that astronomers found direct evidence for
their existence only four years ago. At about the same time, astronomers began
to discover giant planets orbiting other stars.
The smallest brown dwarfs are almost indistinguishable from the giant
planets. No one can say for certain which objects are planets and which are
stars. This could be crucial not only to understanding the nature of the
mysterious “dark matter” that makes up 90 per of the mass of the Universe, but
even in the search for extraterrestrial life.
Advertisement
It’s the desire to find out whether or not we are alone in the Universe that
has driven astronomers to scour the heavens for evidence of planets around other
stars. But if some of the giant planets they have found are really brown dwarfs,
planetary systems such as our own might be much rarer than we think. What’s
more, until we learn the differences between the smallest stars and the largest
planets, and how they form, it will be hard to recognise solar systems that
might hold Earth-like planets capable of supporting life.
Some of the latest research hints at a true demarcation between tiny stars
and giant planets, something that would go a long way towards solving these
problems. But it is controversial, because astronomers are straining to detect
the dimmest of objects.
Brown dwarfs can be no more than 80 times the mass of Jupiter, or about a
thirteenth of the mass of our Sun. In more massive stars, hydrogen nuclei smash
together at such high speeds that they fuse, releasing energy. But below 80
Jupiter masses (MJ) the pressure is too low to trigger these reactions.
Instead, brown dwarfs shine by converting gravitational energy into heat and by
fusing their small stock of deuterium, a heavy isotope of hydrogen. As a result,
they are cool and dim, and most of their feeble radiation is emitted as infrared
rather than visible light.
That makes brown dwarfs maddeningly difficult to detect. Until 1995, their
existence was supported only by theory and indirect observations. Then came the
discovery of Gliese 229B, the first brown dwarf to be imaged directly. Gliese
229B is 100 000 times less luminous than the Sun and has a surface temperature
of just 650 °C—positively chilly compared with the Sun’s 5500 °C.
Its mass is probably around 50 MJ. Although it is 19 light years away,
astronomers were able to capture its feeble light using the 60-inch telescope at
the Palomar Observatory in California.
Since then, about a dozen more possible brown dwarfs have been found within
300 light years of us. Most of them are lightweight companions of stars like the
Sun.
These dwarfs are thought to form in the same way as ordinary stars, when a
cloud of gas and dust collapses under its own weight
(see Diagram).
Physicists believe that the smallest clump of matter that can undergo such a
process is about 10 MJ. Any smaller than that, according to most
computer models, and the pressure of the gas can easily counteract the cloud’s
feeble gravity.
Planets, in contrast, are thought to form where small chunks of rock and
debris come together in a disc of material orbiting a young star
(“Seven planets for seven stars”, 15 June 1996, p 26).
This “bottom up” process, rather than the
“top down” star-formation mechanism, produces a solid core, which eventually
becomes massive enough to draw in gas and dust from neighbouring regions of the
disk.
Unfortunately, no one knows how big a planet can be. According to some
theoretical models, they can grow well into the 10 MJ range, making
them about the same size as some brown dwarfs. Naturally, this leaves many
astronomers frustrated. Just agreeing on a definition of “planet” and “brown
dwarf” is hard enough; telling them apart from dozens of light years away could
be almost impossible.
“It’s pretty confusing,” admits Gibor Basri of the University of California
in Berkeley. “We don’t know yet whether there’s a continuum of processes, or
whether there are really two very distinct processes.”
So it may be futile to try to split planets from brown dwarfs on the basis of
how they were formed. As Alan Boss of the Carnegie Institution of Washington
points out, the “end product” could be similar even if the formation process is
different. “If you take a few Jupiter masses worth of solar-composition gas and
dust, and compress it into a body, it’s going to look pretty much the same
whether you make it by the brown dwarf mechanism or by the giant planet
Ծ.”
For Basri, the most logical way to distinguish the stars from the planets is
to consider what’s going on in their cores. Theorists have shown that brown
dwarfs can sustain deuterium fusion at masses as low as 13 MJ, so
perhaps this value is the best dividing line. “If it’s not even heavy enough to
burn deuterium, then I want to call it a planet,” says Basri. However, many
astronomers want to reserve the label “planet” for objects that form in discs,
even if they outweigh the smallest of the brown dwarfs.
All this confusion would be dispelled if it turned out that planets and stars
occupy different mass ranges. But measuring the mass of dim and distant bodies
is no easy task. The most common method is a radial-velocity study, in which
astronomers analyse the spectrum of light from a normal star. If it has a
companion such as a giant planet, the star will wobble back and forth as the two
orbit each other, producing a Doppler shift in the star’s spectrum. The greater
the wobble, the bigger the invisible companion must be.
Missing middleweights
Unfortunately, astronomers can deduce only a lower limit to the invisible
companion’s mass. That is because this method measures the wobble along our line
of sight. If we happen to be seeing the system edge-on, the measured velocity is
the true speed of the wobble. But if the system is tilted to our line of sight,
the true speed is higher and the invisible companion must be more massive. To
pin down the angle of the orbit, the radial velocity must be combined with an
“astrometric” measurement, in which the position of the main star on the sky is
tracked.
Michel Mayor of the University of Geneva and two French colleagues have done
just that. They carried out an astrometric survey of 10 brown-dwarf candidates
using data from the Hipparcos satellite and presented their results at a
conference in Lisbon last year. They found that most of the candidates have
orbits almost perpendicular to our line of sight —and therefore are much
more massive than previously guessed.
“Most of the candidate brown dwarfs are in fact low-mass compact stellar
companions and not real brown dwarfs,” Mayor says. In other words, many of these
objects are simply normal, hydrogen-fusion powered stars. In the range of 5
to 50 Jupiter masses, very few objects have been found, Mayor says.
Satisfyingly, this implies that there is a true divide between the largest
planets and the smallest stars.
Mayor’s work agrees with an analysis carried out by a team of American and
Israeli scientists. They counted up the planets and brown dwarfs in a few broad
mass ranges to see if there really is a gap in the distribution. Tsevi Mazeh and
Dorit Goldberg of Tel Aviv University and David Latham of the
Harvard-Smithsonian Center for Astrophysics published their plot of frequency
against mass in Astrophysical Journal Letters last July (vol 501, p
199).
Because Mazeh’s team used data only from radial velocity studies, they had to
correct for the “line of sight” effect. They did so by assuming that the objects
had randomly oriented orbits and working out how that would shift their
histogram. Their final result reveals a gap between the planet-like objects on
one side and the star-like objects on the other side, as if two independent
curves had been added together. “The two distributions suggest two distinctive
populations,” they write, with the dividing line between the two groups in the
range of 10 to 30 MJ.
These two studies seem to do away with the idea of a continuum between
Jupiter-sized planets and hydrogen-burning stars. Instead, a gulf divides
them—which suggests that the formation mechanisms are distinct, too.
But not everyone is convinced. David Black of the Lunar and Planetary
Institute in Houston has performed a different statistical analysis, producing a
smooth curve rather than a histogram to describe the mass distribution. His
result implies that giant planets and brown dwarfs are part of a continuum, with
no sign of a gap. He is “very suspicious” of the Hipparcos data, as he thinks
that it is wildly improbable that so many of the brown dwarf candidates would
happen to be in orbits nearly perpendicular to our line of sight.
Other astronomers are more circumspect. Alan Boss, for example, says there’s
a “hint of a gap” in the range of 10 to 20 Jupiter masses. “But more data may
make that gap become stronger, or make it go away—who knows?” Boss says
that the apparent shortage of brown dwarfs could be an artefact of the way some
searches have been carried out.
Many of these searches have concentrated on finding companions to Sun-like
stars, but brown dwarfs could be more common around smaller stars. Indeed,
brown dwarf binaries—which are extremely difficult to detect—may be
more common still. Our existing theories aren’t good enough to predict how many
brown dwarfs should be out there. “There really is no strong expectation of what
we’re going to see, so we’ll just have to see what we find,” says Boss.
Those who yearn to separate the giant planets from the brown dwarfs, however,
are looking beyond mere mass. One clue comes from the shape of a body’s orbit. A
circular orbit is characteristic of a planet, as planets are thought to
originate in circular discs around their host stars
(see Diagram), and the
giant planets in our Solar System have nearly circular orbits. But if an orbit
is highly elliptical, it is more likely to be that of a brown dwarf, formed by a
separate collapsing cloud.FIG-mg21855201.JPG
Most of the extrasolar “planets” discovered so far have highly eccentric
orbits, which makes some astronomers wonder whether they should be reclassified
as brown dwarfs. Such a move, however, may be premature. According to Basri,
planets could develop highly elliptical orbits after they form because of their
interaction with other planets or with the remnants of the disc.
If orbits can’t be used to tell the planets from the dwarfs, what can?
Planets are thought to form as part of complicated systems, whereas brown dwarfs
formed by cloud collapse should usually be found alone or in simple binaries.
Astronomers don’t know of a mechanism that would produce several brown dwarfs
orbiting a larger star. “If you find multiple companions in the system, you’re
probably looking at a planetary system,” says Black. “But if you find only a
basic binary, that suggests that it’s a star-like system, and therefore probably
a brown dwarf.”
True giant
Just last month, a team led by Paul Butler of the Anglo-Australian
Observatory announced that the nearby star Upsilon Andromedae is orbited by
three giant planets. One of them has a mass of at least 4 MJ—a
truly giant planet. But Black is sceptical about the existence of one of these
bodies, and says the system may look “just like a multiple stellar system”. In
other words, the companions could still be brown dwarfs.
In the future, new instruments may allow astronomers to take high-resolution
spectra of all these bodies and compare the chemical compositions of their
atmospheres. A planet should have a higher proportion of heavy elements than its
host star (Saturn, for example, is much richer in carbon, nitrogen, and oxygen
than the Sun). In contrast, a brown dwarf and its host star would have been
formed by similar processes, and should therefore have nearly identical
compositions. But, says Black, “that’s a very, very tough observation to
”.
One planned spacecraft that would be up to the job is the Terrestrial Planet
Finder. Scheduled for launch in 2011, it will combine the light from four big
telescopes—which will be separated by as much as a kilometre—to
create ultra-sharp images of planetary systems. Scientists should be able to
pick out individual planets (even small Earth-like ones) or brown dwarfs, and
take high-quality spectra of their atmospheres.
A true understanding of brown dwarfs would be a breakthrough. For one thing,
brown dwarfs must be part of the “dark matter” whose gravitational effects on
our Galaxy have been observed. But searches for brown dwarfs have already shown
that they cannot form all the dark matter in our galaxy. These latest hints of a
mass “gap” below 30 or 50 MJ would imply that their contribution is
smaller still.
For another thing, understanding how planets form will help us find out
whether the Earth is an aberration, or whether the Universe is teeming with
small Earth-like planets that could be the home of life. If the gap between
stars and planets is really there, we can be more confident that all those
low-mass extrasolar companions mark well-populated planetary systems like our
own.
-
Further reading:
The Extrasolar Planets Encyclopaedia is at www.obspm.fr/planets