Gamma-ray bursters are brief, very bright flashes of gamma rays. They were discovered by accident in 1973 by the Vela series of satellites, which were launched in order to check that countries were not testing nuclear weapons in space, the atmosphere on land or at sea.
Since the objects could only be detected in the gamma-ray band, and no match could be made to any other type of object, the nature of these bursts was very difficult to investigate further.
The nature of these objects remained a great mystery until the 1990s, when they were finally seen also in the X-ray region, followed shortly by identification in the optical and radio. They are now known to at great distances and amongst the most luminous energy releases by any objects in the universe. Theories on just exactly they produce all this energy are still quite fresh and being developed. The main ideas center on the merging of two compact objects, such as neutron stars or black holes, or some kind of new, very energetic supernova explosion.
Before the launch of the Gamma-Ray Observtory, the small number of known gamma ray bursts were distributed isotropically, or completely at random, around the sky. This suggests that they are either very close to us, in the Galaxy (few parsecs to few 10s of parsecs), or they are at large distances (100 to 200 kpc) around the Galaxy, or that they are at very large distances (most of the way across the visible universe, or Gpc) indeed.
GRO strengthened the evidence for the cosmological distance interpretation, since it started detecting a few bursts each day, and the distribution built up to a very isotropic picture after a years. The distribution was so isotropic that a burst origin in the Galactic disk seemed out of the question, since all know Galactic sources are concentrated either into the Galactic disk or toward the Galactic center. An origin in some sort of halo around the Galaxy became harder to imagine since no extra bursts were seen toward the nearby galaxy M31 (Andromeda).
Although many new bursts were found with GRO, the positional accuracy on the sky (exactly where they were coming from) was quite low, making it very hard to pair the bursters up with some other type of object. No Gamma-ray burst was seen to ever come from the same position on the sky however.
The number of faint bursters compared to bright ones gave the best evidence that the bursters were at cosmological distances. If the bursters have some kind of similiar luminosity, then there is a simple relation between the number of bursters seen above some brightness. If they are uniformly distributed in space, the number of bursters should increase with the distance, d as d3, whereas the brightness decreases like the inverse square law with distance, 1/d2. Thus, one would expect the bursters to have a cumulative distribution of brightness B like B-3/2, if we were looking at a uniform distribution of them in space. The actual distribution is a little flatter than this, indicating that we can see beyond the ``edge''of their spatial extent around us.
The bursts typically last between milliseconds and circa 103 seconds. They have a great range of burst profiles, and like snowflakes, one could well say that everyone is different. They show a lot of short term variation however, on the order of milliseconds. This indicates that whatever the root cause of bursters is, they must be very compact, since the light travel time in a few milliseconds is of order a few 100's of kilometers. This suggests neutron star size objects (or low mass black holes) at the core of the burst.
The faint bursters last longer than the bright bursters. This would be expected if the bursters are at cosmological distances, and their light curves are stretched by time dilation effects. This also lowers the count rate from the burst. For sources at a redshift z, the count rate is lowered by a factor 1+z.
Typical bursts results in a few 103 to a few 105 gamma rays per second on the BATSE experiment on the GRO. The typical flux is 10-13 J cm-2 sec-1.
|
|
In the mid-1990s the Italian/Dutch X-ray satellite BEPPO-Sax was launched. The satellite found X-ray bursters, and had sufficient positional accuracy (several arcminutes, rather than degrees) for ground based observers to try and get an optical image of the bursters. This was acheived in 1997, when gamma-ray bursters were linked up with sources in the X-ray, optical, mm and radio wavelengths. The distance problem was rapidly solved, when the optical burster faded away and revealed that the burst has taken place in a faint galaxy, for which a redshift could be measured.
Radio light curves of the bursters can typically be followed for a month after the burst, after which they become too faint. They show variability on times scales of less than a day. The X-ray and optical ``afterglows'' drop in intensity like a power law. The typical power-law exponent is of order 1 to 2. These afterglows provide good evidence for one model of the gamma-ray bursters as highly relativistic blast waves moving out from whatever the central source is and into the surrounding gas.
There is no shortage of ideas for what's doing the work in the core, such as neutron-neutron star mergers, neutron-black hole mergers, white dwarf collapse (!), and core collapse of very massive stars (a kind of failed supernova, called a hypernova). All of these models can somehow be adjusted to provide the central energies, and for all of them it is as yet hard to distinguish which if any is correct on the basis of observations.
One possibility to tell them apart might be that binary mergers could take place quite a long time after the two stars are born, in which case the binary may have moved a long way from star formation regions in the galaxy. On the other hand, if hypernovae ideas are right, the bursts would be expected to be more closely associated with star formation. Present technology and enough bursts for good statistics might be able to sort this issue out in the near future. Some bursts are know to be offset from their host galaxies already, but not enough statistics are in yet. Both ideas might be right, because the time scales of bursts seem to form into two groups, short and long (with a break at about 2 seconds).
A Gamma-ray burster was found to be associated with a supernova in 1998, and following the light curve of the supernova backward seems to indicate that the burst and supernova started within a day or two of each other. This burster was quite faint, and since there are more faint than bright bursters it may turn out that most bursters are associated with supernovae-like explosions.
About 40% of bursters can be detected in the afterglow phase. Why most cannot be detected is still a mystery. It might be because of local effects in the host galaxy (like dust) or because of real physical effects (they decay so fast we cannot catch them with follow-up telescopes) or because they are simply faint in the optical bands. Recently, 8 meter class telescopes have been turned to this task so we can probably expect clarifying results in the near future.
If bursters are associated with young stars or star formation regions, then they might be able to tell us a great deal about the history of star formation in the Universe. Remember that coming out of the big bang, most of the Hydrogen and Helium in the Universe was created along with a tiny trace of other light elements, but that most heavy elements have benn created subsequently inside stars. If bursters trace the local conditions and amount of star formation, then we have a way to see what stars were doing (how many has formed, and how rapidly the gas was being used) in the early universe.
For about 10 bursters an optical identification has been made with a host galaxy, for which the redshift, and hence distance, can be measured. Knowing the distance means we can compute the true brightness (at least for an isotropically distributed radiation) of the burst. If we could find some relationship between the observed properties, such as variability, flux, duration, spectral index etc and the luminosity, then we would be able to measure the distance of all the bursts. This would allow us to make a 3-D map of their distribution around us in space and time. Presently (Feb 2001) there are some indications that such a relation may exist, although time will tell if these claims turn out to be true.