After nearly 30 years of intense debate and scientific inquiry, scientists finally know the answer to the question "Where are the Gamma-Ray Bursts?", and can now move on to answer the question "What causes them?". Recent discoveries in this field by a collection of international astronomers have demonstrated that these bursts are from the most remote parts of the universe, releasing perhaps as much energy in 10 seconds as the Sun emits in its entire 10-billion-year lifetime. This week at the 190th meeting of the American Astronomical Society, astronomers working in the Space Sciences Laboratory of the NASA/Marshall Space Flight Center will participate in a press conference to discuss the flurry of recent activity--both in space and on the ground--that has led us to the resolution of a 30 year mystery in astrophysics.
Gamma-ray bursts (GRBs for short) are brief flashes of high-energy radiation that appear on average about once a day at an unpredictable time from unpredictable directions in the sky. Since their discovery (by accident) in the late 1960's, several thousand bursts have been detected, most of them with BATSE, the Burst and Transient Source Experiment, on board the Compton Gamma Ray Observatory. Their distribution on the sky is completely uniform. In particular, they do not appear to come from the Milky Way. So where do they come from? This is the question that had kept astronomers busy for several decades, with no apparent resolution in sight.
Between the time of the discovery of bursts in the late 1960's, and the launch of the BATSE experiment in 1991, most astronomers were convinced that bursts originated in our own Galaxy, on or near objects called neutron stars. Our Milky Way Galaxy contains many neutron stars, objects as massive as the Sun (about 300,000 times the mass of the Earth) but no bigger than about 10 kilometers in diameter. Their tremendous gravitational field and magnetic field made them an ideal source for the gamma-ray bursts.
The BATSE experiment on GRO was built with every intention of confirming this theory. It was believed that with BATSE's increased sensitivity, we would be able to see the faint gamma-ray bursts map out the Milky Way on the sky, the same way the stars appear to do so when you get away from the city lights and can see fainter objects.
This concept is demonstrated in the accompanying figure. At the top, we see a schematic of our Galaxy's disk, with the Earth and BATSE located at the star. If we were not able to see very far out into space, symbolized by the first circle, we have the situation shown in the first set of plots. We see bursts in every direction equally. At the same time, we see there is no detectable edge and the brightness distribution of bursts follows a special curve indicating homogeneity. This is because our small inner circle is, to first order, filled uniformly with bursts.
On the other hand, if we can see out of the disk of the Galaxy, such as in the case of the outer circle, the bursts do not appear uniform on the sky, and the distribution shows a deficit of weak bursts, because there are not very many bursts located near the edge of the second circle.
This appealing idea was confronted with reality in 1992, when it became clear from BATSE observations that GRBs are distributed uniformly on the sky, without a concentration to the plane of the Milky Way, or towards its center or in other clumps of or concentrations. In fact, the distribution is almost perfectly random.
By itself, the angular distribution could lead one to argue that perhaps we see the bursts coming from quite nearby, from distances small compared to the thickness of the disk of the Milky Way. However, this would imply, as we saw above, that in no direction do we start to see the edge of the GRB distribution, or in this case, the edge of the Milky Way disk. Such an edge would reveal itself in the distribution of brightnesses of bursts, which would show a deficit of faint bursts.
But this is contradicted by the observed brightness distribution of GRBs, which shows a distinct dearth of very weak GRBs: it is as though in all directions we do see that edge.
The combined angular and brightness distributions of bursts eliminates the possibility that GRBs come from the disk of the Milky Way, and left astronomers with a choice between one of two possibilities:
- GRBs originate either in a very large spherical halo (or corona) around the Galaxy, or
- They come from the far depths of the Universe, billions of light years away from us.
The galactic halo would have to be very big, about a million light years across, which is much bigger than the diameter of the known Milky Way system (which is less than a hundred thousand light years), much less the halo that the Milky Way is known to have. The very large halo size is required to avoid an asymmetry in the GRB sky distribution caused by the fact that the Earth is offset from the center of the Galaxy, by about 25,000 light years.
In the case of the other distance scale (referred to by astronomers as "cosmological") it is the smoothness of the GRB sky distribution that tells us that their distances must exceed the length scales over which the distribution of matter appears clumpy, i.e., larger than clusters and superclusters of galaxies, and the voids between them.
This implies distances measured in billions of light years.
At such large distances the effect of the expansion of the Universe becomes noticeable. An important effect of this expansion is the redshift of light (and gamma rays) and the apparent slowing down of clocks, both of which would cause a change in the brightness distribution very similar to what BATSE actually observes. Scientists have looked at the bursts for these cosmological effects, and some have reported seeing time dilation (weak bursts, which are presumably farther away, are, on average, longer than the strong, nearby bursts) and others report energy shifting, where the distribution of gamma-rays in faint bursts appears shifted to lower energies compared to strong bursts. In general the BATSE data seemed to be hinting at a cosmological origin, but it couldn't be proved more directly.
The key to nailing the cosmological hypothesis seemed to hinge on the ability to find a counterpart to a burst in a region of the spectrum outside of the gamma-rays. These counterparts would have the positional accuracies to link the GRB, for instance, to a faint galaxy, or allow the measurement of a redshift from an optical spectrum. Conversely, these would not be seen if GRBs came from a big galactic halo.
Finding such counterparts has recently become possible with the detection of GRBs with the Wide Field Camera (WFC) on board of the Italian-Dutch X-ray satellite BeppoSAX, which can pinpoint a GRB on the sky to within a circle with a diameter of 6 arcminutes (20 percent of the diameter of the full moon), and not only that, but can provide that position within a matter of hours, much faster than was possible before. This error box, if known sufficiently rapidly, is small enough that optical and radio telescopes can be used to search it for the presence of transient emission that may be connected with the GRB.
This WFC is sensitive to X rays, and can view a 40 by 40 degrees field of view, within which it can locate X-ray sources with several arcminute accuracy.
To know whether or not a GRB occurred in the field of view a Gamma-Ray Burst Monitor (GRBM) on BeppoSAX is employed, whose data are searched for GRBs every satellite orbit of 1.5 hours. When a GRB is found the WFC data are searched, and an accurate position can be determined, which is then quickly dispatched to astronomers around the world.
The first time this worked out successfully was on February 28 this year, when a GRB was detected coming from the constellation Orion. Within 21 hours after this burst a team of astronomers from the University of Alabama in Huntsville and the University of Amsterdam used the 4.2 meter William Herschel Telescope on La Palma (one of the Canary Islands in the Atlantic Ocean, West of the North Africa) to make images of the BeppoSAX GRB location. They did the same about a week later. A comparison of these images immediately revealed one star that was present on February 28, which had disappeared a week later.
In the meantime the Beppo SAX astronomers had used a more precise X-ray instrument on their satellite to also observe the GRB location in detail, and they found that it contained a weak decaying X-ray source, with a position (accurate to somewhat less than an arcminute) that coincided with that of the optical star that disappeared. Clearly, the X-ray and optical afterglow of a GRB had been detected for the first time. Further conformation that the decaying X-ray and optical signals came from one and the same source was obtained from X-ray images taken with the ROSAT satellite, which located the X-ray afterglow with an accuracy of some 10 arcseconds: it still coincided with the optical afterglow.
As soon as the disappearing optical afterglow was discovered, very deep optical images were made of it with the ESO New Technology Telescope and the Keck Telescope, which showed that at the location of the optical transient there was a very weak object that looked slightly extended, quite possibly a weak galaxy.
In a matter of weeks two observations were made with the Hubble Space Telescope, which showed that the fading optical transient consists of a point source (its fading between two HST observations shows it is the GRB counterpart) and an extended source ('fuzz') at whose edge it appears to be located. Although they don't constitute solid proof, these observations naturally fit the idea that the GRB of February 28, 1997 went off in a faraway faint galaxy (the fuzz).
The next major step forward occurred with a burst that was observed with the WFC on May 8. A faint optical variable star was detected at the GRB location by H. Bond of the STScI, using a 36 inch telescope at Kitt Peak National Observatory. A spectrum of this star taken by M. Metzger of Caltech and his collaborators with the Keck telescope showed that the presence of absorption lines, which are redshifted by 0.83 (i.e., their wavelengths are 1.83 times their value measured in the laboratory). The case for the long distance scale appears now to be settled.
Finally, for this GRB of May 8, D. Frail of the National Radio Astronomical Observatory discovered the first radio emission associated with any GRB.