v1.0 01 feb 2000 Greg Goebel firstname.lastname@example.org public domain* One of the longest-running mysteries in astronomy has been that of the "gamma-ray bursters" (GRBs): unexplained flashes of gamma rays that last from seconds to hours, occurring at random positions in the sky about once each day. While astronomers have worked very hard to pin down the nature of GRBs, they remain an elusive mystery that gives up clues very reluctantly.
This document provides a history of the GRB search and its current status.
* The gamma ray bursts were discovered in the late 1960s by the US Vela nuclear test detection satellites. The Velas were launched to detect high energy radiation emitted by weapons tests, but they also incidentally sensed occasional bursts of gamma rays from deep space. While the sensors on the Vela satellites had low angular resolution, researchers at the US Los Alamos National Laboratory in New Mexico were able to determine in 1973 that the bursts came from deep space.
Astronomers believed that once better gamma ray detectors were put in orbit, they would be able to quickly pin down the locations of the GRBs. After all, that is what happened with X-ray sources. However, when such improved detectors were sent into space in the 1970s, optical searches of the regions where the bursts originated showed nothing of apparent significance.
Further information on the burst sources proved hard to obtain, and led to more questions than answers. The first question posed by the GRBs was: are they local to our own Galaxy, or do they occur in the distant reaches of the Universe? The second question was: what mechanism causes the bursts? If they do occur in the distant Universe, the mechanism must be producing enormous amounts of energy.
In April 1991, the US National Aeronautics & Space Administration (NASA) launched the Compton Gamma Ray Observatory on board the space shuttle. One of the instruments on board Compton was the Burst & Transient Source Experiment (BATSE), which could detect gamma ray bursts and locate their positions in the sky.
Within a year, BATSE had determined that GRBs occurred about once a day, and were randomly distributed over the entire sky. If they were events occurring in our own Galaxy, they would have been preferentially distributed in the plane of the Milky Way. Even if they were associated with the galactic halo, they would still be preferentially distributed towards the galactic center, 30,000 light years away, unless the halo were truly enormous. Besides, if that were the case, nearby galaxies would be expected to have a similar halo, but they did not show up as "hot spots" of faint gamma-ray bursts.
To many astronomers, this implied that the GRBs originated in the distant Universe, but that led to the problem of finding a mechanism that could generate that much energy. Theorists considered events such as the collision of two neutron stars, the collapse of a neutron star into a black hole, or a new class of supernova explosions.
Other theorists were also still able to come up with "local" models for the GRBs, and BATSE couldn't resolve the issue. While the expansion of the Universe causes a "redshift", a reduction in frequency due to the Doppler effect that increases with distance, and would reduce the energy of the gamma ray bursts, BATSE did not have the ability to provide details of the burst spectra.
* By the late 1990s, the local hypothesis for GRBs had been ruled out. The first clue came from the Italian-Dutch Beppo-SAX satellite, which was launched in 1996.
Beppo-SAX carries a gamma ray detector that works in conjunction with a pair of wide-field X-ray cameras. While the satellite's gamma ray detector has poor resolution, a gamma-ray burst will generally have an X-ray component, which should allow the X-ray cameras to quickly pinpoint the source for observation by optical and other telescopes.
On 28 February 1997, Beppo-SAX managed to pin down the location of an optical counterpart to a gamma-ray burst, which was designated GRB 970228 in accordance with the date of the event. Some observations seemed to show the object was moving rapidly across the sky. That meant it couldn't be too far away, implying the bursters are a local phenomenon.
Then, on 8 May 1997, Beppo-SAX recorded another burst in the constellation Cameleopardis, and the spacecraft's science team sent out an alert over the Internet. Seven hours later, an optical source was detected by astronomer Howard Bond, using a 90-centimeter telescope at Kitt Peak National Observatory in the US state of Arizona.
On 11 May, astronomers used the 10-meter Keck II telescope on the island of Mauna Kea, Hawaii, to obtain a spectrum of the object. The spectrum showed "absorption lines", or frequencies where the light was absorbed by gases between the object and Earth.
The patterns of absorption lines appropriate to different atoms and molecules are very specific, and their displacement easily reveals the redshift of the object and its distance. The spectrum showed a redshift of 0.835, indicating the object was billions of light-years away.
This was baffling. One observation indicated a local origin, the other a distant origin. The only conclusion, aside from the implausible one that there are two different burster mechanisms, was that one of the observations was bogus. At least one of the correlated optical sources may have had nothing to do with a gamma ray burst, and simply happened to be in the right place at the right time. Some astronomers were also unable to detect any proper motion in the object linked to GRB 970228.
Following these observations, astronomers were able to track down more faint visible-light and radio "afterglows" of GRBs, hours or days after the occurrence of the burst. A few more redshifts were obtained, and confirmed that the bursts occurred in the distant cosmos. The high proper motion reported for GRB 970228 was clearly erroneous, and in fact observations made by the Hubble Space Telescope in September 1997 showed no proper motion in the faint afterglow that remained from GRB 970228.
Visible light observations of several of these GRB locations in 1997 and 1998 identified possible links between the bursts and supernovas. The observations were not conclusive, but they were encouraging to the proponents of the supernova theory, and gave astronomers hunting visible components of GRBs something to investigate in more detail.
* Astronomers finally managed to obtain a visible-light image of a GRB as it occurred on 23 January 1999, using an ingenious contraption named ROTSE-1 (Robotic Optical Transient Search Experiment 1), sited in Los Alamos, New Mexico. ROTSE-1 consists of an array of four commercial 200-millimeter telephoto lenses focused on CCD electronic imaging arrays, and mounted on an automated platform.
While the four telephoto lenses are modest instruments even by the standards of amateur astronomy, ROTSE-1 has a wide field of view and can be quickly repositioned to scan any part of the visible sky. ROTSE-1 is operated by a team under Dr. Carl Akerlof of the University of Michigan.
In the dark hours of the morning of 23 January, the Compton satellite recorded a gamma ray burst that lasted for about a minute and a half. There was a peak of gamma and X-ray emission 25 seconds after the event was first detected, followed by a somewhat smaller peak 40 seconds after the beginning of the event. The emission then fizzled out in a series of small peaks over the next 50 seconds, and eight minutes after the event had faded to a hundredth of its maximum brightness. The burst was so strong that it ranked in the top 2% of all bursts detected.
Compton reported the burst to its ground control facility at NASA Goddard Space Flight Center in Maryland the moment it began, and Goddard immediately sent the data out over the Gamma Ray Burst Coordinates Network (GCN). While Compton, as mentioned, cannot provide precise locations of bursts, the location was good enough for the wide-field ROTSE-1. The camera array automatically focused on the region of the sky and obtained an image of the burst 22 seconds after it was detected by Compton, with subsequent images obtained every 25 seconds after that.
ROTSE-1 can image cosmic objects as faint as magnitude 16, and GRB hunters had expected the visible component of a GRB to be very faint. Instead, the visible component reached magnitude 9. It was so bright that it could have been seen by an amateur astronomer with a good pair of binoculars. The object that produced it increased in brightness by a factor of 4,000 in less than a minute.
The news of ROTSE-1's accomplishment didn't make it out on the networks until later in the day, and in the meantime other observatories were focusing on the event, by then designated GRB 990123.
The Beppo-SAX satellite had also seen the burst, and pinned down it's location to within a few arc-minutes. This data was sent out, and four hours after the burst the area was imaged with the 1.52 meter (60 inch) Schmitt camera at Palomar Mountain in California. The image revealed a magnitude 18 optical transient that wasn't on archive images of the same area.
The next night, the fading object, by now down to magnitude 20, was imaged by the Keck telescope, and the 2.6 meter Nordic Optical Telescope in the Canary Islands. The observations revealed absorption lines that indicated a redshift of 1.6, implying a distance of 9 billion light years.
The combination of obvious brightness and implied distance was startling even to GRB hunters. At 9 billion light years, the gamma-ray energy released by the burster was equivalent of converting the entire mass of a star 1.3 times the mass of our Sun completely into gamma radiation. At visual wavelengths, if the burster had gone off in our own Galaxy 2,000 light years away, it would have shined twice as bright in our night sky as the Sun does during the day.
Some astrophysicists believed that such high-energy output was unlikely. One way around the problem was to assume that the burst energy was only sent out in specific directions from the event, rather than in all directions, much as some stars and galaxies emit directional high-energy "cosmic jets" from violent events.
Another explanation for the the great brightness of the burst was that its light had been focused by a "gravitational lens", caused by the distortion of space by a large galaxy between Earth and the GRB. This speculation was fueled by observations that seemed to indicate there was in fact a galaxy between the Earth and the GRB, but the "galaxy" turned out to be a photographic flaw.
This didn't rule out gravitational lensing, and some astronomers speculated that the same event might also be detected by lensing from other galaxies slightly off the line of sight, leading to the observation of other, fainter bursts in the following days or months.
This was a somewhat wild suggestion, and the speculation died down when Bradley E. Schafer of Yale pointed out that at a redshift of 1.6, the density of galaxies made the probability that lensing had occurred only about one in a thousand. Schafer also sensibly pointed out that saying such energy output was impossible was somewhat contrived, since nobody really knew was a GRB really was. In any case, if limits needed to be imposed, the "beaming" hypothesis was much more plausible than lensing.
* The Hubble Space Telescope performed observations on the location of GRB 990123, sixteen days after the event. It had faded by more than a factor of three million in that time. The Hubble was able to pick up the traces of a faint galaxy, whose blue color suggested it was forming new stars at a rapid rate.
This observation tended to discourage theorists who believed that the bursts were due to collisions between neutron stars or a neutron star and a black hole, since that implied a fairly high density of dead stars and that was inconsistent with a young galaxy.
Supernovas, on the other hand, occur frequently in star-forming galaxies, since the big stars that die in supernovas have short lifetimes, and the theorists who preferred the supernova model were correspondingly encouraged. The biggest problem with the supernova model is that current theory can't explain how a supernova can cause a gamma-ray burst, but the theorists are working on that.
* While the mystery of gamma-ray bursters still remains unresolved, astronomers feel they are closing in on a solution, and remain very excited.
One of the latest revelations has been that an analysis of the gamma-ray emission of the half-dozen GRBs whose distances have been determined from redshifts suggests that the amount of flickering is related to the burst's brightness, that is, bright bursters flicker more than dim ones. This implies that the amount of flickering could be used as a "yardstick" to determine the distance of a burst, even if its redshift cannot be determined.
This relationship will have to be supported by further burst observations before it is widely accepted. As the number of detailed observations of GRBs and the instruments and theoretical tools are refined, it seems only a matter of time before the nature of the bursters is understood.
* Sources include:
Greg Goebel (email@example.com)
Return to Contents List Page