Three decades after the discovery of gamma-ray bursts, their causes remain elusive and still offer potential pitfalls for investigators.
"We really need to be careful ... and make sure that we're interpreting the meager data carefully," said Dr. Gerald Fishman of NASA's Marshall Space Flight Center. Speaking at the Fifth biennial Huntsville Gamma Ray Burst Symposium, Fishman reviewed a number of the questions that astrophysicists still have to resolve in what has moved from an oddity to a hot topic.
Gamma-ray bursts were discovered by accident in the late 1960s when the United States launched satellites intended to monitor compliance with the nuclear test ban treaty. They discovered bursts of gamma radiation that came from outside our solar system. Initially these were thought to be associated with our galaxy. But they came and went so quickly that locating bursts was impossible.
With the launch in April 1991 of the Compton Gamma Ray Observatory, scientists hoped to identify the burst sources. Instead, Fishman's Burst and Transient Source Experiment (BATSE), one of the observatory's four instruments, soon showed that the bursts were evenly scattered across the heavens. Combined with the fact that there were fewer weak bursts than anticipated, the discovery implied bursts originated from near the edge of the observable universe.
The suspicion grew that the bursts were deep in the older universe. This was confirmed in early 1997 when telescopes on another satellite, Beppo SAX, happened to quickly and precisely locate a burst and guide ground-based observatories to discover the source buried in a distant galaxy.
(A gentle reminder of that earned a laugh when Fishman put up a plot of the 2512 bursts BATSE has recorded. "For historical reasons we plot this in galactic coordinates," he said, since even he expected them to have galactic origins. "Perhaps we should change it to celestial coordinates.")
Yet even with more than 2,500 bursts recorded by BATSE, and more than a dozen apparent burst sources captured by optical and other telescopes, mystery still shrouds the cause and caution must be used in interpreting the data.
For example, Fishman said, astrophysicists sometimes refer to statistically significant spectral lines appearing in the data.
"We have yet to see a gamma-ray feature that shows up consistently among detectors," Fishman said. "That says to us that we don't understand the systematic errors in our own system."
BATSE actually comprises eight detector modules, each pointing out from the face of an imaginary octahedron or double pyramid. At least two must detect an event before the system is "triggered" to announce a burst. Most events are seen by at least three and sometimes four of the detectors. While the modules are nearly identical, individual variations cause slight differences in how the bursts are measured.
A candidate for the cause of gamma-ray bursts is exceptionally violent supernovae or hypernovae. They would have similar energies, and both could involve energy beamed like water from a fire hose. "We see tantalizing results, but none are definitive," Fishman cautioned.
Another reported finding is that short-duration gamma-ray bursts are anisotropic, meaning they are more numerous in one area. (A good example of anisotropy is our night sky: most stars are seen in the Milky Way, our view across our galaxy.) Not so, Fishman said. Short bursts are isotropic - evenly scattered across the sky.
Meanwhile, scientists have their hands full interpreting the wide variety of burst shapes, the intensities that rise and fall with time. Some are very strong and brief and have no fine-time variations, like a photographer's flash going off, while others have a spiky structure. There are FREDs - fast rise and exponential decay - that appear like a struck match flaring and then fading away. There are odd bursts like one that have a precursor followed 100 seconds later by the main event.
"It's trying to tell us something," Fishman said. "We're just not smart enough to understand it yet."
Some of the uncertainty comes from how BATSE modules detect and measure bursts, admitted Dr. Charles Meegan, also of NASA/Marshall. These "self-inflicted biases" can make the work more challenging. For example, "slow risers" can be missed if the burst rises slowly enough that the BATSE electronics think it's a variation in the background noise.
Or the apparent peak flux can vary. Whether a burst produces 100 counts in 100 milliseconds or 1 second, the electronics will trigger since they are keyed to 100 counts in 1,024 milliseconds (1.024 sec). Meegan said that the BATSE team is working to measure the biases.
Nevertheless, BATSE and other instruments such as the Dutch-Italian Beppo-SAX and Japanese ASCA satellites have produced a wealth of information allowing scientists to try peeling back some of the layers of mystery surrounding gamma-ray bursts. These include:
Dr. Jon Hakkila of Minnesota State University has used pattern recognition programs from an artificial intelligence project to analyze three classes of gamma-ray bursts and determined that there are really two classes. Class 1 bursts are long, have high fluence or total counts, and intermediate hardness in the spectra. Class 2 bursts are short and have low fluence and hard spectra. What appeared to be Class 3 bursts are really Class 1 bursts that mimic Class 3 due to measurement errors or a newly identified bias in BATSE.
The "Holy Grail" of gamma-ray bursts may be in sight, according to Dr. Jay Norris of NASA's Goddard Space Flight Center. He said that it appears that time variations in a burst come directly from the central engine itself. Previously it was thought that the variations were caused by the expanding blast wave of material slamming into clumps of materials in space. Already scientists know that longer bursts have lower peaks. If the burst were from the central engine - whatever it might be - then measuring the distance to a burst would become easier. It would be akin to measuring the scale of the universe by the apparent brightness of variable Cepheid stars. Their brightness and periods are closely linked, so measuring the period and brightness of one allows a quick computation of the distance. A similar method could be applied to gamma-ray bursts - once enough bursts have been connected with visible sources and host galaxies to establish a good unit of measurement.
220,000 electron volts is a magic number that characterizes the energy of gamma-ray bursts, but no one yet knows why, said Dr. Rob Preece of the University of Alabama in Huntsville. "It's a mystery," Preece said. "It falls out around 200 keV. But why it should be that particular value, and why you don't very often have bursts that are 10 times that or one-tenth is a mystery." 220 keV is the intersection of two lines drawn in a power law. It's basically two straight-line graphs, except the axes are logarithmic (the same distance on the line means a tenfold increase in value; this compresses immense events to a manageable graph)
"Power laws are eagerly sought in astrophysics because they tend to say something is happening in a particular way," Preece explained. Most bursts observed by BATSE plot as two power-law lines, one rising, and one falling. The great majority intersects around 220 keV. A speaker will offer possible meanings for that number later in the week.