A new analysis of nearly 2,000 gamma-ray bursts -- the mysterious creators of black holes and the most powerful explosions known in the universe -- has revealed that the two major varieties, long and short bursts, appear to arise from different types of events.

In an analysis of nearly 2,000 bursts, a team of researchers from Europe and Penn State University uncovered new discrepancies in the light patterns in bursts lasting less the two seconds and in bursts lasting longer than two seconds.

"We can now say with a high degree of statistical certainty that the two show a different physical behavior," said Lajos Balazs of Konkoly Observatory in Budapest, lead author on a paper appearing in an upcoming issue of the journal Astronomy & Astrophysics.

The analysis supports the growing consensus that long bursts originate from fantastic explosions of stars over 30 times more massive than our Sun. Short bursts have been variously hypothesized to be fiery mergers of neutron stars, black holes, or both, or perhaps a physically different type of behavior in massive collapses.

"It is suspected that, either way, with each gamma-ray burst we wind up with a brand new black hole," said Peter Meszaros, professor and head of the Penn State Department of Astronomy and Astrophysics. "The puzzle is in trying to identify clues that would help to elucidate whether these two types consist of essentially the same objects with different behaviors, or different objects with somewhat similar behavior."

Gamma-ray bursts are like a 10^45 watt bulb, over a million trillion times as bright as the Sun. Although common -- detectable at a rate of about one per day -- the bursts are fast-fading and random, never occurring in the same place twice. Scientists have been hard pressed to study the bursts in detail, for they last only a few milliseconds to about 100 seconds, with most around 10 seconds long. Most scientists agree that the majority of bursts originate in the distant reaches of the universe, billions of light years away.

Previous results have shown that the short bursts have "harder" spectra, which means that they contain relatively more higher-energy gamma-ray photons than the longer bursts do. Also, in short bursts, the photons hitting a burst detector are closely spaced, or bunched, compared to the longer bursts, suggesting that the source is physically different, as well.

This type of information is valuable because it appears to contain clues about the intrinsic physical mechanism by which the sources produce the gamma rays, but these sources have still not been characterized in enough detail to understand them. Balazs and his colleagues sought to establish what, if any, correlation exists between different pairs of properties, when one considers separately the long and the short bursts.

The team examined the fluence and duration of 1,972 bursts and found a new relationship. The fluence is the total energy of all the photons emitted by the burst during its gamma-ray active stage, a measurement incorporating both the flow and energy of individual photons.

Within both categories, long and short, there is a correlation between fluence and duration: the longer the burst, the greater the fluence. Yet the degree of this relationship is statistically different for the two categories (at a 4.5 sigma significance level). This difference places constraints on what can cause these bursts or how they can operate.

In long bursts, there is a direct proportionality between duration and fluence, suggesting that the energy conversion rate into gamma rays is, on average, more or less constant in time. For the short bursts, there is a weaker dependence, which could, for instance, be due to an energy conversion rate into gamma rays that drops in time, resulting in a less efficient gamma-ray engine.

It seems unlikely that the same engine could produce both types of bursts, the team said. Although not directly addressed in the paper, these results support the notion that if the long bursts originate from massive stellar explosions, then short bursts originate from something entirely different. In the latter scenario, this event could be either mergers or such a drastic Jekyll-and-Hyde-like switch in the stellar explosion mode that the engine appears physically quite different. Such drastic and well-defined differences in the correlation between two of the major variables will need to be addressed quantitatively in future models of the burst physics.

The 1,972 bursts were observed by the BATSE instrument on the NASA Compton Gamma Ray Observatory, a mission active between 1991 and 2000. Coauthors also include Zsolt Bagoly, of the Laboratory for Information Technology at Eotvos University in Budapest; Istvan Horvath, of the Department of Physics at Bolyai Military University in Budapest; and Attila Meszaros, of the Astronomical Institute at Charles University in Prague.

This research was supported by the U. S. National Aeronautics and Space Administration (NASA) and the Hungarian national research foundation (OTKA).