Saturday, January 13, 2007

Gamma-ray burst surprises in 2006

Gamma-ray bursts (GRBs) have been puzzling astrophysicists for a long time. A little over a year ago, the picture seemed to be getting clearer. But during the past year, it seems to have become rather more complicated. And the year saved the best surprise for last. This spurt of activity is largely due to a flood of data, obtained with the help of the Swift satellite, launched in November 2004 specifically to study GRBs. (See also here.)

GRBs first attracted attention because they were unlike any other highly energetic phenomenon known in the universe. A GRB includes a flux of gamma-rays, which are, by definition, the highest energy photons. But unlike other gamma-ray sources, such as occur when matter falls into a black hole, the gamma-ray flux persists only a short time – only a few minutes at most. Gamma-ray emissions associated with black holes persist indefinitely.

At first, it was not even known whether GRBs normally came from something in our own galaxy or something perhaps much farther away. After it was determined around 1992 that GRBs were distributed evenly across the whole sky, it was almost certain that most GRBs were not associated with our galaxy, since in that case they would be mostly located in the direction of the visible Milky Way.

After many more examples had been observed, it was apparent that GRBs were of at least two different types: short duration events ("short GRBs") lasting less than 2 seconds, and longer events ("long GRBs"), lasting from 2 seconds to a few minutes. When it became possible to obtain spectra from the galaxies associated with GRBs, and hence possible to determine approximate distances by means of the redshift, astrophysicists were surprised to find that long GRBs were usually extremely distant – more than 8 billion light-years. This implied that whatever caused the GRB had to be almost incredibly energetic, so that they were visible across so many billion light-years and their photons remained in the gamma-ray energy range, despite being red-shifted by a factor of 2 or more. (That is, the photon wavelength as observed was more than twice its length when it was emitted.) Short-duration GRBs, on the other hand, were found to be somewhat closer, and therefore their source must be 10 or 100 times less energetic.

It was easier to understand long GRBs, since almost the only explanation that could work was for the GRB to be produced in a supernova explosion in which most of the energy was concentrated into jets parallel to the axis around which the supernova's progenitor star rotated. It occurs because material ejected from the dying star just before its collapse forms an accretion disk around the newborn black hole. This material is subsequently sucked back into the black hole and produces particle jets and the blast of gamma-rays. The emitted photons have such high energy since most of the supernova energy is concentrated so narrowly. Eventually, accumulating observations confirmed that the characteristics of the light output in a long GRB was what would be expected in a very energetic supernova event. Such a supernova must result from the death of a very massive stars (more than 40 solar masses), and is sometimes called a "hypernova". This model is sometimes known as the "collapsar" model.

Short GRBs were harder to figure out, but eventually it appeared that they could be explained as the result of a collision between two neutron stars. The amount of energy emitted and the duration of the event appeared to be just about right.

I wrote here about these conclusions a little over a year ago. But the ink was hardly dry (so to speak) on that post before new findings emerged that suggested some short GRBs weren't fully with the program.

Two papers were published in December 2005, suggesting that not all short GRBs had the same origin:

Breakthrough in puzzle of giant explosions in space
The Hertfordshire team’s new result adds a further, unexpected twist to the tale: a significant proportion of short bursts seem to originate from galaxies much more local to us than those previously observed. These nearby short bursts, could, like their more distant brethren, result from the catastrophic collision of neutron stars, though if so then their outbursts must be much weaker. Alternatively they could be a fundamentally different kind of explosion. A prime candidate could be an exotic object called a magnetar — a lone neutron star with a magnetic field a hundred thousand billion times that of the Earth - tearing itself apart due to enormous magnetic stresses.

A second paper published at the same time was more specific about what might cause short GRBs that are more nearby, and hence less energetic. It leaned towards an explanation involving the merger of a neutron star and a black hole, rather than a magnetar disintegrating from magnetic stresses:

Witnessing The Flash From A Black Hole's Cannibal Act
An international team of astronomers reports the discovery of a third short gamma-ray burst, associated with a nearby elliptical galaxy. The low level of star formation in such galaxies and the detection of a second long-lasting flare indicate that this gamma-ray burst is most likely the final scream of a neutron star as it is being devoured by a black hole.

This paper was based on measurements of a GRB observed on July 24, 2005 (hence designated GRB 050724), as well as an earlier one (GRB 050509B). The first of these was located in a galaxy "only" about 3 billion light-years away. This showed that short GRBs might result from the release of 100 to 1000 times less energy than a typical long GRB. GRB 050724 had a longer "afterglow" than would be expected from a merger of neutron stars (which would collapse almost instantly to a black hole). But the afterglow would be consistent with the merger of a neurton star and a black hole, where the process begins with the neutron star being rent asunder, followed by the pieces falling into the black hole over a longer period of time.

Other accounts of these results can be found here, here, here, and here.

To summarize, as of December 2005, the most common type of short GRB was figured to be the result of a merger between two neutron stars, while atypical short GRBs could be either magnetars or the merger of a neurton star with a black hole. But new examples kept showing up.

In February 2006 a computer study showed that about 1% of short GRBs due to neutron star mergers should occur in globular clusters, which are tightly packed with stars, and so the chances of encounter are high. More normally, neutron star mergers should occur between stars that a part of a single binary system. But in fact from 10 to 30% of observed short GRBs occur in globular clusters, far more than would be expected. It was hypothesized that in the latter case, energy output would be less tightly beamed, and hence more likely to be observed. More details are here.

Just a little later, on February 18, a very unusual GRB was observed as part of a supernova event. Named GRB 060218, it was much longer than typical long GRBs – 33 minutes in duration. It was also relatively quite close (440 million light-years) and so much less energetic (by a factor between 10 and 100) than typical long GRBs. Remember this one – its importance will be described later. Details: here, here, here, here, here, here, and here.

In March three papers in Nature announced that observations from a number of ground-based and space-based instruments had confirmed that GRB 050904, which was first seen in September 2005, was the most distant GRB ever seen. Its redshift was measured to be 6.3, making it about 12.8 billion light-years away, and occurring when the universe was only about 900 million years old. The earliest previous GRB to be observed was dated to about 1.4 billion years after the big bang. The characteristics of GRB 050904 were typical of long GRBs, so the results show that such an event was possible at that early date. Details: here, here, here, here.

Starting later in March, additional doubts were expressed that short GRBs had a single, simple explanation in terms of merging neutron stars. First, analysis of short GRBs occurring on July 9 and July 24, 2005 showed X-ray flares minutes after the initial burst. Then a short burst on December 21, 2005 appeared to have the total energy of a typical long GRB – 10 times as much as the most energetic short GRBs known. Several models have been suggested for these and other anomalous bursts. And this is in addition to models involving magnetars, applicable to perhaps 10% of short GRBs. Here's a good summary of the situation: Cosmic Explosion Mystery Deepens.

In May, further analysis of the July 24 event showed that it radiated its energy in all directions. However, the December 21 event appeared to radiate its energy in narrow jets with opening angles between 4° and 8°. Because the energy was narrowly focused, the GRB appeard to be very bright, but the total energy was not as high as if it had radiated in all directions. Thus this event actually had an energy in the normal range for short GRBs. However evidence of jets from other short GRBs is sketchy. And it is difficult to explain jets in a neutron star merger model. The short GRB situation is looking rather messy. Reference: High-energy jets spew from short gamma-ray bursts.

As if all that were not enough, later in May an analysis of a short GRB that occurred on January 21 showed that the event was from 10.1 billion light years to 12.7 billions light years away – where it would be as distant as the long GRB of September 4, 2005. All previously measured short GRBs were no farther than 6.5 billion light years. Consequently, GRB 060121 might be as energetic as the most powerful long GRBs. This might mean the energy of the burst was concentrated in very narrow jets, so it was not as energetic as it appeared – as with GRB 051221. Alternatively, the characteristics of neutron stars in the early universe may have been different than more recent ones, though this seems like a stretch. Or perhaps some entirely different model, involving neither neutron stars nor supernovae is needed. Reference: Distant gamma-ray burst may be in class of its own.

Late in August, 4 papers appeared in Nature that gave detailed analysis of GRB 060218 – the one of very long duration (33 minutes) but low energy. The event has been put in a new class called an X-ray flash. It appears to have a jet structure and result from a Type Ic supernova, which involves the least heavy type of star (about 20 solar masses) that can go supernova when its hydrogen and helium supply is used up. Instead of leaving behind a black hole, its remnant may be a magnetar. (Unlike short GRBs possibly resulting from magnetars that self-destruct from their own magnetic fields, this GRB was produced in the supernova event.) References: here, here, here, here, here, here.

The news flow then went quiet for a few months. And then yet another surprising twist showed up. This involved observations of two more long but nearby and low-energy GRBs. Because they were of the long type, they (presumabley) did not involve neutron stars or magnetars. But these two seemingly did not involve supernovae either, unlike GRB 060218. The events occurred on May 5 and June 14 of this year. The findings were published in the December 21 issue of Nature.

GRB 060614 lasted 102 seconds and occurred at a distance of 1.6 billion light-years. (Nowadays that's considered relatively nearby.) In a long GRB due to a supernova, there is a rebrightening that lasts for days after the initial flare. This is the primary source of light that makes the supernova visible. It comes from the gravitational energy of collapse and the energy of fusion reactions which occur. If anything like that happened in these two cases, it must have been at least 100 times fainter than normal. GRB 060505 lasted only 4 seconds, which is still longer than a short GRB (under 1 second), and was somewhat more than 1 billion light-years distant.

Not only was there no evidence in GRB 060614 of the light normally seen following a supernova explosion, but it occurred in a galaxy with few young stars – the only kind that can go supernova, because they must be massive and short-lived. The problem is that it's very hard to understand the sustained emission of gamma-rays for 102 seconds except in a supernova event. It was also more energetic than the normal short GRB involving neutron stars. There are various speculations about what may have happened in these peculiar GRBs, but as yet no tenable models. It could have been a type of supernova collapse which produced little or no light. Or a merger involving neutron stars that continued to produce gamma-rays for an extended time, perhaps in a system of more than two neutron stars and black holes (a more complex version of GRB 050724). Or perhaps something else entirely. Apart from something exotic, one suggestion is that the distance estimate for GRB 060614 is off, and there was a visible supernova, but it was too far away to be visible.

This mystery will probably spawn various hypothetical models in the next year or two, and may lead to a better understanding of supernovae.

References: here, here, here, here, here, here, here, here, here.

If you have a subscription to Science here's a pretty good summary of the situation: Burst-Hunter's Rich Data Harvest Yields a Cosmic Enigma

Tags: , , , ,


Links to this post:

Create a Link


Post a Comment

<< Home