Thursday, October 06, 2005

What are gamma-ray bursts?

There have been a lot of interesting results in the past few months which seem to be providing a firmer understanding of the mechanisms that underlie gamma ray bursts. The reason is simple: NASA's Swift satellite observatory, launched in November 2004 and designed especially to detect gamma-ray bursts and report them quickly to ground-based astronomers, has been a great success.

According to the Wikipedia article, "Gamma-ray bursts (GRBs) are the most luminous physical phenomena in the universe known to the field of astronomy. They consist of flashes of gamma rays that last from seconds to hours, the longer ones being followed by several days of X-ray afterglow." (This article itself is not especially up-to-date.)

GRBs are scattered uniformly around the sky, and some are known to be extremely distant, implying a very high output of energy, typically around 1051 ergs.

The big question is: what causes them? There have been two main ideas for an answer: very powerful supernovae, or collisions between neutron stars or between a neutron star and a black hole. It's possible, of course, that both processes occur, in different events.

Let's look at some of the recent research reports.

Evidence for neutron star collisions

Creation of Black Hole Detected - May 9, 2005
Scientists Watch Black Hole Born in Split-Second Light Flash - May 11, 2005
Short gamma-ray burst hints at neutron star merger - May 11, 2005
Signs Point to Neutron-Star Crash - May 13, 2005 (subscription required)

On May 9, NASA's Swift satellite detected (as it was designed to do) a sharp gamma ray flare. The energy spike lasted just .05 seconds. This is far too brief to be caused by a supernova event. So it is called a "short" burst, that must be a different sort of event from a "long" burst, which lasts from a few seconds to a few minutes and is generally assumed to be caused by a supernova.

As intended, the Swift satellite notified ground-based astronomers of the event's location. Within a few hours, ground telescopes detected a faint patch of light near an old galaxy 2.7 billion light years away. The galaxy has not been forming new stars for several billion years, and the location, away from the galaxy, of the visual flare was consistent with what would be expected for a pair of neutron stars that could have been ejected in a much earlier supernova event.

Because short burts are so brief, it has not been possible before to locate any optical aftereffect, and therefore not possible to pick up any clues about their nature. The burst has been named GRB050509b.

Link between supernovae and gamma-ray bursts

Naked carbon/oxygen stars linked to gamma-ray bursts - May 26, 2005
The Link Between Supernovae and Gamma Ray Bursts - May 27, 2005 (subscription required)
An Asymmetric Energetic Type Ic Supernova Viewed Off-Axis, and a Link to Gamma Ray Bursts - May 27, 2005 (subscription required)
Double Whammy: Cosmic Fireworks Explained - May 30, 2005
Naked Star's Outburst
The latest in core collapse astronomy - June 16, 2005

Supernovae have been linked to GRBs by observations at visible wavelengths that show evidence of a supernova event at the same location as the GRB. But the precise type and mechanism involved in the supernova has been unclear. The research reported here describes a supernova, located at a distance of 260 million light years, named SN 2003jd, and first observed in 2003, that seems to have been disrupted by an asymmetric process. "Long" GRBs have been hypothesized to result from supernovae which project high-energy jets of matter in the direction of Earth, in what is known as the "collapsar" model. So this supernova could be such an event whose jets were not directed towards us.

The reasoning that originally linked GRBs to supernovae came from observations of an earlier supernova, SN 1998bw, which was found at the site of a GRB. This was not the first such discovery of a supernova in the location of a GRB. (That happened a year earlier.) But 1998bw was only 80 million light years away, so it was possible to learn much more about the nature of the event. In particular 1998bw was found to be a "Type 1c" supernova, meaning its spectrum indicated a lack of hydrogen and helium, which would have been lost before the explosion. If the progenitor star had been spinning very rapidly, perhaps due to collision with another star, then two large opposing jets of matter would be ejected at very high speeds along the axis of rotation, and this would account for a copious production of gamma rays.

Such events, known as "hypernovae", are believed to result from the explosive collapse of a star that weighs between 10 and 100 solar masses and has completely lost its outer layers of hydrogen and helium. If 1998b were a hypernova and one of its jets pointed in our direction, we would see the event as a GRB. (Indeed, the expulsion of matter in a jetlike matter is necessary, as well as sufficient, to explain GRB as some sort of supernova event. That's because if the explosion were spherically symmetric, the same amount of energy would be expelled in all directions, so that the total energy would be too great to be produced in a supernova.)

However, there was one problem. Since 1998bw was only 80 million light years away, while most GRBs had been found at distances of billions of light years, 1998bw must have been much less powerful. Therefore, it seemed hard to accept that much more distant and powerful GRBs might have resulted from the same sort of event. And yet, in 2003 a GRB was observed that was much farther away than 1998bw, hence about 1000 times as powerful. But since this object was much closer than most known GRB, it was possible to determine that it had the same general structure as 1998bw, namely a hypernova. Hence GRB could indeed be explained as hypernovae.

Not all Type 1c supernovae need be hypernovae, i. e. they don't need to have jets. But if GRB are explained as hypernovae, then we would expect there to be many Type 1c hypernovae which aren't GRB, simply because of the unliklihood of a jet pointing in our direction. That's were 2003jd comes in. It is a Type 1c supernova. In order for it to be a hypernova it would need energetic jets. But how would we know, if they weren't directed towards us? The answer is that its spectrum would give evidence of asymmetries. And that is exactly what was found with 2003jd. So we have evidence that hypernovae aren't just theoretical possibilities and do actually occur.

More evidence for neutron star collisions in short bursts

Astronomers unravel cosmic explosion mystery - August 10, 2005

As noted above, "short" and "long" gamma-ray bursts have been suspected to result from different processes. While the evidence is good that long bursts (producing gamma-rays for a couple seconds or more) are caused by a type of supernova in a massive star that lacks hydrogen and helium, short bursts have been more mysterious, partly because the gamma-rays fade so quickly that locating the event at optical wavelenghts has been difficult.

But in July, two more short burst events were detected. One on July 9 lasted a tenth of a second, and another on July 24 lasted a quarter of a second. The first was found to be about 10,000 light years from the center of a star-forming galaxy that is 1.8 billion light years away. The second was inside a galaxy consisting of old stars that is 2.8 billion light years away.

A short burst on May 9 (described above) was found to be associated with a galaxy 2.8 billion light years distant. However, in this case, the burst lasted only one twentieth of a second, so it released less total energy. It is significant that the bursts in July were more energetic, because that apparently rules out an alternate model for the source of short bursts. This model involves a flare driven by a very strong magnetic field of a neutron star (an object called a magnetar). But there is an upper limit to the amount of energy such an event can release, and the July burst events would exceed that limit.

This is more evidence that the event was caused by a collision between a neutron star and another neutron star or a black hole.

Long bursts have a more complex structure than previously supposed

Giant Space Blasts a Two-step Process - August 17, 2005
NASA's Swift Satellite Finds Newborn Black Holes - August 18, 2005
Scientists Watch Baby Black Hole Get to Work Fast - August 18, 2005
Bright X-ray Flares in Gamma-Ray Burst Afterglows - August 18, 2005 (subscription required)
Black holes born through a brutal labour - August 18, 2005
The baby black holes with giant hiccups - August 19, 2005
Black Hole Surprise: Multiple Eruptions Seconds After Birth - August 23, 2005

Although the collapsar model of long gamma-ray bursts now seems fairly well substantiated, the actual sequence of events following the beginning of the burst apparently can be more complicated than previously supposed. This conclusion follows from observations of the burst known as GRB 050502B, which occurred on May 2, 2005, as well as an x-ray flare event known as XRF 050406, which occurred on April 6, 2005. X-ray flares may occur either by themselves or as part of a GRB. (Which suggests that an x-ray flare by itself might be associated with a GRB that doesn't have a jet directed towards the earth and is therefore not observable.)

There were two bright x-ray flares that appeared in the GRB 050502B afterglow and peaked minutes after the main burst. One of these flares carried as much energy as the GRB itself. Evidently, there is a lot going on after the end of the gamma-ray emission.

The standard collapsar model is based on a supernova event in a star which is sufficiently heavy that when the internal thermonuclear fusion process in the core ceases, the star collapses directly to a black hole rather than a neutron star. (Exactly how heavy the star must be isn't clear, but may be around 25 solar masses.) The initial burst of gamma-ray energy is produced by shock waves within the rapidly collapsing stellar matter. This can last from a few seconds to several minutes.

If the star is spinning rapidly, the collapse is not spherically symmetric. Instead, to conserve angular momentum, matter tends to collapse roughly parallel to the spin axis. The result is that at first the star's matter forms a lens-shaped object, in the center of which a black hole develops. This is followed by the contraction of the lens itself under the enormous gravity of the black hole. But not all the matter in the outer parts of the lens falls directly into the black hole. Instead, the matter is forced out at nearly relativistic speeds into opposing jets along the rotation axis, like water squeezed out of a sponge. Shock waves in the collapsing matter then heat the jets to temperatures high enough to generate gamma-rays.

If the progenitor star still has hydrogen and helium in its outer layers, then the jets can be absorbed by this remaining gas and their gamma-ray emission will be extinguished. On the other hand, if the hydrogen and helium layers have been stripped away previously, by a companion star or black hole for example, then the jets can emerge, and the result is a gamma-ray burst.

That's the simple picture. However, in the case of GRB 050502B, the process must have been rather more complex. The gamma-ray emission lasted abou 17 seconds. But about 8 minutes later, there was an enormously powerful burst of x-ray photons, comparable in total energy to the earlier gamma-ray emission. Much smaller x-ray flares have been observed in connection with other GRBs, but his was 100 times as energetic as any seen before. In other GRB events, as many as four separate x-ray flares have been observed.

Evidently the whole process can be much more complex than the simple version sketched above. Exactly what goes on isn't clear. The matter swirling around the black hole may be in such chaotic motion that repeated smaller explosions occur as clumps of matter fall into the black hole. Or perhaps some of the material in the jets falls back upon itself under the huge gravitational pull of the black hole. X-ray flares can be detected in about a third of GRB events. It may be that the speed of rotation of the progenitor star influences how "smoothly" the process occurs.

More precise analysis of some short gamma-ray bursts

Swift Spacecraft Solves Mystery of Short Gamma-Ray Bursts - October 5, 2005
Colliding Stars Behind 35-year-old Mystery - October 5, 2005
HETE-2 satellite solves mystery of cosmic explosions - October 5, 2005
Split-second explosions, so-called short gamma-ray bursts, solidly linked to stellar collisions - October 5, 2005
HETE Satellite Solves Mystery of Short Gamma Ray Bursts - October 5, 2005 (PDF file)
35-year-old mystery solved in a flash of light - October 5, 2005
Ancient Instellar Collision Helps Explain Source of Radiation - October 5, 2005
Discovery of the short g-ray burst GRB 050709 - October 5, 2005 (Nature paper, PDF file)

The observations of short GRBs on May 9, July 9, and July 24 of this year, mentioned above, have been more carefully analyzed, and the results indicate that these events almost certainly involved collisions between an orbiting pair of neutron stars, or a pair of a neutron star with a black hole.

The May 9 event was a GRB that lasted .07 seconds in the vicinity of an elliptical galaxy about 2.9 billion light years away. Since elliptical galaxies are old, they do not contain massive young stars which could produce a supernova.

The July 9 event (GRB 050709) was detected by a satellite named HETE-2, and its X-ray emissions were measured by NASA's Chandra X-ray observatory. This made it possible to detect a visible light afterglow, which placed the event inside a small blue galaxy 1 billion light years away. Given the known distance, it could be calculated that the event was only between a hundredth and a thousandth as intense as a typical long GRB. That amount of energy is consistent with a collision between neutron stars. Also, the event occurred in the outer area of the galaxy, which is where a pair of neutrons stars is most likely to be, and not compatible with a supernova, which is the explosion of a young star. Clinching the case, no supernova was detected at visible wavelengths. However, like long GRBs, the typical short GRB also seems to radiate gamma-rays in jet-like beams that occupy only one 30th of a full sphere.

It's not certain that these events involved a pair of neutron stars rather than a neutron star and a black hole. There are hints that the latter may have been the case, especially in the July 24 event, which was four times as powerful as the one on July 9, as would be expected since a black hole would be more massive than a neutron star. The latter event contained several bursts of x-rays, which suggests it involved a black hole that first consumed most of the neutron star and then consumed smaller fragments separately.



Links to this post:

Create a Link


Blogger Josh said...

Thanks for the trackback. I'm Josh Smith, from over at 3 Quarks, by the way.

If you're interested, Wired has a good article up about all of this. And keep your eye out for those papers in Nature tomorrow. I, for one, am excited. It was at least something to talk about today...

10/06/2005 01:40:00 AM  

Post a Comment

<< Home