Where Have All the Gamma Ray Bursts Gone?
Written by Ian O'Neill
Artist impression of a GRB (ESA)
Astronomers are confused.
As if gamma-ray bursts (GRBs) weren't mysterious enough, there's something else to add to the bag of confusion. GRB events are missing from the furthermost reaches of the Universe. Right around the time when there should be a lot of GRBs, during the "star forming epoch" (when stars were just beginning to evolve after the Big Bang), there appears to be none. Zero. There's no ancient flashes of massive star death to be found. What's more, there doesn't appear to be any afterglow from previous gamma-ray bursts either.
So what's going on? Were there no GRB events before 12.8 billion years ago? Possibly… although there might be another answer. They are out there but we just can't see them.
Gamma-ray bursts are the biggest and brightest explosions in our Universe since the Big Bang. When a GRB detonates, it can easily outshine its host galaxy containing billions of stars. These energetic events have been observed since the 1960's and only until recently have astronomers found an explanation as to what GRBs are. A GRB occurs when a young metal-poor massive star has used up all its fuel and, like a supernova, collapses under its own gravitational field. The rapid-spinning star then funnels intense beams of radiation from its poles in the form of gamma-rays. Should one of these beams be directed toward Earth, we see a disproportionately bright explosion (as a vast amount of energy is channelled through the poles). Until the "collapsar model" was devised, astronomers were at a loss to explain these energetic events.
The collapsar model appears to explain GRBs lasting for two seconds or more. However, there is another class of GRB, of much shorter timescales, that does not fit in with the collapsar model. Short-period GRBs may be the result of violent interactions between black holes and a neutron stars.
So, does this mean GRBs are becoming less mysterious? Actually, GRB theory has just become a little more complicated. It would appear that no GRBs occurred before 12.8 billion years ago. Last month, the most distant (and therefore oldest) GRB was detected 12.8 billion light years away, but that in itself is strange.
During the time when the first stars started to form (around 13.4 billion years ago), they were by definition "metal-poor" stars (heavier elements, such as metals, were only possible after several generations of stellar evolution), so this should be a period of time when GRBs were regularly lighting up the night sky. However, according to observations of the most distant galaxies containing the youngest stars, GRB events seem to be non-existent.
One explanation put forward is the effect of red shift. As the Universe expands, space-time stretches. As light travels from the most distant reaches of the Universe, perhaps the light itself from GRBs has been so stretched (red-shifted) that the electromagnetic emissions simply cannot be detected by our instrumentation. These huge explosions could be happening, but as the emitted light has been so red-shifted, by the time the light reaches us, perhaps the emission does not resemble a GRB. Even the afterglow of one of these massive explosions would be unrecognisable in this case, the light observed would be shifted all the way into the infrared.
So will any GRBs be discovered further away than 12.8 billion light years? I think we'll have to wait until we build some improved infrared observatories or recognise what a distant, ancient GRB looks like…
Source: NASA
Filed under: Astronomy, Physics
Related stories on Universe Today
Comment policy: Be nice and brief. Don't advertise your stuff, or promote your personal theories. We'll delete any comments that break these policies. Click here for more details.
October 23rd, 2008 at 7:41 pm
I'm one of the co-authors on that "12.8 billion year old GRB" paper. I don't get what all the fuss is about. Confirming GRBs at these redshifts is a very tough job (something which is mentioned in deeper detail in the original press release), since all flux moves into the NIR region where the atmosphere starts giving you BIG problems with airglow skylines. No one has EVER published successful NIR spectroscopy of a GRB afterglow, even of the really bright ones!
The discovery of the afterglow of GRB 080913 was sheer luck, since it occured over Chile at nighttime, and was observable immediately with the seven-color GROND camera which got a fast photometric redshift and the ball rolling.
There's just not much instrumentation yet that allows us to capture and confirm these extremely distant GRBs. I see no reason to claim that "they are not there."
People said the same thing about afterglows and host galaxies untill 970228.
October 23rd, 2008 at 9:10 pm
Nice clarification, Don. Wouldn't NIR imaging from space see a lot of these?
October 24th, 2008 at 1:38 am
Here's my 2cents: All stellar axises were perpendicular to our perspective before 12.8 billion years ago.
October 24th, 2008 at 4:26 am
No one has EVER published successful NIR spectroscopy of a GRB afterglow
That'll be why I couldn't find any papers about it, then!
Do you know; have there been any published spectroscopic measurements in the visible…?
October 24th, 2008 at 6:31 am
Could the missing GRBs have all red-shifted out of sight?
October 24th, 2008 at 7:24 am
@Helio George: Yes. of course. This is one of the clues of the proposed JANUS mission.
Couldn't find a project website that really gives details…
But Google for JANUS GRB SMEX.
@Hex: That would seriously violate the Copernican principle, and there is no reason at all for that to happen…
@Invader Xan: Err… Yes? Dozens and dozens of them? I mean, where else would we get the redshifts from?
@Chuck Lam: Well, depends on what you mean with "out of sight". If you just mean "out of the visible spectrum", then yes, that's exactly what's happening, but that's also what I stated above. If you mean "into some kind of total invisibility", then of course not, the most redshifted radiation we see is the CMB from a redshift of roughly 1250, which is way before the first stars.
October 24th, 2008 at 8:25 am
Effectively, we are still looking through straws, regardless of the equipment we make. We'll make better straws in the future, but they will still be straws.
October 24th, 2008 at 9:05 am
Is the expected NIR outside Spitzer's spectral window?
October 24th, 2008 at 9:32 am
It was a really lucky coincidence discovering this GRB with such long wavelengths, I agree.
Still there is a problem to GRB's in general.
For a black hole to produce the GRB's there must be an enormous other source delivering the steady-dust-delivery.
Where does this dust come from? This takes time or the source is very nearby the black-hole.
Is it illogical to assume the dust comes from the source itself (black-hole) itself?
Matter-energy conversion is normal, energy-matter conversion is very unlikely.
But is it impossible?
October 24th, 2008 at 9:33 am
@ bob: Your anology is pretty good. But when LSST comes along, we'll be guzzling out of barrels.
@Helio George: Well, Spitzer's spectral window is pretty wide… So, no. Problem is, Spitzer has, like HST, a rather long reaction time (several days) and afterglows are faint. There've been a few successful afterglow detections with Spitzer (e.g., Heng, K. et al. 2008) but it has been very rarely used, and only for IRAC photometry. MIR observations of GRBs is an as yet uncharted field, the last big hole, actually. Everything from gammas (except extremely hard ones, think TeV) to NIR is fully covered, and then you have many observations from (sub)millimeter to radio. Just not MIR/FIR (for afterglows, at least).
October 24th, 2008 at 10:35 am
Sorry, without being foolish
about Gamma Ray Bursts – strong irradiation indirectly causes the circumstellar matter to ionize immediately around the black hole, and it cools down slowly.
So before the energy-matter conversion from the GRB there must have been another mechanism producing the dust itself.
That matter is what i am referring at.
October 24th, 2008 at 10:54 am
@Hannes: Do you have the first idea how GRBs are even produced?
October 24th, 2008 at 12:27 pm
@ Don Alexander.
No, i do not know for certain how GRB's are produced and i make a lot of mistakes.
And i am a complete noob comparing to somebody like you.
Also sorry – I tend to go sometimes off topic, no offence.
But how sure are you about the nature of GRB's as well about the nature of Quasar's?
I think even the model of an atom is something discussable. How do you explain static friction for example. Seeing an atom as a different space-time conjunction could explain this imho much better.
But I am in no way your's equal scientifically, just someone sceptical.
October 24th, 2008 at 12:51 pm
Universe Today- Not only bringing you the news, but also bringing the news to the people making it who then bring it back … to you (hmmm, sounded funnier in my head…)
@ Don Alexander
I was a bit muddled at first with the article, but crystal clear now, thanks for the clarifications!
October 24th, 2008 at 1:26 pm
@Hannes: It even says so, in a basic way, in the article! I note that "The rapid-spinning star then funnels intense beams of radiation from its poles in the form of gamma-rays." is oversimplified, but otherwise, you have it in a nutshell.
Evidence for (long) GRBs being linked to the deaths of massive stars is very solid. Evidence linking at least some short GRBs to a process that does not need recent star formation is also very solid. That they come from neutron star mergers is a good working model, but we have no smoking gun evidence (specific gravitational wave signal) yet.
There's nothing wrong with being a sceptic. All scientists are sceptical by Nature. On the other hand, we "believe" in the data. And there's a lot of data on GRBs and quasars, and we have models that describe this data very well. Of course, there's devils in the details, and science is about eternally tweaking the models to encompass all observations.
But just going out and saying it's all nonsense, all the mainstream is wrong… Well…
If you have ATM ideas, why not present them in the ATM forum at BAUTforum and see how they fare?
October 24th, 2008 at 1:44 pm
@ Don Alexander
I also believe in the data, like you.
But scientific methods imply also a look at the less obvious.
It is also very interesting what you do, i really admire your job.
October 24th, 2008 at 1:47 pm
And really thanks for the reply!
October 24th, 2008 at 2:21 pm
To Don Alexander, More specifically, could the missing GRBs be caused by the accelerating expansion carrying the GRB sources away from us eventually faster than the speed of light?
October 24th, 2008 at 3:49 pm
Don said: MIR observations of GRBs is an as yet uncharted field, the last big hole, actually.
Thanks, that is newsworthy in itself.
What is the hole filler then? The James Webb? Or will Spitzer get freed, eventually, for such things?
October 24th, 2008 at 11:29 pm
@Chuck Lam: No, this seems to be a common misconception of cosmology. The accelerating expansion of the universe is a relatively "recent" phenomenon, happening, IIRC, at about redshift 0.7. These GRBs are much older. While, of course, there are GRBs that occured beyond the particle horizon (and even the event horizon of the universe), there must be those that occured within, since the surface of last scattering is also within the particle horizon. We are going to be able to see those.
@Helio George. Good question. To my knowledge, there is no MIR/FIR mission flying or in design phase that is somehow conceptualized to react reasonably rapidly to transient events (like XMM Newton, which can observe GRBs within a few hours, whereas Chandra or HST take several days). JWST has MIR capabilities but may be as sluggish as HST. And no idea about Herschel, the upcoming (February 2009, keep your fingers crossed) FIR/Submm mission.
This is actually a pity, as the characteristic frequency (the frequency at which the most flux is emitted) lies in exactly this hole of coverage. So GRBs are actually pretty bright in this regime, but then, so is the dust in the Milky Way surrounding us. So you will be getting background problems even from space…
October 25th, 2008 at 9:04 am
Gamma rays lie at the extreme "blue" side of the spectrum. If they were red-shifted enough to be missed, then virtually ALL the other wavelengths emenating from that galaxy would be shifted to the same degree. I'm certain that this is not so. What am I missing here?
October 25th, 2008 at 9:52 am
@neoguru: No, you are quite correct. The problem in actually detecting the GRBs themselves is twofold: First, most of the emission is shifted into soft gamma-rays and X-rays. For most detectors (Fermi LAT excluded) there is a correlation between the field of view in which events are detectable and the localization precision. That's why Swift follows GRB detections (a few arcminutes) up with the X-ray observations (a few arcseconds) But the XRT FOV is small, less than the full moon. If the GRBs are redshifted strongly, the become X-ray flashes, which are harder for BAT to localize, and basically never pop off in the XRT FOV.
Second point is that time dilation spreads the GRB out by a factor 1+z. If the redshift is nine, a 60 second long GRB becomes 10 minutes long, while at the same time becoming much fainter due to distance too. Swift can surmount this in part by using image triggers, which take long exposure (10 seconds, or 60, or even 10 minutes) shots and compare them with earlier gamma-ray images, looking for new point sources.
But no one says it's easy.
October 25th, 2008 at 1:17 pm
@ Don Alexander, Thanks for the thoughtful and illuminating responses to the various questions posted above. I think interested, thoughtful readers (like me) find your responses well-informed and easy to grasp.My (non-posted) thought of using the JWST for MIR spectra fared as I suspected (too slow to be of much use). But I am heartened to see that LSST and possibly PanStarrs-4 may mitigate the problem of quick detection of the MIR spectra of z>7 GRBs. Thanks again for your insightful responses to the various queries.
October 25th, 2008 at 3:05 pm
@ Don Alexander, I second the response of Jan Hanford above. I really appreciate the clear and detailed answer to my inquery. Thanx! NeoGuru
October 25th, 2008 at 9:31 pm
@Jon Hanford: Um, how do you connect LSST and the full PanSTARRS with MIR spectra??? The two have nothing at all to do with each other. These are wide-field optical survey telescopes. I had just mentioned LSST in a response above because it will not be a "straw" (small FOV small throughput) like today's instruments.
MIR observations of GRBs from the ground (e.g. with VLT VISIR) are not feasable, even 8m class telescopes do not go deep enough, sky background is way too high. Perhaps when we get the 30m class beyond 2015…
October 26th, 2008 at 3:10 pm
@ Don Alexander, in my post above, I didn't mean to imply LSST and PanStarrs would be taking hi-res spectra of these distant GRBs (I know they will employ crude photometric filters as means of making approximate assumptions of the visible-light redshifts of observed objects). I meant to imply that these visible light deep sky surveys may improve the chances for larger ground, air( SOFIA ), and space-based observatories to home in on these extremely distant GRBs. Perhaps JWST will be able to shed some light on the spectra (and distance and composition) of this population of GRBs after all. If not, maybe TMT, OWL etc. will be needed to nail this one down.
October 26th, 2008 at 8:20 pm
@Jon Hanford: Not sure what the filter sets for LSST and PanSTARRS are, but I think they are optical only. Which means that thay will not be able to detect z > 7 GRBs at all (or only faintly in the z band)!!
Furthermore, both are survey telescopes which will, with high probability, not be doing targeted (and especially rapid) follow-up observations at all.
What we need are dedicated 2-3m class robotic scopes (like the RoboNet telescopes) equipped with GROND-like cameras. With GRB 080913, which had a quite faint afterglow (I mean intrinsically, after correcting for the distance), GROND showed it can discover even such "mediocre" GRBs at high redshift.
October 27th, 2008 at 8:47 am
@ Don Alexander, LSST filter groups will include u, g, r,i(750nm), z(865nm), y( unspecified) and eventually add NIR-MIR filters designated Y1, Y2, and Y3. A paper on the instrumentation and science goals is available at arXiv:astro-ph/2366v1. Page 4 of the paper includes section 2.1.3 entitled 'Exploring the Transient Optical Sky' and lays out detailed plans of how GRB searches may be conducted with LSST. Exposures over time will range from 1 minute to many years (cumulative) and the paper also states that the Deep Lens Survey will reach apparent r mags of 24.5 routinely. Even though LSST is primarily a survey telescope, Targets of Opportunity( ToO ) time will be set aside to explore unusual, peculiar of transient objects in some cases. Even if the distant GRB burst may not be optimally imaged in LSST filter passbands, would it still be feasible to locate the parent galaxies ( or protogalaxies) that these first GRBs appeared in, possibly determining their redshift with larger ground, air, and space observatories. By the way, PanStarrs filter lineup will include g, r, i, z, and y (1 micron) filters in its' survey program, with individual exposures to range from 30-60 seconds, but probably not deep enough to pinpoint these first GRB populations. I also looked at the capabilities of the specialized spectrographic telescopes HET and SALT, but found that their spectroscopic range is indeed too limited for distant GRB events. I, too, see the need for dedicated 2-3m class robotic scopes as these too have proven their worth in studies of this type. Obviously, this type of research will need to rely on a panopoly of astronomical instruments over multiwavelength regimes. Thanks again for your insightful replies to my and others queries and speculations. (Hi-res version of the LSST design & science is available also at the LSST website).
October 27th, 2008 at 9:32 pm
@Jon Hanford: Thanks for the info, did not know that about the possible NIR/MIR filters and the ToO follow-up projects.
Notwithstanding the power of LSST, simple physics implies that NIR observations are needed beyond ~ z = 7.5.
HET and SALT can't be on a target for more than ~ 1 hour at a time, which is probably not good enough for faint high-z bursts (independent of iunstrument capability).