As we've discussed a lot recently, gamma ray bursts have the tendency to be just as spectacular as supernova blasts, with none of the glory. Yet comparatively, our amount of knowledge about them remains decidedly small. Especially so when dealing with how they form. The firsts of these high energy events were brought to our attention by the Vela-5B Satellite in the early 1970's. At the time of their discovery, we had a difficult time pinpointing their origin, but over the course of the next few years, we were successful in determining that the events were non-local (meaning, they didn't originated in, or around the sun). Eventually, it was thought that the detected gamma ray-bursts, or GRBs (what these events were eventually called) had a luminosity of about 10^39 ergs, we now know that this is ~13 magnitudes smaller than the actual intensity. Due to the much lower assumed intensity value, astronomers came to the conclusion that the GRB's may take place in the accretion disk around a collapsing star.
Obviously, there was a lot of conflicting data. This resulted in the creation of several different models for GRB events. We had a model that posited the flares are caused by neutron star accretion. Another put forth the idea that they are created by supernova bursts in other galaxies. Additionally, another predicted that they form during the collapse of cores in active galaxies [in reference to very high redshift].
It has since been proposed that millisecond magnetars are another potential source of GRBs and even Soft Gamma Repeaters. It has even been suggested that primordial black holes could impact comets within the Oort Cloud, releasing energy equivalent to ~10^29 ergs or even the desired ~10^39 ergs. (This particular model was never given much credence)
It was also theorized that they could occur in galactic halos (our galaxy is enshrouded by a rather large halo, which could account for a substantial chunk of missing baryon matter). Any such bursts in the galactic halo would most likely come from neutron stars and their magnetic field lines, this idea adequately explained the line strengths and widths, but it did require a super-Eddington flux. White dwarf mergers were also considered as potential cause. As were 'star-quakes' experienced in the belly of neutron stars.
Over the years, many models have been proposed; some more plausible than others, but now, we have narrowed the search down to three progenitors that are widely considered the real cause of GRBs; neutron star mergers [NS-NS], hypernova and black holes. Lets look at them in more depth.
Neutron Star Mergers:
The current model of short GRBs (long duration gamma-ray bursts last about 2 seconds) posits that they are created through the mergers of black holes and neutron stars. The events kickstart when neutron stars initially start to gradually lose orbital energy through the emission of gravitational radiation. Eventually - after an extended period of time - they form a black hole with a temporary accretion disk. This, in turn, fuels the extended burst. It is expected that, in the future, we will be able to detect the gravitational waves emitted as this occurs, giving further credence to this model.
After additional observational evidence was uncovered, the NS-NS and NS-BH models began being used to explain both long and short GRBs (Even before the Compton Gamma Ray Observatory was launched and the BATSE, Burst and Transient Source Experiment [BATSE], tool revealed that they were isotropic across the sky.) Since then, several notable names have been leading the way in GRB theories and research since the mid 1980's. Yet, the favored theory progenitor has always the same; neutron stars. The collisions of these compact objects can release 10^51-10^53 ergs of energy, which is mind boggling.
Regardless, In both the NS-NS and NS-BH events, a black hole is left behind. The black hole is responsible for some of the variability. At the point of the merger, most - but not all - of the matter of both objects falls directly into the BH, but ~0.1 solar masses will make a accretion disk of material before it all falls inwards. This model matches well when coupled with the Fireball model, explaining the initial very short GRB, the near immediate x-ray afterglow and the extended optical afterglow that pervades the region.
The term hypernova originally came about to describe more energetic supernovae blasts, which produce fireballs that travel at relativistic speeds. As it is accepted that Long GRBs are caused by massive stars undergoing a core collapse (subsequently activating supernovae explosions), the amount of GRB to supernova detections that we have, it is reasonable to assume that although most [if not all] LGRBs are related to supernovae (although not all supernovae are GRBs).
The link between LGRBs and supernova is not unfounded, but the first incidence of this occurrence came in the late 90's. A GRB was detected by BeppoSAX on April 25, 1998, shortly after a supernova explosion was detected in the same area of the sky, it was concluded that GRB980425 and SN 1998bw were one and the same.
Through the spectral analysis of SN 1998bw, it was determined that it was a Type 1c supernova with a similar curve to two other supernovas, SNe 1994I & 1997ef. This type of supernova has a lack of hydrogen and helium lines. This is thought to be the result the outer shells having been blown off prior to the ignition of a supernova. Furthermore, our modeling suggests that it comes from the collapse of a star made up mostly of carbon & oxygen. Although these three supernova were very similar to Type 1c supernovae, they had their distinctions; like varying luminosities and curve shapes.
Since SN 1998bw, many other GRBs have had SN counterparts. This has reinforced the notion that a massive, rapidly rotating star, with strong magnetic field lines undergoes a core collapse supernova, resulting in what is called a hypernova. Bi-polar jets then accelerate material to a Lorentz factor of ~100. This is the basis of the Fireball model. Even before the confirmed correlation between SN and GRBs, it was suspected that GRBs 970228, 970508 & 970828 were close to star forming regions, suggesting that the events were the result of"failed supernova events", as proposed by two astronomers in 1999.
We now know that there are a couple of progenitors, one for short GRBs and another for long GRBs. Both events ultimately involve a black hole, NS-NS mergers from a black hole and a hypernova can form one also. Hopefully, one day in the future, we will witness the merger of neutron stars and black holes, we want confirmation! There is enough observed evidence to correlate LGRBs and supernova. SGRBs, however, are only testable through computer modeling. Headway has been made though, through the detection of GRB 130603B, this could be our link between SGRBs and NS mergers. Regardless, the origin of these events may prove to be stranger and more complex than we can currently imagine.
Related Reading About Gamma-ray Bursts:
- Gamma Ray Bursts and The Fireball Model
- The Birth of Gamma-Ray Burst Astronomy
- The Study of Gamma Ray Bursts (part II)