An unusually long burst of γ-rays zapped Earth in December 2011, lasting 4 hours. The cause of this burst is now proposed to be a peculiar supernova produced by a spinning magnetic neutron star. See Letter p.189
The story of γ-ray bursts (GRBs) originates in nuclear-weapons monitoring during the cold war, and has been elaborated by subsequent technological developments and scientific detective work. GRBs were discovered by the Vela satellites launched in the late 1960s by the US Air Force. The spacecraft carried sensitive γ-ray detectors to monitor the Soviet Union's compliance with the Nuclear Test Ban Treaty. No nuclear explosions on Earth were seen. Instead, mysterious γ-ray flashes were detected, randomly distributed on the sky1. On page 189 of this issue, Greiner et al.2 present data for a γ-ray flash that suggest an association with a rare type of supernova, similar to an unusual type of stellar explosion that has been recognized only in the past few years3.
Nearly 50 years after the end of the cold war, following several space missions dedicated to high-energy astronomy and the harnessing of the most powerful ground-based telescopes, we have a clearer picture of the cause of GRBs. They fall into two main types, defined simply by the duration of the γ-ray emission. Short GRBs last between about 0.1 and 1 second, whereas long GRBs last from about 2 seconds to several minutes. The leading model for the production of short GRBs is the merger of a black hole and a neutron star (the dense nucleus of a dead massive star), or of a pair of neutron stars. It is thought that these mergers might also produce bursts of gravitational waves. If these can be recorded by future detectors4, it would directly validate one of general relativity's tenets and revolutionize physics.
Long GRBs are the more commonly detected, and make up about 70% of all the events detected by the Swift satellite, which is dedicated to the discovery of GRBs. The nearest known long GRB was located 40 million parsecs away, far beyond the local group of galaxies that contains the Milky Way. Relatively close events such as this tend to be accompanied by a particular type of supernova explosion. The leading hypothesis for their origin involves the collapse of a massive star followed by the formation of a black hole, which is surrounded by a spinning disk of gas assembled from the star's remains. As material from the disk falls onto the black hole, a jet is launched that accelerates particles close to the speed of light. The relativistic jet is focused in a beam and creates γ-ray emission, followed by X-ray and optical radiation. The signature of the supernova emerges days later, when the emission from the beam, known as the afterglow, fades rapidly.
In the past few years, a rare type of GRB has been discovered5, known as an ultra-long GRB. As the name suggests, the duration of the γ-ray emission from these can be several thousand seconds, but the events seem to be of similar power (energy emitted per second) to the bulk of the known long and short GRBs (Fig. 1). The most likely cause for one of the detected ultra-long GRBs — which lasted several days — was the tidal (gravitational) disruption of a star as it was being consumed by a supermassive black hole at the centre of a galaxy6. However, there are three recorded GRBs for which neither the supernova link nor the cause of the tidal disruption could be established5.
Greiner and colleagues studied the nearest of these, known as GRB 111209A. Its γ-ray emission lasted about 4 hours, and an afterglow detected at optical wavelengths allowed an accurate determination of its redshift (0.677). This tells us that the Universe was about half its current age when the light from the supernova was emitted. The unusual properties of 111209A have been studied by several groups5,7,8, but Greiner and colleagues' deep optical and near-infrared imaging now provide a superior data set suitable for a study of the decaying emission from the explosion.
After forensically detailed examination of the GRB's light curve — which depicts the evolution of its luminosity over time — the authors observed a distinct bump. This bump is the signature of a luminous supernova, corresponding to an explosion that is several times brighter than those typically associated with long GRBs. The spectrum of this supernova is also atypical: it does not show the strong absorption features due to iron that are prominent in the spectra of events associated with long GRBs. The authors suggest that this spectrum is similar to those of a new class of super-luminous supernova discovered in 2011 (ref. 3). Never before has the super-luminous class been directly associated with GRBs.
Super-luminous supernovae are some 10–100 times brighter than all other types, and evolve slowly. Most of them cannot be powered by the radioactive decay of the nickel isotope 56Ni, which is the standard physical model invoked to explain the light emitted from supernovae previously associated with GRBs. However, the light curves of super-luminous supernovae can be quantitatively modelled if extra energy is injected into the explosion from a rapidly spinning neutron star9. Neutron stars are formed when massive stars collapse, and the release of gravitational potential energy is what causes a normal supernova. Neutron stars that are born with spin periods of less than about 10–20 milliseconds have enough energy to power the emission observed by the authors.
The physical mechanism through which rotational energy can be extracted from a rapidly spinning neutron star is the emission of radiation from the object's magnetic-field poles (magnetic dipole radiation); neutron stars known as magnetars, which have magnetic fields of around 1014 gauss, can lose10,11 rotational energy through this mechanism on the timescales required to explain the authors' observations. These are extreme, but physically plausible, field strengths, and it has been shown that simple models of magnetic dipole radiation from magnetars do match the data from super-luminous supernovae quite well12.
Greiner et al. have reached the secure conclusion that an unusual and luminous supernova accompanied the ultra-long GRB 111209A. Their deduction stretches the standard physical model invoked to explain supernova luminosities to breaking point. Although, in this case too, the simple magnetar models fit the light curve quite well, they contain several free parameters that are unconstrained. The magnetic-field strength, spin period and ejecta mass can be chosen to fit many shapes of supernova light curves, not just this one. The models have therefore been criticized for being too flexible. The current analysis cannot confirm that a magnetar is indeed the powering mechanism of the supernova. Also, the low signal-to-noise ratio of the spectrum obtained does not lend itself to an unambiguous conclusion.
This possible link between super-luminous supernovae and GRBs requires further investigation, but the quest is hampered by the rarity of both phenomena. Greiner and colleagues' supernova is only one example, and although its light curve and spectrum bear some resemblance to those of super-luminous supernovae, they are certainly not a perfect match. It is also intriguing that most GRBs and super-luminous supernovae have been found in low-mass dwarf galaxies13. Dwarf galaxies are likely to have lower abundances of elements heavier than helium than the galaxies in which the bulk of star formation occurs in the Universe14. The similarity of the birthplaces of GRBs and super-luminous supernovae has also suggested13 a link between them.
Fifty years after a military mission unexpectedly made one of the most remarkable discoveries in high-energy astronomy, we are still struggling to unify the physical models of GRBs. Greiner and co-workers' findings add another twist to the tale of γ-ray astronomy, which will undoubtedly be followed by others in the next few years, when gravitational-wave detectors start surveying high-energy phenomena in the sky.
About this article
Physical Review D (2016)