Gamma-ray bursts are events that last from less than a second to many thousands of seconds, and as their name implies, they are random bursts of gamma-rays. While we have known of their existence since the late 1960s, the partial understanding we now have is a consequence of multiwavelength observations made within the past decade. The understanding is partial because we only know with certainty the origin of one of the two types of gamma-ray bursts. The long-duration gamma-ray bursts, those of more than one second, are from core-collapse supernovae in very distant galaxies. The short-duration gamma-ray bursts, those of less than a second, are still not associated with an astronomical object.
Gamma-ray bursts emit the bulk of their energies above 50keV; they produce relatively few x-rays, and they produce no immediately observable optical or radio emission. The afterglow of a burst, on the other hand, is very visible. At x-ray energies, a short-lived x-ray source is seen. At radio wavelengths, the afterglow radio emission is observable for many days, and it has the characteristics of synchrotron radiation, which is the type of electromagnetic radiation produced by an electron traveling at close to the speed of light through a magnetic field. At optical wavelengths, the emission is the emission of a core-collapse supernova; for the distances normally encountered with gamma-ray bursts, the optical emission is only about magnitude 24 or 25, making it observable only to the very large telescopes.
The picture that has emerged from this data is of a high-velocity jet of material breaking out from the core of a massive star undergoing core collapse. The jet moves at nearly the speed of light, so that the kinetic energy of the material is hundreds or thousands of times greater than the rest mass of the material in the jet. Some of the energy of the jet is immediately converted into gamma-rays, which are seen promptly. The remainder of the energy is gradually lost as the jet interacts with its surroundings, causing the extended radio emission.
The region emitting gamma-rays in a gamma-ray burst appears to be moving very close to the speed of light. The evidence for this is the shape of the gamma-ray spectrum and the apparent absence of a process call photon-photon pair creation. If the gamma-rays were created in a region that is optically-thick, meaning that the gamma-rays scatter on average more than once before escaping to an observer, then the spectrum would be a thermal spectrum, meaning that it falls-off exponentially at high frequency, and it is characterized by only one parameter—a temperature. Because gamma-ray burst spectra are not thermal spectra, they are produced in an optically-thin region. But gamma-ray bursts are known to come from distant galaxies, and if the gamma-ray emitter were at rest, then the density of gamma-rays at the emitter would be so high that electron-positron pairs would be created through the collision of gamma-rays; this process, photon-photon pair creation, is the inverse of the annihilation of an electron with a positron to create two gamma-rays. The rate at which pair creation would occur would make the area around the emitter optically-thick, which would produce a thermal gamma-ray spectrum. Having the emission region move at close to the speed of light towards the observer resolves this contradiction; the high intensity and short duration of a gamma-ray burst are then consequences of special relativity;.
This hypothesis of a high velocity emission region is reinforced by the observations of radio afterglow from gamma-ray bursts. The dimming of the radio source, and the evolution of its spectrum, is what one expects for emission by a relativistic object that is slowing. The theory that is consistent with the radio observations is that a short-duration jet of material is moving at close to the speed of light and is spreading out into a narrow cone; the material is therefore a thin disk that has a diameter that increases proportionally with the distance from the jet source. The shell sweeps up interstellar material, causing it to slow. When the shell is moving rapidly, only the part of the shell that is within the beaming angle γ-1 to our line of sight is visible. As the shell decelerates, so that the Lorentz factor γ decreases in value, more of the shell is seen, until finally, when the opening angle of the cone equals γ-1, all of the shell is visible. This transition from seeing only part of the shell to seeing all of the shell appears as a change in the rate at which the radio emission is declining; the radio emission falls-off more rapidly after all of the shell is seen than before.
The association of gamma-ray bursts with core collapse supernovae leads to one obvious theory, and that is that when the core of a massive star collapses to form a neutron star (or perhaps a black hole), part of the energy released into the surrounding stellar envelope goes to drive a jet through the star to the outside. The theories that are now being developed resemble the theories for jets from active galaxies: the stellar core spins rapidly as it collapses, creating a jet along the axis of rotation through magnetic and radiative forces. The jetis powered by both gravitational potential energy and rotational energy that is extracted from the collapsing stellar core. The jet forces its way through the outer layers of the star. Once it breaks out of the star, the radiation from the jet propagates to the observer and is seen as a gamma-ray burst. This theory need not be the only theory, and with time others should be developed.
The mechanisms for gamma-ray emission are still not proven, although many theories exist. There are few observational tests that can distinguish these theories, and the theorists themselves have not been fastidious in creating testable theories. The most favored emission mechanism is synchrotron emission. The energy source is usually shocks within the jet; within the shock, electrons are accelerated to high energy, after which they emit synchrotron radiation as they move through the magnetic field carried by the jet. This alone is not sufficient to explain the observed spectra, which have a characteristic energy of 50 to 200keV; this may point to the low-energy gamma-rays being scattered out of the line of sight.
The precise mechanism for accelerating electrons in the relativistic jet has not yet been satisfactorily developed. Often a rather simplistic mechanism called Fermi acceleration is invoke; in this process, it is imagined that electrons travel back and forth across a shock, scattering with waves or turbulence on either side, and gradually building up energy. Effectively, the wave and turbulence are massive objects that the electrons are trying to thermalize with. To come into thermal equilibrium with such massive objects, the electrons must have velocities close to the speed of light. This cartoonish process sweeps under the rug all of the physics of the interaction between electrons and waves or turbulence. A better understanding of the radiative physics would come from studying the plasma processes that accelerate electrons. Such processes, however, are difficult to simulate, and so progress on this front will be slow.