2 What is the trigger?

The photon luminosity, for the few-second duration of a typical burst, is of course colossal: it exceeds by many thousands the most extreme output from any active galactic nucleus (thought to involve supermassive black holes), and is 14 orders of magnitude above the Eddington limit for a stellar-mass object. The total energy, however, is not out of line with some other phenomena encountered in astrophysics - indeed it is reminiscent of the energy released in the core of a supernova, the big difference being that the primary sudden event (with a timescale of seconds) is not smothered by a stellar envelope, as in a supernova, but manifests itself in hard radiation that escapes more promptly.

Unless they are beamed into less than one percent of the solid angle, the triggers for GRBs are thousands of times rarer than supernovae. The most widely favoured and conventional possibility is coalescence of binary neutron stars (see, for example, [28, Narayan et al. 1992)]. Systems such as the famous binary pulsar will eventually coalesce, when gravitational radiation drives them together. When a neutron star (NS) binary coalesces, the rapidly-spinning merged system would be too massive (for most presumed equations of state) to form a single NS; on the other hand, the total angular momentum is probably too large to be swallowed immediately by a black hole. The expected outcome, after a few milliseconds, would therefore be a spinning black hole (BH), orbited by a torus of neutron-density matter.

Other types of progenitor have been suggested - e.g. a NS-BH merger, where the neutron star is tidally disrupted before being swallowed by the hole; the merger of a white dwarf with a black hole [WD-BH]; or a category labeled as hypernovae or collapsars, where the collapsing core is too massive to become a neutron star, but has too much angular momentum to collapse quietly into a black hole (as in a so called "failed supernova''). The details of the latter model are addressed by Woosley in his contribution. The simple point that I wish to stress, however, is that a BH plus debris torus is a common ingredient of all these models; moreover the overall energetics of these various progenitors differ by at most an order of magnitude, the spread reflecting the differing spin energy in the hole and the different masses left behind in an orbiting torus. (There has been some confusion on this point in recent literature, through failure to appreciate that the dominant energy from a NS-NS event comes after a black hole forms, rather than during the precursor stage that [28, Narayan et al. 1992], discussed.) How might such a system generate relativistic outflow or a release of electromagnetic energy?