8 Conclusions and prospects

Simple blast-wave models seem able to accommodate the present data on afterglows. However we can at present only infer the energy per solid angle; there are reasons to suspect that the afterglow is not too narrowly beamed; on the other hand the constraints on the angle-integrated $\gamma$-ray energy are not strong - these could be concentrated in a high-Lorentz-factor beam only a few degrees across. As regards the trigger, there are two key questions. First, does it involve a black hole orbited by a dense torus (which I've advocated as a "best buy'')? Second, if so, can we decide between the various alternative ways of forming it: NS-NS, NS-BH or hypernova?

The locations should help to settle the second question. This is because a hypernova would be expected to lie in a region of recent star formation; on the other hand, a neutron star binary could take hundreds of millions of years to spiral together, and could by then (epecially if given a kick velocity on formation) have moved many kiloparsecs from its point of origin [3, (Bloom et al. 1998)]. There is also already tentative evidence that some detected afterglows arise in relatively dense gaseous environments - e.g. by evidence for dust in GRB 970508 [40, (Reichart 1998)] and the absence of an optical afterglow and strong soft X-ray absorption in GRB 970828 [12, (Groot et al. 1997]; [27, Murakami et al. 1997)]. On the other hand, fits to the observational data on GRB 970508 and GRB 971214 suggest external densities in the range of 0.04-0.4 cm^-1, which would be more typical of a tenuous interstellar medium [56, (Wijers & Galama 1998)].

We must also remain aware of other possibilities. For instance, we may be wrong in supposing that the central object becomes dormant after the gamma-ray burst itself. It could be that the accretion-induced collapse of a white dwarf, or (for some equations of state) the merger of two neutron stars, could give rise to a rapidly-spinning pulsar, temporarily stabilised by rapid rotation. The afterglow could then, at least in part, be due to a pulsar's continuing power output (cf. [49, Usov 1994)]. It could also be that mergers of unequal mass neutron stars, or neutron stars with other compact companions, lead to the delayed formation of a black hole. Such events might also lead to repeating episodes of accretion and orbit separation, or to the eventual explosion of a neutron star which has dropped below the critical mass, all of which would provide a longer time scale, episodic energy output.

And there could be more subclasses of classical GRB than just short ones and long ones. For instance, GRB with no high energy pulses (NHE) appear to have a different (but still isotropic) spatial distribution from those with high energy (HE) pulses [34, (Pendleton et al. 1996)]. Some caution is needed in interpreting this, since selection effects could lead to a bias against detecting HE emission in dim bursts [30, (Norris 1998)]. Then, there is the apparent coincidence of GRB 980425 with the SN Ib/Ic 1998bw [8, (Galama et al. 1998)]. Much progress has been made in understanding how gamma-rays can arise in fireballs produced by brief events depositing a large amount of energy in a small volume, and in deriving the generic properties of the long wavelength afterglows that follow from this. There still remain a number of mysteries, especially concerning the identity of their progenitors, the nature of the triggering mechanism, the transport of the energy and the time scales involved.

Gamma-ray bursts, even if we do not understand them, may still be useful as powerful beacons for probing the high redshift (z > 5) universe. Even if their total energy is reduced by beaming to a "modest" $\sim 10^{52}-10^{52.5}$ ergs in photons, they are the most extreme phenomena that we know about in high energy astrophysics. The modeling of the burst itself - the trigger, the formation of the ultrarelativistic outflow, and the radiation processes - is a formidable challenge to theorists and to computational techniques. It is, also, a formidable challenge for observers, in their quest for detecting minute details in extremely faint and distant sources. And if the class of models that we have advocated here turns out to be irrelevant, the explanation of gamma-ray bursts will surely turn out to be even more remarkable and fascinating.

Acknowledgements

I am especially grateful to Peter Mészáros and Ralph Wijers for extended collaboration, and to Josh Bloom and Stan Woosley for discussions. This research has been supported by the Royal Society.