The Radio After-Glows of Gamma-ray Bursts

The unique status of the supernova explosion has now been challenged by a more energetic, bizarre, and still rather mysterious event, which has been observed for a quarter of a century, though its true distance only became apparent in 1997: the gamma-ray burst (Fishman and Meegan 1995). Discovered (accidentally) by military satellites looking for high-energy emission from nuclear tests (and, like radio emission from the sun 30 years before, first declared top secret), gamma-ray bursts (GRBs) are the brightest objects in the sky outside of the solar system. With a peak magnitude-equivalent at MeV to GeV energies brighter than $v=-\rm 6^m$ (for a few seconds), they would outshine the brightest stars if we could ``see" at such high energies.

Whatever their exact nature, the short timescales involved (with some features lasting <1 ms) betray an origin in a tiny region of space, and most of the models proposed involve neutron stars, in particular the merger of a neutron star pair. In 1997, a concerted effort by optical, gamma- and X-ray astronomers resulted in the first detection of a transient optical (van Paradijs et al. 1997) and X-ray (Costa et al. 1997) event associated with the burst GRB 970228. The key to success was rapid follow-up of the gamma-ray burst with a high-definition X-ray telescope (the BeppoSAX satellite), and the equally quick determination of a position to within a few arc minutes for immediate optical imaging. Optical transient events have now been observed from a handful of GRBs, all have been faint (v>20), they may lie in distant galaxies, and in one case, GRB 970508, absorption lines indicate a redshift of z>0.83 (Metzger et al. 1997). A weak radio transient was also observed in GRB 970508 (Frail et al. 1997), with peak flux densities of about 1 mJy. Unlike the X-ray and optical transients, which decayed on timescales of hours or days, the radio event lasted for weeks, turning on later at longer wavelengths. Besides its general behavior, rapid intensity variations attributed to scintillation set interesting limits to the GRB size. It's also possible that many GRB events will be detectable in the radio even when the optical emission is highly obscured by dust as seems to be indicated in the case of GRB 980703 shown in Fig. 2.30.

 

Radio observations provide critical and unique elements of the physics of gamma-ray bursts.

1.

Rapid Subarcsecond Localization. At the present time the position of the gamma-ray and X-ray transient is only localized to arcminute accuracy. The role of providing sub-arcsecond positions of the GRBs relative to their host galaxies is the role of ground-based radio and optical instruments. Drawing upon historical analogy from the study of supernovae, we note that the location of GRBs within their hosts will provide important clues to one of the key unanswered questions: the identity of the GRB progenitors. For the moment, the detectionand morphology of the host galaxies and determining their star formation rates is exclusively an optical endeavour. With improvements in sensitivity the radio synchrotron from the host galaxy can be detected, allowing unique studies to be made of optically obscurred galaxies (see below).

2.

Dusty Hosts. The radio afterglow can penetrate the dust within the host galaxy, and thus GRBs can be seen both from unobscured and dusty galaxies. There is mounting evidence that at least one third of all GRBs occur in dusty environments (Ramaprakash et al. 1998, Reichart et al. 1998). For example, GRB980329, with an R-K color of 5, might have escaped detection were it not for the initial VLA localization. These findings are especially relevant for distinguishing GRB progenitor models since GRBs naturally arise in dusty regions in the ``microquasar'' model (Paczyski 1998) but not in the neutron star coalescence model (e.g. Bloom et al. 1998b).

Furthermore, if GRBs trace the star formation history of the early Universe then the radio (and X-ray) detection rates relative to the optical detections offer insight into one of the key problems in observational cosmology - the relative amounts of unobscured and obscured star formation at large redshifts.

3.

The Physics of Fireballs. The fireball model has been shown to be correct to first order but more detailed multi-wavelength studies are needed. By combining the radio data with that from optical, infrared and X-ray instruments we derive the instantaneous spectrum of the afterglow (Figure 2.30). Knowing the spectral shape it becomes possible to infer the total energetics, the ambient density and to probe the physics of relativistic shocks, addressing issues such as the degree of equipartition between the magnetic field and particle energy density and the efficiency of particle acceleration.

4.

Expansion and Geometry of the Fireball. Radio observations provide unique diagnostics on the size and geometry of the fireballs. This is possible thanks to the fact that the boundary between strong and weak scattering in the ISM is around 3 $\mu$as at centimeter wavelengths, and this is comparable to the angular size of young fireballs (Narayan 1992). At later times when the scintillation is quenched (due to source expansion) we see the underlying light curve (and spectrum) whose behavior can be used to track the transition from $\Gamma>>1$ to $\Gamma\sim{1}$ (Waxman, Kulkarni & Frail 1998).

5.

New Phenomenology. It is tempting to assume that the GRB phenomenon is homogeneous but historically such simplistic assumptions have been shown to be erroneous. This is already apparent with the (Galactic) soft gamma-ray repeaters and at least one supernova SN1998bw (z=0.0085), which likely produced a burst of gamma-rays GRB 980425 (Galama et al. 1998, Kulkarni et al. 1998). Synchrotron emission at radio wavelengths will continue to be an important means to study these high energy transients discovered through their X-ray and gamma-ray emissions.

The effect which the superior sensitivity of the SKA would have on observations of objects like GRB980703 is easy to imagine: the emission would be detected at an earlier stage, the ``turn on" at several frequencies could be studied in greater detail, variations could be measured more accurately to see how well they conform to scintillation behavior, and it would be possible to detect such an event out to a redshift of 4 or more. In addition, of course, radio emission from many more GRBs would be detected, but in view of the limited statistics at the moment (one detection out of perhaps five well-observed events), it is difficult to estimate their numbers. We can, however, get some idea by considering the likely impact of the SKA upon radio observations of supernovae (SNe).

With the SKA sensitivity, we will be able to extend such work to many tens of GRBs, and for any plausible luminosity function there should be hundreds more radio detections, since current detections must represent only the tip of the iceberg. Detections can be expected to moderately high redshifts (z=4 to 5), and if GRBs occur in gas-rich galaxies, it should be possible to detect HIin absorption. This would provide radio limits to the source distance, as well as another probe of conditions along the line of sight to the GRB. This emission results, like the optical afterglow, from the interaction of burst-induced shocks with the ambient medium, and is thus peripheral to the gamma-ray event itself. There are, however, predictions, based upon rather general arguments, of prompt radio emission from the very burst. The detection of such emission, which is expected to be many times weaker than that observed from the afterglow, would contribute directly to our understanding of the burst process itself, and will only be possible with radio telescopes orders of magnitude more sensitive than what we now have.

It has taken us almost a quarter of a century to come to grips with the gamma-ray burst problem. Now that we have pinned down the essentials required for their detection in other wavebands, and we have the tools for obtaining accurate positions quickly, further breakthroughs are likely to come sooner rather than later. Nonetheless, it is clear that except for the intense gamma-radiation itself, observations of the associated phenomena require radio (and optical) telescopes of the greatest sensitivity and agility. Our understanding of these mysterious objects by the time a SKA sees first light will no doubt be better than it is now, and we may have even singled out their progenitors. Nonetheless, more sensitive radio observations will enable us to probe their environment, observe those enshrouded in dust, determine their distance and size, and detect the very low energy tail of the burst itself.