9 Star-forming regions

Star forming regions consist of a cluster of O/B stars that lie in and around a clumpy cloud of dust and gas. We expect $A_V \gg 1$ for O/B stars embedded in the cloud, and $A_V \approx 0$ for O/B stars that have drifted out of the cloud and/or lie near the surface of the cloud and have expelled the gas and dust in their vicinity. Thus the optical/UV spectrum of star forming regions is a sum of the spectra of many hot (blue) stars, some of which are embedded in the cloud, and therefore heavily extinguished, and some of which lie on the surface or around the cloud, and are therefore essentially un-extinguished. This composite spectrum is rather blue, and yields a value $A_V^{\rm eff} \approx 1$ when a single extinction curve is fitted to it.

The situation is very different when we consider an individual line-of-sight, as is appropriate for the afterglow of a GRB. If the GRB source lies outside and far away from any star-forming region, we expect ; if the GRB source lies outside but near a star-forming region, we expect about half the time and $A_V^{\rm afterglow} \gg 1$ about half the time. Finally, if the GRB source is embedded in the star-forming region, we expect $A_V^{\rm afterglow} \gg 1$.

Thus, if GRB sources actually lie in star-forming regions, one would expect $A_V^{\rm afterglow} \gg 1$ (values of A_V $\sim 10-30$ are not uncommon for dense, cool molecular clouds in the Galaxy). Is this consistent with what we see? No. However, this may not mean that GRB sources do not lie in star-forming regions. The reason is that the soft X-rays and the UV radiation from the GRB and its afterglow are capable, during the burst and immediately afterward, of vaporizing all of the dust in their path (Lamb & Reichart 1999b). Thus the value of $A_V^{\rm afterglow}$ that we measure may have nothing to do with the pre-existing value of the extinction through the star-forming region in which the burst source is embedded, but may instead reflect merely the extinction due to dust and gas in the disk of the host galaxy.

The GRB, and its soft X-ray and UV afterglow, are also capable of ionizing gas in any envelope material expelled by the progenitor of the burst source and in the interstellar medium of the host galaxy. This will produce Strömgren spheres or very narrow cones (if the burst and its afterglow are beamed) in hydrogen, helium and various metals (Bisnovatyi-Kogan & Timokhin 1998; Timokhin & Bisnovatyi-Kogan 1999; Mészáros 1999). Recombination of the ionized hydrogen eventually produces intense [CII], [CIV], [OVI] and [CIII] emission lines in the UV, and intense H${\alpha}$ and H$\beta$ emission lines in the optical. However, the line fluxes may still not be strong enough to be detectable at the large redshift distances of GRB host galaxies. Interaction of the GRB and its soft X-ray afterglow with any envelope material expelled by the progenitor of the burst source and with the surrounding interstellar medium can also produce intense fluorescent iron line emission (see, e.g., Mészáros 1999), but it is again difficult to see how the line flux could be large enough to be detectable or to explain the hints of a fluorescent iron emission line in the X-ray afterglows of GRB 980703 (Piro et al. 1999) and GRB 980828 (Yoshida et al. 1999).