2 Observational strategy

After a gamma-ray burst position has been determined by BeppoSAX or RXTE to an accuracy of about five arcminutes, AXAF can be pointed to the position within less than a day and perhaps within eight hours. Even for the most pessimistic predictions of Panaitescu et al. (1998), a 10 ksec exposure should easily detect the burst (see Fig. 1).

  

Figure 1: The sensitivity of ACIS for detecting GRB afterglows in 104 s. This limit scales as t-1 up to 105 s

For typical X-ray afterglow intensities observed within one day after the gamma-ray event, about 10 ksec of observing time should collect about 10³ counts for spectral analysis. This should be enough to determine spectral slope and absorption cut-off for the burst.

Spectral features from the surrounding medium of the burst are very difficult to predict, since the environment of the preburst object is unknown. Mészáros & Rees (1998) have given some estimates of iron line emission from a variety of environments. They estimate that the flux in the line would be of the order of $3\ 10^{-16}~{\rm ergs/cm}^2/{\rm s}$ for a unit density of the surrounding medium. If the medium were more dense, as in an HII region, then one could get more than an order of magnitude increase in flux. The process that creates the iron lines needs to be examined in some detail. The gamma-ray burst is so intense within about a parsec of the source that it creates what in laser physics is called "self-induced transparency''. In the case of the gamma-ray burst the photon flux is so intense that all of the atoms are stripped of their electrons before the burst photons completely pass a given point in space. Liang & Kargatis (1994) noted this effect. The other point is that the cross section of the atomic electrons decreases rapidly with increasing energy resulting in the removal of the electrons from the outside of the atom toward the inside which reduces the intensity of fluorescent K photons by about a factor of twenty. The electrons from the atoms are ejected as photoelectrons or Compton recoil electrons and are not available to emit fluorescent photons. By formulating the fluorescent flux in terms of the time after the burst, it can be shown that the flux scales as t² for the time after the burst, so that the chance of detecting any fluorescent photons increases with time up to the point where the burst wave escapes from the local medium or the photon flux in the burst falls below the point where it ionizes all of the atoms in the surrounding medium. The decrease in

absorption of the burst spectrum with time through the burst is a good indication that self-induced transparency is playing a role in the burst spectral evolution. To observe the fluorescent photons with the greatest probability, it is best to choose a burst that appears to be within a galaxy, that shows strong spectral evolution, and then wait for a number of months to make the observation. Even so, it will require a very long exposure, $\sim 10^5$ s.