next previous
Up: Testing for synchrotron self-absorption


1 Introduction


The multiwavelength fading counterparts of gamma-ray bursts (GRBs) have been shown to be in agreement with the relativistic blast wave model (Mészáros & Rees 1993). More predictive variations of this model, such as the synchrotron shock model (Katz 1994; Tavani 1996) are consistent with a small number of time-integrated GRB spectra (Cohen et al. 1997), but fail to explain several time-resolved GRB spectra. In particular, the asymptotic photon slope $\alpha$ ($F_{\rm N} \propto \nu^{\alpha}$)below the spectral break is predicted by the synchrotron shock model to be between $-\frac{2}{3}$ and $-\frac{3}{2}$. This was shown to be inconsistent with time-resolved GRB spectra fit with the Band et al. (1993) GRB function (Crider et al. 1997; Crider et al. 1998). Fitting both the Band GRB function and a broken power-law to over 100 bursts, Preece et al. (1998) found roughly a fourth of time-integrated spectra to be inconsistent with the synchrotron shock predictions, as well.

The observed high values of $\alpha$do not easily differentiate between the many possible absorption mechanisms. However, the evolution of $\alpha$, when fitting spectra with the Band GRB function, may favor saturated Comptonization as the absorption mechanism (Crider et al. 1997). This may well be a result of extracting the photon spectra assuming a Band GRB function. In this paper, we fit the time-resolved BATSE LAD spectra of GRB 970111 directly with a self-absorbed synchrotron shock function. We choose this burst because it was very bright, it was seen by many instruments including BeppoSAX and BATSE (trigger 5773), and it had a very high $\alpha = +1.5 \pm 0.2$ (Crider et al. 1998).



next previous
Up: Testing for synchrotron self-absorption

Copyright The European Southern Observatory (ESO)