G-13

Gamma-Ray Burst Spectroscopy

Robert D. Preece
(University of Alabama in Huntsville)

Recent findings from the continuum spectroscopy analyses of gamma-ray bursts can be used to constrain proposed emission models. All the observed richness of behavior comes despite the simplicity of the canonical four-parameter functional form that is so successful for fitting spectra in the BATSE energy band. This has traditionally been cast in the form of the `GRB' function, but other, similar spectral models also work, as long as they consist of a spectral break energy and two limiting power-law components at high and low energies. The distributions of time-resolved fitted spectral parameters, now available for bright bursts in the first BATSE Spectroscopy Catalog, are the most revealing. One can use the width of the distribution of low-energy indices greater than the ``death line'' of -2/3 to rule out the synchrotron shock model for a considerable fraction of the total sample. The peak value of -1should also be accounted for. Any viable model for GRBs must explain the observation of the narrowness of the break energy distribution. While there is room for correction of the shape of this distribution to account for the BATSE response, the main result stands: one half of all spectra fall within a factor of three in energy, while 90% are confined within a factor of seven. The situation is worsened when one considers that factors such as the observed red-shifts for several events, the large Lorentz factors required to avoid pair-production saturation and the observed spectral evolution within each burst all contribute toward widening the observed distribution. With additional data, at lower and higher energies, we can look for departures from the low-and high energy power law portions of the spectra. At lower energies, observed by Ginga, Sax and the BATSE SDs, there is growing evidence for a second spectral break. I will review these and other issues pertaining to the current status of GRB spectroscopy, with an emphasis on observations that must be accounted for by theory.