Confronting Synchrotron Shock and Inverse Comptonization Models with GRB Spectral Evolution

The time-resolved spectra of gamma-ray bursts (GRBs) remain in conflict with many proposed models for these events. After proving that most of the bursts in our sample show evidence for spectral "shape-shifting", we discuss what restrictions that BATSE time-resolved burst spectra place on current models. We find that the synchrotron shock model does not allow for the steep low-energy spectral slope observed in many bursts, including GRB 970111. We also determine that saturated Comptonization with only Thomson thinning fails to explain the observed rise and fall of the low-energy spectral slope seen in GRB 970111 and other bursts. This implies that saturated Comptonization models must include some mechanism which can cause the Thomson depth to increase intially in pulses.