8.8. The Internal - External Scenario
Internal shocks can convert only a fraction of the total energy to radiation [32, 33, 69]. After the flow has produced a GRB via internal shocks it will interact via an external shock with the surrounding medium . This will produce the afterglow - a signal that will follow the GRB. The idea of an afterglow in other wavelengths was suggested earlier [17, 18, 21] but it was suggested as a follow up of the, then standard, external shock scenario. In this case the afterglow would have been a direct continuation of the GRB activity and its properties would have scaled directly to the properties of the GRB.
According to internal-external models (internal shocks for the GRB and external shocks for the afterglow) different mechanisms produce the GRB and the afterglow. Therefore the afterglow should not be scaled directly to the properties of the GRB. This was in fact seen in the recent afterglow observations [25, 26]. In all models of external shocks the observed time satisfy t R / e2 and the typical frequency satisfies e4. Since most of the emission takes place at practically the same radius and all that we see is the variation of the Lorentz factor we expect quite generally : t2 ± . The small parameter reflects the variation of the radius and it depends on the specific assumptions made in the model. We would expect that tx / t ~ 5 and topt / t ~ 300. The observations of GRB970508 show that (topt / t)observed 104. This is in a clear disagreement with the single external shock model for both the GRB and the afterglow.
Under quite general conditions the initial typical synchrotron energy for either the forward or the reverse external shock may fall in the soft GRB band. In this case the initial stage of the afterglow might overlap the -ray emission from the internal shock . The result will be superposition of a rapidly varying signal on top of a long smooth and softening pulse. This possibility should be explored in greater detail.