Abstract. Before 2004, nearly all GRB afterglow data could be understood in the context of the external shocks model. This situation has changed in the past two years, when it became clear that some afterglow components should be attributed to the activity of the central engine; i.e., the central engine afterglow. We review here the afterglow emission that is directly related to the GRB central engine. Such an interpretation proposed by Katz, Piran & Sari, peculiar in pre-Swift era, has become generally accepted now.Keywords: Gamma-ray: bursts-radiation mechanisms: nonthermal PACS: 98.70.Rz
TWO KINDS OF GRB AFTERGLOWSIn the context of standard fireball model of Gamma-ray Bursts (GRBs), the prompt γ−ray emission is powered by internal shocks and the afterglow emission arises due to external shocks (see [21] for a review). In the pre-Swift era, most of the afterglow data was collected hours after the prompt γ−ray emission. These data was found to be consistent with the external forward shock model, though at times energy injection, a wind medium profile, or a structured/patchy jet were needed. We call the emission relevant to the external shocks generated by the GRB remnant as the "fireball afterglow" or "afterglow". An alternative possibility for the production of the afterglow is a continued activity of the central engine (i.e., the central engine afterglow) via either the "internal shocks" or "magnetic dissipation". This idea has been put forward by Katz, Piran & Sari already in 1998, shortly after the discovery of the afterglow of GRB 970228. However, the agreement of the predictions of the external shock afterglow model [24] with most subsequent multi-wavelength afterglows observation strongly disfavors the central engine afterglow interpretation.One disadvantage of the central engine afterglow model is its lack of predictive power. The fireball afterglow model, instead, has predicted smooth light curves and in particular the intrinsic relation between the flux in different bands (e.g., [24]) as well as between the spectral slops and the temporal decay. In 2003, it was already clear that in the case of a fireball afterglow, the variability timescale of the emission δt divided by the occurrence time t has to be in order of 1 or larger [19]. A highly relevant constraint is that the decline of the fireball afterglow emission can not be steeper than t −(2+β ) , (where β is the spectral index) unless the edge of the GRB ejecta is visible. This is because the GRB outflow is curving and emission from high latitude (relative to the observer) will reach us at later