The central engine of Gamma Ray Bursts may live much longer than the duration of the prompt emission. Some evidence of it comes from the presence of strong precursors, post-cursors, and X-ray flares in a sizable fraction of bursts. Additional evidence comes from the fact that often the X-ray and the optical afterglow light curves do not track one another, suggesting that they are two different emission components. The typical steep-flat-steep behavior of the X-ray light curve can be explained if the same central engine responsible for the main prompt emission continues to be active for a long time, but with a decreasing power. The early X-ray afterglow emission is then the extension of the prompt emission, originating at approximately the same location, and is not due to forward shocks. If the bulk Lorentz factor Gamma is decreasing in time, the break ending the shallow phase can be explained, since at early times Gamma is large, and we see only a fraction of the emitting area. Later, when Gamma decreases, we see an increasing fraction of the emitting surface up to the time when Gamma ~ 1/theta_j. This time ends the shallow phase of the X-ray light curve. The origin of the late prompt emission can be the accretion of the fall-back material, with an accretion rate dot M proportional to t^(-5/3). The combination of this late prompt emission with the flux produced by the standard forward shock can explain the great diversity of the optical and the X-ray light curves.
Download