Gamma-ray bursts are the largest electromagnetic explosions known to happen in the Universe and are associated with the collapse of stellar progenitors into blackholes. After an energetic prompt emission phase, lasting typically less than a minute and emitted in the gamma-rays, a long-lived afterglow phase starts. During this phase strong emission is observed at longer wavelengths, e.g., in the X-ray and optical bands. This phase can last several weeks and carries important information about the energetics and structure of the burst as well as about  the circumburst medium (CBM) and its density profile. The standard afterglow model includes a single emission component which comes from synchrotron emission in a blast wave moving into the CBM. Additional factors that could give observable features include prolonged energy injection from the central engine, effects of the jet geometry, and viewing angle effects, which thus constitute an extended standard model.

In this thesis, I study the afterglow emission in a global approach by analysing large samples of bursts in search for general trends and characteristics. In paper I, I compare the light curves in the X-rays and in the optical bands in a sample of 87 bursts. I find that 62% are consistent with the standard afterglow model. Among these, only 9 cases have a pure single power law flux decay in all bands, and are therefore fully described by the model within the observed time window. Including the additional factors described above, I find that 91% are consistent with the extended standard model. An interesting finding is that in nearly half of all cases the plateau phase (energy injection phase) changes directly into the jet decay phase.  In paper II, I study the afterglow by analysing the temporal evolution of color indices (CI), defined as the magnitude difference between two filters. They can be used to study the energy spectrum with a good temporal resolution, even when high-resolution spectra are not available. I find that a majority of the CI do not vary with time, which means that the spectral slope does not change, even between different emission episodes. For the other cases, the variation is found to occur during limited periods. We suggest that they are due to the cooling frequency passing over the observed filter bands and, in other cases, due to the emergence of the underlying supernova emission. In paper III, I study the energetics of the GRBs that can be inferred from the afterglow observations. Using this information I analyse the limits it sets on what the central engine can be, if it is a magnetar or a spinning black hole. Assuming that the magnetar energy is emitted isotropically, I find that most bursts are consistent with a BH central engine and only around 20% are consistent with a magnetar central engine. As a consistency check, we derive the rotational energy and the spin period of the blackhole sample and the initial spin period and surface polar cap magnetic field for the magnetar sample and find them to be consistent to the expected values. We find that 4 of 5 of the short burst belong to the magnetar sample which supports the hypothesis that short GRB come from neutron star mergers.