Among the models proposed to account for the blazar -ray emission, the most accredited are those based on the Inverse Compton process, between the relativistic electrons responsible for the synchrotron emission and seed photons which can either be the synchrotron photons themselves [synchrotron-self Compton (SSC)] or photons produced externally to the jet (EC) (Maraschi et al. 1992; Bloom & Marscher 1993; Sikora et al. 1994, Dermer et al. 1992; Blandford & Levinson 1995; Ghisellini & Madau 1996).
In all cases synchrotron radiation is responsible for the low energy emission (up to the optical-UV band), and the inverse Compton process gives rise to the X- to -ray spectrum. Seed photons of different origin may well coexist and their relative contribution in the same object may be different in different states and in different energy bands. The presence of luminous emission lines and possibly of a blue bump indicates, in the case of PKS 0528+134, that photons outside the jet should play an important role in the IC process (EC scenario). For a detailed discussion of the fitting of the 1994 overall spectrum with SSC and EC models see Sambruna et al. (1997).
As shown in Fig. 6, the source shows correlated X-ray and -ray flux variability. In particular, during the 1997 campaign, the source was at its faintest level both in X- and -rays. In this occasion, the X-ray spectrum was significantly flatter than in higher states. This is in contrast with the "flatter when brighter" behavior of blazars (Ulrich et al. 1997), and therefore likely to give important constraints on the emission models.
Within the framework of the SSC and EC model we can envisage two scenarios to explain the observed behavior:
1) The X-rays are produced by both the self-Compton and EC processes. Therefore there can be two typical frequencies of the seed photons: one corresponding to the peak of the synchrotron emission (in the far IR), and one corresponding to the external radiation. In the case that the emission lines and blue bump photons form the bulk of the external radiation, their typical frequency, as observed in the comoving frame, is in the far UV. If the radiation energy densities of these two components are comparable (within a factor 10), then the self-Compton emission would dominate at lower X-ray energies, while the EC spectrum would be entirely responsible for the emission in the -ray band. This is because the self-Compton spectrum is somewhat steeper than the EC one produced by the same electrons, and because the maximum self-Compton frequency is lower. Let us suppose now that the number of emitting electrons increases. In this case the synchrotron and the EC fluxes vary linearly, while the self-Compton one varies quadratically, making the SSC component to dominate the flux over a larger X-ray energy range.
We have reproduced the 1997 "low state" and the 1995 "high state" spectra along these lines, with a homogeneous one-zone model, taking into account both the SSC and the EC contributions to the high energy spectra. The model finds the equilibrium (steady state) electron distribution by balancing the injection of relativistic particles and their radiative cooling. Photon-photon collisions and electron-positron pair production, Klein-Nishina effects on the scattering cross section and Coulomb collisions are taken into account (see Sambruna et al. 1997 for further details). The input parameters are given in the caption of Fig. 6. We "fitted" both spectra leaving the dimension and the beaming factor of the source unchanged, as well the low energy cut-off of the injected electron distribution. We have instead increased by a factor 3 the injected power in the high state, also characterized by a slightly flatter injected electron distribution. The resulting spectra, shown in Fig. 6, reproduce quite well the different X and -ray slopes.
2) Another process that can in principle account for a flatter X-ray spectral index when the source is fainter, is e pair production. When the source is in a high (and hard) -ray state, a small fraction of the high energy power could be converted in electron-positron pairs via interaction with X-ray photons. This process produces a steepening of the electron spectrum towards lower energies, yielding a brighter and steeper IC spectrum in the X-ray band. The main difficulty with this scenario is the fine tuning required in the amount of the high energy power that is absorbed and reprocessed, which must be of the order of a few per cent. If it is less, then pairs are not produced in a sufficient amount to contribute anywhere in the emitted spectrum. If it is higher, then they could increase the number of target photons for - collisions, inducing a catastrophic cascade, and an overproduction of X-ray with respect to what we observe (Svensson 1987).
For these reasons, we conclude that the required moderate pair reprocessing is unlikely to occur in this source.
© European Southern Observatory (ESO) 1999
Online publication: July 16, 1999