Annu. Rev. Astron. Astrophys. 1997. 35: 445-502 Copyright © 1997 by Annual Reviews. All rights reserved

8.5. The Invisible Jet Core

The EGRET detections of blazars demonstrate that the observed power is in many cases dominated by high-energy gamma rays. Although the degree of dominance could result in part from selection effects - i.e. if there is a scatter, possibly due to variability, in the intrinsic gamma-ray to radio flux ratio of the population, objects with the highest ratio will be "selected" by gamma-ray observations (Impey 1996) - it is clear that gamma rays are a fundamental component of blazar power.

Paradoxically, rapid variability at gamma-ray energies implies that the observed radiation cannot be produced too near the center or the high-energy photons would never escape (Section 6.5). This raises the question of how power is transported from the central engine, and how and why it is radiated at a given distance. In FSRQ the gamma-ray variability time scale is ~1 day, similar to that in the optical for some well-studied objects, so that the minimum radiative zone occurs near ~10¹⁶( / 10) cm. At roughly this scale, a magnetohydrodynamic wind from a rotating disk around a black hole would be collimated and accelerated to relativistic speed (Appl & Camenzind 1993); the energy transport would be magnetohydrodynamic and the radiative region would start "naturally" where a high bulk velocity is achieved.

Alternatively, the energy transport could be either purely electromagnetic, via a Poynting flux or through mildly relativistic ions with low internal entropy (Blandford 1993, Blandford & Levinson 1995). Assuming the emitting region is initially opaque to gamma rays, there will be a "gamma-ray photosphere" (due to interactions with lower energy photons) that is larger for higher energy gamma rays. The concept of a photosphere makes the emergent spectra independent of the details of the energy transport and conversion, and also allows a very strong prediction that gamma-ray flux variations are slower and/or later at higher energies. This prediction is violated for Mrk 421, where TeV gamma rays vary more rapidly and more frequently than GeV gamma rays. That the gamma-ray spectra of FSRQ are harder in brighter states also appears difficult to reconcile with this model.

Another way of transporting energy without dissipation is highly relativistic protons (_p ~ 10⁷) (Mannheim & Biermann 1992, Mannheim 1993, 1996, Protheroe & Biermann 1997). The associated primary electrons, whose break energy is much smaller than that of the protons, produce the low frequency spectral component via the synchrotron mechanism. The very high-energy protons interact with these photons, initiating pair cascades that yield very flat spectra ( 1) in the MeV-TeV range. This model predicts that high-energy variations should always follow low energy ones.

In an intermediate scenario, the jet starts out at high bulk Lorentz factor, _j ~ 10⁴, then decelerates through interactions with the local (disk) radiation field (Coppi et al 1993). This model has a difficulty in that the energy lost in the deceleration phase - i.e. the observed radiation - is much larger than that carried by the decelerated jet, which may therefore be insufficient to power the radio lobes.

All these models have difficulties, and in particular none of them yields a good fit to the overall spectral energy distribution. Nonetheless, they are of interest because they attempt to address crucial points about the origin of the radiating particles, which remain unexplained in the more phenomenological models.