We can now address the main question of this paper: why have blazars been detected by EGRET? There are at least three reasons, which have to do with the fact that blazars are characterized by:
Point number one is obvious. We know that in blazars synchrotron emission reaches at least the infrared/optical range, which reveals the presence of high-energy electrons which can produce -rays via inverse Compton emission (although hadronic processes can also be important or even dominant; Mannheim 1993). Very recent BeppoSAX observations have shown that synchrotron emission can actually reach the X-ray band, namely around 10 keV for 1ES 2344+514 (Giommi et al. 1998) and 100 keV for MKN 501 (Pian et al. 1998). Pian et al. fit a synchrotron-self-Compton (SSC) model to the multifrequency spectrum of MKN 501 and infer a maximum value for the Lorentz factor of the electrons max ~ 3 x 106. With such high values of max, Pian et al. have been able to reproduce the observed TeV flux of this source (see Sect. 4) with an SSC model. Point number two is vital, as described in the previous section, not only to enable the -ray photons to escape from the source, but also to amplify the flux and therefore make the source more easily detectable. Point number three is also very important. -ray emission is clearly non-thermal (although we still do not know for sure which processes are responsible for it) and therefore related to the jet component. The stronger the jet component, the stronger the -ray flux.
Having understood why blazars have been detected by EGRET, one could also ask: why have not all blazars been detected? Many blazars with radio properties similar to those of the detected sources, in fact, still have only upper limits in the EGRET band. This problem has been addressed, for example, by von Montigny et al. (1995) who suggest as possible solutions variability (only objects flaring in the -ray band can be detected) or a -ray beaming cone which either points in a different direction or is more narrowly beamed than the radio one (see also Salamon and Stecker 1994 and Dermer 1995). These can certainly be viable explanations, but one should also note that even a moderate dispersion in the values of the parameters required for -ray emission described above (particle energy, Doppler factor, and non-thermal component strength) can easily imply the non-detection of some sources and the detection of others with similar radio properties.
Do the points discussed above also explain why EGRET has not detected any of the more numerous radio-quiet AGN? Yes, although not all of them might be essential in this case. As radio-emission (at least in radio-loud AGN) is certainly non-thermal, while the optical/ultraviolet emission might be thermal emission associated with the accretion disk (at least in radio-quiet AGN), then the ratio Lr / Lopt could actually be related to the ratio Lnon-thermal / Lthermal. Furthermore, L seems to scale with Lr, although the details of this dependence are still under debate (see e.g., Mattox et al. 1997 and references therein). It could then be that even radio-quiet AGN are -ray emitters, although scaled down by their ratio between radio and optical powers, that is at a level 3-4 orders of magnitude below that of blazars, with typical fluxes F -10 photons cm-2 s-1. In other words, radio-quiet AGN would fulfill requirements number one and two (2) but not number three.
Alternatively, it could be that for some reason the emission mechanisms at work in radio-loud sources are simply not present in the radio-quiet ones, either because there is no jet at all in radio-quiet AGN or because, for example, there is no accelerating mechanism. In this case, either condition number one or number three (or both) would be missing (number two would now be unimportant), and no -ray emission would be expected.
Unfortunately, it might not be possible to test these two alternatives on the
basis of -ray data
for some time: even in the first case, in fact, the
fluxes are below the sensitivity of currently planned
missions, like GLAST
If radio-quiet AGN
however, GLAST might be able to detect some of the nearest sources.
2 One could argue that the relativistic beaming requirement would probably not be very important in radio-quiet AGN as the luminosity/radius ratio in the -ray band in these sources would be much lower anyway and -ray photons would escape even without beaming. However, GeV photons collide preferentially with X-ray photons, which are plentiful in these sources: some beaming might then be required for the (putative) -ray emission in radio-quiet AGN as well.