|| © CAMBRIDGE UNIVERSITY PRESS 1991
3.3. Interactions between jets and the ambient nuclear medium
Because of the huge powers carried by jets they must have significant impacts on the material through which they move at all length scales, many of which were discussed in previous chapters. Here we shall concentrate on several ways in which jets may interact with the surrounding nuclear plasma, other than those already alluded to in our discussion of beam formation.
It has been proposed that essentially all aspects of the BLR are produced by jets (Norman & Miley 1984; Barthel et al. 1984). The pressure exerted by a jet can be estimated as
which is comparable to that necessary to confine putative broad line clouds for a jet of solid angle ~ 10-2 steradians. In this picture initially uniform material would be swept up by the jet into a cocoon which could be unstable to Rayleigh-Taylor type fragmentation. Those fragments could produce many (or all) of the BL clouds if the cloud densities are ~ 1010 cm-3. Further, much of the jet's mechanical energy would be dissipated if interactions with the surrounding medium produce shocks, and this could yield the majority of the local UV and X-ray emission necessary to excite the clouds. In this scenario the cocoon would further function as a natural screen, thereby providing a possible explanation of superluminal motions. One-sided jets could be caused by one-sided blocking of a jet by massive molecular clouds (Norman & Miley 1984). Recent indications of the necessity for higher density BL clouds seem to make this rather extreme picture less attractive, as has the growing list of superluminal sources.
Still, this interaction may be responsible both for "stopping" superluminal components, as in BL Lac, and the rather small overall radio extent of most high redshift quasars (Barthel 1986). Further, the small core polarizations revealed by VLBI (Wardle et al. 1986) may imply the need for an extended depolarizing screen, which could be provided by this material.
Looking at this question from an essentially opposite point of view it can be shown to be very unlikely that the intercloud medium is capable of thermally confining jets at scales much below one parsec, although at larger distances this is certainly possible (Rees 1984b); considering the recent conclusions that the BLR is smaller and the cloud densities higher than typically assumed, this argument is only strengthened. Once out in the line emitting region the jets are unlikely to be able to destroy most clouds passing through them; rather, if the cloud density were high enough the jet could be completely dissipated in the course of its interaction with them (Rees 1984b). Of course, one possible explanation for the knotty structure in VLBI scale jets is the interaction of jets with ambient clouds (e.g., Blandford & Königl 1979), while another would involve shocks propagating down the jet, presumably due to fluctuations in power or velocity (cf. Chapter 4). This is another area where the improved resolution of a RADIOASTRON could be very useful, for the impact of jets upon broad line clouds would probably lead to sharp edges facing the nucleus, while internal shocks would more likely lead to sharper leading edges.
Rees (1984b) has also stressed that flows on sub-pc scales are likely to be much more dissipative than those on larger scales, primarily because the ratio of the radiative cooling time ( d2) to the dynamical time ( d ) scales with d, rather than being independent of distance as is true for larger jets. These inner-most jets are likely to lose their internal pressure more easily and should waste more energy on being bent in comparison to their larger scale incarnations. Unfortunately, the microarcsecond optical resolution needed to see the visible synchrotron radiation emitted through these interactions, similar to that frequently observed on kpc scales (e.g., van Breugel et al. 1984), is unlikely to be achieved.
The possibility that jets passing through the ambient medium will drive large circulation currents that in turn act back upon the jets has been examined from several viewpoints. Fluid flowing along a beam will interact with the surrounding medium through surface instabilities which can provide an effective viscous coupling between the two (cf. Henriksen 1985). The external fluid might be drawn up along the jets and, if so, is likely eventually to separate from them and then descend at greater distances; the circulation would be completed by flowing towards the jets near the equatorial plane. Small asymmetries in these external fluid flow patterns can then significantly alter the density and pressure of the surrounding gas near the equatorial region in such a way as to affect the jets. In general, any asymmetry in jet strength will be amplified, and this process can even cause one of the pair of jets to choke off: this is the so called "clam-shell" mechanism (Icke 1983).
This idea has been modified to stress the role of material entrained by the jets (Allen 1986). If a blob of material is swallowed and dispersed by one jet the net flow speed in that channel is reduced, since the momentum is shared by a larger mass. Allen argues that Kelvin-Helmholtz perturbations will grow more easily in the denser and slower jet. If so, that implies that yet more material is likely to be engulfed on that side, amplifying the asymmetry; further, this increase in entrained matter would increase the radio emission from that choked channel. Such instabilities might act on a wide range of scales and could plausibly explain one-sided sources without the need for Doppler favouritism. If one side of the jet remains relativistic, while the other engulfs enough matter to become subrelativistic, a broader and more diffuse radio structure should emerge from the subrelativistic, choked side.
There are a number of radio sources which have roughly equidistant lobes but exhibit detailed anti-symmetry on the two sides of the central galaxy. It has been argued that such sources are activated in alternate directions, with a single jet "flipping" from one side to another (Rudnick & Edgar 1984; Rudnick 1985). The "clam-shell" type of mechanisms provide a conceivable way of explaining these "flip-flop" sources, although exactly how the switch from one direction to the other might occur with some quasi-fixed periodicity is unclear. Models involving the orbit of the powerhouse within the surrounding medium might also suffice to explain some "flip-flop" sources (Wiita & Siah 1981; Saikia & Wiita 1982), in that the jet that has to propagate through a shorter path-length in the confining medium could escape while the one moving through more material might be temporarily blocked, or much reduced in strength. Half an orbit later, the jet would be expected to escape in the opposite direction and would be stifled on the originally powerful side. However, before spending a great deal of time worrying about this problem, it should be noted that statistical evidence indicates that such candidates for "flip-flop" sources are probably not very common (Ensman & Ulvestad 1984).