The energy and momentum fluxes in FRII jets are
expected to be sufficient to drive a bow shock at supersonic
speed into the ambient medium (e.g.,
134).
Ambient gas crossing
the bow shock will be heated. For a shock advance speed relative to the
speed of light of vadv / c, the Mach number,
, in monatomic gas of
normal cosmic abundances with thermal energy kT in units of keV,
is given by
 |
(15) |
For a non-relativistic equation of state
(
= 5/3),
the jump conditions for a non-radiating shock (e.g.,
189)
find that pressure, density, and temperature ratios between gas that
has crossed the shock and the ambient medium are
 |
(16) |
 |
(17) |
 |
(18) |
where subscripts 2 and 1 refer to post-shock and pre-shock conditions, respectively. For high advance speed and large Mach number the density contrast reaches a factor of four, resulting in enhanced X-ray emissivity from shocked gas. The visibility in observations will depend on the relative volumes of shocked and unshocked gas along given lines of sight.
Complications apply in reality. Firstly, there is observational evidence that in supernova remnants the post-shock electrons are cooler than the ions (e.g., 109, 160). Secondly, a bow shock around a lobe is oblique away from its head, with a consequent change in the jump conditions and the emissivity contrast [210]. The closer a structure is to a spherical expansion, the more normal the shock will be everywhere and the better the applicability of the above equations.
ROSAT data revealed the presence of X-ray cavities coincident with the inner parts of the radio lobes of Cygnus A, and these were interpreted as due to the contrast between undisturbed ambient gas and gas around the lobes that had been heated in the past but has now expanded and cooled to a low emissivity [42], although the parameters of the shock are not effectively constrained by the data. More recent Chandra observations of Cygnus A find gas at the sides of the lobes to have kT ~ 6 keV, slightly hotter than the value of 5 keV from ambient medium at the same cluster radius, but the gas may have cooled after bow-shock heating, and again the data do not usefully constrain model parameters [188]. Evidence of strong shock heating around more distant FRII radio galaxies has yet to be seen.
CSS and GPS sources have been examined for evidence of shock heating. These are good places to look as the radio sources are generally considered to be in an early stage of expansion and they are overpressured with respect to even a cluster ambient medium (e.g., 184). The disadvantage is that source sizes are small so that even Chandra will have difficulty in separating emission from the nuclei, radio structures, and ambient medium from that of any shocked gas. The best evidence for detection of shocked gas thus arises from deep XMM-Newton spectroscopy, and in particular that of the CSS source 3C 303.1 [152]. The X-ray spectrum contains soft emission (associated with the ambient galaxy atmosphere) and a hard component. Since nuclear emission is undetected in the radio, it is reasonable to associate the hard emission with shocked gas, and a model can be constructed [152] that has an expansion velocity consistent with cooling-time arguments for optical emission-line gas [61].
7.2. Dynamics of FRIs in clusters
Low-power sources are closer and more amenable to detailed study, since the various components of X-ray emission are more easily separated. The medium plays an important rôle in the deceleration of the jets, which share momentum and energy with entrained material (Section 4).
The Einstein and ROSAT missions found evidence that the radio lobes of NGC 1275 have pushed Perseus-cluster gas aside (e.g., 28), and now many clusters and groups are found to harbour gas cavities containing radio plasma that originates from active galaxies (e.g., 23). Rather than expansion at high Mach number, the displacement of the gas appears normally to create low-density, rising bubbles in rough pressure balance with the surrounding medium (e.g., 50). NGC 1275, M 87, and Hydra A are showcase examples with deep Chandra exposures and complex bubble and cavity systems [72, 78, 213]. Radio bubbles in clusters are sufficiently common that they are an important heat source today, with enough power to balance the radiative cooling of dense gas in clusters (e.g., 65, 159), although the total energies and lifetimes of individual bubbles are considerably uncertain. An issue of particular interest that follows from this is the potential for the associated heating and cooling to forge the link between black-hole and galaxy growth. A recent review is available [144], and so the topic is not dealt in depth here.
It is noteworthy that the luminosity function of radio sources places the energetically dominant population to be roughly at the FRI/FRII boundary (e.g., 136), rather than within the more numerous but lower power population of FRIs studied in nearby clusters (although there are claims that total jet power scales slightly less than linearly with radio power (e.g., 211, 24). It thus remains possible that the rather gentle heating around currently studied sources does not provide us with the complete picture, and violent shock heating around more powerful sources is energetically important but currently eluding detection.
The best example of supersonic expansion is not in an FRII radio source but associated with the inner southwest radio lobe of Cen A ([124] and see Fig. 17 for a more recent, deeper, Chandra image). Cen A is our nearest radio galaxy, where 1 arcmin corresponds to ~ 1.1 kpc. The full extent of Cen A's radio emission covers several degrees on the sky [117]. Within this lies a sub-galaxy-sized double-lobed inner structure [36] with a predominantly one-sided jet to the NE and weak counter-jet knots to the SW [97] that are embedded in a radio lobe with pressure at least ten times larger than that of the ambient ISM [124]. The lobe should be expanding and be surrounded by a shock. The associated structure is exquisitely seen in Figure 17. Although the capped SW lobe is around the weak counterjet, so it is not evident that the lobe is being thrust forward supersonically with respect to the external interstellar medium (ISM) by the momentum flux of an active jet, the high internal pressure in the radio lobe ensures its strong expansion.
 |
Figure 17. Radio contours on a deep Chandra image of Cen A, showing the core and NE jet crossed by absorption stripes corresponding to NGC 5128's dust lanes, the SW lobe, structures associated with the NE lobe, the position of a merger-related gas discontinuity that shows up better at lower energies, and many XRBs in NGC 5128 (see [100, 114, 223, 127, 187]). |
The density contrast between post-shock and pre-shock gas in Cen A inferred by [124] was larger than four, which is not allowed by Equation 17, and so straightforward modelling was not possible. New modelling is underway using results from the new deep observation. However, the conclusion that the lobe's kinetic energy exceeds its thermal energy, and the thermal energy of the ISM within 15 kpc of the centre of the galaxy, is unlikely to change. As the shell dissipates, most of the kinetic energy should ultimately be converted into heat and this will have a major effect on Cen A's ISM, providing distributed heating.
There is much still to be learned about how gas is displaced by radio structures, and the processes of heat transfer. A new view will be possible with the high-resolution spectroscopic capabilities of the International X-ray Observatory currently under study by ESA and NASA. This will provide the vital ingredient of useful velocity data, giving a handle also on such issues as turbulence and non-perpendicular velocities at shocks.
7.4. The effect of galaxy mergers
It is important to understand what triggers radio activity and what
causes it to cease, particularly since radio sources are now
recognized as an important heat source for large-scale structure
(Section 7.2).
It has long been recognized that mergers may be important in
triggering radio activity, and this is consistent with the preference
for low-power radio galaxies to reside in clusters and rich groups.
For example, NGC 1275 and M 87 (7.2) are
the dominant galaxies of the Perseus and Virgo clusters, respectively.
Cen A (7.3) is hosted by NGC 5128 which in turn
hosts an inner warped disk suspected to be the merger remnant of a small
gas-rich spiral galaxy (e.g.
158).
Mergers leave an imprint on the temperature, density, and metallicity
structures of the gas. Due to good linear resolution it is again
Cen A that shows such effects particularly well,
with clear indications that even the hot X-ray-emitting gas is poorly
mixed. The merger appears to be having an important influence on the
evolution of the northeast radio jet and inner lobe
[127].
In the more extreme case of 3C 442A (Figure 18)
there is evidence that a merger
may have smothered a previously active jet, leaving a large volume of
decaying radio plasma, while at the same time re-starting jet activity
in the nucleus of one of the galaxies
[222].
Here the merger gas has sufficiently high pressure for the radio lobes
to be riding on the pressure front of the merger gas that is sweeping
them apart. The energy in the
merger gas will eventually be dissipated in the outer regions of the
group atmosphere - an additional source of heating to that arising
from both the old and new merger-induced radio activity. The radio
spectrum from the old decaying radio lobes is flatter where they are
being compressed by the expanding merger gas, suggesting that energy
from the gas has a second effect, in re-exciting relativistic electrons
through compression and adiabatic heating
[222].
While it is undoubtedly true that mergers produce messy substructures,
the example of 3C 442A suggests that there is some prospect that the
switching on and off of radio activity
by mergers can be timed (albeit roughly) using the morphology of the
stellar component of the galaxies and spectral changes in the radio
plasma, and that this can be combined with the measured energy content
of the gas and radio plasma to trace the history of radio outbursts and
their effectiveness in heating gas.
Figure 18. Chandra contours
(logarithmic spacing) on a color radio image of 3C 442A, taken from
[222].
Bright X-ray emission from the merger atmospheres of
NGC 7236 and NGC 7237 fills the gap between the radio lobes
which are no longer fuelled by an active jet.
