4.1. Questions and Timescales
There are a number of interesting questions we would like to answer when studying the gaseous X-ray emitting environment around a radio source. Is the external gas pressure greater than or equal to the minimum pressure calculated for the jet, in which case alternative methods of jet confinement are not required? We would like to know if the radio galaxy is moving in the X-ray medium (so that the jet is affected by ram-pressure forces), or if there are large-scale gas motions (cooling-flows, mergers, winds, etc) which affect the production of jets or cause their disruption. Is the gas distribution smooth between large and small scales, or do abrupt transitions in temperature and density induce observable radio deformations? Do we see evidence of direct interaction between the jets and the surrounding medium (e.g., heating of the X-ray gas) which can be tested against model predictions?
We can compare the inferred age of a radio source with the timescale over which the environment is likely to change.
The sound crossing time in gas of size d is 2 (d / kpc) (kT / keV)-1/2 Myr. This means that a medium 100 kpc in size has not had time to change as a result of the presence of a 100 Myr-old radio source, and a young radio source should be in an environment which is similar to that of its older counterparts.
The cooling time of gas of temperature kT and density ne is 3 × 104 (kT / keV)1/2 (ne / 10-3 cm-3)-1 Myr. Wide ranges of temperature and density relate to a wide range of cooling times, in many cases approaching the Hubble time. If a cooling-flow is key to the fuelling of a radio source, as suggested for powerful radio galaxies by Bremer et al. (1997), then it is curious that most radio sources appear to be 100 Myr old or younger, where we might expect some to last for a Gyr or more.
The phases of development of an elliptical-galaxy atmosphere (supernova wind, density increase, cooling, etc; e.g. Ciotti et al. 1991) are long compared with the measured lifetimes of radio galaxies. Are the host galaxies of radio sources in one of these phases, or all?
4.2. The Evidence
4.2.1. FRI radio galaxies
Hot atmospheres have been detected around FRI radio galaxies with Einstein and ROSAT. For representative results I turn to the largest sample of such objects with sensitive pointed X-ray observations: the B2 radio-galaxy sample. This is a 408 MHz flux-limited sample of 50 radio sources identified with elliptical galaxies of mZw 15.4 mag (Colla et al. 1975, Ulrich 1989), of which 40 were observed in ROSAT pointings, 39 being on-axis (Canosa et al. 1999). Apart from one starburst galaxy and one BLRG, all are FRIs at z 0.072. Two of the galaxies are the dominant members of catalogued Abell clusters (A1795 and A2199), and two are galaxies of the Coma cluster. The environments of other sample members are measured through their X-ray observations, with particularly useful data for sources observed with the PSPC (Worrall & Birkinshaw 1999a). Fig. 6 shows four representative X-ray images, illustrating group to cluster scales typical of X-ray emitting atmospheres. Such gas engulfs the radio structures (Fig. 7), but the lack of correlation between radio-source size, and size or central density of the X-ray-emitting medium, means that the gas has indeed not had sufficient time to adjust to the presence of the radio source (Section 4.1), and it must be small-scale processes, on size scales less than those of the overall gaseous environments, which are the major influence on radio-source dynamics and propagation.
Figure 6. Four representative > 8 ks-exposure ROSAT PSPC images entered on B2 radio galaxies not in catalogued clusters. Fields re 1.3 square degrees, and 10 arcmin corresponds to between 600 and 00 kpc. A range of sizes of X-ray-emitting atmosphere (group to luster dimension) is seen. Figure adapted from Worrall & Birkinshaw (1999a).
Figure 7. The X-ray atmospheres (grey scale) of the B2-sample galaxies in Fig. 6 are substantially larger than the VLA radio structures (contours). Note the change of scale from Fig. 6. Radio data are at 5 GHz for NGC 2484 (image kindly provided by M. Birkinshaw) and 1.4 GHz for the other sources (images kindly provided by R. Morganti).
Although the ROSAT PSPC's spectral resolution is poor by the standards of the CCD detectors on ASCA, Chandra, and XMM, and of Astro-E's calorimeters, the energy band is well matched to the typical temperatures of groups and poor clusters. The PSPC-derived luminosities and temperatures of the environments of B2 radio galaxies lie close to an extrapolation of the luminosity-temperature (Lbol - kT) correlation for more-luminous optically-selected clusters (Fig. 8). Since Lbol is principally governed by the gas mass, and kT by the total gravitating mass, this implies that the presence of the radio galaxy does not affect the gas fraction of the environment.
Figure 8. The X-ray-emitting atmospheres of representative B2 radio galaxies with good ROSAT PSPC measurements fit an extrapolation of the luminosity-temperature (Lbol - kT) correlation for more-luminous (~ 1044 - 1046 ergs s-1) optically-selected clusters (Arnaud & Evrard 1999; dotted lines show rms spread). This implies that the radio galaxy does not greatly influence the gas fraction of the environment. Figure from Worrall & Birkinshaw (1999a).
The gas densities for the atmospheres of B2 radio galaxies do not generally suggest the presence of cluster-scale cooling flows - the exceptions being for the two Abell clusters, A2199 and A1795. A2199 is a particularly interesting case, where Owen & Eilek (1998) have pointed out that the rotation measure of the core of B2 1626+39 (3C 338) implies appreciable central magnetic energy density, complicating the interpretation of any cooling flow. Possible galaxy-scale cooling flows, which may play a role in fuelling the radio galaxies, need further investigation using the sensitivity and spatial resolution now available with Chandra.
There is widespread evidence for pressure confinement of the kpc-scale radio structures of FRI sources by the X-ray emitting medium (e.g. Fig. 9 and Morganti et al. 1988, Killeen et al. 1988, Feretti et al. 1995, Trussoni et al. 1997), and in some cases an apparent evacuation of the external medium by the jets argues that additional internal jet pressure is required and must be supplied by something other than thermal gas (Böhringer et al. 1993, Hardcastle et al. 1998c). The exception to this picture, a moderately low-power radio galaxy which appears to require an alternative method of confining its long, straight, jet, is NGC 6251 (Fig. 10 and Birkinshaw & Worrall 1993, Werner et al. 1999).
Figure 9. Thermal pressures in the atmospheres of six B2-sample radio galaxies as deduced from fits to their ROSAT PSPC images (solid line, shown dashed where extrapolated beyond region of clear X-ray detection) compared with minimum internal pressure estimates in the radio sources (horizontal bars). The intergalactic medium is sufficient to confine the outer parts of the radio structures, and in some cases even to within 10 arcsec (5-10 kpc) of the core. In the case of NGC 315 the (extrapolated) pressure of the atmosphere matches the minimum pressure in the radio source over a factor of ~ 100 in linear scale. Figure from Worrall & Birkinshaw (1999a).
Figure 10. NGC 6251. 330 MHz radio contours on ROSAT PSPC image (left) and X-ray radial profile with best-fit model of unresolved emission plus weak group-scale gas described by a -model (right). Radio jet features between 10 arcsec and 4.4 arcmin from the core are all overpressured with respect to the X-ray medium in this giant radio source. Figure from Birkinshaw & Worrall (1993).
This review will not attempt a detailed discussion of how bending and disruption of the kpc-scale jet structures of low-power radio galaxies may relate to the motion of the radio galaxy through the gas or vice versa. However, various factors are likely to be influential, including gas flows and density enhancements resulting from cluster mergers (e.g. Bliton et al. 1998), density and temperature discontinuities at the interface between the galaxy and cluster atmospheres (e.g. Sakelliou & Merrifield 1999), and buoyancy forces (e.g. Worrall et al. 1995).
4.2.2. High-redshift FRII radio galaxies
A major success of ROSAT has been the first detection of high-power radio galaxies at high redshift. Of the 38 radio galaxies at z > 0.6 in the 3CRR sample (Laing et al. 1983), 12 were observed in ROSAT pointed observations and 9 were detected (see summary in Hardcastle & Worrall 1999a), with the four most significant detections exhibiting source extent (Worrall et al. 1994, Hardcastle et al. 1998b, Dickinson et al. 1999). Moreover, extended emission is detected around five 3CRR quasars at redshifts greater than ~ 0.4, one of which is at z > 0.6 (Hardcastle & Worrall 1999a, Crawford et al. 1999). Fig 11 plots the extended luminosities for sources for which the structure can be well modelled, together with upper limits for the other 3CRR FRII sources observed in ROSAT pointings (roughly half the sample). Powerful radio sources are finding some of the highest-redshift X-ray clusters known to date, pointing to deep gravitational potential wells early in the Universe.
Figure 11. The extended soft X-ray luminosity of high-power (FR II) quasars and galaxies from pointed observations of the 3CRR sample (from the work of Hardcastle & Worrall 1999a). Detections, in order of increasing redshift, are galaxies 3C 98 and 3C 388, BLRG 3C 219, CSS quasar 3C 48, quasar 3C 215, galaxy 3C 295, quasars 3C 334 and 3C 275.1, galaxy 3C 220.1, quasar 3C 254, and galaxy 3C 280. Upper limits are uncertain, particularly at low redshifts, due to the need to model spatial extent and adopt a value for the gas temperature.
4.2.3. More local FRII radio galaxies
The nearer a source, the more likely it is that its various X-ray emission components can be separated and the better will be the model fitting to any extended emission. FRII sources are rarer than FRIs and thus typically more distant. Cygnus A and high-redshift FRIIs with good X-ray data have extended X-ray luminosities one to two orders of magnitude higher than a typical FRI, but what about other more local FRIIs? Their extended emission should be as easy to detect if it really is so luminous. The situation appears mixed, with the extended luminosities for 3C 98 (z = 0.0306, Lx ~ 1042 ergs s-1) and 3C 388 (z = 0.0908, Lx ~ 1044 ergs s-1) differing by two orders of magnitude, and atmospheres for many sources not yet detected (Fig 11). This luminosity range spanned by 3C 98 and 3C 388 is similar to that of representative low-redshift FRIs (see Fig 8), although the full distribution of extended X-ray luminosities for FRIIs is uncertain while many nondetections remain. Despite this, an interesting picture emerges. Contrary to earlier work with less sensitive data (Miller et al. 1985), the X-ray atmospheres, where detected, provide sufficient pressure to confine the radio lobes, with no disagreement from the many sources for which only X-ray upper limits currently exist (Hardcastle & Worrall 1999c). In a detailed study of 3C 388, Leahy & Gizani (1999) have argued that that this implies the lobe energy density is higher than given by minimum-energy arguments, and they make the interesting point that if this is the case, jet kinematic luminosities (normally calculated as energy density times volume, divided by spectral age) are underestimated.
4.2.4. Young Radio Galaxies
GHz Peaked Spectrum (GPS) radio sources are believed to be young FRII sources and, even if only ~ 100 pc in size, the sound-crossing time in the surrounding medium (~ 105 years: Section 4.1) is likely to be appreciable compared with the age of the source (Conway 2000). We therefore expect the environments of such sources to be similar to those in the inner parts of their older counterparts. A search with ROSAT and ASCA for X-ray emission in or around the archetypal GPS radio galaxy 2352 + 495, at z = 0.237, has set an upper limit for the soft X-ray band (0.2 - 2 keV) of about 2 × 1042 ergs s-1 (O'Dea et al. 1996, O'Dea et al. 1999). From Fig 11, this is already below the level at which the atmospheres of some FRII radio galaxies are detected, suggesting that slightly more sensitive observations with forthcoming missions should see the atmosphere of this source.