Please Note: the e-mail address(es) and any external links in this paper were correct when it was written in 1995, but may no longer be valid.
Mullard Radio Astronomy Observatory, Cavendish Laboratory, Madingley Road, Cambridge CB3 0HE, UNITED KINGDOM
High-quality radio images show narrow linear features linking central `cores' to extended `lobes' in many radio galaxies and quasars, and these have become known as jets. Bridle & Perley (1984) were the first to state exactly what they meant by `jet' in their well-known review paper. According to Bridle and Perley, a jet is
and they extended this definition to trains of knots if they contained more than two knots or if some knots were elongated along the jet axis. These jets were identified with the `beams' carrying energetic particles from the central engine out into the lobes.
New images presented new problems, however. Kiloparsec-scale jets were not detected in the vast majority of FRII radio galaxies; as recently as 1991 Muxlow and Garrington found that only 10% of these objects had known jets, while FRII quasars of similar power almost all have bright one-sided jets (FRI radio galaxies normally have two-sided jets). This is puzzling if we believe that the jets trace the energy transport in the radio source: it is also puzzling if we believe that quasars and radio galaxies are essentially the same sort of object, as many of their other large-scale properties would suggest (Barthel (1994)). The standard explanation is that the jets in quasars are relativistic and close to the line of sight: this explains their extreme one-sidedness. In this simple unified scheme, radio galaxies would be in the plane of the sky and their jets would be dimmer and less one-sided than those of quasars. However, this model remains no more than speculation without observations of a large number of dim jets in FRII radio galaxies.
Recently these have started to become available. Black et al. (1992) looked at a sample of nearby () powerful () radio galaxies from the 3CR catalogue, imaging them with the full VLA at 8 GHz, and detected jets in up to 70% of them. Typically the contrast between the surface brightness of the jet and the surrounding lobe was only , so that high resolution and sensitivity were needed. Over the limited luminosity range available to him, Black found no relationship between the detectability and brightness of the jets and the luminosity of the parent sources.
To extend the work of Black et al., we have embarked on a project to map the galaxies in the LRL sub-sample of 3CR (Laing, Riley & Longair 1983) with . We have obtained 8 GHz VLA data on 17 of the 21 objects in this redshift range (the remaining four are large and well-studied) at B, C and D configurations, and we expect to obtain A configuration data where necessary. As we expected, the maps show a number of jets, and - like the objects in Black's sample - the features which we describe as jets are remarkably different in appearance, ranging from the twin jets in 3C438 (Figure 1) and the bright one-sided jet in 3C401 to thin, barely perceptible features like the jet in 3C20 (Figure 2).
Figure 1: 3C438 at 0.75 arcsec resolution. Contours at .
Figure 2: 3C20 at 0.6 arcsec resolution. Contours as for Figure 1. The jet appears as a curving ridge in the western lobe.
Our results so far confirm the findings of Black et al. that up to 70% of these sources have identifiable jets, and provide a wealth of information on individual objects. But what can they tell us about the physics of jets in general?
It is immediately obvious from the maps that the jets have a wide range of luminosities, even among sources with similar overall power (compare e.g. 3C401 and 3C436) which highlights the question ``what makes these jets light up?''. Although this large sample of FRII jets will allow us to try to relate the luminosity of the jet to the source luminosity and to properties known to be indicators of the beam power, such as [OIII] narrow line luminosity (Rawlings & Saunders (1991)), it's clear from preliminary work (Figure 3) that there will be a great deal of scatter in such a relation: some additional variables are affecting the fraction of the beam power channelled into jet synchrotron radiation. The fact that jets can bend and change dramatically in surface brightness along their length is suggestive of a strong interaction with the environment, and some authors (e.g. Fraix-Burnet (1992)) have argued that this interaction is entirely responsible for the variations in jet properties. We know that environmental asymmetries can dramatically influence the large-scale properties of a radio source from the work of McCarthy, van Breugel & Kapahi (1991), who showed that the extended emission-line regions in these sources, where their properties are known, are brighter on the side with the shorter lobe; this ties in with the well-known result that the brighter lobe in these sources has a weak but definite tendency to be the shorter. The problem with applying this to jets is that naïve models make predictions - that the brighter or only jet should lie in the shorter or longer lobe - which are not borne out by observation. One result of Black's which seems to be confirmed by our observations is that jets are more readily detected in smaller sources (Figure 4), which seems to be evidence for the large-scale environment influencing the jet visibility. We have applied for time on the Hubble Space telescope to make high-sensitivity observations of the host galaxies of sources in our sample: this may provide an important handle on the environment problem.
Figure 3: 8 GHz flux from the jet as a fraction of total 8 GHz flux, plotted against source luminosity at 178 MHz, for 18 sources in the combined sample. Scale is logarithmic: power is in and assumes , . Some data points are taken from Black (1992). As no attempt was made by Black to correct for background lobe flux, these fractions represent upper limits.
Figure 4: The logarithmic power/linear size diagram for the combined sample. Power and cosmology as in Figure 3; linear size in kpc. Sources which definitely have jets meeting the criteria of Bridle & Perley are denoted by a filled circle.
Another result of Black's which is confirmed by our extension of his sample is the one-sided nature of these jets. With the exception of the remarkable twin-jet source 3C438 - the most powerful such source known - there are no definite counterjets in the new sample and surprisingly few features that could be said to resemble them. The origin of this one-sidedness is debatable. The three standard suggestions to explain it are
Option (c) tends to be less favoured at the moment for several reasons. Observations show many jets which can be traced from the nucleus to the hotspot; there are many compact (and therefore short-lifetime) hotspots in lobes without visible counterjets; and it is hard to imagine a physical mechanism by which the flipping of jet sidedness could take place. Option (b) on its own seems to have problems explaining the enormous predominance of one-sided jets: there seems to be no a priori reason why some radio sources shouldn't be sitting in symmetrical environments. Further, the parsec-scale jets in these sources, when detected, always point towards the brighter kiloparsec-scale jet, which implies that any intrinsic asymmetry in jet emissivity has to begin on very small scales and be maintained over very large distances. Option (a), on the other hand, can't be the whole story; as I've already said, there are wide variations in the surface brightnesses along the jets (and, where visible, the counterjets), together with (and often associated with) significant jet bending, which must indicate some effects of the environment on jet visibility. With our new data, we can try to quantify the extent to which each of the three models apply by looking at limits placed on velocities by observed jet-counterjet asymmetries, by comparing (where possible) the bending and brightness variations in jets and counterjets, and by looking at the relations between jets and hotspots (various authors have suggested that the brighter hotspot is found on the same side as the jet, for example, although this trend was not apparent in the data of Black et al.).
Historically, the jets in quasars and in FRI radio galaxies have tended to be better studied because they are easier to detect. Models for jets in FRI sources are now quite detailed, the most popular ones involving trans- or sub-sonic turbulent flow, and have been applied to a number of sources with success: because these sources are generally nearby, 100-pc spatial resolution is possible with the VLA and excellent maps of the jets can be obtained. It now seems very likely that relativistic velocities are present in the inner parts of these jets, or at least the most powerful ones, and that the crucial difference between FRIs and FRIIs is the rapid deceleration of the jets in the former to trans-sonic velocities. The primary factor affecting this is clearly the jet power, as indicated by the source luminosity, whence the result of Fanaroff & Riley (1974); but intriguingly Owen (1993) has shown that the dividing line between FRI and FRII type is also a function of the host galaxy luminosity, providing clear evidence for a link between large-scale environment and jet properties.
Results are also now starting to emerge about the counterjets in quasars. In a preprint now circulating, Bridle et al. (1994) present very detailed images of 12 quasars made with the VLA at 4.9 GHz and discuss the properties of their jets and counterjet candidates. From the point of view of our sample these results are important as they represent a higher-luminosity sample of FRII objects (there are no FRII quasars with in the LRL sample: the lobe-dominated objects at small angles to the line of sight in unified schemes are probably the N-galaxies or broad-line radio galaxies). Jets in FRII radio galaxies at higher redshifts are harder to detect because of the lack of spatial resolution. By comparing these results with ours it should be possible to put more stringent constraints both on the one-sidedness problem and on the effects of source power on jet type.
We now have an unrivalled database of high-resolution 8 GHz VLA maps of nearby FRII radio galaxies. Jets are detected in up to 70% of the sample, although counterjets are still rare. We expect to be able to put important constraints on the physics of jets in these objects, which can then be compared with results on jets in quasars and FRI radio galaxies.