3.5. Radio Emission and AGN Activity in Ring Galaxies
Using the 3-element Greenbank interferometer, Ghigo (1980) made the first attempt to detect radio sources in ring galaxies. Only the brightest compact radio sources were detected, but these early pioneering measurements showed that compact non-thermal emission was present in the nuclei of some rings. Jeske (1986) made high resolution 20 cm snapshot observations of 28 rings with the A array of the VLA and detected sources in 24 of them. This a is significantly higher detection rate than Ghigo's original Green Bank study, due mainly to the enhanced sensitivity of the VLA. Only in two cases (Arp 147 and the Cartwheel) was extended emission observed from the ring. In most cases the sources were compact low-power radio sources with flux densities of typically 1 mJy at 20 cm. Jeske's thesis observations showed that optical spectra taken of the nuclear regions of rings containing compact radio sources were in almost all cases indicative of a power-law ultraviolet continuum, such as an AGN.
Higdon (1993) made high resolution radio continuum observations of the Cartwheel ring galaxy at 6 and 20 cm with the VLA. Radio emission was detected (at a level of approximately 11.5 mJy at 20 cm) most strongly from the southern quadrant of the outer ring in the region of the most intense star formation. An important result of this work is that large azimuthal variations are observed in the radio continuum observations which are mimicked in Higdon's H emission line observations. The fact that these azimuthal variations are seen at radio wavelengths strongly suggests that the H emission line strengths are not a result of selective extinction in the ring but are mapping real systematic variations in the star formation rate around the ring.
Recently, Ghigo and Appleton (1995) have made short (snapshot) VLA observations of 11 rings with the D-array at a wavelength of 2.3 cm (8.4 GHz). In addition, 4 of the rings were also observed at 6 and 20 cm. Out of their sample of 11 objects observed at 2.3 cm, 9 were detected. As with Jeske's survey, the radio sources associated with the rings tend to be weak (typically 0.5-2 mJy at 2.3 cm). At 20 cm, Ghigo and Appleton detected twice as much flux as Jeske with the D-array. This implies that for those galaxies in common, a large fraction of the flux was missed in the higher resolution A-array observations because of incomplete coverage in the uv (fourier transform) plane. This flux is probably extended on the scale of 5-12 arcsecs, which is significantly larger than the scale of the star forming regions. A good example of this is the radio emission from Arp 10. The emission is seen from both a compact source centered on the galaxy, and extended emission associated with the ring and the inter-ring region. The existence of an extended component is interesting and perhaps not totally unexpected in the collisional picture. As the density wave propagates outwards, it is expected to leave evolving stars in its wake. The more massive stars will, after a few × 106 years, explode as supernovae, generating strong non-thermal radio emission. Given the commensurate timescale for the diffusion and aging of such a relativistic plasma and the propagation time of the expanding ring (both around 108 years), then not only is the radio emission likely to be quite extended behind the ring, but may also show spectral steepening. The steepening in the spectrum might be due to two effects, the change from a mainly thermal to a non-thermal spectrum as one proceeds inwards from the ring, and the aging of the highest energy relativistic electrons, which will also lead to a steeper synchrotron spectrum. In Arp 10 (the only galaxy in the sample near enough to be easily resolved even with the 10 arcsec resolution of the D-array observations) such a change in spectral index is observed. The results are tantalizing, since higher resolution VLA observations will be needed to see whether the effects seen in Arp 10 are universal in the ring galaxy population. On the other hand, the results strongly suggest that the concept of an expanding starburst wave driven outwards through the disk is correct.
The first ring galaxy found to contain an optically identified AGN was NGC 985 (de Vaucouleurs and de Vaucouleurs 1975). The galaxy is of the RK type of Theys and Spiegel (1976). The Seyfert I nucleus is found embedded in the bright knot seen on the SE edge of the ring. It is one of the few ring galaxies so far detected at X-ray wavelengths (Ghigo et al. 1983). A peculiar twisted "arm" extends roughly from the Seyfert nucleus (but see below). de Vaucouleurs and de Vaucouleurs suggested that the ring resulted in the collision between an IGC and a disk galaxy. The galaxy was not studied further until recently in a series of papers by Rodriquez-Espinosa and Stanga (1990) and Stanga and Rodriguez (1991). These authors showed that the peculiar twisted arm contains large numbers of blue massive O stars. Based on H imaging, Rodriguez- Espinosa and Stanga (1990) estimate that the total star formation rate in the entire galaxy is currently 59 M yr-1. The star formation rate is the highest in the peculiar "arm", but large HII region complexes are found around the ring. The star formation rates are quite similar to those found in other ring galaxies (see Section 4.4).
The infrared observations made at 2.2 µm, by Appleton and Marcum (1993) show that a large fraction of the IR light comes from an r1/4 law bulge which is centered on the Seyfert nucleus, and not centered on the ring itself. It is now clear that NGC 985 has a double nucleus (Appleton and Marcum 1993; Garcia & Rodriguez-Espinosa 1995) first hinted at by Rodriquez-Espinosa and Stanga (1990). In addition to the Seyfert nucleus, a very red object is found approximately 3.8 arcsecs from it to the NW. The second nucleus is extended in both the IR and optical luminosity profile and appears like the brighter bulge, to follow an r1/4 law. However, the spectra of Stanga and Rodriguez (1991) show that the second object has an optically featureless continuum, leading Garcia and Rodriguez-Espinosa to suggest it is the bulge of a second galaxy. The extragalactic nature of the second nuclear source is now certain, since, like the merging systems Arp 220 (Norris 1988) and NGC 3256 (Norris and Forbes 1994), both nuclei have now been detected at radio wavelengths (Ghigo and Appleton, in preparation). Appleton and Marcum showed that when the smooth Seyfert bulge light was removed from their IR images, the second nucleus is attached to a straight bar-like structure. Another surprise from the IR observations is that the ring is not closed after bulge subtraction, but appears to take on the appearance of a tightly-wrapped spiral arm attached to the second nucleus by the bar. The evidence is now very strong that NGC 985 is a composite system of two colliding galaxies, one of which contains an active Seyfert nucleus, and both containing weak nuclear radio sources. It is likely that NGC 985 is an example of the kind of transition object first hinted at by Toomre (1978) in which the companion collides a significant distance from the center of the target disk, causing a breakdown of the ring into a spiral shape.
At the time of writing there are three Seyfert 1, (NGC 985, WN1 = Bootes ring, IR0302-5150) and one Seyfert 2 (Arp 118) ring galaxies known (2) (de Vaucouleurs and de Vaucouleurs 1975; Huchra et al. 1982; Wakamatsu and Nishida 1987, 1991; Lipari & Maccetto 1994). It may not be a coincidence that in all these cases the companion appears embedded in the ring or is seen in close contact with it. WN1 (Figure 14) has a ragged companion, exhibiting a high star formation rate. The companion seems to be caught in the act of disintegrating, probably as a result of its recent passage through the target disk. As in the case of NGC 985 above, both galaxies show 2.3 cm radio emission (Ghigo and Appleton, in preparation). The galaxy IR0302-5150 is an ambiguous case. Lipari & Maccetto (1994) favored the view that this was not a collisional system, but a resonant ring, in which the ring formed at the OLR. However, the similarity with WN1 is striking, and we believe that it may be a collisional system, despite the smoothness of the ring and the bar. The only Seyfert 2 system which might be classified as a ring (Arp 118) is a very messy object (Joy and Ghigo 1988), and we are inclined to think that it is not a simple case of a single two body collision. It appears to be a ring formed in a complex group environment. Jeske (1986) showed that radio continuum emission from Arp 118 was of a markedly different morphology from the other rings, again suggesting that it is not a clear-cut case.
Figure 14. The Seyfert 1 ring galaxy WNi discovered by Wakamatsu and Nishida (1987). Note the disrupted companion. This galaxy, like a number of other well known northern rings lies close to a bright star seen prominently in this CCD image (data from Appleton and Marston (1995) B-band- 2.1m KPNO telescope). (See Color Plate IX at the back of this issue.)
Obviously it is dangerous to draw conclusions from the small number of Seyferts and mini-AGN's found in the family of ring galaxies. However, it is tempting to speculate that the AGN activity is more prevalent in "contact" rings than those with well separated companions. This conjecture is supported by the data presented in Figure 15a. Ghigo and Appleton (1995) showed that if the radio luminosity at 2.6 cm was plotted against the separation of the companion galaxy from the ring system, an interesting effect is observed. Those galaxies with companions which are seen in optical contact with the ring (or projected against it) show a larger dispersion in radio fluxes than those at larger separations. This effect is also seen (more weakly) in the H luminosities versus pair separation (Appleton and Marston 1995) as shown in Figure 15b. The implication of Figure 15a, b taken together is that star formation activity is enhanced when the galaxies are in apparent contact. We note that Figure 15 is also in accord with the work of Hummel (1981) and Kennicutt et al. (1987) who showed that large dispersions in radio continuum and optical emission lines are common in paired galaxies in close contact.
Figure 15. a) The 2.3 cm radio luminosity versus the separation of the companion from the ring. A large dispersion in radio luminosity is noted when the companion galaxy is in contact with the ring (when Rsep < Dring/2) perhaps suggesting that enhanced activity may be related to the collision event itself (from Ghigo and Appleton, in preparation). b) The H luminosity of the galaxy versus companion separation. As in a), there is a tendency for the galaxies to show higher luminosities when they are effectively in contact with the ring (from Appleton and Marston 1995).
2 In another case (Davoust, Considere and Poulain (1991)) a large ring structure is found associated with a compact group of galaxies containing a Seyfert galaxy, although in this case it is not clear that the ring has formed by the Lynds and Toomre mechanism. Back.