3.4. Star Formation Properties of Ring Galaxies
Shortly after the IRAS point source catalog became available it was realized that ring galaxies emit a substantial amount of luminosity in the far infrared (Appleton and Struck-Marcell 1987a). Many ring galaxies are IRAS sources at 100 µm and more have been detected at 60 µm, a result of the publication of the faint IRAS point source catalog. Ring galaxies, like other collisional galaxy systems, exhibit higher than normal 60/100 micron color temperatures and range in far IR luminosity from 1010 L to a few × 1011 L. It was argued by Appleton and Struck-Marcell that, in most cases, the bulk of the FIR emission was from extended regions of star formation. A similar conclusion was drawn independently by Jeske (1986) and Wakamatsu and Nishida (1987).
Until recently, very little systematic work has been done on the star formation properties of ring galaxies. The first detailed study of star formation in ring galaxies was made by Fosbury and Hawarden (1977), and they showed that the Cartwheel ring contains large numbers of O and B stars concentrated in bright knots around the ring. The recent study of the Cartwheel by Higdon (1993) confirms these results, considerably adding to our knowledge of the optical emission-line properties of this galaxy. These results will soon be published (Higdon, in preparation). Higdon found that the total energy output of the Cartwheel in the H line was a factor of 10 times larger than that of the most luminous normal late-type spiral studied by Kennicutt (1988). An interesting result of Higdon's study was the discovery that 80% of the H emission from the Cartwheel comes from just one quadrant of the outer ring, the rest coming almost exclusively from the other three quadrants of the outer ring. Like Fosbury and Hawarden before him, Higdon did not detect H emission from the inner ring, despite predictions from numerical models that the inner ring should contain highly compressed gas (Appleton and Struck-Marcell 1987b; Hernquist and Weil 1993). In addition, 15% of the line emission was found to lie in a diffuse component associated with the ring, but not directly attributed to discrete star formation knots.
The brightest HII region complexes in the Cartwheel have H luminosities in excess of 1041 ergs cm-2 s-1, which is at the extreme end of the luminosity function for HII regions in late-type spirals. Higdon found that the total (instantaneous) star formation rate from the galaxy is currently 67.5 M yr-1 (assuming a Miller-Scalo IMF). This is 9.1 times the star formation rate estimated from the B-band luminosity which is sensitive to the integrated star formation rate over the last 15 Gyrs. Given the HI Mass (1.3 × 1010 M) distributed in the ring galaxy, Higdon found that the gas consumption timescale was 430 Myrs, a timescale similar to the ring propagation timescale.
Marston and Appleton (1995) have investigated the distribution and strength of star formation in a larger sample of northern ring galaxies via H imaging. Except for the case of WN1 (Wakamatsu and Nishida 1987), a Seyfert galaxy, H emission is found exclusively in the rings and not interior to them, or within the nucleus. In all cases, a large fraction of the star formation appeared to originate in discrete knots. Like the Cartwheel, a large number of the galaxies studied showed significant azimuthal variation in the H distributions around the ring, often exhibiting major concentrations of star formation in certain regions of the ring. Also, like the Cartwheel, a faint but significant component of H emission was detected from a diffuse component upon which the brighter knots are superimposed. In all cases studied, the diffuse component formed a complete ring, whereas in many cases the bright knots of star formation are concentrated in one quadrant of the ring. Until higher spatial resolution is obtained (via upcoming Hubble Space Telescope observations) it is impossible to determine whether the diffuse H represents a different mode of star formation from the bright knots, or whether the bright knots represent stochastic regions of significantly enhanced star formation in the general background of lower-level star formation in the ring.
In an effort to understand the marked azimuthal asymmetries in the star formation regions in ring galaxies, Charmandaris, Appleton and Marston (1993), made a detailed study of Arp 10, (Figure 13a) which appeared to show a particularly striking example of a non-uniform ring. Figure 13b shows the H distribution in the galaxy. Like the Cartwheel, this galaxy shows a strong enhancement of H flux in one quadrant of the ring. Such global enhancements in star formation rates were predicted in the models of Appleton and Struck-Marcell (1987b) if the star formation rate had a threshold dependence on ring over-density. Charmandaris, Appleton and Marston (1993) were able to show that the star formation rate in the ring jumped suddenly from a low value to a high value in the relatively smooth underlying stellar ring, implying some form of "threshold" process was at work controlling the star formation rates. Observations of the cold gas phase in these galaxies will be needed to confirm these intriguing results.
Figure 13. a) The RN ring galaxy Arp 10 (B-band from Appleton and Marston 1995; KPNO 2.1m telescope) exhibits extremely strong blue emission from one quadrant of the ring. b) B-R color map of Arp 10. Note the extremely blue quadrant of the ring and the color gradient inside the ring. Color coding is similar to that of Figures 7a and 7b. c) The H emission from the galaxy Arp 10 (from Charmandaris, Appleton and Marston 1993). The emission is concentrated in two rings and multiple ring arcs in the outer regions. (2.1m KPNO telescope). (See Color Plate VIII at the hack of this issue.)
Whatever the detailed cause of the azimuthal variations in star formation rates, it is clear that something is regulating the global star formation rates in the rings. Only the most circular rings(e.g. VIIZw466) show a uniform smattering of bright knots around the ring. In most cases, the bright knots are concentrated in one or two major complexes. It is very likely that the appearance of enhancements in star formation relates to the underlying asymmetry of the off-center collision. The distribution of these "hot-spots" may also relate to the position of shocks in three dimensions (Gerber et al. 1994).
One of the remarkable recent numerical results which deserves some comment is that of the different behavior of gas and stars in the SPH models of Gerber (1994). This work has shown that depending on the strength of the perturbation, the gas ring will either lead or trail the stellar ring. In the case of strong perturbations (for example, massive intruders), the gas ring is so dissipative that it lags significantly behind the stellar density wave. In the case of a small perturbation, the gas tends to pile up on the front-edge of the wave, where gas clouds first cross one another (Gerber, personal communication). Such model predictions are of great importance since they illustrate the remarkable usefulness of the ring galaxy as a "cosmic experiment". For example, the fact that star formation is found exclusively on the leading edge of the underlying infrared wave in the recent study of ring galaxies by Marston and Appleton (1995) (see also Appleton, Schombert and Robson 1992) suggests that the ionized gas component leads any underlying density wave. If correct (we don't yet have information on the cool gas) this implies that either most ring-making collisions involve small perturbations, or that the K-band light does not represent a stellar density wave. If the former explanation is correct, then this must imply that many ring galaxy targets have extremely large massive halos, since many of the companions are quite luminous and would be expected to lead to a large rather than small perturbation of the target disk. Further work is clearly needed in this interesting area.