8.8. Emission-Line Luminosities
Gelderman & Whittle (1994) present low-dispersion spectra (and high-dispersion spectra of the region around [O III] 5007) of a sample of 20 CSS sources (both galaxies and quasars). Gelderman & Whittle found that the CSS sources have relatively strong, high equivalent width, high-excitation line emission, with broad, structured [O III] profiles. They suggest that these properties are consistent with strong interactions between the radio source and the ambient line-emitting gas.
At this point, there are only a few GPS galaxies with high quality optical spectra (Lawrence et al. 1996), though work in progress should remedy this. Baum et al. (1998) have compiled emission-line data from the literature and work in preparation. The numbers of objects are small, and some sources are included that are not in the complete samples. Given the redshift distribution of the sources, it is not surprising that [O III] 5007 is the most common bright emission line visible in the spectra. Figure 19 shows the [O III] 5007 luminosity plotted against the projected linear size for the Gelderman & Whittle CSS galaxies and assorted GPS galaxies from the literature. Baum et al. (1998) find that the GPS sources tend to have lower line luminosities than the CSS sources.
Figure 19. [O III] 5007 luminosity vs. projected linear size for a sample of assorted CSS and GPS galaxies (not necessarily from the complete samples). Adapted from Baum et al. (1998). Horizontal lines are upper limits on emission-line luminosity. For these sources without detected emission lines, the redshift is estimated using the R-band Hubble diagram.
Readhead et al. (1996b) noted that four CSOs (also included here) were consistent with the relationship between jet power and emission-line luminosity found for large-scale sources by Rawlings & Saunders (1991; see also Baum & Heckman 1989). Baum et al. have compared the CSS and GPS sources with the large-scale sources studied by Zirbel & Baum (1995). In order to do this, Baum et al. have (1) (in cases where H + [N II] is not available) converted the [O III] fluxes to H + [N II] using the prescription given by Zirbel & Baum, (2) extrapolated the flux densities to 408 MHz, and (3) scaled the results to their values of H0 and q0. The resulting plot of H + [N II] luminosity versus total radio power at 408 MHz is shown in Figure 20. Also shown are the Zirbel & Baum fits to the relationships for FR 2 and FR 1 galaxies. Figure 20 can be compared to Figure 9 of Zirbel & Baum. Baum et al. find that (1) the CSS sources fall on (or even a bit above) the relationship defined by the FR 2 galaxies (cf. Hirst, Jackson, & Rawlings 1996); and (2) the GPS sources have a larger scatter on the emission-line-radio plane than the CSS galaxies or FR 2 sources and extend down to lower emission-line luminosities for a given radio luminosity. These results could be consistent with a model in which at least some of the GPS nuclei are more dust enshrouded than the CSS sources or are intrinsically fainter in ionizing photons for a given radio luminosity. Constraints on radio source evolution (section 12 and Fanti et al. 1995; Readhead et al. 1996a; O'Dea & Baum 1997) suggest that the GPS sources drop strongly in radio luminosity as they grow. O'Dea & Baum have suggested that some of the GPS sources may evolve into FR 1 sources. If this is the case, then the pre-FR 1 GPS sources may have lower ionizing photon luminosities and emission-line luminosities than the pre-FR 2 GPS sources as the large sources do (see, e.g., Baum, Zirbel, & O'Dea 1995).
Figure 20. H + N II luminosity vs. total power at 408 MHz (scaled to H0 = 50 km s-1 Mpc-1 for comparison with the results of Zirbel & Baum 1995). Adapted from Baum et al. (1998). The GPS and CSS galaxies detections are represented by crosses and solid squares, respectively. Upper limits to the emission-line luminosity are the horizontal lines and open squares for the GPS and CSS sources, respectively. The dashed and solid lines are the Zirbel & Baum fits to the FR 2 and FR 1 radio galaxies, respectively. Compare with Fig. 9 of Zirbel & Baum.
There are effects that could in principle influence the relationship between radio and emission-line luminosity. Consider the case where the emission lines are primarily energized by the ionizing nuclear continuum. If the jet thrust and ionizing continuum are constant over the lifetime of the radio sources and the radio luminosity declines by 1-2 orders of magnitude as required in the current evolutionary schemes (Fanti et al. 1995; Readhead 1995; Readhead et al. 1996a; Begelman 1996; O'Dea & Baum 1997), then the ratio of optical line to radio luminosity should increase over the lifetime of the radio source. This would imply that the GPS and CSS sources should have lower optical line luminosities for a given radio luminosity than the large-scale doubles. This is not seen - the CSS sources are at least as bright in emission lines as the extended sources for a given radio luminosity (see also Gelderman 1994). There are several possible implications. (1) Perhaps the simple picture of strong luminosity evolution is not correct. (2) There may be an additional source of ionization in the CSS sources, e.g., shocks generated by the radio source (Bicknell et al. 1997). The broad and structured [O III] profiles seen by Gelderman & Whittle (1994) support such an interaction of the radio source with emission-line gas. (3) The cold gas presumably acquired at the onset of nuclear activity may not yet have settled into a disk perpendicular to the radio source, resulting in a much larger fraction of the gas ionized by either the nuclear continuum or interaction with the radio source. Note that this is also consistent with the alignment effect observed in CSS sources.
In summary, the high emission-line luminosities in CSS and some GPS sources are consistent with an additional source of ionization (interaction with the radio sources) and/or a larger mass for the ionized nebula. The dispersion of emission-line luminosities in GPS sources suggests that at least some sources (1) are very dust enshrouded, (2) are weak sources of ionizing photons, or (3) will evolve into FR 1 sources.