7. FAR-INFRARED PROPERTIES

The infrared properties of the GPS and CSS sources can in principle shed light on the properties of the central engine and its immediate environment. Consider the expected mid-far-infrared (MFIR) properties of GPS/CSS sources compared to those of a sample of classical double radio sources with the same extended radio power and redshift. If the nucleus in GPS sources is surrounded by gas and dust with a higher covering factor than in "normal" extended radio galaxies, then we might expect a higher fraction of the optical-UV continuum light to be reprocessed and to emerge as MFIR radiation. Thus, in this scenario, we might expect the GPS sources to be brighter in the MFIR than "normal" radio sources. On the other hand, it is possible that the jets in the GPS sources convert a higher fraction of their energy into radio luminosity because of the interaction with the dense and clumpy environment (see, e.g., Eilek & Shore 1989; Gopal-Krishna & Wiita 1991; De Young 1993). This enhanced efficiency means that the GPS sources will be intrinsically weaker than a sample matched in radio power. In this case we might expect the GPS sources to be weaker in the MFIR than the comparison sample.

Heckman et al. (1994) constructed samples of GPS and CSS galaxies and quasars and matched them in radio power at 5 GHz and in redshift with comparison samples of classical double 3CR radio galaxies and quasars. They have determined the IRAS flux densities for the ensembles of objects using the SUPERSCANPI algorithm at IPAC. Note that most GPS and CSS sources are not detected by IRAS. Thus, these results only apply to the ensemble.

A summary of the Heckman et al. (1994) results is given in Table 6. Heckman et al. find that the 3CR galaxy comparison sample is detected at wavelengths 25 µm, while the GPS galaxies are detected at only 60 µm in the mean value, but not in the median, consistent with the mean value being dominated by a few larger values. However, the limits on the GPS galaxies are similar to the detections of the comparison sample, so no strong statements can currently be made. The data are consistent with the GPS galaxies and the comparison sample having similar MFIR infrared properties and rest-frame 50 µm luminosities 10¹² L. On the other hand, the quasars, although at higher mean redshift, are detected at higher flux density levels and at more wavelengths than the galaxies. The GPS/CSS quasars are detected at wavelengths 25 µm, while the 3CR comparison sample is detected at 12 µm. The two samples have similar MFIR flux densities to within a factor of about 2. The 50 µm luminosities of the GPS/CSS quasars are roughly 3 × 10¹² L. Note that the MFIR luminosity of the GPS/CSS sources is comparable to that of the ultraluminous IRAS galaxies (see, e.g., Sanders et al. 1988). The plot of the spectral energy distribution for the GPS radio galaxy 1345+125 (Fig. 14) shows that the energy output peaks in the mid-IR around 60 µm similar to the sources studied by Vader et al. (1993).

Figure 14. (a) Broadband spectrum F (Jy) vs. (Hz) of the GPS galaxy 1345+125, including radio, IRAS, near-IR to optical, and X-ray measurements. (b) Same data as (a), but showing F (ergs s-1 cm-2) vs. (Hz). Note that the energy output peaks in the mid-IR.

Hes, Barthel, & Hoekstra (1995) performed a study of the IRAS-detected 3CR quasars and radio galaxies. They also did not find strong evidence for a difference in properties of CSS sources with non-CSS sources. Thus, the similarity of the GPS/CSS samples with their comparison samples is consistent with neither the covering factor of the circumnuclear dust nor the jet efficiency being extremely different (i.e., to within a factor of a few) in the GPS/CSS sources and the classical doubles. Preliminary results from ISO are also consistent with this conclusion (Fanti et al. 1998, in preparation).