Astron. Astrophys. 355, 873-879 (2000)

6. Conclusions

In Chiaberge et al. (1999) we discovered that FR I nuclei lie in the radio-optical luminosity plane along a tight linear correlation. We argued that this is due to a common synchrotron origin for both the radio and optical emission. FR I nuclei must also be unobscured and intrinsically lacking of BLR and of significant thermal emission from any powerful accretion disc.

In order to explore how the differences in radio morphology are related to the optical nuclear properties, we analyzed HST images of 24 extended radio-galaxies morphologically classified as FR II, belonging to the 3C catalog and with . We detected optical cores in 13 sources, which implies that the covering fraction of any obscuring material is less than , or equivalently, the torus has an opening angle of . This can be even larger if at least some of the upper limits are actually just below the detection threshold. Notice that our determination of this critical angle is inconsistent with the division between higher redshift () 3CR quasars and radio galaxies, which has been found to be (Barthel 1989). This might be a problem, however the low redshift selection of our sources does not allow to derive any firm conclusion. We are currently studying a larger and higher redshift sample in order to further investigate this issue (Chiaberge et al. in preparation).

Our results suggest that the radio morphology is not unequivocally connected with the optical properties of the innermost structure of radio galaxies. In fact, at least at low redshifts, there is not a single homogeneous population of FR II: unlike FR I, they show a complex behavior, which is however clearly related to their optical spectral classification.

In BLRG optical nuclei are likely to be dominated by thermal (disc) emission. As discussed above, line emission contamination cannot account for this excess. In agreement with the current unification scheme of radio loud AGNs, we also identify their possible obscured counter-parts. It seems that broad lines and obscuring tori are closely linked and both are present only associated to radiatively efficient accretion.

We also find five FR II sources, spectrally identified as narrow lined objects, which harbor nuclei essentially indistinguishable from those seen in FR I. By analogy with FR I, we argue that their optical nuclear emission is produced primarily by synchrotron radiation, they are not obscured to our line of sight and therefore intrinsically lack a BLR.

Clearly, a classification based on the optical nuclear properties, as seen in these HST images, is more likely to reflect true similarities (or differences) on the nature of the central engine (such as, e.g., the rate of radiative dissipation in the accretion disc) than the traditional dichotomy of radio morphology.

From our data and within the limits of the available statistics, we find no evidence of a continuous transition between the two classes (FR I and FR II), as they are well separated in the vs plane. At this stage we only point out that sources with cores below erg s^-1 Hz^-1 (or equivalently erg s^-1 Hz^-1) have FR I like nuclei, while FR II start above this threshold.

It is of particular interest that a significant fraction of FR II (at least 30%, but can be as large as 50% depending on the nature of the sources without detected optical nuclei) have FR I-like nuclei. The fact that all of these are located in clusters, an environment typical of FR I, might represent an important hint on the origin of the different flavours of radio galaxies, worth exploring through the study of a larger sample of objects.

These results have also interesting bearings from the point of view of the unified models. In fact, this picture argues against the idea that all FR II radio galaxies constitute the parent population of radio-loud quasars. We propose instead that galaxies with FR II morphology and an FR I-like core are possibly mis-aligned counter-parts of BL Lac objects. This can account for the observation that some radio-selected-type BL Lacs show radio morphologies more consistent with FR II than with FR I (e.g. Kollgaard et al. 1992).

To conclude, we note that all of the galaxies included in our sample are low redshift objects with total radio powers not exceeding erg s^-1 Hz^-1: thus a crucial observational issue is to understand whether these results hold to higher power/redshift samples or they are limited to low luminosities FR II. This will be explored in a forthcoming paper.

© European Southern Observatory (ESO) 2000

Online publication: March 21, 2000
helpdesk.link@springer.de