Astron. Astrophys. 329, L45-L48 (1998)

4. The radio-IR spectrum of the inner parsecs of NGC 1068

MPH96 showed that the central have substructures on a scale of 100 mas requiring observations with a resolution of better than to separate the true nuclear spectrum from that of surrounding sources. Only a few published radio flux determinations of the nucleus of NGC 1068 have a sufficiently high angular resolution to allow the separation of individual nuclear components and thus to use it for physical investigations of the nucleus' property. Fortunately, our speckle observations have the required high angular resolution (76 mas). Therefore, our resolution would allow the separation of the individual core components discussed by MPH96 if present in the IR. We assume that the single source that we have observed in the K-band is the same as the true nucleus observed by MPH96. Our observations constitute an upper limit to the volume from which the above determined originate.

Usually, the K band flux from nuclei of Seyfert 2 galaxies is attributed either to a warm dust torus or a compact nuclear stellar cluster or a combination of the two (e.g., Thatte et al. 1997). If our source is the torus that is held responsible for the different appearences of Seyfert 1 and 2 galaxies, our observation constitute the first determination of a torus size. To clarify the nature of the radiating source, further spectroscopic and polarimetric measurements with a similarly high angular resolution are necessary.

However, a combination of the flux measurements and nuclear source identification by MPH96 in the radio frequency regime with our observation makes it intriguing to speculate whether a sizable fraction of could originate from the very nucleus of NGC 1068 rather than from the torus. This could be achieved in a scattering halo above and below the nuclear torus. In this halo a large part of the flux could be isotropically scattered rather than absorbed and thermalized in an opaque torus along our direct line-of-sight to the nucleus. lies only about a factor of two above the extrapolated spectrum derived for the range around 10 GHz: The spectral index between 5 GHz and the K band amounts to . We note that this value is very similar to that of other galactic nuclei, like Sgr A* (BDM96), M 81 (Reuter and Lesch 1996); M 104 (Jauch and Duschl, in prep.), where . If NGC 1068 has the same spectral shape as these other galactic nuclei, then a fraction of could indeed be contributed from the nucleus of NGC 1068.

However, one has to admit that very little is known about the true nuclear spectrum of NGC 1068 in the intermediate frequency range. To persue our speculation, we assume - as a working hypothesis - that also between 22 GHz and the IR range, the spectrum goes like . We then follow BDM96 and interpret this as optically thin synchrotron radiation of quasi-monoenergetic electrons. The mean electron energy then is fairly well constrained since the maximum of has to be at frequencies above the K band, but not much higher as otherwise the total nuclear flux from the center of NGC 1068 would be too large. The situation is less clear with the SSA frequency. We cannot rule out that SSA in fact occurs at frequencies even smaller than 5 GHz. As a consequence of this, the source radius discussed below is only a lower limit. For details we refer the reader to BDM96 ¹.

If we assume that the maximum of is indeed achieved around 2 m, and that SSA of the source becomes important for frequencies below 5 GHz, we find as emitting region a homogeneous sphere of radius cm () with a magnetic field G (assumed to be the same everywhere in this region). The relativistic electrons have a number density , a mean energy GeV and a width of the energy distribution . In Fig. 5 we show a comparison of the observed fluxes of NGC 1068 core and our model spectrum using the above parameters. If our speculation applies, it turns out that the main difference between NGC 1068 and other galactic centers analysed on the basis of the same interpretation (Sgr A*: BDM96; M 81: Reuter and Lesch 1996; M 104: Jauch and Duschl, in prep.) are the source radius and - especially - the energy of the relativistic electrons. The above size of 0.01 mas means that our resolved 30 mas object is not the synchrotron source itself but rather a larger object, most likely the nuclear torus and/or a circumnuclear scattering halo.

Fig. 5. A comparison of the model spectrum (for parameters, see the text) with flux determinations with a sufficiently high spatial resolution. The fluxes were taken from MPH96 (5 and 22 GHz), Ulvestad, Neff, Wilson (1987; 15 GHz), and the present paper (K-band).

© European Southern Observatory (ESO) 1998

Online publication: December 16, 1997
helpdesk.link@springer.de