Astron. Astrophys. 351, 140-146 (1999)

5. Discussion and conclusions

The interior of galactic nuclei is usually hidden to optical and near IR observers by a dust veil of high optical depth. Measurements in the far IR, which are not handicapped by obscuration, are until now always of low resolution and furthermore cannot detect stars directly. If one wishes to determine the spatial configuration of stars and dust in a galactic core, one has to use indirect means. One way is to model the observed IR energy distribution. If one succeeds in building a model that correctly reproduces the IR fluxes of a sufficiently large data base, one has, if not a true, but at least a self-consistent picture. If furthermore maps or observations of different spatial resolution indicate that spherical symmetry is not gravely violated, one has even derived the likely structure of the nucleus.

The data of the four galaxies for which we computed models are certainly not extensive, neither in spatial nor spectral resolution, and are therefore only a first step towards the goal. Nevertheless, the parameters in Table 2 give a first tentative description of the nuclei. Of course, because of the simplified geometry and the small number of parameters, one has to be skeptical about the uniqueness of the models. For example, one may wonder why the transfer code yields such a good fit to Arp 220. In the calculations, it was assumed that the OB star density is constant over a galactic radius of 350 pc, whereas one sees at 2.2 µm and 10 µm that this galaxy is a merger containing two nuclei separated by 1" or 350 pc (Graham et al. 1990, Miles et al. 1996). The real configuration is evidently not spherically symmetric. So the structure of the model is, of course, wrong on a scale pc. But, on the other hand, the data which our model had to fit refer only to regions greater than 1". The more general result that most of the energy in Arp 220 comes from pc and that this region has an optical depth mag and that the total optical depth is about 50 mag are probably correct.

So the method of deriving the structure of an obscured galactic nucleus from modeling its observed IR emission by computing the radiative transfer is cheap and promising, if one does not forget its limitations.

© European Southern Observatory (ESO) 1999

Online publication: November 2, 1999
helpdesk.link@springer.de