The Nature of the Ultraviolet Continuum in Type 2 Seyfert Galaxies

Abstract

Type 2 Seyfert nuclei are well known to contain a ``featureless continuum'' which makes a significant contribution in the optical and ultraviolet. However, the nature of this featureless continuum is not clear. Recent optical spectropolarimetry shows that only a minor part of the optical featureless continuum can be light from a hidden Seyfert 1 nucleus scattered into our line of sight. In this paper, we show that this is also true in the ultraviolet. We have used International Ultraviolet Explorer (IUE) spectra of 20 of the brightest type 2 Seyfert nuclei to construct an ultraviolet spectral template. While the continuum is well detected in the template, there is no detectable broad-line region (BLR). Comparing this template to a similar spectral template of type 1 Seyfert nuclei implies that no more than 20% of the Seyfert 2 template's continuum can be light from a hidden Seyfert 1 nucleus scattered by dust or warm electrons. One obvious possibility is that most of the nuclei in our sample are ``pure'' type 2 Seyfert galaxies that do not contain a hidden type 1 Seyfert nucleus (e.g., we have a clear view of the central engine in the ultraviolet, and it simply lacks a broad-line region). This is not compatible with the evidence that at least some Seyfert 2 galaxies can be unified with Seyfert 1 galaxies on the basis of viewing geometry, unless there are two types of Seyfert 2 galaxies. As an alternative, we consider the possibility of Seyfert 1 light scattered off very hot electrons (T > 10^7^ K). The BLR emission lines can then be broadened beyond recognition in our data. However, a scatterer this hot is inconsistent with optical spectropolarimetry. Optically-thin thermal emission from the type of warm mirror seen in NGC 1068 cannot produce the ultraviolet continuum we observe because the equivalent width of Lyα and He II λ1640 are at least an order of magnitude too small in the Seyfert 2 template, and the observed ultraviolet continuum is generally much too red to be thermal emission from gas warmer than 10^5^ K. We discuss an alternative in which most of the ultraviolet continuum in these Seyfert 2 galaxies may be produced by a reddened starburst: a circumnuclear population of massive stars which is unusually luminous in type 2 Seyfert galaxies compared to normal galaxies of the same Hubble type. We show that our Seyfert 2 template is consistent with existing IUE spectra of metal-rich starbursts in all salient properties. Two consequences of this would be that a significant fraction of the optical Balmer emission lines (as measured in an aperture as large as that of IUE) and most of the far-infrared continuum detected from these Seyfert 2 galaxies would be powered by the starburst. Combining this inference with the evidence that many type 2 Seyfert nuclei contain a ``hidden'' type 1 Seyfert nucleus would then imply that both compact active nuclei and starbursts play important energetic roles in the Seyfert phenomenon. We also show that if NGC 1068 were at the median distance of the Seyfert 2 galaxies in our sample, then about 70% of its ultraviolet continuum as observed through the IUE aperture would arise in its circumnuclear star- forming ring. Thus, the ``anomalous'' behavior of the featureless continuum in NGC 1068 compared to other Seyfert 2 galaxies may be caused at least in part by an aperture effect. Although the type 2 Seyfert template is not of adequate quality to allow the direct spectroscopic detection of massive stars, if such stars are present then this should be possible in the near future with the Hubble Space Telescope (HST). Such observations will also be required to ascertain the relationship between the ultraviolet continuum seen by IUE and the unpolarized component of the optical featureless continuum (the ``FC2''). In an appendix we consider whether any genuine ``type 2 quasars'' have yet been detected at high redshift.