Annu. Rev. Astron. Astrophys. 2000. 38: 761-814 Copyright © 2000 by Annual Reviews. All rights reserved

3.3.3. Re-constructing the Big Blue (EUV) Bump

In the standard paradigm, AGNs are powered by accretion onto massive black holes. From basic theoretical considerations about the size scale of the emitting region and the energy released, one expects that accretion disks emit quasi-thermal radiation with a characteristic temperature of a few 10⁵ K (~ 40 eV). The intrinsic SEDs of AGNs cannot be directly observed between the Lyman edge and a few hundred eV, however, because of Galactic and intrinsic absorption. In Seyfert 1s and QSOs, observations shortward (X-rays) and longward (UV) of the critical region indicate that there is a break in the continuum slope. The data are broadly consistent with the existence of an emission peak [the "Big Blue Bump" (BBB)] in the extreme UV (EUV; Malkan & Sargent 1982, Sanders et al 1989, Walter et al 1994, Elvis et al 1994).

Observations of emission lines from highly excited species enable a different approach to the study of the EUV spectrum of AGNs. These "coronal" lines (Oke & Sargent 1968) originate in the narrow line region and can thus be detected even in type 2 sources, yet they may (and should, in standard unified schemes) probe the intrinsic ionizing continuum. Coronal lines are probably excited by photoionization (Oliva et al 1994). Thus, line ratios from ions in different stages of excitation, along with a photoionization model, can in principle be used to reconstruct the SED of the ionizing continuum. The tricky part is that the emission line flux depends not only on the ionizing luminosity and EUV SED, but also on the ionization parameter in the narrow line region (the ratio of ionizing flux to the local electron density). This makes the derived SED highly model-dependent on the radial and density structure of the narrow line region.

Following a first study of the ionizing continuum in the Circinus galaxy by Moorwood et al (1996a), Alexander et al (1999, 2000) have analyzed in detail three nearby Seyfert nuclei with this technique: Circinus (Seyfert 2 + starburst), NGC 4151 (Seyfert 1.5), and NGC 1068 (Seyfert 2). Starting with a (selected) compilation of ISO lines plus UV/optical/near-IR lines from the literature (typically 20 to 30 in each galaxy), Alexander et al found the best model from a scored fitting approach. In addition to the overall shape of the input SED (parameterized with 6 spectral points in the 10-500 eV range), the model depends on extinction, metallicity, and the density structure/coverage of the gas. Alexander et al explored various models of the narrow line region and concluded that there are well-defined, robust input SEDs for each of the three sources. These are shown in Figure 11. The Circinus data require a pronounced EUV bump peaking at about 70 eV and containing (in L) at least 50% of the AGN's luminosity (5 × 10⁹ L; the other half is in the X-ray power-law; Moorwood et al 1996a). The black-hole mass is less than about 4 × 10⁶ M (Maiolino et al 1998), so the AGN's luminosity must be greater than 10% of the Eddington luminosity. The Circinus data are thus fully consistent with the standard AGN paradigm.

Figure 11. UV spectral energy distributions of three Seyfert galaxies reconstructed from infrared/optical narrow line region line ratios (from Alexander et al 1999, 2000; Moorwood et al 1996). The best model is indicated by a thick continuous line, and the shaded region marks the 99% (90% for Circinus) confidence zone. In the cases of NGC4151 and NGC1068, possible composite models with an absorbed bump are also shown.

The derived SEDs for NGC 4151 and NGC 1068 appear to be different, possibly casting doubt on unified schemes (or the technique used). In both galaxies, the SED turns down sharply just beyond the Lyman edge, comes up again equally sharply for a peak around 100 eV, and then finally drops down again and connects to the X-ray spectrum. Such a structured SED is not predicted by any accretion disk model, and it is not clear which physical mechanism might produce such a sharp emission feature at ~ 100 eV. Alexander et al proposed as a solution that the NLR does not see the intrinsic SED of the AGN. The SEDs for both NGC 1068 and NGC 4151 can alternatively be interpreted by an EUV bump that has a deep absorption notch resulting from intervening absorption with a hydrogen column density of ~ 4 × 10¹⁹ cm^-2. Based on ISO SWS and HST observations, Kraemer et al (1999) concluded as well that an absorber exists between the broad line region and the narrow line region of the very nearby Seyfert 1 galaxy NGC 4395. Further, there is independent evidence for such intervening absorbing gas from UV observations in NGC 4151 (Kriss et al 1992, 1995). Such absorbers may be common in Seyferts (Kraemer et al 1998). These findings are thus consistent with the standard paradigm that all Seyfert galaxies are powered by thin accretion disks with a quasi-thermal EUV bump, if in a (large) fraction of them the narrow line region sees a partially absorbed ionizing continuum.

Nevertheless, caution must be exercized before we can fully trust these very encouraging new results. Although Alexander et al explored a significant range of possible geometries of the narrow line region and density structures, it remains to be shown whether the derived SEDs are robust if more complex models are considered. For instance, Binette et al (1997) concluded that the Circinus data can be fitted with a bumpless power-law continuum if one allows for a component of optically thin (density-bounded) gas clouds.

An additional interesting aspect of the IR coronal line work is that the NGC 4151 SWS line profiles show the same blue asymmetries that are characteristic of the optical emission line profiles in this and other Seyfert galaxies (Sturm et al 1999a). This result excludes the common interpretation - at least for NGC 4151 - that the profile asymmetries are caused by differential extinction in an outflow or inflow of clouds with a modest amount of mixed-in dust. Sturm et al proposed instead a geometrically thin but optically thick obscuring screen close to the nucleus of NGC 4151.