Seyfert 2 Spectra in Polarized Light

In studying the spectra of Seyfert 2 galaxies in the mid-1980s, a very surprising result was uncovered: when observed in polarized light, some Seyfert 2 galaxies revealed that they had hiding within their normal Seyfert 2 spectra a weak Seyfert 1 spectrum. But if a Seyfert 2 spectrum is masking an underlying Seyfert 1 spectrum, perhaps Seyfert 1 and Seyfert 2 galaxies are the same thing, just viewed in different ways. This suggests the possibility of a unified model of Seyfert galaxies.

Seyfert Galaxies (4) ...

The next ingredient in our model of a Seyfert galaxy is an observation that is described in the right box: in polarized light, a faint Seyfert 1 spectrum is seen for some galaxies that normally exhibit the spectrum of a Seyfert 2 galaxy. But what kind of model could account for the broad line region and narrow line region, and under some conditions look like a Seyfert 2 galaxy and under others like a Seyfert 1 galaxy? A strong clue is provided by the restriction that the hidden Seyfert 1 spectrum is unmasked only if we look at a Seyfert 2 galaxy in polarized light.

Conjuring a Seyfert 1 from a Seyfert 2
We have discussed two ways to polarize light that are of importance in astronomy. The first is that synchrotron radiation is naturally polarized. The second is that normal light is polarized by scattering it from surfaces or from clouds of dust particles. Both polarization mechanisms play important roles in understanding Seyfert galaxies, but the second is the key here. What if Seyfert 1 and Seyfert 2 galaxies both have BLRs and NLRs, but something masks our view of the BLR for a Seyfert 2 but not for a Seyfert 1? This could explain the difference in their emission line spectra. Furthermore, it could explain why we might see a weak Seyfert 1 spectrum in the polarized light from a Seyfert 2 if there were a scattering region outside the broad line region that could scatter light from the BLR toward us that was normally blocked from our sight. Notice the important point that this scattered light would be weak, but might be polarized. Therefore, if we used the right polarization filter we might be able to enhance this light enough relative to the normal Seyfert 2 light to see the hidden BLR and the broad lines of a Seyfert 1 spectrum.
A Model for a Seyfert Galaxy
The following diagram illustrates a model for the nucleus of a Seyfert galaxy that implements the above ideas in a concrete way. The central black hole engine is surrounded by a thin accretion disk of swirling matter that is heated to high temperature because of collisions and emits high-energy radiation (largely UV) before it is sucked into the black hole. The black hole and accretion disk are in turn surrounded by a dusty torus (doughnut shaped ring). The torus blocks much of the light coming from the central region in the plane of the torus, but is heated and glows strongly at IR wavelengths.

The broad line region (BLR) consists of clouds (shown schematically in blue in the figure--the colors are not meant to be realistic in this diagram). These clouds are moving at high velocity because they lie near the black hole, deep in its huge gravitational field. The UV radiation ionizes atoms in the BLR clouds (a process called photoionization), and the ions produce broad emission lines when they recombine with electrons. Support for this picture is provided by the observation that the broad line emission flux varies strongly in response to changes in the continuum flux produced by the central engine. The narrow line region (NLR) consists of the clouds shown schematically in red. They also are ionized by the UV radiation from the central engine (and by starlight from hot stars for the outer parts of the NLR), but are much further from the black hole and are less dense and move at lower velocities than the BLR clouds. The ions formed in the NLR produce (relatively) narrow emission lines because of the (relatively) low velocity of the clouds.

Some Caveats

The preceding diagram is meant to be highly schematic. For example, the relative sizes of the BLR, NLR, black hole, torus, and accretion disk are not drawn to scale. Further, always bear in mind that this diagram represents only the tiny central region of the Seyfert. The full galaxy is huge compared with this scale.

Hiding the Monster
Consider two observers, whose positions are indicated by the arrows in the preceding diagram. For the observer at the top, both the BLR and NLR are visible. Thus, this observer sees the characteristics of a Seyfert 1 AGN, with both broad and narrow lines and strong continuum emission across many wavelengths. Now consider the second observer, whose position is indicated by the arrow on the right. This observer can see the NLR region but direct sight of the BLR and the region of the central engine is blocked by the dusty torus. Therefore, this observer sees a Seyfert 2 AGN, which has only narrower lines and strong IR emission (from the torus). However, in polarized light the observer to the right may be able to see light from the blocked central region that is scattered in the observer's direction. In this case, the second observer will begin to see the broad lines characteristic of a Seyfert 1 galaxy emerging from the Seyfert 2 spectrum.