arXiv:astro-ph/0411174v1 7 Nov 2004
February 2, 200814:21WSPC/Trim Size: 9in x 6in for Proceedingsgelbord˙seyfert2model
X-RAY EVIDENCE FOR MULTIPLE ABSORBING
STRUCTURES IN SEYFERT GALAXIES∗
J. M. GELBORD
MIT Center for Space Research, NE80-6091
77 Massachusetts Ave.
Cambridge, MA 02139, USA
K. A. WEAVER AND T. YAQOOB
We have used X-ray spectra to measure attenuating columns in a large sample of
Seyfert galaxies. Over 30 of these sources have resolved radio jets, allowing the
relative orientation of the nucleus and host galaxy to be constrained. We have
discovered that the distribution of absorbing columns is strongly correlated with
the relative orientation of the Seyfert structures. This result is inconsistent with
unification models having only a torus and is instead most readily explained if a
second absorber is included: in addition to a Compton-thick torus there would also
be a larger-scale absorber with NH< 1023cm−2. The second absorber is aligned
with the host galactic plane while the torus is arbitrarily misaligned.
The canonical unified model for Seyfert galaxies invokes chance line-
of-sight obscuration by a single parsec-scale torus to explain different ob-
served phenomena. However, some data are better explained by a model
incorporating a second absorbing structure2,3,4. We test this dual-absorber
(DA) model, assuming one absorber is the canonical parsec-scales torus,
arbitrarily misaligned with the host galactic plane, while the other is at
100-pc scales, aligned with the host galaxy disc (hereafter the galactic-
aligned absorber, or GA). Furthermore, we assume the torus is Compton
thick (NH > 1023.5cm−2) and the GA has a lower attenuating column.
Either absorber is capable of obscuring the central engine.
The relative alignment of the two obscuring structures plays an impor-
tant role in the DA model. Compton-thin Seyfert 2s will be observed only
if the line of sight intercepts the GA and avoids the torus. If the torus and
∗Drawn from Chap. 4 of Gelbord 20021, wherein more details and references may be
found. Online at http://space.mit.edu/∼jonathan/papers/thesis/abstract.html.
February 2, 2008 14:21WSPC/Trim Size: 9in x 6in for Proceedingsgelbord˙seyfert2model Download full-text
< 400 eV. As predicted, Compton-thin Seyfert 2s (GA-only absorption) are only found in
strongly misaligned systems. The distribution of all Compton-thin sources (incl. type 1s)
is insensitive to δ, as expected if only the orientation of the torus is important.
Compton-thick sources have EW > 1 keV, and Compton-thin ones have EW
GA are well aligned much or all of the GA will lie within the shadow cast
by the torus, so randomly-oriented lines of sight are likely to either inter-
cept both or neither, hence most well-aligned Seyferts would be observed as
either unabsorbed type 1s or Compton-thick type 2s. On the other hand,
when the absorbers are misaligned the GA covers part of the opening of
the torus, leaving fewer sight lines with a direct view of the nucleus and
more sight lines with Compton-thin (GA-only) absorption.
To test the DA model, we choose Seyferts for which the relative align-
ment can be constrained and use X-ray spectra to discriminate between
Compton-thick and Compton-thin attenuation. Alignment is indicated by
the δ valuesaof Kinney et al.5; small values indicate strong misalignments.
Fe Kα line equivalent width is measured to constrain absorbing columnsb.
Our measurements (Fig. 1) match the predictions of the DA model.
1. J. Gelbord, PhD Thesis, Johns Hopkins Univ. (2002). Available online at
2. K. K. McLeod, & G. H. Rieke, ApJ 441, 96 (1995).
3. H. Schmitt et al., ApJ 555, 663 (2001).
4. G. Matt, A&A 355, L31 (2000).
5. A. L. Kinney et al., ApJ 537, 152.
aδ is the angle between the (projected) radio jet and major axis of the host galaxy.
bFe Kα EW provides a more robust indicator than measuring NH from continuum
modeling because values outside the range 1022–1024cm−2are not well constrained.