MNRAS 456, L94–L98 (2016) doi:10.1093/mnrasl/slv178
NuSTAR catches the unveiling nucleus of NGC 1068
A. Marinucci,
1‹
S. Bianchi,
1
G. Matt,
1
D. M. Alexander,
2
M. Balokovi
´
c,
3
F. E. Bauer,
4,5,6
W. N. Brandt,
7,8,9
P. Gandhi,
2
M. Guainazzi,
10
F. A. Harrison,
3
K. Iwasawa,
11
M. Koss,
12
K. K. Madsen,
3
F. Nicastro,
13,14,15
S. Puccetti,
13,16
C. Ricci,
4
D. Stern
17
and D. J. Walton
3,17
1
Dipartimento di Matematica e Fisica, Universit
`
a degli Studi Roma Tre, via della Vasca Navale 84, I-00146 Roma, Italy
2
Department of Physics & Astronomy, University of Southampton, Highfield, Southampton, SO17 1BJ, UK
3
Cahill Center for Astronomy and Astrophysics, California Institute of Technology, Pasadena, CA 91125, USA
4
Instituto de Astrof
´
ısica, Facultad de F
´
ısica, Pontificia Universidad Cat
´
olica de Chile, 306, Santiago 22, Chile
5
Millennium Institute of Astrophysics, Vicu
˜
na Mackenna 4860, 7820436 Macul, Santiago, Chile
6
Space Science Institute, 4750 Walnut Street, Suite 205, Boulder, CO 80301, USA
7
Department of Astronomy and Astrophysics, 525 Davey Lab, The Pennsylvania State University, University Park, PA 16802, USA
8
Institute for Gravitation and the Cosmos, The Pennsylvania State University, University Park, PA 16802, USA
9
Department of Physics, 104 Davey Lab, The Pennsylvania S tate University, University Pa rk, PA 16802, USA
10
European Space Astronomy Centre of ESA, PO Box 78, Villanueva de la Canada, E-28691 Madrid, Spain
11
ICREA and Institut de Ci
`
encies del Cosmos, Universitat de Barcelona, IEEC-UB, Mart
´
ı i Franqu
`
es, 1, E-08028 Barcelona, Spain
12
Institute for Astronomy, Department of Physics, ETH Zurich, Wolfgang-Pauli-Strasse 27, CH-8093 Zurich, Switzerland
13
INAF Osservatorio Astronomico di Roma, via Frascati 33, I-00040 Monte Porzio Catone (RM), Italy
14
Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, MS-04 Cambridge, MA 02138, USA
15
Department of Physics, University of Crete, PO Box 2208, GR-710 03 Heraklion, Crete, Greece
16
ASDC-ASI, Via del Politecnico, I-00133 Roma, Italy
17
Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA 91109, USA
Accepted 2015 November 10. Received 2015 October 29; in original form 2015 July 16
ABSTRACT
We present a NuSTAR and XMM–Newton monitoring campaign in 2014/2015 of the Compton-
thick Seyfert 2 galaxy, NGC 1068. During the 2014 August observation, we detect with
NuSTAR a flux excess above 20 keV (32 ± 6 per cent) with respect to the 2012 December
observation and to a later observation performed in 2015 February. We do not detect any
spectral variation below 10 keV in the XMM–Newton data. The transient excess can be
explained by a temporary decrease of the column density of the obscuring material along the
line of sight (from N
H
10
25
cm
−2
to N
H
= 6.7 ± 1.0 × 10
24
cm
−2
), which allows us for the
first time to unveil the direct nuclear radiation of the buried active galactic nucleus in NGC
1068 and to infer an intrinsic 2–10 keV luminosity L
X
= 7
+7
−4
× 10
43
erg s
−1
.
Key words: galaxies: active – galaxies: individual: NGC 1068 – galaxies: Seyfert.
1 INTRODUCTION
Since Antonucci & Miller (1985) proposed the unification scheme
for type-1 and type-2 active galactic nucleus (AGN), it has been
commonly thought that highly absorbed (i.e. Compton-thick, with
N
H
≥ 1.5 × 10
24
cm
−2
) Seyfert 2s are obscured by neutral gaseous
matter embedded in a thick molecular torus located at parsec
distances from the central X-ray source (see Netzer 2015,fora
recent review). Reflection from the torus reveals itself through
a very intense neutral iron K α emission line at 6.4 keV, with
E-mail: marinucci@fis.uniroma3.it
equivalent widths of ∼1 keV, and a prominent Compton hump
peaking at ∼20 keV (Ghisellini, Haardt & Matt 1994).
Recently, both the size and the distance of this thick screen have
been questioned by a number of observations that have measured
significant column density variability of the innermost absorber
over time-scales of days or even hours in nearby bright sources
such as NGC 1365 (Risaliti et al. 2005;Riversetal.2015a), NGC
4388 (Elvis et al. 2004), NGC 4151 (Puccetti et al. 2007) and NGC
7582 (Bianchi et al. 2009,Riversetal.2015b). On the other hand,
spatially resolved iron K α line emission, extended on scales of
hundreds of parsecs, has been detected in the brightest Compton-
thick objects such as NGC 1068 (Young, Wilson & Shopbell 2001;
Brinkman et al. 2002), NGC 4945 (Marinucci et al. 2012a)and
Mrk 3 (Guainazzi et al. 2012). These measurements suggest that
C
2015 The Authors
Published by Oxford University Press on behalf of the Royal Astronomical Society
at University of Durham on June 15, 2016http://mnrasl.oxfordjournals.org/Downloaded from
Unveiling the nucleus of NGC 1068 L95
different absorbers/reflectors, located on a variety of spatial scales,
may contribute (see e.g. Bianchi, Maiolino & Risaliti 2012)tothe
absorption and reprocessing.
NGC 1068 (D
L
= 14.4 Mpc; Tully 1988) is one of the best studied
Seyfert 2 galaxies. Indeed, the unification model was first proposed
to explain the presence of broad optical lines in its polarized light.
In X-rays, it was first studied by Ginga, which detected a strong
(EW ∼ 1.3 keV) neutral iron line (Koyama et al. 1989), an un-
ambiguous sign that we are observing reflected, rather than direct
radiation (Matt, Brandt & Fabian 1996). This result was later con-
firmed by ASCA (Ueno et al. 1994; Iwasawa, Fabian & Matt 1997)
which resolved the iron line into neutral and ionized components.
BeppoSAX (Matt et al. 1997) found no evidence for transmitted ra-
diation up to 100 keV, implying a column density of the absorbing
material in excess of 10
25
cm
−2
.Mattetal.(2004) and Pounds &
Vaughan (2006) studied the XMM–Newton/EPIC spectra, and found
evidence for an iron overabundance with respect to the solar value.
Recently, Bauer et al. (2015) analysed the multi-epoch X-ray
spectra of NGC 1068 using different observatories, including
3–79 keV data from NuSTAR. They interpreted the broad-band cold
reflected emission of NGC 1068 as originating from multiple reflec-
tors with three distinct column densities. The highest N
H
component
(N
H, 1
10
25
cm
−2
) is the dominant contribution to the Compton
hump, while the lowest N
H
component (N
H, 2
∼ 1.5 × 10
23
cm
−2
)
produces much of the line emission. The authors also confirm that
almost 30 per cent of the neutral Fe K α line flux arises from regions
outside the central 140 pc.
Guainazzi et al. (2000) found evidence for variability, in the 3–
10 keV band, comparing two BeppoSAX observations performed
about one year apart. Later on, comparing ASCA, RossiXTE and
BeppoSAX spectra taken at different epochs spanning a few months,
Colbert et al. (2002) claimed variations in both the continuum and
He-like iron line flux on time-scales as short as four months, using
the 2–10 keV energy band. Matt et al. (2004), comparing an XMM–
Newton observation with BeppoSAX observations performed a few
years earlier, found possible evidence for flux variability of both the
cold and the ionized reflectors.
We present a joint XMM–Newton and NuSTAR monitoring cam-
paign of NGC 1068, from 2014 July until 2015 February, and report
on the discovery of a transient excess above 20 keV.
We adopt the cosmological parameters H
0
= 70 km s
−1
Mpc
−1
,
= 0.73 and
m
= 0.27, i.e. the default ones in XSPEC 12.8.1
(Arnaud 1996). Errors correspond to the 90 per cent confidence level
for one interesting parameter (χ
2
= 2.7), if not stated otherwise.
2 OBSERVATIONS AND DATA REDUCTION
NuSTAR: NGC 1068 was observed by NuSTAR with its two co-
aligned X-ray telescopes five times. The first three times were in
2012 December: those data are discussed in Bauer et al. (2015).
After that, NGC 1068 was the target of a monitoring campaign
with XMM–Newton composed of four observations, from 2014 July
until 2015 February. NuSTAR observed the source simultaneously
with the third and fourth XMM–Newton pointings. The Level 1 data
products were processed with the NuSTAR Data Analysis Software
(
NUSTARDAS) package (v. 1.3.0). Cleaned event files (level 2 data
products) were produced and calibrated using standard filtering cri-
teria with the
NUPIPELINE task and the latest calibration files available
in the NuSTAR calibration data base (CALDB 20150316). Since no
spectral variation is found within each observation, we decided t o
use time-averaged spectra for each epoch. Background light curves
are constant within each observation and do not present any flares
due to spurious emission. The background levels are perfectly con-
sistent between the three epochs. We co-added data taken in 2012
December, since no variation was found between those three point-
ings. The extraction radii for the source and background spectra
were 1.5 arcmin each. Net exposure times, after this process, can
be found in Table 1, for both focal plane modules A and B. The
two pairs of NuSTAR spectra were binned in order to oversample
the instrumental resolution by at least a factor of 2.5 and to have a
signal-to-noise ratio (SNR) greater than 5 in each spectral channel.
XMM—Newton: the monitoring campaign of NGC 1068 with
XMM–Newton was composed of four ∼40 ks observations, starting
on 2014 July 10 with the EPIC CCD cameras, the pn (Str
¨
uder et al.
2001) and the two MOS (Turner et al. 2001), operated in small win-
dow and thin filter mode. Details of the XMM–Newton observations
and analysis can be found in Bianchi et al. (in preparation). The
resulting net exposure times can be found in Table 1 for the EPIC-
pn. Spectra were binned in order to oversample the instrumental
resolution by at least a factor of 3 and to have no less than 30 counts
in each background-subtracted spectral channel. Cross-calibration
constants between the NuSTAR-FPMA/B and EPIC-pn are within
10 per cent, in agreement with values presented in Madsen et al.
(2015).
NGC 1068 hosts a strong Ultra Luminous X-ray source (ULX)
near its nucleus already studied in Matt et al. (2004). The source (at a
distance of ∼28 arcmin from the AGN) is present in our 2014/2015
XMM data and we find no differences with respect to the properties
discussed in Matt et al. (2004): its contribution to the 4–10 keV
spectrum is constrained to be ≤5 per cent.
3 SPECTRAL ANALYSIS
We start our analysis by checking for variability in the four XMM–
Newton spectra obtained between 2014 July and 2015 February.
In our analysis, we only consider data above 4 keV due to the
strong contribution at lower energies from distant photoionized,
extra-nuclear emission, which will be discussed in Bianchi et al.
(in preparation). No differences are found between the spectra, nor
do we find differences between the new data set and the one taken
in 2000 July, we refer to Bianchi et al. (in preparation) for details.
Applying the model discussed in Matt et al. (2004)fortheoldXMM
observation to the new 5–10 keV pn spectra, we infer that the flux
of the narrow core of the iron K α line is constant within 5 per cent.
The best fit of the four spectra is shown in Fig. 1, no spectral or flux
variations are apparent.
Since no significant changes are found between the XMM spectra,
we used only the observation simultaneous to the high-energy tran-
sient event (ObsId 0740060401: Table 1, Fig. 2) and then included
NuSTAR FPMA and FPMB s pectra from observations performed in
2012 December, 2014 August and 2015 February. The final data set
is therefore comprised of seven spectra: one XMM spectrum taken
in 2014 August and three pairs of NuSTAR spectra taken in 2012,
2014 and 2015. We only considered NuSTAR data above 8 keV be-
cause XMM spectra have higher spectral resolution and higher SNR
in the Fe Kα energy range.
We then apply the best-fitting model discussed in Bauer et al.
(2015) to our XMM+NuSTAR 4–79 keV data set. This model
fits data from 1996 until 2012 from multiple X-ray observato-
ries. The authors found, using
MYTORUS tables (model M2d), that
the reflecting material is composed of three distinct components
with N
H, 1
10
25
cm
−2
, N
H, 2
= (1.5 ± 0.1) × 10
23
cm
−2
and
N
H,3
= (5.0
+4.5
−1.9
) × 10
24
cm
−2
. N
H, 1
is the absorbing column den-
sity along the line of sight. Chandra observations show that the
MNRASL 456, L94–L98 (2016)
at University of Durham on June 15, 2016http://mnrasl.oxfordjournals.org/Downloaded from
L96 A. Marinucci et al.
Tab le 1. Observation log for the NuSTAR and XMM–Newton monitoring of NGC 1068.
Obs. ID Date Exp. time (ks) 4–20 keV count rate (cts s
−1
) 20–80 keV count rate (cts s
−1
)
FPMA FPMB FPMA FPMB
60002030002 2012-12-18 57
60002030004 2012-12-20 48
0.1579 ± 0.0011 0.1494 ± 0.0011 0.0324 ± 0.0006 0.0286 ± 0.0006
60002030006 2012-12-21 19
60002033002 2014-08-18 52 0.1556 ± 0.0018 0.1456 ± 0.0018 0.0417 ± 0.0010 0.0387 ± 0.0010
60002033004 2015-02-05 53 0.1573 ± 0.0018 0.1444 ± 0.0017 0.0329 ± 0.0009 0.0285 ± 0.0008
4–10 keV Count rate (cts s
−1
)
EPIC-pn
0740060201 2014-07-10 44 0.1593 ± 0.0022
0740060301 2014-07-18 39 0.1540 ± 0.0021
0740060401 2014-08-19 37 0.1537 ± 0.0021
0740060501 2015-02-03 37 0.1625 ± 0.0025
Figure 1. Best fit of the four EPIC-pn spectra, with residuals. Black, red,
green and blue data points indicate observations performed on 2014 July 10,
July 18, August 19 and 2015 February 3, respectively.
spectral features attributed to the N
H, 1
and N
H, 2
components arise
from the central 2 arcsec only, while N
H, 3
corresponds to regions
outside the central 2 arcsec. N
H, 1
and N
H, 3
contribute primarily to
the Compton hump, while N
H, 2
and N
H, 3
provide dominant contri-
butions to the Fe K lines (Bauer et al. 2015).
For this analysis, the normalizations of the absorbed, scattered
and line emission in
MYTORUS tables were kept tied together (coupled
reprocessor solution; Yaqoob 2012): further details about this solu-
tion can be found in Bauer et al. (2015). Applying their best-fitting
model to our data set, we find χ
2
/dof = 1166/896 = 1.30 because
the model does not reproduce well the 2014 August NuSTAR data
(Fig. 2, left-middle panel). Indeed, Fig. 2 and Table 1 show that
there is a 32 ± 6 per cent flux increase above 20 keV and a clear
additional spectral feature in the NuSTAR observation taken in 2014
August.
We first model this excess by leaving the column density N
H, 1
(along the line of sight) free to vary in the 2014 spectra, which
results in a significant improvement of the fit (χ
2
= 177 for
one additional free parameter). A marginal improvement is found
when we also leave the normalization of the primary component
free to vary (χ
2
= 10 for one additional free parameter), for
afinalχ
2
/dof = 979/894 = 1.09; no strong residuals are seen
Figure 2. Best-fitting models, spectra and residuals are shown. The 2014 August EPIC-pn spectrum and the 2012 FPMA/B spectra are plotted as grey circles.
The 2014 FPMA and FPMB data are overplotted in the top-left panel in red and orange, respectively. The new 2015 FPMA and FPMB data are overplotted in
the top-right panel in blue and cyan, respectively. Middle panels show residuals of the two 2014 and 2015 data set to the 2012 best-fitting model (black solid
line). Bottom panels show residuals to a model in which the column density and the nuclear flux are left free in the new observations (red and blue solid lines).
See the text for details.
MNRASL 456, L94–L98 (2016)
at University of Durham on June 15, 2016http://mnrasl.oxfordjournals.org/Downloaded from
Unveiling the nucleus of NGC 1068 L97
Figure 3. Contour plots between the column density along the line of sight
and the intrinsic 2–10 keV luminosity. Grey, red and blue lines represent
contours for 2012, 2014and 2015 data sets, respectively. Boldest to thinnest
contours indicate 99 per cent, 90 per cent and 68 per cent confidence levels.
Vertical orange and purple stripes indicate the 2–10 keV intrinsic luminosi-
ties inferred from the mid-IR and [O
III] observed luminosities with their
relative dispersions from Gandhi et al. (2009) and Lamastra et al. (2009).
throughout the whole 4–79 keV energy band (Fig. 2, bottom
panels). Best-fitting values for the column density along the line of
sight and for the nuclear component normalization are N
H, 1
= (6.7
± 1.0) × 10
24
cm
−2
and A
nucl
= 0.9
+1.0
−0.5
ph cm
−2
s
−1
keV
−1
at
1 keV, respectively. This normalization leads to an unabsorbed
2–10 keV luminosity L
X
= 7
+7
−4
× 10
43
erg s
−1
, which is consistent
with the value presented in Bauer et al. (2015), within the error bars
(Fig. 3). The intrinsic luminosity presented in Bauer et al. (2015)
is L
X
= 2.2 × 10
43
erg s
−1
and indeed, the authors state that it is a
factor of ∼1.6 lower than the one derived from the mid-IR to X-ray
relation in Gandhi et al. (2009) (which is the orange vertical stripe in
Fig. 3). However, we note that at high column densities, the derived
intrinsic luminosity is highly dependent on the (unknown) geome-
try of the absorber (e.g. Matt, Pompilio & La Franca 1999). The fit
does not improve if we leave the normalization and column density
of the 2015 spectra free to vary, indicating that there is no difference
between the 2012 and 2015 data sets. Residuals around 25–30 keV
in both 2014 August and 2015 February observations (Fig. 2, middle
and bottom panels) may be ascribed to residual instrumental fea-
tures in the NuSTAR ARFs (see figs 7 and 8; Madsen et al. 2015). If
we include the instrumental background emission lines between 22
and 35 keV, no significant variations in the best-fitting parameters
are found. This effect represents ∼2 per cent of the total 20–80 keV
flux in the 2014 August observation and ∼7 per cent of the observed
flux excess, in the 20–80 keV energy band. Fig. 3 shows the contour
plots between the column density along the line of sight (the one di-
rectly obscuring the primary continuum) and the intrinsic 2–10 keV
nuclear luminosity extrapolated from the de-absorbed best fit from
the primary continuum for the three NuSTAR data sets (colours as
in Fig. 2). The inferred X-ray luminosity L
X
∼ 10
43
–10
44
erg s
−1
is almost four orders of magnitude greater than usually observed in
ULXs (Swartz et al. 2004; Walton et al. 2011): the lack of variations
below 10 keV and the sharp cutoff in ULX spectra above 20 keV
(Bachetti et al. 2013; Walton et al. 2013, 2014, 2015) lead us to
conclude that this transient excess cannot be attributed to the ULX
in the NGC 1068 FOV.
We then considered the possibility that the high-energy excess in
the 2014 August observation might be due to a rise in the reflected
emission only, due to matter with N
H
∼ 10
25
cm
−2
. The only way
to have a variation in the Compton hump without an associated
variation in the iron line is for the reflector to be almost completely
self-obscured. Indeed, fits show that the inclination angle of the
reflector would have to be larger than 87
◦
, assuming a toroidal
configuration. We therefore conclude that this interpretation, while
not impossible, is unlikely.
4 DISCUSSION
We interpret the high-energy excess detected in the 2014 August
NuSTAR spectra as the first unveiling event ever observed in NGC
1068, in which there is a drop in the column density along our line
of sight. If we take into account the mid-IR and [O
III] luminosities
as proxies of the intrinsic nuclear luminosity, we have additional
pieces of information to add to the contour plots shown in Fig. 3.
The vertical orange lines represent the intrinsic 2–10 keV lumi-
nosity inferred from the mid-IR luminosity (Gandhi et al. 2009).
Purple lines indicate the intrinsic 2–10 keV luminosity calculated
from the extinction-corrected [O
III] luminosity (Marinucci et al.
2012b) using the [O
III]–X-ray relation from Lamastra et al. (2009).
Contour plots show that the intrinsic X-ray luminosity for the three
observations is consistent with those inferred using other proxies,
and all the spectral difference can be attributed to a change in the
absorbing column density, from N
H
> 8.5 × 10
24
cm
−2
in the 2012
observation to N
H, 1
= (5.9 ± 0.4) × 10
24
cm
−2
in 2014 (Fig. 3,
using 90 per cent confidence level for two interesting parameters).
Assuming the bolometric correction from Marconi et al. (2004),
we infer L
bol
= 2.1
+3.2
−1.4
× 10
45
erg s
−1
, in agreement with H
¨
onig,
Prieto & Beckert (2008). The black hole mass of NGC 1068 is esti-
mated to be 1 × 10
7
M
(Greenhill et al. 1996; Lodato & Bertin
2003). For consistency with the mid-IR and [O
III] luminosities
(Fig. 3), we take the lower value L
bol
= 7 × 10
44
erg s
−1
, leading
to an accretion rate η
edd
∼ 0.55, confirming the highly accreting
nature of the source.
Absorption variability is common when observations performed
months to years apart are compared (Risaliti, Elvis & Nicastro
2002), and has been found on time-scales of hours to days in sev-
eral sources. However, even in the so-called changing-look AGN
(sources that switched from the Compton-thick to the Compton-thin
state and vice versa) an eclipsing/unveiling event affecting only the
spectrum above 10 kev has never been observed: we emphasize
that this is the first time that a Compton-thick unveiling event of
this kind has been reported. We note that this is different from the
intrinsic variability recently reported for the Compton-thick AGN
NGC 4945 (Puccetti et al. 2014). Our finding is supporting a clumpy
structure of the obscuring material along the line of sight (Nenkova
et al. 2008).
In this scenario, we do not have a single, monolithic obscuring
wall, but the total column density along the line of sight is the
sum of the contributions from a discrete number of clouds. The
NuSTAR sensitivity above 10 keV allowed us to infer only a lower
limit on the column density variation (N
H
2.5 × 10
24
cm
−2
)
but greater changes could have occurred (top-left corner of Fig. 3:
the parameter space with N
H
> 8.5 × 10
24
cm
−2
) but were not
measurable with our data. Further monitoring of NGC 1068 could
provide constraints on the number of clouds and their distance from
the illuminating s ource.
MNRASL 456, L94–L98 (2016)
at University of Durham on June 15, 2016http://mnrasl.oxfordjournals.org/Downloaded from
L98 A. Marinucci et al.
5 CONCLUSIONS
We presented a spectral analysis of the 4–79 keV NuSTAR and
XMM–Newton monitoring campaign of NGC 1068 obtained be-
tween 2014 July and 2015 February. We found a clear transient
excess above 20 keV in the 2014 August NuSTAR observation,
while no variations are found in the XMM data below 10 keV. The
most plausible explanation is an unveiling event, in which for a
short while the total absorbing column, probably composed by a
number of individual clouds, became less thick so as to permit to
the nuclear radiation to pierce through it. Our result provides fur-
ther evidence that the obscuring material along our line of sight is
clumpy, and enables us to infer a 2–10 keV intrinsic luminosity of
L
X
= 7
+7
−4
× 10
43
erg s
−1
.
ACKNOWLEDGEMENTS
We thank the referee for her/his comments. AM, SB and GM
acknowledge financial support from Italian Space Agency un-
der grant ASI/INAF I/037/12/0-011/13. FEB acknowledges sup-
port from CONICYT-Chile (PFB-06/2007, FONDECYT 1141218,
AC T1101), and grant IC120009, awarded to The Millennium In-
stitute of Astrophysics, MAS. WNB acknowledges Caltech NuS-
TAR subcontract 44A-1092750. PG thanks STFC for support (grant
reference ST/J003697/1). This work was supported under NASA
Contracts No. NNG08FD60C, NNX10AC99G, NNX14AQ07H and
made use of data from the NuSTAR mission, a project led by the
California Institute of Technology, managed by the Jet Propulsion
Laboratory, and funded by the National Aeronautics and Space
Administration. We thank the NuSTAR Operations, Software and
Calibration teams for support with the execution and analysis of
these observations. This research has made use of the NuSTAR
Data Analysis Software (
NUSTARDAS) jointly developed by the ASI
Science Data Center (ASDC, Italy) and the California Institute of
Technology (USA).
REFERENCES
Antonucci R. R. J., Miller J. S., 1985, ApJ, 297, 621
Arnaud K. A., 1996, in Jacoby G. H., ed., ASP Conf. Ser. Vol. 101, As-
tronomical Data Analysis Software and Systems V. Astron. Soc. Pac.,
San Francisco, p. 17
Bachetti M. et al., 2013, ApJ, 778, 163
Bauer F. E. et al., 2015, ApJ, 812, 116
Bianchi S., Piconcelli E., Chiaberge M., Bail
´
on E. J., Matt G., Fiore F., 2009,
ApJ, 695, 781
Bianchi S., Maiolino R., Risaliti G., 2012, Adv. Astron., 2012, 17
Brinkman A. C., Kaastra J. S., van der Meer R. L. J., Kinkhabwala A., Behar
E., Kahn S. M., Paerels F. B. S., Sako M., 2002, A&A, 396, 761
Colbert E. J. M., Weaver K. A., Krolik J. H., Mulchaey J. S., Mushotzky
R. F., 2002, ApJ, 581, 182
Elvis M., Risaliti G., Nicastro F., Miller J. M., Fiore F., Puccetti S., 2004,
ApJ, 615, L25
Gandhi P., Horst H., Smette A., H
¨
onig S., Comastri A., Gilli R., Vignali C.,
Duschl W., 2009, A&A, 502, 457
Ghisellini G., Haardt F., Matt G., 1994, MNRAS, 267, 743
Greenhill L. J., Gwinn C. R., Antonucci R., Barvainis R., 1996, ApJ, 472,
L21
Guainazzi M., Molendi S., Vignati P., Matt G., Iwasawa K., 2000, New
Astron., 5, 235
Guainazzi M., La Parola V., Miniutti G., Segreto A., Longinotti A. L., 2012,
A&A, 547, A31
H
¨
onig S. F., Prieto M. A., Beckert T., 2008, A&A, 485, 33
Iwasawa K., Fabian A. C., Matt G., 1997, MNRAS, 289, 443
Koyama K., Inoue H., Tanaka Y., Awaki H., Takano S., Ohashi T., Matsuoka
M., 1989, PASJ, 41, 731
Lamastra A., Bianchi S., Matt G., Perola G. C., Barcons X., Carrera F. J.,
2009, A&A, 504, 73
Lodato G., Bertin G., 2003, A&A, 398, 517
Madsen K. K. et al., 2015, ApJS, 220, 8
Marconi A., Risaliti G., Gilli R., Hunt L. K., Maiolino R., Salvati M., 2004,
MNRAS, 351, 169
Marinucci A., Risaliti G., Wang J., Nardini E., Elvis M., Fabbiano G.,
Bianchi S., Matt G., 2012a, MNRAS, 423, L6
Marinucci A., Bianchi S., Nicastro F., Matt G., Goulding A. D., 2012b, ApJ,
748, 130
Matt G., Brandt W. N., Fabian A. C., 1996, MNRAS, 280, 823
Matt G. et al., 1997, A&A, 325, L13
Matt G., Pompilio F., La Franca F., 1999, New Astron., 4, 191
Matt G., Bianchi S., Guainazzi M., Molendi S., 2004, A&A, 414, 155
Nenkova M., Sirocky M. M., Nikutta R., Ivezi
´
c
ˇ
Z., Elitzur M., 2008, ApJ,
685, 160
Netzer H., 2015, ARA&A, 53, 365
Pounds K., Vaughan S., 2006, MNRAS, 368, 707
Puccetti S., Fiore F., Risaliti G., Capalbi M., Elvis M., Nicastro F., 2007,
MNRAS, 377, 607
Puccetti S. et al., 2014, ApJ, 793, 26
Risaliti G., Elvis M., Nicastro F., 2002, ApJ, 571, 234
Risaliti G., Elvis M., Fabbiano G., Baldi A., Zezas A., 2005, ApJ, 623, L93
Rivers E. et al., 2015a, ApJ, 804, 107
Rivers E. et al., 2015b, preprint (arXiv:1511.01951)
Str
¨
uder L. et al., 2001, A&A, 365, L18
Swartz D. A., Ghosh K. K., Tennant A. F., Wu K., 2004, ApJS, 154, 519
Tully R. B., 1988, Nearby Galaxies Catalog. Cambridge Univ. Press,
Cambridge
Turner M. J. L. et al., 2001, A&A, 365, L27
Ueno S., Mushotzky R. F., Koyama K., Iwasawa K., Awaki H., Hayashi I.,
1994, PASJ, 46, L71
Walton D. J., Roberts T. P., Mateos S., Heard V., 2011, MNRAS, 416, 1844
Walton D. J. et al., 2013, ApJ, 779, 148
Walton D. J. et al., 2014, ApJ, 793, 21
Walton D. J. et al., 2015, ApJ, 806, 65
Yaqoob T., 2012, MNRAS, 423, 3360
Young A. J., Wilson A. S., Shopbell P. L., 2001, ApJ, 556, 6
This paper has been typeset from a T
E
X/L
A
T
E
X file prepared by the author.
MNRASL 456, L94–L98 (2016)
at University of Durham on June 15, 2016http://mnrasl.oxfordjournals.org/Downloaded from
![Figure 3. Contour plots between the column density along the line of sight and the intrinsic 2–10 keV luminosity. Grey, red and blue lines represent contours for 2012, 2014and 2015 data sets, respectively. Boldest to thinnest contours indicate 99 per cent, 90 per cent and 68 per cent confidence levels. Vertical orange and purple stripes indicate the 2–10 keV intrinsic luminosities inferred from the mid-IR and [O III] observed luminosities with their relative dispersions from Gandhi et al. (2009) and Lamastra et al. (2009).](/figures/figure-3-contour-plots-between-the-column-density-along-the-3iv7kvr0.webp)