s00159-017-0102-9

Astron Astrophys Rev (2017) 25:2
https://doi.org/10.1007/s00159-017-0102-9
REVIEW ARTICLE
Active galactic nuclei: what’s in a name?
P. Padovani1 · D. M. Alexander2 · R. J. Assef3 · B. De Marco4,5 ·

P. Giommi6 · R. C. Hickox7 · G. T. Richards8 · V. Smolčić9 ·
E. Hatziminaoglou1 · V. Mainieri1 · M. Salvato4
Received: 12 June 2017 / Published online: 23 August 2017

© Springer-Verlag GmbH Germany 2017
Abstract Active galactic nuclei (AGN) are energetic astrophysical sources powered
by accretion onto supermassive black holes in galaxies, and present unique observa-
tional signatures that cover the full electromagnetic spectrum over more than twenty
orders of magnitude in frequency. The rich phenomenology of AGN has resulted in a
large number of different “flavours” in the literature that now comprise a complex and
confusing AGN “zoo”. It is increasingly clear that these classifications are only par-
B P. Padovani
ppadovan@eso.org
http://www.eso.org/∼ppadovan
D. M. Alexander
d.m.alexander@durham.ac.uk
R. J. Assef
roberto.assef@mail.udp.cl
B. De Marco
bdemarco@camk.edu.pl
P. Giommi
paolo.giommi@asdc.asi.it
R. C. Hickox
ryan.c.hickox@dartmouth.edu
G. T. Richards
gtr@physics.drexel.edu
V. Smolčić
vs@phy.hr
E. Hatziminaoglou
ehatzimi@eso.org
V. Mainieri
vmainier@eso.org
M. Salvato
mara@mpe.mpg.de
123
2 Page 2 of 91 P. Padovani et al.
tially related to intrinsic differences between AGN and primarily reflect variations in a
relatively small number of astrophysical parameters as well the method by which each
class of AGN is selected. Taken together, observations in different electromagnetic
bands as well as variations over time provide complementary windows on the physics
of different sub-structures in the AGN. In this review, we present an overview of
AGN multi-wavelength properties with the aim of painting their “big picture” through
observations in each electromagnetic band from radio to γ -rays as well as AGN vari-
ability. We address what we can learn from each observational method, the impact of
selection effects, the physics behind the emission at each wavelength, and the potential
for future studies. To conclude, we use these observations to piece together the basic
architecture of AGN, discuss our current understanding of unification models, and
highlight some open questions that present opportunities for future observational and
theoretical progress.
Keywords Galaxies: active · Quasars: supermassive black holes · Radio continuum:

galaxies · Infrared: galaxies · X-rays: galaxies · Gamma-rays: galaxies
1 The active galactic nuclei zoo
The discovery of quasars (Schmidt 1963) opened up a whole new branch of astronomy
(e.g. D’Onofrio et al. 2012; Kellermann 2015, for historical details). Twenty years
earlier Seyfert (1943) had reported the presence of broad and strong emission lines in
the nuclei of six spiral nebulae (including some by now “classical” AGN, like NGC
1068 and NGC 4151). However, his work remained largely ignored until Baade and
Minkowski (1954) pointed out the similarities between the spectra of the galaxies
studied by Seyfert and that of the galaxy they had associated with the Cygnus A radio
source.
As implicit in the name, AGN are stronger emitters than the nuclei of “normal”
galaxies. This “extra” component is unrelated to the nuclear fusion powering stars
and is now universally accepted to be connected instead to the presence of an actively
accreting central supermassive (106 M ) black hole (SMBH).
1 European Southern Observatory, Karl-Schwarzschild-Str. 2, 85748 Garching bei München,

Germany
2 Department of Physics, Centre for Extragalactic Astronomy, Durham University, Durham, UK
3 Núcleo de Astronomía de la Facultad de Ingeniería, Universidad Diego Portales, Santiago, Chile
4 Max Planck Institute for Extraterrestrial Physics, Garching bei München, Germany
5 Present Address: Nicolaus Copernicus Astronomical Center, 00-716 Warsaw, Poland
6 Italian Space Agency, ASI, via del Politecnico snc, 00133 Rome, Italy
7 Department of Physics and Astronomy, Dartmouth College, Hanover, NH, USA
8 Department of Physics, Drexel University, Philadelphia, PA, USA
9 Department of Physics, Faculty of Science, University of Zagreb, Bijenička cesta 32, 10000 Zagreb,
Croatia
123
Active galactic nuclei: what’s in a name? Page 3 of 91 2
Fig. 1 A schematic representation of an AGN spectral energy distribution (SED), loosely based on the
observed SEDs of non-jetted quasars (e.g. Elvis et al. 1994; Richards et al. 2006a). The black solid curve
represents the total emission and the various coloured curves (shifted down for clarity) represent the individ-
ual components. The intrinsic shape of the SED in the mm-far infrared (FIR) regime is uncertain; however,
it is widely believed to have a minimal contribution (to an overall galaxy SED) compared to star formation
(SF), except in the most intrinsically luminous quasars and powerful jetted AGN. The primary emission
from the AGN accretion disk peaks in the UV region. The jet SED is also shown for a high synchrotron
peaked blazar (HSP, based on the SED of Mrk 421) and a low synchrotron peaked blazar (LSP, based on
the SED of 3C 454.3; see Sect. 6.1). Adapted from Harrison (2014). Image credit: C. M. Harrison
AGN have many interesting properties, including the following: (1) very high
luminosities (up to L bol ≈ 1048 erg s−1 ), which make them the most powerful non-
explosive sources in the Universe and, therefore, visible up to very high redshifts
(currently z = 7.1: Mortlock et al. 2011); (2) small emitting regions in most bands,
of the order of a milliparsec, as inferred from their rapid variability (e.g. Ulrich et al.
1997), implying high energy densities; (3) strong evolution of their luminosity func-
tions (LFs; e.g. Merloni and Heinz 2013); (4) detectable emission covering the whole
electromagnetic spectrum (this review).
The latter property means that AGN are being discovered in all spectral bands.
Different methods are employed in different bands to identify these sources but,
most importantly, the various wavelength regimes provide different windows on AGN
physics. Namely, the infrared (IR) band is mostly sensitive to obscuring material and
dust, the optical/ultraviolet (UV) band is related to emission from the accretion disk,
while the X-ray band traces the emission of a (putative) corona. γ -ray and (high flux
density) radio samples, on the other hand, preferentially select AGN emitting strong
non-thermal (jet [or associated lobe] related) radiation (see Fig. 1).1 The surface den-
1 The mm/sub-mm band is missing from this paper because it mostly probes molecular gas that resides
in the AGN host galaxy. However, with the high-resolution capabilities of the Atacama Large Millime-
ter/submillimeter Array (ALMA) we are starting to resolve the innermost parts of AGN down to parsec
scales (Sect. 8.4).
123
sities of AGN also vary strongly across the electromagnetic spectrum due to a complex
mix of physical processes, selection effects, and technological limitations (as shown
in Sect. 8).
The past years have seen a proliferation of AGN classes, which outsiders to the
field (but insiders as well!) find mesmerising. A (possibly incomplete) list is given in
Table 1, which gives the class or acronym in col. (1), its meaning in col. (2), and the
main properties or a reference to a relevant paper in col. (3).
Reality is much simpler, however, as we know that most of these seemingly different
classes are due to changes in only a small number of parameters, namely orientation
(e.g. Antonucci 1993; Urry and Padovani 1995; Netzer 2015), accretion rate (e.g.
Heckman and Best 2014), the presence (or absence) of strong jets (e.g. Padovani
2016), and possibly the host galaxy and the environment. Sorting out these issues is a
pre-requisite to understand AGN physics and the role AGN play in galaxy evolution
(e.g. Alexander and Hickox 2012).
To go beyond taxonomy and paint the AGN “big picture”, which comes out of multi-
wavelength surveys, and understand the truly intrinsic and fundamental properties of
AGN, the workshop “Active Galactic Nuclei: what’s in a name?” was organised at
ESO, Garching, between June 27 and July 1, 2016. This was done by discussing AGN
selection and physics in all bands and by addressing the following:
– the different types of AGN selected in the various spectral bands;
– the similarities and differences they display;
– the impact of selection effects on the interpretation of the results;
– the physical mechanism(s) behind emission in a given band;
– the effective range of black hole (BH) mass (MBH ) and Eddington ratios2 (L/L Edd )
probed by each selection method;
– the possible limitations of current observations and/or facilities.
The workshop consisted of seven different sessions: radio, IR, optical, X-ray, γ -ray,
variability, and multi-frequency. All of the sessions (with the exception of the multi-
frequency one) were introduced by a review talk which set the scene, followed by
contributed talks, for a total of 86 speakers, 48% of whom were women. Sixty-seven
posters completed the programme. A summary talk and a discussion session were
held at the end of the workshop.3 The workshop was very well attended, with 165
participants, covering 5 continents and 31 different countries; 60 of the participants
were students. This review was inspired by the workshop.
In this paper we review our progress in addressing these key issues by discussing
radio-, IR-, optical-, X-ray-, γ -ray-, and variability-selected AGN (Sects. 2–7). Sec-
tion 8 summarizes our understanding of AGN and examines some open issues.
Throughout this paper, spectral indices are defined by Sν ∝ ν −α and the values
H0 = 70 km s−1 Mpc−1 , m = 0.27, and = 0.73 have been used. The acronyms
of the AGN classes mentioned in this paper are defined in Table 1.
2 The ratio between the observed luminosity and the Eddington luminosity, L
Edd = 1.3 ×
1046 (M/108 M ) erg s−1 , where M is one solar mass. This is the maximum isotropic luminosity a
body can achieve when there is balance between radiation pressure (on the electrons) and gravitational
force (on the protons).
3 Most presentations and posters can be found at http://www.eso.org/sci/meetings/2016/AGN2016.html.
123
Table 1 The AGN zoo: list of AGN classes
The top part of the table relates to major/classical classes. The last column describes the main properties.
When these are too complex, it gives a reference to the first paper, which defined the relevant class or, when
preceded by “see”, a recent paper, which gives up-to-date details on it. Reference key: 1. Padovani (2017);
2. Fanaroff and Riley (1974); 3. Giommi et al. (2012); 4. Weymann et al. (1981); 5. Ghisellini (2010); 6.
O’Dea et al. (1991); 7. Padovani and Giommi (1995); 8. Laing et al. (1994); 9. Heckman and Best (2014);
10. Ho (2008); 11. Osterbrock and Pogge (1985); 12. Sulentic et al. (2002); 13. Osterbrock (1981)
123
2 Radio-selected AGN
In this section we describe radio-selected AGN. The types of, and physical mechanisms
powering, radio-AGN are outlined in Sect. 2.1. Selection effects are discussed in
Sect. 2.2, while the evolution of radio-AGN and future prospects are addressed in
Sects. 2.3 and 2.4, respectively. Typical observing frequencies in the radio regime
range from about 10 MHz (λ ∼ 30 m) up to a few tens of GHz (e.g. 30 GHz, λ = 1 cm).
For other recent reviews of radio-selected AGN we refer to Heckman and Best (2014),
Padovani (2016), Tadhunter (2016) and Smolčić (2016).
2.1 Physical mechanism(s) and types of AGN selected in the radio
The dominant emission process in the radio band is synchrotron emission, i.e. radi-
ation by charged particles gyrating at relativistic velocities through magnetic fields.4
Being non-thermal in origin, this emission is usually parametrized by a power law
of the form Sν ∝ ν −α where Sν is the flux density [expressed in Jy, mJy, etc.] at
frequency ν, and α is the spectral index. Supernova remnants and processes related to
the central SMBH are the main sources of synchrotron radiation in galaxies, resulting
in two dominant galaxy populations identified in extragalactic radio continuum sur-
veys, namely star forming galaxies (SFGs) and AGN (e.g. Miley 1980; Condon 1992).
The former, being intrinsically weaker radio sources, become more prominent at the
faintest radio flux densities (e.g. Wilman et al. 2008; Padovani et al. 2015; Padovani
2016; Smolčić et al. 2017a). We outline here the main classes of radio emitting
AGN.
2.1.1 Radio galaxies
Fanaroff and Riley (1974) differentiated radio galaxies (RGs) into type I and II based
on the ratio of the separation of the highest surface brightness regions on opposite
sides of the central galaxy, and the extent of the source measured from the lowest
surface brightness contour5 (see Fig. 2). These classes are commonly referred to
in the literature as FR I (ratio <0.5) and FR II (ratio >0.5) RGs, where the first
(second) are often described as core- (edge-) brightened. They also separate in the
radio luminosity vs. optical host galaxy luminosity plane (Ledlow and Owen 1996;
see also Fig. 11 in Buttiglione et al. 2010). Recently a third class, FR 0, was suggested
by Baldi et al. (2015, see also Baldi and Capetti 2010), which includes RGs sharing the
properties of FR Is but lacking prominent extended radio emission, being a factor of
∼30 more core-dominated. Further morphological classes often found in the literature
refer to bent or warped appearances of the RGs (e.g. wide-angle tail, narrow-angle
4 Free-free emission originating in H II regions may substantially contribute to the overall radio spectrum
of SFGs. This, however, is expected to occur at rest-frame frequencies higher than ∼20 GHz (e.g. Fig. 1
in Condon 1992). The contribution of this emission process to the overall radio spectrum is taken to be
negligible for galaxies dominated in the radio regime by an AGN.
5 Any compact component coincident with the central galaxy was not taken into account.
123
Fig. 2 Examples of FR I (panel

‘a’; 3C 449; Perley et al. 1979),
FR II (panel ‘b’; 3C 175; Bridle
et al. 1994), wide-angle tail
(panel ‘c’; 3C83.1B; Owen et al.
1978), and X-shaped (panel ‘d’;
3C 223.1; Dennett-Thorpe et al.
2002) RGs
tail, X-shaped RGs; see Fig. 2; see also Miley 1980 for a review, and Figs. 6 and 7
therein).
2.1.2 Flat- and steep-spectrum sources
When multiple radio continuum frequency observations are available, radio sources
are usually separated into two main groups of steep (α ≥ 0.5) and flat (α < 0.5) radio
spectrum sources (e.g. Wall 1975; Peacock and Gull 1981; Willott et al. 2001; Kim-
ball and Ivezić 2008), which roughly correspond, with some exceptions, to extended
and compact sources. They can further be classified as, e.g. USS (e.g. De Breuck
et al. 2001), GPS, and CSS sources (see Sadler 2016 and O’Dea 1998 for reviews).
FSRQs and BL Lacs, both of which are flat spectrum sources, make up the blazar
class, which includes AGN hosting jets oriented at a very small angle (15−20◦ )
with respect to the line of sight (e.g. Urry and Padovani 1995; Giommi et al. 2013;
see also Sect. 6). The two sub-classes’ main difference lies in their optical spec-
tra, with FSRQs displaying strong, broad emission lines just like standard quasars,
and BL Lacs instead showing at most weak emission lines, sometimes exhibiting
absorption features, and in many cases being completely featureless. Blazars dom-
inate the bright radio (e.g. Padovani 2016) and the γ -ray sky (Sect. 6.1). In this
respect, we note that a strong radio flux density is one of the most efficient (albeit
incomplete) AGN selection criteria: out of the 527 sources with 5 GHz flux den-
sity >1 Jy and |bII | ≥ 10◦ (Kühr et al. 1981) only one, M 82, is not an AGN (and
only two do not belong to the RG, radio quasar, or blazar classes: M 82 and NGC
1068).
123
2.1.3 RL and RQ AGN
One of the most used classifications of radio AGN is their division into RL and RQ
AGN. Initially, this distinction was defined in the context of quasars (MB < −23) with
the threshold between the two classes set either in (1) radio flux density or luminosity
(e.g. Peacock et al. 1986), (2) or the ratio of radio-to-optical flux density or luminosity
(e.g. Schmidt 1970). Radio loudness has since then been defined in different ways
in the literature but in any case these “classical” definitions apply only to type 1
AGN (Padovani et al. 2011; Bonzini et al. 2013). Following Baloković et al. (2012)
the parametrization of radio loudness for type 1 AGN can be summarized as R K =
log L radio − K · log L λ . K = 0 for a simple radio flux density or luminosity threshold
(where L radio may refer to flux density or luminosity measurements at frequencies
∼1−6 GHz in the observed- or rest-frame, respectively; Peacock et al. 1986; Miller
et al. 1990; Ivezić et al. 2002), and K = 1 for a threshold in the logarithm of the ratio
of flux density or luminosity in the radio band and that within a wavelength range Δλ,
which can either be in the optical (e.g. B-band; Kellermann et al. 1989; r-, i-, or z-band;
Ivezić et al. 2002) or IR (e.g. 24 µm; Padovani et al. 2011; Bonzini et al. 2013).
Padovani (2016) (see also Padovani 2017) has argued that the distinction between
these two types of AGN is not simply a matter of semantics but rather that the two
classes represent intrinsically different objects, with RL AGN emitting a large fraction
of their energy non-thermally and in association with powerful relativistic jets, while
the multi-wavelength emission of RQ AGN is dominated by thermal emission, directly
or indirectly related to the accretion disk (see Fig. 1). Moreover, he pointed out that
the “radio-loud/radio-quiet” classes are obsolete, misleading, and inappropriate. Since
the major physical difference between these two classes is the presence or lack of
strong relativistic jets, which also implies that the two classes reach widely different
maximum photon energies (see his Sect. 2.1.2 and Fig. 1), we will be using in this
review the terms “jetted” and “non-jetted” instead of RL and RQ AGN. We discuss
this further in Sect. 8.
Note that, although we know that jetted AGN represent a minority, their exact frac-
tion is still not well determined. The oft-quoted value of ≈15% comes from optically
selected samples of quasars (e.g. Kellermann et al. 1989). Padovani (2011), by inte-
grating the radio LFs of jetted and non-jetted AGN, has suggested a much smaller
fraction (<1%).
2.1.4 Low- and high-excitation AGN
The classification of radio AGN into two main classes based on their optical spec-
troscopic properties goes back to Hine and Longair (1979). They divided the Third
Cambridge Catalogue of Radio Sources (3CR) into objects characterized by strong
emission lines in their spectra ([O ii] λ3727, [O iii] λ5007, [Ne ii] λ3867), and sources
which exhibited either absorption line spectra typical of giant elliptical galaxies
(absorption line galaxies hereafter) or weak [O ii] λ3727 emission lines. Since then this
has been expanded and refined by using also [O iii] λ5007 equivalent widths (EWs; e.g.
Tadhunter et al. 1998) or high/low excitation/ionization emission line criteria (Laing
et al. 1994; Kewley et al. 2006; Buttiglione et al. 2009, 2010; see Fig 3). In general,
123
Fig. 3 Optical spectroscopic diagnostic diagrams (see Kewley et al. 2001; Kauffmann et al. 2003b; Kewley
et al. 2006) that separate emission-line galaxies into star forming, composite galaxies, and Seyfert and
LINER AGN. Small grey dots represent galaxies from the Sloan Digital Sky Survey (SDSS) DR4 “main”
spectroscopic sample. Large open squares (filled dots) denote z < 0.1 Revised Third Cambridge Catalogue
of Radio Sources (3CRR; Laing et al. 1983) RGs independently classified based on their core X-ray
emission as systems with radiatively efficient (inefficient) BH accretion (Evans et al. 2006). © AAS. Figure
reproduced from Smolčić (2009), Fig. 1, with permission
objects without and with high-excitation emission lines in their optical spectra are
referred to as LEGs and HEGs, respectively. The LEG/HEG classification holds not
only for radio selected AGN but applies also to AGN selected in other bands (for
which an optical spectroscopic classification is available). More specifically, quasars
and Seyferts belong to the HEG category, while LINERs and absorption line galaxies
are classified as LEGs (see Fig. 3 and Sect. 4; Baldwin et al. 1981; Veilleux and Oster-
brock 1987; Kewley et al. 2001; Kauffmann et al. 2003a; Kewley et al. 2006. But see
Sarzi et al. 2010 for evidence that the nebular emission of most objects in the LINER
part of Fig. 3 is not powered by an AGN).
As illustrated in Fig. 4 fundamental physical differences between these two types
of radio AGN (at z < 1) have been found (e.g. Nagar et al. 2005; Evans et al. 2006;
Hardcastle et al. 2007; Smolčić et al. 2009; Smolčić 2009; Buttiglione et al. 2010;
Best and Heckman 2012; Padovani et al. 2015; see also Fig. 6 in Smolčić 2016).
Namely, LEGs (LINERs and absorption line systems) have on average redder optical
g −r colours, larger values of the 4000 Å break strength, and higher stellar masses than
HEGs (Seyferts; top panel). LEGs are also shown (bottom panel) to exhibit radiatively
inefficient accretion related to low L/L Edd (0.01), possibly fuelled by the hot phase
of the inter-galactic medium (IGM), and are typically highly efficient in collimated
jet production. HEGs accrete in a radiatively efficient manner, at high Eddington rates
(0.01 L/L Edd 1), are fuelled by the cold IGM phase, and (on average) less
likely to launch collimated jets. From a theoretical aspect, the observed difference
in L/L Edd can be related to the switch between a standard accretion, i.e. radiatively
efficient, geometrically thin (but optically thick) disk accretion flow (Shakura and
Sunyaev 1973), and a radiatively inefficient, geometrically thick (but optically thin)
disk accretion flow (Esin 1997; Narayan et al. 1998). The switch occurs at accretion
rates below a certain L/L Edd (≈0.01; Rees et al. 1982; Narayan and Yi 1994; Meier
2002; Fanidakis et al. 2011) as then the decreased accreting gas density lowers the
123
Fig. 4 From left to right the top panels show the distribution of the observed g − r colour, 4000 Å break
strength, and stellar mass, while the bottom panels show the distribution of the velocity dispersion, σ 4 ,
proportional to BH mass (Tremaine et al. 2002), and log(L [O iii] /σ 4 ) proportional to BH accretion rate (in
Eddington units; Heckman et al. 2004) for various galaxy populations drawn from the FIRST-NVSS-SDSS
sample with 0.04 < z < 0.1 (SFGs: dotted lines and empty histograms; Seyferts: dash-dotted lines ad
diagonally hatched histograms; absorption line galaxies: dashed lines and vertically hatched histograms;
LINERs: full lines and filled histograms). © AAS. Figure reproduced from Smolčić (2009), Figs. 2 and 3,
with permission
cooling rate, and a substantial amount of heat can be carried along, i.e. advected, rather
than irradiated.6
2.1.5 Radio-selected AGN classes and unification
There is some overlap amongst the classes described above. For example, almost
all FR Is are LEGs, while most FR IIs usually display strong emission lines and
would thus be classified as HEGs. However, there is no one-to-one correspondence
between FR class and emission lines as quite a few FR IIs (20%: e.g. Gendre
6 The switch in L/L

Edd is not to be taken as sharp but as a transition in a statistical sense. The fundamental
physical separation of the various AGN types may be a function of more parameters (such as spin and BH
mass) and one should keep in mind that the observational data used to constrain this separation are subject to
measurement and computational uncertainties and biases (e.g. the role of environment in kinetic luminosity
determinations, contamination of selection proxies by stellar, rather than AGN related processes, etc.; see,
e.g. Mingo et al. 2014).
123
et al. 2013) have been found to be LEGs (see also Evans et al. 2006; Buttiglione
et al. 2010; Mingo et al. 2014). Powerful radio quasars and RGs are generally of the
HEG type, while the less powerful (and more common) RGs are mostly LEGs (e.g.
Padovani 2016). “Classical” non-jetted AGN (i.e. Seyferts and quasars) are always
HEGs.
Moreover, radio quasars are intrinsically the same sources as some RGs. Namely
they are simply FR II/HEG RGs with their jets at an angle 45◦ with respect to the
line of sight (Orr and Browne 1982; Barthel 1989; Antonucci 1993; Urry and Padovani
1995). The fact that radio quasars display strong and Doppler broadened lines in their
optical spectra (with full width half maximum [FWHM] 1000 km s−1 ), unlike RGs,
requires also the presence of dust in a flattened configuration roughly perpendicular to
the jet (see also Sect. 3.1). This so-called “unification model” explains in a natural way
why the (projected) sizes of the jets of RGs are larger than those of quasars (Barthel
1989).
With regard to FR I/LEG RGs, obscuration towards their nuclei appears to be much
smaller than that of their FR II/HEG relatives (e.g. Chiaberge et al. 2002; Evans et al.
2006), which indicates that dust might be not present (see also Sect. 3.2.4). This applies
also to the population of FR II/LEG RGs. LEG RGs, therefore, irrespective of their
radio morphology, are “unified” with BL Lacs (e.g. Giommi et al. 2013).
2.1.6 Classification of radio-selected AGN
The classification of radio-selected AGN is a complex matter, especially so for the faint
radio sources routinely studied these days. Getting optical spectra for their counter-
parts is very time consuming, prohibitively so for the very faint tail [R 26] even with
8–10 m class telescopes. But even if we had optical spectra for all radio sources, optical-
based classification is well known to be prone to obscuration biases (Sect. 4.3) and for
this reason spectra need to be complemented with information from multi-wavelength
ancillary data. Multi-wavelength methods used in the literature to identify AGN
detected also in the radio band include the following: (1) rest-frame optical colours of
the host galaxies, which have been shown to correlate with the galaxies spectral emis-
sion line properties (Strateva et al. 2001; Smolčić et al. 2006, 2008; Smolčić 2009;
Ilbert et al. 2010); (2) X-ray luminosity (L X > 1042 erg s−1 ; e.g. Szokoly et al. 2004;
see also Sect. 5.2); (3) mid-IR (MIR) colours (e.g. Lacy et al. 2004; Stern et al. 2005;
Donley et al. 2012; see also Sect. 3.2.1); (4) multiple component SED fitting using sets
of galaxy and AGN templates (Berta et al. 2013; Delvecchio et al. 2014, 2017; Calistro
et al. 2017) (5) excess of radio luminosity relative to a tracer of the SF rate (SFR) in the
host galaxy (e.g. far or total IR luminosity; Condon 1992; Del Moro et al. 2013; Delvec-
chio et al. 2017). A detailed discussion of these methods (and others) can be found in
Padovani (2016).
2.2 Selection effects: SF contribution to the total radio luminosity output
In the local Universe LEGs occupy the red sequence of galaxies, while HEGs are
hosted by bluer galaxies, populating the so-called green valley (see Fig. 4 in Smolčić
2009). As, generally, bluer host galaxy colours imply higher SFRs (e.g. Ilbert et al.
123
2009), in the latter case the observed radio emission could be partially or entirely due
to synchrotron radiation generated by supernova remnants, rather than by the central
SMBH. As the classification of radio AGN often relies on non-radio properties, such
as the optical emission line properties or multi-wavelength proxies (Sect. 2.1.6), an
a posteriori assessment of the (SF- or AGN-related) origin of the observed radio
emission is needed as X-ray, optical, or IR signatures of an accreting central SMBH
do not necessarily imply AGN-related radio emission. Thus, one of the most severe
selection effects in the radio band is the contribution of SF-related processes to the
total radio power7 (especially given the limited angular resolution available in many
radio continuum surveys).
Detailed high angular resolution radio continuum studies often reveal a mixture
of SF- and accretion-related radio emission in non-jetted HEGs (e.g. Chi et al. 2013;
Maini et al. 2016; Herrera Ruiz et al. 2016). For example, Maini et al. (2016) used the
Australian Long Baseline Array to search for compact radio cores in four non-jetted
AGN located in the Extended Chandra Deep Field-South (E-CDFS). They find that
some such sources contain an active AGN that can contribute significantly (>50%)
to the total radio emission. Herrera Ruiz et al. (2016) studied three non-jetted quasars
(MB < −23) in the COSMOS field observed with the Very Large Array (VLA) and
the Very Long Baseline Array (VLBA) at 1.5 and <20 mas resolutions, respectively.
Comparing the core (VLBA) and total (VLA) radio flux densities they infer that
50–75% of the radio emission in these sources is powered by AGN activity. Such
high-angular resolution studies are, however, still limited to rather small samples, and
to-date the main source of radio emission in the most powerful AGN (i.e. Mi < −22)
is still debated. While agreement exists that radio emission in jetted quasars is powered
by SMBH accretion-related processes (e.g. Miller et al. 1990), opposing (statistical)
evidence can be found in the literature related to the origin of radio emission in non-
jetted quasars (the so-called quasar radio loudness dichotomy problem): SF in the host
galaxies (Kimball et al. 2011; Condon et al. 2013), or AGN activity (White et al. 2015;
Zakamska et al. 2016; see also Padovani 2016 for a detailed discussion of this topic).
From a statistical point of view, studies of radio AGN drawn from large radio-
continuum surveys, combined with spectroscopic and/or multi-wavelength data find
that (z < 1) HEGs are much more likely to be associated with SF in their host galaxies,
with SFRs at least a factor 3–4 higher than those in LEGs (Hardcastle et al. 2013; see
also Gürkan et al. 2015). The inferred statistical contribution of SF-related processes
to the observed radio emission in HEGs and LEGs (0.04 < z 0.2) is estimated
to be ∼60% (∼10%) for HEGs (LEGs; Morić et al. 2010). Similarly, Bonzini et al.
(2015) find that the radio luminosities of non-jetted AGN at z ∼ 1.5−2 in the E-CDFS
survey are consistent with the galaxies’ SFRs inferred from their FIR luminosities.
Identifying AGN in a slightly different way, Delvecchio et al. (2017) observe that for
about 70% of their AGN the radio luminosities are consistent (within 3σ from the
average) with those expected based on the SFR in the host galaxies inferred from the
total IR emission (corrected for the AGN component).
7 Note that, while it might be safe to assume that anything above P 24 −1 has nothing
1.4 GHz ≈ 10 W Hz
to do with SF, this is only valid at low redshifts given the strong evolution of SFGs (e.g. Padovani 2016).
123
2.3 Cosmic evolution of radio-selected AGN
Past research has shown that radio AGN evolve via a “downsizing” effect, i.e. low
radio luminosity sources evolve less strongly than high-luminosity ones (e.g. Longair
1966; Willott et al. 2001; Rigby et al. 2015; see also Sect. 5.3 for the X-ray perspec-
tive). Studies of powerful radio AGN (L 1.4 GHz 2 × 1026 W Hz−1 ) have found a
strong positive density evolution at z 2, beyond which their comoving volume den-
sity declines (Dunlop and Peacock 1990; Willott et al. 2001). A substantially slower
evolution, with a lower redshift (z ∼ 1−1.5) comoving volume density turnover, has
been found for weaker radio AGN (L 1.4 GHz > 2 × 1025 W Hz−1 ; Waddington et al.
2001). Studies of even lower luminosity radio AGN (L 1.4 GHz 1025 W Hz−1 ) find a
mild evolution out to z ∼ 1 (e.g. Smolčić et al. 2009; Sadler et al. 2007; Donoso et al.
2009; Padovani et al. 2011; Smolčić et al. 2017b).
The local radio LFs derived separately for HEGs and LEGs have been presented
by, e.g. Filho et al. (2006), Best and Heckman (2012), Gendre et al. (2013), Pracy
et al. (2016). While it is clear that HEGs dominate the volume densities at high radio
luminosities (L 1.4 GHz > 1026 W Hz−1 ; Heckman and Best 2014; Pracy et al. 2016),
the slope of the low-luminosity end of the HEG radio LF is still somewhat unclear.
Best and Heckman (2012) find a significantly flatter slope than Pracy et al. (2016),8
and Filho et al. (2006)9 (see Fig. 8 in Pracy et al. 2016). As discussed in Pracy et al.
(2016) this is likely due to the difficulty of disentangling the real contribution of AGN-
related radio emission in HEGs within the faint radio luminosity regime, dominated
by SFGs. Regardless of the faint end slope, studies out to z ∼ 1 consistently find
that HEGs evolve more rapidly than LEGs. For example, Pracy et al. (2016) find
that their LEG population displays little or no evolution over the observed redshift
+0.17 +0.22
range (0.005 < z < 0.75), evolving as (1 + z)0.06−0.18 [(1 + z)0.46−0.24 ] assuming
pure density [pure luminosity] evolution, while their HEG population evolves more
+0.46 +0.79
rapidly, as (1 + z)2.93−0.47 [(1 + z)7.41−1.33 ] assuming pure density [pure luminosity]
evolution.
Constraining the cosmic evolution of the two dominant AGN types detected in
the radio band beyond redshift 1 (z 6) is not trivial given the difficulty of (1)
separating the two distinct AGN types so that they can quantitatively be related to
the HEG/LEG populations identified via optical spectroscopy at z < 1, (2) isolating
the fraction of radio luminosity arising from the AGN (rather than SF in the host
galaxy), which is challenging even in the lower (z < 1) redshift Universe. Recently,
Padovani et al. (2015) have constrained the cosmic evolution of their selected non-
jetted and jetted AGN identified in the E-CDFS survey out to z ∼ 4. These two
populations can qualitatively be roughly related to the HEG and LEG populations,
respectively (see Bonzini et al. 2013; Padovani et al. 2015). Padovani et al. find a
strong evolution, similar to that for SFGs, of their non-jetted sample throughout the
8 Derived for FIRST-SDSS AGN at z < 0.3 from the ∼900 deg2 LARGESS survey and spectroscopically
classified.
9 Derived for Seyfert galaxies using 1 angular resolution radio continuum data; see also Padovani et al.
(2015).
123
redshift range probed, and a peak at z ∼ 0.5 in the number density of their jetted AGN,
with a decline at higher redshifts. The first results using the VLA-COSMOS 3 GHz
Large Project (Ceraj et al., in prep.), combined with the COSMOS multi-wavelength
data (Laigle et al. 2016; Marchesi et al. 2016), yield a stronger cosmic evolution for
AGN with the highest bolometric (radiative) luminosities throughout the entire redshift
range (z 6; and qualitatively broadly consistent with HEG samples), relative to that
of the AGN sample with lower bolometric (radiative) luminosities (and qualitatively
broadly consistent with LEG samples).10 We stress, however, that both the E-CDFS
and COSMOS fields are not large enough to constrain the broad radio luminosity
range, encompassing also the rare, highest luminosity AGN (detectable in shallower,
wide-area surveys). Hence, for a full, quantitative assessment of the evolution of the
AGN radio LF separated into the two types out to high redshift and over a broad
luminosity range a combination of surveys with various areal coverages is needed,
with access to a robust AGN classifier, and methods to isolate AGN-related radio
emission in (low radio luminosity) AGN, that can be uniformly applied throughout
the entire redshift range considered.
2.4 The future of AGN studies in the radio band
Studies such as those discussed in the previous section are becoming feasible only
now, and will be invigorated with the onset of the Square Kilometre Array (SKA11 ),
offering an observing window between 50 MHz and 20 GHz extending well into
the nanoJy regime with unprecedented versatility, in combination with contempo-
raneous projects over the entire electromagnetic spectrum (such as, e.g., the Large
Synoptic Survey Telescope [LSST], Euclid, the Extended ROentgen Survey with
an Imaging Telescope Array (eROSITA), and the James Webb Space Telescope
[JWST]).
A revolution has in fact started in radio astronomy, which has entered an era of
large area surveys reaching flux density limits well below current ones. The Jansky
Very Large Array (JVLA12 ), the LOw Frequency ARray (LOFAR13 ), the Murchison
Widefield Array,14 are already taking data, and are being joined by the Australian
Square Kilometre Array Pathfinder (ASKAP15 ), MeerKAT,16 e-MERLIN,17 and A-
10 The former class has been selected through X-ray, IR, and SED-criteria, while the latter has been
identified via >3σ radio-excess relative to the host galaxies’ IR-based SFRs, and red rest-frame optical
colours, and lacking X-ray, IR, and SED-based signatures of AGN activity (Smolčić et al. 2017a; Delvecchio
et al. 2017). In this study the SF related contribution to the total radio luminosity was statistically subtracted.
11 http://www.skatelescope.org.
12 science.nrao.edu/facilities/vla.
13 http://www.astron.nl/radio-observatory/astronomers/lofar-astronomers.
14 http://www.mwatelescope.org.
15 http://www.atnf.csiro.au/projects/askap/.
16 http://www.ska.ac.za/meerkat.
17 http://www.e-merlin.ac.uk.
123
PERTIF.18 These projects will survey the sky vastly faster than it is possible with
existing radio telescopes producing surveys covering large areas of the sky down to
fainter flux densities than presently available, as fully detailed in Norris et al. (2013).
This, amongst other things, will revolutionise AGN studies. The Evolutionary Map
of the Universe (EMU; Norris et al. 2011), one of the ASKAP surveys, for example,
is expected to detect ∼70 million sources, about half of which will likely be AGN
unaffected by the problems of obscuration, which plague the optical (Sect. 4.3) and
soft X-ray (Sect. 5.2) bands. Identifying AGN in these new radio surveys, however,
will not be straightforward and will require many synergies with facilities in other
bands (e.g. Padovani 2016).
3 Infrared-selected AGN
IR studies have had a strong impact on our understanding of AGN structure, their
evolution through cosmic time, and their role in galaxy evolution. In Sect. 3.1 we
discuss the mechanisms that give rise to the IR emission in AGN and the advantages
of AGN identification in the MIR wavelengths. In Sect. 3.2 we examine in detail how
the MIR selection of AGN works and what the characteristics of the selected samples
are. In Sect. 3.3 and 3.4 we explore additional AGN selection criteria that rely on IR
observations. Finally, in Sect. 3.5 we analyse the future of AGN studies in the IR in
light of upcoming facilities. For the purpose of this section, we divide the IR SED of
AGN in three wavelength regimes: the near-IR (NIR; 1−3 µm), the MIR (3−50 µm),
and the FIR (50−500 µm).
3.1 Physical mechanism behind IR emission
Despite many drawbacks, the “dusty torus” paradigm has been quite successful in
explaining the appearance of a wide variety of AGN. The basis of this paradigm is the
presence of dust surrounding the accretion disk on scales larger than that of the broad
line region (BLR), with an inner boundary set by the sublimation temperature of the
dust grains (Barvainis 1987). This dust reprocesses the emission of the accretion disk
into the IR and dominates the AGN SED from wavelengths longer than ∼1 µm up to
a few tens of micron (see Fig. 1). It plays a fundamental role in the AGN unification
scheme (Antonucci 1993; Urry and Padovani 1995, see also Sect. 2.1.5), as through
polarisation studies it was established that the difference between type 1 and 2 AGN
is simply an effect of orientation with respect to the dust. In type 2 AGN the dust
obscures the line of sight towards the accretion disk and the BLR and therefore only
narrow emission lines can be observed in the optical spectrum (e.g. Antonucci and
Miller 1985; Antonucci 1993, although see Elitzur and Netzer 2016 for a discussion
about possible real type 2 AGN where the difference is not caused by dust obscuration;
see also Sect. 5.3)
18 http://www.astron.nl/general/apertif/apertif.
123
There has been a significant debate in the literature over whether the dust is smoothly
distributed in the torus (Pier and Krolik 1992; Dullemond and van Bemmel 2005;
Fritz et al. 2006), whether it is clumpy in the form of optically and geometrically thick
clouds (Krolik and Begelman 1988; Nenkova et al. 2002, 2008; Elitzur and Shlosman
2006; Tristram et al. 2007), or a combination of the two (Stalevski et al. 2012; Assef
et al. 2013). Observations of the strength of the silicate feature at 9.7 µm in AGN, for
example, seem to favour models where the dust is most prominently clumpy (Nenkova
et al. 2008; Nikutta et al. 2009; Hatziminaoglou et al. 2015), but Feltre et al. (2012)
has argued that observations are not yet able to discriminate between the different
models. Recent ground-based MIR observations, on the other hand, suggest that a
large proportion of the dust might, instead, reside in the walls of the ionization cone
(Asmus et al. 2016, and references therein), at least in a fraction of nearby AGN
(López-Gonzaga et al. 2017). For simplicity and compatibility with the rest of the
literature, we will refer to this structure as the “dusty torus” throughout this section,
despite evidence that this dust component may have a significantly more complex
distribution (e.g. Nenkova et al. 2008).
A number of authors have studied the fraction of lines-of-sight that are obscured by
the dusty torus, either by comparing the relative fraction of type 1 and 2 AGN at a given
redshift, or by modelling the SED of individual objects. Average obscured fractions
of 40–75% are reported in the literature (see,e.g. Treister et al. 2004; Hatziminaoglou
et al. 2009; Assef et al. 2013; Roseboom et al. 2013). However, a single number does
not englobe the diversity of AGN in Nature. Some authors have found that the fraction
of obscured lines-of-sight diminishes with increasing luminosity of the accretion disk
(Ueda et al. 2003; Hasinger 2004; Simpson 2005; Hatziminaoglou et al. 2008; Assef
et al. 2013; Mateos et al. 2016), an effect typically referred to as the receding torus
(Lawrence 1991), although others have found no evidence of such an effect (Wang and
Jiang 2006; Lawrence and Elvis 2010; Hönig et al. 2011; Lacy et al. 2013; Stalevski
et al. 2016, see also Sect. 5.3). Additionally, some authors have found a significant
variance in the amount of dust in AGN (e.g. Roseboom et al. 2013), with some showing
little to no dust (see Sect. 3.2.4) and some showing very large amounts (e.g. Mateos
et al. 2016). Furthermore some authors have found a larger fraction of obscured sources
at the highest luminosities (Banerji et al. 2012; Assef et al. 2015), suggesting a more
complex scenario, and possibly consistent with models where AGN dust obscuration
evolves through time (Sanders et al. 1988; Hopkins et al. 2008).
As already mentioned, the emission of the dusty torus is very prominent in the MIR
for both type 1 and type 2 AGN. Dust emission from SF can rival in luminosity the
AGN but with typically much cooler temperatures 40 K (e.g. Magnelli et al. 2012).
As significant SF activity is regularly ongoing in the host galaxies of many AGN, it
is more likely to dominate the FIR (e.g. Hatziminaoglou et al. 2010). At rest-frame
NIR wavelengths, where the AGN emission has a local minimum at the cross-over
between the dropping accretion disk emission and the rising dust emission, the stellar
1.6 µm peak can severely hamper AGN identification. As the stellar emission drops
steeply longward of the 1.6 µm peak, stellar contamination is less of an issue in
the MIR (although see Sect. 3.2.5). Hence, the MIR wavelengths are the optimal IR
wavelengths for AGN identification.
123
3.2 AGN in the MIR
3.2.1 Broad-band MIR AGN identification
The large sky background and water absorption by the Earth’s atmosphere make
ground-based MIR observations challenging. Spaceborne telescopes are, therefore,
better suited for the identification of large AGN samples. In what follows, we focus
solely on selection using space-based broad-band photometry, as they account for the
great majority of MIR identified AGN, although most implications and many of the
caveats also apply to ground-based and to spectroscopic observations. We explore
those further in Sect. 3.4.
A number of AGN MIR selection criteria have been proposed over the years. The
first ones were already developed for the Infrared Astronomical Satellite (IRAS) mis-
sion (de Grijp et al. 1985, 1987; Leech et al. 1989), and their number has grown
enormously since, with the advent, in the past decade, of the Spitzer Space Tele-
scope (Werner et al. 2004; see, e.g. Lacy et al. 2004, 2007, 2013; Stern et al. 2005;
Hatziminaoglou et al. 2005; Donley et al. 2012), AKARI (Murakami et al. 2007;
Oyabu et al. 2011), and the Wide-field Infrared Survey Explorer (WISE: Wright et al.
2010; see, e.g. Stern et al. 2012; Mateos et al. 2012; Wu et al. 2012b; Assef et al.
2013). These selection criteria have typically been calibrated against independent
AGN selection methods and rely primarily on colours to separate AGN from stars or
galaxies with inactive nuclei, as AGN are expected to be significantly redder in the
shorter wavelength MIR bands (e.g. Stern et al. 2005). We note that the latter is not
necessarily true for AGN at redshifts where the Hα emission line contaminates the
shortest wavelength channels, which can lead to significant biases against 4 z 5
AGN in some cases (Richards et al. 2009a; Assef et al. 2010), but is generally true
otherwise.
Assef et al. (2013) presented an interesting comparison between the WISE- and
Spitzer-based19 selection criteria. These are listed in Table 2 and shown in Fig. 5.20
Using a sample of large, multi-wavelength, spectroscopically identified AGN they
determined how reliable and complete each criterion is. The results are shown for
two W 2 limiting Vega magnitudes, namely W 2 < 15.05 on the left, representative
of the shallowest fields in the WISE mission, and W 2 < 17.1 on the right, that
probes down to a 3σ depth. Figure 5 shows that shallow and deep surveys need
to be analysed separately. For the shallow surveys, most selection criteria recover
samples with high reliability and completeness. For deeper surveys, however, the
situation is different. Most selection criteria line up diagonally in the diagram trading
completeness for reliability. One should also keep in mind that due to their lower
sensitivity, the completeness of selection criteria relying on the W 3 and W 4 bands is
19 For Spitzer we refer specifically to the four broad bands of the IRAC instrument (Fazio et al. 2004)
centred at 3.6, 4.5, 5.8 and 8 µm (and referred to as [3.6], [4.5], [5.8] and [8.0] respectively), and to the
24 µm band of the MIPS instrument (Rieke et al. 2004). For WISE we refer to all its four bands, centred
at 3.4, 4.6, 12 and 22 µm, usually referred to as W 1–W 4.
20 Adapted from Assef et al. (2013) to include the criteria of Donley et al. (2012) and Lacy et al. (2013).
Note that the WISE selection criteria still use the data from the All-Sky data release (Cutri et al. 2012).
123
Table 2 AGN MIR selection criteria shown in Fig. 5
ID References Criteria
WISE only criteria

(1) Assef et al. (2013)—R90a W 1 − W 2 > 0.662 exp{0.232 (W 2 − 13.97)2 }
(2) Assef et al. (2013)—R75a W 1 − W 2 > 0.530 exp{0.183 (W 2 − 13.76)2 }
(3) Assef et al. (2013)—C90a W 1 − W 2 > 0.50
(4) Assef et al. (2013)—C75a W 1 − W 2 > 0.77
(5) Stern et al. (2012)b W 1 − W 2 > 0.80
(6) Jarrett et al. (2011)c W 2 − W 3 > 2.2 ∧ W 2 − W 3 < 4.2 ∧
W 1 − W 2 > (0.1 (W 2 − W 3) + 0.38 ∧
W 1 − W 2 < 1.7, except
(i) W 1 < 10.5 ∧ W 2 − W 3 < 1.5 ∧
W 1 − W 2 < 0.4 or
(ii) W 3 − W 4 < 1.2
(7) Mateos et al. (2012)d —3-band W 1 − W 2 > −3.172 (W 2 − W 3) + 7.624 ∧
W 1 − W 2 < 0.315 (W 2 − W 3) + 0.796 ∧
W 1 − W 2 > 0.315 (W 2 − W 3) − 0.222
(8) Mateos et al. (2012)d —4-band W 1 − W 2 > −2.00 (W 3 − W 4) + 4.33 ∧
W 1 − W 2 < 0.50 (W 3 − W 4) + 0.979 ∧
W 1 − W 2 > 0.50 (W 3 − W 4) − 0.405
(9) Assef et al. (2010)—2-band W 1 − W 2 > 0.85
(10) Assef et al. (2010)—4-band W 3 − W 4 > 2.1 ∧ W 1 − W 2 > 0.85 ∧
W 1 − W 2 > 1.67 (W 3 − W 4) − 3.41
(11) Wu et al. (2012b) W 1 − W 2 > 0.57
Spitzer only criteria
(12) Stern et al. (2005) [5.8] − [8.0] > 0.6 ∧
[3.6] − [4.5] > 0.2 ([5.8] − [8.0]) + 0.18 ∧
[3.6] − [4.5] > 2.5 ([5.8] − [8.0]) − 3.5
(13) Lacy et al. (2004) log f 8.0 / f 4.5 > −0.2 ∧ log f 5.8 / f 3.6 > −0.2 ∧
log f 8.0 / f 4.5 < 0.8 log f 5.8 / f 3.6 + 0.5
log f 8.0 / f 4.5 < 0.8 log f 5.8 / f 3.6 + 0.5
log f 8.0 / f 4.5 < 0.8 log f 5.8 / f 3.6 + 0.5
(16) Donley et al. (2012) log f 5.8 / f 3.6 ≥ 0.08 ∧ log f 8.0 / f 4.5 ≥ 0.15 ∧
log f 8.0 / f 4.5 ≥ 1.21 log f 5.8 / f 3.6 − 0.27 ∧
log f 8.0 / f 4.5 ≤ 1.21 log f 5.8 / f 3.6 + 0.27 ∧
f 4.5 > f 3.6 ∧ f 5.8 > f 4.5 ∧ f 8.0 > f 5.8
123
Table 2 continued
ID References Criteria
NIR + Spitzer criteria

(17) Messias et al. (2012)—KI K s − [4.5] > 1.42 ∧ [4.5] − [8.0] > 1.14
(18) Messias et al. (2012)—KIM K s − [4.5] > 1.42 ∧ [8.0] − [24] > 2.87 ∧
[8.0] − [24] > −2.9 ([4.5] − [8.0]) + 8.47
In all the criteria above, the name of a band represents its magnitude in the Vega system, while f X represents
the flux density of band X . Note that [24] refers to the MIPS 24 µm band Vega magnitude, for which we
have assumed a flux density zero point of 7.14 Jy. For empirically calibrated WISE selection criteria we
indicate the magnitude range of the calibration sample if one was applied, as the comparison in Figure 5
may be extrapolated to fainter magnitudes in some cases
a Calibrated for galaxies with W 1 < 18.50 and W 2 < 17.11
b Calibrated for galaxies with W 2 < 15.05
c Calibrated for galaxies with W 1 < 18.1, W 2 < 17.2, W 3 < 13.4
d Calibrated for galaxies detected with S/N > 5 in all WISE bands used
Fig. 5 Comparison of completeness and reliability of popular photometric MIR AGN selection criteria,
assessed by the test developed by Assef et al. (2013) and discussed in Sect. 3.2.1. The left and right panels
show the case of a shallow and deep survey, respectively. The criteria shown correspond to the references
listed in Table 2. The red points show the WISE only selection criteria, the blue points show the Spitzer
only selection criteria, and the black points show the NIR+Spitzer criteria
lower than the completeness obtained using criteria derived based on a W 2-limited
sample. A selection relying on these longer wavelengths can be of particular use in
the WISE fields closer to the ecliptic poles, where the survey scan pattern is denser
and the shorter wavelength bands reach the confusion limit (e.g. Jarrett et al. 2011).
3.2.2 Comparison with general AGN identification at other wavelengths
MIR AGN identification is considerably less sensitive to obscuration of the central

engine by dust compared to optical identification, as dust opacity is lower at longer
wavelengths and is, therefore, better for the selection of obscured AGN than optical
identification, although its sensitivity to obscured AGN decreases with increasing
redshift due to the K-correction (e.g. Assef et al. 2011).
123
As discussed in Sect. 5.2, X-rays are also sensitive to obscured sources, especially in
the harder X-ray energies, which are less affected by neutral hydrogen absorption. The
main advantage of the MIR over the X-rays is that the integration times needed for AGN
identification are much shorter, and hence allow for faster survey speeds. For example,
Gorjian et al. (2008) finds that 97.5% of all sources in the 5 ks integrations of Chandra
X-ray Observatory of the XBoötes survey (0.5–7.0 keV flux > 8×10−15 erg cm−2 s−1 ;
Murray et al. 2005) have counterparts detected in the 90 s integrations of the IRAC
Shallow Survey ( f 3.6 µm > 12.3 µJy; Eisenhardt et al. 2004). Of the X-ray sources
without an IR match, 40% might be spurious. Furthermore, MIR identification can
find AGN that are hard to detect in the X-rays, either due to obscurations or to intrinsic
X-ray faintness (e.g. Stern et al. 2012, see also Sect. 5.2).
However, MIR identification is affected by contaminants and biases that are only
marginally relevant to X-ray or optical selections. In the next sections we discuss these
issues, which need to be taken into account when drawing statistical conclusions about
the AGN population from MIR selected samples.
3.2.3 Contaminants
As mentioned earlier, AGN selection using MIR broad-bands mostly relies on the
typically redder colours of AGN, particularly in the observed 3–5 µm wavelength
range. However, there are a number of different populations that can mimic the colours
of AGN in these bands and will affect most selection criteria, although the extent will
depend on each specific selection. At redshift ∼0.2, strongly SFGs with powerful
polycyclic aromatic hydrocarbon (PAH) emission can have red enough colours to
be confused with AGN in some identification schemes (see e.g. Stern et al. 2005;
Assef et al. 2010; Hainline et al. 2016). Such galaxies can appear as contaminants in
shallow and deep observations (see previous section). As they are uncommon and the
co-moving volume is low enough at the respective redshift range, they are typically
only a minor contaminant. However, Hainline et al. (2016) has recently pointed out
that such contaminants can be confused with AGN hosted in dwarf galaxies and hence
represent a major problem for their identification (e.g. Satyapal et al. 2014, 2016).
For deeper surveys, the most serious contaminants are high redshift (z 1) massive
galaxies. At those redshifts the 1.6 µm stellar bump is shifted into the MIR, and their
colours can become red enough to mimic those of AGN in many selection schemes (see
Donley et al. 2007, 2012; Yun et al. 2008; Assef et al. 2010, 2013; Mendez et al. 2013).
The selection technique developed by Donley et al. (2012) for Spitzer observations is
specifically aimed at avoiding these galaxies, resulting in high reliability at faint fluxes
although at the cost of completeness. Using W 2 magnitude dependent prescriptions,
as described in Assef et al. (2013), has the same effect resulting also in high reliability
at faint fluxes with a very low completeness level (see Fig. 5).
In addition to extragalactic contaminants, there are a number of Galactic sources
that can mimic the colours of AGN in the MIR, such as brown dwarfs or young
stellar objects. Brown dwarfs are rare and typically only account for a small fraction
of the contaminants but might still affect the identification of high-redshift (z > 5)
AGN (Stern et al. 2007). Young stellar objects also have MIR colours that can be
123
confused with those of AGN (see, e.g. Koenig et al. 2012), and they can be numerous
contaminants close to the Galactic plane.
3.2.4 Dust-free AGN and hot dust poor quasars
Broad-band MIR AGN selection primarily relies on the detection of the hot dust
emission at low and intermediate redshifts. While hot dust emission is rather prominent
in most AGN, its ratio to the accretion disk emission (usually described as the torus’
apparent covering fraction) shows a wide distribution (see e.g. Roseboom et al. 2013;
Mateos et al. 2016; Hernán-Caballero et al. 2016). Therefore, AGN with low hot dust
emission relative to that of their host could escape identification, especially if they
reside within luminous hosts. Such objects account for ∼10% of the quasar population
selected in the X-rays, optical or MIR (Hao et al. 2010, 2011, but see also Lyu et al.
2017) and their fraction may be increasing with redshift (Hao et al. 2010; Mor and
Trakhtenbrot 2011; Jun and Im 2013). According to Hao et al. (2011), their small
amount of dust seems to be sufficient to place them to the borders of the Lacy et al.
(2004) criteria, but they would be systematically missed by a more stringent selection.
MIR selections would of course completely miss dust-free AGN, such as the local
analogues of the two sources identified by Jiang et al. (2010) among a sample of 21
z ∼ 6 quasars observed with Spitzer. Although such analogues may not necessarily
exist, they might occur more often at lower luminosities. In fact, Elitzur and Shlosman
(2006) showed that if the torus is populated by outflowing clouds of material from
the accretion disk, at L 1042 erg s−1 the accretion would be too low to sustain the
required outflow rate resulting in the disappearance of the torus. This may apply to
some LEG RGs (see also Sect. 2.1.5). A clear case among the observational evidence
(e.g. Chiaberge et al. 1999; Maoz et al. 2005; Müller-Sánchez et al. 2013) is the nearby
LEG M87, that has been shown to lack the thermal radiation of the torus (Whysong and
Antonucci 2004; Perlman et al. 2007). Such objects would be missed systematically
by all MIR selection criteria at low redshifts.
3.2.5 Eddington ratios and BH Masses
The luminosity of the spheroidal component of a galaxy correlates with the mass of
its SMBH (at least at relatively low redshifts: e.g. Marconi and Hunt 2003; Gültekin
et al. 2009) and hence with its Eddington limit. The ratio of the specific luminos-
ity of the AGN to that of the host, L ν,AGN /L ν,Host , therefore directly depends on
the AGN L/L Edd . In other words, the incompleteness due to host dilution directly
translates into a bias of MIR AGN selection against low L/L Edd . Such a bias
has been discussed by, e.g., Hickox et al. (2009) (see also Fig. 25) and Mendez
et al. (2013), who showed that MIR AGN selection only probes the upper end
(1%) of the L/L Edd distribution compared to X-ray selection (this applies also
to optically-selected AGN: see Sect. 4.5). Quantifying this bias is difficult, as it
strongly depends on the selection function being used, but it needs to be taken into
account and the complete selection function needs to be modelled in order to be
able to give a physical and statistical interpretation of results based on MIR-selected
AGN.
123
Although in an indirect manner, this selection effect can also bias the distribution of
BH masses in MIR identified AGN. As BH masses correlate only with the luminosity
of the host galaxy’s spheroidal component, the selection effect discussed above will be
stronger in galaxies with important disk or irregular components, as the starlight from
them will increase L ν,Host for a fixed BH mass. In other words, AGN in galaxies with
disk components need to accrete at higher L/L Edd to be identified by MIR selection
techniques. As such non-bulge components are typically more prominent in lower
mass galaxies, and these host lower mass BHs (Magorrian et al. 1998), the latter will
be systematically underrepresented in MIR AGN samples.
3.3 Red, reddened and high-redshift AGN
NIR photometry has also been used to select AGN (e.g. Warren et al. 2000; Francis
et al. 2004; Kouzuma and Yamaoka 2010). Such methods, however, offer little gain
compared to MIR and optical selection. Nevertheless, NIR has been useful in the
identification of red AGN samples, e.g. Glikman et al. (2007) using 2MASS, FIRST
and R-band photometry, as well as the heavily reddened quasars of Banerji et al. (2012,
2015) found in the UKIDSS survey.
Populations of heavily reddened AGN have also been found by means of MIR
photometry, often in combination with optical observations. For example, using the
colour between the optical R-band and the MIPS 24 µm band, Dey et al. (2008) found
a new class of objects, dubbed Dust-Obscured Galaxies or DOGs, many of which host
heavily reddened AGN (see e.g. Melbourne et al. 2012). Recently Ross et al. (2015)
presented a selection of Extremely Red Quasars relying on SDSS and WISE data.
Similar is the case of the Hot Dust Obscured galaxies or Hot DOGs (Eisenhardt et al.
2012; Wu et al. 2012a), selected based purely on their WISE colours, for which a
number of studies suggest they probe an important stage of galaxy evolution (see e.g.
Jones et al. 2014; Assef et al. 2015; Tsai et al. 2015; Díaz-Santos et al. 2016).
Finally, NIR and MIR wavelengths are very important for the identification of
the earliest quasars in the Universe, which are not observable in the optical due to
Lyα forest absorption and the Lyman break. For example, the highest redshift quasar
currently known at z = 7.1 was found by Mortlock et al. (2011) relying on the IR
coverage of the UKIDSS survey, and Bañados et al. (2016) has recently used NIR
and MIR observations from 2MASS, VHS and WISE to help in the identification of
z > 5.6 quasars candidates selected from the optical PanSTARRS21 survey. Upcoming
NIR surveys such as those that will be provided by Euclid and the Wide-Field Infrared
Survey Telescope (WFIRST) will allow to better probe the quasar population at the
earliest cosmic times (see Sect. 3.5).
3.4 MIR spectroscopy
MIR spectroscopy, particularly with the InfraRed Spectrograph (IRS; Houck et al.
2004) on board the Spitzer Space Telescope, provided new insights into the physics
21 panstarrs.stsci.edu/.
123
and classification of AGN. The unambiguous observations of the silicate feature at

9.7 µm in emission in many known AGN (Hao et al. 2005; Siebenmorgen et al. 2005;
Sturm et al. 2005; Buchanan et al. 2006; Shi et al. 2006) came as the long sought
confirmation of the unified scheme. At the same time, however, IRS observations
indicated that in some cases the source of obscuration resides in the host rather than
the torus (e.g. Goulding et al. 2012; Hatziminaoglou et al. 2015).
Identification through MIR spectroscopy is very powerful, allowing to detect
obscured AGN components even when the MIR is dominated by the host galaxy.
Several classification diagrams have been developed to determine the AGN contribu-
tion to an observed spectrum based on certain spectral features, such as high ionisation
emission lines like [Ne v], [Ne ii] and [O iv], the EW of PAH features and the strength
of the silicate feature at 9.7 µm (see, e.g. Spoon et al. 2007; Armus et al. 2007;
Veilleux et al. 2009; Hernán-Caballero and Hatziminaoglou 2011). A number of tech-
niques have also been developed to model the observed MIR spectra and constrain the
AGN and starburst contributions (see e.g. Schweitzer et al. 2008; Nardini et al. 2008;
Deo et al. 2009; Feltre et al. 2013).
Although MIR spectroscopy has had a great impact on our understanding of AGN,
the number of objects studied through these techniques is limited when compared to
photometric studies, as spectroscopic observations require significantly longer inte-
gration times. Ground-based observations are generally limited to the brightest targets
due to the effects of the Earth’s atmosphere (e.g. Alonso-Herrero et al. 2016), while
deeper observations were possible with the IRS during its cryogen-cooled phase. For
the most part, such observations were limited to z 1 luminous IR galaxies (LIRGs),
ultraluminous IR galaxies (ULIRGs), and quasars (Hernán-Caballero and Hatzimi-
naoglou 2011, and references therein) although a number of higher redshift ULIRGs
were also studied by IRS (see e.g. Kirkpatrick et al. 2012). The impact of these tech-
niques will be greatly expanded by the upcoming JWST (Gardner et al. 2006) and
Space Infrared Telescope for Cosmology and Astrophysics (SPICA; Nakagawa et al.
2015), that will probe significantly fainter targets and will allow us to select new,
currently inaccessible, sets of objects, as discussed next.
3.5 The future of AGN studies in the IR band
The upcoming generation of ground-based giant telescopes will significantly expand

upon the current NIR and MIR capabilities, as most of them will have significant
focus on these wavelengths. The Giant Magellan Telescope (GMT22 ) is planning on
first generation instruments with imaging capabilities from 0.9 to 2.5 µm and spec-
troscopic capabilities in the 1–5 µm range, and its first light is currently expected for
2021. The Thirty Meter Telescope (TMT23 ) is planning on first-light photometric and
spectroscopic instruments in the 0.8–2.5 µm range, with the goal of extending this
range to 28 µm within its first decade of operations. The first generation instrumen-
tation of the largest of the upcoming giant telescopes, the Extremely Large Telescope
22 http://www.gmto.org.
23 http://www.tmt.org.
123
(ELT24 ), will allow for imaging and spectroscopy all the way to 19 µm. Finally, the
University of Tokyo Atacama Observatory (TAO25 ) 6.5m telescope will focus on the
IR, with planned first light instrumentation probing wavelengths as long as 38 µm.
The only space-based observatories with MIR imaging capabilities currently in
operation are Spitzer, whose operations have recently been extended until 2019 by
the 2016 NASA Senior Review, and WISE, whose NEOWISE-R phase is planned to
extend until the end of 2017. In the near future, NASA’s JWST,26 expected to launch
in 2018, will offer unprecedented imaging and spectroscopic capabilities in the 0.6–
28.3 µm wavelength range thanks to its 6.5 m diameter primary mirror. Observations
with the JWST will probe with high angular resolution a number of targets that are
not accessible from the ground, likely having a major impact in our understanding of
AGN. The proposed joint JAXA/ESA mission SPICA,27 that will be launched in 2028
if approved by the two agencies, will have a 2.5 m aperture and will provide low-to-
high resolution spectroscopy in the wavelength range between 35 and 240 µm, and
imaging capabilities. Finally, the upcoming Euclid and WFIRST missions will have a
significant impact in NIR AGN identification by mapping large areas of the sky to very
faint depths. Euclid,28 expected to launch in 2020, will have a 1.2 m primary mirror
and will survey 15 000 deg2 of extragalactic sky down to limiting AB magnitudes
of 24 in Y , J and H , as well as 24.5 in a very wide optical broad-band. It will also
observe a smaller region of 40 deg2 to limiting AB magnitudes of 26 in the NIR bands
and 26.5 in the optical band. The WFIRST mission,29 expected to launch by 2024,
will, on the other hand, have a 2.4m primary mirror and a similar set of bands, and is
planned to image 2227 deg2 down to a limiting AB magnitude of 26.7 in J (Spergel
et al. 2015). Both telescopes will also obtain NIR slitless grism spectroscopy in their
survey areas. Through their unique combination of area and depth, both surveys will
probe AGN activity during the formation of the first galaxies in the Universe.
4 Optically-selected AGN
This section discusses the selection and properties of optically-selected AGN as con-
trasted with investigations at other wavelengths. The focus here is on the more luminous
subsets that would typically be classified as quasars or Seyfert 1 galaxies. We will not
cover objects like the host-galaxy dominated AGN (Kauffmann et al. 2003a), the
LINERs (Heckman 1980), or XBONGs (Hornschemeier et al. 2005). While optical
surveys are able to identify quantitatively more AGN than other wavelengths (through
a combination of area and depth), this size comes with a bias towards brighter, unob-
scured sources with L/L Edd > 0.01 (see Sect. 4.5). Even though optical surveys
are not ideal for probing obscured AGN, we discuss how they can guide our search
24 http://www.eso.org/sci/facilities/eelt.
25 http://www.ioa.s.u-tokyo.ac.jp/TAO/en.
26 http://www.stsci.edu/jwst.
27 http://www.ir.isas.jaxa.jp/SPICA/SPICA_HP/index-en.html.
28 http://www.euclid-ec.org.
29 wfirst.gsfc.nasa.gov.
123
for them. The bias towards unobscured sources in the optical is partially mitigated,
however, by an increase in information content for the sources that are identified—in
the form of physics probed by the combination of optical continuum, absorption, and
emission. An example is the ability to estimate the mass of the SMBHs based on
the optical/UV emission lines. We discuss the physical mechanisms behind optical
emission in Sect. 4.1 and AGN identification, selection effects, information content,
and the range of masses (and accretion rates) probed by the optical band in Sect. 4.2–
4.5. While the redshift/luminosity evolution of optically-selected luminous quasars
would seem to be well constrained from today to redshift ∼6, we review evidence
suggesting that the Hopkins et al. (2007) bolometric LF needs to be updated. How
next-generation surveys such as LSST can bridge the evolution of luminous quasars to
lower-luminosity AGN (that are typically better probed at other wavelengths) will be
addressed in Sect. 4.6. In this section we cover the ∼1000−8000 Å range (rest-frame).
4.1 Physical mechanisms behind optical emission
AGN are believed to be powered by accretion onto a SMBH, that gives rise to high
X-ray-to-optical luminosities, a characteristic rest-frame UV/optical power-law con-
tinuum (very distinct from the continuum of non-active galaxies) as well as the so
called “big blue bump”, and a break of this continuum blueward of ≈1000 Å (Fig. 1).
Many models, usually assuming a geometrically thin, optically thick accretion disk,
have been developed in order to explain this emission (e.g. Sun and Malkan 1989; Laor
and Netzer 1989; Hubeny et al. 2001, and references therein). AGN that have a line of
sight to the central engine that is not obscured show broad emission lines with typical
gas velocities of a few 1000 km s−1 covering a large range in strength and profiles,
whose properties correlate with the luminosity of the AGN (e.g. Baldwin 1997). The
source of the broad emission lines is the BLR, believed to be located between the
SMBH and the inner wall of the dusty torus (see Sect. 3), with photoionized gas that
has been heated by the radiation coming from the accretion disk around the SMBH.
Finally, AGN display narrow emission lines, with gas velocities between 300 and
1000 km s−1 , originating in the Narrow Line Region (NLR) with sizes ≈ hundreds of
parsec, located above (and below) the plane of the dust.
4.2 Photometric and spectroscopic identification
The features described in the previous section are the basis of optical AGN identifi-
cation, be it photometric or spectroscopic. Broad band photometry is sensitive to the
presence of broad emission lines in the various filters as a function of redshift, as they
alter the otherwise very typical colours of the AGN that separate them from the stellar
locus (see, e.g. Richards et al. 2001, Fig. 4). Narrow band surveys such as COMBO17
(Wolf et al. 2003), ALHAMBRA (Moles 2008), and now J-PAS (Benitez et al. 2014)
make use of the spectral features to not only identify AGN but to also estimate their
(photometric) redshifts with a precision that reaches below 1% (e.g. Salvato et al. 2009;
Hsu et al. 2014), a great improvement with respect to early attempts (Hatziminaoglou
et al. 2000). Finally, we are generally reliant on optical spectroscopy to provide con-
123
Fig. 6 Number of quasars as a

function of time. The dashed
line and triangles give the
number of quasars in the largest
heterogeneous samples to date.
The solid line and squares give
the largest homogeneous quasar
samples (including photometric
quasar candidates) to
date—extrapolating to
10 000 000 expected AGN for
LSST (LSST Science
Collaboration et al. 2009).
Adapted from Richards et al.
(2009b); the 2015 photometric
point is from DiPompeo et al.
(2015)
firmation of a source as an AGN and to determine its redshift, while the presence of
narrow emission lines in the spectra of galaxies and their ratios are indicative of the
presence of an AGN (e.g. Feltre et al. 2016).
The problem with the optical band, as compared to, say, the hard X-rays, is
that bright optical sources are not necessarily AGN. The same is true for the radio
(Sect. 2.1.2) and IR to some extent (Sect. 3.2)—the brightest sources on the sky have
a high probability of being an AGN. This point is illustrated in the comparison of
number counts between the X-ray (Lehmer et al. 2012, Fig. 5) and optical (Shanks
et al. 2015, Fig. 24): in the optical, stars far outnumber AGN at typical survey depths.
Optical selection can make up for this (photometric) uncertainty with sheer num-
bers. Deep, high-resolution X-ray, IR and radio fields provide a much higher AGN
density—up to ∼24 000 deg−2 (Sect. 5.1; e.g. Luo et al. 2017, see also Sect. 8), but only
over minuscule areas. The density of the largest-area photometric and spectroscopic
quasar samples is only ∼150 deg−2 , but over a large fraction of the sky, which results
in larger samples of AGN overall (e.g. Richards et al. 2009b; D’Abrusco et al. 2009;
Bovy et al. 2011; Flesch 2015; Brescia et al. 2015; Pâris et al. 2017). Figure 6 shows
the growth in quasar numbers with time for both heterogeneous and homogeneous
quasar samples. We discuss the impact of future facilities in Sect. 4.6.
4.3 Selection effects
Paradoxically, many (optically) unobscured AGN are missed by optical surveys. These
are objects whose colours put them in (or close to) the stellar locus. Since luminous
quasars are point sources, but are outnumbered by stars in our galaxy by ∼100:1 at
the SDSS depth, it is very difficult to create a complete sample of quasars at certain
123
redshifts—without considerable stellar contamination. The redshifts affected span a

large range around z ∼ 2.6 with very low completeness, and a smaller range around
z ∼ 3.5 (Richards et al. 2002, 2006b; Worseck and Prochaska 2011). Even the SDSS-
III/BOSS quasar sample is only 60% complete at these redshifts despite being designed
to identify 2.2 < z < 3.5 quasars (Ross et al. 2013). It is worth noting that most MIR
selections have a similar redshift “hole” over 4 z 5 (Sect. 3.2.1) but this can be
filled in by using optical and MIR data simultaneously (Richards et al. 2015).
Lower-luminosity AGN are also a challenge for imaging-only optical surveys (like
the Dark Energy Survey [DES30 ] and LSST31 ) for the reasons noted above: without
spectroscopy, it is difficult to distinguish a normal galaxy from an active galaxy.
Variability selection may not help all that much for such sources considering that, while
the amplitude of variability increases at lower luminosity (Vanden Berk et al. 2004,
see also Sect. 7.2.3), the fraction of optical emission from the central engine decreases.
Yet this is the population that we most need to probe, especially for comparison to
X-ray and MIR samples.
The biggest hole in the selection of AGN via optical photometry is certainly
obscured (or type 2) AGN. It does not matter whether the optical obscuration is by
a smooth torus or a clumpy one (see Sect. 3.1 for a full list of references on smooth
and clumpy distributions): the traditional AGN model has a region where most of the
optical emission is obscured. The irony is that a large fraction of known type 2 AGN
are still selected in the optical (Zakamska et al. 2003; Reyes et al. 2008; Alexandroff
et al. 2013; Yuan et al. 2016). This result is due to a combination of effects: the host
galaxy is not (always) obscured and both strong emission lines and scattering can
result in non-negligible optical flux producing unusual (or even AGN-like) colours,
which can cause them to be identified as potential type 1 sources despite them being
type 2.
Thus the question of the relative fraction of obscured and unobscured AGN is still
much debated/investigated (see also Sect. 3.1), particularly as a function of luminosity
(e.g. Gilli et al. 2007; Ueda et al. 2014). Crucially, differences in the definition of
“obscured” between the optical and X-ray (e.g. Hickox et al. 2007) make it more
difficult than one might imagine to paint a full picture (see also Sect. 5.3).
4.4 Information content
In Sect. 3 we saw that IR-detected quasars are relatively unbiased against type 2
AGN, whereas most of the optical light comes from the central accretion disk, which
is blocked in these sources. However, while the optical may be missing a crucial
component of the AGN zoo in terms of obscured AGN, it more than makes up for
that loss in terms of information content of those AGN that are detected. Moreover,
the information content in the continuum, emission lines, and absorption lines from
optical spectroscopy is particularly rich. For example, BAL quasars (Weymann et al.
30 http://www.darkenergysurvey.org.
31 http://www.lsst.org.
123
Fig. 7 Diversity of UV emission line properties in SDSS quasars. Quadrants refer to the C iv EW-blueshift
plane with red objects having low C iv blueshifts and large EW and blue objects having large C iv blueshift
and small EW. These correlated features may be indicative of differences in the hardness of the spectral
energy distributions (e.g. Leighly 2004)
1991) have provided us with direct evidence of winds in AGN systems with outflow
velocities extending to tens of thousands km s−1 (e.g. Hamann et al. 2011).
While the BAL sub-class represents only ∼20% of the population of luminous
quasars (Hewett and Foltz 2003), the advent of principal component analysis, both
using photometry (Boroson and Green 1992) and spectroscopy (Francis et al. 1992;
Yip et al. 2004) and large data sets from SDSS (Schneider et al. 2010; Pâris et al.
2017), has enabled the community to treat quasars as diverse systems. For example,
Richards et al. (2011) argue that we can learn about winds using emission lines in
addition to absorption lines, which potentially turns every quasar into powerful probe
of AGN outflows.
Indeed, one way to illustrate the diversity of quasars is presented in Fig. 7, which
shows how physical trends are manifested in the different emission lines: quasars with
harder spectra have stronger emission lines and less “blueshift” of the C iv emission
line. The probability of a quasar being radio detected or having strong BAL troughs is
a strong function of the appearance of these emission lines (Richards et al. 2011). The
trends shown in Fig. 7 appear to be related to those that define the “Eigenvector 1”32
parameter space (Boroson and Green 1992; Brotherton and Francis 1999; Sulentic et al.
2000, 2007), which together highlight the great diversity of AGN even when consid-
ering only those that are both optically-selected and very luminous. Often overlooked
is the fact that this diversity has important implications for accurate determination of
quasar redshifts (Hewett and Wild 2010).
4.5 L/L Edd and MBH
Arguably the best example of where optical provides additional information content
and makes up for selection effects is in our ability to utilise BH mass scaling relations
to estimate the masses of the BHs powering quasars (e.g. Vestergaard and Peterson
2006). Again, this process requires optical spectroscopy.33 By: (1) knowing the width
of the emission lines; (2) assuming that width is dominated by gravitational effects; (3)
32 This is a set of correlations between properties observed in quasar spectra, which comes out of principal
component analysis.
33 NIR spectroscopy can be used as well, but the sample size of objects with appropriate data is relatively
small by comparison (but see, Ricci et al. 2017).
123
having an estimate of the characteristic radius of the emitting gas assuming R ∝ L ∼0.5
(Bentz et al. 2009); and (4) calibrating this information against the few dozen objects
for which there exists “reverberation mapping” (Peterson 1993) data, we can then
estimate the masses of the BH in every type 1 quasar.
Generally speaking the types of AGN discussed herein have masses of 106 –1010 M
and accretion rates (in terms of L/L Edd ) of 0.01–1 (Greene and Ho 2007; Vestergaard
et al. 2008; Shen and Kelly 2012; Trakhtenbrot and Netzer 2012). See Schulze et al.
(2015) for a recent, comprehensive analysis that combines data from the VIMOS-VLT
Deep Survey (VVDS), zCOSMOS, and SDSS.
Scaling relations for estimating the masses of BHs in relatively local AGN (up to
z ∼ 0.7) make use of the Hβ emission line and are thought to be relatively robust (at
the level of ∼0.3 dex). Using the Mg ii emission line these scaling relations have been
extended to z ∼ 1.9. However, beyond that redshift other broad emission lines need
to be used. As such, attempts have been made to calibrate C iv to produce BH mass
estimates, though it is becoming clear that the winds discussed above can significantly
bias the estimates for high-redshift quasars which rely on the C iv emission line
(Baskin and Laor 2005; Richards et al. 2011; Denney 2012; Shen 2013). As the
sample size of high-redshift quasars with both optical and IR spectroscopic coverage
grows, corrections to this scheme might help bringing the BH masses into alignment
(e.g. Runnoe et al. 2013; Coatman et al. 2016). However, the uncertainties associated
with the determination of SMBH based on C iv are not solely due to low spectral
resolution and/or signal-to-noise ratio (S/N) but rather point towards differences in
the physics of the BLR. In other words, either the C iv-emitting gas is non-virialised
or objects with low and high FWHM(Hβ) have different ionisation structure, since
the FWHM(C iv) has only a loose correlation, if any, with FWHM(Hβ) (Denney et al.
2013; Mejia-Restrepo et al. 2016; Coatman et al. 2017).
4.6 The evolution with redshift and the impact of future facilities
The SDSS quasar LF (Richards et al. 2006b) was unique not in the redshift range or
luminosity that it probed, but rather because it probed such a large range with just
one uniform data set with a large number of quasars. One of the shortcomings of the
wide, but shallow SDSS work was that it generally only probed the bright end of the
LF, whereas narrow, but deep X-ray surveys were better able to probe the faint end.
Hopkins et al. (2007) combined the best of both worlds from various multi-wavelength
surveys to create a bolometric LF.
Recent works, however (e.g. Assef et al. 2011; Ross et al. 2013; McGreer et al.
2013), have shown that the evolution of the break luminosity (where the slope of the
LF changes rapidly) may be very different from that predicted by Hopkins et al. (2007)
and that the faintest SDSS quasars might start probing the faint end (rather than the
bright end, as was assumed) of the LF at high z. What is needed here are both deeper
surveys in the optical and wider surveys in the X-ray (Sect. 5.4), or similarly deep/wide
NIR surveys (Sect. 3.5). But even without new data there is a decade of observations
that could be incorporated into an updated bolometric LF.
Improving our knowledge of the quasar LF has consequences beyond the study
of quasars: it also has important consequences for reionization in the early Universe.
123
For example, the results of McGreer et al. (2013) and Glikman et al. (2011) suggest
photoionization rates at z > 4 that do not fully agree.
It is important to understand that the bias in the optical band is towards the high
luminosity end of the AGN distribution, i.e. bona fide quasars. In part that is because
that is the population that optical surveys themselves are biased to. However, current
and upcoming experiments like Pan-STARRS,34 DES, and LSST (will) probe to fainter
limits that are more compatible with the lower-luminosity AGN (where the host emis-
sion is more comparable to the central engine) that surveys at other wavelengths most
commonly identify. These new optical surveys will help complete a multi-wavelength
bridge that will allow AGN astronomers to more fully sample luminosity-redshift
space across the full electromagnetic spectrum, as illustrated by LaMassa et al. (2016,
Fig. 14).
5 X-ray-selected AGN
We discuss in this section X-ray-selected AGN. The physical mechanism behind X-

ray emission is examined in Sect. 5.1, while Sect. 5.2 deals with AGN selection and
its challenges. Sect. 5.3 reviews the types of AGN selected in the X-ray band, the
range of MBH and L/L Edd they sample, and their evolution, while Sect. 5.4 gives an
overview of the future of this field. In this review we define the X-ray band as covering
the energy range of 0.2–200 keV.
5.1 Physical mechanism behind X-ray emission
X-ray observations provide a near complete selection of AGN with low contamination
from non-AGN systems. The primary reasons for this are the following: (1) X-ray
emission from AGN appears to be (near) universal; (2) X-rays are able to penetrate
through large column densities of gas and dust (particularly at high X-ray energies);
(3) X-ray emission from host-galaxy processes are typically weak when compared to
the AGN (see also Sect. 1.1 of Brandt and Alexander 2015). On the basis of these
advantages, the deepest blank-field cosmic X-ray surveys have identified the largest
reliable AGN source density to date (≈24 000 deg−2 ; Lehmer et al. 2012; Luo et al.
2017 and Sect. 8).
The intrinsic X-ray emission from AGN is due to processes related to the accretion
disk (see Mushotzky et al. 1993; Done 2010; Gilfanov and Merloni 2014 for reviews;
note that in jetted AGN the jet can make a major contribution in the X-ray band
as well).35 The primary process is thought to be inverse Compton scattering of the
accretion-disk photons to X-ray energies via the accretion-disk “corona” (see Fig. 1;
34 pswww.ifa.hawaii.edu/pswww/.
35 In this section we focus on high accretion-rate AGN (with an optically thick and geometrically thin
accretion disk; i.e. L/L Edd > 0.01), which account for the majority of the BH growth in the Universe (e.g.
Ueda et al. 2014; Aird et al. 2015). Low accretion-rate AGN (with an optically thin, geometrically thick, hot
accretion flow; i.e. L/L Edd < 0.01) can also be selected at X-ray energies, although the accretion process
is driven by advection of a hot plasma; see Sect. 2.1.4, Done et al. (2007), and Yuan and Narayan (2014)
for more details.
123
this is generally depicted as an atmosphere above the inner accretion disk, though its
exact geometry is unknown). However, thermal X-ray emission due to the inner regions
of the accretion disk can also be produced at the lowest X-ray energies (e.g. Sobolewska
et al. 2004). The X-ray emission is then modified due to the interaction with matter in
the nuclear region (e.g. reflection, scattering, and photo-electric absorption of photons
from the accretion disk and/or the obscuring AGN torus: see Sect. 3.1). The relative
strength of these components can vary quite significantly from source to source, mostly
due to differences in the geometry and inclination angle of the torus to the line of sight,
leading to a broad range of X-ray spectral shapes. The intrinsic X-ray emission from the
“corona” is tightly connected to the accretion-disk emission (as parameterised by αOX ;
e.g. Steffen et al. 2006; Lusso and Risaliti 2016) for almost all systems (for exceptions
see, e.g. Wu et al. 2011; Luo et al. 2014), demonstrating a causal relationship and
showing that X-ray emission from AGN is (near) universal.
A large number of X-ray observatories have been launched since the first pioneer-
ing rocket flights of the 1960s (see Giacconi 2009 for a review). The majority of the
results presented here have been obtained from the most sensitive X-ray observatories
in operation, all of which employ grazing-incidence optics to focus X-ray photons
and achieve high sensitivity at high spatial resolution: Chandra, XMM-Newton, and
NuSTAR. Chandra (launched July 1999; Weisskopf et al. 2002) provides up-to sub-
arcsecond imaging at ≈0.3–8 keV with sufficient collecting area for high S/N X-ray
spectra of bright X-ray sources. The low background, high spatial resolution, and good
collecting area allows Chandra to detect sources three orders of magnitude fainter than
previous-generation X-ray observatories. XMM-Newton (launched December 1999;
Jansen et al. 2001) provides lower spatial resolution imaging than Chandra (≈5
FWHM) over ≈0.2–12 keV but with substantially larger collecting area, allowing for
higher S/N X-ray spectra of bright X-ray sources than Chandra. NuSTAR (launched
June 2012; Harrison et al. 2013) provides lower spatial resolution imaging than Chan-
dra and XMM-Newton (≈18 FWHM and ≈58 half-power diameter) but is sensitive
to much higher energy photons (≈3–79 keV). In comparison to previous generation
observatories with sensitivity at >10 keV NuSTAR is able to detect sources two orders
of magnitude fainter.
5.2 Selection of AGN in the X-ray band: identification challenges
The (near) universality of X-ray emission from AGN, the high penetrating power of X-
rays, and the low contamination from host-galaxy emission, mean that AGN selection
effects at X-ray energies are generally modest, particularly at high energies (rest-frame
energies >10 keV). Figure 8 demonstrates the broad L X –z plane coverage for AGN
selected from blank-field cosmic X-ray surveys: AGN with L X > 1044 erg s−1 , L X =
1042 –1044 erg s−1 , and L X < 1042 erg s−1 are typically classified as high luminosity
(broadly corresponding to “quasars”), moderate luminosity (broadly corresponding to
“Seyfert galaxies”), and LLAGN, respectively.
For the majority of AGN, the most significant selection effect is that of absorption,
which is a function of rest-frame X-ray energy: lower energy X-rays are more eas-
ily absorbed than higher energy X-rays (Wilms et al. 2000). The impact of varying
123
Fig. 8 X-ray luminosity versus redshift for sources detected in a selection of blank-field cosmic sur-
veys undertaken with Swift-BAT (red), ASCA (magenta), XMM-Newton (blue), and Chandra (cyan and
black). The open and filled symbols indicate X-ray unabsorbed (NH < 1022 cm−2 ) and X-ray absorbed
(NH > 1022 cm−2 ) systems, respectively. © AAS. Figure reproduced from Ueda et al. (2014), Fig. 3, with
permission
Fig. 9 The impact of varying amounts of absorption on the fraction of X-ray detected counts (with respect
to the case for no absorption) for an AGN at z = 0 for a variety of X-ray observatories: ROSAT (grey
≈0.2–2.4 keV), eROSITA (burgundy ≈0.2–5 keV), Chandra (red ≈0.3–8 keV), XMM-Newton (green
≈0.2–12 keV), NuSTAR (blue ≈3–79 keV), and Swift-BAT (black ≈14–195 keV). A cutoff power law with
= 1.9 and E c > 200 keV has been assumed. © AAS. Figure reproduced from Koss et al. (2016), Fig. 2,
with permission
amounts of absorption on the detection of X-ray photons for a z = 0 AGN from a selec-
tion of X-ray observatories is shown in Fig. 9. The effect of absorption on the fraction
of X-ray detected counts is clear for all of the X-ray observatories but is only significant
123
at >10 keV for heavily obscured AGN (NH > 3 × 1023 cm−2 ). Since an increase in
redshift leads to an increase in the rest-frame energies probed within a given observed-
frame band, the effect of absorption is less significant at high redshift than low redshift.
However, CT AGN (i.e. AGN where the absorbing column density exceeds the inverse
of the Thomson scattering cross section: NH > 1/σT > 1.5 × 1024 cm−2 ) are
challenging to detect even at high rest-frame energies due to Compton recoil and
subsequent absorption of the X-ray photons (see Comastri 2004 for a review). Other
wavebands that are less sensitive to the effects of heavy obscuration (e.g. the radio
and IR wavebands; see Sects. 2 and 3) will be less biased towards the detection of CT
AGN. However, the unambiguous identification of the signatures of CT absorption
and the measurement of absorbing column densities requires X-ray observations. This
is most effectively achieved from broad-band X-ray spectral fitting (i.e. the combi-
nation of [ideally simultaneous] data from, for example, NuSTAR and Chandra or
XMM-Newton), which can break the degeneracy between the contributions from vari-
ous X-ray emitting components (e.g. photo-electric absorption; reflection; scattering;
Baloković et al. 2014; Del Moro et al. 2014; Gandhi et al. 2014; Akylas et al. 2016).
For AGN with low observed X-ray luminosities (e.g. intrinsically LLAGN, includ-
ing LEGs, or heavily obscured AGN), the host galaxy can start to contaminate the
emission from the AGN, leading to challenges in the identification and characterisation
of the AGN. The dominant host-galaxy phenomenon at X-ray energies is commonly
referred to as X-ray binaries (see Fabbiano 2006 and Remillard and McClintock 2006
for reviews). The emission from X-ray binaries is due to mass accretion onto a degen-
erate star (neutron star or BH) from a companion star in a binary system: X-ray
binaries are sub-classified into low-mass X-ray binaries (LMXBs) and high-mass
X-ray binaries (HMXBs), depending on the mass of the companion star. The emis-
sion from LMXBs is closely tied to the stellar mass of the galaxy and can reach
L X ≈ 1041 erg s−1 for a massive galaxy of ≈1011 M (e.g. Lehmer et al. 2010; Boro-
son et al. 2011) while the emission from HMXBs is closely tied to the SFR and can
reach L X ≈ 3×1042 erg s−1 for an extreme SFG with a SFR of ≈ 1000 M year−1 (e.g.
Lehmer et al. 2010; Mineo et al. 2012a). To accurately identify or characterise the AGN
requires taking account of both of these X-ray binary components. However, since the
X-ray emission from host-galaxy processes rarely exceeds L X > 1042 erg s−1 , and
has not been known to exceed L X > 1043 erg s−1 (e.g. Alexander et al. 2005; Wang
et al. 2013), an X-ray source with L X > 1042 erg s−1 is likely to be an AGN. Further-
more, the integrated emission from a population of X-ray binaries is mostly produced
at low energies (<10 keV) and therefore AGN can be more reliably identified and
characterised at higher energies (see Sect. 2.3 of Brandt and Alexander 2015, for a list
of additional criteria to identify X-ray emission from an AGN).
Another potential component of non-AGN contamination at X-ray energies is
emission from hot gas, either from the host galaxy or a galaxy cluster. The hot-gas
component can be up-to L X ≈ 1041 –1042 erg s−1 from galaxies (e.g. Boroson et al.
2011; Mineo et al. 2012b) and can be as high as L X ≈ 1044 –1045 erg s−1 in the cores
of massive galaxy clusters (e.g. Stanek et al. 2006; Ebeling et al. 2010). However, since
the X-ray emission from the hot gas is thermal, it is mostly produced at low energies
(<2–5 keV) and therefore the AGN can be more reliably identified and characterised
at higher energies, even in the most massive galaxy cluster cores.
123
On the basis of the aforementioned factors, it is clear that X-ray observations provide
an efficient and reliable selection of the overall AGN population. Furthermore, the
modest X-ray selection biases are well understood and can be reliably modelled to
allow for robust measurements on the evolution of AGN and the growth of BHs
(e.g. Gilli et al. 2007; Georgantopoulos et al. 2013; Ueda et al. 2014; Aird et al.
2015; Buchner et al. 2015; Miyaji et al. 2015; Merloni 2016). However, significant
uncertainties remain on the contributions to the cosmic BH growth from LLAGN
and CT AGN. Various complementary approaches can be utilised to make further
progress. For example, combining constraints in the X-ray band with AGN and host-
galaxy measurements in other wavebands (e.g. IR and radio wavelengths) to assess
the expected strength of the AGN and host-galaxy components at X-ray energies (e.g.
Alexander et al. 2008; Gandhi et al. 2009; Georgantopoulos et al. 2011; Del Moro et al.
2013, 2016; Asmus et al. 2015) or utilising high spatial resolution X-ray imaging and
spectroscopy to directly identify the AGN and host-galaxy components in nearby
low-luminosity systems (e.g. Annuar et al. 2015, 2017; Ricci et al. 2016).
5.3 Selection of AGN in the X-ray band: AGN types, MBH , L/L Edd , and cosmic
evolution
Essentially all types of AGN are selected at X-ray energies: absorbed and unabsorbed
AGN of low, moderate, and high luminosity (see Fig. 8). Overall there is also good
agreement between the optical and X-ray signatures of absorption (e.g. Malizia et al.
2012; Merloni et al. 2014; Burtscher et al. 2016): the majority (>80–90%) of optical
type 1 AGN are X-ray unabsorbed (NH < 1022 cm−2 ) while the majority (>80–90%)
of optical type 2 AGN are X-ray absorbed (NH > 1022 cm−2 ; see Fig. 10). Careful
Fig. 10 Distribution of
absorbing column densities for
X-ray selected AGN from the
INTEGRAL observatory
(20–100 keV) with type 1
(optically unobscured; top) and
type 2 (optically obscured;
bottom) optical spectral
signatures. The dashed bins
indicate column density upper
limits and the dashed lines
indicate the typical adopted
column density threshold
between X-ray absorbed and
unabsorbed AGN. Figure
reproduced from Malizia et al.
(2012), Fig. 5, with permission.
© The Authors
123
consideration of the thresholds in the classification of X-ray and optical absorbed

and unabsorbed AGN can provide even closer agreement (Burtscher et al. 2016). The
overall consistency between the X-ray and optical absorption indicators provide some
of the strongest observational support for the basic unified AGN model (e.g. Antonucci
1993; Urry and Padovani 1995).
However, the clear disagreements between absorption indicators for a small subset
of the AGN population (e.g. X-ray unabsorbed type 2 AGN and X-ray absorbed type
1 AGN; e.g. Panessa and Bassani 2002; Bianchi et al. 2012) also provide interesting
insight into the overall universality of the basic unified model (see Netzer 2015 for
a review). Time-series observations have shown that the X-ray and optical spectral
properties of AGN can vary on relatively short timescales as a result of changes
in the absorbing column density along the line of sight (e.g. Risaliti et al. 2002,
2005; Matt et al. 2003; MacLeod et al. 2016; Ruan et al. 2016; see also Sect. 7.1.3).
Therefore, some of the occasional disagreements between absorption indicators are
due to non-simultaneous X-ray and optical observations; however, AGN variability
does not explain the differences for all cases and some AGN appear to genuinely depart
from the basic unified model (see also Sect. 3.1).
Due to the (near) universality of X-ray emission from AGN, its production should
not be inherently biased towards specific ranges in MBH and L/L Edd . Indeed, X-ray
emission has been detected from AGN with a broad range of MBH out to high redshifts
(≈105 –109 M ; e.g. Hickox et al. 2009; Page et al. 2014; Baldassare et al. 2015; Chen
et al. 2017). Despite the lack of an intrinsic bias towards detecting low-mass BHs at X-
ray energies, the majority of the AGN detected in blank-field cosmic X-ray surveys are
nevertheless found to reside in massive galaxies (stellar masses >1010 M , implying
MBH > 107 M ; e.g. Brusa et al. 2009; Hickox et al. 2009; Xue et al. 2010; see Brandt
and Alexander 2015 for a review). However, the deficiency of X-ray AGN in low-mass
galaxies (i.e., a low AGN fraction) appears to be due to selection effects: for a fixed X-
ray flux limit (i.e. a luminosity limit at a given redshift), the probability of detecting a
low L/L Edd AGN with a massive BH is significantly higher than a high L/L Edd AGN
with a low-mass BH (e.g. Aird et al. 2012; Bongiorno et al. 2012). Taking account
of these observational biases, the distribution of L/L Edd for X-ray AGN appears to
be broadly consistent with a power law and is largely mass independent (e.g. Aird
et al. 2012; Bongiorno et al. 2012). However, there is evidence for differences in the
normalisation (and potentially the shape) of the L/L Edd distribution between quiescent
and star-forming host galaxies (e.g. Aird et al. 2012; Azadi et al. 2015; see Fig. 11).
Blank-field cosmic X-ray surveys have provided some of the most detailed and
sensitive constraints on the evolution of the AGN population. It is now clear from a
large suite of studies that the evolution in the space density of X-ray selected AGN is
dependent on luminosity: moderate-luminosity AGN peak at z ≈ 0.5–1 while high-
luminosity AGN peak at z ≈ 2–3 (e.g. Ueda et al. 2014; Aird et al. 2015; Buchner
et al. 2015; Miyaji et al. 2015); see Fig. 12. This luminosity dependent evolution is
commonly referred to as “AGN downsizing” (see also Sect. 2.3) and, to first order,
is likely to be driven by the availability of a cold-gas supply in the vicinity of the
accreting BH; however, the processes that influence the availability of the gas are
likely to be manifold (e.g. SF; AGN and stellar feedback; large-scale environment; see
Alexander and Hickox 2012 for a general review). In terms of the cosmic BH growth
123
Fig. 11 Specific accretion rate

(a proxy for L/L Edd )
distributions for X-ray selected
AGN at z = 0.2−1.2 residing in
a range of host galaxy types
dividing the sample into four
populations according to their
epoch-normalized specific SFRs
(quiescent galaxies with low
SFR, quiescent galaxies with
higher SFR, SFGs below the
star-forming main sequence
(MS), and SFGs above the
star-forming MS). © AAS.
Figure reproduced from Azadi
et al. (2015), Fig. 12, with
permission
Fig. 12 Space density versus

redshift for AGN selected across
a wide range in X-ray
luminosity. The symbols indicate
the data for AGN from Miyaji
et al. (2015) in different X-ray
luminosity bins while the curves
and shaded regions indicate the
models for the evolution of
X-ray selected AGN from Aird
et al. (2015). Figure reproduced
from Aird et al. (2015), Fig. 18,
with permission. © The Authors
(essentially the product of the space density and AGN luminosity), the majority has
occurred at z ≈ 1–2, with broadly similar amounts of growth at z < 1 and z > 2: high-
luminosity AGN dominate the BH growth density at z > 1.5 (L X ≈ 1044 –1045 erg s−1
AGN contribute the majority) while moderate-luminosity AGN dominate at z < 0.5
(L X ≈ 1043 –1044 erg s−1 AGN contribute the majority).
The fraction of X-ray absorbed AGN is found to be a function of luminosity, with
a decreasing fraction towards higher X-ray luminosities, and appears to also increase
123
with redshift (e.g. La Franca et al. 2005; Ueda et al. 2014; Aird et al. 2015; Buchner et al.
2015; but see also Sect. 3.1 for IR-selected AGN). The redshift evolution in absorption
may be a consequence of the luminosity dependent X-ray absorbed AGN fraction
shifting to higher luminosities at higher redshifts (e.g. Aird et al. 2015; Buchner et al.
2015). These results suggest that the covering factor of the obscuring material (i.e.
the AGN torus) might change as a function of luminosity and redshift; however, it is
currently unclear what drives this behaviour.
5.4 The future of AGN studies in the X-ray band
We now look towards the scientific gains that can be anticipated from several future
X-ray facilities; see Sect. 6.4 of Brandt and Alexander (2015) for a more detailed
discussion of these observatories and other proposed X-ray facilities.
eROSITA36 (Merloni et al. 2012) is a joint Russian-German mission planned for
launch in 2018. The principle objective of eROSITA is to undertake a sensitive all-
sky survey, achieving sensitivity limits ≈20 and ≈200 times deeper than ROSAT
(0.5–2 keV; Voges et al. 1999) and HEAO 1 A-2 (2–10 keV; Piccinotti et al. 1982),
respectively. The great advance that eROSITA will provide over the more sensitive
Chandra and XMM-Newton observatories is huge AGN statistics (≈3 million AGN
out to z ≈ 6), effectively providing an X-ray equivalent of the SDSS (York et al.
2000) or WISE, to explore the cosmic growth of BHs and large-scale structure at
X-ray energies.
The Advanced Telescope for High ENergy Astrophysics (Athena37 ; Nandra et al.
2013) is an ESA-led mission planned for launch in 2028. Athena will be a revolutionary
general-purpose X-ray observatory with a wide field of view (FoV; 40 × 40 ), large
collecting area (2 m2 at 1 keV), good spatial resolution (5 ), and excellent spectral
resolution (2.5 eV). Athena has many key scientific aims but, from the point of view
of this review, two of the main advances will come from the following: (1) excellent-
quality X-ray spectroscopy (both in terms of unprecedented sensitivity and spectral
resolution), to elucidate the physics of AGN activity (e.g. Cappi et al. 2013; Dovciak
et al. 2013); (2) unsurpassed sensitivity for wide–deep blank-field cosmic surveys,
to construct a near-complete census of AGN activity out to z ≈1–3 and identify
moderately luminous AGN out to z ≈6–10 (down to L X ≈ 1043 erg s−1 ; e.g. Aird
et al. 2013; Georgakakis et al. 2013).
Finally, the X-ray Imaging Polarimetry Explorer (XIPE38 ; Soffitta et al. 2016) is a
new mission concept selected by ESA in 2015 to undergo a 2 year-long assessment
phase in the context of the Cosmic Vision competition. XIPE is dedicated to undertake
temporally, spatially, and spectrally resolved X-ray polarimetry. From the point of view
of AGN the polarised emission is expected to be due to the scattering and reflection of
photons and to originate in the vicinity of the accretion disk, the absorbing region (the
AGN torus), and larger-scale ionisation “cones”. Therefore, XIPE will provide unique
36 http://www.mpe.mpg.de/eROSITA.
37 http://www.the-athena-x-ray-observatory.eu.
38 http://www.isdc.unige.ch/xipe.
123
physical insight on the geometry and connections between these different regions
(Goosmann and Matt 2011).
6 γ -ray-selected AGN
We discuss in this section γ -ray-selected AGN. The γ -ray band is conventionally split
into the High Energy (HE) band, between 100 MeV and ∼100 GeV, and the Very
High Energy (VHE) band, covering the ∼50 GeV to ∼10 TeV range. The types of
AGN selected, their SEDs, and the physical mechanism(s) behind γ -ray emission are
detailed in Sect. 6.1, while γ -ray detectors are described in Sect. 6.2. Sects. 6.3 and
Sect. 6.4 deal with the HE and VHE bands, respectively, while Sect. 6.5 addresses the
blazar population and its contribution to the γ -ray background. Sect. 6.6 discusses
selection biases and, finally, Sect. 6.7 addresses the future of this field.
6.1 γ -ray AGN, their SEDs, and physical mechanism(s) behind γ -ray emission
The γ -ray band is the most energetic part of the electromagnetic spectrum and, as such,
is beyond the reach of most astronomical sources. This applies also to most types of
extragalactic objects, including the non-jetted AGN detected in large numbers in IR,
optical, and X-ray surveys thanks to the radiation resulting from accretion onto the
central SMBH (Sects. 3, 4, and 5; see Fig. 1).
Blazars, despite being intrinsically very rare (orders of magnitude less abundant
than non-jetted AGN of the same optical magnitude), dominate the extragalactic γ -
ray sky, which include also a few other AGN, mostly nearby RGs (blazars are also
prevalent in the bright radio sky: Sect. 2.1.2). Non-jetted AGN have not been detected
in the γ -rays39 (Ackermann et al. 2012a). The power output of blazars covers the entire
electromagnetic spectrum and is dominated by non-thermal, blue-shifted, and Doppler
boosted radiation arising in a relativistic jet pointed in the direction of the observer (e.g.
Urry and Padovani 1995, see also Sect. 2.1.5). It is because of this intense non-thermal
emission, and the very special geometrical conditions, that blazars reach the most
extreme parts of the electromagnetic spectrum with fluxes well above the sensitivity
of the instruments operating today. Blazars, as discussed in Sect. 2.1.2, are further
divided into two subgroups defined by their optical spectra, i.e. BL Lacs and FSRQs.
The SED of blazars (see Figs. 13, 14 for two examples of well-known objects)
covers the entire electromagnetic spectrum, from radio waves to γ -ray energies, and
is characterized by a typical “double humped” shape (in a ν L(ν) vs. ν [or ν f (ν)
vs. ν] space). The low energy component, peaking between the IR and the X-ray
band, is generally attributed to synchrotron radiation produced by relativistic electrons
moving in a magnetic field. Sources where this component peaks at low energies
(νsynch peak < 1014 Hz) are called LSPs, while objects with SED peaking at high
energies (νsynch peak > 1015 Hz) are called HSPS (Padovani and Giommi 1995; Abdo
et al. 2010a). Objects with νsynch peak located at intermediate energies are called ISPs.
39 With the exception of NGC 1068 and NGC 4945, two Seyfert 2 galaxies in which the γ -ray emission is
thought to be related to their starburst component (Ackermann et al. 2012b).
123
Fig. 13 The SED of the BL Lac Mrk 421. Strong and variable emission is present at all energies, from
the radio band to γ -rays. Variability is most pronounced in the X-rays and γ -rays where the two SED
components peak. The green line represents the expected emission from a typical blazar host galaxy. A
hard spectrum is present in the HE band with variability of approximately a factor 50 in the 200 MeV (light
green), 1 GeV (orange) and 10 GeV (purple) data. The γ -ray SED peaks in the VHE band, where large
variability is also present. VHE data are from Bartoli et al. (2011), Biteau and Williams (2015), Aharonian
et al. (2005), Chandra et al. (2010), Sharma et al. (2015), Amenomori et al. (2003) and Albert et al. (2007).
The SEDs here and in Fig. 14 have been generated using the SED builder tool of the ASI Science Data Center
(ASDC), available at tools.asdc.asi.it/SED, while the light curves have been produced at the ICRANet site
of Yerevan as part of a collaboration with ASDC, using Fermi public software and archival data
Fig. 14 The SED of the FSRQ 3C 273. The strong and highly variable non-thermal radiation from the jet
encompasses the entire electromagnetic spectrum, but is not dominant in the optical-UV and the soft X-ray
bands where most of the emission is due to accretion onto the SMBH and to the BLR. The γ -ray spectrum
is steep with extremely high variable intensity (up to a factor of 10 000: see the light green, orange and
purple light curves at 0.2, 1.0 and 10.0 GeV, respectively) in the HE band, while very little or nothing is
present in the VHE band
123
The nature of the second SED component, that extends well into the γ -ray band, is
still debated as two alternative (or complementary) interpretations are being consid-
ered. In leptonic models (e.g. Maraschi et al. 1992) the emission is explained as inverse
Compton scattering between the electrons in the jet and their own synchrotron emission
(synchrotron self-Compton) or an external photon field (external inverse Compton). In
hadronic scenarios (e.g. Böttcher et al. 2013) γ -rays are instead assumed to originate
from high-energy protons either loosing energy through synchrotron emission (Aharo-
nian 2000) or through photo–meson interactions (Mannheim 1993). In this case blazars
would also be neutrino emitters (from the decay of charged pions) extending their SEDs
outside the electromagnetic spectrum into newly explored multi-messenger scenarios,
which might even include cosmic rays (CRs) (e.g. Padovani et al. 2016; Resconi et al.
2017).
Figures 13 and 14 show the SEDs of the HSP BL Lac Mrk 421 and of the LSP FSRQ
3C 273 built using large amounts of archival data from many space- and ground-based
observatories covering almost all frequencies and a time span of several years. The
HE and VHE γ -ray bands are highlighted. A clear difference in the γ -ray emission
of the two objects is apparent: a flat spectrum in the HE band extending to VHE
energies for Mrk 421, a steep spectrum in the HE band with almost no emission in the
VHE band for 3C 273. Strong variability, another defining characteristics of blazars,
is clearly visible in both objects at all energies, with the largest amplitude occurring
in the X-rays for Mrk 421 and in the γ -ray band for 3C 273.
6.2 γ -ray detectors
Different γ -ray detectors operate in different bands. The HE band is where γ -rays are
detected in electron pair-conversion telescopes. These instruments operate in space
and are characterized by a very large FoV (thousands of square degrees; e.g. Tavani
2009; Atwood et al. 2009). The VHE band is where the present detection capability
is provided by Imaging Atmospheric Cherenkov Telescopes (IACTs) and Extensive
Air Shower (EAS) observatories (e.g. de Naurois and Mazin 2015). These instruments
observe from the ground the particle showers that are produced by the impact of VHE
γ -ray photons on the top layers of the atmosphere, either through the Cherenkov light
they generate, or via the direct detection of the charged particles in the shower. The
small overlap between the two bands (between 50 and 100 GeV) allows for inter-
calibration between space and ground-based observatories. This is particularly impor-
tant as this is where spectral breaks occur in the most extreme astrophysical sources.
6.3 The HE band
Currently four γ -ray space telescopes or instruments sensitive to photons in the

100 MeV–100 GeV band are operational: the “Astro-rivelatore Gamma a Immag-
ini LEggero” (AGILE), which was launched in 2007 and is the oldest in operation,
Fermi (launched in 2008), the Alpha Magnetic Spectrometer (AMS-2, launched in
2011), and the DArk Matter Particle Explorer (DAMPE, launched in 2015). Published
results or publicly available data are, however, so far available only from AGILE and
Fermi. The other facilities, although sensitive to cosmic γ -rays, have different prime
123
objectives, such as the detailed measurement of charged CRs and dark matter, and
have not contributed to the γ -ray astronomy literature so far.
The best sensitivity is provided by the Large Area Telescope on board Fermi (Fermi-
LAT, Atwood et al. 2009), which is approximately two orders of magnitude more
sensitive than the previous generation of γ -ray telescopes like the Energetic Gamma
Ray Experiment Telescope (EGRET) on board of the Compton Gamma Ray Obser-
vatory (Thompson et al. 1993).
The size of the point spread function and the effective area of Fermi-LAT strongly
depend on energy, resulting in a sensitivity limit that significantly depends on the
intrinsic source spectrum.40 This has a strong impact on the blazar types that are
detected in different γ -ray bands, introducing selection biases, as discussed below
(Sect. 6.6).
6.3.1 AGN in the HE γ -ray band
Two main approaches have been followed to produce catalogues of γ -ray sources
in the HE band. In the one adopted by the project teams the detection is based on
γ -ray data only and is therefore “blind” with respect to any additional information
about astrophysical sources. For this reason they use conservative statistical thresh-
olds (e.g. AGILE: Pittori et al. 2009; Fermi-LAT 0,1,2,3FGL: Acero et al. 2015, and
references therein, where the numbers denote the n-th Fermi-LAT source catalogue;
Fermi 1,2,3FHL: The Fermi-LAT Collaboration 2017, and references therein, where
the numbers denote the n-th catalogue of hard sources). Other lists have instead been
obtained by searching for γ -ray excesses at the positions of previously known blazars,
or blazar candidates (e.g. Arsioli and Chang 2017, Arsioli et al., in preparation) in
Fermi-LAT archival data. The two approaches are complementary as they make use
of different information in the application of the statistical detection methods, use
different thresholds and integration times, and are sensitive in different ways to source
confusion, a problem that is becoming an issue as the sensitivity improves after many
years of Fermi-LAT data.
The largest catalogue of γ -ray sources published so far, the 3FGL, is based on the
first 4 years of Fermi-LAT data and includes 3033 objects. Outside the Galactic plane
almost all γ -ray detections that have been associated to known objects are jetted AGN,
the large majority of them being blazars with only a handful of RGs.
Figure 15 shows a plot in Galactic coordinates of the subsample of 1563 high
Galactic latitude (|b| > 10◦ ) sources in the 3FGL catalogue that have been firmly
associated with AGN (Ackermann et al. 2015b). This includes 415 FSRQs (all of
the LSP type), 657 BL Lacs (mostly of the HSP type), and 402 blazars of uncertain
type (assumed to be blazars because of their radio to γ -ray SED but with no optical
spectrum yet available to classify them as FSRQs or BL Lacs).
An example of the second approach to the detection of γ -ray sources is the Brazil
ICRANet Gamma-ray Blazar catalogue (1BIBG; Arsioli and Chang 2017) where the
40 See http://www.slac.stanford.edu/exp/glast/groups/canda/lat_Performance.htm.
123
Fig. 15 The third catalogue of AGN detected by the Fermi-LAT (3LAC; Ackermann et al. 2015b) plotted
in Galactic coordinates. FSRQs (all of the LSP type) are plotted as red filled circles, BL Lacs (mostly of
the HSP type) are shown as blue points; green filled circles represent blazars of uncertain type
authors report 150 new γ -ray sources in the sample of multi-frequency selected Second
WISE High Synchrotron Peaked (2WHSP) blazars.
A comprehensive list of AGN detected by Fermi, compiled at the ASDC from all
published catalogues and other publications, is available on-line.41 At the time of
writing it includes 1959 distinct sources, 536 of which are FSRQs, 687 are BL Lacs,
15 are RGs, and 75 are AGN of unknown type.
6.4 The VHE band
The VHE γ -ray band is where currently operating IACTs and EAS are sensitive, with
Fermi-LAT also partly covering this band up to ∼2 TeV. Indeed, the largest catalogue
of VHE sources, the 3FHL (The Fermi-LAT Collaboration 2017), is based on Fermi-
LAT data.
This energy region is particularly challenging for AGN observations as VHE γ -
rays are subject to pair production interactions with the extragalactic background light
(EBL; Ackermann et al. 2012c) causing strong flux attenuations that are energy and
redshift dependent, thus modifying the observed spectra and limiting the horizon to
the relatively low redshift Universe.
The present generation of IACTs (e.g. MAGIC, H.E.S.S., VERITAS) and the
upcoming Cherenkov Telescope Array (CTA) are characterized by relatively small
FoV (a few square degrees), and very good sensitivity. IACTs are excellent instru-
ments for detecting fast variability, catching flaring states, and measure the status of
the object during the (typically few hours) of observations. Several years of observa-
tions led to the detection of many blazars, especially during large flares, often following
pointings triggered by high states discovered in other wavebands.
41 http://www.asdc.asi.it/fermiagn.
123
Fig. 16 The sample of γ -ray sources detected in the VHE band as reported in TeVCat at the time of writing
plotted in Galactic coordinates. Red and black points represent extragalactic and Galactic plus unidentified
sources respectively
Water Cherenkov/EAS detectors (e.g. HAWK, ARGO, TIBET, and the future Large
High Altitude Air Shower Observatory [LHAASO]) are instead characterized by a
large FoV (thousands of square degrees), moderate sensitivity, and operate nearly
100% of the time (day and night). These observatories are well suited to carry out long-
integration surveys of large parts of the sky, with typical output being fluxes averaged
over long integration periods or the discovery of strong flares in bright objects.
6.4.1 AGN in the VHE γ -ray band
The main lists of objects in this energy band detected by IACTs are available on-
line42 as interactive tables that are updated periodically as new sources are detected.
At present they include approximately 180 sources distributed in the Galaxy as shown
in Fig. 16. These samples do not represent uniform surveys of the VHE sky as they
only include sources detected during pointings of known sources, often during flaring
states. Only a fraction of the high Galactic latitude sky has been observed so far.
At high Galactic latitudes almost all sources in VHE catalogues are blazars, the
majority of which (50 sources) are of the HSP or ISP type, while only 8 are of the LSP
type (6 FSRQs and 2 BL Lacs); two of these, namely S3 0218+35 and PKS1441+25,
are located at z > 0.9, a remarkably high value for this energy band. Four RGs are
also detected (Centaurus A, NGC 1275, PKS 0625−35, and M 87).
The 2FHL catalogue (Ackermann et al. 2016a) is based on 80 months of Fermi-LAT
data and is a real full sky survey compiled with photons having energies between 50
GeV and 2 TeV. It includes 360 sources, many more that those detected by IATCs.
In fact only about 25% of 2FHL sources were previously detected by Cherenkov
telescopes. Similarly to the Fermi-3FGL catalogue off the Galactic plane nearly all
42 tevcat.uchicago.edu and http://www.asdc.asi.it/tgevcat.
123
sources are blazars. However, in this case BL Lacs of the HSP type are the large
majority; only 10 FSRQs have been found so far. This difference with respect to the
HE band is easily explained as due to the much steeper γ -ray spectral slope of FRSQs
(all of which are LSPs) and to EBL absorption that further steepens and reduces the
intensity of the spectrum of the more distant FSRQs.
To complement the 2FHL sample, and in an effort to construct a much larger list
of targets for VHE observations, Chang et al. (2017) assembled the 2WHSP cata-
logue, a very large sample of confirmed and candidate blazars of the HSP type that
are expected to emit in the VHE band. The sample was selected using radio, IR
(WISE), optical, and X-ray survey data, imposing that the SED of the candidates is
similar to that of known HSP blazars, which are the most abundant type of AGN
found in the VHE band. The 2WHSP sample includes 1691 sources and is available
on-line.43 The already mentioned study by Arsioli and Chang (2017), who found
150 γ -ray detections that were never reported before using over 7 years of Fermi-
LAT data, confirms that 2WHSP blazars indeed constitute a very good reservoir of
candidate VHE γ -ray sources for the next generation of detectors such as CTA and
LHAASO.
6.5 Blazar population properties and contribution to the γ -ray extragalactic

background
The large samples that can be assembled from Fermi-LAT catalogues and other pub-
lications have been used to determine the population properties of blazars such as the
γ -ray number counts, the LF and cosmological evolution for both BL Lacs and FSRQs
(Abdo et al. 2010b; Ajello et al. 2015).
Ajello et al. (2015) found that the blazar γ -ray LF can be represented by a bro-
ken power law and that its evolution with redshift is strong for all types of evolution
models (luminosity, density and luminosity dependent density evolution) considered.
By integrating the best fit LF, taking into account the estimated amount of cosmo-
logical evolution, the distribution of blazar spectral slopes and EBL attenuation, the
authors estimated the contribution of AGN to the extragalactic γ -ray background
(EBG, Ackermann et al. 2015a). The results show that the integrated emission from
blazars and RGs can explain both the intensity and the spectral shape of the extra-
galactic background in the 100 MeV–820 GeV energy band. In particular, above
100 GeV a very large fraction of the EGB is due to HSP blazars with SED simi-
lar to that shown in Fig. 13. The fact that AGN are responsible for most, if not all,
of the EGB is particularly important also because it leaves little room to the con-
tribution of diffuse components such as γ -rays from the annihilation of dark matter
particles and from the interaction of ultra HE CRs with the cosmic microwave back-
ground.
43 http://www.asdc.asi.it/2whsp.
123
6.6 Blazars and selection biases in the γ -ray and other energy bands
The complex broad-band SEDs of blazars result from the superposition of many spec-
tral components, such as the double humped non-thermal emission, light from the host
galaxy, the BLR, and the accretion onto the SMBH. This mix, combined with different
viewing angles and a wide range of maximum particle acceleration energies leads to
SEDs with largely different shapes, causing very strong selection effects when looking
at blazars in widely separated regions of the electromagnetic spectrum. For instance,
blazars selected at radio or microwave frequencies are mostly of the FSRQ/LSP type,
whereas X-ray selection leads to samples largely dominated by BL Lacs of the HSP
type. In the γ -ray band selection effects are not less important, especially because
the sensitivity of Fermi-LAT strongly depends on the spectral slope of the detected
sources: steep γ -ray spectra FSRQs/LSPs (like 3C 273, see Fig. 14) are, therefore,
detected less efficiently than hard γ -ray spectra HSP BL Lacs (e.g. Fig. 13), leading
to percentages of the two blazar types that do not represent the real cosmic abundance.
The ∼40 to ∼60% mix between FSRQs and BL Lacs in the 3LAC sample and the
over 95% of HSP BL Lacs in the 2FHL catalogue are just the result of this effect in
the HE and VHE γ -ray energy bands. To understand the intrinsic population proper-
ties of blazars is therefore essential to control all the selection biases. To this purpose
Giommi, Padovani and collaborators published a series of papers (Giommi et al. 2012,
2013; Padovani and Giommi 2015; Giommi and Padovani 2015), where, through a
detailed Monte Carlo approach, called the blazar simplified view (BSV), they showed
that the widely different statistical population properties of blazars observed in the
radio, X-ray and γ -ray bands can be reproduced in detail starting from simple mini-
mal assumptions. In particular, the BSV correctly predicts the different composition
of γ -ray catalogues in the HE and VHE bands, including percentages of FSRQs/LSPs
and BL Lacs/HSPs, spectral slopes, redshift distributions, and contribution to the
EGB.
6.7 The future of AGN studies in the γ -ray band
Building on the technology of current generation IACTs, CTA44 will be ten times
more sensitive and will have unprecedented accuracy in its detection of VHE γ rays,
with more than 100 telescopes located in the northern and southern hemispheres cov-
ering the 20 GeV–300 TeV energy range. Construction should start in 2018, with the
first telescopes on site in 2019. CTA will provide a systematic approach to blazar
studies, as our knowledge of VHE γ -ray emission from blazars is very biased and
patchy, since many of them have been detected only because they were in outburst
(Sect. 6.4).
LHAASO45 (LHAASO Collaboration 2016) is a new generation instrument, to be
built at 4410 m of altitude in the Sichuan province of China, with the aim of studying
with unprecedented sensitivity the energy spectrum, the elemental composition and
44 http://www.cta-observatory.org.
45 english.ihep.cas.cn/ic/ip/LHAASO/.
123
the anisotropy of CRs in the energy range between 1012 and 1017 eV, as well as to act
simultaneously as a wide aperture (∼2 sr), continuously-operated γ -ray telescope in
the energy range between 100 GeV and 1 PeV.
In the BSV paper dedicated to the VHE γ -ray band Padovani and Giommi (2015)
estimated the number of blazars of the FSRQ and BL Lac type and their redshift distri-
butions expected in future deep VHE surveys. The expectations are largely consistent
with the 2FHL catalogue (Ackermann et al. 2016b), which suggests that the BSV
might also be a reliable predictor of the average VHE sky that will be probed by the
upcoming large new VHE facilities like CTA and LHAASO.
A study based on the subset of 2WHSP and 2FHL common sources, shows that the
average (EBL unabsorbed) VHE flux F(E > 50 GeV) of HSP blazars can be predicted
with an uncertainty of the order of a factor two based on the shape and intensity of
their SED synchrotron component. The following preliminary relationship
F(E > 50 GeV) ≈ 2 × 1011 Speak flux 10(−0.154Log(νsynch peak )−8.03) (1)
gives F(E > 50 GeV), the VHE flux in units of photons cm−2 s−1 , as a function
of Speak flux , which is the flux at the peak of the SED synchrotron hump in units of
erg cm−2 s−1 and of νsynch peak in units of Hz.
To emphasize the rapid evolution of the VHE AGN sky, Fig. 17 shows the distribu-
tion of the currently known and predicted VHE AGN in galactic coordinates, including
TevCAT sources and Fermi 2FHL objects. 2WHSP sources that are predicted (on the
basis of Eq. 1) to be at or above the sensitivity of current IACTs and also detectable
by the future CTA are also shown.
Fig. 17 The comparison of this plot with that of Fig. 16 illustrates the rapid evolution of the VHE sky. VHE
sources detected by IACTs are shown as black dots as in Fig. 16, those from the Fermi 2FHL catalogue
(E > 50 GeV) are shown as orange filled circles, while blazars in the 2WHSP sample that are predicted
by Eq. 1 to be VHE emitters with intensity larger than the sensitivity of current IACTs and that of CTA,
appear as large and small green circles, respectively
123
7 Variability-selected AGN
We discuss in this section variability-selected AGN. An overview of AGN variability

in the local Universe is given in Sect. 7.1, while Sect. 7.2 deals with AGN variability
in extragalactic surveys. The future of this field is discussed in Sect. 7.3.
7.1 An overview of AGN variability in the local Universe
Variability of the emitted multi-wavelength (from radio to γ -ray) radiation has been
recognized as one of the main characteristics of AGN as a class (e.g. Angione 1973;
Marshall et al. 1981). Therefore, just like other diagnostics, variability can be used as
a tool to select AGN in extragalactic surveys.
AGN display erratic, aperiodic flux variability over a wide range of timescales (from
years to minutes). The distribution of AGN variability power (i.e. the power spectral
density [PSD] defined as the squared amplitude of the flux, e.g. Uttley et al. 2002) over
timescales (or, equivalently, temporal frequencies) strongly depends on the observing
waveband. In other words, for example, much faster variability is observed in the X-
ray band than in the optical band, where the same variability amplitude is reached only
over longer timescales (Fig. 18). The minimum timescale of variability measured in a
given waveband provides us with an estimate of the linear size of the source compo-
nent emitting in that waveband (e.g. Terrell 1967). The X-ray band is where some of
the most rapid (hours-minutes), largest-amplitude flux variations are measured. This
Fig. 18 Multiwavelength light curves of the non-jetted AGN MR 2251−178. From top to bottom: 2−10 keV
X-rays from RXTE; B band Wise Observatory data in blue open circles and B band SMARTS data in pink
filled circles; V band SMARTS data, and R band Wise Observatory data. Figure reproduced from Arévalo
et al. 2008, Fig. 1, with permission. © The Authors
123
variability is thought to originate in the innermost regions of the accretion flow (corona
and inner disk). Moreover, it is responsible for driving (at least part of) the variabil-
ity from the outer accretion disk, observed at longer wavelengths (UV and optical).
Processes occurring in the jet can also contribute to the observed AGN variability
(e.g. shocks or bulk injection of new particles; Marscher and Gear 1985; Böttcher and
Dermer 2010). Variability associated with the jet can dominate in the radio-to-γ -ray
bands in jetted AGN (e.g. Max-Moerbeck et al. 2014; see also Sect. 6.1 and Figs. 13
and 14). These variations may ultimately originate from stochastic instabilities within
the accretion flow (Malzac 2014). However, whether and how accretion-driven vari-
ability is transferred into the jet is not definitely known. In the following we will focus
on discussing variability directly associated with the accretion process.
Note that one can indirectly probe AGN variability on much longer timescales by
studying powerful past events in nearby galaxies through studies of extended emission
line regions. These can then trace the history of AGN emission over the light travel
time from the nucleus to the gas (typically 104 −105 years: e.g. Dadina et al. 2010;
Keel et al. 2012; Gagne et al. 2014; Davies et al. 2015).
7.1.1 X-ray variability
The availability of long-term monitoring and continuous sampling X-ray observations

has allowed us to make significant progress towards a good characterization of X-
ray variability in low-redshift AGN. These studies played a key role in our general
understanding of AGN variability.
The main property of AGN variability is its “red noise” character, namely the
occurrence of larger amplitude variations on longer timescales. This property became
evident from early EXOSAT long-looks (e.g. Lawrence et al. 1987). The more recent
long monitorings performed by the Rossi X-ray Timing Explorer (RXTE) enabled a
detailed characterization of the X-ray PSD. Particularly when combined with continu-
ous but shorter X-ray observations (e.g. by XMM-Newton) they allowed sampling AGN
variability over a broad range of timescales (from several years to seconds, Fig. 19,
left panel). These studies revealed that the X-ray PSD of AGN has a universal shape,
characterized by a steep (spectral index ∼ − 2) high-frequency (i.e. short-timescales)
slope. Moreover, they showed that in many AGN the PSD flattens out to a slope of
∼ − 1 below a characteristic frequency, dubbed “break frequency”, νb (e.g. Edelson
and Nandra 1999; Uttley et al. 2002). Apart from this characteristic frequency, the
PSD of AGN is mostly featureless over several decades in frequency.46 The existence
of a break in the PSD is indicative of νb corresponding to a physical timescale of
the flow at some characteristic radius. There are a number of fundamental physical
timescales that might be associated with the timescales of accretion flow variability
(e.g. Treves et al. 1988). However, these timescales depend also on some unknown
parameters such as the viscosity and scale-height of the disk. Therefore, it is difficult
to unequivocally associate νb with one of these timescales. Nonetheless, all the above
46 Notable exceptions are: Ark 564, which shows evidences of a second break at lower frequencies and a
more structured PSD (McHardy et al. 2007); REJ 1034+396 and MS 2254.9−3712, where the presence of
quasi-periodic oscillations, QPO, has been reported (Gierliński et al. 2008; Alston et al. 2015).
123
Fig. 19 Left panel X-ray PSDs from combined RXTE and XMM-Newton observations of NGC 4051 (filled
circles) and NGC 3516 (open squares), and comparison with the PSD of the BHXRB Cygnus X-1 in different
accretion states. Figure reproduced from McHardy et al. 2004, Fig. 118, with permission. Right panel The
scaling of the X-ray PSD break (expressed in terms of timescale 1/νb ) as a function of MBH in a sample
of AGN observed by XMM-Newton. Open and filled circles are Narrow Line Seyferts, squares are Seyfert
1s, the green star is a Seyfert 2, the red star is Cygnus X-1. The continuous and dashed lines show the
best-fitting linear model and the ±1 dex region around this model. Figure reproduced from González-Martín
and Vaughan 2012, Fig. 5, with permission
mentioned timescales depend linearly on MBH . Early indications that also νb scales
(inversely) with MBH were reported in several studies (Edelson and Nandra 1999;
McHardy et al. 2004) and finally confirmed by McHardy et al. (2006, see also Körd-
ing et al. 2007), who showed that the correlation becomes tighter after correcting for
the accretion rate/bolometric luminosity. Although the exact functional dependence
on the accretion rate/bolometric luminosity is currently debated (e.g. González-Martín
and Vaughan 2012; Ponti et al. 2012), it is widely accepted that stellar-mass BHs in
X-ray binary systems (BHXRBs) lie on the extension of the νb − MBH relation, which
holds for SMBHs (Fig. 19, right panel; a similar scaling occurs also in accreting neu-
tron stars and white dwarfs: Körding et al. 2007 and Scaringi et al. 2015). Overall, the
striking similarities of the X-ray timing properties of AGN and BHXRBs suggest the
existence of a common physical process driving the observed variability.
7.1.2 UV/optical variability
The UV/optical emission of AGN displays highly correlated variability, as expected if

produced in the same physical region, the accretion disk. A detailed characterization
of UV/optical variability is made difficult by the sparse and irregular sampling of
ground-based observations, as well as the relatively long timescales associated with
the regions of the disk emitting at these wavelengths (days-months). In recent years,
dedicated ground-based campaigns (e.g. Sánchez et al. 2016; Caplar et al. 2017),
the use of new powerful analysis approaches (Kelly et al. 2014), and the increasing
exploitation of data from space observatories (Kepler and Swift; e.g. Smith et al. 2015)
are opening the way to detailed studies of UV/optical variability of AGN. These studies
have been revealing properties significantly different from those characterizing X-ray
123
variability (e.g. Mushotzky et al. 2011; Simm et al. 2016), including steeper PSD high-
frequency slopes (ranging between ∼ − 3 and −4) and the detection of the PSD break
at systematically longer timescales than expected from X-rays (∼100−300 days) with
no clear dependence on MBH .
An important breakthrough in the understanding of UV/optical variability came
from the detection of inter-band, red lags (i.e. with longer wavelengths lagging behind
shorter wavelengths) on day or sub-day timescales (e.g. Sergeev et al. 2005). These
lags are too short to be explained by some form of outwards diffusion in the flow.
Therefore, heating from a central X-ray source is considered the most plausible mech-
anism. Indeed, the derived lag amplitude-wavelength dependence is consistent with
that expected from reprocessing of X-rays in a standard disk (e.g. Cackett et al. 2007;
Edelson et al. 2015). In this scenario the lags are dominated by the light crossing time
from the X-ray source to the region of the disk emitting at a given wavelength, with
longer (red) wavelengths coming from larger radii, thus producing longer lags. Recent
analyses of simultaneous X-ray, UV/optical, and IR observations confirmed the disk
reprocessing hypothesis, and extended it to explain also the variable emission from
the surrounding dusty torus (e.g. McHardy et al. 2014; Vazquez et al. 2015).
Reprocessing of variable X-ray photons in the UV/optical emitting regions of the
disk explains why, on timescales of hours-to-days, AGN are more variable in the
X-rays than in the UV/optical. However, the availability of long X-ray observations
and simultaneous optical monitoring has allowed the extension of these studies to
longer timescales, revealing that on month-to-year timescales variations in the optical
are more intense than in the X-rays (e.g. Arévalo et al. 2009; Breedt et al. 2010;
Uttley et al. 2003). This behaviour cannot be explained by X-ray reprocessing only,
but it requires some additional source of intrinsic disk variability dominating on long
timescales (Sect. 7.1.3).
7.1.3 What does variability tell us about the driving physical mechanism?
The X-ray PSD of AGN has been extensively studied to gather information about
the driving physical mechanism. However, the sole analysis and modelization of the
PSD has not led to a clear identification of the underlying process. Indeed, most of
the originally proposed models (such as shot-noise models: e.g. Lehto 1989) had
difficulties in explaining other X-ray timing properties of both AGN and BHXRBs,
such as the observed “rms-flux” relation, whose existence was first shown by Uttley
and McHardy (2001) and later confirmed by several studies (e.g. Vaughan et al. 2003;
Gaskell 2004; McHardy et al. 2004). This relation indicates that the absolute amplitude
of X-ray variability (rms, e.g. Vaughan et al. 2003) is linearly correlated with the flux
level, meaning that the source is more variable when it is brighter. The existence of
an rms-flux relation implies that the underlying physical process is multiplicative and
that the long-term variations should modulate the short-term variations (Gaskell 2004;
Uttley et al. 2005).
Such findings argue against models (such as additive shot-noise models) invoking
the presence of independent flares or active regions. Rather, it appears that the class
of “propagating-fluctuation” models, first introduced by Lyubarskii (1997), are more
suitable to explain a vast range of X-ray timing properties, including the rms-flux
123
relation (Arévalo and Uttley 2006; Ingram and Klis 2013). These models assume the
emergence of local perturbations of disk parameters triggering variations of the accre-
tion rate, with long timescale perturbations produced at larger radii. If the timescale of
the perturbations is longer than the radial diffusion time, the perturbations can propa-
gate inward, combining multiplicatively with perturbations produced at smaller radii,
and reach the innermost zones where most of the energy is released (e.g. Churazov
et al. 2001; Hogg and Reynolds 2016).
These models can explain the presence of large amplitude X-ray variability on a
wide range of timescales, even orders of magnitude longer than the viscous-timescale
of the compact X-ray emitting regions of the flow. Moreover, the perturbations pro-
duced throughout the disk could be responsible for driving intrinsic UV/optical
disk variability (Uttley et al. 2003). This component would explain the excess
long-timescales optical variability detected in some AGN (Sect. 7.1.2). Propagating-
fluctuation models are also invoked to explain the common detection of “hard” X-ray
lags (time-delays of the hard X-ray band variations with respect to soft X-ray band
variations) in both AGN and BHXRBs (e.g. Miyamoto et al. 1988; McHardy et al.
2004; Uttley et al. 2011; De Marco et al. 2013, 2015). These lags have large ampli-
tudes, typically of the order of 1% of the variability timescale. By interpreting the lags
as due to light-travel time effects, the inferred size of the emitting regions would be
too large (∼103 gravitational radii) to be plausible (e.g. Nowak et al. 1999). However,
if due to the slower, viscous propagation in the flow the large lags can be easily recov-
ered in standard disk-corona geometries (e.g. Kotov et al. 2001; Arévalo and Uttley
2006).
Variations of photoelectric absorption, e.g. associated with temporary obscuration
of the central engine as a consequence of gas clouds located in the BLR or the torus
crossing the observer’s line of sight can also contribute to the observed AGN variability
(Risaliti et al. 2011; Cappi et al. 2016). However, this is unlikely to be responsible
for the bulk of it, as several arguments and observational evidences rather favour an
intrinsic origin. These include: the ubiquity of AGN variability; the universal shape of
the PSD; the scaling of characteristic timescales with MBH ; the similarities of timing
properties in AGN and BHXRBs; the fact that the inner/outer disk and the BLR
directly respond to continuum variations as inferred from results of X-ray (Fabian
et al. 2009; Zoghbi et al. 2012; De Marco et al. 2013; Kara et al. 2016) and UV/optical
reverberation mapping (e.g. Peterson et al. 2004; Edelson et al. 2015). Moreover, recent
studies of long-term (∼15 years) X-ray variability of distant AGN (z = 0.6−3.1: Yang
et al. 2016) showed that intrinsic X-ray variability prevails over absorption variability
(see also Hernández-García et al. 2015; Soldi et al. 2014), further supporting this
conclusion.
7.2 AGN variability in extragalactic surveys
7.2.1 Selecting AGN by variability
Extragalactic surveys are typically characterized by uneven sampling, with large

gaps between consecutive cycles. This makes PSD techniques unsuitable for study-
123
ing variability in these datasets. The identification of variable sources requires

assessing whether the variability observed between observations exceeds the vari-
ations due to statistical fluctuations. This is usually done by computing the quantity
Nobs
X 2 = i=1 (xi − x̄)2 /σerr,i
2 , where x denotes the photon flux (e.g. Lanzuisi et al.
i
2014; De Cicco et al. 2015; Simm et al. 2016; Kozłowski et al. 2016) during each
observation, x̄ is the average photon flux, and σerr,i
2 is the measurement error. The X 2
statistics tends to a χ 2 distribution for large photon fluxes (in a low-photon flux regime
assessing whether a source is significantly variable requires carrying out MonteCarlo
simulations; e.g. Paolillo et al. 2004; Young et al. 2012). Thus, the ability to detect
significant variability in single sources depends on the error of each measurement. At
a given source flux and exposure, this in turn depends on the collecting area of the
detector and the background contribution, which affects the uncertainty on the flux
measurements (e.g. Lanzuisi et al. 2014).
Selection on the basis of variability can return a relatively high number of AGN
candidates. The latest estimates from optical variability over a temporal baseline of
3 years (De Cicco et al. 2017, in preparation) are 275 deg−2 AGN at a magnitude limit
r < 23. This corresponds to ∼48% completeness with respect to X-ray confirmed
AGN at the depth reached in the COSMOS field (Fig. 20, left panel). However, the
resulting AGN density strongly depends on the characteristics of the survey—mainly
the spanned temporal baseline and sampling cadence, the observing wavelength, the
depth of the single exposures—as well as on the AGN type. Some of these aspects are
discussed in Sect. 7.2.2.
Fig. 20 Left panel Fraction of optically variable AGN in the VST-SUDARE/VOICE survey of the COSMOS
field as a function of temporal baseline. Figure reproduced from De Cicco et al. 2017, in preparation, with
permission. Right panel Distribution of X-ray variability index V (which parametrizes the confidence level
of variability) for sources selected from the XMM-COSMOS survey divided by class. The dashed vertical
line marks the value V = 1.3 used to discriminate between variable and non-variable sources. Sources with
V > 1.3 are intrinsically variable at the 95% confidence level. © AAS. Figure reproduced from Lanzuisi
et al. 2014, Fig. 5, with permission
123
7.2.2 Selection effects
As the bulk of AGN variability appears to be intrinsically associated with the accretion
process, any AGN can in principle be selected by variability provided a significant
fraction of the emission from the nucleus is directly observable. This requirement
makes the X-ray band particularly suitable for the identification of AGN, indepen-
dently of their classification, thanks to the penetrating power of X-rays and the
relatively weak host galaxy emission (Sect. 5.2). As a matter of fact, the number
of X-ray variable AGN appears to be type-independent (see Fig. 20, right panel),
with X-ray variable type 2 AGN typically showing less variability at longer wave-
lengths and on similar time scales (e.g. Lanzuisi et al. 2014; Hernández-García
et al. 2015). Indeed, in the optical band the completeness of AGN samples selected
through variability decreases significantly for type 2 AGN (De Cicco et al. 2015
estimate 25 and 6% completeness, respectively, for type 1 and type 2 AGN over a
temporal baseline of 5 months), while in the X-ray band it is similar for the two
classes (e.g. ∼40 and ∼32% completeness respectively for type 1 and type 2 AGN,
Lanzuisi et al. 2014). Nonetheless, it is important to note that high column den-
sities of circumnuclear absorbing material can significantly affect the fraction of
detected X-ray variable AGN, e.g. as a consequence of the increased contribution
from a large scale, constant reflector (Paolillo et al. 2004; Hernández-García et al.
2015).
X-ray variability is also suitable for identifying LLAGN, which are missed by other
selection criteria (Young et al. 2012). Systematic X-ray variability studies showed
that, given good statistics, variability can be detected in a high fraction of AGN (e.g.
80−90% Paolillo et al. 2004; Lanzuisi et al. 2014; Soldi et al. 2014; Yang et al. 2016)
independently of the optical classification of the object or its radio loudness (note that
in jetted AGN the X-ray variability can contain a very significant contribution from
the jet: e.g. Lichti et al. 2008; see also Figs. 13 and 14).
The effectiveness of variability as a tool to select AGN is strongly related to the
temporal baseline spanned by the observations and the depth of a single frame. Indeed,
given the red-noise character of the PSD of AGN, the ability of detecting variability
is higher at longer timescales (see Fig. 20, left panel), particularly when the low-
frequency part of the PSD below νb is sampled. In the X-ray band, νb scales inversely
with MBH and directly with the accretion rate/bolometric luminosity (but see discus-
sion in Sect. 7.1.1), thus the sampling of longer timescales also allows detection of
variability over a wider range of MBH and L/L Edd . In the UV/optical this scaling
is not clearly observed (Simm et al. 2016), but the break is measured at systemat-
ically longer timescales than in the X-rays (Sect. 7.1.2). Therefore, the longer the
maximum sampled timescales, the higher the detection fraction and, at least as far
as the X-rays are concerned, the larger the range of MBH and L/L Edd probed. Note
that time dilation effects reduce the effective range of sampled timescales. There-
fore, given a total exposure time, the more distant a quasar is the more likely it is
to intercept the high-frequency part of the PSD, where the variability power drops
rapidly, thus making detection of variability more difficult and limited by statis-
tics.
123
7.2.3 Variability dependence on luminosity and redshift
In the presence of uneven, gappy data, variability can be quantified using the so-called
normalized excess variance (σNXV 2 ; e.g. Nandra et al. 1997; Vaughan et al. 2003; Ponti
et al. 2012; see also Kelly et al. 2014), which provides an estimate of the amplitude of
intrinsic source variability over the sampled timescales. Biases associated with mea-
surements of σNXV2 have been extensively studied by Allevato et al. (2013, see also
Almaini et al. 2000). These biases can be greatly reduced choosing an appropriate
observing strategy (e.g. planning observations separated by constant temporal gaps),
while “ensemble” σNXV 2 estimates (i.e. obtained by averaging over single σNXV 2 esti-
mates from multiple light curves of the same source, or from single light curves of
many sources with similar properties) should be preferred.
Deriving a reliable estimate of σNXV
2 is important in order to properly analyse global
variability properties of different AGN classes, to study correlations with fundamental
parameters, and to investigate the evolution of variability with redshift. These studies
have shown that variability is anti-correlated with luminosity in several wavebands
(e.g. in the X-rays: Ponti et al. 2012; Lanzuisi et al. 2014; Vagnetti et al. 2016 and
references therein; in the optical/MIR: Hook et al. 1994; Kelly et al. 2009; MacLeod
et al. 2010; Zuo et al. 2012; Kozłowski et al. 2016; Simm et al. 2016). The origin of
this anti-correlation is currently unknown, but, at least in the X-rays and on relatively
short timescales (shorter than ∼1/νb ), this correlation seems to be the byproduct of
a more fundamental anti-correlation of variability with MBH ultimately driven by the
scaling of the PSD break (e.g. Czerny et al. 2001; Papadakis 2004; O’Neill et al.
2005; Nikołajuk et al. 2006; Zhou et al. 2010; Ponti et al. 2012; Lanzuisi et al. 2014).
The σNXV2 –MBH relation is very tight, with an estimated scatter of the order of the
uncertainties on MBH (Zhou et al. 2010; Ponti et al. 2012). These findings reinforce
early suggestions that σNXV2 could be used to estimate MBH (Czerny et al. 2001;
Nikolajuk et al. 2004; Nikołajuk et al. 2006; Gierliński et al. 2008).
Several studies also showed that both X-ray and optical variability tend to be
stronger at higher redshifts, but this appears to be a consequence of selection effects
(e.g. Cid Fernandes et al. 1996; Almaini et al. 2000; Manners et al. 2002; Paolillo et al.
2004; Lanzuisi et al. 2014; Morganson et al. 2014; Simm et al. 2016; Yang et al. 2016;
Paolillo et al. 2017). Therefore, to date there are no clear indications of an evolution
of AGN variability with redshift.
7.3 The future of AGN variability studies
Variability is a defining property of AGN, therefore a fundamental tool to understand

the AGN engine and map its close environments. Due to its diagnostic power and
complementarity to other probes, variability is gaining increasing importance in the
census of the AGN population. The recognized relevance of AGN variability studies
is mirrored by the growing number of future missions and facilities that include the
investigation of the variable Universe amongst their science goals. Among these is
eROSITA (Merloni et al. 2012; see also Sect. 5.4), due to launch in spring 2018.
During its four years of operation eROSITA will perform a deep survey of the entire
123
sky at X-ray wavelengths. The planned monitoring strategy (eight all-sky surveys and
a cadence of six months, with the ecliptic poles being monitored at higher pace) will
provide repeated observations of the same portions of the sky. Ultimately, this will
allow sampling AGN X-ray variability over time scales from 250 s (corresponding to
the daily eROSITA exposure) up to 4 years (corresponding to the scheduled duration
of the survey phase). Based on AGN X-ray variability characteristics and scaling
properties (Sect. 7.1.1), eROSITA is expected to detect around 60 000 variable AGN
in the full sky (and many more overall: Sect. 5.4).
In the far future, Athena (Nandra et al. 2013; see also Sect. 5.4) is expected to greatly
advance our knowledge of AGN X-ray variability. Thanks to its large collecting area,
the sensitivity of X-ray variability measurements will be significantly enhanced. This
will have a great impact on our understanding of the accretion process variability, it
will allow the mapping of the close environments of AGN and will assess the role
of AGN winds and outflows and their influence on the observed X-ray variability
properties.
The exploration of the transient and variable sky is one of the driving science
theme of several future facilities that will operate at longer wavelengths. In the optical
band these include the Zwicky Transient Facility (ZTF47 ), a high-cadence, 3 year,
time-domain survey, starting in summer 2017. However, the currently most ambitious
planned survey in the optical band is the LSST, starting operations in 2022 (LSST
Science Collaboration et al. 2009). LSST is designed to uniformly observe an area
of ∼20 000 deg2 , with about 50−200 visits per source for each of the six filters,
over an operation period of ten years. This cadence will probe optical variability on
time scales ranging between 1 minute and one decade, allowing selection of tens of
millions of AGN and the construction of highly-complete catalogues with minimum
contamination. Finally, variability studies of radio-emitting AGN will be one of the
main science goals of the SKA (Sect. 2.4), which is expected to provide a deluge of
AGN radio variability information.
8 Summary
This review highlights the extraordinary diversity of observational signatures produced

by growing SMBHs, and illustrates the challenges of placing all these observations
into a coherent picture. AGN are complex systems exhibiting a wide range of phe-
nomena, and different classes of AGN can be selected across the full range of the
electromagnetic spectrum (Fig. 21). The efficiency of selection of AGN in various
wavebands is the result of: (1) variations in the underlying physical properties of
AGN; (2) observational capabilities; (3) selection effects.
The first consideration relates to physical effects (introduced in Sect. 1 and discussed
throughout this review). The most straightforward of these is the presence or absence
of a strong relativistic jet, leading to the distinction between jetted and non-jetted AGN
as discussed in Sect. 2. As detailed in Sect. 2.1.3 and shown in Fig. 21, the latter class
is fainter in the radio and absent in the γ -ray band (Sect. 6), highlighting how AGN
47 http://www.ptf.caltech.edu/ztf.
123
Fig. 21 The largest AGN surface density over the whole electromagnetic spectrum. Black filled points
refer to all AGN, open red points are for non-jetted AGN. The latter are actually directly measured only
in the radio band, while in the NIR to X-ray bands they have been derived by simply subtracting 10%
from the total values. Non-jetted AGN have not been detected in the γ -ray band. The blue square indicates
variability-selected AGN (Sect. 7.2). Updated from Fig. 11 of Padovani (2016), where one can find the
references to the relevant samples and facilities, to include variability-selected AGN and the results of the
CDFS 7 Ms sample (Luo et al. 2017)
selection can be affected by variations in the physical properties of different AGN

populations. Other physical effects include changes in the accretion flow with L/L Edd
(which can modify the shape of the intrinsic SED, in particular the relative outputs in
the optical/UV and X-ray bands; e.g. Vasudevan and Fabian 2009; Yuan and Narayan
2014), and obscuration by gas and dust that can highly suppress emission in the NIR
through soft X-rays (Sects. 3–5).
Second, observations in some wavebands are more sensitive than others relative to
the typical AGN SED. For example, the current generation of radio telescopes (Sect. 2)
can probe AGN with far smaller bolometric fluxes, compared to contemporary γ -ray
observatories (Sect. 6), yielding a far higher surface density of detected AGN.
Third, AGN selection techniques (due to luminosity, colour, morphology, vari-
ability, etc.) have different levels of efficiency in different wavebands. Of particular
importance is the relative emission of the AGN compared to its host galaxy. At wave-
lengths at which stellar processes in the host galaxy can be particularly luminous
(e.g. optical, IR, and to some extent the radio), it is challenging to select extremely
faint AGN even in the deepest observations (although well-designed techniques using
colours or emission line characteristics can help minimize contamination). By contrast,
host contamination is particularly low in the X-ray band (Sect. 5), and cutting-edge
123
X-ray observatories are capable of extremely deep observations. Simple X-ray lumi-
nosity selection can therefore probe AGN at very faint fluxes, leading the detected
AGN surface density to be highest in this band (Fig. 21).
We note that by multiplying the largest entry in Fig. 21 by the area of the sky we esti-
mate that there are at least ≈1 billion AGN in the Universe that could be detected with
current technology. This number needs to be compared with the number of currently
known AGN, of the order of a million (Fig. 6), which shows the exciting potential for
discovery in future AGN surveys (as detailed in Sects. 2.4, 3.5, 4.6, 5.4, 6.7, and 7.3).
We can also compare it with the number of currently observable galaxies in the Uni-
verse, ≈200 billion (Conselice et al. 2016), indicating that bright AGN are found in
approximately 1% of galaxies.
Table 3 summarizes the selection biases and key capabilities (in other words, the
weaknesses and the strengths) in different bands, as described above and in detail
throughout this review. Despite the significant differences between AGN classes
detected in some bands, the observations highlighted in this review illustrate that
AGN share some common components over a wide dynamic range (up to 12 orders
of magnitude) in physical scales. In what follows, we will describe this basic “archi-
tecture” of an AGN in the context of the observational signatures described in the
review. We will then discuss our present understanding of AGN unification models,
highlighting the need for a more complex picture than the classic “strict” unification,
and discussing the importance of understanding relevant timescales in interpreting
observations. Finally, we will present a broad framework for an emerging picture of
AGN and their host galaxies in light of the exciting recent advances such as those pre-
sented in this review and will conclude by highlighting some future prospects and open
questions.
8.1 Fitting together the components of an AGN
Based on the current observational knowledge of AGN as discussed throughout this

review, we can now assemble the different components, covering a wide range of
physical scales, that together make up the complete picture of the AGN phenomenon.
Black hole (10−7 –10−3 pc) The fundamental element common to all AGN is of course
the central BH itself. Astrophysical BHs are simple objects whose only distinguishing
characteristics are mass and spin, but as we have seen throughout this review, these
two parameters are critically important in determining the observable features of an
AGN. It is a nearly universal feature of AGN selection that more massive BHs are
easier to detect, because the mass of the BH sets the Eddington limit, and so massive
BHs are more luminous at a given L/L Edd (Sect. 3.2.5 and 5.3). (The exception here is
selection via X-ray variability, for which characteristic frequencies vary inversely with
MBH , so that detecting more massive BHs requires longer timescale observations, as
described in Sect. 7.1.1.) The effects of BH spin are far more difficult to understand
observationally, but rapidly spinning BHs may have accretion disks that extend closer
to the event horizon, producing a bluer SED (see below) with a higher radiative effi-
ciency, and may influence the ability of accreting BHs to launch relativistic jets and
radiation line-driven winds.(e.g. Tchekhovskoy et al. 2011; Narayan et al. 2014).
123
123
2 Page 58 of 91
Table 3 A multi-wavelength overview of AGN highlighting the different selection biases (weaknesses) and key capabilities (strengths)
Band Type Physics Selection biases/weaknesses Key capabilities/strengths
Radio, f r 1 mJy Jetted Jet Non-jetted sources High efficiency, no obscuration bias
Radio, f r 1 mJy Jetted and non-jetted Jet and SF Host contamination Completeness, no obscuration bias
IR Type 1 and 2 Hot dust and SF Completeness, reliability, host con- Weak obscuration bias, high effi-
tamination, no dust ciency
Optical Type 1 Disk Completeness, low-luminosity, High efficiency, detailed physics from
obscured sources, host contamination lines
X-ray Type 1 and (most) 2 Corona Very low-luminosity, heavy obscura- Completeness, low host contamina-
tion tion
γ -ray Jetted Jet Non-jetted, unbeamed sources High reliability
Variability All (in principle) Corona, disk, jet Host contamination, obscuration, Low-luminosity
cadence and depth of observations
The definitions of some of the terms used in the bias and capability columns are as follows: Efficiency: ability to identify a large number of AGN with relative small total
exposure times (this is thus a combination of the nature of AGN emission and the capabilities of current telescopes in a given band). Reliability: the fraction of sources that
are identified as AGN using typical criteria that are truly AGN. Completeness: the ability to detect as much as possible of the full underlying population of AGN
P. Padovani et al.
Accretion flow (10−7 –1 pc) The accretion flow onto the BH is the ultimate energy
source in AGN and the most important component in determining their observational
characteristics. Indeed, turning a normal galaxy into an AGN merely requires a SMBH
accreting at a relatively high accretion rate. There is compelling theoretical and obser-
vational evidence that the structure of the flow changes with L/L Edd (e.g. Narayan
and Yi 1994; Vasudevan and Fabian 2009; Yuan and Narayan 2014), with higher
L/L Edd systems producing stronger emission from an optically-thick accretion disk
in the optical and UV (e.g. Slone and Netzer 2012; Netzer and Trakhtenbrot 2014;
Capellupo et al. 2015; Castelló-Mor et al. 2016; Bertemes et al. 2016) and stronger
radiation line-driven winds (e.g. Laor and Davis 2014).
The structure of the flow also appears to influence emission from the magnetic
corona (with lower L/L Edd systems having stronger X-rays relative to the disk emis-
sion: Sect. 5.1) and is also connected to the launching of relativistic jets, with luminous
radio emission observed to be more common in lower L/L Edd systems (see references
in Sect. 2). Changes in the accretion flow are also the source of the bulk of observed
variability in AGN. As discussed in Sect. 7, the variability is observed over a wide
range of timescales from the optical to the X-rays, likely corresponding to character-
istic (e.g. viscous) time scales at various radii in the accretion disk (although there
is also some interplay due to reprocessing of emission from different regions). The
nature of the variability changes with L/L Edd further illustrating the dependence on
the structure of the flow.
Torus (1–10 pc) The putative torus of gas and dust surrounding the central engine in
the AGN is critical to understanding two key observational phenomena: emission from
the AGN in the IR, and obscuration of the accretion disk and corona emission (see
also Netzer 2015 for a review). As discussed in Sect. 3.1, the majority of the MIR light
from luminous AGN is emission from hot dust. The geometric distribution (smooth or
clumpy, axisymmetric or polar) and kinematics (static, inflowing, or outflowing) of the
dusty material are however still uncertain, as is their connection to the accretion flow
on larger and smaller scales. At minimum, it is now well-established that the structure
of the torus is not universal but vary significantly for AGN types and even within an
individual class (Sect. 3.2.4). This variation in the dusty torus also has implications for
AGN obscuration, as the covering factors of obscuring material can differ significantly
for different AGN, and variations in torus structure can affect the observed column
densities, extent, and time variability of the obscuration. We can ultimately conclude
that the dusty torus is a common component of AGN that is critical for understanding
observations, but its physical nature is uncertain, variable, and complex.
Jet (10−7 –106 pc) Another important but poorly understood component of the AGN
phenomenon is the presence and strength of a relativistic jet. Jets dominate the emis-
sion from AGN in the γ -rays and often in the radio band (as discussed in Sects. 2
and 6), and can make important contributions in the X-rays and (in the case of
blazars) across the whole SED. The structure and orientation of the jet can influ-
ence a range of rich observational phenomena, particularly in the radio and γ -rays;
indeed the majority of the AGN classes presented in Table 1 relate in some way
to the presence of a jet. However, despite jet emission being among the best- (and
earliest-) studied aspects of AGN, the physical nature of AGN jets and the preva-
123
lence of relativistic outflows, particularly low-luminosity jets on small scales, remain

uncertain.
Host galaxy and dark matter halo (10–106 pc) In understanding the observational
signatures of AGN it is essential to account for the host galaxies and large-scale struc-
tures in which the BHs reside (see Hickox et al. 2016 for a review). One important
consideration is the effect of host galaxy dilution (or equivalently, “contamination”)
on AGN selection, which are spelled out in detail in this review, particularly in the
radio, IR, optical, and X-ray bands (Sects. 2–5). In all cases, this contamination leads
to uncertainties in the population of faint AGN. It also produces a bias in AGN selec-
tion toward higher L/L Edd , an effect which is strongest in the IR and optical bands,
and serves as motivation for observations at high angular resolution that can separate
the AGN emission from the stellar processes in the surrounding galaxy (e.g. Gandhi
et al. 2009; Simmons et al. 2011). The properties of gas in host galaxies can also have
a physical influence on the fueling and emission from AGN. Some AGN obscuration
originates on galaxy scales much larger than the putative dusty torus (e.g. Chen et al.
2015; Buchner et al. 2017), and connections between physical AGN properties (for
example, the distinction between HEG and LEG line classifications, indicating dif-
ferent L/L Edd : Sect. 2.1.4) and the stellar populations of host galaxies indicate that
the gas supply for BH accretion can be connected to the larger-scale reservoir from
which stars form in the host galaxy (e.g. Hickox et al. 2009; Smolčić et al. 2009;
Smolčić 2009; Azadi et al. 2015). Finally, the interaction between AGN and gas in
surrounding dark matter halos can have important effects on observed phenomena,
particularly with hot spots, lobes, and wide-angle tails in large-scale relativistic jets
(e.g. Hardcastle 2015).
8.2 Unification and the importance of timescales
Piecing together our current understanding of the various components of an AGN and
their observational signatures, we can begin to re-assess the extent to which the full
AGN population can be explained by a “unified” model that reduces the complex AGN
“zoo” described in Sect. 1 (Table 1) to a relatively small set of parameters.
In the most straightforward “strict” unified model, the only parameter is orientation
relative to an axisymmetric dusty torus and relativistic jet (Antonucci 1993; Urry and
Padovani 1995). In broad terms this model has been remarkably successful in con-
necting type 1 and 2 Seyferts, and FSRQs and HEGs. In addition to the presence or
absence broad lines, orientation can also help explain the details of the line profiles;
for example, Shen and Ho (2014) suggest FWHM(Hβ) is more of an indicator of ori-
entation than mass. However, the evidence presented in this review and elsewhere (see
Netzer 2015 for a comprehensive overview) suggests that a picture based primarily on
orientation and obscuring material is incomplete. While the structure of AGN clearly
deviates from spherical symmetry and so there is some orientation dependence on
observed properties, other parameters also have key effects. Of particular importance
is L/L Edd , which can change the structure and emission properties of the accretion
flow and thus the SED shape and excitation properties of surrounding gas (Sect. 8.1),
and can also impact the detectability of AGN over emission from the host galaxy.
123
Fig. 22 Observational signatures in a “weak” unification model with two parameters (radiative efficiency
[related to L/L Edd ], and relativistic jet strength), showing the classes associated with each broad range in
this parameter space. This schematic illustrates the need for parameters beyond orientation to explain the
wide range of AGN properties. A theme that emerges throughout this review is that these classifications
do not represent distinct groups of AGN, but rather general regions in a continuous distribution of AGN
properties. Adapted from schematic by P. Hopkins
Even at a fixed L/L Edd , the structure of the AGN is not universal. For example,
the dusty torus can have a range of covering factors (producing a statistical differ-
ence in the intrinsic properties of objects that are observed to be type 2 as opposed
to type 1: Sect. 3.2.4), and the strength of a relativistic jet can vary, yielding very dif-
ferent multi-wavelength signatures particularly in the radio and γ -rays (as discussed
in Sects. 2 and 6). Further, obscuration of the AGN or dilution of stellar processes
can be produced in the host galaxy, independent of the small-scale orientation of
the central engine (e.g. Chen et al. 2015; Buchner et al. 2017). A more complete
“weak” unified model is one in which all these components (L/L Edd , torus covering
factor, jet strength, and host galaxy properties as well as orientation) come together
to produce the observed signatures of an AGN. A schematic showing the effects of
just two of these parameters, BH accretion rate (BHAR) and jet strength, is shown
in Fig. 22. An important distinction is that these parameters, as well as the others
described above, comprise a continuous distribution. Therefore the separate AGN
classes sample different parts of this distribution rather than representing truly dis-
tinct populations, extending the picture for the dominance of radio jets presented by
Padovani (2016).
It is also important to note that the parameters that determine AGN properties will
naturally be expected to vary with time. This has motivated another sort of “unifica-
tion”, in which AGN represent the same underlying objects viewed at different points
in their evolution. One example of this is the classic merger evolution scenario, in
which a galaxy merger triggers gas flows that fuel an AGN, starting off obscured until
feedback clear the surroundings of the black hole producing an unobscured object
(e.g. Sanders et al. 1988; Hopkins et al. 2008). While there is compelling evidence
123
Fig. 23 An illustration of the stochastic fluctuations in AGN accretion rates (taken from the hydro sim-
ulations of Novak et al. 2011) on timescales shorter than those for typical galaxy evolution processes.
This schematic highlights the weak connection between the instantaneous properties of the AGN and the
evolution of the larger host galaxy. © AAS. Figure reproduced from Hickox et al. (2014), Fig. 1, with
permission
for time evolution in some cases (for example large-scale radio jets are likely to
have evolved from more compact relativistic outflows) it is challenging to construct
a successful unification picture based on time evolution alone. This difficulty arises
from the relative time and spatial scales: while galaxy evolution processes can take
108 years or more and on scales of kpc, variations in nuclear structures, gas flows, and
BHARs can occur on timescales many orders of magnitude shorter (e.g. Hickox et al.
2014; Fig. 23) as evidenced by hydrodynamic simulations (e.g. Novak et al. 2011)
and AGN light echoes (Keel et al. 2012; Schawinski et al. 2015) that point toward
large-amplitude fluctuations in AGN luminosity on timescales of ∼105 –106 years.
Thus an AGN cannot directly “know” about a galaxy merger or other processes that
happen on large galactic scales. While there may be some connection between galaxy-
scale processes and the instantaneous observed signatures of an AGN, these will be
primarily statistical in nature (e.g. Hickox et al. 2009; Goulding et al. 2014; Azadi
et al. 2015).
8.3 The emerging technicolor picture
With this reasoning in mind, there is emerging a general picture to describe the “tech-
nicolor” properties of AGN and their host galaxies that is based on the statistical
relationships between BH and galaxy evolution. In this picture, BHARs are broadly
correlated with the supply of gas (and thus SFR) in the centres of galaxies, but can vary
dramatically over several orders of magnitude in response to processes (instabilities,
feedback, etc.) that occur on small scales. A schematic showing the range of physical
123
Fig. 24 The range of physical processes involved with BH fuelling, in different regimes of BHAR and
gas supply in the host galaxy. Many observed properties of AGN can be attributed to the different physics
operating in these various regimes. Adapted from schematic by P. Hopkins
processes involved in different regimes of BHAR and a galaxy’s central gas supply is
shown in Fig. 24.
The key point is that any type of galaxy can in principle host AGN with a wide range
of observational signatures, but broad correlations between gas supply and BHAR
introduce some important general trends (e.g. high L/L Edd , IR-selected AGN being
found in SFGs, while low L/L Edd LERGs are found primarily in passive galaxies;
Fig. 25).
8.4 The future and open questions
There are exciting prospects for future progress in studies of AGN over the coming
decades. Many of the potential breakthroughs come from new observational resources,
in particular observations at high resolution and sensitivity. On the theoretical side,
increasingly sophisticated and complex simulations of AGN will inform the nature
of the central engine and of AGN feedback. Here we briefly discuss each of these
avenues for progress and conclude with a survey of interesting open questions in the
field.
New observations at high resolution and sensitivity One critical observational capa-
bility in studying the AGN central engine is the ability to directly resolve the relevant
small scales with correspondingly high sensitivity. Recently, NIR adaptive optics and
particularly interferometric observations have provided new insights about the struc-
ture of the AGN torus (Sect. 3.1). ALMA is already resolving the innermost parts of
nearby AGN, such as NGC 1068 (Gallimore et al. 2016; García-Burillo et al. 2016;
Imanishi et al. 2016) and NGC 1097 (Hatziminaoglou et al. in preparation, Izumi et al.
in preparation) down to a few parsec scales, providing new insights into the dynamics
near the SMBH by resolving its sphere of influence, the existence of the torus and the
123
Fig. 25 Left panel L/L Edd distributions for radio, IR, and X-ray selected AGN. Adapted from Hickox
et al. (2009), where one can find details on the area and depth of each survey. Right panel A schematic
picture for evolution of galaxies and their observed AGN classes, based on observations of AGN pop-
ulations to z ∼ 1. All galaxies appear to begin as star-forming blue-cloud systems and end as passive
red-sequence sources, once their dark matter halos have grown sufficiently. However, we show that those
galaxies hosting IR, X-ray, and/or radio AGN appear to follow a similar evolutionary path: radiatively
efficient rapid BH growth (IR/X-ray AGN) appears to be linked with those galaxies with large sup-
plies of cool gas, while mechanically dominated (radio) accretion is associated with passive galaxies,
which may also be responsible for preventing late SF. The red and green stars represent schemat-
ically the subset of galaxies that are observed as IR- and X-ray bright AGN, respectively, in the
different phases of evolution. © AAS. Figure reproduced from Goulding et al. (2014), Fig. 7, with
permission
feeding of the SMBH, and is promising more ground-breaking results once the high
frequency bands are offered in the longest baselines (∼16 km). ALMA, in collabo-
ration with the Global mm-Very Large Baseline Interferometry (VLBI) Array48 and
the Event Horizon Telescope49 consortia, is also trying to directly observe for the first
time the immediate environment of the SMBH in, amongst others, M 87, OJ 287, and
Cen A, with angular resolution comparable to the event horizon, as part of the Cycle
4 mm-VLBI campaign, building on previous work (e.g. Johnson et al. 2015; Fish et al.
2016). With the launch of JWST and the advent of >30 m class ground-based tele-
scopes, we will see a dramatic improvement in the ability to study the morphologies
and spectral properties of the central engine on small scales. In the radio and submm,
advances in interferometry (including along Earth-scale baselines) have the potential
to bring further insights into the nature of AGN jet acceleration, mechanical feed-
back, and gas content and kinematics. Planned or large space missions or concepts in
the X-rays (e.g. Athena, Lynx50 ) and optical/UV (e.g. HabEx,51 LUVOIR52 ), along
with a range of other potential missions (some of which are discussed throughout this
48 http://www3.mpifr-bonn.mpg.de/div/vlbi/globalmm/.
49 http://www.eventhorizontelescope.org.
50 wwwastro.msfc.nasa.gov/lynx/.
51 http://www.jpl.nasa.gov/habex/.
52 asd.gsfc.nasa.gov/luvoir/.
123
review), will enable the study of AGN phenomena with high sensitivity and spectral
and/or spatial resolution.
Large sky survey data sets All areas of astronomy are being revolutionized by increas-
ingly larger and deeper sky surveys at all wavelengths, and this flood of data has the
potential for major breakthroughs in our understanding of AGN. As shown in Fig. 21,
X-ray surveys are the most efficient for detecting a high surface density of AGN
with minimal host galaxy contamination. The next major X-ray survey will be the
eRosita All Sky Survey (eRASS) by the eROSITA instrument (Sect. 5.4), which will
detect ≈3 million AGN over most of the sky, with ground-based optical spectroscopic
follow-up from multiple planned surveys. Further deep, wide surveys with Athena,
Lynx, and other future X-ray missions would probe large numbers of AGN over a
wider parameter space in redshift and luminosity.
In the optical and NIR, the Euclid and WFIRST satellite missions will perform
wide-area surveys characterizing millions of AGN (Sect. 3.5 and 4.6), while dedi-
cated ground-based surveys (for example from the LSST, the Subaru Prime Focus
Spectrograph,53 and the Multi Object Optical and Near-infrared Spectrograph for the
VLT [MOONS]54 ) will build on the legacy of ongoing surveys (for example DES and
SDSS-IV). At longer wavelengths, the growing number of sensitive, wide-area radio
telescopes (leading ultimately toward the SKA) will provide increasing numbers (in
the tens of millions: Sect. 2.4) of radio-selected AGN across most of the sky. Together,
these surveys will produce massive data sets with huge statistical samples of images,
spectra, and time variability of AGN over a wide range in redshift and luminosity, and
will enable statistical analyses of AGN with precision comparable to what is currently
possible for “normal” galaxies. Harnessing these large data sets represents a major
challenge and opportunity for the future of AGN science.
Hydrodynamic simulations Major theoretical challenges in understanding the AGN
arise from the complexity of the physical processes involved and the enormous
dynamic range in the relevant physical scales. With continual advances in computing
power and improvements in astrophysical codes, theorists will continue to approach the
ultimate goal of resolving the physics of AGN on all the scales described in Sect. 8.1,
the structure and kinematics of the torus and central engine (e.g. Chan and Krolik
2016; Namekata and Umemura 2016; Wada et al. 2016), the physics of the inner
accretion flow and the launching of winds and relativistic jets (e.g. Tchekhovskoy
2015; Waters and Proga 2016), and the impact of this feedback on host galaxies and
halos (e.g. Dubois et al. 2016; McAlpine et al. 2017; Weinberger et al. 2017). Coupled
with an improved understanding of the underlying physics, emission mechanisms,
and prescriptions for radiative transfer (e.g. Jud et al. 2017; Hopkins et al. 2017),
these simulations will be critical for interpreting the high-resolution, high-sensitivity
observations described above.
Forward modelling of selection effects As discussed in depth throughout this review,
an important difficulty in piecing together the AGN population is understanding the
wide array of selection effects due to physical variation among AGN, host galaxy
53 pfs.ipmu.jp/.
54 http://www.eso.org/sci/facilities/develop/instruments/MOONS.html.
123
contamination, and observational limitation. Many previous studies have attempted

to invert these selection effects to recover the intrinsic AGN population studied in a
particular waveband. However, with a proliferation of multi-wavelength data on large
samples of AGN, it is increasingly challenging to successfully account for selection
effects across multiple bands simultaneously. There is therefore great potential for
progress from forward-modelling these effects by producing a simulated population
of AGN with various physical characteristics (e.g. luminosities, redshifts, L/L Edd ,
dust obscuration, and host galaxy properties) and then producing simulated SEDs and
other observables for direct comparisons to data (for recent examples of this method
see Jones et al. 2016).
This approach is made particularly powerful in combination with large hydrody-
namic or semi-analytic simulations of galaxy formation (for just a few examples, see
Hirschmann et al. 2014; Schaye et al. 2015; Dubois et al. 2016; Di Matteo et al.
2017), which provide realistic populations of galaxies that enable a robust accounting
for host contamination and connections between AGN and galaxy properties. This
forward-modelling approach, informed by a better understanding of AGN physics and
future high-resolution and high-sensitivity AGN observations, will move us closer to
a complete “unification” of AGN that can account for the properties of the large AGN
populations to be discovered in the next generation of surveys.
These exciting observational and theoretical resources in the coming decades will
be critical in answering the many interesting open questions that remain about the
AGN phenomenon. Following the structure of this review, we discuss some outstand-
ing problems related to AGN observed using each technique, plus general questions
regarding the full AGN population:
– Radio: What is the physical driver of AGN outflows, and in particular relativistic
jets? How does the nature of these outflows change as a function of BH mass
and accretion rate, and what processes (for example spin) explain the existence of
jetted AGN that otherwise look (almost) identical to non-jetted AGN?
– IR: What is the composition, geometry, and morphology of the AGN torus, and
how is its structure related to the accretion flow and/or AGN feedback?
– Optical: What is the physics of AGN emission lines and how can they be used to
probe the physics of outflows? What are the emission line characteristics of high-
redshift obscured quasars, and why are these currently rare in optical surveys?
– X-rays: What is the origin, geometrical configuration, and energy source of the
corona? How do the AGN detected in X-ray surveys (which are the most efficient
method for uncovering AGN) trace the underlying full population of AGN?
– γ -rays: What is the connection between γ -ray emission from AGN and high-
energy particles (neutrinos and CRs), and how can these observations constrain
theoretical models for particle acceleration in AGN?
– Variability: What is the physical driver of AGN variability as observed in different
wavebands? Can X-ray variability be used as an independent indicator of BH
mass?
– General: What is the complete census of the AGN population as observed at
all wavebands (and those still undetected)? How does the cosmic history of BH
accretion, as traced by the complete AGN population, compare to the history of
123
SF? What are the physical connections between the evolution of BHs and that
of their host galaxies and halos, and are these driven by common sources of gas,
feedback processes, or other effects?
In summary, the future of AGN studies is very bright and we will soon be flooded
with exciting new data. We need to be ready to extract as much information as possible
from these observations by asking the right questions and using the appropriate tools,
so that we can most effectively piece together the physical nature of AGN and uncover
the full cosmic evolution of growing black holes.
Acknowledgements We thank the participants of the AGN Workshop “Active Galactic Nuclei: what’s in
a name?” (see full list at http://www.eso.org/sci/meetings/2016/AGN2016/participants.html) for their pre-
sentations, Lisa Kewley, John Silverman, and Sylvain Veilleux for their role in the SOC, Phil Hopkins for his
summary talk, Chris Harrison for producing Fig. 1, and Sarah Gallagher, Darshan Kakkad, Andrea Merloni,
Kevin Schawinski, and an anonymous referee for reading the paper and providing helpful comments. DMA
thanks the Science and Technology Facilities Council (STFC) for support through Grant ST/L00075X/1, and
James Aird and Bret Lehmer for useful critical feedback. RJA was supported by FONDECYT Grant number
1151408. BDM thanks D. De Cicco for helpful discussions and acknowledges support from the European
Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agree-
ment No. 665778 via the Polish National Science Center Grant Polonez UMO-2016/21/P/ST9/04025. RCH
acknowledges support from the National Science Foundation through grants 1515364 and 1554584, and
from NASA through Grants NNX16AN48G, NNX15AP24G, and NNX15AU32H. GTR is grateful for the
support of NASA-ADAP Grant NNX12AI49G, NSF Grant 1411773, and the Alexander von Humboldt
Foundation. VS acknowledges support from the European Union’s Seventh Framework Programme under
Grant agreement 337595 (ERC Starting Grant, ’CoSMass’).
References
Abdo AA, Ackermann M, Ajello M et al (2010a) The spectral energy distribution of Fermi bright blazars.
Astrophys J 716:30–70. doi:10.1088/0004-637X/716/1/30
Abdo AA, Ackermann M, Ajello M et al (2010b) The Fermi-LAT high-latitude survey: source count
distributions and the origin of the extragalactic diffuse background. Astrophys J 720:435. doi:10.
1088/0004-637X/720/1/435
Acero F, Ackermann M, Ajello M, Albert A, Atwood WB, Axelsson M, Baldini L, Ballet J et al (2015)
Fermi large area telescope third source catalog. Astrophys J Suppl 218:23. doi:10.1088/0067-0049/
218/2/23
Ackermann M, Ajello M, Allafort A, Baldini L, Ballet J, Barbiellini G, Bastieri D, Bechtol K et al (2012a)
Search for gamma-ray emission from X-ray-selected Seyfert galaxies with Fermi-LAT. Astrophys J
747:104. doi:10.1088/0004-637X/747/2/104
Ackermann M, Ajello M, Allafort A, Baldini L, Ballet J, Bastieri D, Bechtol K, Bellazzini R et al (2012b)
GeV observations of star-forming galaxies with the Fermi large area telescope. Astrophys J 755:164.
doi:10.1088/0004-637X/755/2/164
Ackermann M et al (2012c) The imprint of the extragalactic background light in the gamma-ray spectra of
blazars. Science 338:1190. doi:10.1126/science.1227160
Ackermann M, Ajello M, Albert A et al (2015a) The spectrum of isotropic diffuse gamma-ray emission
between 100 MeV and 820 GeV. Astrophys J 799:86. doi:10.1088/0004-637X/799/1/86
Ackermann M, Ajello M, Atwood WB, Baldini L, Ballet J, Barbiellini G, Bastieri D, Becerra Gonzalez J
et al (2015b) The third catalog of active galactic nuclei detected by the Fermi large area telescope.
Astrophys J 810:14. doi:10.1088/0004-637X/810/1/14
Ackermann M, Ajello M, Atwood WB et al (2016a) 2FHL: the second catalog of hard Fermi-LAT sources.
Astrophys J Suppl 222:5. doi:10.3847/0067-0049/222/1/5
Ackermann M, Ajello M, Albert A et al (2016b) Resolving the extragalactic gamma-ray background above
50 GeV with the Fermi large area telescope. PhRvL 116:1105. doi:10.1103/PhysRevLett.116.151105
Aharonian FA (2000) TeV gamma rays from BL Lac objects due to synchrotron radiation of extremely high
energy protons. New Astron. 5:377. doi:10.1016/S1384-1076(00)00039-7
123
Aharonian F, Akhperjanian AG, Aye KM et al (2005) Observations of Mkn 421 in 2004 with HESS at large
zenith angles. Astron Astrophys 437:95. doi:10.1051/0004-6361:20053050
Aird J, Coil AL, Moustakas J, Blanton MR, Burles SM, Cool RJ, Eisenstein DJ, Smith MSM et al (2012)
PRIMUS: the dependence of AGN accretion on host stellar mass and color. Astrophys J 746:90. doi:10.
1088/0004-637X/746/1/90
Aird J, Comastri A, Brusa M, Cappelluti N, Moretti A, Vanzella E, Volonteri M, Alexander D et al (2013)
The hot and energetic Universe: the formation and growth of the earliest supermassive black holes.
arXiv:1306.2325
Aird J, Coil AL, Georgakakis A, Nandra K, Barro G, Pérez-González PG (2015) The evolution of the X-
ray luminosity functions of unabsorbed and absorbed AGNs out to z ∼ 5. MNRAS 451:1892–1927.
doi:10.1093/mnras/stv1062
Ajello M, Gasparini D, Sanchez-Conde M et al (2015) The origin of the extragalactic gamma-ray background
and implications for dark matter annihilation. Astrophys J 800:L27–L34. doi:10.1088/2041-8205/800/
2/L27
Akylas A, Georgantopoulos I, Ranalli P, Gkiokas E, Corral A, Lanzuisi G (2016) Compton-thick AGN in
the 70-month swift-BAT all-sky hard X-ray survey: a Bayesian approach. Astron Astrophys 594:A73.
doi:10.1051/0004-6361/201628711
Albert J et al (2007) Observations of Markarian 421 with the MAGIC telescope. Astrophys J 663:125.
doi:10.1086/518221
Alexander DM, Hickox RC (2012) What drives the growth of black holes? New Astron Rev 56:93–121.
doi:10.1016/j.newar.2011.11.003
Alexander DM, Bauer FE, Chapman SC, Smail I, Blain AW, Brandt WN, Ivison RJ (2005) The X-ray
spectral properties of SCUBA galaxies. Astrophys J 632:736–750. doi:10.1086/444342
Alexander DM, Chary R-R, Pope A, Bauer FE, Brandt WN, Daddi E, Dickinson M, Elbaz D et al (2008)
Reliable identification of compton-thick quasars at z ≈ 2: Spitzer mid-infrared spectroscopy of HDF-
oMD49. Astrophys J 687:835–847. doi:10.1086/591928
Alexandroff R, Strauss MA, Greene JE, Zakamska NL, Ross NP, Brandt WN, Liu G, Smith PS et al (2013)
Candidate type II quasars at 2<z<4.3 in the Sloan Digital Sky Survey III. MNRAS 435:3306–3325.
doi:10.1093/mnras/stt1500
Allevato V, Paolillo M, Papadakis I, Pinto C (2013) Measuring X-ray variability in faint/sparsely sampled
active galactic nuclei. Astrophys J 771:9. doi:10.1088/0004-637X/771/1/9
Almaini O, Lawrence A, Shanks T, Edge A, Boyle BJ, Georgantopoulos I, Gunn KF, Stewart GC et al
(2000) X-ray variability in a deep, flux-limited sample of QSOs. MNRAS 315:325–336. doi:10.1046/
j.1365-8711.2000.03385.x
Alonso-Herrero A, Esquej P, Roche PF, Ramos Almeida C, González-Martín O, Packham C, Levenson
NA, Mason RE et al (2016) A mid-infrared spectroscopic atlas of local active galactic nuclei on sub-
arcsecond resolution using GTC/CanariCam. MNRAS 455:563–583. doi:10.1093/mnras/stv2342
Alston WN, Parker ML, Markevičiūtė J, Fabian AC, Middleton M, Lohfink A, Kara E, Pinto C et al (2015)
Discovery of an ∼2-h high-frequency X-ray QPO and iron K α; reverberation in the active galaxy MS
2254.9−3712. MNRAS 449:467–476. doi:10.1093/mnras/stv351
Amenomori M, Ayabe S, Cui SW et al (2003) Multi-TeV gamma-ray flares from Markarian 421 in 2000
and 2001 observed with the Tibet air shower array. Astrophys J 598:242. doi:10.1086/378350
Angione RJ (1973) QSO historical light curves. Astronom J 78:353–368. doi:10.1086/111424
Annuar A, Gandhi P, Alexander DM, Lansbury GB, Arévalo P, Ballantyne DR, Baloković M, Bauer FE et al
(2015) NuSTAR observations of the compton-thick active galactic nucleus and ultraluminous X-ray
source candidate in NGC 5643. Astrophys J 815:36. doi:10.1088/0004-637X/815/1/36
Annuar A, Alexander DM, Gandhi P, Lansbury GB, Asmus D, Ballantyne DR, Bauer FE, Boggs SE et al
(2017) A new Compton-thick AGN in our cosmic backyard: unveiling the buried nucleus in NGC
1448 with NuSTAR. Astrophys J 836:165. doi:10.3847/1538-4357/836/2/165
Antonucci R (1993) Unified models for active galactic nuclei and quasars. Annu Rev Astron Astrophys
31:473–521. doi:10.1146/annurev.aa.31.090193.002353
Antonucci RRJ, Miller JS (1985) Spectropolarimetry and the nature of NGC 1068. Astrophys J 297:621–
632. doi:10.1086/163559
Arévalo P, Uttley P (2006) Investigating a fluctuating-accretion model for the spectral-timing properties of
accreting black hole systems. MNRAS 367:801–814. doi:10.1111/j.1365-2966.2006.09989.x
Arévalo P, Uttley P, Kaspi S, Breedt E, Lira P, McHardy IM (2008) Correlated X-ray/optical variability in
the quasar MR2251-178. MNRAS 389:1479–1488. doi:10.1111/j.1365-2966.2008.13719.x
123
Arévalo P, Uttley P, Lira P, Breedt E, McHardy IM, Churazov E (2009) Correlation and time delays of the
X-ray and optical emission of the Seyfert Galaxy NGC 3783. MNRAS 397:2004–2014. doi:10.1111/
j.1365-2966.2009.15110.x
Armus L, Charmandaris V, Bernard-Salas J et al (2007) Observations of ultraluminous infrared galaxies
with the infrared spectrograph on the Spitzer space telescope. II. The IRAS bright galaxy sample.
Astrophys J 656:148–167. doi:10.1086/510107
Arsioli B, Chang YL (2017) Searching for gamma-ray signature in WHSP blazars: Fermi-LAT detection
of 150 excess signal in the 0.3-500 GeV band. Astron Astrophys 598:A134. doi:10.1051/0004-6361/
201628691
Asmus D, Gandhi P, Hönig SF, Smette A, Duschl WJ (2015) The subarcsecond mid-infrared view of local
active galactic nuclei—II. The mid-infrared-X-ray correlation. MNRAS 454:766–803. doi:10.1093/
mnras/stv1950
Asmus D, Hönig SF, Gandhi P (2016) The subarcsecond mid-infrared view of local active galactic nuclei.
III. Polar dust emission. Astrophys J 822:109. doi:10.3847/0004-637X/822/2/109
Assef RJ, Kochanek CS, Brodwin M, Cool R, Forman W, Gonzalez AH, Hickox RC, Jones C et al (2010)
Low-resolution spectral templates for active galactic nuclei and galaxies from 0.03 to 30 μm. Astrophys
J 713:970–985. doi:10.1088/0004-637X/713/2/970
Assef RJ, Kochanek CS, Ashby MLN, Brodwin M, Brown MJI, Cool R, Forman W, Gonzalez AH et al
(2011) The mid-IR- and X-ray-selected QSO luminosity function. Astrophys J 728:56. doi:10.1088/
0004-637X/728/1/56
Assef RJ, Stern D, Kochanek CS et al (2013) Mid-infrared selection of active galactic nuclei with the wide-
field infrared survey explorer. II. Properties of WISE-selected active galactic nuclei in the NDWFS
Boötes field. Astrophys J 772:26. doi:10.1088/0004-637X/772/1/26
Assef RJ, Eisenhardt PRM, Stern D et al (2015) Half of the most luminous quasars may be obscured:
investigating the nature of WISE-selected hot dust-obscured galaxies. Astrophys J 804:27. doi:10.
1088/0004-637X/804/1/27
Atwood WB, Abdo AA, Ackermann M, Althouse W, Anderson B, Axelsson M, Baldini L, Ballet J, Band
DL, Barbiellini G et al (2009) The large area telescope on the Fermi gamma-ray space telescope
mission. Astrophys J 697:1071–1102. doi:10.1088/0004-637X/697/2/1071
Azadi M, Aird J, Coil AL, Moustakas J, Mendez AJ, Blanton MR, Cool RJ, Eisenstein DJ et al (2015)
PRIMUS: the relationship between star formation and AGN accretion. Astrophys J 806:187. doi:10.
1088/0004-637X/806/2/187
Baade W, Minkowski R (1954) On the identification of radio sources. Astrophys J 119:215–231. doi:10.
1086/145813
Baldassare VF, Reines AE, Gallo E, Greene JE (2015) A ∼50,000 solar mass black hole in the nucleus of
RGG 118. Astrophys J Lett 809:L14. doi:10.1088/2041-8205/809/1/L14
Baldi RD, Capetti A (2010) Spectro-photometric properties of the bulk of the radio-loud AGN population.
Astron Astrophys 519:A48. doi:10.1051/0004-6361/201014446
Baldi RD, Capetti A, Giovannini G (2015) Pilot study of the radio-emitting AGN population: the emerging
new class of FR 0 radio-galaxies. Astron Astrophys 576:A38. doi:10.1051/0004-6361/201425426
Baldwin JA (1997) Luminosity indicators in the spectra of quasi-stellar objects. Astrophys J 214:679–684.
doi:10.1086/155294
Baldwin JA, Phillips MM, Terlevich R (1981) Classification parameters for the emission-line spectra of
extragalactic objects. Publ Astron Soc Pac 93:5–19. doi:10.1086/130766
Baloković M, Smolčić V, Ivezić Ž, Zamorani G, Schinnerer E, Kelly BC (2012) Disclosing the radio loudness
distribution dichotomy in quasars: an unbiased Monte Carlo approach applied to the SDSS-FIRST
quasar sample. Astrophys J 759:30. doi:10.1088/0004-637X/759/1/30
Baloković M, Comastri A, Harrison FA, Alexander DM, Ballantyne DR, Bauer FE, Boggs SE, Brandt WN
et al (2014) The NuSTAR view of nearby Compton-thick active galactic nuclei: the cases of NGC
424, NGC 1320, and IC 2560. Astrophys J 794:111. doi:10.1088/0004-637X/794/2/111
Bañados E, Venemans BP, Decarli R, Farina EP, Mazzucchelli C, Walter F, Fan X, Stern D et al (2016)
The pan-STARRS1 distant z>5.6 quasar survey: more than 100 quasars within the first Gyr of the
Universe. Astrophys J Suppl 227:11. doi:10.3847/0067-0049/227/1/11
Banerji M, McMahon RG, Hewett PC et al (2012) Heavily reddened quasars at z ∼ 2 in the UKIDSS Large
Area Survey: a transitional phase in AGN evolution. MNRAS 427:2275–2291. doi:10.1111/j.1365-
2966.2012.22099.x
123
Banerji M, Alaghband-Zadeh S, Hewett PC, McMahon RG (2015) Heavily reddened type 1 quasars at z>2—
I. Evidence for significant obscured black hole growth at the highest quasar luminosities. MNRAS
447:3368–3389. doi:10.1093/mnras/stu2649
Barthel PD (1989) Is every quasar beamed? Astrophys J 336:606–611. doi:10.1086/167038
Bartoli B, Bernardini P, Bi XJ et al (2011) Long-term monitoring of the TeV emission from Mrk 421 with
the ARGO-YBJ experiment. Astrophys J 734:110. doi:10.1088/0004-637X/734/2/110
Barvainis R (1987) Hot dust and the near-infrared bump in the continuum spectra of quasars and active
galactic nuclei. Astrophys J 320:537–544. doi:10.1086/165571
Baskin A, Laor A (2005) What controls the CIV line profile in active galactic nuclei? MNRAS 356:1029–
1044. doi:10.1111/j.1365-2966.2004.08525.x
Benitez N, Dupke R, Moles M, Sodre L, Cenarro J, Marin-Franch A, Taylor K, Cristobal D et al (2014)
J-PAS: the Javalambre-physics of the accelerated universe astrophysical survey. arXiv:1403.5237
Bentz MC, Peterson BM, Netzer H, Pogge RW, Vestergaard M (2009) The radius-luminosity relationship
for active galactic nuclei: the effect of host-galaxy starlight on luminosity measurements. II. The full
sample of reverberation-mapped AGNs. Astrophys J 697:160–181. doi:10.1088/0004-637X/697/1/
160
Berta S, Lutz D, Santini P, Wuyts S, Rosario D, Brisbin D, Cooray A, Franceschini A et al (2013) Panchro-
matic spectral energy distributions of Herschel sources. Astron Astrophys 551:A100. doi:10.1051/
0004-6361/201220859
Bertemes C, Trakhtenbrot B, Schawinski K, Done C, Elvis M (2016) Testing the completeness of the SDSS
colour selection for ultramassive, slowly spinning black holes. MNRAS. doi:10.1093/mnras/stw2207
Best PN, Heckman TM (2012) On the fundamental dichotomy in the local radio-AGN population: accretion,
evolution and host galaxy properties. MNRAS 421:1569–1582. doi:10.1111/j.1365-2966.2012.20414.
x
Bianchi S, Panessa F, Barcons X, Carrera FJ, La Franca F, Matt G, Onori F, Wolter A et al (2012) Simulta-
neous X-ray and optical observations of true type 2 Seyfert galaxies. MNRAS 426:3225–3240. doi:10.
1111/j.1365-2966.2012.21959.x
Biteau J, Williams DA (2015) The extragalactic background light, the hubble constant, and anomalies:
conclusions from 20 years of TeV gamma-ray observations. Astrophys J 812:60. doi:10.1088/0004-
637X/812/1/60
Bongiorno A, Merloni A, Brusa M, Magnelli B, Salvato M, Mignoli M, Zamorani G, Fiore F et al (2012)
Accreting supermassive black holes in the COSMOS field and the connection to their host galaxies.
MNRAS 427:3103–3133. doi:10.1111/j.1365-2966.2012.22089.x
Bonzini M, Padovani P, Mainieri V, Kellermann KI, Miller N, Rosati P, Tozzi P, Vattakunnel S et al (2013) The
sub-mJy radio sky in the extended Chandra Deep Field-South: source population. MNRAS 436:3759–
3771. doi:10.1093/mnras/stt1879
Bonzini M, Mainieri V, Padovani P, Andreani P, Berta S, Bethermin M, Lutz D, Rodighiero G et al (2015)
Star formation properties of sub-mJy radio sources. MNRAS 453:1079–1094. doi:10.1093/mnras/
stv1675
Boroson TA, Green RF (1992) The emission-line properties of low-redshift quasi-stellar objects. Astrophys
J Suppl 80:109–135. doi:10.1086/191661
Boroson B, Kim D-W, Fabbiano G (2011) Revisiting with Chandra the scaling relations of the X-ray emission
components (binaries, nuclei, and hot gas) of early-type galaxies. Astrophys J 729:12. doi:10.1088/
0004-637X/729/1/12
Böttcher M, Dermer CD (2010) Timing signatures of the internal-shock model for blazars. Astrophys J
711:445–460. doi:10.1088/0004-637X/711/1/445
Böttcher M, Reimer A, Sweeney K, Prakash A (2013) Leptonic and hadronic modeling of Fermi-detected
blazars. Astrophys J 768:54. doi:10.1088/0004-637X/768/1/54
Bovy J, Hennawi JF, Hogg DW, Myers AD, Kirkpatrick JA, Schlegel DJ, Ross NP, Sheldon ES et al (2011)
Think outside the color box: probabilistic target selection and the SDSS-XDQSO quasar targeting
catalog. Astrophys J 729:141. doi:10.1088/0004-637X/729/2/141
Brandt WN, Alexander DM (2015) Cosmic X-ray surveys of distant active galaxies. The demographics,
physics, and ecology of growing supermassive black holes. Astron Astrophys Rev 23:1. doi:10.1007/
s00159-014-0081-z
Breedt E, McHardy IM, Arévalo P, Uttley P, Sergeev SG, Minezaki T, Yoshii Y, Sakata Y et al (2010)
Twelve years of X-ray and optical variability in the Seyfert galaxy NGC 4051. MNRAS 403:605–619.
doi:10.1111/j.1365-2966.2009.16146.x
123
Brescia M, Cavuoti S, Longo G (2015) Automated physical classification in the SDSS DR10. A catalogue
of candidate quasars. MNRAS 450:3893–3903. doi:10.1093/mnras/stv854
Bridle AH, Hough DH, Lonsdale CJ, Burns JO, Laing RA (1994) Deep VLA imaging of twelve extended
3CR quasars. Astron J 108:766–820. doi:10.1086/117112
Brotherton MS, Francis PJ (1999) The intermediate line region and the Baldwin effect. Quasars Cosmol
162:395
Brusa M, Fiore F, Santini P, Grazian A, Comastri A, Zamorani G, Hasinger G, Merloni A et al (2009)
Black hole growth and starburst activity at z = 0.6−4 in the Chandra Deep Field South. Host galaxies
properties of obscured AGN. Astron Astrophys 507:1277–1289. doi:10.1051/0004-6361/200912261
Buchanan CL, Gallimore JF, O’Dea CP, Baum SA, Axon DJ, Robinson A, Elitzur M, Elvis M et al (2006)
Spitzer IRS spectra of a large sample of Seyfert galaxies: a variety of infrared spectral energy distri-
butions in the local active galactic nucleus population. Astron J 132:401–419. doi:10.1086/505022
Buchner J, Georgakakis A, Nandra K, Brightman M, Menzel M-L, Liu Z, Hsu L-T, Salvato M et al (2015)
Obscuration-dependent evolution of active galactic nuclei. Astrophys J 802:89. doi:10.1088/0004-
637X/802/2/89
Buchner J, Schulze S, Bauer FE (2017) Galaxy gas as obscurer—I. GRBs x-ray galaxies and find an
NH3 ∝ M relation. MNRAS 464:4545–4566. doi:10.1093/mnras/stw2423

Burtscher L, Davies RI, Graciá-Carpio J, Koss MJ, Lin M-Y, Lutz D, Nandra P, Netzer H et al (2016) On the
relation of optical obscuration and X-ray absorption in Seyfert galaxies. Astron Astrophys 586:A28.
doi:10.1051/0004-6361/201527575
Buttiglione S, Capetti A, Celotti A, Axon DJ, Chiaberge M, Macchetto FD, Sparks WB (2009) An optical
spectroscopic survey of the 3CR sample of radio galaxies with z<0.3. I. Presentation of the data.
Astron Astrophys 495:1033–1060. doi:10.1051/0004-6361:200811102
Buttiglione S, Capetti A, Celotti A, Axon DJ, Chiaberge M, Macchetto FD, Sparks WB (2010) An optical
spectroscopic survey of the 3CR sample of radio galaxies with z<0.3. II. Spectroscopic classes and
accretion modes in radio-loud AGN. Astron Astrophys 509:A6. doi:10.1051/0004-6361/200913290
Cackett EM, Horne K, Winkler H (2007) Testing thermal reprocessing in active galactic nuclei accretion
disks. MNRAS 380:669–682. doi:10.1111/j.1365-2966.2007.12098.x
Capellupo DM, Netzer H, Lira P, Trakhtenbrot B, Mejía-Restrepo J (2015) Active galactic nuclei at z ∼
1.5—I. Spectral energy distribution and accretion discs. MNRAS 446:3427–3446. doi:10.1093/mnras/
stu2266
Caplar N, Lilly SJ, Trakhtenbrot B (2017) Optical Variability of AGNs in the PTF/iPTF survey. Astrophys
J 834:111. doi:10.3847/1538-4357/834/2/111
Cappi M, Done C, Behar E, Bianchi S, Braito V, Costantini E, Dadina M, Feruglio C et al (2013) The hot
and energetic universe: astrophysics of feedback in local AGN. arXiv:1306.2330
Cappi M, De Marco B, Ponti G, Ursini F, Petrucci P-O, Bianchi S, Kaastra JS, Kriss GA et al (2016)
Anatomy of the AGN in NGC 5548. VIII. XMM-Newton’s EPIC detailed view of an unexpected
variable multilayer absorber. Astron Astrophys 592:A27. doi:10.1051/0004-6361/201628464
Castelló-Mor N, Netzer H, Kaspi S (2016) Super- and sub-Eddington accreting massive black holes: a
comparison of slim and thin accretion discs through study of the spectral energy distribution. MNRAS
458:1839–1858. doi:10.1093/mnras/stw445
Calistro RG, Williams WL, Hardcastle MJ, Duncan K, Rottgering HJA, Best PN, Bruggen M, Chyzy KT,
et al (2017) The LOFAR window on star-forming galaxies and AGNs - curved radio SEDs and IR-radio
correlation at 0<z<2.5. MNRAS 469:3468–3488. doi:10.1093/mnras/stx1040
Chan CH, Krolik JH (2016) Radiation-driven outflows from and radiative support in dusty tori of active
galactic nuclei. Astrophys J 825:67. doi:10.3847/0004-637X/825/1/67
Chandra P, Yadav KK, Rannot RC et al (2010) TeV observations of Mrk 421 with the TACTIC gamma-ray
telescope during 2006–8. J Phys G 37:125201. doi:10.1088/0954-3899/37/12/125201
Chang Y-L, Arsioli B, Giommi P, Padovani P (2017) 2WHSP: a multi-frequency selected catalogue of high
energy and very high energy γ -ray blazars and blazar candidates. Astron Astrophys 598:A17. doi:10.
1051/0004-6361/201629487
Chen C-TJ, Hickox RC, Alberts S et al (2015) A connection between obscuration and star formation in
luminous quasars. Astrophys J 802:50. doi:10.1088/0004-637X/802/1/50
Chen C-TJ, Brandt WN, Reines AE, Lansbury G, Stern D, Alexander DM, Bauer FE, Del Moro A et al
(2017) Hard X-ray selected AGNs in low-mass galaxies from the NuSTAR serendipitous survey.
Astrophys J 837:48. doi:10.3847/1538-4357/aa5d5b
123
Chi S, Barthel PD, Garrett MA (2013) Deep, wide-field, global VLBI observations of the Hubble deep
field north (HDF-N) and flanking fields (HFF). Astron Astrophys 550:A68. doi:10.1051/0004-6361/
201220783
Chiaberge M, Capetti A, Celotti A (1999) The HST view of FR I radio galaxies: evidence for non-thermal
nuclear sources. Astron Astrophys 349:77–87
Chiaberge M, Macchetto FD, Sparks WB, Capetti A, Allen MG, Martel AR (2002) The nuclei of radio
galaxies in the ultraviolet: the signature of different emission processes. Astrophys J 571:247–255.
doi:10.1086/339846
Churazov E, Gilfanov M, Revnivtsev M (2001) Soft state of Cygnus X-1: stable disk and unstable corona.
MNRAS 321:759–766. doi:10.1046/j.1365-8711.2001.04056.x
Cid Fernandes R, Jr Aretxaga I, Terlevich R (1996) The QSO variability-luminosity-redshift relation.
MNRAS 282:1191–1202. doi:10.1093/mnras/282.4.1191
Coatman L, Hewett PC, Banerji M, Richards GT (2016) C IV emission-line properties and systematic trends
in quasar black hole mass estimates. MNRAS 461:647–665. doi:10.1093/mnras/stw1360
Coatman L, Hewett PC, Banerji M, Richards GT, Hennawi JF, Prochaska JX (2017) Correcting C IV-based
virial black hole masses. MNRAS 465:2120–2142. doi:10.1093/mnras/stw2797
Comastri A (2004) Compton-thick AGN: the dark side of the X-ray background. Supermassive Black Holes
Distant Universe 308:245. doi:10.1007/978-1-4020-2471-9_8
Condon JJ (1992) Radio emission from normal galaxies. Annu Rev Astron Astrophys 30:575–611. doi:10.
1146/annurev.aa.30.090192.003043
Condon JJ, Kellermann KI, Kimball AE, Ivezić Ž, Perley RA (2013) Active galactic nucleus and starburst
radio emission from optically selected quasi-stellar objects. Astrophys J 768:37. doi:10.1088/0004-
637X/768/1/37
Conselice CJ, Wilkinson A, Duncan K, Mortlock A (2016) The evolution of galaxy number density at z<8
and its implications. Astrophys J 830:83. doi:10.3847/0004-637X/830/2/83
Cutri RM, Wright EL, Conrow T et al (2012) Explanatory supplement to the WISE all-sky data release
products. Technical report
Czerny B, Nikołajuk M, Piasecki M, Kuraszkiewicz J (2001) Black hole masses from power density spectra:
determinations and consequences. MNRAS 325:865–874. doi:10.1046/j.1365-8711.2001.04522.x
D’Abrusco R, Longo G, Walton NA (2009) Quasar candidates selection in the virtual observatory era.
MNRAS 396:223–262. doi:10.1111/j.1365-2966.2009.14754.x
Dadina M, Guainazzi M, Cappi M, Bianchi S, Vignali C, Malaguti G, Comastri A (2010) X-ray imaging
of the ionisation cones in NGC 5252. Astron Astrophys 516:A9. doi:10.1051/0004-6361/200913727
Davies RL, Schirmer M, Turner JEH (2015) The ‘Green Bean’ galaxy SDSS J224024.1-092748: unravelling
the emission signature of a quasar ionization echo. MNRAS 449:1731–1752. doi:10.1093/mnras/
stv3432801d2799
De Breuck C, van Breugel W, Röttgering H, Stern D, Miley G, de Vries W, Stanford SA, Kurk J et al
(2001) Spectroscopy of ultra-steep-spectrum radio sources. Astronom J 121:1241–1265. doi:10.1086/
319392
De Cicco D, Paolillo M, Covone G, Falocco S, Longo G, Grado A, Limatola L, Botticella MT et al (2015)
Variability-selected active galactic nuclei in the VST-SUDARE/VOICE survey of the COSMOS field.
de Grijp MHK, Miley GK, Lub J, de Jong T (1985) Infrared Seyferts—a new population of active galaxies?
Nature 314:240–242. doi:10.1038/314240a0
de Grijp MHK, Lub J, Miley GK (1987) Warm IRAS sources. I. A. Catalogue of AGN candidates from the
point source catalog. Astron Astrophys Suppl 70:95–114
De Marco B, Ponti G, Cappi M, Dadina M, Uttley P, Cackett EM, Fabian AC, Miniutti G et al (2013)
Discovery of a relation between black hole mass and soft X-ray time lags in active galactic nuclei.
MNRAS 431:2441–2452. doi:10.1093/mnras/stt339
De Marco B, Ponti G, Muñoz-Darias T, Nandra K (2015) Tracing the reverberation lag in the hard state of
black hole X-ray binaries. Astrophys J 814:50. doi:10.1088/0004-637X/814/1/50
Del Moro A, Alexander DM, Mullaney JR, Daddi E, Pannella M, Bauer FE, Pope A, Dickinson M et al (2013)
GOODS-Herschel: radio-excess signature of hidden AGN activity in distant star-forming galaxies.
Del Moro A, Mullaney JR, Alexander DM, Comastri A, Bauer FE, Treister E, Stern D, Civano F et al (2014)
NuSTAR J033202–2746.8: direct constraints on the compton reflection in a heavily obscured quasar
at z ≈ 2. Astrophys J 786:16. doi:10.1088/0004-637X/786/1/16n
123
Del Moro A, Alexander DM, Bauer FE, Daddi E, Kocevski DD, McIntosh DH, Stanley F, Brandt WN et al
(2016) Mid-infrared luminous quasars in the GOODS-Herschel fields: a large population of heavily
obscured, Compton-thick quasars at z ≈ 2. MNRAS 456:2105–2125. doi:10.1093/mnras/stv2748
Delvecchio I, Gruppioni C, Pozzi F, Berta S, Zamorani G, Cimatti A, Lutz D, Scott D et al (2014) Tracing
the cosmic growth of supermassive black holes to z ∼ 3 with Herschel. MNRAS 439:2736–2754.
doi:10.1093/mnras/stu130
Delvecchio I, Smolčić V, Zamorani G, Lagos CDP, Berta S, Delhaize J, Baran N, Rosario DJ et al (2017) The
VLA-COSMOS 3 GHz large project: AGN and host-galaxy properties out to z 6. Astron Astrophys
602:A3. doi:10.1051/0004-6361/201629367
de Naurois M, Mazin D (2015) Ground-based detectors in very-high-energy gamma-ray astronomy. CR
Phys 16:610–627. doi:10.1016/j.crhy.2015.08.011
Dennett-Thorpe J, Scheuer PAG, Laing RA, Bridle AH, Pooley GG, Reich W (2002) Jet reorientation in
active galactic nuclei: two winged radio galaxies. MNRAS 330:609–620. doi:10.1046/j.1365-8711.
2002.05106.x
Denney KD (2012) Are outflows biasing single-epoch C IV black hole mass estimates? Astrophys J 759:44.
doi:10.1088/0004-637X/759/1/44
Denney KD, Pogge RW, Assef RJ, Kochanek CS, Peterson BM, Vestergaard M (2013) C IV line-width
anomalies: the perils of low signal-to-noise spectra. Astrophys J 775:60. doi:10.1088/0004-637X/
775/1/60
Deo RP, Richards GT, Crenshaw DM, Kraemer SB (2009) The mid-infrared continua of Seyfert galaxies.
Astrophys J 705:14–31. doi:10.1088/0004-637X/705/1/14
Dey A, Soifer BT, Desai V et al (2008) A significant population of very luminous dust-obscured galaxies
at redshift z ∼ 2. Astrophys J 677:943–956. doi:10.1086/529516
Díaz-Santos T, Assef RJ, Blain AW et al (2016) The strikingly uniform, highly turbulent interstellar medium
of the most luminous galaxy in the universe. Astrophys J Lett 816:L6. doi:10.3847/2041-8205/816/
1/L6
Di Matteo T, Croft RAC, Feng Y, Waters D, Wilkins S (2017) The origin of the most massive black holes at
high-z: BlueTides and the next quasar frontier. MNRAS 467:4243–4251. doi:10.1093/mnras/stx319
DiPompeo MA, Bovy J, Myers AD, Lang D (2015) Quasar probabilities and redshifts from WISE mid-IR
through GALEX UV photometry. MNRAS 452:3124–3138. doi:10.1093/mnras/stv1562
Done C (2010) Observational characteristics of accretion onto black holes. arXiv:1008.2287
Done C, Gierliński M, Kubota A (2007) Modelling the behaviour of accretion flows in X-ray binaries.
Everything you always wanted to know about accretion but were afraid to ask. Astron Astrophys Rev
15:1–66. doi:10.1007/s00159-007-0006-1
Donley JL, Rieke GH, Pérez-González PG, Rigby JR, Alonso-Herrero A (2007) Spitzer power-law active
galactic nucleus candidates in the Chandra Deep Field-North. Astrophys J 660:167–190. doi:10.1086/
512798
Donley JL, Koekemoer AM, Brusa M, Capak P, Cardamone CN, Civano F, Ilbert O, Impey CD et al (2012)
Identifying luminous active galactic nuclei in deep surveys: revised IRAC selection criteria. Astrophys
J 748:142. doi:10.1088/0004-637X/748/2/142
D’Onofrio M, Marziani P, Sulentic JW (2012) Fifty years of quasars: from early observations and ideas to
future research. Springer, Berlin. doi:10.1007/978-3-642-27564-7
Donoso E, Best PN, Kauffmann G (2009) Evolution of the radio-loud galaxy population. MNRAS 392:617–
629. doi:10.1111/j.1365-2966.2008.14068.x
Dovciak M, Matt G, Bianchi S, Boller T, Brenneman L, Bursa M, D’Ai A, di Salvo T et al (2013) The hot
and energetic universe: the close environments of supermassive black holes. arXiv:1306.2331
Dubois Y, Peirani S, Pichon C et al (2016) The HORIZON-AGN simulation: morphological diversity of
galaxies promoted by AGN feedback. MNRAS 463:3948–3964. doi:10.1093/mnras/stw2265
Dullemond CP, van Bemmel IM (2005) Clumpy tori around active galactic nuclei. Astron Astrophys 436:47–
56. doi:10.1051/0004-6361:20041763
Dunlop JS, Peacock JA (1990) The redshift cut-off in the luminosity function of radio galaxies and quasars.
MNRAS 247:19–42
Ebeling H, Edge AC, Mantz A, Barrett E, Henry JP, Ma CJ, van Speybroeck L (2010) The X-ray brightest
clusters of galaxies from the massive cluster survey. MNRAS 407:83–93. doi:10.1111/j.1365-2966.
2010.16920.x
Edelson R, Nandra K (1999) A cutoff in the X-ray fluctuation power density spectrum of the Seyfert 1
galaxy NGC 3516. Astrophys J 514:682–690. doi:10.1086/306980
123
Edelson R, Gelbord JM, Horne K, McHardy IM, Peterson BM, Arévalo P, Breeveld AA, De Rosa G et al
(2015) Space telescope and optical reverberation mapping project. II. Swift and HST reverberation
mapping of the accretion disk of NGC 5548. Astrophys J 806:129. doi:10.1088/0004-637X/806/1/
129
Eisenhardt PR, Stern D, Brodwin M et al (2004) The infrared array camera (IRAC) shallow survey. Astrophys
J Suppl 154:48–53. doi:10.1086/423180
Eisenhardt PRM, Wu J, Tsai CW et al (2012) The first hyper-luminous infrared galaxy discovered by WISE.
Astrophys J 755:173. doi:10.1088/0004-637X/755/2/173
Elitzur M, Netzer H (2016) Disc outflows and high-luminosity true type 2 AGN. MNRAS 459:585–594.
doi:10.1093/mnras/stw657
Elitzur M, Shlosman I (2006) The AGN-obscuring torus: the end of the “Doughnut” paradigm? Astrophys
J Lett 648:L101–L104. doi:10.1086/508158
Elvis M, Wilkes BJ, McDowell JC, Green RF, Bechtold J, Willner SP, Oey MS, Polomski E et al (1994)
Atlas of quasar energy distributions. Astrophys J Suppl 95:1–68. doi:10.1086/192093
Esin AA (1997) Heating and cooling of hot accretion flows by nonlocal radiation. Astrophys J 482:400–413.
doi:10.1086/304129
Evans DA, Worrall DM, Hardcastle MJ, Kraft RP, Birkinshaw M (2006) Chandra and XMM-Newton
observations of a sample of low-redshift FR I and FR II radio galaxy nuclei. Astrophys J 642:96–112.
doi:10.1086/500658
Fabbiano G (2006) Populations of X-ray sources in galaxies. Annu Rev Astron Astrophys 44:323–366.
doi:10.1146/annurev.astro.44.051905.092519
Fabian AC, Zoghbi A, Ross RR, Uttley P, Gallo LC, Brandt WN, Blustin AJ, Boller T et al (2009) Broad line
emission from iron K- and L-shell transitions in the active galaxy 1H0707-495. Nature 459:540–542.
doi:10.1038/nature08007
Fanaroff BL, Riley JM (1974) The morphology of extragalactic radio sources of high and low luminosity.
MNRAS 167:31P–36P
Fanidakis N, Baugh CM, Benson AJ, Bower RG, Cole S, Done C, Frenk CS (2011) Grand unification of
AGN activity in the ΛCDM cosmology. MNRAS 410:53–74. doi:10.1111/j.1365-2966.2010.17427.
x
Fazio GG, Hora JL, Allen LE et al (2004) The infrared array camera (IRAC) for the Spitzer space telescope.
Astrophys J Suppl 154:10–17. doi:10.1086/422843
Feltre A, Hatziminaoglou E, Fritz J, Franceschini A (2012) Smooth and clumpy dust distributions in AGN:
a direct comparison of two commonly explored infrared emission models. MNRAS 426:120–127.
doi:10.1111/j.1365-2966.2012.21695.x
Feltre A, Hatziminaoglou E, Hernán-Caballero A et al (2013) The roles of star formation and AGN activity
of IRS sources in the HerMES fields. MNRAS 434:2426–2437. doi:10.1093/mnras/stt1177
Feltre A, Charlot S, Gutkin J (2016) Nuclear activity versus star formation: emission-line diagnostics at
ultraviolet and optical wavelengths. MNRAS 456:3354–3374. doi:10.1093/mnras/stv2794
Filho ME, Barthel PD, Ho LC (2006) A radio census of nuclear activity in nearby galaxies. Astron Astrophys
451:71–83. doi:10.1051/0004-6361:20054510
Fish V, Akiyama K, Bouman K et al (2016) Observing and imaging active galactic nuclei with the event
horizon telescope. Galaxies 4:54. doi:10.3390/galaxies4040054
Flesch EW (2015) The half million quasars (HMQ) catalogue. Publ Astron Soc Aust 32:e010. doi:10.1017/
pasa.2015.10
Francis PJ, Hewett PC, Foltz CB, Chaffee FH (1992) An objective classification scheme for QSO spectra.
Astrophys J 398:476–490. doi:10.1086/171870
Francis PJ, Nelson BO, Cutri RM (2004) An unbiased census of active galactic nuclei in the two micron all
sky survey. Astronom J 127:646–655. doi:10.1086/380939
Fritz J, Franceschini A, Hatziminaoglou E (2006) Revisiting the infrared spectra of active galactic nuclei
with a new torus emission model. MNRAS 366:767–786. doi:10.1111/j.1365-2966.2006.09866.x
Gagne JP, Crenshaw DM, Kraemer SB, Schmitt HR, Keel WC, Rafter S, Fischer TC, Bennert VN et al
(2014) Spatially resolved spectra of the “Teacup” active galactic nucleus: tracing the history of a dying
quasar. Astrophys J 792:72. doi:10.1088/0004-637X/792/1/72
Gallimore JF, Elitzur M, Maiolino R, O’Dea CP, Lutz D, Baum SA, Nikutta R, Marconi A et al (2016)
High-velocity bipolar molecular emission from an AGN torus. Astrophys J Lett 829:L7. doi:10.3847/
2041-8205/829/1/L7
123
Gandhi P, Horst H, Smette A, Hönig S, Comastri A, Gilli R, Vignali C, Duschl W (2009) Resolving the mid-
infrared cores of local Seyferts. Astron Astrophys 502:457–472. doi:10.1051/0004-6361/200811368
Gandhi P, Lansbury GB, Alexander DM, Stern D, Arévalo P, Ballantyne DR, Baloković M, Bauer FE et al
(2014) NuSTAR unveils a Compton-thick type 2 quasar in Mrk 34. Astrophys J 792:117. doi:10.1088/
0004-637X/792/2/117
García-Burillo S, Combes F, Ramos Almeida C, Usero A, Krips M, Alonso-Herrero A, Aalto S, Casasola
V et al (2016) ALMA resolves the torus of NGC 1068: continuum and molecular line emission.
Astrophys J Lett 823:L12. doi:10.3847/2041-8205/823/1/L12
Gardner JP, Mather JC, Clampin M et al (2006) The James Webb space telescope. Space Sci Rev 123:485–
606. doi:10.1007/s11214-006-8315-7
Gaskell CM (2004) Lognormal X-ray flux variations in an extreme narrow-line Seyfert 1 galaxy. Astrophys
J Lett 612:L21–L24. doi:10.1086/424565
Gendre MA, Best PN, Wall JV, Ker LM (2013) The relation between morphology, accretion modes and
environmental factors in local radio AGN. MNRAS 430:3086–3101. doi:10.1093/mnras/stt116
Georgakakis A, Carrera F, Lanzuisi G, Brightman M, Buchner J, Aird J, Page M, Cappi M et al (2013) The
hot and energetic universe: understanding the build-up of supermassive black holes and galaxies at
the heyday of the universe. arXiv:1306.2328
Georgantopoulos I, Rovilos E, Akylas A, Comastri A, Ranalli P, Vignali C, Balestra I, Gilli R et al (2011)
On the L x -L 6μm ratio as a diagnostic for Compton-thick AGN. Astron Astrophys 534:A23. doi:10.
1051/0004-6361/201117400
Georgantopoulos I, Comastri A, Vignali C, Ranalli P, Rovilos E, Iwasawa K, Gilli R, Cappelluti N et al
(2013) The XMM deep survey in the CDF-S. IV. Compton-thick AGN candidates. Astron Astrophys
555:A43. doi:10.1051/0004-6361/201220828
Ghisellini G (2010) The jet/disk connection in blazars. Am Inst Phys Conf Ser 1242:43–54. doi:10.1063/
1.3460151
Giacconi R (2009) History of X-ray telescopes and astronomy. Exp Astron 25:143–156. doi:10.1007/
s10686-009-9139-8
Gierliński M, Middleton M, Ward M, Done C (2008) A periodicity of ∼1 hour in X-ray emission from the
active galaxy RE J1034+396. Nature 455:369–371. doi:10.1038/nature07277
Gierliński M, Nikołajuk M, Czerny B (2008) High-frequency X-ray variability as a mass estimator of stellar
and supermassive black holes. MNRAS 383:741–749. doi:10.1111/j.1365-2966.2007.12584.x
Gilfanov M, Merloni A (2014) Observational appearance of black holes in X-ray binaries and AGN. Space
Sci Rev 183:121–148. doi:10.1007/s11214-014-0071-5
Gilli R, Comastri A, Hasinger G (2007) The synthesis of the cosmic X-ray background in the Chandra and
XMM-Newton era. Astron Astrophys 463:79–96. doi:10.1051/0004-6361:20066334
Giommi P, Padovani P, Polenta G, Turriziani S, D’Elia V, Piranomonte S (2012) A simplified view of blazars:
clearing the fog around long-standing selection effects. MNRAS 420:2899–2911. doi:10.1111/j.1365-
2966.2011.20044.x
Giommi P, Padovani P, Polenta G (2013) A simplified view of blazars: the γ -ray case. MNRAS 431:1914–
1922. doi:10.1093/mnras/stt305
Giommi P, Padovani P (2015) A simplified view of blazars: contribution to the X-ray and gamma-ray cosmic
backgrounds. MNRAS 450:2404. doi:10.1093/mnras/stv793
Glikman E, Helfand DJ, White RL et al (2007) The FIRST-2MASS red quasar survey. Astrophys J 667:673–
703. doi:10.1086/521073
Glikman E, Djorgovski SG, Stern D, Dey A, Jannuzi BT, Lee K-S (2011) The faint end of the quasar
luminosity function at z ∼ 4: implications for ionization of the intergalactic medium and cosmic
downsizing. Astrophys J Lett 728:L26. doi:10.1088/2041-8205/728/2/L26
González-Martín O, Vaughan S (2012) X-ray variability of 104 active galactic nuclei. XMM-Newton power-
spectrum density profiles. Astron Astrophys 544:A80. doi:10.1051/0004-6361/201219008
Goosmann RW, Matt G (2011) Spotting the misaligned outflows in NGC 1068 using X-ray polarimetry.
MNRAS 415:3119–3128. doi:10.1111/j.1365-2966.2011.18923.x
Gorjian V, Brodwin M, Kochanek CS et al (2008) The mid-infrared properties of X-ray sources. Astrophys
J 679:1040–1046. doi:10.1086/587431
Goulding AD, Alexander DM, Bauer FE, Forman WR, Hickox RC, Jones C, Mullaney JR, Trichas M et al
(2012) Deep silicate absorption features in Compton-thick active galactic nuclei predominantly arise
due to dust in the host galaxy. Astrophys J 755:5. doi:10.1088/0004-637X/755/1/5
123
Goulding AD, Forman WR, Hickox RC et al (2014) Tracing the evolution of active galactic nuclei host
galaxies over the last 9 Gyr of cosmic time. Astrophys J 783:40. doi:10.1088/0004-637X/783/1/40
Greene JE, Ho LC (2007) The mass function of active black holes in the local universe. Astrophys J
667:131–148. doi:10.1086/520497
Gültekin K, Richstone DO, Gebhardt K et al (2009) The M-σ and M-L relations in galactic bulges, and
determinations of their intrinsic scatter. Astrophys J 698:198–221. doi:10.1088/0004-637X/698/1/
198
Gürkan G, Hardcastle MJ, Jarvis MJ, Smith DJB, Bourne N, Dunne L, Maddox S, Ivison RJ et al (2015)
Herschel-ATLAS: the connection between star formation and AGN activity in radio-loud and radio-
quiet active galaxies. MNRAS 452:3776–3794. doi:10.1093/mnras/stv1502
Hainline KN, Reines AE, Greene JE, Stern D (2016) Mid-infrared colors of dwarf galaxies: young starbursts
mimicking active galactic nuclei. Astrophys J 832:119. doi:10.3847/0004-637X/832/2/119
Hamann F, Kanekar N, Prochaska JX, Murphy MT, Ellison S, Malec AL, Milutinovic N, Ubachs W et al
(2011) A high-velocity narrow absorption line outflow in the quasar J212329.46−005052.9. MNRAS
410:1957–1974. doi:10.1111/j.1365-2966.2010.17575.x
Hao L, Spoon HWW, Sloan GC, Marshall JA, Armus L, Tielens AGGM, Sargent B, van Bemmel IM
et al (2005) The detection of silicate emission from quasars at 10 and 18 microns. Astrophys J Lett
625:L75–L78. doi:10.1086/431227
Hao H, Elvis M, Civano F et al (2010) Hot-dust-poor type 1 active galactic nuclei in the COSMOS survey.
Astrophys J Lett 724:L59–L63. doi:10.1088/2041-8205/724/1/L59
Hao H, Elvis M, Civano F, Lawrence A (2011) Hot-dust-poor quasars in mid-infrared and optically selected
samples. Astrophys J 733:108. doi:10.1088/0004-637X/733/2/108
Hardcastle M (2015) Kiloparsec-scale AGN jets. In: Contopoulos I, Gabuzda D, Kylafis N (eds) The
formation and disruption of black hole jets, vol 414. Astrophysics and Space Science Library, p 83.
doi:10.1007/978-3-319-10356-3_4
Hardcastle MJ, Evans DA, Croston JH (2007) Hot and cold gas accretion and feedback in radio-loud active
galaxies. MNRAS 376:1849–1856. doi:10.1111/j.1365-2966.2007.11572.x
Hardcastle MJ, Ching JHY, Virdee JS, Jarvis MJ, Croom SM, Sadler EM, Mauch T, Smith DJB et al (2013)
Herschel-ATLAS/GAMA: a difference between star formation rates in strong-line and weak-line radio
galaxies. MNRAS 429:2407–2424. doi:10.1093/mnras/sts510
Harrison CM (2014) Observational constraints on the influence of active galactic nuclei on the evolution of
galaxies. Ph.D. Thesis. doi:10.1007/978-3-319-28454-5
Harrison FA, Craig WW, Christensen FE, Hailey CJ, Zhang WW, Boggs SE, Stern D, Cook WR et al
(2013) The nuclear spectroscopic telescope array (NuSTAR) high-energy X-ray mission. Astrophys J
770:103. doi:10.1088/0004-637X/770/2/103
Hasinger G (2004) The X-ray background and AGNs. Nucl Phys B Proc Suppl 132:86–96. doi:10.1016/j.
nuclphysbps.2004.04.127
Hatziminaoglou E, Mathez G, Pelló R (2000) Quasar candidate multicolor selection technique: a different
approach. Astron Astrophys 359:9–17
Hatziminaoglou E, Pérez-Fournon I, Polletta M et al (2005) Sloan Digital Sky Survey quasars in the Spitzer
wide-area infrared extragalactic survey (SWIRE) ELAIS N1 field: properties and spectral energy
distributions. Astronom J 129:1198–1211. doi:10.1086/428003
Hatziminaoglou E, Fritz J, Franceschini A, Afonso-Luis A, Hernán-Caballero A, Pérez-Fournon I, Ser-
jeant S, Lonsdale C et al (2008) Properties of dusty tori in active galactic nuclei—I. The case of
SWIRE/SDSS quasars. MNRAS 386:1252–1264. doi:10.1111/j.1365-2966.2008.13119.x
Hatziminaoglou E, Fritz J, Jarrett TH (2009) Properties of dusty tori in active galactic nuclei—II. Type 2
AGN. MNRAS 399:1206–1222. doi:10.1111/j.1365-2966.2009.15390.x
Hatziminaoglou E, Omont A, Stevens JA et al (2010) HerMES: far infrared properties of known AGN in
the HerMES fields. Astron Astrophys 518:L33. doi:10.1051/0004-6361/201014679
Hatziminaoglou E, Hernán-Caballero A, Feltre A, Piñol Ferrer N (2015) A complete census of silicate
features in the mid-infrared spectra of active galaxies. Astrophys J 803:110. doi:10.1088/0004-637X/
803/2/110
Heckman TM (1980) An optical and radio survey of the nuclei of bright galaxies—activity in normal galactic
nuclei. Astron Astrophys 87:152–164
Heckman TM, Best PN (2014) The coevolution of galaxies and supermassive black holes: insights from
surveys of the contemporary universe. Annu Rev Astron Astrophys 52:589–660. doi:10.1146/annurev-
astro-081913-035722
123
Heckman TM, Kauffmann G, Brinchmann J, Charlot S, Tremonti C, White SDM (2004) Present-day growth
of black holes and bulges: the Sloan Digital Sky Survey perspective. Astrophys J 613:109–118. doi:10.
1086/422872
Hernán-Caballero A, Hatziminaoglou E (2011) An atlas of mid-infrared spectra of star-forming and active
Hernán-Caballero A, Hatziminaoglou E, Alonso-Herrero A, Mateos S (2016) The near-to-mid infrared
spectrum of quasars. MNRAS 463:2064–2078. doi:10.1093/mnras/stw2107
Hernández-García L, Masegosa J, González-Martín O, Márquez I (2015) X-ray spectral variability of Seyfert
2 galaxies. Astron Astrophys 579:A90. doi:10.1051/0004-6361/201526127
Herrera Ruiz N, Middelberg E, Norris RP, Maini A (2016) Unveiling the origin of the radio emission in
radio-quiet quasars. Astron Astrophys 589:L2. doi:10.1051/0004-6361/201628302
Hewett PC, Foltz CB (2003) The frequency and radio properties of broad absorption line quasars. Astronom
J 125:1784–1794. doi:10.1086/368392
Hewett PC, Wild V (2010) Improved redshifts for SDSS quasar spectra. MNRAS 405:2302–2316. doi:10.
1111/j.1365-2966.2010.16648.x
Hickox RC, Jones C, Forman WR, Murray SS, Brodwin M, Brown MJI, Eisenhardt PR, Stern D et al (2007)
A large population of mid-infrared-selected, obscured active galaxies in the Bootes field. Astrophys J
671:1365–1387. doi:10.1086/523082
Hickox RC, Jones C, Forman WR, Murray SS, Kochanek CS, Eisenstein D, Jannuzi BT, Dey A et al (2009)
Host galaxies, clustering, Eddington ratios, and evolution of radio, X-ray, and infrared-selected AGNs.
Astrophys J 696:891–919. doi:10.1088/0004-637X/696/1/891
Hickox RC, Mullaney JR, Alexander DM et al (2014) Black hole variability and the star formation-active
galactic nucleus connection: do all star-forming galaxies host an active galactic nucleus? Astrophys J
782:9. doi:10.1088/0004-637X/782/1/9
Hickox RC, Lamassa SM, Silverman JD, Kolodzig A (2016) Host galaxies and large-scale structures of
active galactic nuclei. IAU Focus Meet 29:113–123. doi:10.1017/S1743921316004592
Hine RG, Longair MS (1979) Optical spectra of 3CR radio galaxies. MNRAS 188:111–130. doi:10.1093/
mnras/188.1.111
Hirschmann M, Dolag K, Saro A et al (2014) Cosmological simulations of black hole growth: AGN
luminosities and downsizing. MNRAS 442:2304–2324. doi:10.1093/mnras/stu1023
Ho LC (2008) Nuclear activity in nearby galaxies. Annu Rev Astron Astrophys 46:475–539. doi:10.1146/
annurev.astro.45.051806.110546
Hogg JD, Reynolds CS (2016) Testing the propagating fluctuations model with a long. Global accretion
disk simulation. Astrophys J 826:40. doi:10.3847/0004-637X/826/1/40
Hönig SF, Leipski C, Antonucci R, Haas M (2011) Quantifying the anisotropy in the infrared emission of
powerful active galactic nuclei. Astrophys J 736:26. doi:10.1088/0004-637X/736/1/26
Hook IM, McMahon RG, Boyle BJ, Irwin MJ (1994) The variability of optically selected quasars. MNRAS
268:305. doi:10.1093/mnras/268.2.305
Hopkins PF, Richards GT, Hernquist L (2007) An observational determination of the bolometric quasar
luminosity function. Astrophys J 654:731–753. doi:10.1086/509629
Hopkins PF, Hernquist L, Cox TJ, Kereš D (2008) A cosmological framework for the co-evolution of quasars,
supermassive black holes, and elliptical galaxies. I. Galaxy mergers and quasar activity. Astrophys J
Suppl 175:356–389. doi:10.1086/524362
Hopkins PF, Wetzel A, Keres D et al (2017) FIRE-2 simulations: physics versus numerics in galaxy forma-
tion. MNRAS (submitted). arXiv:1702.06148
Hornschemeier AE, Heckman TM, Ptak AF, Tremonti CA, Colbert EJM (2005) Chandra-SDSS normal and
star-forming galaxies. I. X-ray source properties of galaxies detected by the Chandra X-ray observatory
in SDSS DR2. Astronom J 129:86–103
Houck JR, Roellig TL, van Cleve J et al (2004) The infrared spectrograph (IRS) on the Spitzer space
telescope. Astrophys J Suppl 154:18–24. doi:10.1086/423134
Hsu L-T, Salvato M, Nandra K, Brusa M, Bender R, Buchner J, Donley JL, Kocevski DD et al (2014)
CANDELS/GOODS-S, CDFS, and ECDFS: photometric redshifts for normal and X-ray-detected
galaxies. Astrophys J 796:60. doi:10.1088/0004-637X/796/1/60
Hubeny I, Blaes O, Krolik JH, Agol E (2001) Non-LTE models and theoretical spectra of accretion disks in
active galactic nuclei. IV. Effects of Compton scattering and metal opacities. Astrophys J 559:680–702.
doi:10.1086/322344
123
Ilbert O, Capak P, Salvato M, Aussel H, McCracken HJ, Sanders DB, Scoville N, Kartaltepe J et al (2009)
Cosmos photometric redshifts with 30-bands for 2-deg2 . Astrophys J 690:1236–1249. doi:10.1088/
0004-637X/690/2/1236
Ilbert O, Salvato M, Le Floc’h E, Aussel H, Capak P, McCracken HJ, Mobasher B, Kartaltepe J et al
(2010) Galaxy stellar mass assembly between 0.2<z<2 from the S-COSMOS survey. Astrophys J
709:644–663. doi:10.1088/0004-637X/709/2/644
Imanishi M, Nakanishi K, Izumi T (2016) ALMA 0.1-0.2 arcsec resolution imaging of the NGC (1068)
nucleus: compact dense molecular gas emission at the putative AGN location. Astrophys J Lett
822:L10. doi:10.3847/2041-8205/822/1/L10
Ingram A, van der Klis M (2013) An exact analytic treatment of propagating mass accretion rate fluctuations
in X-ray binaries. MNRAS 434:1476–1485. doi:10.1093/mnras/stt1107
Ivezić Ž, Menou K, Knapp GR, Strauss MA, Lupton RH, Vanden Berk DE, Richards GT, Tremonti C et al
(2002) Optical and radio properties of extragalactic sources observed by the FIRST survey and the
Sloan Digital Sky Survey. Astronom J 124:2364–2400. doi:10.1086/344069
Jansen F, Lumb D, Altieri B, Clavel J, Ehle M, Erd C, Gabriel C, Guainazzi M et al (2001) XMM-Newton
observatory. I. The spacecraft and operations. Astron Astrophys 365:L1–L6. doi:10.1051/0004-6361:
20000036
Jarrett TH, Cohen M, Masci F et al (2011) The Spitzer-WISE survey of the ecliptic poles. Astrophys J
735:112. doi:10.1088/0004-637X/735/2/112
Jiang L, Fan X, Brandt WN et al (2010) Dust-free quasars in the early Universe. Nature 464:380–383.
doi:10.1038/nature08877
Johnson MD, Fish VL, Doeleman SS et al (2015) Resolved magnetic-field structure and variability near
the event horizon of Sagittarius A*. Science 350:1242–1245. doi:10.1126/science.aac7087
Jones SF, Blain AW, Stern D et al (2014) Submillimetre observations of WISE-selected high-redshift,
luminous, dusty galaxies. MNRAS 443:146–157. doi:10.1093/mnras/stu1157
Jones ML, Hickox RC, Black CS et al (2016) The intrinsic Eddington ratio distribution of active galactic
nuclei in star-forming galaxies from the Sloan Digital Sky Survey. Astrophys J 826:12. doi:10.3847/
0004-637X/826/1/12
Jud H, Schartmann M, Mould J, Burtscher L, Tristram KRW (2017) Radiative transfer modelling of parsec-
scale dusty warped discs. MNRAS 465:248–259. doi:10.1093/mnras/stw2755
Jun HD, Im M (2013) Physical properties of luminous dust-poor quasars. Astrophys J 779:104. doi:10.
1088/0004-637X/779/2/104
Kara E, Alston WN, Fabian AC, Cackett EM, Uttley P, Reynolds CS, Zoghbi A (2016) A global look at
X-ray time lags in Seyfert galaxies. MNRAS 462:511–531. doi:10.1093/mnras/stw1695
Kauffmann G, Heckman TM, Tremonti C, Brinchmann J, Charlot S, White SDM, Ridgway SE, Brinkmann
J et al (2003) The host galaxies of active galactic nuclei. MNRAS 346:1055–1077. doi:10.1111/j.
1365-2966.2003.07154.x
Kauffmann G, Heckman TM, White SDM, Charlot S, Tremonti C, Brinchmann J, Bruzual G, Peng EW
et al (2003) Stellar masses and star formation histories for 105 galaxies from the Sloan Digital Sky
Survey. MNRAS 341:33–53. doi:10.1046/j.1365-8711.2003.06291.x
Keel WC, Lintott CJ, Schawinski K, Bennert VN, Thomas D, Manning A, Chojnowski SD, van Arkel H
et al (2012) The history and environment of a faded quasar: hubble space telescope observations of
Hanny’s Voorwerp and IC 2497. Astronom J 144:66. doi:10.1088/0004-6256/144/2/66
Kellermann KI (2015) The road to quasars. IAU Symp 313:190–195. doi:10.1017/S1743921315002185
Kellermann KI, Sramek R, Schmidt M, Shaffer DB, Green R (1989) VLA observations of objects in the
Palomar bright quasar survey. Astronom J 98:1195–1207. doi:10.1086/115207
Kelly BC, Bechtold J, Siemiginowska A (2009) Are the variations in quasar optical flux driven by thermal
fluctuations? Astrophys J 698:895–910. doi:10.1088/0004-637X/698/1/895
Kelly BC, Becker AC, Sobolewska M, Siemiginowska A, Uttley P (2014) Flexible and scalable methods for
quantifying stochastic variability in the era of massive time-domain astronomical data sets. Astrophys
J 788:33. doi:10.1088/0004-637X/788/1/33
Kewley LJ, Dopita MA, Sutherland RS, Heisler CA, Trevena J (2001) Theoretical modeling of starburst
galaxies. Astrophys J 556:121–140. doi:10.1086/321545
Kewley LJ, Groves B, Kauffmann G, Heckman T (2006) The host galaxies and classification of active
galactic nuclei. MNRAS 372:961–976. doi:10.1111/j.1365-2966.2006.10859.x
Kimball AE, Ivezić Ž (2008) A unified catalog of radio objects detected by NVSS, first, WENSS, GB6,
and SDSS. Astronom J 136:684–712. doi:10.1088/0004-6256/136/2/684
123
Kimball AE, Kellermann KI, Condon JJ, Ivezić Ž, Perley RA (2011) The two-component radio luminosity
function of quasi-stellar objects: star formation and active galactic nucleus. Astrophys J Lett 739:L29.
doi:10.1088/2041-8205/739/1/L29
Kirkpatrick A, Pope A, Alexander DM et al (2012) GOODS-Herschel: impact of active galactic nuclei and
star formation activity on infrared spectral energy distributions at high redshift. Astrophys J 759:139.
doi:10.1088/0004-637X/759/2/139
Koenig XP, Leisawitz DT, Benford DJ, Rebull LM, Padgett DL, Assef RJ (2012) Wide-field infrared survey
explorer observations of the evolution of massive star-forming regions. Astrophys J 744:130. doi:10.
1088/0004-637X/744/2/130
Körding EG, Migliari S, Fender R, Belloni T, Knigge C, McHardy I (2007) The variability plane of accreting
compact objects. MNRAS 380:301–310. doi:10.1111/j.1365-2966.2007.12067.x
Koss MJ, Assef R, Baloković M, Stern D, Gandhi P, Lamperti I, Alexander DM, Ballantyne DR et al
(2016) A new population of Compton-thick AGNs identified using the spectral curvature above 10
keV. Astrophys J 825:85. doi:10.3847/0004-637X/825/2/85
Kotov O, Churazov E, Gilfanov M (2001) On the X-ray time-lags in the black hole candidates. MNRAS
327:799–807. doi:10.1046/j.1365-8711.2001.04769.x
Kouzuma S, Yamaoka H (2010) Near-infrared colours of active galactic nuclei. Astron Astrophys 509:A64.
doi:10.1051/0004-6361/200912863
Kozłowski S, Kochanek CS, Ashby MLN, Assef RJ, Brodwin M, Eisenhardt PR, Jannuzi BT, Stern D et al
(2016) Quasar variability in the mid-infrared. Astrophys J 817:119. doi:10.3847/0004-637X/817/2/
119
Krolik JH, Begelman MC (1988) Molecular tori in Seyfert galaxies—feeding the monster and hiding it.
Astrophys J 329:702–711. doi:10.1086/166414
Kühr H, Witzel A, Pauliny-Toth IIK, Nauber U (1981) A catalogue of extragalactic radio sources having
flux densities greater than 1 Jy at 5 GHz. Astron Astrophys Suppl 45:367–430
Lacy M, Storrie-Lombardi LJ, Sajina A, Appleton PN, Armus L, Chapman SC, Choi PI, Fadda D et al
(2004) Obscured and unobscured active galactic nuclei in the Spitzer space telescope first look survey.
Lacy M, Petric AO, Sajina A et al (2007) Optical spectroscopy and X-ray detections of a sample of quasars
and active galactic nuclei selected in the mid-infrared from two Spitzer space telescope wide-area
surveys. Astronom J 133:186–205. doi:10.1086/509617
Lacy M, Ridgway SE, Gates EL et al (2013) The Spitzer mid-infrared active galactic nucleus survey.
I. Optical and near-infrared spectroscopy of obscured candidates and normal active galactic nuclei
selected in the mid-infrared. Astrophys J Suppl 208:24. doi:10.1088/0067-0049/208/2/24
La Franca F, Fiore F, Comastri A, Perola GC, Sacchi N, Brusa M, Cocchia F, Feruglio C et al (2005) The
HELLAS2XMM survey. VII. The hard X-ray luminosity function of AGNs up to z = 4: more absorbed
AGNs at low luminosities and high redshifts. Astrophys J 635:864–879. doi:10.1086/497586
Laigle C, McCracken HJ, Ilbert O, Hsieh BC, Davidzon I, Capak P, Hasinger G, Silverman JD et al (2016)
The COSMOS2015 catalog: exploring the 1<z<6 universe with half a million galaxies. Astrophys J
Suppl 224:24. doi:10.3847/0067-0049/224/2/24
Laing RA, Riley JM, Longair MS (1983) Bright radio sources at 178 MHz—flux densities, optical identi-
fications and the cosmological evolution of powerful radio galaxies. MNRAS 204:151–187. doi:10.
1093/mnras/204.1.151
Laing RA, Jenkins CR, Wall JV, Unger SW (1994) Spectrophotometry of a complete Sample of 3CR radio
sources: implications for unified models. Phys Act Galaxies 54:201
LaMassa SM, Urry CM, Cappelluti N, Böhringer H, Comastri A, Glikman E, Richards G, Ananna T et al
(2016) The 31 deg2 release of the stripe 82 X-ray survey: the point source catalog. Astrophys J 817:172.
doi:10.3847/0004-637X/817/2/172
Lanzuisi G, Ponti G, Salvato M, Hasinger G, Cappelluti N, Bongiorno A, Brusa M, Lusso E et al (2014)
Active galactic nucleus X-ray variability in the XMM-COSMOS survey. Astrophys J 781:105. doi:10.
1088/0004-637X/781/2/105
Laor A, Davis SW (2014) Line-driven winds and the UV turnover in AGN accretion discs. MNRAS
438:3024–3038. doi:10.1093/mnras/stt2408
Laor A, Netzer H (1989) Massive thin accretion discs. I-Calculated spectra. MNRAS 238:897–916. doi:10.
1093/mnras/238.3.897
Lawrence A (1991) The relative frequency of broad-lined and narrow-lined active galactic nuclei—
implications for unified schemes. MNRAS 252:586–592. doi:10.1093/mnras/252.4.586
123
Lawrence A, Elvis M (2010) Misaligned disks as obscurers in active galaxies. Astrophys J 714:561–570.
doi:10.1088/0004-637X/714/1/561
Lawrence A, Watson MG, Pounds KA, Elvis M (1987) Low-frequency divergent X-ray variability in the
Seyfert galaxy NGC4051. Nature 325:694–696. doi:10.1038/325694a0
Ledlow MJ, Owen FN (1996) 20 CM VLA survey of Abell clusters of galaxies. VI. Radio/optical luminosity
functions. Astron J 112:9. doi:10.1086/117985
Leech KJ, Penston MV, Terlevich R et al (1989) High-luminosity IRAS galaxies. II—Optical spectroscopy,
modelling of starburst regions and comparison with structure. MNRAS 240:349–372. doi:10.1093/
mnras/240.2.349
Lehmer BD, Alexander DM, Bauer FE, Brandt WN, Goulding AD, Jenkins LP, Ptak A, Roberts TP et al
(2010) A Chandra perspective on galaxy-wide X-ray binary emission and its correlation with star
formation rate and stellar mass: new results from luminous infrared galaxies. Astrophys J 724:559–
571. doi:10.1088/0004-637X/724/1/559
Lehmer BD, Xue YQ, Brandt WN, Alexander DM, Bauer FE, Brusa M, Comastri A, Gilli R et al (2012) The
4 Ms Chandra deep field-south number counts apportioned by source class: pervasive active galactic
nuclei and the ascent of normal galaxies. Astrophys J 752:46. doi:10.1088/0004-637X/752/1/46
Lehto HJ (1989) A model for 1/f-type variability in active galactic nuclei. Two topics in X-ray astronomy,
volume 1: X ray binaries. Volume 2: AGN and the X ray. Background 296:499–503
Leighly KM (2004) Hubble space telescope STIS ultraviolet spectral evidence of outflow in extreme narrow-
line Seyfert 1 galaxies. II. Modeling and interpretation. Astrophys J 611:125–152. doi:10.1086/422089
Collaboration LHAASO (2016) Gamma ray astronomy with LHAASO. J Phys Conf Ser 718:052043.
doi:10.1088/1742-6596/718/5/052043
Lichti GG, Bottacini E, Ajello M, Charlot P, Collmar W, Falcone A, Horan D, Huber S et al (2008)
INTEGRAL observations of the blazar Mrk 421 in outburst. Results of a multi-wavelength campaign.
Longair MS (1966) On the interpretation of radio source counts. MNRAS 133:421–436. doi:10.1093/mnras/
133.4.421
López-Gonzaga N, Asmus D, Bauer FE, Tristram KRW, Burtscher L, Marinucci A, Matt G, Harrison FA
et al (2017) NGC 1068: No change in the mid-IR torus structure despite X-ray variability. Astron
Astrophys Astron Astrophys 602:A78. doi:10.1051/0004-6361/201629600
LSST Science Collaboration, Abell PA, Allison J, Anderson SF, Andrew JR, Angel JRP, Armus L, Arnett
D et al (2009) LSST science book, version 2.0. arXiv:0912.0201
Luo B, Brandt WN, Alexander DM, Stern D, Teng SH, Arévalo P, Bauer FE, Boggs SE et al (2014) Weak hard
X-ray emission from broad absorption line quasars: evidence for intrinsic X-ray weakness. Astrophys
J 794:70. doi:10.1088/0004-637X/794/1/70
Luo B, Brandt WN, Xue YQ, Lehmer B, Alexander DM, Bauer FE, Vito F, Yang G et al (2017) The Chandra
deep field-south survey: 7 Ms source catalogs. Astrophys J Suppl 228:2. doi:10.3847/1538-4365/228/
1/2
Lusso E, Risaliti G (2016) The tight relation between X-ray and ultraviolet luminosity of quasars. Astrophys
J 819:154. doi:10.3847/0004-637X/819/2/154
Lyu J, Rieke GH, Shi Y (2017) Dust-deficient Palomar-green quasars and the diversity of AGN intrinsic IR
emission. Astrophys J 835:257. doi:10.3847/1538-4357/835/2/257
Lyubarskii YE (1997) Flicker noise in accretion disks. MNRAS 292:679. doi:10.1093/mnras/292.3.679
MacLeod CL, Ivezić Ž, Kochanek CS, Kozłowski S, Kelly B, Bullock E, Kimball A, Sesar B et al (2010)
Modeling the time variability of SDSS stripe 82 quasars as a damped random walk. Astrophys J
721:1014–1033. doi:10.1088/0004-637X/721/2/1014
MacLeod CL, Ross NP, Lawrence A, Goad M, Horne K, Burgett W, Chambers KC, Flewelling H et al
(2016) A systematic search for changing-look quasars in SDSS. MNRAS 457:389–404. doi:10.1093/
mnras/stv2997
Magnelli B, Lutz D, Santini P et al (2012) A Herschel view of the far-infrared properties of submillimetre
galaxies. Astron Astrophys 539:A155. doi:10.1051/0004-6361/201118312
Magorrian J, Tremaine S, Richstone D et al (1998) The demography of massive dark objects in galaxy
centers. Astronom J 115:2285–2305. doi:10.1086/300353
Maini A, Prandoni I, Norris RP, Giovannini G, Spitler LR (2016) Compact radio cores in radio-quiet active
galactic nuclei. Astron Astrophys 589:L3. doi:10.1051/0004-6361/201628305
123
Malizia A, Bassani L, Bazzano A, Bird AJ, Masetti N, Panessa F, Stephen JB, Ubertini P et al (2012)
The INTEGRAL/IBIS AGN catalogue—I. X-ray absorption properties versus optical classification.
MNRAS 426:1750–1766. doi:10.1111/j.1365-2966.2012.21755.x
Malzac J (2014) The spectral energy distribution of compact jets powered by internal shocks. MNRAS
443:299–317. doi:10.1093/mnras/stu1144
Manners J, Almaini O, Lawrence A (2002) The X-ray variability of high-redshift QSOs. MNRAS 330:390–
398. doi:10.1046/j.1365-8711.2002.05065.x
Mannheim K (1993) The proton blazar. Astron Astrophys 269:67
Maoz D, Nagar NM, Falcke H, Wilson AS (2005) The murmur of the sleeping black hole: detection of
nuclear ultraviolet variability in LINER galaxies. Astrophys J 625:699–715. doi:10.1086/429795
Maraschi L, Ghisellini G, Celotti A (1992) A jet model for the gamma-ray emitting blazar 3C 279. Astrophys
J Lett 397:L5. doi:10.1086/186531
Marchesi S, Civano F, Elvis M, Salvato M, Brusa M, Comastri A, Gilli R, Hasinger G et al (2016) The
Chandra COSMOS legacy survey: optical/IR identifications. Astrophys J 817:34. doi:10.3847/0004-
637X/817/1/34
Marconi A, Hunt LK (2003) The relation between black hole mass, bulge mass, and near-infrared luminosity.
Astrophys J Lett 589:L21–L24. doi:10.1086/375804
Marscher AP, Gear WK (1985) Models for high-frequency radio outbursts in extragalactic sources, with
application to the early 1983 millimeter-to-infrared flare of 3C 273. Astrophys J 298:114–127. doi:10.
1086/163592
Marshall N, Warwick RS, Pounds KA (1981) The variability of X-ray emission from active galaxies.
MNRAS 194:987–1002. doi:10.1093/mnras/194.4.987
Mateos S, Alonso-Herrero A, Carrera FJ et al (2012) Using the bright ultrahard XMM-Newton survey to
define an IR selection of luminous AGN based on WISE colours. MNRAS 426:3271–3281. doi:10.
1111/j.1365-2966.2012.21843.x
Mateos S, Carrera FJ, Alonso-Herrero A et al (2016) X-ray absorption, nuclear infrared emission, and dust
covering factors of AGNs: testing unification schemes. Astrophys J 819:166. doi:10.3847/0004-637X/
819/2/166
Matt G, Guainazzi M, Maiolino R (2003) Changing look: from Compton-thick to Compton-thin, or the
rebirth of fossil active galactic nuclei. MNRAS 342:422–426. doi:10.1046/j.1365-8711.2003.06539.
x
Max-Moerbeck W, Hovatta T, Richards JL, King OG, Pearson TJ, Readhead ACS et al (2014) Time corre-
lation between the radio and gamma-ray activity in blazars and the production site of the gamma-ray
emission. MNRAS 445:428–436. doi:10.1093/mnras/stu1749
McAlpine S, Bower RG, Harrison CM et al (2017) The link between galaxy and black hole growth in the
eagle simulation. MNRAS 468:3395–3407. doi:10.1093/mnras/stx658
McGreer ID, Jiang L, Fan X, Richards GT, Strauss MA, Ross NP, White M, Shen Y et al (2013) The z = 5
quasar luminosity function from SDSS Stripe 82. Astrophys J 768:105. doi:10.1088/0004-637X/768/
2/105
McHardy IM, Papadakis IE, Uttley P, Page MJ, Mason KO (2004) Combined long and short time-scale
X-ray variability of NGC 4051 with RXTE and XMM-Newton. MNRAS 348:783–801. doi:10.1111/
j.1365-2966.2004.07376.x
McHardy IM, Koerding E, Knigge C, Uttley P, Fender RP (2006) Active galactic nuclei as scaled-up galactic
black holes. Nature 444:730–732. doi:10.1038/nature05389
McHardy IM, Arévalo P, Uttley P, Papadakis IE, Summons DP, Brinkmann W, Page MJ (2007) Discovery
of multiple Lorentzian components in the X-ray timing properties of the narrow line Seyfert 1 Ark
564. MNRAS 382:985–994. doi:10.1111/j.1365-2966.2007.12411.x
McHardy IM, Cameron DT, Dwelly T, Connolly S, Lira P, Emmanoulopoulos D, Gelbord J, Breedt E et al
(2014) Swift monitoring of NGC 5548: X-ray reprocessing and short-term UV/optical variability.
MNRAS 444:1469–1474. doi:10.1093/mnras/stu1636
Meier DL (2002) Grand unification of AGN and the accretion and spin paradigms. New Astron Rev 46:247–
255. doi:10.1016/S1387-6473(01)00189-0
Mejia-Restrepo JE, Trakhtenbrot B, Lira P, Netzer H, Capellupo DM (2016) Active galactic nuclei at
z ∼ 1.5—II. Black hole mass estimation by means of broad emission lines. MNRAS 460:187–211.
doi:10.1093/mnras/stw568
123
Melbourne J, Soifer BT, Desai V et al (2012) The spectral energy distributions and infrared luminosities of
z ∼ 2 dust-obscured galaxies from Herschel and Spitzer. Astron J 143:125. doi:10.1088/0004-6256/
143/5/125
Mendez AJ, Coil AL, Aird J et al (2013) PRIMUS: infrared and X-ray AGN selection techniques at
0.2<z<1.2. Astrophys J 770:40. doi:10.1088/0004-637X/770/1/40
Merloni A (2016) Observing supermassive black holes across cosmic time: from phenomenology to physics.
In: Lecture notes in physics, vol 905. Springer, Berlin, p 101. doi:10.1007/978-3-319-19416-5_4
Merloni A, Heinz S (2013) Evolution of active galactic nuclei. Planets, stars and stellar systems. Volume
6: Extragalactic astronomy and cosmology 6:503–566. doi:10.1007/978-94-007-5609-0_11
Merloni A, Predehl P, Becker W, Böhringer H, Boller T, Brunner H, Brusa M, Dennerl K et al (2012)
eROSITA science book: mapping the structure of the energetic universe. arXiv:1209.3114
Merloni A, Bongiorno A, Brusa M, Iwasawa K, Mainieri V, Magnelli B, Salvato M, Berta S et al (2014)
The incidence of obscuration in active galactic nuclei. MNRAS 437:3550–3567. doi:10.1093/mnras/
stt2149
Messias H, Afonso J, Salvato M, Mobasher B, Hopkins AM (2012) A new infrared color criterion for the
selection of 0<z<7 AGNs: application to deep fields and implications for JWST surveys. Astrophys
J 754:120. doi:10.1088/0004-637X/754/2/120
Miley G (1980) Annu Rev Astron Astrophys 18:165. doi:10.1146/annurev.aa.18.090180.001121
Miller L, Peacock JA, Mead ARG (1990) The bimodal radio luminosity function of quasars. MNRAS
244:207–213
Mineo S, Gilfanov M, Sunyaev R (2012a) X-ray emission from star-forming galaxies—I. High-mass X-ray
binaries. MNRAS 419:2095–2115. doi:10.1111/j.1365-2966.2011.19862.x
Mineo S, Gilfanov M, Sunyaev R (2012) X-ray emission from star-forming galaxies—II. Hot interstellar
medium. MNRAS 426:1870–1883. doi:10.1111/j.1365-2966.2012.21831.x
Mingo B, Hardcastle MJ, Croston JH, Dicken D, Evans DA, Morganti R, Tadhunter C (2014) An X-ray
survey of the 2 Jy sample—I. Is there an accretion mode dichotomy in radio-loud AGN? MNRAS
440:269–297. doi:10.1093/mnras/stu263
Miyaji T, Hasinger G, Salvato M, Brusa M, Cappelluti N, Civano F, Puccetti S, Elvis M et al (2015) Detailed
shape and evolutionary behavior of the X-ray luminosity function of active galactic nuclei. Astrophys
J 804:104. doi:10.1088/0004-637X/804/2/104
Miyamoto S, Kitamoto S, Mitsuda K, Dotani T (1988) Delayed hard X-rays from Cygnus X-1. Nature
336:450–452. doi:10.1038/336450a0
Moles M et al (2008) The Alhambra survey: a large area multimedium-band optical and near-infrared
photometric survey. Astronom J 136:1325–1339. doi:10.1088/0004-6256/136/3/1325
Mor R, Trakhtenbrot B (2011) Hot-dust clouds with pure-graphite composition around type-I active galactic
nuclei. Astrophys J Lett 737:L36. doi:10.1088/2041-8205/737/2/L36
Morganson E, Burgett WS, Chambers KC, Green PJ, Kaiser N, Magnier EA, Marshall PJ, Morgan JS et al
(2014) Measuring quasar variability with pan-STARRS1 and SDSS. Astrophys J 784:92. doi:10.1088/
0004-637X/784/2/92
Morić I, Smolčić V, Kimball A, Riechers DA, Ivezić Ž, Scoville N (2010) A closer view of the radio-
FIR correlation: disentangling the contributions of star formation and active galactic nucleus activity.
Astrophys J 724:779–790. doi:10.1088/0004-637X/724/1/779
Mortlock DJ, Warren SJ, Venemans BP, Patel M, Hewett PC, McMahon RG, Simpson C, Theuns T et al
(2011) A luminous quasar at a redshift of z = 7.085. Nature 474:616–619. doi:10.1038/nature10159
Müller-Sánchez F, Prieto MA, Mezcua M et al (2013) The central molecular gas structure in LINERs with
low-luminosity active galactic nuclei: evidence for gradual disappearance of the torus. Astrophys J
Lett 763:L1. doi:10.1088/2041-8205/763/1/L1
Murakami H, Baba H, Barthel P et al (2007) The infrared astronomical mission AKARI. Publ Astron Soc
Jpn 59:S369. doi:10.1093/pasj/59.sp2.S369
Murray SS, Kenter A, Forman WR et al (2005) XBootes: an X-ray survey of the NDWFS Bootes field. I.
Overview and initial results. Astrophys J Suppl 161:1–8. doi:10.1086/444378
Mushotzky RF, Done C, Pounds KA (1993) X-ray spectra and time variability of active galactic nuclei.
Annu Rev Astron Astrophys 31:717–761. doi:10.1146/annurev.aa.31.090193.003441
Mushotzky RF, Edelson R, Baumgartner W, Gandhi P (2011) Kepler observations of rapid optical variability
in active galactic nuclei. Astrophys J Lett 743:L12. doi:10.1088/2041-8205/743/1/L12
123
Nagar NM, Falcke H, Wilson AS (2005) Radio sources in low-luminosity active galactic nuclei. IV. Radio
luminosity function, importance of jet power, and radio properties of the complete Palomar sample.
Nakagawa T, Shibai H, Onaka T, Matsuhara H, Kaneda H, Kawakatsu Y (2015) The next-generation infrared
astronomy mission SPICA under the new framework. Publ Korean Astron Soc 30:621–624. doi:10.
5303/PKAS.2015.30.2.621
Namekata D, Umemura M (2016) Sub-parsec-scale dynamics of a dusty gas disc exposed to anisotropic
AGN radiation with frequency-dependent radiative transfer. MNRAS 460:980–1018. doi:10.1093/
mnras/stw862
Nandra K, George IM, Mushotzky RF, Turner TJ, Yaqoob T (1997) ASCA observations of Seyfert 1 galaxies.
I. Data analysis, imaging, and timing. Astrophys J 476:70–82
Nandra K, Barret D, Barcons X, Fabian A, den Herder J-W, Piro L, Watson M, Adami C et al (2013) The hot
and energetic universe: a white paper presenting the science theme motivating the Athena+ mission.
arXiv:1306.2307
Narayan R, Yi I (1994) Advection-dominated accretion: a self-similar solution. Astrophys J Lett 428:L13–
L16. doi:10.1086/187381
Narayan R, Mahadevan R, Grindlay JE, Popham RG, Gammie C (1998) Advection-dominated accretion
model of Sagittarius A*: evidence for a black hole at the Galactic center. Astrophys J 492:554–568.
doi:10.1086/305070
Narayan R, McClintock JE, Tchekhovskoy A (2014) Energy extraction from spinning black holes via
relativistic jets, p 523. doi:10.1007/978-3-319-06349-2_25
Nardini E, Risaliti G, Salvati M et al (2008) Spectral decomposition of starbursts and active galactic nuclei
in 5–8 μm Spitzer-IRS spectra of local ultraluminous infrared galaxies. MNRAS 385:L130–L134.
doi:10.1111/j.1745-3933.2008.00450.x
Nenkova M, Ivezić Ž, Elitzur M (2002) Dust emission from active galactic nuclei. Astrophys J Lett 570:L9–
L12. doi:10.1086/340857
Nenkova M, Sirocky MM, Nikutta R, Ivezić Ž, Elitzur M (2008) AGN dusty tori. II. Observational impli-
cations of clumpiness. Astrophys J 685:160–180. doi:10.1086/590483
Netzer H (2015) Revisiting the unified model of active galactic nuclei. Annu Rev Astron Astrophys 53:365–
408. doi:10.1146/annurev-astro-082214-122302
Netzer H, Trakhtenbrot B (2014) Bolometric luminosity black hole growth time and slim accretion discs in
active galactic nuclei. MNRAS 438:672–679. doi:10.1093/mnras/stt2238
Nikolajuk M, Papadakis IE, Czerny B (2004) Black hole mass estimation from X-ray variability measure-
ments in active galactic nuclei. MNRAS 350:L26–L30. doi:10.1111/j.1365-2966.2004.07829.x
Nikołajuk M, Czerny B, Ziółkowski J, Gierliński M (2006) Consistency of the black hole mass determination
in AGN from the reverberation and the X-ray excess variance method. MNRAS 370:1534–1540.
doi:10.1111/j.1365-2966.2006.10576.x
Nikutta R, Elitzur M, Lacy M (2009) On the 10 μm silicate feature in active galactic nuclei. Astrophys J
707:1550–1559. doi:10.1088/0004-637X/707/2/1550
Norris RP, Hopkins AM, Afonso J, Brown S, Condon JJ, Dunne L, Feain I, Hollow R et al (2011) EMU:
evolutionary map of the universe. Publ Astron Soc Aust 28:215–248. doi:10.1071/AS11021
Norris RP, Afonso J, Bacon D, Beck R, Bell M, Beswick RJ, Best P, Bhatnagar S et al (2013) Radio
continuum surveys with square kilometre array pathfinders. Publ Astron Soc Aust 30:e020. doi:10.
1017/pas.2012.020
Novak GS, Ostriker JP, Ciotti L (2011) Feedback from central black holes in elliptical galaxies: two-
dimensional models compared to one-dimensional models. Astrophys J 737:26. doi:10.1088/0004-
637X/737/1/26
Nowak MA, Wilms J, Vaughan BA, Dove JB, Begelman MC (1999) Rossi X-ray timing explorer observation
of Cygnus X-1. III. Implications for Compton corona and advection-dominated accretion flow models.
Astrophys J 515:726–737. doi:10.1086/307039
O’Dea CP (1998) The compact steep-spectrum and gigahertz peaked-spectrum radio sources. Publ Astron
Soc Pac 110:493–532. doi:10.1086/316162
O’Dea CP, Baum SA, Stanghellini C (1991) What are the gigahertz peaked-spectrum radio sources? Astro-
phys J 380:66–77. doi:10.1086/170562
O’Neill PM, Nandra K, Papadakis IE, Turner TJ (2005) The relationship between X-ray variability amplitude
and black hole mass in active galactic nuclei. MNRAS 358:1405–1416. doi:10.1111/j.1365-2966.
2005.08860.x
123
Orr MJL, Browne IWA (1982) Relativistic beaming and quasar statistics. MNRAS 200:1067–1080. doi:10.
1093/mnras/200.4.1067
Osterbrock DE (1981) Seyfert galaxies with weak broad H alpha emission lines. Astrophys J 249:462–470.
doi:10.1086/159306
Osterbrock DE, Pogge RW (1985) The spectra of narrow-line Seyfert 1 galaxies. Astrophys J 297:166–176.
doi:10.1086/163513
Owen FN, Burns JO, Rudnick L (1978) VLA observations of NGC 1265 at 4886 MHz. Astrophys J Lett
226:L119–L123. doi:10.1086/182845
Oyabu S, Ishihara D, Malkan M et al (2011) AKARI detections of hot dust in luminous infrared galax-
ies. Search for dusty active galactic nuclei. Astron Astrophys 529:A122. doi:10.1051/0004-6361/
201014221
Padovani P (2011) The microjansky and nanojansky radio sky: source population and multiwavelength
properties. MNRAS 411:1547–1561. doi:10.1111/j.1365-2966.2010.17789.x
Padovani P (2016) The faint radio sky: radio astronomy becomes mainstream. Astron Astrophys Rev 24:13.
doi:10.1007/s00159-016-0098-6
Padovani P (2017) On the two main classes of active galactic nuclei. Nat Astron 1:0194. doi:10.1038/
s41550-017-0194
Padovani P, Giommi P (1995) The connection between x-ray- and radio-selected BL Lacertae objects.
Astrophys J 444:567–581. doi:10.1086/175631
Padovani P, Giommi P (2015) A simplified view of blazars: the very high energy γ -ray vision. MNRAS
446:L41–L45. doi:10.1093/mnrasl/slu164
Padovani P, Miller N, Kellermann KI, Mainieri V, Rosati P, Tozzi P (2011) The VLA survey of Chandra
Deep Field South. V. Evolution and luminosity functions of sub-millijansky radio sources and the
issue of radio emission in radio-quiet active galactic nuclei. Astrophys J 740:20. doi:10.1088/0004-
637X/740/1/20
Padovani P, Bonzini M, Kellermann KI, Miller N, Mainieri V, Tozzi P (2015) Radio-faint AGN: a tale of
two populations. MNRAS 452:1263–1279. doi:10.1093/mnras/stv1375
Padovani P, Resconi E, Giommi P, Arsioli B, Chang YL (2016) Extreme blazars as counterparts of IceCube
astrophysical neutrinos. MNRAS 457:3582–3592. doi:10.1093/mnras/stw228
Page MJ, Simpson C, Mortlock DJ, Warren SJ, Hewett PC, Venemans BP, McMahon RG (2014) X-rays
from the redshift 7.1 quasar ULAS J1120+0641. MNRAS 440:L91–L95. doi:10.1093/mnrasl/slu022
Panessa F, Bassani L (2002) Unabsorbed Seyfert 2 galaxies. Astron Astrophys 394:435–442. doi:10.1051/
0004-6361:20021161
Paolillo M, Schreier EJ, Giacconi R, Koekemoer AM, Grogin NA (2004) Prevalence of X-ray variability
in the Chandra Deep Field-South. Astrophys J 611:93–106. doi:10.1086/421967
Paolillo M, Papadakis I, Brandt WN, Luo B, Xue YQ, Tozzi P, Shemmer O, Allevato V, et al (2017) Tracing
the accretion history of supermassive Black Holes through X-ray variability: results from the Chandra
Deep Field-South. MNRAS, in press. arXiv:1707.05332
Papadakis IE (2004) The scaling of the X-ray variability with black hole mass in active galactic nuclei.
MNRAS 348:207–213. doi:10.1111/j.1365-2966.2004.07351.x
Pâris I, Petitjean P, Ross NP, Myers AD, Aubourg É, Streblyanska A, Bailey S, Armengaud É et al (2017)
The sloan digital sky survey quasar catalog: twelfth data release. Astron Astrophys 597:A79. doi:10.
1051/0004-6361/201527999
Peacock JA, Gull SF (1981) Multifrequency models for the cosmological evolution of extragalactic radio
sources. MNRAS 196:611–633. doi:10.1093/mnras/196.3.611
Peacock JA, Miller L, Longair MS (1986) The statistics of radio emission from quasars. MNRAS 218:265–
278. doi:10.1093/mnras/218.2.265
Perley RA, Willis AG, Scott JS (1979) The structure of the radio jets in 3C449. Nature 281:437–442. doi:10.
1038/281437a0
Perlman ES, Mason RE, Packham C et al (2007) The mid-infrared emission of M87. Astrophys J 663:808–
815. doi:10.1086/518781
Peterson BM (1993) Reverberation mapping of active galactic nuclei. Publ Astron Soc Pac 105:247–268.
doi:10.1086/133140
Peterson BM, Ferrarese L, Gilbert KM, Kaspi S, Malkan MA, Maoz D, Merritt D, Netzer H et al (2004)
Central masses and broad-line region sizes of active galactic nuclei. II. A homogeneous analysis of a
large reverberation-mapping database. Astrophys J 613:682–699. doi:10.1086/423269
123
Piccinotti G, Mushotzky RF, Boldt EA, Holt SS, Marshall FE, Serlemitsos PJ, Shafer RA (1982) A complete
X-ray sample of the high-latitude/absolute value of B greater than 20 deg/sky from HEAO 1 A-2 −
Log N-log S and luminosity functions. Astrophys J 253:485–503. doi:10.1086/159651
Pier EA, Krolik JH (1992) Infrared spectra of obscuring dust tori around active galactic nuclei. I-
Calculational method and basic trends. Astrophys J 401:99–109. doi:10.1086/172042
Pittori C, Verrecchia F, Chen AW et al (2009) First AGILE catalog of high-confidence gamma-ray sources.
Astron Astrophys 506:1563–1574. doi:10.1051/0004-6361/200911783
Ponti G, Papadakis I, Bianchi S, Guainazzi M, Matt G, Uttley P, Bonilla NF (2012) CAIXA: a catalogue of
AGN in the XMM-Newton archive. III. Excess variance analysis. Astron Astrophys 542:A83. doi:10.
1051/0004-6361/201118326
Pracy MB, Ching JHY, Sadler EM, Croom SM, Baldry IK, Bland-Hawthorn J, Brough S, Brown MJI et al
(2016) GAMA/WiggleZ: the 1.4 GHz radio luminosity functions of high- and low-excitation radio
galaxies and their redshift evolution to z = 0.75. MNRAS 460:2–17. doi:10.1093/mnras/stw910
Rees MJ, Begelman MC, Blandford RD, Phinney ES (1982) Ion-supported tori and the origin of radio jets.
Nature 295:17–21. doi:10.1038/295017a0
Remillard RA, McClintock JE (2006) X-ray properties of black-hole binaries. Annu Rev Astron Astrophys
44:49–92. doi:10.1146/annurev.astro.44.051905.092532
Resconi E, Coenders S, Padovani P, Giommi P, Caccianiga L (2017) Connecting blazars with ultrahigh-
energy cosmic rays and astrophysical neutrinos. MNRAS 468:597–606. doi:10.1093/mnras/stx498
Reyes R, Zakamska NL, Strauss MA, Green J, Krolik JH, Shen Y, Richards GT, Anderson SF et al (2008)
Space density of optically selected type 2 quasars. Astron J 136:2373–2390. doi:10.1088/0004-6256/
136/6/2373
Ricci C, Bauer FE, Treister E, Romero-Cañizales C, Arevalo P, Iwasawa K, Privon GC, Sanders DB et al
(2016) NUSTAR unveils a heavily obscured low-luminosity active galactic nucleus in the luminous
infrared galaxy NGC 6286. Astrophys J 819:4. doi:10.3847/0004-637X/819/1/4
Ricci F, La Franca F, Onori F, Bianchi S (2017) Novel calibrations of virial black hole mass estimators in
active galaxies based on X-ray luminosity and optical/near-infrared emission lines. Astron Astrophys
598:A51. doi:10.1051/0004-6361/201629380
Richards GT, Fan X, Schneider DP, Vanden Berk DE, Strauss MA, York DG, Anderson JE Jr et al (2001)
Colors of 2625 quasars at 0<z<5 measured in the Sloan digital sky survey photometric system. Astron
J 121:2308–2330. doi:10.1086/320392
Richards GT, Fan X, Newberg HJ, Strauss MA, Vanden Berk DE, Schneider DP, Yanny B, Boucher A
et al (2002) Spectroscopic target selection in the Sloan digital sky survey: the quasar sample. Astron
J 123:2945–2975. doi:10.1086/340187
Richards GT, Lacy M, Storrie-Lombardi LJ, Hall PB, Gallagher SC, Hines DC, Fan X, Papovich C et al
(2006a) Spectral energy distributions and multiwavelength selection of type 1 quasars. Astrophys J
Suppl 166:470–497. doi:10.1086/506525
Richards GT, Strauss MA, Fan X, Hall PB, Jester S, Schneider DP, Vanden Berk DE, Stoughton C et al
(2006b) The Sloan digital sky survey quasar survey: quasar luminosity function from data release 3.
Astron J 131:2766–2787. doi:10.1086/503559
Richards GT, Deo RP, Lacy M, Myers AD, Nichol RC, Zakamska NL, Brunner RJ, Brandt WN et al (2009a)
Eight-dimensional mid-infrared/optical Bayesian quasar selection. Astron J 137:3884–3899. doi:10.
1088/0004-6256/137/4/3884
Richards GT, Myers AD, Gray AG, Riegel RN, Nichol RC, Brunner RJ, Szalay AS, Schneider DP et al
(2009b) Efficient photometric selection of quasars from the Sloan digital sky survey. II. 1,000,000
quasars from data release 6. Astrophys J Suppl 180:67–83. doi:10.1088/0067-0049/180/1/67
Richards GT, Kruczek NE, Gallagher SC, Hall PB, Hewett PC, Leighly KM, Deo RP, Kratzer RM et al
(2011) Unification of luminous type 1 quasars through C IV emission. Astron J 141:167. doi:10.1088/
0004-6256/141/5/167
Richards GT, Myers AD, Peters CM, Krawczyk CM, Chase G, Ross NP, Fan X, Jiang L et al (2015) Bayesian
high-redshift quasar classification from optical and mid-IR photometry. Astrophys J Suppl 219:39.
doi:10.1088/0067-0049/219/2/39
Rieke GH, Young ET, Engelbracht CW et al (2004) The multiband imaging photometer for Spitzer (MIPS).
Rigby EE, Argyle J, Best PN, Rosario D, Röttgering HJA (2015) Cosmic downsizing of powerful radio
galaxies to low radio luminosities. Astron Astrophys 581:A96. doi:10.1051/0004-6361/201526475
123
Risaliti G, Elvis M, Nicastro F (2002) Ubiquitous variability of X-ray-absorbing column densities in Seyfert
2 galaxies. Astrophys J 571:234–246. doi:10.1086/324146
Risaliti G, Elvis M, Fabbiano G, Baldi A, Zezas A (2005) Rapid Compton-thick/Compton-thin transitions
in the Seyfert 2 galaxy NGC 1365. Astrophys J Lett 623:L93–L96. doi:10.1086/430252
Risaliti G, Nardini E, Salvati M, Elvis M, Fabbiano G, Maiolino R, Pietrini P, Torricelli-Ciamponi G et al
(2011) X-ray absorption by broad-line region clouds in Mrk 766. MNRAS 410:1027–1035. doi:10.
1111/j.1365-2966.2010.17503.x
Roseboom IG, Lawrence A, Elvis M et al (2013) IR-derived covering factors for a large sample of quasars
from WISE-UKIDSS-SDSS. MNRAS 429:1494–1501. doi:10.1093/mnras/sts441
Ross NP, McGreer ID, White M, Richards GT, Myers AD, Palanque-Delabrouille N, Strauss MA, Anderson
SF et al (2013) The SDSS-III baryon oscillation spectroscopic survey: the quasar luminosity function
from data release nine. Astrophys J 773:14. doi:10.1088/0004-637X/773/1/14
Ross NP, Hamann F, Zakamska NL et al (2015) Extremely red quasars from SDSS, BOSS and WISE:
classification of optical spectra. MNRAS 453:3932–3952. doi:10.1093/mnras/stv1710
Ruan JJ, Anderson SF, Cales SL, Eracleous M, Green PJ, Morganson E, Runnoe JC, Shen Y et al (2016)
Toward an understanding of changing-look quasars: an archival spectroscopic search in SDSS. Astro-
phys J 826:188. doi:10.3847/0004-637X/826/2/188
Runnoe JC, Brotherton MS, Shang Z, DiPompeo MA (2013) Rehabilitating C IV-based black hole mass
estimates in quasars. MNRAS 434:848–861. doi:10.1093/mnras/stt1077
Sadler EM (2016) GPS/CSS radio sources and their relation to other AGN. Astron Nachr 337:105. doi:10.
1002/asna.201512274
Sadler EM, Cannon RD, Mauch T, Hancock PJ, Wake DA, Ross N, Croom SM, Drinkwater MJ et al
(2007) Radio galaxies in the 2SLAQ luminous red galaxy survey—I. The evolution of low-power
radio galaxies to z ∼ 0.7. MNRAS 381:211–227. doi:10.1111/j.1365-2966.2007.12231.x
Salvato M, Hasinger G, Ilbert O, Zamorani G, Brusa M, Scoville NZ, Rau A, Capak P et al (2009) Pho-
tometric redshift and classification for the XMM-COSMOS sources. Astrophys J 690:1250–1263.
doi:10.1088/0004-637X/690/2/1250
Sánchez P, Lira P, Cartier R (2016) Characterization of the AGN variability in the optical and near infrared
regimes. Zenodo. doi:10.5281/zenodo.60353
Sanders DB, Soifer BT, Elias JH et al (1988) Ultraluminous infrared galaxies and the origin of quasars.
Astrophys J 325:74–91. doi:10.1086/165983
Sarzi M, Shields JC, Schawinski K, Jeong H, Shapiro K, Bacon R, Bureau M, Cappellari M et al (2010) The
SAURON project—XVI. On the sources of ionization for the gas in elliptical and lenticular galaxies.
MNRAS 402:2187–2210. doi:10.1111/j.1365-2966.2009.16039.x
Satyapal S, Secrest NJ, McAlpine W et al (2014) Discovery of a population of bulgeless galaxies with
extremely red mid-IR colors: obscured AGN activity in the low-mass regime? Astrophys J 784:113.
doi:10.1088/0004-637X/784/2/113
Satyapal S, Secrest NJ, Rothberg B et al (2016) Obscured AGNs in bulgeless hosts discovered by WISE:
the case study of SDSS J1224+5555. Astrophys J 827:58. doi:10.3847/0004-637X/827/1/58
Scaringi S, Maccarone TJ, Kording E, Knigge C, Vaughan S, Marsh TR, Aranzana E, Dhillon VS et al
(2015) Accretion-induced variability links young stellar objects, white dwarfs, and black holes. Sci
Adv 1:e1500686–e1500686. doi:10.1126/sciadv.1500686
Schawinski K, Koss M, Berney S, Sartori LF (2015) Active galactic nuclei flicker: an observational estimate
of the duration of black hole growth phases of ≈105 yr. MNRAS 451:2517–2523. doi:10.1093/mnras/
stv1136
Schaye J, Crain RA, Bower RG et al (2015) The EAGLE project: simulating the evolution and assembly of
galaxies and their environments. MNRAS 446:521–554. doi:10.1093/mnras/stu2058
Schmidt M (1963) 3C 273: a star-like object with large red-shift. Nature 197:1040. doi:10.1038/1971040a0
Schmidt M (1970) Space distribution and luminosity functions of quasars. Astrophys J 162:371. doi:10.
1086/150668
Schneider DP, Richards GT, Hall PB, Strauss MA, Anderson SF, Boroson TA, Ross NP, Shen Y et al (2010)
The sloan digital sky survey quasar catalog. V. Seventh data release. Astron J 139:2360. doi:10.1088/
0004-6256/139/6/2360
Schulze A, Bongiorno A, Gavignaud I, Schramm M, Silverman J, Merloni A, Zamorani G, Hirschmann M
et al (2015) The cosmic growth of the active black hole population at 1<z<2 in zCOSMOS, VVDS
and SDSS. MNRAS 447:2085–2111. doi:10.1093/mnras/stu2549
123
Schweitzer M, Groves B, Netzer H et al (2008) Extended silicate dust emission in Palomar-green QSOs.
Astrophys J 679:101–117. doi:10.1086/587097
Sergeev SG, Doroshenko VT, Golubinskiy YV, Merkulova NI, Sergeeva EA (2005) Lag-luminosity relation-
ship for interband lags between variations in B, V, R, and I bands in active galactic nuclei. Astrophys
J 622:129–135. doi:10.1086/427820
Seyfert CK (1943) Nuclear emission in spiral nebulae. Astrophys J 97:28. doi:10.1086/144488
Shakura NI, Sunyaev RA (1973) Black holes in binary systems. Observational appearance. Astron Astrophys
24:337–355
Shanks T, Metcalfe N, Chehade B, Findlay JR, Irwin MJ, Gonzalez-Solares E, Lewis JR, Yoldas AK et al
(2015) The VLT survey telescope ATLAS. MNRAS 451:4238–4252. doi:10.1093/mnras/stv1130
Sharma M, Nayak J, Koul MK et al (2015) Validation of a new background discrimination method for the
TACTIC TeV gamma-ray telescope with Markarian 421 data. Nucl Inst Methods Phys Res A 770:42.
doi:10.1016/j.nima.2014.10.012
Shen Y (2013) The mass of quasars. Bull Astron Soc India 41:61–115
Shen Y, Ho LC (2014) The diversity of quasars unified by accretion and orientation. Nature 513:210–213.
doi:10.1038/nature13712
Shen Y, Kelly BC (2012) The demographics of broad-line quasars in the mass-luminosity plane. I. Testing
FWHM-based virial black hole masses. Astrophys J 746:169. doi:10.1088/0004-637X/746/2/169
Shi Y, Rieke GH, Hines DC, Gorjian V, Werner MW, Cleary K, Low FJ, Smith PS et al (2006) 9.7 μm
Silicate features in active galactic nuclei: new insights into unification models. Astrophys J 653:127–
136. doi:10.1086/508737
Siebenmorgen R, Haas M, Krügel E, Schulz B (2005) Discovery of 10 μm silicate emission in quasars.
Evidence of the AGN unification scheme. Astron Astrophys 436:L5–L8. doi:10.1051/0004-6361:
200500109
Simm T, Salvato M, Saglia R, Ponti G, Lanzuisi G, Trakhtenbrot B, Nandra K, Bender R et al (2016)
Pan-STARRS1 variability of XMM-COSMOS AGN. II. Physical correlations and power spectrum
analysis. Astron Astrophys 585:A129. doi:10.1051/0004-6361/201527353
Simmons BD, Van Duyne J, Urry CM et al (2011) Obscured GOODS active galactic nuclei and their host
galaxies at z<1.25: the slow black hole growth phase. Astrophys J 734:121. doi:10.1088/0004-637X/
734/2/121
Simpson C (2005) The luminosity dependence of the type 1 active galactic nucleus fraction. MNRAS
360:565–572. doi:10.1111/j.1365-2966.2005.09043.x
Slone O, Netzer H (2012) The effects of disc winds on the spectrum and black hole growth rate of active
galactic nuclei. MNRAS 426:656–664. doi:10.1111/j.1365-2966.2012.21699.x
Smith KL, Boyd PT, Mushotzky RF, Gehrels N, Edelson R, Howell SB, Gelino DM, Brown A, Young S
(2015) KSwAGS: a swift X-ray and UV survey of the Kepler field. I. Astron J 150:126. doi:10.1088/
0004-6256/150/4/126
Smolčić V (2009) The radio AGN population dichotomy: green valley Seyferts versus red sequence low-
excitation active galactic nuclei. Astrophys J Lett 699:L43–L47. doi:10.1088/0004-637X/699/1/L43
Smolčić V (2016) Radio continuum surveys and galaxy evolution: the AGN view. arXiv:1603.05687
Smolčić V, Ivezić Ž, Gaćeša M, Rakos K, Pavlovski K, Ilijić S, Obrić M, Lupton RH et al (2006) The rest-
frame optical colours of 99000 Sloan Digital Sky Survey galaxies. MNRAS 371:121–137. doi:10.
1111/j.1365-2966.2006.10662.x
Smolčić V, Schinnerer E, Scodeggio M, Franzetti P, Aussel H, Bondi M, Brusa M, Carilli CL et al (2008) A
new method to separate star-forming from AGN galaxies at intermediate redshift: the submillijansky
radio population in the VLA-COSMOS survey. Astrophys J Suppl 177:14–38. doi:10.1086/588028
Smolčić V, Zamorani G, Schinnerer E, Bardelli S, Bondi M, Bîrzan L, Carilli CL, Ciliegi P et al (2009)
Cosmic evolution of radio selected active galactic nuclei in the cosmos field. Astrophys J 696:24–39.
doi:10.1088/0004-637X/696/1/24
Smolčić V, Delvecchio I, Zamorani G, Baran N, Novak M, Delhaize J, Schinnerer E, Berta S et al (2017a)
The VLA-COSMOS 3 GHz large project: multiwavelength counterparts and the composition of the
faint radio population. Astron Astrophys 602:A2. doi:10.1051/0004-6361/201630223
Smolčić V, Novak M, Delvecchio I, Ceraj L, Bondi M, Delhaize J, Marchesi S, Murphy E et al (2017b) The
VLA-COSMOS 3 GHz large project: cosmic evolution of radio AGN and implications for radio-mode
feedback since z 5. Astron Astrophys 602:A6. doi:10.1051/0004-6361/201730685
123
Sobolewska MA, Siemiginowska A, Życki PT (2004) High-redshift radio-quiet quasars: exploring the
parameter space of accretion models. II. Patchy corona model. Astrophys J 617:102–112. doi:10.
1086/425262
Soffitta P, Bellazzini R, Bozzo E, Burwitz V, Castro-Tirado A, Costa E, Courvoisier T, Feng H et al (2016)
XIPE: the x-ray imaging polarimetry explorer. Proc SPIE 9905:990515. doi:10.1117/12.2233046
Soldi S, Beckmann V, Baumgartner WH, Ponti G, Shrader CR, Lubiński P, Krimm HA, Mattana F et al
(2014) Long-term variability of AGN at hard X-rays. Astron Astrophys 563:A57. doi:10.1051/0004-
6361/201322653
Spergel D, Gehrels N, Baltay C et al (2015) Wide-field infrared survey telescope-astrophysics focused
telescope assets WFIRST-AFTA 2015 report. arXiv e-prints
Spoon HWW, Marshall JA, Houck JR et al (2007) Mid-infrared galaxy classification based on silicate
obscuration and PAH equivalent width. Astrophys J Lett 654:L49–L52. doi:10.1086/511268
Stanek R, Evrard AE, Böhringer H, Schuecker P, Nord B (2006) The X-ray luminosity-mass relation for
local clusters of galaxies. Astrophys J 648:956–968. doi:10.1086/506248
Steffen AT, Strateva I, Brandt WN, Alexander DM, Koekemoer AM, Lehmer BD, Schneider DP, Vignali C
et al (2006) The X-ray-to-optical properties of optically selected active galaxies over wide luminosity
and redshift ranges. Astron J 131:2826–2842. doi:10.1086/503627
Stern D, Eisenhardt P, Gorjian V, Kochanek CS, Caldwell N, Eisenstein D, Brodwin M, Brown MJI et al
(2005) Mid-infrared selection of active galaxies. Astrophys J 631:163–168. doi:10.1086/432523
Stern D, Kirkpatrick JD, Allen LE et al (2007) Mid-infrared selection of brown dwarfs and high-redshift
quasars. Astrophys J 663:677–685. doi:10.1086/516833
Stern D, Assef RJ, Benford DJ et al (2012) Mid-infrared selection of active galactic nuclei with the wide-
field infrared survey explorer. I. Characterizing WISE-selected active galactic nuclei in COSMOS.
Astrophys J 753:30. doi:10.1088/0004-637X/753/1/30
Strateva I, Ivezić Ž, Knapp GR, Narayanan VK, Strauss MA, Gunn JE, Lupton RH, Schlegel D et al (2001)
Color separation of galaxy types in the Sloan Digital Sky Survey imaging data. Astron J 122:1861–
1874. doi:10.1086/323301
Sturm E, Schweitzer M, Lutz D, Contursi A, Genzel R, Lehnert MD, Tacconi LJ, Veilleux S et al (2005)
Silicate emissions in active galaxies: from LINERs to QSOs. Astrophys J Lett 629:L21–L23. doi:10.
1086/444359
Sulentic JW, Marziani P, Dultzin-Hacyan D (2000) Phenomenology of broad emission lines in active galactic
nuclei. Annu Rev Astron Astrophys 38:521–571. doi:10.1146/annurev.astro.38.1.521
Sulentic JW, Marziani P, Zamanov R, Bachev R, Calvani M, Dultzin-Hacyan D (2002) Average quasar
spectra in the context of eigenvector 1. Astrophys J Lett 566:L71–L75. doi:10.1086/339594
Sulentic JW, Bachev R, Marziani P, Negrete CA, Dultzin D (2007) C IV λ1549 as an eigenvector 1 parameter
for active galactic nuclei. Astrophys J 666:757–777. doi:10.1086/519916
Sun W-H, Malkan MA (1989) Fitting improved accretion disk models to the multiwavelength continua of
quasars and active galactic nuclei. Astrophys J 346:68–100. doi:10.1086/167986
Stalevski M, Fritz J, Baes M, Nakos T, Popović LČ (2012) 3D radiative transfer modelling of the dusty tori
around active galactic nuclei as a clumpy two-phase medium. MNRAS 420:2756–2772. doi:10.1111/
j.1365-2966.2011.19775.x
Stalevski M, Ricci C, Ueda Y, Lira P, Fritz J, Baes M (2016) The dust covering factor in active galactic
nuclei. MNRAS 458:2288–2302. doi:10.1093/mnras/stw444
Szokoly GP, Bergeron J, Hasinger G, Lehmann I, Kewley L, Mainieri V, Nonino M, Rosati P et al (2004)
The Chandra deep field-south: optical spectroscopy. I. Astrophys J Suppl 155:271–349. doi:10.1086/
424707
Tadhunter C (2016) Radio AGN in the local universe: unification, triggering and evolution. Astron Astrophys
Rev 24:10. doi:10.1007/s00159-016-0094-x
Tadhunter CN, Morganti R, Robinson A, Dickson R, Villar-Martin M, Fosbury RAE (1998) The nature
of the optical-radio correlations for powerful radio galaxies. MNRAS 298:1035–1047. doi:10.1046/j.
1365-8711.1998.01706.x
Tavani M et al (2009) The AGILE mission. Astron Astrophys 502:995–1013. doi:10.1051/0004-6361/
200810527
Tchekhovskoy A (2015) Launching of active galactic nuclei jets. In: Contopoulos I, Gabuzda D, Kylafis
N (eds) The formation and disruption of black hole jets, vol 414. Astrophysics and Space Science
Library, p 45. doi:10.1007/978-3-319-10356-3_3
123
Tchekhovskoy A, Narayan R, McKinney JC (2011) Efficient generation of jets from magnetically arrested
accretion on a rapidly spinning black hole. MNRAS 418:L79–L83. doi:10.1111/j.1745-3933.2011.
01147.x
Terrell J (1967) Luminosity fluctuations and a local model of quasi-stellar objects. Astrophys J 147:827.
doi:10.1086/149065
The Fermi-LAT Collaboration (2017) 3FHL: the third catalog of hard Fermi-LAT sources. arXiv:1702.00664
Thompson DJ, Bertsch DL, Fichtel CE et al (1993) Calibration of the energetic gamma-ray experiment
telescope (EGRET) for the Compton gamma-ray observatory. Astrophys J Suppl 86:629. doi:10.
1086/191793
Trakhtenbrot B, Netzer H (2012) Black hole growth to z = 2 − I . Improved virial methods for measuring
MBH and L/LEdd . MNRAS 427:3081–3102. doi:10.1111/j.1365-2966.2012.22056.x
Treister E, Urry CM, Chatzichristou E et al (2004) Obscured active galactic nuclei and the X-ray, optical, and
far-infrared number counts of active galactic nuclei in the GOODS fields. Astrophys J 616:123–135.
doi:10.1086/424891
Tremaine S, Gebhardt K, Bender R, Bower G, Dressler A, Faber SM, Filippenko AV, Green R et al (2002)
The slope of the black hole mass versus velocity dispersion correlation. Astrophys J 574:740–753.
doi:10.1086/341002
Treves A, Maraschi L, Abramowicz M (1988) Basic elements of the theory of accretion. Publ Astron Soc
Pac 100:427–451. doi:10.1086/132189
Tristram KRW, Meisenheimer K, Jaffe W et al (2007) Resolving the complex structure of the dust torus in
the active nucleus of the Circinus galaxy. Astron Astrophys 474:837–850. doi:10.1051/0004-6361:
20078369
Tsai CW, Eisenhardt PRM, Wu J et al (2015) The most luminous galaxies discovered by WISE. Astrophys
J 805:90. doi:10.1088/0004-637X/805/2/90
Ueda Y, Akiyama M, Ohta K, Miyaji T (2003) Cosmological evolution of the hard X-ray active galactic
nucleus luminosity function and the origin of the hard X-ray background. Astrophys J 598:886–908.
doi:10.1086/378940
Ueda Y, Akiyama M, Hasinger G, Miyaji T, Watson MG (2014) Toward the standard population synthesis
model of the X-ray background: evolution of X-ray luminosity and absorption functions of active
galactic nuclei including Compton-thick populations. Astrophys J 786:104. doi:10.1088/0004-637X/
786/2/104
Ulrich M-H, Maraschi L, Urry CM (1997) Variability of active galactic nuclei. Annu Rev Astron Astrophys
35:445–502. doi:10.1146/annurev.astro.35.1.445
Urry CM, Padovani P (1995) Unified schemes for radio-loud active galactic nuclei. Publ Astron Soc Pac
107:803–845. doi:10.1086/133630
Uttley P, McHardy IM (2001) The flux-dependent amplitude of broadband noise variability in X-ray binaries
and active galaxies. MNRAS 323:L26–L30. doi:10.1046/j.1365-8711.2001.04496.x
Uttley P, McHardy IM, Papadakis IE (2002) Measuring the broad-band power spectra of active galactic
nuclei with RXTE. MNRAS 332:231–250. doi:10.1046/j.1365-8711.2002.05298.x
Uttley P, Edelson R, McHardy IM, Peterson BM, Markowitz A (2003) Correlated long-term optical and
X-ray variations in NGC 5548. Astrophys J Lett 584:L53–L56. doi:10.1086/373887
Uttley P, McHardy IM, Vaughan S (2005) Non-linear X-ray variability in X-ray binaries and active galaxies.
MNRAS 359:345–362. doi:10.1111/j.1365-2966.2005.08886.x
Uttley P, Wilkinson T, Cassatella P, Wilms J, Pottschmidt K, Hanke M, Böck M (2011) The causal connection
between disc and power-law variability in hard state black hole X-ray binaries. MNRAS 414:L60–L64.
doi:10.1111/j.1745-3933.2011.01056.x
Vagnetti F, Middei R, Antonucci M, Paolillo M, Serafinelli R (2016) Ensemble X-ray variability of active
galactic nuclei. II. Excess variance and updated structure function. Astron Astrophys 593:A55. doi:10.
1051/0004-6361/201629057
Vanden Berk DE, Wilhite BC, Kron RG, Anderson SF, Brunner RJ, Hall PB, Ivezić Ž, Richards GT et al
(2004) The ensemble photometric variability of ∼25,000 quasars in the Sloan digital sky survey.
Astrophys J 601:692–714. doi:10.1086/380563
Vasudevan RV, Fabian AC (2009) Simultaneous X-ray/optical/UV snapshots of active galactic nuclei from
XMM-Newton: spectral energy distributions for the reverberation mapped sample. MNRAS 392:1124–
1140. doi:10.1111/j.1365-2966.2008.14108.x
Vaughan S, Edelson R, Warwick RS, Uttley P (2003) On characterizing the variability properties of X-ray
light curves from active galaxies. MNRAS 345:1271–1284. doi:10.1046/j.1365-2966.2003.07042.x
123
Vazquez B, Galianni P, Richmond M, Robinson A, Axon DJ, Horne K, Almeyda T, Fausnaugh M et al

(2015) Spitzer space telescope measurements of dust reverberation lags in the Seyfert 1 galaxy NGC
6418. Astrophys J 801:127. doi:10.1088/0004-637X/801/2/127
Veilleux S, Osterbrock DE (1987) Spectral classification of emission-line galaxies. Astrophys J Suppl
63:295–310. doi:10.1086/191166
Veilleux S, Rupke DSN, Kim DC et al (2009) Spitzer quasar and Ulirg evolution study (QUEST). IV.
Comparison of 1 Jy ultraluminous infrared galaxies with Palomar-green quasars. Astrophys J Suppl
182:628–666. doi:10.1088/0067-0049/182/2/628
Vestergaard M, Peterson BM (2006) Determining central black hole masses in distant active galaxies and
quasars. II. Improved optical and UV scaling relationships. Astrophys J 641:689–709. doi:10.1086/
500572
Vestergaard M, Fan X, Tremonti CA, Osmer PS, Richards GT (2008) Mass functions of the active black
holes in distant quasars from the Sloan digital sky survey data release 3. Astrophys J Lett 674:L1.
doi:10.1086/528981
Voges W, Aschenbach B, Boller T, Bräuninger H, Briel U, Burkert W, Dennerl K, Englhauser J et al (1999)
The ROSAT all-sky survey bright source catalogue. Astron Astrophys 349:389–405
Wada K, Schartmann M, Meijerink R (2016) Multi-phase nature of a radiation-driven fountain with nuclear
starburst in a low-mass active galactic nucleus. Astrophys J Lett 828:L19. doi:10.3847/2041-8205/
828/2/L19
Waddington I, Dunlop JS, Peacock JA, Windhorst RA (2001) The LBDS Hercules sample of mJy radio
sources at 1.4 GHz—II. Redshift distribution, radio luminosity function, and the high-redshift cut-off.
MNRAS 328:882–896. doi:10.1046/j.1365-8711.2001.04934.x
Wall JV (1975) The quasars from the Parkes 2700-MHz survey. Observatory 95:196–204
Wang JX, Jiang P (2006) On the fraction of X-ray-obscured quasars in the local universe. Astrophys J Lett
646:L103–L106. doi:10.1086/507100
Wang SX, Brandt WN, Luo B, Smail I, Alexander DM, Danielson ALR, Hodge JA, Karim A et al (2013) An
ALMA survey of submillimeter galaxies in the extended Chandra deep field-south: the AGN fraction
and X-ray properties of submillimeter galaxies. Astrophys J 778:179. doi:10.1088/0004-637X/778/
2/179
Warren SJ, Hewett PC, Foltz CB (2000) The KX method for producing K-band flux-limited samples of
quasars. MNRAS 312:827–832. doi:10.1046/j.1365-8711.2000.03206.x
Waters T, Proga D (2016) On the efficient acceleration of clouds in active galactic nuclei. MNRAS 460:L79–
L83. doi:10.1093/mnrasl/slw056
Weinberger R, Springel V, Hernquist L et al (2017) Simulating galaxy formation with black hole driven
thermal and kinetic feedback. MNRAS 465:3291–3308. doi:10.1093/mnras/stw2944
Weisskopf MC, Brinkman B, Canizares C, Garmire G, Murray S, Van Speybroeck LP (2002) An overview
of the performance and scientific results from the Chandra X-ray observatory. Publ Astron Soc Pac
114:1–24. doi:10.1086/338108
Werner MW, Roellig TL, Low FJ et al (2004) The Spitzer space telescope mission. Astrophys J Suppl
154:1–9. doi:10.1086/422992
Weymann RJ, Carswell RF, Smith MG (1981) Absorption lines in the spectra of quasistellar objects. Annu
Rev Astron Astrophys 19:41–76. doi:10.1146/annurev.aa.19.090181.000353
Weymann RJ, Morris SL, Foltz CB, Hewett PC (1991) Comparisons of the emission-line and continuum
properties of broad absorption line and normal quasi-stellar objects. Astrophys J 373:23–53. doi:10.
1086/170020
White SV, Jarvis MJ, Häußler B, Maddox N (2015) Radio-quiet quasars in the VIDEO survey: evidence
for AGN-powered radio emission at S1.4GHz <1 mJy. MNRAS 448:2665–2686. doi:10.1093/mnras/
stv134
Whysong D, Antonucci R (2004) Thermal emission as a test for hidden nuclei in nearby radio galaxies.
Astrophys J 602:116–122. doi:10.1086/380828
Willott CJ, Rawlings S, Blundell KM, Lacy M, Eales SA (2001) The radio luminosity function from the
low-frequency 3CRR, 6CE and 7CRS complete samples. MNRAS 322:536–552. doi:10.1046/j.1365-
8711.2001.04101.x
Wilman RJ, Miller L, Jarvis MJ, Mauch T, Levrier F, Abdalla FB, Rawlings S, Klöckner H-R et al (2008)
A semi-empirical simulation of the extragalactic radio continuum sky for next generation radio tele-
scopes. MNRAS 388:1335–1348. doi:10.1111/j.1365-2966.2008.13486.x
123
Wilms J, Allen A, McCray R (2000) On the absorption of X-rays in the interstellar medium. Astrophys J
542:914–924. doi:10.1086/317016
Wolf C, Wisotzki L, Borch A, Dye S, Kleinheinrich M, Meisenheimer K (2003) The evolution of faint AGN
between z = 1 and z = 5 from the COMBO-17 survey. Astron Astrophys 408:499–514. doi:10.1051/
0004-6361:20030990
Worseck G, Prochaska JX (2011) GALEX far-ultraviolet color selection of UV-bright high-redshift quasars.
Astrophys J 728:23. doi:10.1088/0004-637X/728/1/23
Wright EL, Eisenhardt PRM, Mainzer AK et al (2010) The wide-field infrared survey explorer (WISE):
mission description and initial on-orbit performance. Astron J 140:1868–1881. doi:10.1088/0004-
6256/140/6/1868
Wu J, Brandt WN, Hall PB, Gibson RR, Richards GT, Schneider DP, Shemmer O, Just DW et al (2011)
A population of X-ray weak quasars: PHL 1811 analogs at high redshift. Astrophys J 736:28. doi:10.
1088/0004-637X/736/1/28
Wu J, Tsai CW, Sayers J et al (2012a) Submillimeter follow-up of WISE-selected hyperluminous galaxies.
Astrophys J 756:96. doi:10.1088/0004-637X/756/1/96
Wu XB, Hao G, Jia Z, Zhang Y, Peng N (2012b) SDSS quasars in the WISE preliminary data release and
quasar candidate selection with optical/infrared colors. Astron J 144:49. doi:10.1088/0004-6256/144/
2/49
Xue YQ, Brandt WN, Luo B, Rafferty DA, Alexander DM, Bauer FE, Lehmer BD, Schneider DP et al
(2010) Color-magnitude relations of active and non-active galaxies in the Chandra deep fields: high-
redshift constraints and stellar-mass selection effects. Astrophys J 720:368–391. doi:10.1088/0004-
637X/720/1/368
Yang G, Brandt WN, Luo B, Xue YQ, Bauer FE, Sun MY, Kim S, Schulze S et al (2016) Long-term X-ray
variability of typical active galactic nuclei in the distant universe. Astrophys J 831:145. doi:10.3847/
0004-637X/831/2/145
Yip CW, Connolly AJ, Vanden Berk DE, Ma Z, Frieman JA, SubbaRao M, Szalay AS, Richards GT et al
(2004) Spectral classification of quasars in the Sloan digital sky survey: eigenspectra, redshift, and
luminosity effects. Astron J 128:2603–2630. doi:10.1086/425626
York DG, Adelman J, Anderson JE Jr, Anderson SF, Annis J, Bahcall NA, Bakken JA, Barkhouser R
et al (2000) The Sloan digital sky survey: technical summary. Astron J 120:1579–1587. doi:10.1086/
301513
Young M, Brandt WN, Xue YQ, Paolillo M, Alexander DM, Bauer FE, Lehmer BD, Luo B et al (2012)
Variability-selected low-luminosity active galactic nuclei in the 4 Ms Chandra deep field-south. Astro-
phys J 748:124. doi:10.1088/0004-637X/748/2/124
Yuan F, Narayan R (2014) Hot accretion flows around black holes. Annu Rev Astron Astrophys 52:529–588.
doi:10.1146/annurev-astro-082812-141003
Yuan S, Strauss MA, Zakamska NL (2016) Spectroscopic identification of type 2 quasars at z<1 in SDSS-
III/BOSS. MNRAS 462:1603–1615. doi:10.1093/mnras/stw1747
Yun MS, Aretxaga I, Ashby MLN et al (2008) Spitzer IRAC infrared colours of submillimetre-bright
Zakamska NL, Strauss MA, Krolik JH, Collinge MJ, Hall PB, Hao L, Heckman TM, Ivezić Ž et al (2003)
Candidate type II quasars from the Sloan digital sky survey. I. Selection and optical properties of a
sample at 0.3<z<0.83. Astron J 126:2125–2144. doi:10.1086/378610
Zakamska NL, Lampayan K, Petric A, Dicken D, Greene JE, Heckman TM, Hickox RC, Ho LC et al
(2016) Star formation in quasar hosts and the origin of radio emission in radio-quiet quasars. MNRAS
455:4191–4211. doi:10.1093/mnras/stv2571
Zhou X-L, Zhang S-N, Wang D-X, Zhu L (2010) Calibrating the correlation between black hole mass
and X-ray variability amplitude: X-ray only black hole mass estimates for active galactic nuclei and
ultra-luminous X-ray sources. Astrophys J 710:16–23. doi:10.1088/0004-637X/710/1/16
Zoghbi A, Fabian AC, Reynolds CS, Cackett EM (2012) Relativistic iron K X-ray reverberation in NGC
4151. MNRAS 422:129–134. doi:10.1111/j.1365-2966.2012.20587.x
Zuo W, Wu X-B, Liu Y-Q, Jiao C-L (2012) The correlations between optical variability and physical
parameters of quasars in SDSS Stripe 82. Astrophys J 758:104. doi:10.1088/0004-637X/758/2/104
123

s00159-017-0102-9

Uploaded by

Copyright:

Available Formats

s00159-017-0102-9

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

s00159-017-0102-9

Uploaded by

Copyright:

Available Formats

Astron Astrophys Rev (2017) 25:2

Active galactic nuclei: what’s in a name?

P. Padovani1 · D. M. Alexander2 · R. J. Assef3 · B. De Marco4,5 ·

Received: 12 June 2017 / Published online: 23 August 2017

Keywords Galaxies: active · Quasars: supermassive black holes · Radio continuum:

1 The active galactic nuclei zoo

1 European Southern Observatory, Karl-Schwarzschild-Str. 2, 85748 Garching bei München,

Table 1 The AGN zoo: list of AGN classes

2.1 Physical mechanism(s) and types of AGN selected in the radio

2.1.1 Radio galaxies

Fig. 2 Examples of FR I (panel

2.1.2 Flat- and steep-spectrum sources

2.1.3 RL and RQ AGN

2.1.4 Low- and high-excitation AGN

2.1.5 Radio-selected AGN classes and unification

6 The switch in L/L

2.1.6 Classification of radio-selected AGN

2.2 Selection effects: SF contribution to the total radio luminosity output

2.3 Cosmic evolution of radio-selected AGN

2.4 The future of AGN studies in the radio band

3.1 Physical mechanism behind IR emission

3.2 AGN in the MIR

3.2.1 Broad-band MIR AGN identification

Table 2 AGN MIR selection criteria shown in Fig. 5

WISE only criteria

NIR + Spitzer criteria

3.2.2 Comparison with general AGN identification at other wavelengths

MIR AGN identification is considerably less sensitive to obscuration of the central

3.2.4 Dust-free AGN and hot dust poor quasars

3.2.5 Eddington ratios and BH Masses

3.3 Red, reddened and high-redshift AGN

3.4 MIR spectroscopy

and classification of AGN. The unambiguous observations of the silicate feature at

3.5 The future of AGN studies in the IR band

The upcoming generation of ground-based giant telescopes will significantly expand

4.1 Physical mechanisms behind optical emission

4.2 Photometric and spectroscopic identification

Fig. 6 Number of quasars as a

4.3 Selection effects

redshifts—without considerable stellar contamination. The redshifts affected span a

4.4 Information content

4.5 L/L Edd and MBH

We discuss in this section X-ray-selected AGN. The physical mechanism behind X-

5.1 Physical mechanism behind X-ray emission

5.2 Selection of AGN in the X-ray band: identification challenges

consideration of the thresholds in the classification of X-ray and optical absorbed

Fig. 11 Specific accretion rate

Fig. 12 Space density versus

5.4 The future of AGN studies in the X-ray band

6.2 γ -ray detectors

6.3 The HE band

Currently four γ -ray space telescopes or instruments sensitive to photons in the

6.3.1 AGN in the HE γ -ray band

6.4 The VHE band

6.4.1 AGN in the VHE γ -ray band

42 tevcat.uchicago.edu and http://www.asdc.asi.it/tgevcat.

6.5 Blazar population properties and contribution to the γ -ray extragalactic

6.7 The future of AGN studies in the γ -ray band

We discuss in this section variability-selected AGN. An overview of AGN variability

7.1 An overview of AGN variability in the local Universe