Probing Relativistic Astrophysics Around SMBHs: The Suzaku AGN Spin Survey

Abstract

In addition to providing vital clues as to the formation and evolution of
black holes, the spin of black holes may be an important energy source in the
Universe. Over the past couple of years, tremendous progress has been made in
the realm of observational measurements of spin. In particular, detailed
characterization and modeling of X-ray spectral features emitted from the inner
regions of black hole accretion disks have allowed us to probe the location of
the innermost stable circular orbit and hence the spin. In this contribution, I
will describe the Suzaku AGN Spin Survey, an on-going Cycle 4--6 Suzaku Key
Project that aims to probe five nearby AGN with sufficient depth that strong
gravity effects and spin can be robustly assessed. Of the three objects for
which we currently have our deep datasets, we can constrain spin in two
(NGC3783, a>0.9; Fairall 9, $a=0.45\pm 0.15$) whereas complexities with the
warm absorber prevent us from reporting results for NGC3516. We conclude with a
brief discussion of spin-dependent selection biases in flux-limited surveys.

Full text

Page 1

arXiv:1112.0036v1 [astro-ph.HE] 30 Nov 2011
Probing Relativistic Astrophysics Around
SMBHs: The Suzaku AGN Spin Survey
C.S.Reynolds∗, L.W.Brenneman†, A.M.Lohfink∗, M.L.Trippe∗,
J.M.Miller∗∗, R.C.Reis∗∗, M.A.Nowak‡and A.C.Fabian§
∗Dept. of Astronomy, University of Maryland, College Park, MD20742, USA
†Harvard Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA, USA
∗∗Department of Astronomy, University of Michigan, Ann Arbor, MI 48109, USA
‡MIT Kavli Institute for Astrophysics, Cambridge, MA 02139, USA
§Institute of Astronomy, Madingley Road, Cambridge, CB3 OHA, UK
Abstract. In addition to providing vital clues as to the formation and evolution of black holes, the
spin of black holes may be an important energy source in the Universe. Over the past couple of
years, tremendous progress has been made in the realm of observational measurements of spin. In
particular, detailed characterization and modeling of X-ray spectral features emitted from the inner
regions of black hole accretion disks have allowed us to probe the location of the innermost stable
circular orbit and hence the spin. In this contribution, I will describe the Suzaku AGN Spin Survey,
an on-going Cycle 4–6 Suzaku Key Project that aims to probe five nearby AGN with sufficient
depth that strong gravity effects and spin can be robustly assessed. Of the three objects for which
we currently have our deep datasets, we can constrain spin in two (NGC 3783, a > 0.9; Fairall 9,
a=0.45±0.15)whereas complexities with the warm absorberprevent us from reporting results for
NGC 3516. We conclude with a brief discussion of spin-dependent selection biases in flux-limited
surveys.
Keywords: Active Galactic Nuclei, Seyfert galaxies, accretion disks, black holes
PACS: 98.54.Cm, 95.30.Sf, 95.85.Nv
INTRODUCTION
Probing the physics of the strong gravity region close to astrophysical black holes has
been a long held promise of X-ray astronomy. Starting with ASCA studies of the Seyfert
galaxy MCG–6-30-15 [23, 8], studies of relativistically broadened and redshifted X-
ray spectral features (aka broad iron lines) from the innermost regions of black hole
accretion disks have allowed this promise to be realized. In this contribution, I shall
give a very brief introduction to broad iron line studies of active galactic nuclei (AGN)
before giving a progress report on the Suzaku AGN Spin Survey, a Cycle 4–6 Suzaku
Key Project whose primary aim is to measure the black hole spin in a sample of nearby
AGN. I shall conclude with a brief discussion of spin-dependent selection biases that
may affect future flux-limited surveys of black hole spin.
What are the origin of the spectral features that we study? There is strong evidence
that luminous AGN accrete from radiatively-efficient disks [6, 22] in which the vast
majority of the accretion energy is released within r ∼ 20GM/c2of the supermassive
black hole (SMBH). Much of this energy is radiated as thermal optical/UV emission.
However, of course, AGN are known to be powerful sources of hard (variable) X-

Page 2

110 100
0.1
1
10
100
1000
νF(ν)
Energy (keV)
FIGURE 1.
red (top) is the rest-frame X-ray reflection spectrum assuming that the disk has an ionization parameter
ξ = 10ergs−1cm, solar abundances and is irradiated by a Γ = 2 power-law spectrum of X-rays. The
blue (bottom) curve shows this same spectrum convolved with relativistic smearing relevant for a disk
extending down to the ISCO of rapidly spinning (a = 0.9) black hole viewed at 30degrees inclination
with an irradiation profile of ε ∝ r−3. For clarity, the two spectra have been offset in normalization by a
factor of 10.
Theoretical X-ray reflection spectra (tabulated in the reflionx model; [19]). Shown in
ray emission and so we conclude that a significant fraction of the accretion energy is
channeled into an X-ray emitting corona situated above the cold, optically-thick disk.
Given this geometry, the inner accretion disk will be strongly irradiated by the X-ray
corona resulting in the so-called X-ray reflection spectrum consisting of fluorescence
lines, radiative recombination emission, and Compton backscattered radiation (Fig. 1).
The presence of these X-ray reflection features is a natural consequence of the basic
AGN paradigm. As expected, we see very similar features in the spectra of Galactic
Black Hole Binaries [11] and accreting weakly-magnetized neutron stars [3].
Characterizing the X-ray reflection spectrum provides a our best window to date
on the immediate environment of the SMBH. Fitting full reflection models convolved
with appropriate relativistic smearing kernels (Fig. 1) to high-quality broad-band X-
ray spectra allow us to determine the photospheric ionization state, the photospheric
iron abundance, the inclination, and the irradiation profile of the inner disk. We can also
determinetheinner“reflection" edge ofthedisk[9] from thestrength ofthegravitational
redshifting. Identifying this edge with the general relativistic innermost stable orbit then
allows us to constrain the spin of the SMBH [18].
THE SUZAKU AGN SPIN SURVEY
Motivation and basic goals
Extracting the relatively subtle signatures of black hole spin from X-ray spectra
requires high-quality data. For this reason, prior to the initiation of Suzaku Key Projects

Page 3

1 1025 20
0.5
1
1.5
Ratio
Energy (keV)
0.4 0.6 0.81
0
20
40
∆χ2
a
FIGURE 2.
a simple powerlaw continuum. XIS/BI data are shown in red, XIS/FI data are in black (co-added for the
purposesofdisplay),andPINdataare ingreen.Note thestrongwarm absorberbelow3keV andthestrong
narrow iron line at 6.4keV (rest frame). Despite these spectral complexities, broadened disk reflection is
strongly required by these data. Figure from [17]. Right panel : Goodness of fit parameter as a function of
dimensionless spin parameter for a fiducial spectral model which includes a multi-zone warm absorber,
distant reflection, soft excess, disk reflection, and a fixed XIS/PIN cross-normalizationfactor (black line).
Also shown are variants of this model where the warm absorber parameters are frozen (red; to assess
trades with absorption parameters), the XIS/PIN cross-normalization is a free parameter (blue), and only
hard band (> 3keV) data are modeled (green). From bottom to top, the dotted lines show the 90%, 95%
and 99% confidence levels for one interesting parameter. Figure from [2].
Left panel : 0.5–40keV Suzaku/XIS+PIN spectrum of NGC 3783 shown as a ratio against
in 2008, only one AGN possessed a useful constraint on the spin of its SMBH; multiple
analyzes of MCG–6-30-15 found a rapidly rotating black hole, with dimensionless spin
parameter a > 0.9 [4, 1, 12]. Even this single datapoint proves interesting; the discovery
of such a rapidly rotating black hole in a radio-quiet AGN rules out simple spin-driven
models for the radio-quiet/radio-loud dichotomy [21]. Clearly, expanding a spin study
to a sample of AGN is a high priority.
The Suzaku AGN Spin Survey is an on-going Cycle 4–6 Suzaku Key Project that,
when complete, will provide high-quality spectral datasets for five nearby AGN that had
previous indications for broad iron lines (NGC3516, NGC 3783, Fairall 9, 3C120, and
Mrk 841). The primary goals of this project are to study the physics of the innermost
accretion disk and the black hole spin by characterizing the relativistically broadened
reflection features. However, these datasets also allow studies of the circumnuclear
environment (via the emission/absorption signatures of photoionized matter) as well
as the physics of the X-ray continuum (via the study of spectral curvature and energy-
dependent temporal lags) and hence will be a valuable legacy to AGN researchers.
In the rest of this contribution, we describe the current results from this program.
NGC 3783
The Seyfert 1.5 galaxy NGC 3783 was observed quasi-continuouslyby Suzaku in July
2009, with a total on-source exposure of 210ks. Detailed analysis of these data can be

Page 4

Count per second
P6 P5P4P3 P2P1
0.2-12 keV
6
7
8
9
10
11
15-40 keV
0
0.1
0.2
0.3
Low
High
Colour
2.0-10.0 keV
0.3-1.0 keV
Normalized
−2
−1
0
1
2
Time (seconds)
0 5×104
105
1.5×105
2×105
2.5×105
3×105
FIGURE 3.
band light curves and the softness ratio (bottom panel). Complex patterns of variability are revealed;
during some periods (P1 and P5–P6), color variations track the count rate variations perfectly whereas
during other periods (P3 and P4), significant color variations are seen which are uncorrelated with overall
count rate. Detailed analysis shows that the spectral variability can be explained as being due, principally,
to changes in the ionization and strength of the inner disk reflection. Figure from [17].
Summed XIS light curve for NGC 3783 (top panel) together with normalized restricted
found in [2, 17]; here we give a brief summary of these findings. Figure 2 (left) shows
the 0.5–40keV Suzaku/XIS+PIN spectrum of this object. Here, and in all Suzaku spec-
tra presented here, we ignore XIS data in the 1.5–2.5keV range due to some obvious
calibration problems associated with modeling the absorption edges in the X-ray mir-
ror. The soft X-ray spectrum (< 3keV) is strongly affected by a warm absorber that has
been well studied in previous works, including a 900ks Chandra High-Energy Trans-
mission Grating (HETG) campaign in 2000–2001 [13]. A narrow fluorescent 6.4keV
iron line from low velocity, distant material is also readily apparent. Despite these com-
plexities,thespectrumstronglyrequires relativisticallysmearedreflection fromtheinner
accretion disk [2]. Describing the disk reflection with the reflionx ionized reflection
model [19], and using the relconv relativistic smearing model [5], we find a disk
inclination of i = 22+3
relativistic broadening is very strong, requiring black hole spin of a > 0.9 at the 90%
confidence level (CL) even allowing for possible systematic errors introduced by the
analysis/modeling procedure (Fig. 2 [right]).
We note briefly the independent analysis of these data by Patrick and collaborators
[16] which, contrary to our conclusions, finds that NGC 3783 has a low (or even retro-
grade) spin, a < 0.37. This alternative analysis differs from ours in several key ways.
Probably most importantly, they include an additional continuum component (described
−8degrees and supersolar iron abundance Z = 3.7±0.9Z⊙. The

Page 5

125
1
1.2
1.4
data/model
Energy (keV)
0 0.20.4 0.60.8
3650
3652
3654
3656
3658
χ2
a
FIGURE 4.
simple unabsorbed power-law model. Note the strong narrow and broad iron lines in the spectrum of this
remarkable “clean" (i.e. unabsorbed) AGN. Right panel : Preliminary spin constraint on Fairall 9. Shown
here is the goodness of fit parameter as a function of black hole spin in a multi-epoch analysis of the Key
Project Suzaku data along with data from the long 2010 XMM-Newton observation. The spectral model
includes FeXXV and FeXXVI line emission from photoionizedplasma in the circumnuclearenvironment
(assumed to remain constant between the Suzaku and XMM-Newton observations) and a soft excess
component described by an additional thermal comptonization model. See A.Lohfink’s contribution in
this volume for further details.
Left panel : XIS0 (black), XIS1 (red) and XIS3 (blue) spectra of Fairall 9 ratioed against a
by a “warm" comptonization model), and make the ansatz that the iron abundance is
solar. We have confirmed that, fixing the iron abundance to unity, significantly decreases
the inferred spin (90% confidence range a = 0.28−68). However, even when we in-
clude the additional comptonization component, we find that supersolar abundance is
strongly required (90% confidence range Z = 3.3−3.7Z⊙; ∆χ2= 35 for 1 extra degree
of freedom) and the need for rapid prograde spin is restored. Thus, we argue that the
solar abundance ansatz can be falsified (we also note that the cores of Seyfert galaxies
are atypical environments and hence supersolar abundances are not unreasonable [14]).
A more complete comparison of our analysis with the Patrick analysis, including a dis-
cussion of some important differences in data handling and modeling, will be presented
in a forthcoming publication.
Significant flux and spectral variabilityis seen duringthis Suzakuobservation(Fig. 3);
our detailed analysis of the variability is described in [17]. We find that the spectral
variability(includinglarge amplitudechanges in thePIN-band flux) isdrivenby changes
in the ionization and strength of the X-ray reflection from the inner accretion disk,
whereas the warm absorber is remarkably constant during the 400ks duration of this
observation. Astrophysically, these changes must be a result of changes in both the
power and geometry/location of the irradiating X-ray source.
Fairall 9
The luminous Seyfert 1 galaxy Fairall 9 was observed quasi-continuously in May
2010, with a total on-source exposure of 190ks. Fairall 9 is a prototypical example of a
“bare" Seyfert nucleus [15]; previous studies have failed to find any signatures of either