ataleoftwomonsters ...hiddenmonsters/talks/robinson.pdfgeometry,$cloud$distribution$etc. 1.0 0.8 0.6...
TRANSCRIPT
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
Norm
aliz
ed fl
ux
56500560005550055000
MJD
Optical 3.6 µm 4.5 µm
50 100 150 200 250
A TALE OF TWO MONSTERS: EMBEDDED AGN IN NGC6418 AND
IRAS16399−0937
Andy RobinsonRochester Institute of Technology
Photometry Aperture
0 100 200 300
N
E
Figure 3.1: NGC6418 false color image from the Spitzer space telescope at3.6 µm. The white circle has a radius of 6 pixels and represents the aperture(diameter 7.2”) used for the photometry for the cross-correlation analysis.
flux. Therefore, the flux vs. aperture function should plateau with increasing
radii. All Figures for all AGNs can be found in Appendix A.
All the mean flux density plots exhibit, as expected, a rapidly increasing
flux density that plateaus at larger aperture sizes with some exceptions.
NGC6418, UGC10697, MRK885, AKN524 and KAZ163 exhibit aperture
analysis plots that increase in mean flux without reaching a plateau. This
behavior is explained by the extended emission from their host galaxies as
seen in Figures 3.1, A.49, A.41, A.17 and 3.2. In addition, KAZ163 is the
southern member of an interacting galaxy pair. The companion galaxy is
the fainter object visible in 3.2. The companion galaxy is at a distance
greater than 20 pixels (12”). This aperture size is the largest used in the
analysis.
The “knee” of the flux vs. aperture function is used to determine the best
aperture. To determine the “knee”, the simple definition of the intersection
of the slopes of the first two data points with the last two data points is used.
On average, the obtained knee values agree with the ideal aperture found
via visual inspection. It is important to note that there is no significant
di↵erence between the aperture analysis for the 3.6 µm and 4.5 µm channels.
As an example, Figures 3.3 and A.4 show the analysis done for both channels
for NGC6418. There is less than a pixel di↵erence between the two channels.
22
Outline• Two (contrasting) examples of dust enshrouded AGN• NGC6418 – isolated Seyfert• IR reverberation mapping with Spitzer• “changing look” AGN
• IRAS16399-‐0937 merger system• Multiwavelength SED modeling• Evidence for a deeply embedded AGN
Sales et al. 2015, ApJ, 799, 25
Vazquez et al. 2015, ApJ, 801, 127Robinson et al., in preparation
50 100 150 200 250
Billy Vazquez; Michael Richmond; Triana Almeyda; Shawn Foster; + Spitzer reverberation mapping collaboration
Dinalva Sales; Jack Gallimore; Moshe Elitzur +
2
Spitzer monitoring campaign
2.5-‐year IR – optical monitoring campaign Aug. 2011 – Jan. 2014• 12 type 1 AGN monitored at 3.6 and 4.5 µm with Spitzer• Optical data from Liverpool Telescope, CSS, PTF• Some results presented in Billy Vazquez’s talk (Tuesday)• Triana Almeyda poster on reverberation models
����
��
����
��
����
��
����
������ ������ ����� ����� ����� ����� ����� �� �� ����� ���� ����� �����
���������� �������
����������
B. Vazquez, 2015 PhD3
Dust reverberation mapping• Response of torus dust emission to UV-‐optical variations depends on size, geometry, cloud distribution etc.
1.0
0.8
0.6
0.4
0.2
0.0
Relat
ive am
plitu
de
2520151050Delay
1.8
1.6
1.4
1.2
1.0
0.8
0.6
Rela
tive
flux
302520151050
Time
optical dust emission
At short wavelengths, IR lag ~ inner radius of torus, usually taken to be dust sublimation radius
Rd ≈ 0.4L
1045erg−1#
$%
&
'(
0.51500KTsub
#
$%
&
'(
2.6
pc
Transfer function encodes torus properties…
…convolution with driving optical light curve → IR light curve
(Nenkova et al. 2008b, Barvainis 1987)4
Almeyda et al., in prep.
NGC6418 optical & IR light curves
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
Norm
aliz
ed fl
ux
56500560005550055000
MJD
Optical 3.6 µm 4.5 µm
Seyfert 1 in Sab host; z = 0.0285
Cycle 8 (3 days) Cycle 9 (30 days)
Vazquez et al. 2015
5
NGC6418 optical & IR light curves
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
Norm
aliz
ed fl
ux
56500560005550055000
MJD
Optical 3.6 µm 4.5 µm
Seyfert 1 in Sab host; z = 0.0285
Cycle 8 (3 days) Cycle 9 (30 days)
6
��
�����
����
�����
����
�����
����
�����
����
�� �� �� �� �� � � ��
������������ ���
� ���� ���
Cross-‐correlation analysis
• Increase in IR– optical lags following flare
��
�����
�����
�����
�����
����
�����
�����
�����
��� �� �� � ��� �� ��
������������ ���
�� �������
3.6 µm vs optical
��
�����
�����
�����
�����
����
�����
�����
�� � ��� �� �� �
������������ ���
�� �������
��
�����
�����
�����
�����
����
�����
�����
�����
�����
����
�� �� �� ��� ��� ��� ��� ��� ��
������������ ���
�� �������
𝜏 = 33.3%&.'('.)
𝜏 = 64.5%-..(-.)
𝜏 = 41.4%&.0(&.)
𝜏 = 41.4%&.0(&.)
𝜏 = 80.3%'.0('..
4.5 µm vs optical
Cycle 9
Cycle 8
7
��
�����
����
�����
����
�����
����
�����
�� � ��� ��� ��� ��� �� ��� ���
������������ ���
� ���� ���
Cross-‐correlation analysis
• 4.5 µm lags 3.6 µm; lag increased following flare
Implications of the Full Campaign on NGC6418
!"
!"#$
!"#%
!"#&
!"#'
!"#(
!"#)
!"#*
!"#+
!"#,
!$
-'" -&" -%" -$" !" !$" !%" !&" !'" !(" !)" !*" !+"
!"##$%&'(")*+&!'"#
./0!12/345
Figure 6.14: An instance of the CCF (⌧) of the 4.5 vs 3.6 µm light curves.
!"
!"#"$
!"#%
!"#%$
!"#&
!"#&$
!"#'
!%" !%$ !&" !&$ !'"
!"#$%$&'&()*+,-.&()
()*!+,)-./
Figure 6.15: 3.6 µm and 4.5 µm CCCD for Spitzer Cycle 9. The centroid ofthe CCCD is 20.3+1.0
�1.0 days.
116
Cycle 9
Cycle 8
𝜏 = 20.3%'.-('.-
𝜏 = 12.4%'.-(-.4
4.5 µm vs 3.6 µm
For dust grains in radiative equilibrium: 𝑅6 𝑅7~ 𝑇:;< 𝑇6⁄ >⁄ ; a ≈ 2.6 for ISM composition
Expect: t4.5/t3.6 ~ 1.8
Measured: t4.5/t3.6 ≈ 1.2(in both cycles)
Favours clumpy dust distribution
8
�����
����
���� ���� ���� ���� ����
��������������
������� ���������
����������������� ��� ������������������������
������������������� !������������� !����
�����"����
Torus radius-‐luminosity relation
Opt. – IR lags small compared to predicted sublimation radius for standard ISM dust composition, but…
increase in lags following flare ~ consistent with increase in dust sublimation radius,R ~ L0.5
9
Optical spectra: 2001
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
Rela
tive
Flux
575005700056500560005550055000MJD
Optical 3.6 µm 4.5 µm
120
100
80
60
40
20
Relat
ive F
λ
7000600050004000Wavelength (Å)
SDSS Apr 2001
Broad line Ha/Hb ≥ 6; AV ≥ 2
11
Optical spectra: Jan. 2014
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
Rela
tive
Flux
575005700056500560005550055000MJD
Optical 3.6 µm 4.5 µm
80
60
40
20
Rela
tive
F λ
7000600050004000
Wavelength (Å)
APO Jan 2014
Broad line Ha/Hb ≈ 3; AV ≈ 0
12
Optical spectra: Aug. 2015
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
Rela
tive
Flux
575005700056500560005550055000MJD
Optical 3.6 µm 4.5 µm
50
40
30
20
10
Rela
tive
F λ
7000600050004000
Wavelength (Å)
APO Aug 2015
Broad line Ha/Hb ≈ 6; AV ≈ 2
13
Optical spectra: Feb. 2016
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
Rela
tive
Flux
575005700056500560005550055000MJD
Optical 3.6 µm 4.5 µm
80
70
60
50
40
30
20
10
Rela
tive
F λ
7000600050004000
Wavelength (Å)
WHT 17 Feb 2016
14
NGC6418 as a “changing look” AGN
• Factor ≥ 2 increase in optical luminosity accompanied by similar increase in IR luminosity and followed by…• Emergence of Sy 1 spectrum• Increase in optical -‐ IR lags, consistent with L0.5, suggesting increase in sublimation radius• Line-‐of-‐sight extinction to BLR decreased from AV ≥ 2 → ≈0• Timescales: opt.-‐IR flare ~ 100 days; change in spectrum ≤ 1 year• Seems to be returning to low activity state (Jan 2014 – present)
Change in “look”
Flare in accretion disk luminosity
Increase in BLR
luminosity
Increase in torus inner radius
Destruction of line-‐of-‐sight dust 15
OH Megamaser Galaxies
~ 20% of (U)LIRGs contain extremely luminous OH masers • emitting primarily in the 1667 and 1665 MHz lines • luminosities ∼102–4 L⊙• Represent distinct evolutionary phase in gas-‐rich mergers?• Probe of high z star formation?
• Multiwavelength study of ~80 OHMG• Optical – NIR: HST observations archive data; Gemini integral field spectroscopy• IR-‐sub mm: Spitzer + Herschel archive data• Radio: VLA observations + archive data
IRAS 16399-‐0937 (z = 0.027)• LIRG (LFIR ≈ 1011.2 L⊙; LOH ≈ 101.7 L⊙) • mid-‐late stage merger
19
Double nuclei in common, tidally distorted envelope è mid-stage major, gas rich, merger.
Multiwavelength Morphology of IRAS16399-‐0937
HST/ACS 0.4 µm HST/ACS 0.8 µm HST/NICMOS 1.6 µm
HST/ACS Hα+[NII] Spitzer/IRAC 8.0 µm PAH VLA 1.49 GHz
z = 0.027012 => 111.5 Mpc
21
50 100 150 200 250
Optical spectrum
• Nucleus separation ≈3.4 kpc
5
10
15
20
25
30 IRAS 15268-7757
0.00.5
1.0
1.5
2.0
2.5
3.0 IRAS 15361-0313
0
5
10
15 IRAS 15437+0234 obj 1
01
2
3
4
5
6 IRAS 15437+0234 obj 2
0.5
1.0
1.5
2.0
2.5
3.0 IRAS 15456-1336
0.00
0.10
0.20
0.30
0.40IRAS 16399-0937 obj 1
0.0
0.2
0.4
0.6
0.8 IRAS 16399-0937 obj 2
0
1
2
3
4IRAS 16504+0228
0.00.20.40.60.81.01.21.4 IRAS 17138-1017
0.0
0.5
1.0
1.5
2.0
2.5 IRAS 17324-6855
2
4
6
8IRAS 17467+0807
20
40
60
80
100 IRAS 18078-5815
5000 5500 6000 650002
4
6
8
10
12 IRAS 18093-5744 obj 1
5000 5500 6000 650002468
101214 IRAS 18093-5744 obj 2
FIG. 2.ÈContinued
53
5
10
15
20
25
30 IRAS 15268-7757
0.00.5
1.0
1.5
2.0
2.5
3.0 IRAS 15361-0313
0
5
10
15 IRAS 15437+0234 obj 1
01
2
3
4
5
6 IRAS 15437+0234 obj 2
0.5
1.0
1.5
2.0
2.5
3.0 IRAS 15456-1336
0.00
0.10
0.20
0.30
0.40IRAS 16399-0937 obj 1
0.0
0.2
0.4
0.6
0.8 IRAS 16399-0937 obj 2
0
1
2
3
4IRAS 16504+0228
0.00.20.40.60.81.01.21.4 IRAS 17138-1017
0.0
0.5
1.0
1.5
2.0
2.5 IRAS 17324-6855
2
4
6
8IRAS 17467+0807
20
40
60
80
100 IRAS 18078-5815
5000 5500 6000 650002
4
6
8
10
12 IRAS 18093-5744 obj 1
5000 5500 6000 650002468
101214 IRAS 18093-5744 obj 2
FIG. 2.ÈContinued
53
5
10
15
20
25
30 IRAS 15268-7757
0.00.5
1.0
1.5
2.0
2.5
3.0 IRAS 15361-0313
0
5
10
15 IRAS 15437+0234 obj 1
01
2
3
4
5
6 IRAS 15437+0234 obj 2
0.5
1.0
1.5
2.0
2.5
3.0 IRAS 15456-1336
0.00
0.10
0.20
0.30
0.40IRAS 16399-0937 obj 1
0.0
0.2
0.4
0.6
0.8 IRAS 16399-0937 obj 2
0
1
2
3
4IRAS 16504+0228
0.00.20.40.60.81.01.21.4 IRAS 17138-1017
0.0
0.5
1.0
1.5
2.0
2.5 IRAS 17324-6855
2
4
6
8IRAS 17467+0807
20
40
60
80
100 IRAS 18078-5815
5000 5500 6000 650002
4
6
8
10
12 IRAS 18093-5744 obj 1
5000 5500 6000 650002468
101214 IRAS 18093-5744 obj 2
FIG. 2.ÈContinued
53
5
10
15
20
25
30 IRAS 15268-7757
0.00.5
1.0
1.5
2.0
2.5
3.0 IRAS 15361-0313
0
5
10
15 IRAS 15437+0234 obj 1
01
2
3
4
5
6 IRAS 15437+0234 obj 2
0.5
1.0
1.5
2.0
2.5
3.0 IRAS 15456-1336
0.00
0.10
0.20
0.30
0.40IRAS 16399-0937 obj 1
0.0
0.2
0.4
0.6
0.8 IRAS 16399-0937 obj 2
0
1
2
3
4IRAS 16504+0228
0.00.20.40.60.81.01.21.4 IRAS 17138-1017
0.0
0.5
1.0
1.5
2.0
2.5 IRAS 17324-6855
2
4
6
8IRAS 17467+0807
20
40
60
80
100 IRAS 18078-5815
5000 5500 6000 650002
4
6
8
10
12 IRAS 18093-5744 obj 1
5000 5500 6000 650002468
101214 IRAS 18093-5744 obj 2
FIG. 2.ÈContinued
53
l (Å)
“North” nucleus: Low Ionization nuclear emission region – weak AGN(?)
“South” nucleus: starburst
(Spectra from Kewley et al. 2001)
22
Green: Hα+[NII] (HST ACS)Red: 1.49 GHz VLAContours: ISM PAH-dust 8µm emission (fromSpitzer IRAC 8.0 & 3.6 µm images)
1''
Red: 1.49 GHz VLA Blue: Chandra 0.5–2 keV X-ray
Multiwavelength Morphology of IRAS16399-‐0937
23
Extended and compact components of radio emission consistent with star formation
Compact X-‐ray source associated with N nucleus, but weak: Lx(0.5–2 keV) ~ 5x1040 erg/s
Mid-‐Infrared Spectrum (nuclei not resolved)
Low-resolution Spitzer IRS spectrum. Vertical solid lines indicateabsorption bands of water ice (6.0µm) and HACs (6.85µm and7.25µm)
[NeV], [OIV] not detected
24
Spectral energy distribution fits
• SED fits using MCMC code clumpyDREAM (Gallimore)• Simultaneous fits to both nuclei, unresolved points at l >14 µm treated as upper limits
Clumpy torus model (Nenkovaet al. 2008a,b)
ISM dust/PAH model (Draine & Li 2007)
Stellar population model (GRASIL Silva et al. 1998)
27
ModelLAGN(erg/s)
3.4x1044
LISM(erg/s)
2.9x1044 erg/s
SFRFIR(M¤/yr)
11.6
Bayes informationcriterion (BIC) = 872
Measured (M¤/yr)SFRX-ray 10.3±3.7SFR8μm 4.2±0.6SFR1.4GHz 6.0±0.7
SED fit – North nucleus with AGN
28
ModelLISM(erg/s)
5.1x1044
SFRFIR(M¤/yr)
20.3
Bayes informationcriterion (BIC) = 1225
Measured(system)
(M¤/yr)
SFR8μm 19.4±2.9SFR24μm 23.2±3.2SFR1.4GHz 13.7±1.5
SED fit – North nucleus without AGN
Measured (M¤/yr)SFRX-ray 10.3±3.7SFR8μm 4.2±0.6SFR1.4GHz 6.0±0.7
29
Measured (M¤/yr)SFR8μm 3.0±0.4SFR1.4GHz 2.7±0.2
SED fit – South nucleus (no AGN)
ModelLAGN(erg/s)
–
LISM(erg/s)
8.9x1043 erg/s
SFRFIR(M¤/yr)
3.6
30
North nucleus – “torus” parameters
On the other hand, the torus dominates the emission at shortwavelengths; at 2 !m, more than 80% of the flux measuredwith apertures!100 comes from the torus even though its imagesize is less than 0.04 00 (Weigelt et al. 2004).
These difficulties highlight a problem that afflicts all IR stud-ies of AGNs. The torus emission can be expected to dominate theAGN observed flux at near-IR because such emission requireshot dust that exists only close to the center. But longer wave-lengths originate from cooler dust, and the torus contribution canbe overwhelmed by the surrounding regions. Unfortunately, thereare not toomany sources like NGC 1068. No other AGN has beenobserved as extensively and almost no other observations havethe angular resolution necessary to identify the torus component,making it impossible to determine in any given source which arethe wavelengths dominated by torus emission. There are no easysolutions to this problem. One possible workaround is to forgofitting of the spectral energy distribution (SED) in individualsources and examine instead the observations of many sourcesto identify characteristics that can be attributed to the torus sig-nature. One example for the removal of the starburst componentis the Netzer et al. (2007) composite SED analysis of the Spitzerobservations of PG quasars. Netzer et al. identify two subgroupsof ‘‘weak FIR’’ and ‘‘strong FIR’’ QSOs and a third group of far-IR (FIR) nondetections. Assuming a starburst origin for the FIR,they subtract a starburst template from the mean SED of eachgroup. The residual SEDs are remarkably similar for all threegroups, and thus can be reasonably attributed to the intrinsicAGN contribution, in spite of the many uncertainties. However,while presumably intrinsic to the AGN, it is not clear what frac-tion of this emission originates from the torus as opposed to theionization cones. An example of a sample analysis that may haveidentified the torus component is the Hao et al. (2007) compi-lation of Spitzer IR observations. In spite of the large aperture ofthese measurements, Seyfert 1 and 2 galaxies show a markedlydifferent behavior for the 10 !m feature, both in their mean IRSEDs and in their distributions of feature strength. Furthermore,ultraluminous IR galaxies (ULIRGs) that are not associated withAGNs show yet another, entirely different behavior, indicatingthat the observed mean behavior of Seyfert galaxies is intrinsicto the AGN. Accepting the framework of the unification scheme,the differencesHao et al. find between the appearances of Seyfert 1and 2 galaxies can be reasonably attributed to the torus contribu-
tion; the ionization cones’ dust is optically thin, and therefore itsIR emission is isotropic and cannot generate the observed differ-ences between types 1 and 2.
Here we invoke both approaches in comparing our model pre-dictions with observations. We start by assembling dusty cloudsinto complete models of the torus, as described in x 2. Our modelpredictions for torus emission and the implications for IR obser-vations are presented in xx 3Y5, while in x 6 we discuss aspectsof clumpiness that are unrelated to the IR emission, such as thetorus mass and unification statistics. In x 7 we conclude with asummary and discussion.
2. MODEL OF A CLUMPY TORUS
Consider an AGN with bolometric luminosity L surroundedby a toroidal distribution of dusty clouds (Fig. 1). The ‘‘naked’’AGN flux at distance D is FAGN ¼ L/4"D2 at any direction, butbecause of absorption and reemission by the torus clouds the ac-tual flux distribution is anisotropic, with the level of anisotropystrongly dependent on wavelength. The grain mix has standardinterstellar properties (see x 3.1.1 of Paper I for details), and theoptical depth of each cloud is #V at visual.
2.1. Dust Sublimation
The distribution inner radius Rd is set by dust sublimation attemperature Tsub. From x 3.1.2 in Paper I,
Rd ’ 0:4L
1045 erg#1
! "1=2 1500 K
Tsub
! "2:6
pc: ð1Þ
Barvainis (1987) derived an almost identical relation for Rd. Hisequation (5) has the same normalization and only a slight dif-ference in the power of Tsub (2.8 instead of 2.6); this differencereflects the more detailed radiative transfer calculations we per-form. Here the distance Rd is determined from the temperatureon the illuminated face of an optically thick cloud of compositedust representing the grain mixture. The sharp boundary we em-ploy is an approximation. In reality, the transition between thedusty and dust-free environments is gradual because individualcomponents of the mix sublimate at slightly different radii, withthe largest grains surviving closest to the AGN (Schartmann et al.2005). From near-IR reverberation measurements, Minezaki et al.
Fig. 1.—Model geometry. Dusty clouds, each with an optical depth #V at visual, occupy a toroidal volume from inner radius Rd , determined by dust sublimation(eq. [1]), to outer radius Ro ¼ YRd . The radial distribution is a power law r#q, and the total number of clouds along a radial equatorial ray is N 0. Various angulardistributions, characterized by a width parameter $, were considered. The angular distribution has a sharp edge on the left and a smooth boundary (e.g., a Gaussian) onthe right.
AGN DUSTY TORI. II. OBSERVATIONAL IMPLICATIONS 161
From Nenkova et al. 2008
Parameter Value
“Size” (Ro) 20 pc
Angular height (σ) ≥ 66°
Inclination (i) ≥ 60°
N° clouds (Nd) ≥ 14
Opt. depth (τV) ≈30
Covering frac. (Cf) ≥ 0.998
Quasi-spherical
distribution of optical
thick clouds
31
Origin of optical line emission?
• North nucleus covering fraction => ≈ 0.1% of AGN photons (QAGN) escape
• LINER spectrum not due to AGN photoionization – probably results from shock ionization
Qesc = (1–Cf)QAGN ≈ 3.6x1053 ionizing photons/s
Hα luminosity due to AGN photoionization:
LHα,AGN ≈ CISM pHα hνHα Qesc ≤ 4.9x1039 erg/s
(since CISM ≤ 1)
…only 2% of observed Hα luminosity of N nucleus (LHα,obs ≈ 3x1041 erg/s)
CISM = fraction ionizing photons absorbed by ISM
pHα ≈ 0.45, probability Hα photon emitted per H recombination
33
Summary
NGC6418
• Isolated Seyfert galaxy• Changing look AGN “caught in the act”• Intrinsic increase in AGN luminosity AND decrease in extinction• Evidence for increase in torus inner radius, following optical flare• Timescale < 1 yr
IRAS16399-‐0937
• Gas-‐rich, mid-‐stage, major merger, nuclei separated by ≈3 kpc• North nucleus contains moderately luminous AGN, embedded in ~spherical dust cloud distribution • Embedded AGN cannot produce LINER spectrum; probably shocks associated with merger driven gas flows
35
Torus covering fraction may not scale simply with instantaneous
luminosity
LINER spectrum does not necessarily indicate
presence of AGN