Transcript Slide 1

Are AGN ‘just’ scaled-up
stellar-mass black holes?
Rob Fender (Southampton)
+ Guy Pooley, Elena Gallo, Simone Migliari, Elmar Koerding, Sebastian Jester,
Stephane Corbel, Ralph Spencer, Dave Russell, Valeriu Tudose, Catherine
Brocksopp, Christian Kaiser, Tomaso Belloni, Jeroen Homan, Sera Markoff, Paolo
Soleri, Tom Maccarone, James Miller-Jones, Clement Cabanac, Robert Dunn,
Martin Bell…
The importance of
black hole accretion in
the universe
6<z<14 first ‘AGN’ reionize
universe
1<z<6 peak of AGN activity:
feedback regulates galaxy
growth, reheats cooling
flows, creates X-ray
background
z<1 AGN accretion rates
drop, accretion luminosity of
universe dominated by
binaries, jets dominate
radiation
Scaling black hole accretion with mass (naively!)
M / R = const
 constant accretion
efficiency
BUT density
and temp very
different
When jets are formed
(patterns of radio:X-ray coupling)
Patterns of outbursts in the
Hardness-Intensity diagram
600 day outburst of the black
hole GX 339-4 in 60 seconds
Data from Homan & Belloni
(2006)
Cause: disc instability cycle
(probably)
~30% Eddington 
Belloni, Corbel, Fender, Gallo, Hanke, Kalemci, McHardy, Maitra, Markoff, Nowak, Petrucci, Pottschmidt, Wilms
Do AGN behave
anything like this ?
Belloni, Corbel, Fender, Gallo, Hanke, Kalemci, McHardy, Maitra, Markoff, Nowak, Petrucci, Pottschmidt, Wilms
Relation to AGN: what would an ensemble of X-ray binaries look like … ?
Same source, different outburst…
Different source…
But the X-ray HID is no good
for AGN, because their discs
are cooler and peak at lower
frequencies – need a more
physically meaningful method
of comparison
The Disc Fraction Luminosity Diagram (DFLD):
Koerding, Jester & Fender (2006)
Simulated ensemble
of X-ray binaries
(Colour
scale is+ radio
loudness)
SDSS
quasars
LLAGN
All disc
All power-law
The real ensemble of BH X-ray binaries (Dunn, Fender et al. in prep)
>10 000 points – equivalent to
a large AGN sample
We do not have radio
loudness for the vast majority
of these points…
.. but with the new generation
of radio observatories (SKA
pathfinders) we will get these
and be able to make direct
comparison with AGN
Marscher et
al. suggest
3C120 is
behaving like
XRB GRS
1915+105 with
ejections
associated
with X-ray
colour
changes
GRS 1915+105: black hole accretion at ~Eddington
 unstable, quasi-periodic state changes and jet formation
One hour of (patchy)
data on GRS 1915+105,
persistently accreting
at ~Eddington
‘State’ changes can be
as rapid as seconds
(apparent disc radius
changes faster than
the viscous timescale)
This side sensitive to power-law
This movie is sped-up by 60x
This side sensitive
to disc
Is this what’s
happening in the
most luminous
AGN?
Perhaps timing properties are a better tracer of ejection?
Dips (‘zones’) of low variability
Gradual state transition
Ejection of the corona ? Fender, Homan & Belloni (in prep)
The power of jets
Cygnus X-1: a jet-blow bubble  calorimeter for jet power
zoom out x 50 000: jet-ISM interaction
(external shock over 106 years) (WSRT)
zoom out x 10: transient jet at state change
(internal shock over several hours)
Hard state jet (steady state)
(MERLIN)
(VLBA)
(Stirling et al. 2001; Fender et al. 2005; Gallo et al. 2006)
Optical confirmation of shocked nebula
H-alpha and O[III] Line-emitting nebula.
Russell, Fender et
al. (2006, 07)
 Narrow
bowshock with high
O[III]:Halpha
ratio
Analysis
indicates
LJET ~ LX
(at Eddington
ratio of ~0.02)
Blue = V-band
Red[O
= Hα
III] Green
/ Hα = [O III] (500.7nm)
Calibrating core
radio luminosity
to accretion rate
for X-ray binaries
.
Lradio a m1.4
(as predicted
for
.
Ljet a m)
Koerding, Fender &
Migliari (2006)
.
Then if we can rely on LRADIO calibration, we can see how LX varies with m
(X-ray luminosity)
.
Up here LJET ~ LX and falls off linearly with m
LJET
Direct evidence for
radiatively inefficient
accretion and jetdominated states (with
advection…)
LX
(mass accretion rate calculated from Lradio)
Koerding, Fender &
Migliari (2006)
‘Fundamental’ plane(s)
(Quantitatively linking X-ray
binaries and AGN)
Merloni, Heinz & di Matteo (2003)
Falcke, Koerding & Markoff (2004)
The ‘fundamental plane’
of black hole activity
What does the fundamental plane mean ?
.
Lradio a m1.4
if [a]
(which we’ve just shown)
and [b]
. .
LX/LEdd a (m/mEdd)2
.
 LX / M a (m / M)2
. radiatively inefficient accretion
(which is a general approximate solution for
where the accretion flow knows what Eddington ratio its at…)
then simple re-arranging gives us:
Lradio a LX0.7 M0.7
(c.f. LX0.6 M0.8 fit)
 The fundamental plane is almost perfectly recovered (extra tweaks required
at the level of +/- 0.1 in power law indices)
This implies that the plane is dominated by radiatively inefficient sources which
are jet-dominated (and that hard  soft state transitions do not have a strong
dependence on M, both in agreement with Koerding, Fender & Migliari 2006)
X-ray power
spectra
In XRBs break
frequency correlates
with jet power
(Migliari, Fender &
van der Klis 2006)
Fourier transform
of X-ray lightcurve
.. and scaling with
AGN is approximately
linear in M
(McHardy et al. 2006)
… so…..
Fundamental plane #2 !
Tbreak ~ M2 / L
.
.
Tbreak ~ M / (m / mEdd)
AGN
XRBs
McHardy, Koerding, Knigge, Uttley & Fender
(Nature, 2006); Koerding et al. (2007)
Accretion disc radii as
a function of luminosity
~0.1 Edd
~10-3 Edd
This line is the slope
predicted by the timing
plane
Rinner a L-1/3bol
Hysteretical
Zone
~10-5 Edd
Hard
State
~10-7 Edd
Lbol
~ISCO
Cabanac, Fender et al.
(in prep)
~100 ISCO
So what do these planes mean ?
The `fundamental plane’ means (we think) that
• all black holes produce the same amount of kinetic power output
(jet) per unit mass of accreted material
• the radiation produced is a function of Eddington ratio (the ratio
of accretion rate to the maximum rate), so for a given accretion
rate in kg, more massive black holes produce less radiation
(unless you’re at or close to Eddington limit)
The new `timing’ plane seems to mean that
• variability timescales depend linearly (as expected) with black
hole mass, and inversely on accretion rate (in Eddington units)
 We have found extremely simple scalings between objects
differing in both mass and accretion rate by eight orders of
magnitude !
Beware of cheap
imitations
(or: who needs an event horizon)
Neutron stars and White Dwarfs do it too
radio
flare
Radio
flaring and
hysteretical
patterns
observed
from
Cataclysmic
Variable
SS Cyg
(Koerding et
al. 2008)
Conclusions:
Patterns: the qualitative relation between spectral states and jet production
may be independent of black hole mass
Planes: Jet power and mass accretion rate may be quantified as a function of
radio luminosity, and demonstrate jet-dominated advective states are the
norm
.
0.6
0.8
plane #1
Lradio a LX M
( LJet a m a (LX/LEdd )0.5 )
. .
plane #2
Tbreak a M / (m/mEdd)
… so everything looks simple as long as you’re happy that accretion flows
know what Eddington ratio they are at….
• What next ? we can use patterns from XRBs to estimate e.g. the kinetic
luminosity function of Active Galactic Nuclei (Koerding, Jester & Fender 2008
 LLAGN dominate kinetic feedback in local universe)
• Even though the scaling laws are very nice, beware of attributing any of the
propertes of the ‘disc jet’ coupling to specific physical properties of black
hole…
… and there are of course differences between AGN and X-ray binaries
Environment
AGN are in a messy environment that is both:
Extrinsic (wide distribution of temp, density, angular momentum in the fuel
supply – very different to nearly all binaries), and
Intrinsic (broad line region  X-ray binaries don’t launch line-driven winds
from inner disc)
These effects result in obscuration, modification of spectra, and – possibly –
different outburst cycles
Spin ?
AGN and X-ray binaries may have a different distribution of black hole spin
(but NB there is no direct evidence yet that spin strongly affects jet)
THE END
What we don’t know very well
• How fast the jets are
The jets may
be just as
relativistic as
those from
AGN
XRB
AGN
(Jorstad)
Miller-Jones,
Fender & Nakar
(2006)
LRADIO is not particularly fundamental, being less than 10-4 of LJET …
… but we now know
how to calibrate it
to jet power and
accretion rate….
Koerding, Fender &
Migliari (2006)
 So we can calibrate the fundamental plane…
4 x 1024
4 x 1044
1021
1 x 1041
4 x 1017
4 x 1037
1x
(jet power erg/sec)
(mass accretion rate g/sec)
Fundamentaler and fundamentaler…
X-ray binaries
Do we know how the jets are formed ? No
Do we know when (in terms of accretion state) and
how much power ? Yes (approximately)
XRB:AGN similarities… not a new idea
Observed BH mass range 5 MO < MBH< 109 MO
Shakura & Sunyaev (1976) and other disc models realised that
accretion onto black holes might scale in a simple way
Pounds, Done & Osbourne (1995) suggested Seyfert X-ray
emission was like the soft state of galactic BHC
Sams, Eckart & Sunyaev (1996) discussed scaling of BH jets
with mass
Falcke & Biermann (1995, 96, 99…): jet-disc ‘symbiosis’
Mirabel & Rodriguez (1992, 94, 99): ‘microquasar’
Heinz & Sunyaev (2002) calculated detailed scalings for jets
McHardy et al. (2006)
Timing plane
Koerding et al. (2007)
Extended timing
plane – includes
How do the plane and states relate to each other?
For a given mass,
this part results in
the plane…
Introduce a range of masses  a very broad plane
The lack of very large deviations from the plane
indicates transitions to radiatively efficient, jet-quiet
states occurs in the same small range of Eddington
ratios (ie. 0.01 < L < 1) for all black hole mass
.. and this part results
in small deviations
from the plane
What information do we get from this ?
There is a ‘soft state’
which only occurs at
high X-ray luminosity
There is hysteresis
There is a
‘hard state’
in which
the source
begins
and
finishes
the
outburst
Cir X-1
… and, we have observed highly
relativistic (Lorentz factor >10 !!)
jets from neutron stars too (Fender
et al. 2004) …
That was an outburst of the neutron star X-ray binary Aql X-1 … (Maitra & Bailyn 03)
What is the relation
to jet formation ?
Here we see
no jet
Here we
see
major
ejections
jet behaviour (like other
properties) is hysteretical
with luminosity
Here we
see a
steady jet
(LR a LX0.7)
Hard state: Lradio a LX0.7
Apparent tightness
of this correlation
for different sources
probably means G<2
Soft
State
LX (Edd)
Gallo,
Fender
Pooley
(2003) Gallo,Gallo,
Fender
et al. et
(2006)
Gallo,
Fender
&&
Pooley
(2003)
Fender
al. (2006
What is the relation
to the accretion disc
?
In fading soft state
disc cools at or close
to
L
a T4
i.e. a black body with
fixed size
(slightly
controversial!)
As source
fades in the
hard state
the
accretion
disc
recedes
Accretion disc temperature in soft state
Disc T (keV)
(Dunn et al. in prep)
Accretion disc radii as
a function of luminosity
~0.1 Edd
~10-3 Edd
Hysteretical
Zone
~10-5 Edd
Hard
State
~10-7 Edd
Lbol
~ISCO
Cabanac et al.
(in prep)
~100 ISCO
Towards a unified model…
More powerful,
hard sources have
more powerful,
steady jets…
As source softens,
jet velocity increases
abruptly, causing
internal shock in jet
Subsequently,
soft states
show no jet
Only crossing the ‘jet
line’ from hard to soft
makes an outburst !!
Crossing from soft to
hard (e.g.  quiescence)
there is no shock
Faint, hard
source have
steady, G~1
jets
Fender, Belloni & Gallo (2004)
Why we expect black hole accretion to be essentially
scale free:
The extreme mathematical simplicity of black holes:
Physical size scales linearly with black hole mass
M / R is the same (within a factor of a few, depending on spin) for
all black holes – no other object in the universe scales so perfectly.
The only other parameter is spin ( ‘giant elementary particles’)
Why we do not expect black hole accretion to be
essentially scale free:
Microphysics ! The matter at the inner edge of an X-ray binary
accretion disc is much hotter and much denser than that in an
accretion disc around a supermassive black hole… (and who knows
about conditions in magnetic field)
(and certainly neutron star and white dwarf accretion should be
much messier, with solid surfaces, central dipole fields etc)