SUPERMASSIVE
BLACK HOLES:
ACCRETION AND
OUTFLOWS
Mitch Begelman
JILA, University of Colorado
COLLABORATORS:
• Roger Blandford (KIPAC)
• Mateusz Ruszkowski (JILA/MPA)
• Marcus Brüggen (IU Bremen)
• Eric Hallman (CASA, U. Colo.)
 WHY ARE OUTFLOWS INEVITABLE?
 HOW DO SMBHs AFFECT THEIR
SURROUNDINGS?
 WHY ARE OUTFLOWS INEVITABLE?
 HOW DO SMBHs AFFECT THEIR
SURROUNDINGS?
DISKS PRODUCE JETS/WINDS
BECAUSE THEY HAVE TROUBLE
RADIATING AWAY ENOUGH OF
THE LIBERATED ENERGY
TWO KEY POINTS:
1. ENERGY, NOT ANGULAR
MOMENTUM, IS THE CRUCIAL
FACTOR
2. ENERGY ARGUMENT IS GENERIC,
BUT THE DETAILS ARE DIVERSE
AND SPECIFIC
The role of angular momentum is that
it must be transported outward – this
can be done directly in a wind, or
internally through torques
TORQUE TRANSPORTS ENERGY
Angular Momentum Flux:
G
Torque G ~ M  outward
Energy Flux:
G outward
IN A THIN ACCRETION DISK:
Local rate of energy release:
G
GM

M
2R
Local rate of dissipation:
  GM 
3  M

2R 

2/3 of energy dissipated at R transported
from <R by viscous torque
IN A RADIATIVELY INEFFICIENT
ACCRETION FLOW (RIAF):
Energy Transport:

G ~ M B
G
2
v
B h 0
2
Bernoulli Function
Energy transport from small R by torque unbinds
gas at large R unless radiative efficiency > 2/3
RIAFs are EXPLOSIVE
1 g of gas accreting at r ~ m
can liberate 1 kg of gas at r ~ 1000 m*
• Torque a “conveyor belt” for liberated energy
• Flow must lose energy OR limit accretion
– Mass loss or circulation
– Small fraction of supplied mass reaches BH
ADIOS =
ADIABATIC INFLOW-OUTFLOW
SOLUTION (Blandford & Begelman 99)
3 effects contribute to radiative
inefficiency:
3 effects contribute to radiative
inefficiency:
• High accretion rate
– gas highly opaque, radiates but photons can’t escape
M  M Eddington
3 effects contribute to radiative
inefficiency:
• High accretion rate
– gas highly opaque, radiates but photons can’t escape
M  M Eddington
• Low accretion rate
• Dissipated energy goes mainly into protons
If:
• Protons-electron thermal coupling weak
– gas is tenuous, falls into BH before radiating
2 

M   M Eddington;   1
3 effects contribute to radiative
inefficiency:
• High accretion rate
– gas highly opaque, radiates but photons can’t escape
M  M Eddington
• Low accretion rate
• Dissipated energy goes mainly into protons
• Protons-electron thermal coupling weak
– gas is tenuous, falls into BH before radiating
2 

M   M Eddington;   1
• Everything in-between?
– Coronae are common, rad. eff. is a function of z
Energy argument is generic…
PHYSICALLY, WHAT DRIVES THE
OUTFLOW or CIRCULATION?
• Candidate mechanisms:
– magnetic torques, flares - relativistic jets
– radiation pressure - BALs, SS433
– radiative heating
• Convection = “minimal” mechanism
– all else being equal, adiabatic flows must be
convectively unstable
ADIOS behavior in MHD...
Hawley & Balbus 02
2-D, adiabatic α-model
M in
M out
M acc  M in  M out
Stone, Pringle & Begelman 99
2 complications to this picture…
IF TORQUE IS EXTERNAL…
HYDROMAGNETIC WIND removes energy and
angular momentum with no mass loss and negligible
entropy generation.
Blandford & Payne 82
-d EW =  dLW
IF TORQUE IS EXTERNAL…
HYDROMAGNETIC WIND removes energy and
angular momentum with no mass loss and negligible
entropy generation.
Blandford & Payne 82
-d EW =  dLW
 close to supplied
…accreted M
even larger wind power
M
BLACK HOLE SPIN:
THE “OTHER” ENERGY SOURCE?

• BH spin extractible magnetically
– up to 0.29 M available
• Field could connect ergosphere to disk, wind


50rg
Amplification
of B in plunge
region
Reynolds et al. 2006
Weak mag. diffusion :
Pm  20
Magnetic Prandtl no. :
Pm   / 
h/r=0.1
ISCO
15 pc
SgrA* Bipolar
X-Ray Lobes
WEAK SOURCE: ADIOS?
M. Morris et al. 2003
Energy range
3.3 - 4.7 keV
Adaptively smoothed,
point-sources removed
Lobes are centered on the black hole,
and perpendicular to the MW plane
M. Morris et al. 2003
There is also a Sgr A* Jet...
~ 1 pc
F. Baganoff et al. 2003
There is also a Sgr A* Jet...
~ 1 pc
F. Baganoff et al. 2003
M87 RADIO
MONTAGE
F. Owen (NRAO),
J. Biretta (STScI),
J. Eilek (NMIMT)
1999
Jet already exists at
R  100m
INTERMEDIATE
POWER: JET
POWERED BY BH
SPIN OR INNER
DISK?
www.nrao.edu/~fowen/M87.html
BAL Outflows in X-ray Absorption
...driven by
radiation pressure?
NH ~ 1023 cm-2
v ~ 0.2 c
 ~ 4p/10
LKE
 rabs 
~ 0.1 17 
LBol
 10 cm 
QUASAR:
POWERFUL
OUTFLOW
PG1115
Chartas et al. 2003
 WHY ARE OUTFLOWS INEVITABLE?
 HOW DO SMBHs AFFECT THEIR
SURROUNDINGS?
Jets inject kinetic energy into
their surroundings, but…
…energy use depends on…
• Rate of deposition
– power L vs. total output  KE Mc 2
– intermittency?
• Importance of cooling
• Uniformity of gas
– Clumpy: energy “goes around” the clumps
– Gas in a disk may be immune
• Directionality of outflow
– Collimated jet vs. broad disk wind
Why is “porosity” important?
McKee & Ostriker 77 (ISM theory)
Outflow only interacts with
diffuse medium
blasts out supersonically
but has little effect on most of
the mass
Directionality: analogous effect
– how can a powerful jet source
continue to accrete?
THE INTRACLUSTER MEDIUM
A CASE STUDY
 GAS RELATIVELY SMOOTH
 COOLING TIME RELATIVELY LONG
(COMPARED TO DYNAMICAL TIME)
VIOLENT HEATING IS FAMILIAR…
What does AGN
look
like?
CYGNUSheating
A in radio
(VLA,
6cm)
Energy injection concentrated into “hotspots”
Elongated structure (momentum-dominated)
CYGNUS A – Exception rather than rule?
undisturbed
intergalactic gas
bow shock
“cocoon” (shocked
jet gas)
splash point
backflow
Very powerful source in relatively tenuous ICM
Most powerful RGs this age have: ~spherical cocoons
subsonic expansion
3 STAGES of ENERGY INJECTION:
• MOMENTUM-DRIVEN
– ELONGATED COCOON
SUPERSONIC,
OVERPRESSURED
• ENERGY-DRIVEN
– SPHERICAL BLAST WAVE
• BUOYANCY-DRIVEN
– RISING BUBBLES
SUBSONIC
Cygnus A in X-rays (Chandra) – entering energy-driven
phase
3 STAGES of ENERGY INJECTION:
• MOMENTUM-DRIVEN
SUPERSONIC,
OVERPRESSURED
• ENERGY-DRIVEN
• BUOYANCY-DRIVEN
SUBSONIC
MORPHOLOGICAL EVIDENCE
OF GENTLE HEATING
• Radio-filled bubbles and “ghost
cavities”
• Ripples
• Plumes, eddies - connected to AGN
3C 84 and
Perseus Cluster
Fabian et al. 2000
3C 84 and
Perseus Cluster
Fabian et al. 2000
Radio emitting
plasma fills holes
Chandra
image +
radio
BUBBLES
3C 84 and
Perseus Cluster
Fabian et al. 2000
Multiple “fossil”
bubbles(?), not
aligned with
current radio jets
Chandra
image
3C 84 and Perseus Cluster
Fabian et al. 2000
Cool rims
Chandra
image
RIPPLES: PERSEUS CLUSTER
Unsharp masked Chandra
image
X-ray temperatures
131 kpc
Fabian et al. 2006
RIPPLES: VIRGO CLUSTER
Forman et al. 2004
Two types of radio sources: FRI vs FRII
PLUMES AND EDDIES
M87
Cygnus A
Carilli
Cygnus A
Young
Forman
M87 RADIO
HALO
80 kpc
Owen, Eilek & Kassim
1999
VLA, 327
MHz
www.nrao.edu/~fowen/M87.html
M87 + core
of Virgo
cluster
Chandra image +
radio
A. Young et al. 2002
THE ICM NEEDS GENTLE,
WIDESPREAD HEATING TO…
• OFFSET “COOLING FLOWS”
• EXPLAIN ENTROPY EXCESS IN
CLUSTERS AND GROUPS
SIMULATIONS:
• Crucial to model mixing and weak shocks
accurately
– Use PPM code with Adaptive Mesh Refinement, e.g.,
FLASH
FLASH 2D, 8 levels of mesh refinement
Current 3D dynamic range: 1000:1 (FLASH), 33,000:1 (ENZO)
DENSITY
Ruszkowski, Brüggen & Begelman 2004
DISSIPATION RATE
Ripples in the Perseus Cluster
Unsharp masked Chandra
image
Viscous dissipation map:
pulsed source
131 kpc
Fabian et al. 2003
Ruszkowski, Brüggen & Begelman 2004
cluster extracted from a full
COSMOLOGICAL SIMULATION
~2.5Mpc
Evolved for ~ 250 Myr
~400kpc
3D Heating rate/ Cooling rate
Periodic behavior reflects periodicity of the source
O(1)
Different annuli (10 kpc)
Weakly nonlinear: Mach no. ~ 1.3
Ruszkowski, Brϋggen & Begelman 2004
CONCLUSIONS I:
THE EXISTENCE OF OUTFLOWS
FROM ACCRETING BLACK HOLES
IS GENERIC
BUT THE DETAILS ARE DIVERSE
AND SPECIFIC
CONCLUSIONS II:
SMBH OUTFLOWS CAN STRONGLY
AFFECT THEIR ENVIRONMENTS
SURPRISINGLY, THEY CAN PROVIDE
WIDESPREAD, GENTLE HEATING OF
INTRACLUSTER GAS
Rough consistency with entropy floor,
cooling flows, morphology ….