Black Hole Accretion

1
(No Transcript)
2
Steady Accretion Solutions(Frank, King Raine
2002)

• Spherical accretion (Bondi 1952) X (no angular
  mmtm)
• Thin accretion disk (Shakura Sunyaev 1973) ?
• Two-Temperature solution (Shapiro, Lightman
  Eardley 1976) X (thermally unstable)
• Advection-Dominated Accretion Flow, ADAF
  (Ichimaru 1977 Narayan Yi 1994, 1995
  Abramowicz et al. 1995) ?

3
Energy Equationq q - qadv

• Thin Accretion Disk
• (Shakura Sunyaev 1973 Novikov Thorne
  1973)
• Most of the viscous heat energy is radiated

• Advection-Dominated Accretion Flow (ADAF)
• (Ichimaru 1977 Narayan Yi 1994, 1995
  Abramowicz et al. 1995)
• Most of the heat energy is retained in the gas

4
Conditions for a Thin Disk

• Two conditions must be met for a thin disk
• The gas must be able to lose its energy to
  radiation in less than an accretion time. This
  requires Mdot ? 10-2 MdotEdd, where MdotEdd
  10-8 (M/M?) M?/yr
• The radiation must be able to escape in less than
  an accretion time. This requires Mdot ? MdotEdd
• If either condition is not satisfied, we have an
  ADAF
• Normally, one discusses electromagnetic
  radiation, but neutrino emission may also be of
  interest (NDAF)

5
Mdot Regimes Thin Disk vs ADAF

• Two regimes for thin disk
• Lower regime corresponds to bright XRBs and AGN
• Upper regime corresponds to SNe and GRBs
• Everywhere else (shaded) we have an ADAF
• Radiation-trapped ADAF (slim disk)
• Radiatively inefficient ADAF (RIAF)

Narayan Quataert (2005) M 3M?
6
Thin Disk, ADAF and BH Spectral States
Esin et al. (1997)
7
Properties of a Thin Disk

• Radiatively efficient ? 0.057 0.42 Luminous
• Geometrically thin H ? R (except near Eddington)
• Inner edge at Innermost Stable Circular Orbit
  (RISCO) ?
• Optically thick --- blackbody-like radiation ?
• Gas strongly bound specific energy ?GM/2R
• Not easy to generate a thermal outflow (but
  line-driven wind possible?) relatively Weak Wind

8
Radiation from Thin Disk

• Most of the accretion luminosity in the universe
  comes from thin disks (AGN)
• Most mass accretion in AGN is also via thin disks
• BH XRBs in high soft state give excellent fit
  with multi-color blackbody spectrum (MCD)
• AGN spectra (Blue Bump) less well-fit --- why?
• Peak of emission (EUV) not observed
• Complex opacity effects
• Radiatively driven winds
• Reprocessing

9
High Soft State of BHXRBs

• L scales as T4
• Multicolor blackbody spectrum fits very well

Gierlinski Done (2002) Kubota et al. (2004)
Grove et al.
10
1995 ASCA data from Ueda et al. (1998)
1997 ASCA data from Yamaoka et al. (2001)
Analysis by Shafee et al. (2005)
11
Where is the Disk Inner Edge?

• Classical thin disk theory zero torque inner bc
  ? well-defined inner edge
• But magnetic fields may exert a torque at inner
  edge (Krolik 1999, Gammie 1999)
• Modifies temperature profile T(R)
• Energy dissipation inside RISCO
• No well-defined inner edge
• Bad news for almost everything!
• Perhaps zero torque is okay when the disk is thin
  (Afshordi Paczynski 2003)
• Consistent with spectra of high soft state BHXRBs

12
Multi Color Disk (MCD) Spectral Models in XSPEC

• DISKBB (Mitsuda et al. 1984) Newtonian model
  very approximate, huge torque at Rin (Gierlinski
  et al. 1999)
• EZDISKBB (Zimmerman et al. 2005) Newtonian model
  with zero torque at Rin
• DISKPN (Gierlinski et al. 1999) Pseudo-Newtonian
  model
• GRAD (Hanawa 1989) Relativistic model for
  non-spinning hole some errors (Li et al. 2005)
• KERRBB (Li et al. 2005) Includes all
  relativistic effects, all BH spins, inclinations,
  adjustable torque

13
Measuring the Spin of an Accreting BH

• Assume accretion disk terminates at RISCO, and
  gas free-falls inside this radius
• Disk emission is dominated by gas near RISCO, so
  observations can be used to estimate this radius
• RISCO/M uniquely gives a
• Several techniques
• Blackbody spectral fit ?
• Iron line fit
• QPO frequency

14
Measuring the Radius of a Star

• Measure the distance D and flux F
• Get luminosity L 4?D2F
• From the spectrum, which is nearly a blackbody,
  get the temperature T of the stellar photosphere
• Then, stellar radius R is obtained from the
  relation L (?T4) (4?R2)

15
BH Spin From Continuum Fitting

• Assume disk emission is locally blackbody-like
• From the accretion luminosity (requires distance,
  disk inclination) and temperature (from
  spectrum), obtain the radius of disk inner edge
  Rin RISCO
• A little more complicated than stellar case since
  T varies with R, but functional form of T(R) is
  known
• From RISCO/(GM/c2), obtain a (Zhang et al. 1997)
• Requires BH mass, distance and disk inclination

16
Spectral Hardening Factor

• Unfortunately, disk emission is not a perfect
  blackbody
• It is approximately a modified blackbody with a
  spectral hardening factor f Tcolor f Teff
• f 1.7 - 0.2 (Shimura Takahara 1995)
• f 1.4 1.65 (Davis et al. 2005)
• Including the effect of f, spectral fit gives the
  ratio RISCO/f1/2(GM/c2)
• Need to know f to solve for RISCO/M and hence a
• But there is a degeneracy between f and RISCO

17
KERRBB Analysis of ASCA Data on GRO J1655-40

• Data give excellent consistency between ASCA GIS2
  and GIS3, as well as with RXTE
• Robust spectral fit with very few parameters
• BUT, the data cannot determine f
• Need an independent theoretical estimate of f to
  solve for BH spin

1997 outburst Data from Yamaoka et al. (2001)
18
Theoretical Estimate of f

• Davis et al. (2005) estimate f using a detailed
  disk atmosphere model
• Their f values are lower than Shimura
  Takaharas (1995) estimates because they include
  metal opacity
• Luminosity-dependent f
• f 1.5 for L/LEdd 0.1
• f 1.6 for L/LEdd 0.3
• Use this to obtain BH spin

19
Spin of GRO J1655-40

• Fit ASCA and RXTE data using KERRBB Good
  consistency between ASCA and RXTE
• Very good consistency among results from
  different epochs
• If constant f 1.5 is used, get trend of a vs l
• Davis et al. model gives slightly more uniform a
• a 0.80.9
• Definitely not 0 or 0.998

GRO J1655-40 ASCA (1995,1997), RXTE
(1997) Shafee, McClintock, Narayan et al. (2005)
20
Spin of 4U 1543-47

• Fit RXTE data (Park et al. 2004) using KERRBB
• Very good consistency among results from
  different observations
• Davis et al. model gives slightly less uniform a
  compared to constant f
• a 0.850.95
• Definitely not 0

4U1543-47 RXTE (2002) Shafee et al. (2005)
21
Assessment of the Spectral Method of Estimating a

• The method demands a lot
• Must know M, D, i (e.g., J1550 poor distance)
• Needs carefully calibrated data
• Needs a believable estimate of f (Davis et al.)
• Given all these, the method is capable of giving
  a robust measurement of a (-0.05)
• Very straightforward method in principle
• Data analysis tools available (KERRBB)

22
Properties of an ADAF

• Radiatively inefficient ? ? 1
• Low-mdot ADAF (RIAF) is very hot (virial), since
  gas loses hardly any energy through radiation
• Large pressure cs vK
• Geometrically thick H/R 1
• Low-mdot ADAF branch exists below critical mdot
  0.01-0.1 (NY95)

23
Thin Disk, ADAF and BH Spectral States
X-ray Binaries different spectral states
identified with combinations of thin disk and
ADAF (Narayan 1996, Esin et al. 1997) AGN
something similar must happen, though there is no
clear correspondence between spectral states in
the two cases Roughly, AGN with Lgt10-1.5LEdd
probably have thin disks, and LLAGN probably have
ADAFs plus outer thin disks Near transition (int.
state), radiative efficiency expected to be high
Esin et al. (1997)
24
AGN with ADAFs
NGC 5548 Chiang Blaes (2003)
Quataert et al. (1999)
Yuan et al. (2003)

• LLAGN do not have a Big Blue Bump --- have a red
  bump instead Suggests an inner ADAF plus an outer
  thin disk transition at rtr 100
• Some Seyferts may be similar, with smaller value
  of rtr ? (high rad. eff.)
• At extremely low luminosities, e.g., Sgr A,
  there is a clear conflict between observed
  luminosity and available Mdot. Here there is
  compelling evidence for radiatively inefficient
  accretion and advection

25
Boundary Between Thin Disk and Low-Mdot ADAF
Yuan Narayan (2004)

• Boundary can be estimated as a function of
  luminosity from observations (Yuan Narayan
  2004)
• With some modeling, can be converted to boundary
  as a function of Mdot
• Consistent with basic theoretical models (NY95)

26
Yuan Narayan (2003)

• Lower boundary from observations coupled with
  modeling (also theory Narayan Yi 1995)
• Upper boundary from theory (slim disk Abramowicz
  et al. 1988)

27
ADAFs, Winds and Jets

• Narayan Yi (1994, Abstract)
• the Bernoulli parameter is positive, implying
  that advection-dominated flows are susceptible to
  producing outflows We suggest that
  advection-dominated accretion may provide an
  explanation for the widespread occurrence of
  outflows and jets in accreting systems
• Blandford Begelman (1999)
• outflows, with or without magnetic fields, can
  be self-collimating and form jets

28
ADAFs Have Strong Winds

• Positive Bernoulli parameter ? Strong Wind
  (Narayan Yi 1994, 1995 Narayan, Kato Honma
  1997)
• Confirmed by numerical simulations (Stone,
  Pringle Begelman 1999 Igumenshchev
  Abramowicz 1999, 2000 Igumenshchev et al. 2000,
  2003 Stone Pringle 2001 Hawley, Balbus
  Stone 2001 Hawley Balbus 2002 McKinney
  Gammie 2004 de Villiers et al. 2005 )
• MdotBH is much less than Mdotsupply (Blandford
  Begelman 1999)

29
Disk Solutions and Jets

• Jet/wind/accretion connection
• Relativistic jet may be
• inner part of wind X
• collimated by wind ?
• Either case, strong wind needed for strong jet
• Thus, strong jets should be mostly associated
  with ADAFs

30
MHD Jet Simulations
Numerical MHD simulations of ADAFs around
rotating BHs produce impressive jets (Koide et
al. 2002 de Villiers et al. 2003 McKinney
Gammie 2004 Komissarov 2004 Semenov et al.
2004 McKinney 2005 )
McKinney (2005)
31
Jet-Wind Link
Take a stable jet simulation (McKinney 2005) and
artificially cool the corona to quench the
coronal wind (McKinney Narayan 2005) Jet is
immediately disrupted within a few dynamical
times Illustrates the importance of a strong wind
for confining the jet
Repeat with Cooling
Original simulation McKinney (2005)
32
BHXRB Spectral States and Jets
Definite pattern of jet activity seen as a
function of spectral state (Fender et al.
2004) Hysteresis present (no good theoretical
explanation yet)

• Text

Narayan (1996) Esin et al. (1997)
Fender, Belloni Gallo (2004)
33
Jet Regimes in AGN

• Radiatively inefficient ADAF present in LINERs
  (Lasota et al. 1996), FRI sources (Reynolds et
  al. 1996 Begelman Celotti 2004), BL Lacs
  (Maraschi Tavecchio 2003), XBONGs (Yuan
  Narayan 2004), quiescent nuclei like Sgr A (Yuan
  et al. 2003, 2004), etc.
• Jets expected and seen ?
• ? L/LEdd
• R 6 cm /B band
• Slim disk ADAF and jets --- which sources? FRII
  sources ??
• Radio-quiet AGN probably have no ADAF, only thin
  disk

Ho (2002)
34
Summary

• Broadly, two distinct kinds of accretion
• Thin accretion disk
• Very luminous bulk of accretion luminosity in
  the universe
• Spectrum often agrees very well with MCD model
• Can estimate BH spin a 0.8--0.9 (J1655-40,
  4U1543-47)
• Very weak wind ? no jet
• ADAF RIAF and slim disk
• BHs spend most of their time in this state
• But do not produce much radiation
• Very strong winds ? can confine jets
• Most jets probably associated with ADAFs
• Quantitative predictions not possible at this time