Transcript Summary - Chandra X

Science Summary
Omer Blaes (UCSB)
Accretion Theory – Simulations (Hawley)
MRI is now nearly 20 years old, and is the only game in town as far as ab initio
treatment of turbulence. But has it yet realized its promise?
• Important to distinguish DYNAMICS from THERMODYNAMICS:
• Dynamics is reasonably well-treated by simulations and there are many results.
• Not (yet!) so much with thermodynamics (energy dissipation and heat transport).
Local (Shearing Box) Results:
Magnetorotational turbulence transports angular momentum robustly with stress
proportional to magnetic pressure. MRI does NOT go away with increasing numerical
resolution! Turbulence level is influenced by viscosity, resistivity, presence or absence
of net vertical field and perhaps other things still not understood.
As yet, no locally generated corona with significant power, but surface layers are strongly
magnetized. (Disk reflection modelers take note: density is NOT constant, nor is material in
hydrostatic equilibrium with thermal pressure.)
No radiation pressure driven thermal instability, because thermal pressure is only correlated
with stress after a time lag (turbulence dissipates which then produces thermal pressure).
Prad>>Pgas
Pgas>>Prad
Radiation dominated branch is thermally STABLE. (“Viscous” instability is still
an open question.)
-Turner (2004); Hirose, Krolik & Blaes (2009); Hirose, Blaes, & Krolik (2009)
Global Simulation Results for Black Hole Accretion (Hawley)
• Collimated, Poynting-flux dominated jets are produced by
Blandford-Znajek mechanism. Spinning black hole threaded
by a dipole field entrained by MRI accretion flow is necessary.
• No diskoseismic modes, but there are spiral acoustic waves.
• Stress is nonzero at ISCO, implying an enhanced accretion
efficiency. Results depend on field geometry.
• Fe Ka line may not be affected “too much” by plunging region.
Observers: be aware that available XSPEC models with simple
ISCO physics must have systematic errors at SOME level. Debate
rages on among theorists as to how big a level that is.
RIAF’s, Sgr A* and M87 (Moscibrodzka)
Moscibrodzka
• Sub-mm polarization based on GRMHD can also give spin (a/M~0.9) as well as axis
direction. (Shcherbakov poster)
• Rotating hot spots for NIR and X-ray flares: NIR polarization also gives spin
magnitude (a/M~0.4-1) and direction. (Eckart)
But WHY should accretion flow angular momentum be aligned with black hole spin axis?
(Bardeen-Petterson alignment probably doesn’t work in this regime [H/R>a].)
More on Black Hole Jet Theory (Tchekhovskoy)
  C
1/ 2
{
C 1
C  20
Confined
Deconfined
Outflows/Winds
Mechanisms (Proga): Thermal, Radiation, Magnetic
-Fast outflow in GRO J1655-40 does indeed appear to be
magnetically driven by process of elimination. Still, a low
density slow thermal wind caused by irradiation heating should
produce a mass loss rate 7 times higher than the accretion rate!
New simulations of large (~10 pc) scale radiation-driven
outflows in AGN (Kurosawa):
3D simulations – clumpy bipolar outflow generated by
cos i dependence of disk emission. Possibly related to NLR
on larger scales?
Also new 2D models with self consistent accretion luminosity
based on what mass is left flowing through inner boundary.
X-ray BAL’s can be observed in lensed objects (Chartas):
-Variability implies wind launched from NEAR the hole. Radiative driving probably
responsible (v/c correlated with X-ray photon index). Requires steep spectra
and/or shielding (what’s the geometry?). Implied energy outflow efficiencies ~1!
Multi-D Radiative Transport (Sim):
3 classes of spectra depending on viewing angle:
Direct continuum plus
reflection (polar).
Weaker continuum plus
reflection plus narrow
absorption lines
(intermediate).
NO transmitted continuum,no
narrow absorption lines,
just complex scattered and
reprocessed photons
(high inclination).
Fe Ka Lines in AGN (Brenneman)
AGN
EW
Rin (rg)
a
q1
Fe/solar
~400
<1.6
>0.98
30±
1
6.0±0.
3
2.0±0.
5
104±1
~130
3.6±0.
4
0.65±0.05
44±
1
5
0.8±0.
2
4±1
~220
3.8±0.
8
0.6±0.2
46±
4
5.3±1.
7
1.5±0.
3
40±35
~1200
<1.3
>0.98
59±
1
6.6±1.
9
>7
50±40
---
3.4±0.
4
0.7±0.1
24±
1
4.4±1.
2
1.2
119±6
6
4
1.5±0.
4
<58
(eV)
MCG—6-30-15
(Brenneman & Reynolds
2006; Miniutti+ 2007)
Fairall 9
(Schmoll+ 2009)
SWIFT J2127
(Miniutti+ 2009)
1H0707
(Fabian+ 2009; Zoghbi+ 2010)
Mrk 79*
(Gallo+ 2010, submitted)
i (°)
NGC 3783**
~85
<2.2
>0.92
24±
(Brenneman+ 2010, in prep.)
3
Superb data which, however, can be challenging to model!
ξ (ergs
cm/s)
Exciting implications: black hole demographics, radio loud/quiet dichotomy, …
Theorists need to work harder to provide observers with more physically motivated models
of hard X-ray emission.
More Reflection/Reprocessing/Absorption
• Reflection in AGN also produces a forest of soft X-ray emission which, when
when relativistically smeared, can fit the RGS spectra of NLS1’s. In future,
such fitting may be able to probe gravitational field further from hole. (Boller)
• New XSPEC models becoming available:
MYTorus: X-ray absorption and reprocessing in a torus geometry (Murphy)
High energy resolution reflection models with newer atomic data (Garcia
poster)
Empirical Attempts to Probe of Nature of AGN Accretion Flow
• SDSS/XMM-Newton Quasar Survey (Young):
473 Sources!
aOX-Lopt real.
Within patchy disk corona model, suggests that covering factor of corona
increases with luminosity.
• Use of BHXRB states to simulate aOX (Sobolewska):
Spread in AGN BH masses gives aOX-L2500 correlation.
Type 1 AGN and NLS1 data overlap with soft state simulations.
LINERS are most probably hard state.
• X-ray binaries and AGN (CHAMP) (Constantin):
X-ray photon spectral index G is negatively correlated with L/Ledd in low
luminosity AGN (i.e. softer when dimmer), opposite to what is seen in QSO’s
- same behavior seen in hard/soft states of BHXRB’s. Agrees with ADAF models.
• X-ray Eclipses (Risaliti):
NH variations on hour-day time scales in 10 Seyferts!
Inferred velocities and distances of absorbing clouds match
BLR, and X-ray source size must be a few gravitational radii.
Clouds have cometary shape.
Is this interpretation of the data unique?
Chandra Imaging of Seyfert Nuclear Regions (Wang)
NGC 4151
Red: HRC 0.1-10 keV
Green: optical [OIII]
Blue: radio (1.4 GHz)
Wang J. et al. 2009, ApJ, 704, 1195
YSO’s
X-ray Irradiation of Protoplanetary Disks (Ercolano):
>1 keV photons penetrate deeply and affect ionization and
therefore MRI and magnetic physics in general (active, dead, and
undead zones). Complex chemistry affected by recombination on
dust grains.
0.1-1 keV photons heat outer layers and drive photo-evaporation.
Finally explains rapid inside-out dispersal of disk after million
year time scale. Warm winds agree with observed emission lines.
(Magnetic driving might also be important, but simpler and
more predictive photo-evaporation wind appears to work just fine.)
YSO’s – Interaction Between Magnetospheric Accretion Stream and Star
• TW Hya and V4046 Sgr (Brickhouse, Kastner)
BEAUTIFUL X-ray line diagnostics distinguish hot coronal emission from accretion
shock structure. Reasonable densities and temperatures inferred, but simple 1D
post-shock settling models do NOT reproduce observations.
• Theory Informed by Solar Corona (Cranmer)
Blobs in accretion stream splash into star, generating waves which shake
magnetic field lines. Alfven wave turbulence on closed field lines cascade and
dissipate – and provide coronal heating at about the right level. Wave pressure on
open field lines can also launch winds, but not quite enough to explain observation?
• MHD Simulations of Accretion Column (Sacco poster)
Both 1D and 2D, complete with spectral line synthesis.
Disk Accretion Onto Magnetized Stars
(Romanova, Kulkarni, Long Poster)
Two Accretion Modes:
Stable (Funnel Streams)
Unstable (Tongues)
Disk Accretion Onto Magnetized Stars – Some Highlights
• Tilted magnetosphere drives large scale, trailing spiral waves in disk.
• MRI is now being included – produces more variability and funnel streams
are more episodic. Variability depends on relative directions between disk
and stellar field.
• Reconnection between disk and stellar magnetic fields, if collisionless, can
accelerate particles to high energies.
• Funnel streams need not rotate with star, causing QPO’s at frequencies
reflecting inner disk rather than stellar spin frequency.
• Conical winds form when matter comes in faster than B-field can diffuse
outward for zero or slow stellar rotation. Rapid rotation can add a Poynting
flux dominated jet.
• Possible scenario for NS LMXB kHz QPO’s (small magnetospheres):
upper QPO due to tongues, lower QPO due to funnel streams.
CV’s
• Magnetic (Mauche):
Chandra HETG spectrum of EX Hya – He-like forbidden lines missing due to
photo-excitation. Fe L-shell lines provide better density diagnostics (MASSIVE
amount of theoretical investment) and give n~1014 cm-3. Radial velocity variations
of X-ray emission lines can measure white dwarf mass!
HETG spectrum of AE Aqr: X-ray plasma is high density and/or close to white
dwarf (“It’s accretion, stupid.”) Radial velocities vary with white dwarf spin phase!
• Non-magnetic (Wheatley):
Kepler data on V344 Lyr – WOW!
In quiescence, X-rays (eclipses imply from WD boundary layer) are 100 times
brighter than disk instability model predicts.
In outburst, hard X-rays are extended – what’s causing them?
Non-magnetic CV’s contribute to Galactic ridge emission.
• Compton cooling may help explain various temperature discrepancies (Mukai)
-plea for boundary layer models including this.
Other Miscellaneous Sources
• X-ray Pulsars (Suzaki):
Detailed fits to pulse profiles consistent with distinct (and occasionally
time-varying) accretion column geometries in different sources.
• Symbiotics:
Mira – Can use accretion disk flickering to DETERMINE (!) nature of companion
from characteristic time scales in innermost disk. It’s a white dwarf, NOT a main
sequence star. (Sokoloski)
Mira AB can be resolved with imaging! Something is focusing the wind to make it
appear to be Roche lobe overflow. (Karovska poster)
Extremely successful Swift program has more than doubled (from 4 to 10) known
hard X-ray symbiotics in last 6 months! Contain massive (1.3 Msun) white dwarfs
– possible SN Ia progenitors. (Luna)
• PSR J1023+0038 (Bogdanov):
Did have an optical disk in 2000-01. If only we had X-ray observations then.
Now a ms pulsar. Possible transition from LMXB to MSP? One of several systems
that we might catch in the act of changing its identity.
Neutron Star LMXB’s
(Z, Atoll, Banana, “Flaring branch” that isn’t really flaring – ugh!
I guess you have to be in this field to master the phenomenology.)
• Bright LMXB’s (Balucinska-Church):
Flaring in Cyg X-2 – like Z-sources: unstable nuclear burning
Flaring in Sco X-1 – like Z-sources: unstable nuclear burning plus
Mdot variations
No flaring in Atoll sources because no unstable nuclear burning.
• Chandra grating spectra (Schulz):
Phase-resolved with orbital period – wealth of detail when
adequately observed! Can determine geometry of ionization layers
(thin on top of disk in 4U 1822-37), disk outer edge (4U 0614+091),
and accretion coronae (extended in Cyg X-2).
Variability (Uttley)
-PSD in AGN looks like Cyg X-1 soft state, and high frequency
break scales inversely with black hole mass and directly with
Eddington ratio.
-2008 discovery of first (and still the only) believable QPO
in AGN: RE J1034+396
-XMM-Newton observations of hard state of GX 339-4: disk
component DOES vary on long time scales, and LEADS variations
in hard photon energies. At high frequencies, disk LAGS variations
in hard photon energies (reverberation – finally!).
X-Ray Binaries
Sivakoff: X-ray Binaries in Cen A
Most of the point sources (90%) turn out not to be transients. Persistent sources
also dominate X-ray binary population in other galaxies.
Homan: BHXRB State Transitions
Mdot does not uniquely determine spectral state at high luminosities.
Peak luminosity is determined by hard to soft transition luminosity.
Brighter sources show flaring, resulting in additional structure in hardness-intensity
diagram. Associated with steep power law state. This and intermediate “state” have
HFQPO’s. In latter state, they appear when rms noise drops in hardsoft transition,
near time (~few days) of radio flare.
Type C LFQPO frequency related to hardness. Type B LFQPO frequency related
to luminosity.
Winds appear to be more associated with softer states.
Sub-Eddington neutron star XRB hardness-intensity diagrams are similar to those
of black holes. There now exists a single neutron star that spans BH-like HID’s to
Sco X-1-like and Cyg X-2-like Z-source HID’s.
Malzac: Cyg X-1 Corona and Jet
BELM – a new code that self-consistently handles evolution of lepton and photon
distributions, including synchrotron emission and absorption (thermalization by
synchrotron boiler).
Proton temperature must be low compared to hot two-temperature accretion flow
models of hard state.
Strongly magnetized plasma (Pmag>Pgas) might produce a self-consistent hot plasma
for hard state. But nonthermal high energy excess requires a weak B-field. I.e. must
have two spatially separated regions.
Nowak: Cyg X-1 hard state in all (6) X-ray satellites – simultaneously!
Modeling requires accounting for absorption by blobs in accretion stream and dust
scattering halo.
VERY broad (0.5-500 keV) spectrum, with very gratifying agreement between all
instruments!
Broad (as well as narrow) Fe Ka line is present. Required inner radius varies between
6M and 40M, depending on continuum model. (Multiple continuum models adequately
fit spectrum, including one involving a condensation in inner hot flow that acts as a
seed photon source.)
Lots of interesting science that touches on
all of astrophysics, from planet formation to
galaxies to cosmology.
Simulations are getting more powerful and
including more and more of the relevant
physics.
New missions (Smith): Astro-H, GEMS, and
(I hope!) IXO, AXTAR,…