Title: AGN Outflows: Part II
1AGN Outflows Part II
- Outflow Generation Mechanisms
- Models and Observations
- Leah Simon
- May 4, 2006
2Review Unified Model
3Review Outflows exist
- BALs (Broad Absorption Lines)
- Large velocity widths V(FWHM) gt 3000 km/s
- Within 60,000km/s of quasar redshift (v 0.2c)
- Variability timescales of year(s)
- Caused by continuum source variability affecting
photoionized clouds - Or caused by cloud (outflow) motion across LOS
- Partial coverage
- Continuum source is small!
- Cloud must be nearby if some continuum source
can pass around cloud to our eye
4Review Acceleration Mechanisms
- Radiation Pressure (Photoionization)
- Line Driving momentum from radiation field
through line opacity - Expect vtransverse small
- Require very high L/LEdd
- Thermal Pressure (Parker Wind)
- Not strong enough
- Requires Isothermal wind...
- Magnetic Pressure (Magnetocentrifugal Driving)
- 'Beads on a string'
- See John Everett (CITA)
5MHD vs LD
- MagnetoHydroDynamics
- Does not necessitate shielding (over-ionization
unimportant) - Expected from collimated radio jets
- Predicts high velocity flows, and can move
high-density gas
- Line Driving
- Requires shield to protect wind from inner x-ray
radiation - UV flux and wind velocities correlate
- Radiative momentum lost from continuum found in
BALs - Can explain relative X-ray and UV flux well
- Predicts high velocity outflows, but maybe
densities too low
6Probably a combination of the the two methods
(Everett 2005, Proga, 2003). Need to constrain
models to distinguish between them!
7Fluid angular-momentum-conservation Not
magneto-centrifugal wind Mass loss through LD
at inner disk (fast stream) through MHD at outer
disk (slow stream)
Proga 2003 simulates MHDLD using both poloidal
and toroidal B-fields Similar to LD, but with
faster (slow) dense wind at outer disk
8Observational Evidence General Results
What's all the buzz?
- CIV width relates to Lxray Proga 2005, Proga
Kallman 2004 - Are UV and and X-ray radiatively coupled?
- X-ray absorption Gallagher et al. 2006
- Hardest X-ray spectra are also weakest
intrinsic absorption? - Shielding and/or Over-ionization Proga, Everett,
Murray et al. 1995 - Line driving requires shielding to protect from
over-ionization - Hot corona?
9Using Gravitational Lensing
- Use multiple LOS to compare structural models for
BLR - Virialized clouds (Kaspi Netzer 1999)
- Continuously outflowing wind ( Murray et al.
1995) - How it works
- observe lensed BALQSOs
- compare 2 observations
- Infer geometry based on
- variation among LOS
D. Chelouche, ApJ 2003
10Gravitational Lensing Results
- Chelouche finds lensed troughs are similar to
within S/N for all but 2 quasars - Single Cloud Model
- lateral size of clouds must be smaller than RS -
expected based on partial coverage - For non-varying clouds, must have lateral to
radial aspect ratio 10-3 - Would be destroyed
on dynamical timescale no coherent acceleration
--NO - Tube model - many (n) identical clouds with
aspect ratio also ltlt 1 - alignment of tube over
numerous LOS unlikely --NO - Clumpy Wind Model
- Cloudlets imply statistical isotropy different
LOS views same distribution variation should
follow Poissonian distribution - similarities imply nv gtgt1 and ntotgtgt100
- changes imply change in cloud distribution
function YES - implies isotropy on few arcsec scale BAL
Outflow probably one or many sheets or cones with
large lateral size not time- dependent
dynamical wind
11Evidence for Multiphase Flows
- de Kool et al. 2001 observe disparate ionization
states at similar velocities-conclude shielded
gas at large distances (1kpc) - Everett et al. 2002 re-evaluate and conclude
multiphase flow, with continuous low-density wind
and embedded high density clouds at small
distances (4pc) - Inner continuous region acts as shield, driven by
MHD or failed LD - Outer region is LD outflow, with lower
ionizations - Lowest ionizations found in dense embedded clouds
? Centrifugally driven disk wind? Turbulence?
Shocks?
12Multiphase Flow in NALs?
- Observe CIV and CII at same velocities
- Initial distance determinations locate SiII very
far from source (150 kpc) - Combine with partial coverage in CIV!
- Could multiphase flow be a solution?
13Variability Test
Approximate Variability Timescales Accretion disk
size .1pc Light crossing time .35
years Viscous time 200 years Dynamical time
0.3 days Using M108Msun, R2x1014 3RS(X-ray
source size)
Observation Separation PKS 2204 13 years Q 0401
7 years PKS 2044 17 years Q 0249 14 years Q
0334 14 years
14Thanks!