Probing quasar accretion discs in anomalous lensed quasars Nick Bate David Floyd, Rachel Webster, Stuart Wyithe The University of Melbourne, Australia October 2nd, Oz Lens 2008
Dec 19, 2015
Probing quasar accretion discs in anomalous
lensed quasarsNick Bate
David Floyd, Rachel Webster, Stuart Wyithe
The University of Melbourne, AustraliaOctober 2nd, Oz Lens 2008
2
Motivation
• Quasar accretion disc models poorly constrained by observations– (Blaes 2007 review)
• Why? Accretion discs are very small!
• Direct imaging not currently possible
Gravitational telescopes!
Lewis 1995
3
Standard microlensing analyses
• Aim to constrain the size of quasar emission regions
• Require monitoring campaigns– (eg: Kochanek 2004; Morgan et al
2007, 2008; Anguita et al 2008; Poindexter et al 2008)
• But: – Expensive in terms of telescope
time– Degeneracy between size and
velocity
• Alternative: lensed quasars displaying a flux ratio anomaly
4
Flux ratio anomalies• Multiply imaged quasars with a pair
of images straddling a critical curve
• Simple theoretical arguments suggest we should see magnification ratio A2/A1 ~ 1, equivalently m ~ 0 (Blandford & Narayan 1986)
• We don’t!• An example: MG 0414+0534
– zsource = 2.64– zlens = 0.96– A2/A1 = 0.45 0.06
(mA1/A2 = -0.9 0.1)(Schechter & Moore 1993)
0 = 0.01 parsec cfa-www.harvard.edu/castles
A2
A1
MG 0414+0534
5
The cause: microlensing
• Schechter & Wambsganss 2002: images at minima (A1) and saddle points (A2) in the time delay surface behave differently when microlensed by a combination of smooth and clumpy matter
• See also: Congdon, Keeton & Osmer 2007; Bate, Wyithe & Webster 2008
Minimum Saddle
m m
pro
bab
ility
6
How do we use this information?
• Probability of observing a flux ratio anomaly depends on source size and smooth matter content of the lens
• Conduct microlensing simulations for a range of source sizes () and smooth matter percentages (s)
• Invert using Bayes’ Theorem:QuickTime™ and a decompressor
are needed to see this picture.
7
Probing accretion discs
• Accretion disc models: longer wavelengths are emitted at larger radii
• Multi-band observations size constraints for multiple emission regions in the source
• We can fit a power-law: QuickTime™ and a
decompressorare needed to see this picture.
source size
wavelength
8
QuickTime™ and a decompressor
are needed to see this picture.
Observational data
Filter r’ i’ z’ J H
central (Å) 6231 7625 9134 12500 16500
A2/A1 0.210.1 0.260.1 0.340.10.60.
20.670.0
5
Filter F110W F205W
central (Å) 11250 20650
A2/A10.640.0
40.830.03
Filter F675W F814W
central (Å) 6714 7940
A2/A10.400.0
10.470.01
Magellan 6.5-m Baade Telescope, 2007 November 3
IMACS and PANIC
HST, 1997 August 14
NICMOS
HST, 1994 November 8
WFPC2
CASTLES Survey: cfa-www.harvard.edu/castles
9
• Use an inverse ray-shooting technique (Kayser, Refsdal & Stabell 1986; Wambsganss, Paczynski & Katz 1990)
• Lens model from Witt, Mao & Schechter 1995
• Generate magnification maps for images A1 and A2, for smooth matter 0% -- 99%
• Microlenses selected from a Salpeter mass function, Mmax/Mmin = 50
Microlensing simulations I
QuickTime™ and a decompressor
are needed to see this picture.
A1, 0%
QuickTime™ and a decompressor
are needed to see this picture.
A2, 0%
QuickTime™ and a decompressor
are needed to see this picture.
A1, 90%
QuickTime™ and a decompressor
are needed to see this picture.
A2, 90%
Minimum
Saddle Point
10
Microlensing simulations II
• Use Gaussian source brightness profiles, characteristic size
• See Mortonson, Schechter and Wambsganss 2005
• Create mock A1 and A2 observations
• Divide A2 magnifications by A1 magnifications
• Result: a library of 25500 Fsim() for each smooth matter percentage
QuickTime™ and a decompressor
are needed to see this picture.
11
Microlensing simulations III
• Go from Fsim() to Fsim() using:
• Thus, we have a library of 25500 Fsim() for each combination of 0, and smooth matter percentage s
• compare with Fobs(), apply Bayes’ Theorem probability distribution for 0, and s
QuickTime™ and a decompressor
are needed to see this picture.
source size
wavelength
12
QuickTime™ and a decompressor
are needed to see this picture.
The accretion disc in 0414
0 = 0.01 parsec for MG 0414+0534• Shakura-Sunyaev disc (1973): R (4/3)
0 = 0.01 parsec for MG 0414+0534
r’-band
size
QuickTime™ and a decompressor
are needed to see this picture.
13
The accretion disc in 0414
• Smooth matter percentage in this lens:unconstrained
• r’-band emission region size (95%):
0 ≤ 1.80 1016 h-1/2 (M/M)1/2 cm
• power-law index (95%):0.77 ≤ ≤ 2.67
• power-law index (68%): 1.05 ≤ ≤ 2.08
• HE 1104-1805, Poindexter et al 2008 (68%):1.18 ≤ ≤ 2.16
14
Conclusions
• Demonstrated a method for probing quasar accretion discs with multi-band imaging– No need for monitoring cheap!– Independent of unknown source transverse velocity
• The accretion disc in MG 0414+0534 (95%):– r’-band emission region ≤ 1.80 1016 cm– 0.77 ≤ ≤ 2.67
• Bate, Floyd, Webster & Wyithe 2008 (MNRAS, accepted)
• More systems to follow