New constraints on accretion

from fast optical and X-ray

timing of X-ray binaries

Poshak Gandhi (RIKEN)

Martin DurantAndy FabianJulien Malzac

Jon MillerTariq Shahbaz

Kazuo MakishimaVik DhillonTom MarshHenk Spruit

What is the source of the optical power of X-ray binaries?

GX 339-4 ESO (2008) / L. Calçada

Optical and X-ray simultaneous timing

u’

(Gandhi+10 submitted)

(CC

F)

2.5 s

GX 339-4 : 2007 low/hard state

+VLT RXTE

PCA

New fast red component in CCF?

r’

g’

u’

u’

g’

r’

(Gandhi+10 submitted)

(CC

F)

2.5 s

GX 339-4 : 2007 low/hard state

+VLT RXTE

PCA

Sub-second X-O Cross Correlation Function (CCF)‏

Avera

ge

Optical Lag (seconds) (Gandhi+ 2008)

150 ms

50 ms

r’r’r’

GX 339-4 : 2007 low/hard state

+VLT RXTE

PCA

ULTRACAM:ultra-fast, triple-beam CCD camera

High-speed light-weight camera

Frame-transfer CCDs withnegligible dead-time

Speeds ~ 500 frames / sec

3 simultaneous optical filters

Absolute timing ~ 1 ms

http://www.shef.ac.uk/physics/people/vdhillon/ultracam/

ug

r / i / z

ULT

RA

CA

M @

VLT

Speedy optical variability

Swift J1753.5-0127

GX 339-4

de

ΔT

=50 m

sΔ

T=

39 m

sU

LT

RA

CA

M

Anim

atio

ns: b

insim

Speedy optical variations in X-ray binaries

(XTE J1118+480: Kanbach et al. 2001)‏

(V404 Cyg: Uemura+02)

(GX 339-4: Motch+82)‏

mag

1s

0.5 s

Swift J1753.5-0127: Durant et al. 2008

XTE J1118+480:Kanbach et al. 2001‏

GX 339-4: Gandhi et al . 2008

Asymmetric and complex optical/X-ray cross-correlation functions in the low/hard state

Rapid variability origin?:

Simple reprocessing models

All X-ray data: RXTE PCA

1. Small time delayA

vera

ge

Optical Lag (seconds) (Gandhi+ 2008)

150 ms

Swift J1753.5-0127: Durant et al. 2008

XTE J1118+480:Kanbach et al. 2001‏

GX 339-4: Gandhi et al . 2008

2. Anti-correlation

All X-ray data: RXTE PCA

3. Small optical coherence times

X-rays (PCA)

Optical‏(r’,‏g’,‏u’)‏

Optical ACFs narrower than X-ray ACF for two sources

Lag (s)

Auto Correlation Function(Poisson corrected)

Optical

(white)

X-rays (PCA)

GX 339-4 XTE J1118+480

How much reprocessing is expected?(v

an P

arad

ijs&

McC

linto

ck 1

99

4)

All sources in low/hard state with

LX/LEdd<~0.01

(Gan

dh

i+2

01

0, D

ura

nt+

10

)

Speedy optical variability: scales and power

• Fastest flares have timescales <~ 20-50 ms.

=> R <~ 104 km <~ 103 RG

• Brightness temperature

=> TB >~ 107 K

Models for XTE J1118+480

“The physical origin of the

variability

Is likely to be complicated.”

• Merloni+00 Magnetic corona

• Esin+01 Advection flow (ADAF)

• Markoff+01 Pure jet

• Malzac+04 Magnetic‏‘reservoir’‏jet/corona

• Yuan+05 ADAF+jet

• …

Optical (cyclo)synchrotron.

Interaction between components.

CCFs characteristic times =>interacting components

150 ms

X-ray coronal flares lead to reordering of jet poloidal field

(~ tens x tdynamical ~ 100 ms; Livio+03; Malzac+04)

Ima

ge cred

it: Pu

nsly+

09

(Gandhi+2008)

CCFs characteristic times =>interacting components

150 ms

Jet /

Outflow

Schem

atic ca

rtoo

n

Disk

Propagation of perturbations : from disc/corona to jet

(150 ms 50 RG @ Alvenic speed, or 5000 RG @ lightspeed)

(Gan

dh

i+2

00

8, 2

01

0)

B field dissipation anti-correlation

150 ms

Jet /

Outflow

Schem

atic ca

rtoo

n

Disk

Release of coronal B energy density

optical cyclosynchrotron

(Gandhi+2008)

Optical power spectra

reminiscent of X-rays

(Gandhi 2009)

X-ray: PCAOptical: Either

r’ or g’ or white

Broad band optical noiseand QPO

r’

g’

u’

X-ray PCA

(Gandhi+10 submitted)GX 339-4 : 2007 low/hard state

Optical QPO quite “stable” in GX 339-4

(Motch+83)white light

(Motch+83)Ariel data

(Steiman-Cameron+97)

(Gandhi+10): PCA(Gandhi+10):

r’, g’

(Motch+85)

Optical X-ray

Fre

q x

Po

wer

(arb

itra

ry)

cf. also Uttley talk: 100s QPO

Anything

special

about

these

binaries?

Swift J1753.5-0127

XTE J1118+480

GX 339-4

V4641 Sgr

(J. O

ros

z)

(Durant+10, in prep)

X-r

ay :

Op

tica

l Cro

ss-C

orr

ela

tio

n F

un

ctio

n (

CC

F)

Optical Delay (s)

Neutron star binaries?

Fast optical timing constraints• Optical timing complementary constraints to X-

ray timing.

(Gandhi 2009)

Flux-dependent variability

(Gandhi 2009, Uttley et al. 2001…2005)

R.M.S. variability amplitude increases with flux=> Argues against additive shot noise

Fast optical timing constraints• Optical timing complementary constraints to X-

ray timing.

• How “close in” can optical timing take us?

(Gandhi 2009)

Summary

• Rapid optical flaring in

low/hard state observations

of several binaries.

• Optical not reprocessed simply.

• Complex CCF

=> jet/corona/disk interaction.

• Optical timing gives

independent constraints on

inner accretion processes.