1. Optical/infrared polarised emission in X-ray binaries Dave Russell
New York University Abu Dhabi
In collaboration with Tariq Shahbaz (IAC, Tenerife)
I’d also like to thank Rob Fender, Elena Gallo, Poshak Gandhi, James Miller-Jones
Maria Cristina Baglio, Karri Koljonen, Marion Cadolle Bel, Richard Plotkin
Polarised Emission from Astrophysical Jets, Ierapetra, Crete, 14th June 2017

2. X-ray Binary Jets Black hole XB: GRO J1655-40 Neutron star XB: Sco X-1
Radio emission:  is synchrotron in nature
 collimated outflows (2 types of jet)
The approximate spectrum of a steady, hard state jet:
Optically thick Optically thin
Turnover 
“jet break”
log
log
Radio X-ray?
The power carried in the jets is uncertain
Depends on the position of the “jet break”
How does the jet spectrum evolve during outbursts?  Time evolution (impossible for AGN)
What are the conditions in the inner regions of the jets?  Polarisation 2

3. Can we see the jet in the optical/infrared?
In the last decade or so, evidence shows that:
 the jet is sometimes visible in optical and NIR
 the turnover in the jet spectrum probably lies somewhere in the IR
JET ON
----JET OFF—
JET ON
Data from Homan et al. 2005, Jain et al. 2001, Buxton & Bailyn 2004 3

4. We need polarisation data in the hard state
The radio and optical/infrared emission drops when the source leaves the hard state as jet is quenching (Cadolle Bel et al. 2011)
This happens in every outburst of GX in which there are state transitions (Buxton et al. 2012)
The infrared component is highly variable (Casella et al. 2010, Kalamkar et al. 2016)
2010 outburst:
Soft X-ray
Hard X-ray
Optical
Time resolution
typically ~100 sec
Faulkes Telescope South 4

5. Jet spectral breaks
Jet break found in the black hole XRB GX in the mid-IR and varying rapidly:
WISE data
Gandhi et al identify the jet break in GX 339-4:
In the mid-IR: moving between 4 and 22 microns (probably further)
See also Corbel & Fender (2002), Migliari et al. (2010), Russell et al. (2013)

6. Polarisation of optically thin synchrotron emission
Gliozzi et al. 1998
Some radio data exist:
A few % polarised
(James Miller-Jones’ talk)
Some optical data exist:
A few % polarised due to scattering
(e.g. Dolan, Gliozzi, Baglio)
Very little opt/NIR data exist, but growing field (Shahbaz, Russell, Baglio, Chaty)
Shahbaz et al. 2008
In NIR, the observed emission of X-ray binaries can be highly polarised
Depends on magnetic field configuration
Ordered field  up to ~80% polarised
Tangled field  ~ no net polarisation (low f)

7. VLT observations of GX 339-4 in the hard state
We observed GX during a hard state with VLT+ISAAC
 We detect significant, variable linear polarisation in the near-infrared (when the jet dominated)
Polarisation variability timescale: < 60 sec
Resolved radio jet of GX (Gallo et al. 2004)
We infer a predominantly tangled, variable magnetic field near the jet base: 1 – 3 % polarised
The PA of polarisation is ~ perpendicular to the PA of the resolved radio jet
 The magnetic field is approximately parallel to the jet axis  variability dampens pol.?

8. Time-resolved optical polarimetry of V404 Cyg during its flaring state in 2015 (discrete ejections)
Shahbaz et al. 2016:
On 23 June 2015: time resolved optical polarimetry with the Telescopio Nazionale Galileo (TNG) telescope
2 sec exposure times in r'-band, for 2 hours
Intrinsic polarisation light curve determined by subtracting the interstellar quiescent polarisation level
Variable polarisation: 4.5 to 3.5 % in 20 mins PA is -9 degrees E of N

9. Time-resolved optical polarimetry of V404 Cyg during its flaring state in 2015 (discrete ejections)
Outburst VLBI data show resolved radio jet: PA between ~ -30 and +5 degrees E of N (James Miller-Jones’ talk)
The optical PA implies that the E-field vector near the base of the jet is parallel to the jet axis
B-field is orthogonal to the jet axis
Evidence for magnetic field compression in shocks within the jet, resulting in a partially ordered transverse field

10. Time-resolved optical polarimetry of V404 Cyg during its flaring state in 2015 (discrete ejections)
A huge X-ray flare was observed with INTEGRAL JEM X-1 reaching about 40 Crab
A prominent radio flare at 16 GHz was observed with the AMI, peaking just a few hours after the optical polarisation flare
The optical polarisation flare occurs during the initial stages of the rise of the radio flux

11. Time-resolved optical polarimetry of V404 Cyg during its flaring state in 2015 (discrete ejections)
The 16 GHz and 5 GHz flares (Fender et al. in prep) peak at 2 h and 4 h respectively after the r’-band polarisation flare
We first observe the jet base in polarised light at optical wavelengths, and then the radio flares, which arise from emission along the jet downstream
The three flares describe a classically evolving synchrotron flare from an ejection
It appears we are witnessing an individual ejection evolving from optical to radio wavelengths, over a frequency range spanning 4.5 orders of magnitude

12. Time-resolved optical polarimetry of V404 Cyg during its flaring state in 2015 (discrete ejections)
Lupinov et al. 2016:
Found 2 flares of optical polarisation in V404 Cyg
Changes by ~ 4-6 % in polarisation in 1 hour
Seen when the optical flux was faint, before an optical flare

13. All detections are stronger at low frequencies
Polarisation of neutron star XRBs
4U : M.C. Baglio’s talk
Cyg X-2 and Sco X-1
NIR spectropolarimetry (Shahbaz et al. 2008)
All detections are stronger at low frequencies
The results imply a predominantly tangled, likely variable magnetic field near the jet base
Cyg X-2 has an infrared excess (Wang & Wang 2014)
Sco X-1
NIR (Russell & Fender 2008) and optical (Schultz et al. 2004) polarisation
The radio jet of Cyg X-2 has now been resolved (Spencer et al. 2013)

14. A multiwavelength campaign on Cyg X-2
We took time-resolved NIR polarisation observations with WHT + LIRIS of Cyg X-2, simultaneously with X-ray (Swift and RXTE) in 2010
4.2 m William Herschel Telescope LIRIS 14

15. A multiwavelength campaign on Cyg X-2
We took time-resolved NIR polarisation observations with WHT + LIRIS of Cyg X-2, simultaneously with X-ray (Swift and RXTE) in 2010 15

16. A multiwavelength campaign on Cyg X-2
We took time-resolved NIR polarisation observations with WHT + LIRIS of Cyg X-2, simultaneously with X-ray (Swift and RXTE) in 2010 16

17. BH XRBs in quiescence have jets
Radio IR optical
Swift J1357.2–0933 has a steep IR–optical spectrum, high rms variability (20 – 30%)
Optical, NIR, WISE mid-IR (3.4 to 22 mu)
power-law with index (Shahbaz et al. 2013)
Could be a thermal, possibly Maxwellian distribution of electrons in a weaker jet
Jet break is seen by Plotkin et al. 2016
V404 Cyg has flat radio spectrum
(Gallo et al. 2005, 2007) with instabilities (Rana et al. 2015)
Jets exist in quiescence 17

18. New results from quiescent jets
We took NIR polarisation observations with WHT + LIRIS of Swift J1357.2–0933 in quiescence
4.2 m William Herschel Telescope LIRIS
The synchrotron emission is polarised at a level of 8.0 ± 2.5 % (J to K) (a detection of intrinsic polarisation at the 3.2σ level)
The mean magnitude and rms variability of the flux agree with previous observations (fractional rms of 15–21 per cent)
These properties imply a continuously launched (stable on long timescales), highly variable (on short timescales) jet, which has a moderately tangled magnetic field close to the jet base 18

19. Comparing to polarisation in AGN jets
We need to compare to polarisation measurements from the core in AGN jets
For example:
Lopez-Rodriguez et al. (2014) report polarisation of % at 12 μm and % at 9 μm in Cyg A
This is from the steep spectrum originating near the base of the jet
Above broadband spectrum from Koljonen et al. 2015
Similar levels of polarisation in M87 at these size scales (~103 Rg) 19

20. Intrinsic polarisation Band Reference HARD STATE BLACK HOLE X-RAY
Source
Intrinsic polarisation
Band
Reference
HARD STATE BLACK
HOLE X-RAY
BINARIES:
XTE J
~2.4 % K
Chaty et al. 2011
GRO J
5 – 6 %
H and K
Russell & Fender 2008
GX 339-4
0 – 3% flickering
J - K
Russell et al. 2011
QUIESCENT BLACK
Swift J1357.2–0933
8.0 ± 2.5 %
Russell et al. 2016 A
~0.5 – 1 %
NEUTRON STAR
X-RAY
Sco X-1
0.2 – 1 % flickering
Cyg X-2
2 – 10 %
Shahbaz et al. 2008 4U
2.85 ± 0.96 % r
Baglio et al. 2014
AGN CORES:
GPS / CSS sources
1 – 7 %
Radio
O’Dea 1998
Flat-spectrum AGN,XRBs
0 – 10 %
eg.Lister&Homan 2005
NON-COMPACT JETS
(DISCRETE EJECT I
ONS):
Curran; Brocksopp, etc
X-ray binaries
1 – 50 %
Radio, opt.
James’ talk +V404 Cyg
AGN
10 – 50 %
Opt,radio
e.g.Perlman et al. 2006
GPS / CSS sources (O’Dea 1998)

21. Conclusions X-RAY POLARISATION: COULD BE UP TO A FEW % FROM THE JET?
NIR-optical synchrotron emission from jets in X-ray binaries is polarised
The results so far suggest:
Near the jet base the magnetic field is probably:
 generally turbulent (only partially ordered) and rapidly changing
 (this seems to be similar to AGN jet cores)
 parallel to the jet axis (except in V404 Cyg: perpendicular: shocks?)
Open questions:
 What are the timing properties of the variable polarisation?
 Does polarisation correlate with anything in the inflow?
 What drives the magnetic field changes?
More data and more models are needed to explain the observations
Thanks for listening
X-RAY POLARISATION: COULD BE UP TO A FEW % FROM THE JET?
May be detectable with experiments:
POGOLITE, X-CALIBUR, IXPE
NEED FUTURE MISSIONS!

22. Gamma-ray polarisation detected in Cyg X-1
Polarised γ-ray emission from Cygnus X-1 might be from the jet (Laurent et al. 2011, Science)
Polarisation strength is very high: % !! (0.4-2 MeV)
Derived from 58 days of exposure time with INTEGRAL
This would imply a very highly ordered, constant B field at the base of the jet of Cyg X-1
Jourdain et al confirmed the result using a different instrument on INTEGRAL
% at keV
<20% at keV
(Jet) synchrotron is the only plausible origin
BUT THERE IS A SPANNER IN THE WORKS….

23. Broadband polarisation measurements of Cyg X-1
We observed Cyg X-1 in near-IR with WHT + LIRIS and mid-IR with GTC + CanariCam
No previous IR polarisation measurements published
Its bright: achieved polarisation accuracy of 0.07 % in NIR
The 10.4 m Gran Telescopio Canarias (GTC) on La Palma

24. Broadband polarisation measurements of Cyg X-1
A simple model can reproduce the broadband fractional linear polarisation (FLP) given the input SED (Russell & Shahbaz 2014)
Components:
Self-absorbed synchrotron (radio to IR)
Optically thin synchrotron (IR to gamma-ray) with cut-off in gamma-ray
Max FLP = 82%
Comptonized corona, assumed here to be unpolarised (chaotic corona geometry, no net aligned field?)
We find f ~ 0.9
This implies a stable, highly ordered magnetic field on long timescales
Magnetic field is perpendicular to the jet axis in inner regions of the jet
Needs confirmation from mid-IR or radio!

26. Polarisation of optically thin synchrotron emission
Gliozzi et al. 1998
Some radio data exist:
A few % polarised
(James Miller-Jones’ talk)
Some optical data exist:
A few % polarised due to scattering
(e.g. Dolan, Gliozzi, Baglio)
Very little opt/NIR data exist, but growing field (Shahbaz, Russell, Baglio, Chaty)
Shahbaz et al. 2008
In NIR, the observed emission of X-ray binaries can be highly polarised
Depends on magnetic field configuration
Ordered field  up to ~80% polarised
Tangled field  ~ no net polarisation (low f)