1. Polarisation across the H-R DiagramII
Polarisation across the H-R Diagram
III IV
Daniel Cotton (UNSW)
with
Jeremy Bailey, Lucyna Kedziora-Chudczer, Kimberly Bott & Jonty Marshall V

2. Overview Polarisation Basics Stellar Polarimetry History
Why do people care about polarisation now: Exoplanets!
Polarisation across the H-R Diagram
Close Binary – Spica
Rotational Flattening – Regulus
FGK Dwarf Stars
Debris Disks
Interstellar Polarisation in the Local Hot Bubble (< 75 – 150 pc from Sun).

3. Limb Polarisation p = 0 q r
Light is scattered in the atmosphere of hot stars by electrons, and in cool stars by molecules above the photosphere.
The light reaching an observer from the limb of the star is scattered at 90 degrees and thus is more polarised.
Chandrasekhar was the first to calculate this in 1946.
When light is scattered polarisation is angle dependant. q r
Polarisation is a vector quantity.
Q & U describe photon oscillation orientation.
p = 0
Limb polarisation has been observed for the Sun and Algol*.
(*more on that later)

4. Aperture Polarimetry p = 0 -q r q r
As a result of spherical symmetry, when we look at the starlight within an aperture, there is net zero polarisation.
We see a signal when the symmetry is broken by, for e.g.:
Scattering by particles/molecules in/around a stellar/planetary atmosphere.
Reflection from a surface like Lunar regolith.
Synchrotron generation of light, and magnetic fields.
Scattering by interstellar dust aligned to magnetic fields. -q r q r
p = 0

5. Stellar Polarimetry History
Chandrasekhar (1946) predicted eclipse polarisation. Test of stellar models.
Many people looked for this, but the effect is smaller than predicted.
Only detection so far: Algol (Kemp et al. 1983).
Detections for similar effects in rotationally flattened stars not yet seen.
40s/50s Astronomers surprised to find interstellar polarisation.
Polarisation increases with distance (Hiltner 1949, Hall 1951) due to oblate dust particles aligned to B fields.
Maximum wavelength depends on particle size (Serkowski et al. 1975).
p/pmax
lmax/l

6. HIPPI HIPPI is an optical linear polarimeter.
Two key technologies: FLC modulator, high sensitivity PMTs allow us to beat seeing!
Precision of 4 x 10-6 on bright stars.
Mini HIPPI operating on 14” Celestron 5 mins from Sydney CBD! (Seen here with full sized Jeremy Bailey.)

7. + = Pmax = 0.06 (Rp /a)2 Unpolarised 10-5 polarised 5-10% polarised
+ =
10-5 polarised
5-10% polarised
Pmax for some planets discovered is expected to be around 70 ppm, but the flux from the star also plays a role. The precision of HIPPI is essentially count statistics limited. So, the flux of the star is also important, for around 2 hour observations V=7 gives 5ppm, V=10 gives 20ppm.
F < 10-4 FStar
Pmax = 0.06 (Rp /a)2
Seager et al. (2000)

8. Are stars unpolarised really?
Image Credit: Karen Teraumra, UH IfA

9. Bright Star Surveys
The brightest stars within 100 pc: 50 in the southern sky, 49 in the northern.
Least interstellar polarisation.
Our sensitivity 4x10-6 (Cotton 2016), previous best 7x10-5 (Tinbergen 1982).
Cotton et al., 2015

10. Close Binary – Spica
Primary hypothesis: Proto-stellar material entrained between components leads to asymmetric scattering.
Tkachenko et al. (2016)
Our data.

11. Regulus
Polarisation by rapid rotation. Gravity darkening creates asymmetry.
Polarisation b/c of rotation predicted 70 yrs ago.
B7IV, 86 deg, 86% critical speed.
Sonneborn Model (1982) for 95%.
Data from PlanetPol and HIPPI.
A test of stellar atmosphere models.

12. Starspots & Late Type Stars
HD a BY Dra variable ie starspots:
1% surface coverage = 2-3 ppm effect in polarisation from scattering.
Magnetic Intensification = larger effect?
Total polarisation of the Sun is 2x10-7 (Kemp).
Kostogryz et al. 2014 B P

13. Bright Star Surveys – Debris Disks
Disk systems have higher polarisation.
Disk scatters so that the E-vector is parallel to the plain of the disk at small angles, and perpendicular at larger angles.
Light is more strongly polarised from the limb.
If we observe the whole disk light will be polarised perpendicular to the disk major axis.
Marshall, Cotton et al. in prep., Cotton, Bailey et al
Larger grains around higher luminosity stars. Should give a more redward slope.
Binary disk system e Sgr will have (slowly) varying polarisation.
Cotton, Marshall et al
Clumpy disks may have similar behaviour!
Image: ALMA (NRAO/ESO/NAOJ); C. Brogan, B. Saxton (NRAO/AUI/NSF)

14. Interstellar Polarisation within 100 pc
Using A-K stars only:
The interstellar medium is clumpy.
Interstellar polarisation is greater in the southern hemisphere.
The maximum interstellar polarisation is ~2 x 10-6 pc-1 within ~50 pc.
lmax 315 nm +/- 280 nm
We need better colour information!
When we looked at close bright stars, and excluded those we believed to be intrinsically polarised we found that the interstellar medium is patchy. This however, amounts only to a shifted baseline on our Hot Jupiter polarisation phase curves. The patchiness means that it is difficult to get an independent determination of the baseline polarisation in the system, but that’s something that would be nice, rather than necessary.