Dust polarization expectations The PILOT experiment J.-Ph. Bernard CESR Toulouse Dust polarization at long wavelengths J.-Ph. Bernard, Orsay, Bpol meeting.

• Dust polarization expectations• The PILOT experiment

J.-Ph. Bernard CESR Toulouse

Dust polarization at long wavelengths

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

° Measured in absorption in the visible

Matthewson & Ford 1970

° Measured in emission in the FIR/mm(ground, Archeops, Boomerang)

850 m OMC-3 in Orion A

Dust Polarization

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

UV extinction UV rise (PAH) is not polarized

Bump (VSG) is not polarized

IR extinction

- Visible extinction (BG) is polarized- 10 m absorption feature is polarized- 3.4 m ice feature is polarized

So, small (Carbonaceous) grain emission is not expected to be polarized while large (Silicate) grains is.

Lessons from extinction

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006Desert et al 1990

cold dust (silicate ?)cold dust (silicate ?)

warm dust (graphite ?)warm dust (graphite ?)

Model by Finkbeiner, Davis, Schlegel (FDS)mm excess:- FIRAS data show =1.7near the Galactic plane- WMAP data show thatthis extends down to >2-3 mm

Complications ...

T1=9.6 K, 1=1.5T2=16.4 K, 2=2.6

100 m 1 mm 10 mm

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

cold dust (silicate ?)cold dust (silicate ?)

warm dust (graphite ?)warm dust (graphite ?)

Model by Finkbeiner, Davis, Schlegel (FDS)mm excess:- FIRAS data show =1.7near the Galactic plane- WMAP data show thatthis extends down to >2-3 mm

Complications ...

T1=9.6 K, 1=1.5T2=16.4 K, 2=2.6

Would imply a polarizationdegree highly variablewith wavelength

X-foreground = magnetic dipoleemission from vibrating dust

X-foreground =spinning small dust

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

Alternative ModelsMeny et al. in prep.

Two-Level Systems

In principle, such models predict no large change of polarization with wavelength

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

DCD+TLS Model

€

T =17.9 ± 0.22 K

lc = 2.6 ± 0.17 nm

Ahop = 63.8 ±1.3

Ares =1.06 ± 0.4

DCD

TLS: hoping

TLS: resonant

FDSFDS

TLSTLSBest fit:

Alternative Models

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

From Vaillancourt et al.

Dust polarization certainly varies with in some (dense to very dense) objects ...

FIR/submm Observations ?

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

... but the observed variations match none of the models.Indeed, they could be entirely due to confusion on line-of-sight, so ..We have no clue if it varies in the diffuse ISM

From Vaillancourt et al.

FIR/submm Observations ?

TLS Hoping

TLS Res.DCD

Planck will be very sensitive to the shape of the dust emissivity and therefore, should allow to select a dust model

Dust Temperature is an important parameter. For B-pol, we may consider measuringit (with B-pol itself, or otherwise) in order to understand dust foreground.

FIR/submm Planck

Predictions of TLS model with dust Temperature

Planck: unpolarized

Planck: polarized

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

From Pelkonen et al.

For a resolution of 15'

0.5 mag

Planck sensitivity to dust polar

FIR/submm Planck

Planck should allow to detect -variations of the dust polarization(at 3- every 1° below Av=0.5 mag).

Detecting more precisely smaller variations toward more diffuse clouds may require using a polarized instrument at higher frequency like PILOT

Planck: unpolarized

Planck: polarized

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

Planck: unpolarized

PILOT

Planck: polarized

PILOT vs Planck -coverage

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

- Today, we don't really know if dust polarization will vary with in the sub-mm, and if yes, how ...- So, it is difficult to devise how well we will be able to subtract it from CMB channels for B-pol (and even Planck, actually ...).- WMAP has only one dust-dominated channel (90GHz). We could compare it to Archeops polarization at 353 GHz to start constraining variations.- Planck (and PILOT) should answer the question ... but actual life could be significantly more complicated than described here.

° Other open questions: + Is there l-variations appart from dust distribution ? + Is there a contribution from spinning dust ? + Is there a contribution from HVCs ? + Is there a contribution from ZL ?

J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

Conclusion (Dust polarization)

The END

PILOT is the pilot fishFuture CMB-pol mission ?

PILOT

http://pilot.cesr.fr J.-Ph. Bernard, Orsay, Bpol meeting 2006, Oct26th 2006

• PILOT will attempt at measuring the linear polarization from dust emission in our Galaxy to very high accuracy levels at high frequencies• The goal is to understand the physics of dust polarization, to ultimately help with component separation for B-pol• PILOT is a test-bed for the use of compact multiplexed bolometer arrays for polarization measurements• PILOT is funded by CNES. Detailed definition is ending and realization of the instrument is starting. • First flight expected in end-2008 (Kiruna)