From CMB-S4 wiki
Jump to: navigation, search

HI-based Dust Polarization model (HiDPol)

T. Ghosh writing


We construct a HI-based Dust Polarization (HiDPol) model that incorporates HI column density maps as tracers of the dust intensity structures and a phenomenological description of the Galactic magnetic field (GMF). The two main ingredients of the model are: (a) line of sight (LOS) depolarization arising from fluctuations of the GMF orientation and (b) alignment of the filamentary structures in the cold neutral medium with the magnetic field at high Galactic latitude. By adjusting the parameters of the dust model, we are able to reproduce the statistical properties of the dust polarization over a selected low column density region comprising 34% of the southern Galactic cap (b \le -30 deg) or SGC34 region. Realistic simulations of the polarized dust emission enabled by such a dust model are useful for testing the accuracy of component separation methods, and studying non-Gaussianity. SGC34 covers 3500 deg^2 region containing the cleanest sky accessible to ongoing ground and ballon-based CMB experiments. Unlike PSM, this dust model is not noise-limited in low column density regions over the southern Galactic cap.

Please refer to ArXiv version of the paper (https://arxiv.org/abs/1611.02418) or A&A accepted version (http://www.aanda.org/component/article?access=doi&doi=10.1051/0004-6361/201629829) for more details. The same dust model at 353 GHz is extended to other microwave frequencies, assuming a modified blackbody spectrum for each of the three HI layers: cold neutral medium (CNM), thermally unstable neutral medium (UNM) and warm neutral medium (WNM). The mean dust spectral index (\beta_d=1.59) and temperature (T_d=19.6 K) are fixed for three HI layers, but we introduce uncorrelated beta_d fluctuations around the mean value.


Our dust model is able to match the following statistical properties of the dust polarization over the SGC34 region:

a) observed distribution of the dust polarization fraction at 1\deg resolution.

b) dispersion of the dust polarization angle around the mean GMF at 353 GHz.

c) power spectrum of the dust polarization D_\ell (TE), D_\ell (EE), and D_\ell (BB) at 353 GHz .

d) observed TE correlation and E-B asymmetry.

e) includes dust decorrelation where the dust polarization angle changes slightly with frequency.


Dust model.png

Left Column: Orthographic projections of the Planck 353 GHz intensity map (top row) and the square of the polarization intensity P^2 (bottom row) at 1 deg resolution over the southern Galactic cap. Right column is similar to left column, but for a given realization of the HiDPol model at 353 GHz. The black contour highlights the SGC34 region where the model parameters are fitted. The HiDPol model produces noiseless dust simulations.

Decorrelation dust.png

The dust polarization angle changes slightly with frequency as we go from 353 to 70 GHz in the BICEP2 field. The background images show the filtered dust intensity maps at 353 and 70 GHz for the HiDPol model. Each bar represents the dust polarization angle averaged over 10x4 pixels.

Correlation between different dust realizations:

The total GMF is a vector sim of fixed mean large-scale and a random turbulent component. Because the model of the turbulent GMF is statistical, we can simulate multiple dust sky realizations for a given set of model parameters. As the three HI layers are fixed and a perfect alignment of the CNM structures with the magnetic field, the HiDPol simulated maps are highly correlated between different realizations at a reference frequency 353 GHz. For other frequencies, the correlation decreases linearly with increasing multipoles due to the introduction of the uncorrelated dust spectral index map.