Linear Polarization of the 21 cm Line from the Epoch of Reionization

TL;DR

This study revisits the 21 cm linear polarization during reionization, incorporating nonlinear effects and realistic simulations, revealing complex scale-dependent features and correlations with ionization, but highlighting observational challenges.

Contribution

It provides a more accurate estimation of 21 cm polarization power spectra during reionization by including nonlinear effects and realistic simulations, improving upon previous toy models.

Findings

Power spectra are enhanced on sub-bubble scales and suppressed on super-bubble scales.

The TE power spectrum shows a zero-crossing at low multipoles, sensitive to H I bias.

Ratios of power spectra to their maxima correlate with the global ionized fraction.

Abstract

The 21 cm linear polarization due to Thomson scattering off free electrons can probe the distribution of neutral hydrogen in the intergalactic medium during the epoch of reionization, complementary to the 21 cm temperature fluctuations. Previous study (Babich & Loeb 2005) estimated the strength of polarization with a toy model and claimed that it can be detected with 1-month observation of the Square Kilometre Array (SKA). Here we revisit this investigation with account of nonlinear terms due to inhomogeneous reionization, using seminumerical reionization simulations to provide the realistic estimation of the 21 cm TE and EE angular power spectra ($C^{\rm TE}_\ell$ and $C^{\rm EE}_\ell$). We find that (1) both power spectra are enhanced on sub-bubble scales but suppressed on super-bubble scales, compared with previous results; (2) $C^{\rm TE}_\ell$ displays a zero-crossing at $\ell<100$, and its angular scale is sensitive to the scale-dependence of H I bias on large scales; (3) the ratios of the power spectrum to its maximum value during reionization at a given $\ell$, i.e. $C^{\rm TE}_\ell / C^{\rm TE}_{\ell,{\rm max}} $ and $C^{\rm EE}_{\ell}/C^{\rm EE}_{\ell,{\rm max}}$, show robust correlations with the global ionized fraction. However, measurement of this signal will be very challenging not only because the overall strength is weaker than the sensitivity of SKA, but also because of the polarized foregrounds from diffuse synchrotron emission, and Faraday rotation which modifies the observed polarization. Nevertheless, the 21 cm linear polarization signal may still likely be detectable through other approaches, e.g. its cross-correlation with other probes.