Polarized Synchrotron Foreground Assessment for CMB Experiments

TL;DR

This study analyzes polarized synchrotron emission across multiple frequencies to map its spectral index variation, aiding in foreground removal for CMB experiments, and highlights the need for improved future surveys.

Contribution

It provides a composite map of the polarized synchrotron spectral index and assesses its spatial variation, addressing data limitations and inconsistencies in existing datasets.

Findings

Spectral index generally between -3.2 and -3 in the Galactic plane

Steepening of spectral index with increasing Galactic latitude

Constant spectral index of -3.25 in the BICEP2/Keck survey region

Abstract

Polarized Galactic synchrotron emission is an undesirable foreground for cosmic microwave background (CMB) experiments observing at frequencies $< 150$ GHz. We perform a combined analysis of observational data at 1.4, 2.3, 23, 30 and 33 GHz to quantify the spatial variation of the polarized synchrotron spectral index, $\beta^{pol}$, on $\sim3.5^\circ$ scales. We compare results from different data combinations to address limitations and inconsistencies present in these public data, and form a composite map of $\beta^{pol}$. Data quality masking leaves 44% sky coverage (73% for $|b|> 45^\circ$). Generally $-3.2 < \beta^{pol} \lesssim -3$ in the inner Galactic plane and spurs, but the Fan Region in the outer Galaxy has a flatter index. We find a clear spectral index steepening with increasing latitude south of the Galactic plane with $\Delta \beta^{pol}=0.4$, and a smaller steepening of $0.25$ in the north. Near the south Galactic pole the polarized synchrotron spectral index is $\beta^{pol} \approx -3.4$. Longitudinal spectral index variations of $\Delta \beta^{pol} \sim 0.1$ about the latitudinal mean are also detected. Within the BICEP2/Keck survey footprint, we find consistency with a constant value, $\beta^{pol} = -3.25 \pm 0.04$ (statistical) $\pm 0.02$ (systematic). We compute a map of the frequency at which synchrotron and thermal dust emission contribute equally to the total polarized foreground. The limitations and inconsistencies among datasets encountered in this work make clear the value of additional independent surveys at multiple frequencies, especially between $10-20$ GHz, provided these surveys have sufficient sensitivity and control of instrumental systematic errors.