# The bispectrum of polarized galactic foregrounds

## TL;DR

This paper analyzes the bispectrum of galactic dust and synchrotron emissions across multiple frequencies, revealing strong non-Gaussian features that are crucial for accurate CMB studies and foreground mitigation.

## Contribution

It provides the first comprehensive measurement of the galactic foreground bispectrum, including polarization and parity-odd components, across a wide frequency range.

## Key findings

- Strong temperature bispectra in dust and synchrotron emissions, especially in the squeezed limit.
- Detection of significant polarized bispectra in galactic dust, peaking in the squeezed configuration.
- No residual foreground bispectra found in Planck component separated maps after masking bright sources.

## Abstract

Understanding the properties of the galactic emission at millimetre wavelengths is important for studies of the cosmic microwave background (CMB). In this work we explore the bispectrum, the harmonic equivalent of the three point function, from galactic dust and synchrotron emission. We investigate these effects across a broad range of frequencies using the synchrotron dominated S-band Polarization All Sky Survey (SPASS) maps at 2.3 GHz, the Planck satellite maps (30-857 GHz) and dust dominated Infrared Astronomical Satellite (IRAS) maps at 3 THz. We measure bispectra of total intensity fields, T, as well as the gradient, E, and curl modes, B, of the polarization field. We find that the synchrotron and galactic dust have strong temperature bispectra with significant contributions in the squeezed limit, which probes the correlations between two small scale modes with a large scale mode. Additionally, we find that the dust also has strong polarised bispectra that also peak in the squeezed configuration. We explore parity odd bispectra, such as BTT bispectra, and find strong parity odd bispectra for the galactic dust notably in BTT, BTE and BEE configurations. After masking bright sources, we find no evidence for polarised synchrotron bispectra and no evidence for cross bispectra between the dust and synchrotron emission. The strong foreground bispectra discussed here need to be carefully controlled to avoid biasing measurements of primordial non-Gaussianity. Finally we use these bispectra tools to test for residual foregrounds in the component separated Planck maps and find no evidence of residual foregrounds. These tools will be useful for characterizing residual foregrounds in component separated maps, particularly for experiments with less frequency coverage than the Planck satellite.