Galactic foregrounds: Spatial fluctuations and a procedure of removal

TL;DR

This paper proposes a new method for removing Galactic foreground emissions from CMB data by leveraging the statistical properties of their power-law spectra, which vary with Galactic latitude, to improve foreground separation.

Contribution

It introduces a novel spatial filtering technique based on the universality of turbulence spectra, enhancing foreground removal in CMB studies by using templates of spatial fluctuations.

Findings

The power-law spectra of Galactic foregrounds change with latitude due to emission region geometry.

The proposed filtering method can effectively separate foregrounds from CMB signals at different scales.

Combining multi-scale measurements improves the accuracy of foreground removal.

Abstract

Present day cosmic microwave background (CMB) studies require more accurate removal of Galactic foreground emission. In this paper, we consider a way of filtering out the diffuse Galactic fluctuations on the basis of their statistical properties, namely, the power-law spectra of fluctuations. We focus on the statistical properties of two major Galactic foregrounds that arise from magnetized turbulence, namely, diffuse synchrotron emission and thermal emission from dust and describe how their power laws change with the Galactic latitude. We attribute this change to the change of the geometry of the emission region and claim that the universality of the turbulence spectrum provides a new way of removing Galactic foregrounds. We discuss and demonstrate how we can make use of our findings to remove Galactic foregrounds using a template of spatial fluctuations. In particular, we consider examples of spatial filtering of a foreground at small scales, when the separation into CMB signal and foregrounds is done at larger scales. We demonstrate that the new technique of spatial filtering of foregrounds may be promising for recovering the CMB signal in a situation when foregrounds are known at a scale different from the one under study. It can also improve filtering by combining measurements obtained at different scales.