Foreground removal for Square Kilometre Array observations of the Epoch of Reionization with the Correlated Component Analysis

TL;DR

This paper demonstrates that the Correlated Component Analysis method can effectively remove diffuse foreground contamination from simulated 21 cm reionization signals in Square Kilometre Array data, even with complex foreground models.

Contribution

The study adapts the CCA method, originally for CMB, to EoR data, showing its robustness without assuming spectral smoothness and across complex foreground simulations.

Findings

Effective foreground cleaning across 100-200 MHz range

CCA performs well with complex, non-smooth foreground models

Method shows promise for Epoch of Reionization component separation

Abstract

We apply the Correlated Component Analysis (CCA) method on simulated data of the Square Kilometre Array, with the aim of accurately cleaning the 21 cm reionization signal from diffuse foreground contamination. The CCA has been developed for the Cosmic Microwave Background, but the application of the Fourier-domain implementation of this method to the reionization signal is straightforward. The CCA is a parametric method to estimate the frequency behaviour of the foregrounds from the data by using second-order statistics. We test its performance on foreground simulations of increasing complexity, designed to challenge the parametric models adopted. We also drop the assumption of spectral smoothness that most of the methods rely upon. We are able to clean effectively the simulated data across the explored frequency range (100-200 MHz) for all the foreground simulations. This shows that the CCA method is very promising for EoR component separation.