Will point sources spoil 21 cm tomography?

TL;DR

This paper evaluates the challenges of removing unresolved extragalactic point source foregrounds in 21 cm cosmology, emphasizing the importance of instrument design and algorithms to ensure successful tomography of the early universe.

Contribution

It analyzes how instrumental and algorithmic parameters affect foreground removal, especially considering the frequency-dependent beam effects, providing guidelines for experiment design.

Findings

Current experiments can effectively clean point source contamination

Frequency-dependent beam effects introduce complex spectral structures

Design specifications are identified to optimize foreground removal

Abstract

21 cm tomography is emerging as a promising probe of the cosmological dark ages and the epoch of reionization, as well as a tool for observational cosmology in general. However, serious sources of foreground contamination must be subtracted for experimental efforts to be viable. In this paper, we focus on the removal of unresolved extragalactic point sources with smooth spectra, and evaluate how the residual foreground contamination after cleaning depends on instrumental and algorithmic parameters. A crucial but often ignored complication is that the synthesized beam of an interferometer array shrinks towards higher frequency, causing complicated frequency structure in each sky pixel as "frizz" far from the beam center contracts across unresolved radio sources. We find that current-generation experiments should nonetheless be able to clean out this points source contamination adequately, and quantify the instrumental and algorithmic design specifications required to meet this foreground challenge.