HERA Mock Observations: Looking for Closure HERA Memorandum Number 13

TL;DR

This study explores the potential of closure phase analysis in detecting the HI 21cm signal from cosmic reionization, demonstrating that while the signal shows distinctive frequency structure, continuum contamination complicates its extraction.

Contribution

It introduces a method using closure phases for HI 21cm detection and evaluates its effectiveness with realistic sky models and HERA configurations.

Findings

Closure spectra of HI signal show distinctive frequency structure.

Continuum signals dominate, making line detection challenging.

Smooth fitting cannot reliably separate line from continuum in closure spectra.

Abstract

We investigate the use of closure phase as a method to detect the HI 21cm signal from the neutral IGM during cosmic reionzation. Closure quantities have the unique advantage of being independent of antenna-based calibration terms. We employ realistic, large area sky models from Sims et al. (2016). These include an estimate of the HI 21cm signal generated using 21cm FAST, plus continuum models of both the diffuse Galactic synchrotron emission and the extragalactic point sources. We employ the CASA simulator and adopt the Dillon-Parsons HERA configuration to generate a uv measurement set. We then use AIPS to calculate the closure phases as a function of frequency ('closure spectra'), and python scripts for subsequent analysis. We find that the closure spectra for the HI signal show dramatic structure in frequency, and based on thermal noise alone, the redundant HERA-331 array should detect these fluctuations easily. Comparatively, the frequency structure in the continuum closure spectra is much smoother than that seen in the HI closure spectra. Unfortunately, when the line and continuum signals are combined, the continuum dominates the visibilities at the level of 10^3 to 10^4, and the line signal is lost. We have investigated fitting and removing smooth curves in frequency to the line plus continuum closure spectra, and find that the continuum itself shows enough structure in frequency in the closure spectra to preclude separation of the continuum and line based on such a process. We have also considered the subtraction of the continuum from the visibilities using a sky model, prior to calculation of the closure spectra. TRUNCATED.