The Necessity of Individually Validated Beam Models for an Interferometric Epoch of Reionization Detection

TL;DR

Accurately modeling and calibrating with individually validated beam models is crucial for detecting the Epoch of Reionization, as beam variations significantly affect foreground leakage and signal recovery.

Contribution

This paper demonstrates the importance of using physically motivated, individually validated beam models for calibration to mitigate foreground leakage in EoR detection.

Findings

Beam-induced foreground leakage can exceed the cosmological signal by over 1000 times.

Using a full set of deformed beam models reduces leakage to sub-dominant levels.

Calibrating with measured, varied beam models improves EoR detection prospects.

Abstract

A first statistical detection of the 21-cm Epoch of Reionization (EoR) is on the horizon, as cosmological volumes of the Universe become accessible via the adoption of low-frequency interferometers. We explore the impact which non-identical instrumental beam responses can have on the calibrated power spectrum and a future EoR detection. All-sky satellite measurements of Murchison Widefield Array (MWA) beams have revealed significant sidelobe deviations from cutting-edge electromagnetic simulations at the ~10% zenith power level. By generating physically motivated deformed beam models, we emulate real measurements of the MWA which inherently encode the imprints of varied beams. We explore two calibration strategies: using a single beam model across the array, or using a full set of deformed beams. Our simulations demonstrate beam-induced leakage of foreground power into theoretically uncontaminated modes, at levels which exceed the expected cosmological signal by factors of over ~1000 between the modes k=0.1-1 $hMpc^{-1}$. We also show that this foreground leakage can be mitigated by including measured models of varied beams into calibration frameworks, reducing the foreground leakage to a sub-dominant effect and potentially unveiling the EoR. Finally, we outline the future steps necessary to make this approach applicable to real measurements by radio interferometers.