Accurately accounting for spectral structure in spectrometer data induced by instrumental chromaticity on scales relevant for detection of the 21-cm signal is among the most significant challenges in global 21-cm signal analysis. In the publicly available EDGES low-band data set, this complicating structure is suppressed using beam-factor based chromaticity correction (BFCC), which works by dividing the data by a sky-map-weighted model of the spectral structure of the instrument beam. Several analyses of this data have employed models that start with the assumption that this correction is complete. However, while BFCC mitigates the impact of instrumental chromaticity on the data, given realistic assumptions regarding the spectral structure of the foregrounds, the correction is only partial. This complicates the interpretation of fits to the data with intrinsic sky models (models that assume no instrumental contribution to the spectral structure of the data). In this paper, we derive a BFCC data model from an analytic treatment of BFCC and demonstrate using simulated observations that, in contrast to using an intrinsic sky model for the data, the BFCC data model enables unbiased recovery of a simulated global 21-cm signal from beam-factor chromaticity corrected data in the limit that the data is corrected with an error-free beam-factor model.
@article{arxiv.2212.03875,
title = {A Bayesian approach to modelling spectrometer data chromaticity corrected using beam factors -- I. Mathematical formalism},
author = {Peter H. Sims and Judd D. Bowman and Nivedita Mahesh and Steven G. Murray and John P. Barrett and Rigel Cappallo and Raul A. Monsalve and Alan E. E. Rogers and Titu Samson and Akshatha K. Vydula},
journal= {arXiv preprint arXiv:2212.03875},
year = {2023}
}
Comments
26 pages, 8 figures; accepted for publication in MNRAS