Formulating and Critically Examining the Assumptions of Global 21 cm Signal Analyses: How to Avoid the False Troughs That Can Appear in Single-spectrum Fits

Abstract

Abstract The assumptions inherent to global 21 cm signal analyses are rarely delineated. In this paper, we formulate a general list of suppositions underlying a given claimed detection of the global 21 cm signal. Then, we specify the form of these assumptions for two different analyses: (1) the one performed by the team for the Experiment to Detect the Global Epoch-of-Reionization Signature (EDGES) showing an absorption trough in brightness temperature that they modeled separately from the sky foreground and (2) a new, so-called minimum assumption analysis (MAA) that makes the most conservative assumptions possible for the signal. We show fits using the EDGES analysis on various beam-weighted foreground simulations from the EDGES latitude with no signal added. Depending on the beam used, these simulations produce large false troughs because of the invalidity of the foreground model in describing the combination of beam chromaticity and the shape of the Galactic plane in the sky, the residuals of which are captured by the ad hoc flattened Gaussian signal model. On the other hand, the MAA provides robust fits by including many spectra at different time bins and allowing any possible 21 cm spectrum to be modeled exactly. We present uncertainty levels and example signal reconstructions found with the MAA for different numbers of time bins. With enough time bins, one can determine the true 21 cm signal with the MAA to <10 times the noise level.