The multifrequency angular power spectrum in parameter studies of the cosmic 21-cm signal

Abstract

The light-cone effect breaks the periodicity and statistical homogeneity (ergodicity) along the line-of-sight direction of cosmological emission/absorption line surveys. The spherically averaged power spectrum (SAPS), which by definition assumes ergodicity and periodicity in all directions, can only quantify some of the second-order statistical information in the 3D light-cone signals, and therefore, gives a biased estimate of the true statistics. The multifrequency angular power spectrum (MAPS), by extracting more information from the data, does not rely on these assumptions. It is therefore better aligned with the properties of the signal. We have compared the performance of the MAPS and SAPS metrics for parameter estimation of a mock 3D light-cone observation of the 21-cm signal from the Epoch of Reionization. Our investigation is based on a simplified 3-parameter 21cmFAST model. We find that the MAPS produces parameter constraints, which are a factor of ∼2 more stringent than when the SAPS is used. The significance of this result does not change much even in the presence of instrumental noise expected for 128 h of SKA-Low observations. Our results therefore suggest that a parameter estimation framework based on the MAPS metric would yield superior results over one using the SAPS metric.