Detecting Galaxies in the Epoch of Reionization

Abstract

The aim of this project was to develop a technique that allows us to measure the redshift of radio galaxies in the Epoch of Reionization. We achieved this by constructing a parameterization of the 21-cm absorption feature which is expected to be seen in the spectra of high redshift radio galaxies. A Bayesian framework for parameter estimation and model selection was applied on the candidate models found. The method's proficiency is demonstrated with two data sets: SIMFAST21-generated & model-generated data. The SIMFAST21-generated data was constructed using a simulation of the Epoch of Reionization. The model-generated data was synthesized from our best analytical model for the 21-cm absorption feature. A spectrum template from radio galaxies at low redshifts was then combined with the data sets. The resulting spectra are fitted using a model that accounts for 21-cm absorption feature. The parameter fitting was done using the expected SKA1-LOW noise. A strong detection can be achieved for galaxies with redshifts zg >12 and luminosities L > 1026[W/Hz]. Objects at lower z, although with brighter flux, are harder to detect due to the weak absorption feature (e.g. less intervening HI). A detection proxy grid space was assembled to quantify model evidence found when fitting data with the best model and null hypothesis (no absorption). To interpret the detection proxy, the Jeffreys scale was used. The grid for both data sets was overlaid with the SKADS simulated sources. In a 25 deg2 sky area, a total of 62206 "SKADS" radio galaxies above redshift 7 are expected to be detected in continuum by SKA-LOW. From these, about 1% should be confirmed to be in the epoch of reionization by SKA-LOW (with the same survey) through the detection of the 21 cm absorption feature using the proposed technique