Extragalactic radio-source evolution under the dual-population unification scheme

Abstract

We show that a dual-population unification scheme provides a successful paradigm with which to describe the evolution and beaming of all bright extragalactic radio sources. The paradigm consists of two intrinsic radio-source populations, based on the two distinct radio-galaxy morphologies of Fanaroff-Riley (FR) classes I and II. These represent the 'unbeamed' or 'side-on' parent populations of steep radio spectra; the 'beamed' source types, including flat-spectrum quasars and BL Lac objects, arise through the random alignment of their radio axis to our line of sight where Doppler beaming of the relativistic radio jets produces highly anisotropic radio emission. We develop the model in two stages. In the first stage the source space density as a function of cosmic epoch is determined for the two parent populations, and for this we use low-frequency source-count and identification data to avoid biases due to Doppler-enhanced radio emission. The second stage defines the beaming models for each population, using high-frequency survey data and in particular the 5-GHz source count in which at high flux densities the flat- and steep-spectrum sources contribute in similar measures. We assume that the flat-spectrum objects, quasars and BL Lac objects are 'beamed' versions of FR I and FR II objects in which the close alignment of the radio axis with the line of sight has changed the radio appearance into a core-dominated (flat-spectrum) object. We adopt a simple parametrization of the beaming, orient the parent populations at random with a Monte Carlo process, and use a minimization process to determine beaming parameters that yield a best fit to the 5-GHz source count. The best-fitting parameters are found to be in good agreement with those measured observationally for individual radio sources. In this, the model accurately reproduces the change in source-count form with frequency. Indeed the unified-scheme paradigm has great predictive power, and we show how the model successfully describes several additional and independent data sets.