The clustering of radio galaxies: biasing and evolution versus stellar mass

We study the angular clustering of $\sim 6\times 10^5$ NVSS sources on scales $\gtrsim 50 h^{-1}$ Mpc in the context of the $\Lambda$CDM scenario. The analysis partially relies on the redshift distribution of 131 radio galaxies, inferred from the Hercules and CENSORS survey, and an empirical fit to the stellar to halo mass (SHM) relation. For redshifts $z\lesssim 0.7$, the fraction of radio activity versus stellar mass evolves as $f_{_{\rm RL}}\sim M_*^{\alpha_0+\alpha_1 z}$ where $\alpha_0=2.529{\pm0.184}$ and $\alpha_1=1.854^{+0.708}_{-0.761}$. The estimate on $\alpha_0$ is largely driven by the results of Best et al. (2005), while the constraint on $\alpha_1$ is new. We derive a biasing factor $b(z=0.5)=2.093^{+0.164}_{-0.109}$ between radio galaxies and the underlying mass.The function $b(z)=0.33z^2+0.85 z +1.6$ fits well the redshift dependence. We also provide convenient parametric forms for the redshift dependent radio luminosity function, which are consistent with the redshift distribution and the NVSS source count versus flux.