The Clustering of Active Galactic Nuclei and Star Forming Galaxies in the LoTSS DeepFields

Abstract

Using deep observations across three of the LOFAR Two-metre Sky Survey Deep Fields, this work measures the angular clustering of star forming galaxies (SFGs) and low-excitation radio galaxies (LERGs) to $z$$\lesssim$1.5 for faint sources, $S_{\textrm{144 MHz}}$$\geq$200 $μ$Jy. We measure the angular auto-correlation of LOFAR sources in redshift bins and their cross-correlation with multi-wavelength sources {to} measure the evolving galaxy bias for SFGs and LERGs. Our work shows the bias of the radio-selected SFGs increases from $b=0.90^{+0.11}_{-0.10}$ at $z \sim 0.2$ to $b = 2.94^{+0.36}_{-0.36}$ at $z \sim 1.2$; faster than the assumed $b(z)$$\propto$$1/D(z)$ models adopted in previous LOFAR cosmology studies (at sensitivities where AGN dominate), but in broad agreement with previous work. We further study the luminosity dependence of bias for SFGs and find little evidence for any luminosity dependence at fixed redshift, although uncertainties remain large for the sample sizes available. The LERG population instead shows a weaker redshift evolution with $b=2.33^{+0.28}_{-0.27}$ at $z \sim 0.7$ to $b=2.65^{+0.57}_{-0.55}$ at $z \sim 1.2$, though it is also consistent with the assumed bias evolution model ($b(z)$$\propto$$1/D(z)$) within the measured uncertainties. For those LERGs which reside in quiescent galaxies (QLERGs), there is weak evidence that they are more biased than the general LERG population and evolve from $b = 2.62^{+0.33}_{-0.33}$ at $z \sim 0.7$ to $b = 3.08^{+0.85}_{-0.84}$ at $z \sim 1.2$. This suggests the halo environment of radio sources may be related to their properties. These measurements can help constrain models for the bias evolution of these source populations, and can help inform multi-tracer analyses.

Figures (18)

• Figure 1: Flowchart outlining the steps to make the catalogue of random sources associated with the radio data which are used to measure the clustering, divided into three stages. The first (yellow) relates to the creation of the general simulated source properties, as in Section \ref{['sec:randoms_radio']}. Second (in blue) describes the method of applying completeness effects and measurement errors, as in Section \ref{['sec:randoms_radio']}(i). Finally, is the effect of applying corrections for the intrinsic luminosity distribution (pink), as in Section \ref{['sec:randoms_radio']}(ii).
• Figure 2: Comparison plots of the flux density distributions (1st row); redshift distributions (2nd row); luminosity distribution (3rd row); signal-to-noise (SNR, 4th row); and integrated-to-peak flux density ratio ($S_I/S_P$; 5th row) for SFGs in the different redshift bins considered in this work, increasing in redshift from left to right. In each panel, the data catalogue with redshift cuts applied on the Z_BEST column are shown as black dots, whilst each blue shaded region represents the output distribution from the data samples given in the range of the 16th - 84th percentiles of the values from the $p(z)$ resamples. The randoms for the full sample are shown as red stars. These have associated red shaded regions with the range of randoms from those associated with each of the data $p(z)$ resample (to ensure a constant ratio of random sources to data), though these are small as they are drawn from the same random sample and only have small differences reflecting the number of data per $p(z)$ sample.
• Figure 3: LERGs
• Figure 4: QLERGs
• Figure 6: Auto correlation of the multi-wavelength sample of galaxies in each of the redshifts bins used to measure $\omega(\theta)$ for SFGs. Black open pentagons indicate the combined TPCF across the three fields, with their individual $\omega(\theta)$ shown for Boötes (red stars), ELAIS-N1 (blue squares) and Lockman Hole (gold triangles). The dashed vertical lines highlight the region used to fit the correlation function over in order to measure the bias.