We previously reported a putative detection of a radio galaxy at $z=10.15$ , selected from the GaLactic and Extragalactic All-sky Murchison Widefield Array (GLEAM) survey. The redshift of this source, GLEAM J0917–0012, was based on three weakly detected molecular emission lines observed with the Atacama Large Millimetre Array (ALMA). In order to confirm this result, we conducted deep spectroscopic follow-up observations with ALMA and the Karl Jansky Very Large Array (VLA). The ALMA observations targeted the same CO lines previously reported in Band 3 (84–115 GHz) and the VLA targeted the CO(4-3) and [CI(1-0)] lines for an independent confirmation in Q-band (41 and 44 GHz). Neither observation detected any emission lines, removing support for our original interpretation. Adding publicly available optical data from the Hyper Suprime-Cam survey, Widefield Infrared Survey Explorer (WISE), and Herschel Space Observatory in the infrared, as well as $<$ 10 GHz polarisation and 162 MHz inter-planetary scintillation observations, we model the physical and observational characteristics of GLEAM J0917–0012 as a function of redshift. Comparing these predictions and observational relations to the data, we are able to constrain its nature and distance. We argue that if GLEAM J0917–0012 is at $z<3,$ then it has an extremely unusual nature, and that the more likely solution is that the source lies above $z=7$ .

We present the results of a new selection technique to identify powerful ($L_{\rm 500\,MHz} \gt 10^{27}\,\text{WHz}^{-1}$) radio galaxies towards the end of the Epoch of Reionisation. Our method is based on the selection of bright radio sources showing radio spectral curvature at the lowest frequency (${\sim}100\,\text{MHz}$) combined with the traditional faintness in K-band for high-redshift galaxies. This technique is only possible, thanks to the Galactic and Extra-galactic All-sky Murchison Wide-field Array survey which provides us with 20 flux measurements across the 70–$230\,\text{MHz}$ range. For this pilot project, we focus on the GAMA 09 field to demonstrate our technique. We present the results of our follow-up campaign with the Very Large Telescope, Australian Telescope Compact Array, and the Atacama Large Millimetre Array to locate the host galaxy and to determine its redshift. Of our four candidate high-redshift sources, we find two powerful radio galaxies in the $1<z<3$ range, confirm one at $z=5.55$, and present a very tentative $z=10.15$ candidate. Their near-infrared and radio properties show that we are preferentially selecting some of the most radio luminous objects, hosted by massive galaxies very similar to powerful radio galaxies at $1<z<5$. Our new selection and follow-up technique for finding powerful radio galaxies at $z>5.5$ has a high 25–50% success rate.

We report the first detection of [NII] 122 μm line toward a QSO-SMG pair, BRI 1202-0725, at z = 4.7 using ALMA. Combining with [NII] 205 μm line detection and taking the line ratio of [NII]122/[NII]205, we constrain electron densities of both galaxies. The derived electron densities are ${26_{ - 11}^{ + 12}$ and ${134_{ - 39}^{ + 50}$ cm−3 for the SMG and the QSO, respectively, which are the first measurements for galaxies at z > 4. The electron density of the SMG is comparable to the Galactic plane and the average of local spiral galaxies, while the value for the QSO is comparable to local starbursts and optical-line based measurements for star-forming galaxies at z ∼ 2–3. Considering the similar star-formation rates (SFRs) of ≍ 1000 M⊙ yr−1 for both galaxies, our results suggest a large scatter of electron densities at fixed SFR and caution against using optical lines for dusty starbursts. The details of this report are presented in Lee et al. 2019 (submitted).

The HeRGÉ (Herschel Radio Galaxy Evolution) project consists of a sample of 70 radio galaxies in the range 1 < z < 5.2. They benefit from continuous coverage from 3 to 870μm with Spitzer, Herschel and sub-mm ground-based instruments (SCUBA, LABOCA). As a calorimeter, IR is an excellent proxy to estimate the contribution of both AGN and starburst, making of radio galaxies perfect candidates to provide new insights into the relationship between AGN and their host galaxies. The IR SED fitting with empirical templates reveals that radio galaxies are luminous and that their black holes and their host galaxies are not growing simultaneously. Extending the SED to optical/near-IR on a subsample of 12 radio galaxies spanning 1 < z < 4 reveal the necessity of three components to reproduce the observations. Making use of the evolutionary code PEGASE.3 and an AGN torus model, we are able to estimate parameters from the AGN torus, the evolved stellar population and the starburst (SB). They reveal that radio galaxies are massive, evolved, forming the bulk of their mass at very high redshift in a short timescale, but experience episodic, strong SB events, often associated with an AGN activity.

Radio galaxies are identified with the most massive host galaxies known out to z = 5 and can put strong constraints on galaxy evolution models, provided their space density is accurately determined. Here, I present the important role low-frequency radio surveys will have in this by selecting the highest redshift radio galaxies using the steepness of their radio spectra.

The Sydney University Molonglo Sky Survey (SUMSS) is a radio continuum imaging survey of the southern sky at 843 MHz, with similar sensitivity and resolution to the northern NRAO VLA Sky Survey (NVSS). We have combined radio data from SUMSS and NVSS with optical spectra from the 2dF Galaxy Redshift Survey and 6dF Galaxy Survey to study the space distribution and properties of both AGN and star-forming galaxies in the local universe (redshift range 0 < z < 0.3). We also discuss new results on radio-source clustering in the more distant universe (z ∼ 1) and present the first results from a search for the most distant (z > 3) southern radio galaxies.

We present a short overview of radio surveys for AGN, including the ‘complete’ flux limited surveys and ‘filtered’ surveys. We also describe our ultra-steep spectrum search for the highest redshift radio galaxies, and our follow-up VLA and ATCA observations of the most distant (z = 5.19) and the most luminous z <2 radio galaxy known.