Discovery of a radio galaxy at z = 5.72
A. Saxena, M. Marinello, R.A. Overzier, P.N. Best, H.J.A Rottgering, K.J. Duncan, I. Prandoni, L. Pentericci, M. Magliocchetti, D. Paris, F. Cusano, F. Marchi, H.T. Intema, G.K. Miley
MMNRAS , 1–11 (2018) Preprint 11 October 2018 Compiled using MNRAS L A TEX style file v3.0
Discovery of a radio galaxy at z = 5.72
A. Saxena (cid:63) , M. Marinello , , R. A. Overzier , P. N. Best , H. J. A. R¨ottgering ,K. J. Duncan , I. Prandoni , L. Pentericci , M. Magliocchetti , D. Paris , F. Cusano ,F. Marchi , H. T. Intema and G.K. Miley Leiden Observatory, Leiden University, P.O. Box 9513, 2300 RA Leiden, The Netherlands Observat´orio Nacional, Rua General Jos´e Cristino, 77, S˜ao Crist´ov˜ao, Rio de Janeiro, RJ, CEP 20921-400, Brazil Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, EH9 3HJ Edinburgh, UK INAF-Instituto di Radioastronomia, Via P. Gobetti 101, I-40129 Bologna, Italy INAF-Osservatorio Astronomico di Roma, Via Frascati 33, I-00040 Monteporzio (RM), Italy IAPS-INAF, Via Fosso del Cavaliere 100, I-00133 Rome, Italy INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via P. Gobetti 93/3, I-40129 Bologna, Italy
Accepted 2018 July 12. Received 2018 July 5; in original form 2018 June 4
ABSTRACT
We report the discovery of the most distant radio galaxy to date, TGSS J1530+1049at a redshift of z = . , close to the presumed end of the Epoch of Reionisation. Theradio galaxy was selected from the TGSS ADR1 survey at 150 MHz for having anultra-steep spectral index, α
150 MHz1.4 GHz = − . and a compact morphology obtained usingVLA imaging at 1.4 GHz. No optical or infrared counterparts for the radio source werefound in publicly available sky surveys. Follow-up optical spectroscopy at the radioposition using GMOS on Gemini North revealed the presence of a single emissionline. We identify this line as Lyman alpha at z = . , because of its asymmetric lineprofile, the absence of other optical/UV lines in the spectrum and a high equivalentwidth. With a Ly α luminosity of . × erg s − and a FWHM of km s − ,TGSS J1530+1049 is comparable to ‘non-radio’ Lyman alpha emitters (LAEs) at asimilar redshift. However, with a radio luminosity of log L
150 MHz = . W Hz − anda deconvolved physical size . kpc, its radio properties are similar to other knownradio galaxies at z > . Subsequent J and K band imaging using LUCI on the LargeBinocular Telescope resulted in non-detection of the host galaxy down to σ limits of J > . and K > . (Vega). The K band limit is consistent with z > from the K − z relation for radio galaxies and helps rule out low redshifts. The stellar mass limitderived using simple stellar population models is M stars < . M (cid:12) . Its relatively lowstellar mass and small radio and Ly α sizes suggest that TGSS J1530+1049 may be aradio galaxy in an early phase of its evolution. Key words: galaxies: individual: TGSS J1530+1049 – galaxies: high-redshift
High-redshift radio galaxies (HzRGs) are thought to be theprogenitors of local massive elliptical galaxies and generallycontain large amounts of dust and gas (Best et al. 1998;Carilli et al. 2002a; Reuland et al. 2004; De Breuck et al.2010). They are also among the most massive galaxies attheir redshift (Overzier et al. 2009) and are often found to belocated at the centre of clusters and proto-clusters of galaxies(Pentericci et al. 2000; Venemans et al. 2002; R¨ottgeringet al. 2003; Miley et al. 2004; Hatch et al. 2011; Orsi et al. (cid:63)
E-mail: [email protected] z > , in the Epoch of Reionisation(EoR), are of particular interest as they could be used asunique tools to study the process of reionisation in detail.At these redshifts, the hyper-fine transition line from neu-tral hydrogen atoms, with a rest-frame wavelength of 21cm, falls in the low-frequency radio regime and can be ob-served as absorption signals in spectra of luminous back-ground radio sources such as radio galaxies (Carilli et al. © a r X i v : . [ a s t r o - ph . GA ] O c t A. Saxena et al. z > − (Dunlop & Peacock1990; Willott et al. 2001; Rigby et al. 2011, 2015). Althougha number of quasars at z > are known, with a few alsobeing radio-loud (see Ba˜nados et al. 2015, for example), thesame cannot be said for radio galaxies - previously, the onlyknown radio galaxy at z > was TN J0924 − z = . (van Breugel et al. 1999). If the orientation-based unifica-tion of radio galaxies and quasars is valid (see Morabitoet al. 2017, for example), the number of radio quasars andgalaxies at any given epoch should be comparable and itmay be possible that many of the already unidentified radiosources are at z > . The scarcity of z > radio galax-ies could therefore be due to the relative difficulty in firstidentifying these sources amongst the wider radio sourcepopulation, and then obtaining spectroscopic redshifts forthese radio galaxies, which are optically much fainter thanquasars. Dedicated spectroscopic follow-up of radio sourcessuch as the WEAVE-LOFAR survey (Smith et al. 2016), andthe upcoming major optical facilities such as the ExtremelyLarge Telescope (ELT), the Thirty Meter Telescope (TMT),the Giant Magellan Telescope (GMT) and the James WebbSpace Telescope (JWST) will help overcome these difficul-ties and potentially help characterise a number of USS radiosources.The key requirement for gathering enough statistics formeaningful studies of radio galaxies at high redshifts aredeep low-frequency radio surveys covering large areas on thesky. Surveys such as the TIFR GMRT Sky Survey Alterna-tive Data Release 1 (TGSS; Intema et al. 2017) and the cur-rently ongoing surveys using LOFAR (Shimwell et al. 2017)are opening up new parameter spaces for searches for radiogalaxies at z ≥ . Using TGSS, which covers the entire ra-dio sky north of − declination at a frequency of 150 MHzand achieving a median noise level of . mJy beam − , welaunched a campaign to hunt for fainter and potentially moredistant HzRGs, with the ultimate aim of discovering radiogalaxies that could be suitable probes of the EoR (Saxenaet al. 2018). In this paper, we report the discovery of a radiogalaxy at a redshift of z = . , TGSS J1530+1049, whichwas pre-selected as part of our sample of high-redshift radiogalaxy candidates.The layout of this paper is as follows. In Section 2we present details about the initial source selection crite-ria and follow-up radio observations at high resolution forTGSS J1530+1049. In Section 3 we present the new opti-cal spectroscopy and infrared imaging obtained for TGSSJ1530+1049 and expand upon the data reduction methods. In Section 4 we describe how the redshift for this source wasdetermined. In Section 5 we study the emission line and ra-dio properties of this source and set constraints on its stellarmass. We also compare the observed properties to galaxiesat the same epoch from the literature. Finally, in Section 6we summarise the findings of this paper. Throughout thispaper we assume a flat Λ CDM cosmology with H = 70 kms − Mpc − and Ω m = . . Using this cosmology, at a redshiftof 5.72 the age of the Universe is 0.97 Gyr, and the angularscale per arcsecond is 5.86 kpc. Our two stage selection process is based on first isolatingcompact radio sources with an ultra-steep spectrum (USS; α < − . , where S ν ∝ ν α ) at radio wavelengths, that has his-torically been very successful at finding HzRGs from widearea radio surveys (R¨ottgering et al. 1994; Blundell et al.1999; De Breuck et al. 2000; Afonso et al. 2011), and thencombining it with optical and/or infrared faintness require-ments. The relation that exists between the apparent K -band magnitude of radio galaxies and their redshift, knownas the K − z relation, (Lilly & Longair 1984; Jarvis et al. 2001;Willott et al. 2003; Rocca-Volmerange et al. 2004) gives fur-ther strength to the argument of selecting USS sources thatare also faint at near-infrared wavelengths in a bid to iso-late HzRGs (Ker et al. 2012). Deep near-infrared imaging ofpromising USS candidates can therefore serve as an indepen-dent way to set constraints on the redshifts of radio sources.HzRGs are expected to be very young and therefore, havesmall sizes at the highest redshifts (Saxena et al. 2017): im-plementing an additional criterion that puts an upper limiton the angular sizes of radio sources has the potential toincrease the efficiency of pin pointing the highest-redshiftsources in a wide area radio survey.Combining all of these selection methods, we firstshortlisted 588 candidates with an ultra-steep spectrum( α
150 MHz1.4 GHz < − . ) and compact morphologies, out of a to-tal of 65,996 sources that had spectral index informationfrom TGSS and FIRST/NVSS. This sample probes fainterflux densities than previous large area searches and has fluxlimits that ensure that a new parameter space in flux den-sity and spectral index is probed, where a large number ofundiscovered HzRGs are expected to lie (Ishwara-Chandraet al. 2010). From this shortlist, we then retained in oursample only those radio sources that are blank in all avail-able optical surveys such as the Sloan Digital Sky SurveyDR12 (SDSS; Alam et al. 2015) and the Pan-STARRS1 sur-vey (PS1; Chambers et al. 2016), and infrared surveys suchas ALLWISE using the WISE satellite (Wright et al. 2010)and the UKIDSS surveys (Lawrence et al. 2007) to maximisethe chances of finding radio galaxies at the highest redshifts.This led to a final sample of 32 very promising HzRG candi-dates. Details of the sample selection can be found in Saxenaet al. (2018).High resolution imaging using the Karl G. Jansky VeryLarge Array (VLA) for the 32 candidates, including forTGSS J1530+1049 (RA: 15:30:49.9, Dec: + MNRAS , 1–11 (2018) adio galaxy at z = 5.72 Figure 1.
The location of TGSS J1530+1049 in the fluxdensity − spectral index parameter space. The large orange pointsshow the parameter space probed by the Saxena et al. (2018)sample and the smaller grey points show radio sources from DeBreuck et al. (2000), scaled to an observed frequency of 150 MHzusing the spectral indices provided for individual sources. Alsoshown for comparison is TN J0924 − z = . (van Breugelet al. 1999). TGSS J1530+1049 is fainter than the previouslystudied large area samples and offers a new window into fainterradio galaxies at high redshifts. scopic follow-up. TGSS J1530+1049 in particular showed acompact morphology, which was fitted with a single Gaus-sian. With a flux density of S
150 MHz = ± mJy, TGSSJ1530+1049 is one of the brightest sources in the sample.At 1.4 GHz, it has a flux density S = . ± . mJy,giving a spectral index of α = − . ± . . With a relativelysmall (deconvolved) angular size of . ± . arcsec, TGSSJ1530+1049 was deemed to be a promising HzRG candi-date. We show the location of TGSS J1530+1049 in the fluxdensity − spectral index parameter space in Figure 1.TGSS J1530+1049 is not detected in any of the PS1bands ( g , r , i , z and y ). This source also happens to liein the sky area covered by the UKIDSS Large Area Survey(LAS), and is not detected down to (Vega) magnitude limitsof y > . , J > . , H > . and K > . . We show theimage obtained from stacking all of the LAS bands withradio contours overlaid in Figure 2. Lastly, this source isalso not detected in any of the ALLWISE bands. These non-detections coupled with the ultra-steep radio spectral indexand compact radio morphology are in line with expectationsof a high-redshift host galaxy and made TGSS J1530+1049a prime candidate for follow-up spectroscopy. A long-slit spectrum of TGSS J1530+1049TGSSJ1530+1049 was taken using GMOS on Gemini North on28 April, 2017 (Program ID: GN-2017A-Q-8; PI: Overzier)using the filter GG455 G0305 and the R400 G5305 gratinggiving a wavelength coverage of − ˚A and a Figure 2.
Stacked y , J , H and K band image from the UKIDSSLarge Area Survey, with contours (starting from 0.5 mJy, in ageometric progression of √ ) from the 1.4 GHz VLA map (Saxenaet al. 2018) overplotted for TGSS J1530+1049. The radio sourceis compact and has an ultra-steep spectral index. A non-detectionin the UKIDSS LAS K band down to a magnitude limit of 18.4Vega ( ∼ . AB) made TGSS J1530+1049 a promising HzRGcandidate and a prime target for spectroscopic follow-up. resolution of roughly R ∼ . The central wavelength wasset to 7000 ˚A. The total length of the slit was 300 arcsecondsand the slit width was chosen to be 1.5 arcseconds so thatit covers the entire radio emission footprint detected in theVLA image. As the host galaxy of the radio source wasundetected in all available all-sky optical/IR surveys, weperformed blind offsetting from a bright star, which ensurespositional accuracy to within 0.1 arcseconds, to the centroidof the radio emission. The VLA observations ensuredsub-arcsecond localisation of the expected position of thehost galaxy and the relatively large slit-width providedinsurance against minor positional uncertainties. We took3 exposures of 800 seconds each, giving a total of 2400seconds of on-source exposure time. The standard starEG131 was observed for flux calibration.We used the Gemini IRAF package for reducing thedata, which includes the standard steps for optical spectrumreduction. Briefly, the bias frames were mean stacked in amaster bias which was subtracted from all other images ac-quired. Pixel-to-pixel sensitivity was corrected through theflat field image taken during the day of the observations. Thewavelength solution was derived from the arc lamp frametaken immediately after the science observations, and ap-plied to the science frame and standard star. The 2D imageswere then combined in a single frame, rejecting possible cos-mic rays. The sky lines were removed and flux calibrationwas achieved using the standard star spectrum.A single emission line with a peak at 8170 ˚A and a spa-tial extent of ∼ arcsecond was detected in the reduced 2Dspectrum at the expected position of the radio galaxy. Noother line associated with this source was detected. No con-tinuum was detected either bluewards or redwards of this MNRAS000
Stacked y , J , H and K band image from the UKIDSSLarge Area Survey, with contours (starting from 0.5 mJy, in ageometric progression of √ ) from the 1.4 GHz VLA map (Saxenaet al. 2018) overplotted for TGSS J1530+1049. The radio sourceis compact and has an ultra-steep spectral index. A non-detectionin the UKIDSS LAS K band down to a magnitude limit of 18.4Vega ( ∼ . AB) made TGSS J1530+1049 a promising HzRGcandidate and a prime target for spectroscopic follow-up. resolution of roughly R ∼ . The central wavelength wasset to 7000 ˚A. The total length of the slit was 300 arcsecondsand the slit width was chosen to be 1.5 arcseconds so thatit covers the entire radio emission footprint detected in theVLA image. As the host galaxy of the radio source wasundetected in all available all-sky optical/IR surveys, weperformed blind offsetting from a bright star, which ensurespositional accuracy to within 0.1 arcseconds, to the centroidof the radio emission. The VLA observations ensuredsub-arcsecond localisation of the expected position of thehost galaxy and the relatively large slit-width providedinsurance against minor positional uncertainties. We took3 exposures of 800 seconds each, giving a total of 2400seconds of on-source exposure time. The standard starEG131 was observed for flux calibration.We used the Gemini IRAF package for reducing thedata, which includes the standard steps for optical spectrumreduction. Briefly, the bias frames were mean stacked in amaster bias which was subtracted from all other images ac-quired. Pixel-to-pixel sensitivity was corrected through theflat field image taken during the day of the observations. Thewavelength solution was derived from the arc lamp frametaken immediately after the science observations, and ap-plied to the science frame and standard star. The 2D imageswere then combined in a single frame, rejecting possible cos-mic rays. The sky lines were removed and flux calibrationwas achieved using the standard star spectrum.A single emission line with a peak at 8170 ˚A and a spa-tial extent of ∼ arcsecond was detected in the reduced 2Dspectrum at the expected position of the radio galaxy. Noother line associated with this source was detected. No con-tinuum was detected either bluewards or redwards of this MNRAS000 , 1–11 (2018)
A. Saxena et al. line either. To ensure that the line detection is indeed realand not due to an artefact or contamination by cosmic rays,we looked at the individual frames, both raw and sky sub-tracted, to ensure that the detection (although marginal)was present in each science frame. The three frames areshown in Figure 3. The top panels show the raw frames andthe bottom panels the sky subtracted frames. The emissionline is clearly present in all three frames, ensuring that thedetection is real. The extracted 1D spectrum with a 1 arcsec-ond aperture showing the detected emission line is shown inFigure 4. We give details about line identification in Section4.
Imaging in the J and Ks bands using LUCI (formerly knownas LUCIFER; Seifert et al. 2003) on the Large BinocularTelescope (LBT) was carried out in two separate runs, withthe first on 1 February 2018 and the second on 11 May2018 (Program ID 2017 2018 43; PI: Prandoni). The averageseeing throughout the observations was . − . arcseconds.In the first run, the on-source exposure time was 720 ( × s) seconds in J (central wavelength of 1.247 microns) and1200 ( × s) seconds in Ks (central wavelength of 2.194microns). In the second run we obtained additional 3600( × s) seconds in J and 3000 ( × s) seconds in Ks ,giving a total on-source exposure time of 4320 seconds in J band and 4200 seconds in Ks band.The LUCI data reduction pipeline developed at INAF-OAR was used to perform the basic reduction such as darksubtraction, bad pixel masking, cosmic ray removal, flatfielding and sky subtraction. Astrometric solutions for in-dividual frames were obtained and the single frames werethen resampled and combined using a weighted co-additionto form a deeper image. The (cid:48) × (cid:48) field-of-view of LUCIcontained many bright objects detected in both 2MASS andthe UKIDSS Large Area Survey, which were used to cali-brate the photometry of the images in both bands.The median and standard deviation of the backgroundin both images was calculated by placing 5000 random aper-tures with a diameter of 1.5 arcseconds. We measure σ depths of J = . and Ks = . . Aperture photometryperformed on both J and Ks (from here on we denote Ks as simply K ) images using photutils (Bradley et al. 2017)at the peak of the radio emission using an aperture of di-ameter 1.5 arcseconds yield magnitudes that are lower thanthe σ depths in both images. Smoothing the K band imagewith a × pixel Gaussian kernel reveals a faint source veryclose to the peak radio pixel, as shown in Figure 5, but itis not entirely clear if this indeed the host galaxy and thereis no faint detection even in the smoothed J band image. Asummary of the observations is given in Table 1. We identify the single emission line detected in the GMOSspectrum as Ly α λ , giving a redshift of z = . ± . ,which is shown in Figure 6. Other plausible identificationsof this emission line could be [O iii ] λ , giving a redshiftof z ≈ . or H α λ at z ≈ . . These can be ruled outgiven the non-detection of other bright lines expected in the Table 1.
Observation log.Telescope Instrument Date Exp. time (sec)Gemini N GMOS long-slit 28-04-2017 2400 ( × s)LBT LUCI J-band 01-02-2018 720 ( × s)09-05-2018 3600 ( × s)Total 4320LBT LUCI Ks-band 01-02-2018 1200 ( × s)09-05-2018 3000 ( × s)Total 4200 wavelength range covered. For example, if the observed lineis [O iii ] at z = . , then the [O ii ] λ line would beobserved at ∼ ˚A, unless the galaxy has a particularlyhigh [O iii ]/[O ii ] ratio as in the case of very low metallic-ity objects. An unresolved [O ii ] λλ , doublet at aredshift of z ≈ . could be a possibility, but the absence ofother expected UV/optical lines common in AGN and radiogalaxy spectra, such as C ii ] λ or Mg ii λλ , ,which are on average a factor of − times fainter than [O ii ] (De Breuck et al. 2001), makes this possibility unlikely.Another possibility could be C iv λ giving z = . , butthis can be ruled out because in this case Ly α , which is oftenmuch brighter (for example, there are no radio galaxies inDe Breuck et al. (2001) with Ly α flux lower than C iv λ flux), would be expected at ∼ ˚A and we do not see anysigns of an emission line in that region, which is free frombright sky lines.We fit a Gaussian to the emission line (shown in Figure6) to measure an integrated line flux of F Ly α = . ± . × − erg s − cm − . The total measured Ly α luminosity is L Ly α = . ± . × erg s − . The full width at half maxi-mum (FWHM) after correcting for the instrumental FWHMis ± km s − . Since no continuum is detected in the spec-trum (down to σ depth of . × − erg s − cm − ), we canonly put a lower limit on the rest-frame equivalent width(EW) of the line, EW > ˚A. Table 2 presents a summaryof emission line measurements for TGSS J1530+1049. To further confirm our redshift determination, we quantifythe asymmetry of the emission line following the prescrip-tions laid out by Kashikawa et al. (2006), by calculating theS-statistic and the weighted skewness parameter. A mea-sure of the skewness of the emission line is particularly use-ful when dealing with spectra with a single emission lineand can help differentiate Ly α emission at high redshiftsfrom [O ii ], [O iii ] or H α emission from lower redshift galax-ies (Rhoads et al. 2003; Kurk et al. 2004; Kashikawa et al.2006). We measure the skewness S = . ± . and theweighted skewness S w = . ± . ˚A, which are consistentwith what is observed for confirmed Ly α emitters at highredshift (Kashikawa et al. 2006, 2011; Matthee et al. 2017).To check what possible values of skewness could be ob-tained from an unresolved [O ii ] doublet, we simulated thedoublet with all possible ratios ( . < j λ / j λ < . ),convolved with the instrument resolution. We find that theskewness measured for the emission line seen in the spectrum MNRAS , 1–11 (2018) adio galaxy at z = 5.72 Figure 3.
Raw (top panels) and sky subtracted (bottom panels) 2D frames shown for the three individual exposures taken using GMOSon Gemini. Traces of the emission line are visible in all three frames, ensuring that the detected line is real and not a consequence ofcosmic rays or artefacts. There is some cosmic ray residual left over in the second frame but that does not contaminate the emission linesignal.
Figure 4.
Extracted 1D spectrum showing the single emissionline detection centred at 8170 ˚A from the GMOS 2D spectrum.No other line or continuum is detected. Shaded regions mark thepresence of sky lines in the spectrum. ( S = . ) is only possible for j λ / j λ < . . These lineratios correspond to the high electron density regime whenthe line would be collisionally de-excited, and hence unlikelyto be as strong as observed, with previous studies of the [O ii ] doublet in high-z galaxies (Steidel et al. 2014; Shimakawaet al. 2015; Sanders et al. 2016) also finding much higher lineratios on average. This helps drive the interpretation of theobserved emission line more towards a Ly α at high redshift. Further, an EW > ˚A for an [O ii ] line originating froma presumably massive radio galaxy at z ≈ . would be atthe extreme end of the EW distribution (Bridge et al. 2015),including for radio-loud quasars (Kalfountzou et al. 2012).This EW value is also incompatible with the line ratios thatwould give rise to the observed skewness, as in regions of veryhigh electron densities the [O ii ] line is expected to be weakerdue to collisional de-excitation. Therefore, we can practicallyrule out the [O ii ] doublet as a possible identification of thisemission line. An EW > ˚A, however, is typical for Ly α emission seen in galaxies at z ≈ . (see Kashikawa et al.2011, for example) and generally consistent with the z ∼ galaxy population (De Barros et al. 2017). Finally, a strong indicator of a high-redshift nature of thehost galaxy is the non-detection in K band down to a σ lim-iting magnitude of 22.4 (Figure 5) using aperture photome-try at the peak pixel of the radio emission. For comparison,TN J0924 − z = . has a magnitude of K = . ± . and our measurement of K > . is consistent with z > and helps rule out lower redshifts owing to the K − z relationfor radio galaxies (note that the luminosity and the spectralindex rule out that it is a star-forming galaxy). We expandupon this point in Section 5.3. The additional non-detectionin J band down to a σ limit of 24.3 further favours a highredshift galaxy and supports the argument that the line wesee is indeed Ly α and not [O ii ]. MNRAS000
Extracted 1D spectrum showing the single emissionline detection centred at 8170 ˚A from the GMOS 2D spectrum.No other line or continuum is detected. Shaded regions mark thepresence of sky lines in the spectrum. ( S = . ) is only possible for j λ / j λ < . . These lineratios correspond to the high electron density regime whenthe line would be collisionally de-excited, and hence unlikelyto be as strong as observed, with previous studies of the [O ii ] doublet in high-z galaxies (Steidel et al. 2014; Shimakawaet al. 2015; Sanders et al. 2016) also finding much higher lineratios on average. This helps drive the interpretation of theobserved emission line more towards a Ly α at high redshift. Further, an EW > ˚A for an [O ii ] line originating froma presumably massive radio galaxy at z ≈ . would be atthe extreme end of the EW distribution (Bridge et al. 2015),including for radio-loud quasars (Kalfountzou et al. 2012).This EW value is also incompatible with the line ratios thatwould give rise to the observed skewness, as in regions of veryhigh electron densities the [O ii ] line is expected to be weakerdue to collisional de-excitation. Therefore, we can practicallyrule out the [O ii ] doublet as a possible identification of thisemission line. An EW > ˚A, however, is typical for Ly α emission seen in galaxies at z ≈ . (see Kashikawa et al.2011, for example) and generally consistent with the z ∼ galaxy population (De Barros et al. 2017). Finally, a strong indicator of a high-redshift nature of thehost galaxy is the non-detection in K band down to a σ lim-iting magnitude of 22.4 (Figure 5) using aperture photome-try at the peak pixel of the radio emission. For comparison,TN J0924 − z = . has a magnitude of K = . ± . and our measurement of K > . is consistent with z > and helps rule out lower redshifts owing to the K − z relationfor radio galaxies (note that the luminosity and the spectralindex rule out that it is a star-forming galaxy). We expandupon this point in Section 5.3. The additional non-detectionin J band down to a σ limit of 24.3 further favours a highredshift galaxy and supports the argument that the line wesee is indeed Ly α and not [O ii ]. MNRAS000 , 1–11 (2018)
A. Saxena et al.
Figure 5. K -band image from the Large Binocular Telescope(LBT), which has been smoothed with a × pixel Gaussiankernel, with radio contours (same as Figure 2) from the VLA at1.4 GHz overlaid. The measured magnitude at the radio positionwith a 1.5 arcsecond aperture is fainter than the σ depth of theimage, giving K > . . When the image is smoothed, however,faint emission is visible around the peak of the radio emission.The magnitude limit is consistent with z > following the K − z relation for radio galaxies. For comparison, the z = . radiogalaxy TN J09224 − Table 2.
Spectroscopic redshift and emission line measurementsfor TGSS J1530+1049 through GMOS spectroscopy.Property Measurement z spec . ± . F Ly α . ± . × − erg s − cm − L Ly α . ± . × erg s − FWHM ± km s − EW obs > ˚A The Ly α luminosity and FWHM measured for TGSSJ1530+1049 are lower than what is seen for typical HzRGsat z > (see Spinrad et al. 1995; De Breuck et al. 1999; vanBreugel et al. 1999; Miley & De Breuck 2008, for examples)and more consistent with those measured for ‘non-radio’Ly α emitting galaxies (LAEs) at this redshift (Rhoads et al.2003; Ouchi et al. 2008; Kashikawa et al. 2011; Lidman et al.2012; Matthee et al. 2017). However, the FWHM for TGSSJ1530+1049 is consistent with that of a very faint radiogalaxy VLA J123642 + S = . mJy, discovered at z = . (Waddingtonet al. 1999). This galaxy has a FWHM of ≈ km s − anda Ly α luminosity ≈ × erg s − , which is weaker thanTGSS J1530+1049. VLA J123642 + Figure 6.
Rest-frame 1D spectrum showing the asymmetric Ly α line profile at a redshift of z = . . Also shown is the best-fitGaussian to the emission line. The peak of the fitted Gaussian isslightly redder than the peak of the line, suggesting asymmetryin the emission line. This is also clear from the excess towardsthe redder parts of the Gaussian. Top:
The 2D GMOS spectrumshowing the detected Ly α line. The spatial extent of the emissionis roughly 1 arcsecond, which is also the aperture size used toextract the 1D spectrum. detected in TGSS at 150 MHz down to a noise level of 3.5mJy beam − , suggesting a relatively flat spectral index or aspectral turnover at low radio frequencies. We present somecomparisons of the Ly α properties we measure for TGSSJ1530+1049 with other HzRGs at z > and also non-radioLAEs at z = . in Table 3.A statistical sample of radio galaxies at z ∼ is neededto understand whether they are more like LAEs at high red-shift or whether a majority of them continue being very dif-ferent systems, surrounded by extremely overdense regionsand forming stars intensively. The relatively underluminousLy α would be one signature of a significantly neutral inter-galactic medium (IGM) during the late stages of the EoR.Weaker Ly α emission may also be caused by significant ab-sorption in a cold and dusty medium surrounding the radiogalaxy. The presence of cold gas and dust has been reportedin many HzRGs, including TN J0924 − z > are required to better characterisetheir surrounding medium. TGSS J1530+1049 has a flux density of 170 mJy at a fre-quency of 150 MHz and 7.5 mJy at 1.4 GHz (Saxena et al.2018). Using the standard K -corrections in radio astronomyand assuming a constant spectral index of α = − . , wecalculate a rest-frame radio luminosity of log L
150 MHz = . and log L = . W Hz − , which places this MNRAS , 1–11 (2018) adio galaxy at z = 5.72 Table 3.
Comparison of Ly α emission line properties of TGSS J1530+1049 reported in this paper with some of the known radio galaxiesat z > , and the much fainter radio galaxy at z = . (all marked as RG), in addition to several confirmed LAEs at z ≈ . from theliterature. F Ly α L Ly α FWHM Ly α Name z ( × − erg s − cm − ) ( × erg s − ) (km s − ) ReferenceTGSS J1530+1049 5.72 1.6 5.7 370 This workTN J0924 − + + + + source at the most luminous end of the radio luminosityfunction at this epoch (Saxena et al. 2017). For compar-ison, TN J0924 − K -corrected radio luminosityof log L = . W Hz − using a spectral index of α = − . (van Breugel et al. 1999). TGSS J1530+1049 isclose to an order of magnitude fainter than TN J09224 − − z ∼ from Saxena et al. (2017), as radio galaxies inthe early Universe are expected to be young and very com-pact (Blundell et al. 1999). In Table 4 we compare the radioproperties of TGSS J1530+1049 with all currently known ra-dio galaxies at z > . This was done by querying the TGSSADR1 catalog to determine flux densities for all z > radiogalaxies at 150 MHz, which were then used to calculate ra-dio powers using the standard K -corrections. We find thatTGSS J1530+1049 is comparable to many of the z > radiogalaxies when looking at radio properties alone.TGSS J1530+1049 has a spectral index of α
150 MHz1.4 GHz = − . , which is ultra-steep but flatter than TN J0924 − z = . , which was selected because of its spectral in-dex of α
365 MHz1.4 GHz = − . . Interestingly, at lower radio frequen-cies the spectral index of TN J0924 − − ± mJy,giving a low frequency spectral index α
150 MHz365 MHz = − . . Ifthe spectral index were to be calculated only using the fluxdensities at frequencies of 150 MHz and 1.4 GHz, the in-ferred spectral index would be α
150 MHz1.4 GHz = − . , making itnot strictly ultra-steep ( α < − . ). This implies that in asearch for ultra-steep spectrum radio sources using data at150 MHz and 1.4 GHz, such as Saxena et al. (2018), TNJ0924 − z > radio galaxies between frequencies of 150 MHz and 1.4 GHzin Table 4. A large majority of these radio sources were se-lected for having an ultra-steep spectral index in a higher frequency range, but appear to have a flatter spectral indexwhen calculated between 150 MHz and 1.4 GHz.Spectral flattening or even a turnover at low radio fre-quencies is expected in radio galaxies at increasingly higherredshifts due to: a) Inverse Compton (IC) losses due to thedenser cosmic microwave background that affect the higherfrequencies and result in a steeper high frequency spec-tral index, and b) free-free or synchrotron self absorptiondue to the compact sizes of radio sources at high redshiftsthat can lead to a turnover in the low frequency spectrum(see Callingham et al. 2017, and references therein). Sax-ena et al. (2018) have reported evidence of flattening of thelow-frequency spectral index in candidate HzRGs and ob-servations at intermediate radio wavelengths for sources likeTGSS J1530+1049 are essential to measure spectral flatten-ing and constrain various energy loss mechanisms that dom-inate the environments of radio galaxies in the early Uni-verse. Additionally, search techniques for radio galaxies ateven higher redshifts could be refined by possibly using radiocolours instead of a simple ultra-steep spectral index selec-tion. The LOFAR Two Metre Sky Survey (Shimwell et al.2017, Shimwell et al. submitted) will eventually provide in-band spectral indices at 150 MHz and could potentially beused to identify HzRG candidates more efficiently.We also draw attention towards the radio galaxyJ163912.11 + z = . (Jarvis et al. 2009), thathas a spectral index of α
325 MHz1.4 GHz = − . and is not anultra-steep spectrum radio source. Interestingly, there is ev-idence of spectral flattening at lower frequencies with a 150MHz flux density of 103.5 mJy, giving a spectral index α
150 MHz325 MHz = − . , which is flatter than that at higher fre-quencies. This source was targeted for spectroscopic follow-up owing to the faintness of its host galaxy at . µ m . Thevery faint radio galaxy VLA J123642 + z = . (Waddington et al. 1999) is also not strictly an ultra-steepspectrum source at high radio frequencies ( α = − . )and is too faint to be detected in TGSS. This source wasalso selected based on its optical and infrared faintness, sug-gesting that a considerable fraction of HzRGs may not beultra-steep at all and therefore, be missed in samples con-structed using the ultra-steep spectrum selection technique.Indeed Ker et al. (2012) have shown that selectinginfrared-faint radio sources (IFRS) could be more efficient atisolating HzRGs from large samples when compared to ra- MNRAS , 1–11 (2018)
A. Saxena et al.
Table 4.
A comparison of the radio properties of TGSS J1530+1049 with other known radio galaxies at z > . Flux densities at 150MHz are measured from the TGSS catalog. S log L S SizeName z (mJy) (W Hz − ) (mJy) α (kpc) ReferenceTGSS J1530+1049 5.72 170 29.1 7.5 − . − − . + − . − (Jarvis et al. 2009)RC J0311 + − . + − − − (Waddington et al. 1999)6C 0140 +
326 4.41 860 29.4 91.0 − . +
635 4.25 8070 30.4 497.0 − . − − . − − . dio selection alone. This has been confirmed observationallythrough many previous studies (Norris et al. 2006; Collieret al. 2014; Herzog et al. 2014; Maini et al. 2016; Singh et al.2017). However, the caveat is that deep infrared photome-try over large sky areas is required to effectively implementsuch a selection, which can be expensive. The recently con-cluded UKIRT Hemisphere Survey (UHS; Dye et al. 2018)has the potential to be extremely useful in the identifica-tion of promising HzRG candidates in the Northern Hemi-sphere, particularly from the LOFAR surveys (Shimwellet al. 2017, Shimwell et al. submitted), as a combinationof a relaxed spectral index steepness criterion and infrared-faintness could be more effective at isolating HzRGs. The non-detection of the host galaxy down to a σ depthof K = . can be used to set limits on the stellar massfor TGSS J1530+1049 using simple stellar population syn-thesis modelling. To do this, we make use of the python package smpy , which is designed for building compositestellar populations in an easy and flexible manner, allowingfor synthetic photometry to be produced for single or largesuites of models (see Duncan & Conselice 2015). To buildstellar populations, we use the Bruzual & Charlot (2003)model with a Chabrier (2003) initial mass function (IMF)and solar metallicity (Willott et al. 2003), a formation red-shift z f = and assume a maximally old stellar populationthat has been forming stars at a constant rate (Lacy et al.2000). We follow the Calzetti et al. (2000) law for dust at-tenuation and use values of A v = . (moderate extinction)and . mag (dusty), which are commonly seen in massivegalaxies at < z < (McLure et al. 2006). The syntheticphotometry is produced for different stellar masses, whichwe then convolve with the K band filter to calculate appar-ent K magnitudes over a redshift range − .The K magnitude limit for TGSS J1530+1049 fits wellwith a stellar mass limit of M stars < ∼ . M (cid:12) for A v = . mag, and M stars < ∼ . for A v = . mag.We note here that thanks to the excellent seeing for K -bandobservations ( . − . arcseconds), and since at z ∼ thehost galaxy is expected to be small, any aperture correction https://github.com/dunkenj/smpy Figure 7.
The ‘ K − z ’ diagram for radio galaxies, showing stellarmass limits derived from stellar population synthesis modellingfor TGSS J1530+1049 (black triangle). The K -band σ limit gives M stars < . M (cid:12) for A v = . mag, and M stars < . for A v = . mag. Also shown are K -band magnitudes and redshiftsfor known radio galaxies in the literature (grey points; see text),with TN J0924 − z = . (orange circle). The K -bandlimits for TGSS J1530+1049 further help exclude lower redshiftmeasurements from incorrect line identification. is only expected to be at the level of a few tenths of a mag-nitude at most, or . − . in the logarithmic stellar mass,which is smaller than the uncertainty from dust extinctioncorrections.We find that the stellar mass limits we infer are in agree-ment with the J band σ limit from LBT. The photometrypredicted by the models in the optical bands from PS1 ( g , r , i , z , y ) is also consistent with the non-detections thatwe report. This stellar mass limit places TGSS J1530+1049towards the > M ∗ end of the galaxy stellar mass functionat z ∼ (see Duncan et al. 2014, for example). For com-parison, we show the apparent K band magnitudes of otherradio galaxies in the literature, taken from the 3CRR, 6CE,6C* and 7C − I/II/III samples (Willott et al. 2003), in Figure7. Also shown is the K magnitude for TN J09224 − z = . (van Breugel et al. 1999), which is best fit with astellar mass of . M (cid:12) for A v = . mag and M (cid:12) for A v = . mag. MNRAS , 1–11 (2018) adio galaxy at z = 5.72 We also show the K -band magnitude limit for theUKIDSS LAS as a dashed black line in Figure 7. TGSSJ1530+1049 was initially selected due to its non-detectionin LAS. However, these magnitude limits alone were not suf-ficient to constrain the very high redshift nature of the hostgalaxy. With deeper LBT observations in K band, we showthat TGSS J1530+1049 follows the trend in the K − z plotfor radio galaxies. It is also clear that a low redshift solutionthat would arise if the detected emission line in Section 4is not Lyman alpha (for example, z ≈ . if the line is [O ii ]) would be hard to explain using galaxy evolution modelswith the inputs and assumptions outlined above and thosegenerally used to model radio galaxy spectra (Overzier et al.2009).A stellar mass limit of M (cid:63) < . M (cid:12) for TGSSJ1530+1049 (based on the outlined assumptions) is almostan order of magnitude lower than the sample of radio galax-ies studied by Rocca-Volmerange et al. (2004), where thehighest stellar masses are seen for radio galaxies in the red-shift range − . This suggests that TGSS J1530+1049 isstill evolving and is in the process of building up its stellarmass. This interpretation is in line with TGSS J1530+1049being a relatively young radio galaxy owing to its small radiosize, and lower stellar masses may be expected from radiogalaxies close to or into the epoch of reionisation. Robustdetections at optical and infrared wavelengths are requiredto properly characterise the galaxy spectral energy distribu-tion in order to understand better the star formation historyof TGSS J1530+1049. In this paper we have presented the discovery of the highestredshift radio galaxy, TGSS J1530+1049, at z = . . Thegalaxy was initially selected at 150 MHz from TGSS (Saxenaet al. 2018) and was assigned a high priority for spectroscopicfollow-up owing to its compact morphology and faintness atoptical and near-infrared wavelengths. The conclusions ofthis study are listed below:(i) Long-slit spectroscopy centered at the radio positionof the source revealed an emission line at 8170 ˚A, which weidentify as Lyman alpha at z = . . We rule out alter-native line IDs owing to the absence of other optical/UVlines in the spectrum, the asymmetrical nature of the emis-sion line characteristic of Lyman alpha at high redshifts thatwe quantify using the skewness parameter and the high ob-served equivalent width of the emission line.(ii) Deep J and K band imaging using the Large BinocularTelescope led to no significant detection of the host galaxydown to σ limits of K > . and J > . . The limits in K can be used as an additional constraint on the redshift, owingto the relation that exists between K band magnitude andredshift of radio galaxies. The magnitude limit is consistentwith z > , practically ruling out a redshift of z ≈ . thatwould be expected if the emission line were an unresolved [O ii ] λλ , doublet, which is the most likely alternativeline identification.(iii) The emission line is best fitted with a skewed Gaus-sian, giving an integrated line flux of F Ly α = . × − ergs − cm − , a Ly α luminosity of . × erg s − , an equiva- lent width of EW > ˚A and a FWHM of km s − . Thesevalues are more consistent with those observed in non-radioLyman alpha emitting galaxies at this redshift and muchlower than those corresponding to typical radio galaxies at z > .(iv) The radio luminosity calculated at 150 MHz is log L = . W Hz − , which places it at the most lu-minous end of the radio luminosity function at this epoch.The deconvolved angular size is 3.5 kpc, which is in linewith the compact morphologies expected at high redshifts.We find that the radio properties of TGSS J1530+1049 arecomparable to other known radio galaxies at z > but thecompact size suggests that it is a radio galaxy in an earlyphase of its evolution. A joint study of the Ly α halo andthe radio size of this source may provide one of the earliestconstraints on the effects of radio-mode feedback.(v) We use the K band limit to put constraints on thestellar mass estimate using simple stellar population syn-thesis models. Assuming a constant star formation historyand a maximally old stellar population, we derive a stel-lar mass limit of M stars < ∼ . M (cid:12) for A v = . mag,and M stars < ∼ . for A v = . mag. These limits are al-most an order of magnitude lower than typical radio galaxymasses in the redshift range − and suggest that TGSSJ1530+1049 may still be in the process of assembling itsstellar mass, which is in line with it being a relatively youngradio galaxy.An effective application of deep radio surveys coveringvery large areas on the sky has been demonstrated by thisdiscovery of the first radio galaxy at a record distance afteralmost 20 years. With the more sensitive, large area surveyscurrently underway with LOFAR (LoTSS; Shimwell et al.2017, Shimwell et al. submitted), there is potential to pushsearches for radio galaxies to even higher redshifts. Discoveryof even a single bright radio galaxy at z > would open upnew ways to study the epoch of reionisation in unparalleleddetail, through searches for the 21cm absorption features leftbehind by the neutral hydrogen that pervaded the Universeat high redshifts. ACKNOWLEDGEMENTS
The authors thank the referee for useful comments and sug-gestions. AS would like to thank Jorryt Matthee, David So-bral, Reinout van Weeren and Nobunari Kashikawa for fruit-ful discussions. AS, HJR and KJD gratefully acknowledgesupport from the European Research Council under the Eu-ropean Unions Seventh Framework Programme (FP/2007-2013)/ERC Advanced Grant NEWCLUSTERS-321271.RAO and MM received support from CNPq (400738/2014-7,309456/2016-9) and FAPERJ (202.876/2015). PNB is grate-ful for support from STFC via grant ST/M001229/1. IP ac-knowledges funding from the INAF PRIN-SKA 2017 project1.05.01.88.04 (FORECaST).This paper is based on results from observations ob-tained at the Gemini Observatory, which is operated by theAssociation of Universities for Research in Astronomy, Inc.,under a cooperative agreement with the NSF on behalf ofthe Gemini partnership: the National Science Foundation(United States), the National Research Council (Canada),
MNRAS000
MNRAS000 , 1–11 (2018) A. Saxena et al.
CONICYT (Chile), Ministerio de Ciencia, Tecnolog´ıa e In-novaci´on Productiva (Argentina), and Minist´erio da Ciˆen-cia, Tecnologia, Inova¸c˜ao e Comunica¸c˜oes (Brazil). This pa-per also contains data from the Large Binocular Telescope(LBT), an international collaboration among institutions inthe United States, Italy and Germany. LBT Corporationpartners are: The University of Arizona on behalf of theArizona university system; Istituto Nazionale di Astrofisica,Italy; LBT Beteiligungsgesellschaft, Germany, representingthe Max-Planck Society, the Astrophysical Institute Pots-dam, and Heidelberg University; The Ohio State University,and The Research Corporation, on behalf of The Universityof Notre Dame, University of Minnesota, and University ofVirginia.This work has made extensive use of ipython (P´erez& Granger 2007), astropy (Astropy Collaboration et al.2013), aplpy (Robitaille & Bressert 2012), matplotlib (Hunter 2007) and topcat (Taylor 2005). This work wouldnot have been possible without the countless hours put inby members of the open-source developing community allaround the world.
REFERENCES
Afonso J., et al., 2011, ApJ, 743, 122Alam S., et al., 2015, ApJS, 219, 12Astropy Collaboration et al., 2013, A&A, 558, A33Ba˜nados E., et al., 2015, ApJ, 804, 118Best P. N., et al., 1998, MNRAS, 301, L15Blundell K. M., Rawlings S., Willott C. J., 1999, AJ, 117, 677Bradley L., et al., 2017, astropy/photutils: v0.4,doi:10.5281/zenodo.1039309, https://doi.org/10.5281/zenodo.1039309
Bridge J. S., et al., 2015, ApJ, 799, 205Bruzual G., Charlot S., 2003, MNRAS, 344, 1000Callingham J. R., et al., 2017, ApJ, 836, 174Calzetti D., Armus L., Bohlin R. C., Kinney A. L., Koornneef J.,Storchi-Bergmann T., 2000, ApJ, 533, 682Carilli C. L., Harris D. E., Pentericci L., R¨ottgering H. J. A.,Miley G. K., Kurk J. D., van Breugel W., 2002a, ApJ, 567,781Carilli C. L., Gnedin N. Y., Owen F., 2002b, ApJ, 577, 22Chabrier G., 2003, PASP, 115, 763Chambers K. C., et al., 2016, preprint, ( arXiv:1612.05560 )Ciardi B., et al., 2015, MNRAS, 453, 101Collier J. D., et al., 2014, MNRAS, 439, 545De Barros S., et al., 2017, A&A, 608, A123De Breuck C., van Breugel W., Minniti D., Miley G., R¨ottgeringH., Stanford S. A., Carilli C., 1999, A&A, 352, L51De Breuck C., van Breugel W., R¨ottgering H. J. A., Miley G.,2000, A&AS, 143, 303De Breuck C., et al., 2001, AJ, 121, 1241De Breuck C., et al., 2010, ApJ, 725, 36Dewdney P. E., Hall P. J., Schilizzi R. T., Lazio T. J. L. W., 2009,IEEE Proceedings, 97, 1482Duncan K., Conselice C. J., 2015, MNRAS, 451, 2030Duncan K., et al., 2014, MNRAS, 444, 2960Dunlop J. S., Peacock J. A., 1990, MNRAS, 247, 19Dye S., et al., 2018, MNRAS, 473, 5113Ewall-Wice A., Dillon J. S., Mesinger A., Hewitt J., 2014, MN-RAS, 441, 2476Furlanetto S. R., Loeb A., 2002, ApJ, 579, 1Hatch N. A., et al., 2011, MNRAS, 410, 1537Herzog A., Middelberg E., Norris R. P., Sharp R., Spitler L. R.,Parker Q. A., 2014, A&A, 567, A104 Hunter J. D., 2007, Computing In Science & Engineering, 9, 90Intema H. T., Jagannathan P., Mooley K. P., Frail D. A., 2017,A&A, 598, A78Ishwara-Chandra C. H., Sirothia S. K., Wadadekar Y., Pal S.,Windhorst R., 2010, MNRAS, 405, 436Jarvis M. J., Rawlings S., Eales S., Blundell K. M., Bunker A. J.,Croft S., McLure R. J., Willott C. J., 2001, MNRAS, 326,1585Jarvis M. J., Teimourian H., Simpson C., Smith D. J. B., RawlingsS., Bonfield D., 2009, MNRAS, 398, L83Kalfountzou E., Jarvis M. J., Bonfield D. G., Hardcastle M. J.,2012, MNRAS, 427, 2401Kashikawa N., et al., 2006, ApJ, 648, 7Kashikawa N., et al., 2011, ApJ, 734, 119Ker L. M., Best P. N., Rigby E. E., R¨ottgering H. J. A., GendreM. A., 2012, MNRAS, 420, 2644Klamer I. J., Ekers R. D., Sadler E. M., Weiss A., Hunstead R. W.,De Breuck C., 2005, ApJ, 621, L1Kurk J. D., Cimatti A., di Serego Alighieri S., Vernet J., DaddiE., Ferrara A., Ciardi B., 2004, A&A, 422, L13Lacy M., et al., 1994, MNRAS, 271, 504Lacy M., Bunker A. J., Ridgway S. E., 2000, AJ, 120, 68Lawrence A., et al., 2007, MNRAS, 379, 1599Lidman C., Hayes M., Jones D. H., Schaerer D., Westra E., Tap-ken C., Meisenheimer K., Verhamme A., 2012, MNRAS, 420,1946Lilly S. J., Longair M. S., 1984, MNRAS, 211, 833Mack K. J., Wyithe J. S. B., 2012, MNRAS, 425, 2988Maini A., Prandoni I., Norris R. P., Spitler L. R., Mignano A.,Lacy M., Morganti R., 2016, A&A, 596, A80Matthee J., Sobral D., Darvish B., Santos S., Mobasher B.,Paulino-Afonso A., R¨ottgering H., Alegre L., 2017, MNRAS,472, 772McLure R. J., et al., 2006, MNRAS, 372, 357Miley G., De Breuck C., 2008, A&ARv, 15, 67Miley G. K., et al., 2004, Nature, 427, 47Morabito L. K., et al., 2017, submittedNorris R. P., et al., 2006, AJ, 132, 2409Orsi ´A. A., Fanidakis N., Lacey C. G., Baugh C. M., 2016, MN-RAS, 456, 3827Ouchi M., et al., 2008, ApJS, 176, 301Overzier R. A., et al., 2009, ApJ, 704, 548Parijskij Y. N., et al., 2014, MNRAS, 439, 2314Pentericci L., et al., 2000, A&A, 361, L25P´erez F., Granger B. E., 2007, Computing in Science and Engi-neering, 9, 21Rawlings S., Lacy M., Blundell K. M., Eales S. A., Bunker A. J.,Garrington S. T., 1996, Nature, 383, 502Reuland M., R¨ottgering H., van Breugel W., De Breuck C., 2004,MNRAS, 353, 377Rhoads J. E., et al., 2003, AJ, 125, 1006Rigby E. E., Best P. N., Brookes M. H., Peacock J. A., DunlopJ. S., R¨ottgering H. J. A., Wall J. V., Ker L., 2011, MNRAS,416, 1900Rigby E. E., Argyle J., Best P. N., Rosario D., R¨ottgering H. J. A.,2015, A&A, 581, A96Robitaille T., Bressert E., 2012, APLpy: Astronomical Plot-ting Library in Python, Astrophysics Source Code Library(ascl:1208.017)Rocca-Volmerange B., Le Borgne D., De Breuck C., Fioc M., MoyE., 2004, A&A, 415, 931R¨ottgering H. J. A., Lacy M., Miley G. K., Chambers K. C.,Saunders R., 1994, A&AS, 108R¨ottgering H., Daddi E., Overzier R., Wilman R., 2003, New As-tron. Rev., 47, 309Sanders R. L., et al., 2016, ApJ, 816, 23Saxena A., R¨ottgering H. J. A., Rigby E. E., 2017, MNRAS, 469,4083 MNRAS , 1–11 (2018) adio galaxy at z = 5.72 Saxena A., et al., 2018, MNRAS, 475, 5041Seifert W., et al., 2003, in Iye M., Moorwood A. F. M., eds,Proc. SPIEVol. 4841, Instrument Design and Performancefor Optical/Infrared Ground-based Telescopes. pp 962–973,doi:10.1117/12.459494Shibuya T., et al., 2018, PASJ, 70, S15Shimakawa R., et al., 2015, MNRAS, 451, 1284Shimwell T. W., et al., 2017, A&A, 598, A104Singh V., Wadadekar Y., Ishwara-Chandra C. H., Sirothia S.,Sievers J., Beelen A., Omont A., 2017, MNRAS, 470, 4956Smith D. J. B., et al., 2016, in Reyl´e C., Richard J., Cam-br´esy L., Deleuil M., P´econtal E., Tresse L., Vauglin I.,eds, SF2A-2016: Proceedings of the Annual meeting of theFrench Society of Astronomy and Astrophysics. pp 271–280( arXiv:1611.02706 )Spinrad H., Dey A., Graham J. R., 1995, ApJ, 438, L51Steidel C. C., et al., 2014, ApJ, 795, 165Swarup G., Ananthakrishnan S., Kapahi V. K., Rao A. P., Sub-rahmanya C. R., Kulkarni V. K., 1991, Current Science,Vol. 60, NO.2/JAN25, P. 95, 1991, 60, 95Taylor M. B., 2005, in Shopbell P., Britton M., Ebert R., eds,Astronomical Society of the Pacific Conference Series Vol. 347,Astronomical Data Analysis Software and Systems XIV. p. 29Tingay S. J., et al., 2013, Publ. Astron. Soc. Australia, 30, e007Venemans B. P., et al., 2002, ApJ, 569, L11Waddington I., Windhorst R. A., Cohen S. H., Partridge R. B.,Spinrad H., Stern D., 1999, ApJ, 526, L77Willott C. J., Rawlings S., Blundell K. M., Lacy M., Eales S. A.,2001, MNRAS, 322, 536Willott C. J., Rawlings S., Jarvis M. J., Blundell K. M., 2003,MNRAS, 339, 173Wright E. L., et al., 2010, AJ, 140, 1868Xu Y., Chen X., Fan Z., Trac H., Cen R., 2009, ApJ, 704, 1396van Breugel W., De Breuck C., Stanford S. A., Stern D., R¨ottger-ing H., Miley G., 1999, ApJ, 518, L61van Haarlem M. P., et al., 2013, A&A, 556, A2This paper has been typeset from a TEX/L A TEX file prepared bythe author.MNRAS000