Remnant Radio Galaxies Discovered in a Multi-frequency Survey
B. Quici, N. Hurley-Walker, N. Seymour, R. J. Turner, S. S. Shabala, M. Huynh, H. Andernach, A. D. Kapi?ska, J. D. Collier, M. Johnston-Hollitt, S. V. White, I. Prandoni, T. J. Galvin, T. Franzen, C. H. Ishwara-Chandra, S. Bellstedt, S. J. Tingay, B. M. Gaensler, A. O'Brien, J. Rogers, K. Chow, S. Driver, A. Robotham
PPublications of the Astronomical Society of Australia (PASA)doi: 10.1017/pas.2021.xxx.
Remnant Radio Galaxies Discovered in aMulti-frequency Survey
B. Quici , N. Hurley-Walker , N. Seymour , R. J. Turner , S. S. Shabala , , M. Huynh ,H. Andernach , A. D. Kapińska , J. D. Collier , , M. Johnston-Hollitt , S. V. White , ,I. Prandoni , T. J. Galvin , T. Franzen , , C. H. Ishwara-Chandra , S. Bellstedt , S. J. Tingay ,B. M. Gaensler , A. O’Brien , , , J. Rogers , K. Chow , S. Driver , and A. Robotham International Centre for Radio Astronomy Research, Curtin University, Bentley, WA 6102, Australia School of Natural Sciences, University of Tasmania, Private Bag 37, Hobart, 7001, Australia ARC Centre of Excellence for All Sky Astrophysics in 3 Dimensions (ASTRO 3D) CSIRO Astronomy and Space Science, 26 Dick Perry Avenue, Kensington, WA 6151, Australia IFUG, Universidad de Guanajuato, Guanajuato, C.P. 36000, Mexico" for "DCNE, Universidad de Guanajuato, Cjón. deJalisco s/n, Guanajuato, C.P. 36023, Mexico National Radio Astronomy Observatory, 1003 Lopezville Road, Socorro, NM 87801, USA CSIRO Astronomy and Space Science (CASS), Marsfield, NSW 2122, Australia School of Science, Western Sydney University, Locked Bag 1797, Penrith, NSW 2751, Australia The Inter-University Institute for Data Intensive Astronomy (IDIA), Department of Astronomy, University of Cape Town,Private Bag X3, Rondebosch, 7701, South Africa Department of Physics and Electronics, Rhodes University, PO Box 94,Grahamstown, 6140, South Africa Istituto di Radioastronomia, Via P. Gobetti 101, 40129, Italy ASTRON: the Netherlands Institute for Radio Astronomy, PO Box 2, 7990 AA, Dwingeloo, The Netherlands National Centre for Radio Astrophysics, TIFR, Post Bag No. 3, Ganeshkhind Post, 411007 Pune, India International Centre for Radio Astronomy Research, M468, University of Western Australia, Crawley, WA 6009, Australia Dunlap Institute for Astronomy and Astrophysics, University of Toronto, 50 St. George Street, Toronto ON M5S 3H4,Canada CSIRO Astronomy and Space Science, PO Box 76, 1710, Epping, NSW, Australia Center for Gravitation, Cosmology, and Astrophysics, Department of Physics, University of Wisconsin-Milwaukee, P.O.Box 413, Milwaukee, WI 53201, USA
Abstract
The remnant phase of a radio galaxy begins when the jets launched from an active galactic nucleusare switched off. To study the fraction of radio galaxies in a remnant phase, we take advantage of a8 .
31 deg sub-region of the GAMA 23 field which comprises of surveys covering the frequency range0.1–9 GHz. We present a sample of 104 radio galaxies compiled from observations conducted by theMurchison Wide-field Array (216 MHz), the Australia Square Kilometer Array Pathfinder (887 MHz),and the Australia Telescope Compact Array (5.5 GHz). We adopt an ‘absent radio core’ criterion toidentify 10 radio galaxies showing no evidence for an active nucleus. We classify these as new candidateremnant radio galaxies. Seven of these objects still display compact emitting regions within the lobes at5.5 GHz; at this frequency the emission is short-lived, implying a recent jet switch-off. On the other hand,only three show evidence of aged lobe plasma by the presence of an ultra-steep spectrum ( α < − . (cid:46) f rem (cid:46) Keywords: galaxies: active – radio continuum: galaxies – methods: statistical a r X i v : . [ a s t r o - ph . GA ] J a n Quici et al.
The jets launched from a radio-loud active galacticnucleus (AGN) arise from the accretion onto a super-massive black hole, and form synchrotron-emitting radiolobes in the intergalactic environments of their hostgalaxies (Scheuer, 1974). Whilst the jets are active, aradio galaxy will often display compact features such asan unresolved radio core coincident with its host galaxy,bi-polar jets, and hotspots in the lobes of Fanaroff Rileytype II ( FR-II; Fanaroff & Riley, 1974) radio galax-ies. The radio continuum spectrum arising from thelobes is usually well approximated by a broken power-law for radio frequencies between 100 MHz and 10 GHz;the observed spectral index, α , typically ranges within − . < α < − .
5. Steepening in the radio lobe spectrumis allowed by ∆ α ≤ . ∼
150 MHz) frequencies, which is comparablebut shorter than the duration of their previous activephase (Shulevski et al., 2017; Turner, 2018; Brienza et al.,2020). Jets are also known to restart after a period ofinactivity (e.g; Roettiger et al., 1994), giving rise to arestarted radio galaxy. Several observational classes existto describe such sources; double-double radio galaxies(DDRG; Schoenmakers et al., 2000) describe sourcesin which two distinct pairs of lobes can be observed,however restarting jets can also appear as compact steepspectrum sources embedded within larger-scale remnantlobes (e.g; Brienza et al., 2018).Compiling samples of remnant (Saripalli et al., 2012;Godfrey et al., 2017; Brienza et al., 2017; Mahatmaet al., 2018) and restarted (Saripalli et al., 2012; Ma-hatma et al., 2019) radio galaxies sheds new light on theirdynamics and evolution, and by extension, the AGN jetduty cycle. Jurlin et al. (2020) present a direct analy-sis of the radio galaxy life cycle, in which their sampleis decomposed into active, remnant and restarted ra-dio galaxies. To complement these observational works,Shabala et al. (2020) present a new methodology inwhich uniformly-selected samples of active, remnant andrestarted radio galaxies are used to constrain evolution- The spectral index α is defined through S ν ∝ ν α ary models describing the AGN jet duty cycle.However, using radio observations to confidently iden-tify radio sources in these phases is a challenging task,even in the modern era of radio instruments. Remnantradio galaxies, which are the focus of this work, displayvarious observational properties that correlate with age,presenting a challenge for identifying complete samples ofsuch sources. Various selection techniques exist amongstthe literature, each with their selection biases. Due tothe red preferential cooling of higher-energy synchrotron-radiating electrons (Jaffe & Perola, 1973; Komissarov& Gubanov, 1994), many authors have identified rem-nants by their ultra-steep radio spectrum ( α < − . α high - α low such that SPC > > . ∼
10 GHz) frequencies.Morphological selection offers a complementary wayto identify remnants, independent of spectral ageingof the lobes. This technique often involves searchingfor low surface brightness (SB) profiles (SB <
50 mJyarcmin − ), amorphous radio morphologies, and an ab-sence of hotspots (e.g; Saripalli et al., 2012; Brienza et al.,2017). However, young remnants in which the hotspotshave not yet disappeared due to a recent switch-off ofthe jets, (e.g 3C 028; Feretti et al., 1984; Harwood et al.,2015), will be missed with these techniques.An alternative approach is to identify remnants basedon an absent radio core; a radio core should be absent ifthe AGN is currently inactive (Giovannini et al., 1988).This property is often invoked to confirm the statusof a remnant radio galaxy, e.g. see Cordey (1987), andis recently employed by Mahatma et al. (2018) as acriterion to search for remnant candidates in a LOwFrequency ARray (LOFAR)-selected sample. The caveathere is the plausibility for a faint radio core to exist belowthe sensitivity of the observations, meaning this methodselects only for remnant candidates. This method willalso miss remnant lobes from a previous epoch of activityin restarted radio galaxies. A likely example of such asource is MIDAS J230304-323228, discussed in Sect 3.These sources are beyond the scope of this work, howeverare a promising avenue for future work.A common aspect of almost all previously-mentionedobservational studies of remnant radio galaxies, is their emnant Radio Galaxies Discovered in a Multi-frequency Survey H =67 . − Mpc − , Ω M =0.307 and Ω Λ =1-Ω M (PlanckCollaboration et al., 2014). All coordinates are reportedin the J2000 equinox. Our sample is focused on a 8.31 deg sub-region of theGAMA 23 field (RA = 345 ◦ , Dec = − . ◦ ), e.g. seeFigure 1. The rich multi-wavelength coverage makes theabsent radio-core criterion a viable method to searchfor remnant radio galaxies, and allows us to match theradio sample with their host galaxies. Below we describe the radio observations used through-out this work, summarised in Table 1 and 2.
From 2013 to 2015, the MWA observed ∼ ofthe sky south of Dec = 30 ◦ . The GaLactic and Extra-galactic All-sky MWA (GLEAM; Wayth et al., 2015)survey adopted a drift scan observing method using the D e c . ( d e g ) GAMA 23VIKINGMIDASuGMRT obs.EMUESGLASS reg. A,B,C,E,FGLASS reg. D (This work)
Figure 1.
Sky coverage of radio surveys dedicated to observingGAMA 23, described in Sect. 2.1. Near-infrared VIKING obser-vations (Sect. 2.2.1) are also displayed. Thick-red and thin-greenarrows indicate the directions in which MIDAS ES and VIKINGextend beyond the represented footprint.
MWA Phase I configuration and a maximum baselineof ∼ . GLEAM spansa 72–231 MHz frequency range, divided evenly into 30.72MHz wide-bands centered at ν c = (cid:8)
88, 119, 154, 185,216 (cid:9)
MHz. Observations conducted during 2014–2015(year 2) covered the same sky area as GLEAM, but witha factor of two increase in the observing time due tooffset ( ± ◦ ≤ RA ≤ ◦ , -48 ◦ ≤ Dec ≤ − ◦ ( ∼ , sky coverage), Franzen et al. (in prep) indepen-dently reprocessed the overlapping observations fromboth years to produce the GLEAM South Galactic Pole(SGP) survey, achieving a factor ∼ ) study. The point spread function (PSF) variesslightly across the survey; at 216 MHz the PSF sizewithin GAMA 23 is approximately 150 × , andan RMS of σ ∼ ± . − is achieved. Thisis consistent with a factor 2 increase in sensitivity overGLEAM. We use an internal GLEAM SGP component GLEAM catalogue publicly available on VizieR: http://cdsarc.u-strasbg.fr/viz-bin/Cat?VIII/100
Quici et al.
Table 1
Summarised properties of the radio surveys spanning the GAMA 23 field. Each column, in ascending order, detailsthe telescope used to conduct the observations, the name of the radio survey, the dates observations were conducted, thecentral frequency of the observing band, the bandwidth available in each band, the average noise properties within region D,and the shape of the restoring beam.
Telescope Survey Date Frequency Bandwidth Noise Beam shapeobserved [MHz] [MHz] [mJy beam − ] Bmaj[ ], Bmin[ ], PA[ ◦ ] ( MWA Phase I GLEAM SGP 2013–2015 119 30.72 16.2 241, 202, -57154 9.3 199, 157, -51186 6.7 162, 129, -47MWA Phase II MIDAS ES 2018–2020 216 30.72 0.9 54, 43, 157uGMRT – 2016 399 200 0.1 15.8, 6.7, 9.5ASKAP EMU-ES 2019 887.5 288 0.035 10.5, 7.8, 87VLA NVSS 1993–1996 1400 50 0.45 45, 45, 0ATCA GLASS 2016–2020 5500, 9500 2000, 2000 0.024, 0.04 (4, 2, 0), (3.4, 1.7, 0) catalogue, developed by the method used for GLEAM.Franzen et al. (in prep) quote an 8% absolute uncer-tainty in the GLEAM SGP flux density scale, consistentwith the value reported for GLEAM by Hurley-Walkeret al. (2017).
In 2018 the MWA was reconfigured into an extendedbaseline configuration (MWA Phase II; Wayth et al.,2018). The MWA Phase II provides a maximum base-line of 5.3 km which, at 216 MHz, improves the angularresolution to 54 × . This drives down the classicaland sidelobe confusion limits, allowing for deeper imag-ing to be made with an RMS of σ (cid:28) − .MWA Phase II science is presented in Beardsley et al.(2020). The MWA Interestingly Deep Astrophysical Sur-vey (MIDAS, Seymour et al. in prep) will provide deepobservations of six extra-galactic survey fields, includingGAMA 23, in the MWA Phase II configuration. Here,we use an early science (MIDAS ES herein) image ofthe highest frequency band centered at 216 MHz. Datareduction followed the method outlined by Franzen etal. (in prep) where each snapshot was calibrated usinga model derived from GLEAM. Imaging was performedusing the W S C l e a n imaging software (Offringa et al.,2014) using a robustness of 0 (Briggs, 1995). Altogether,54 two-minute snapshot images, each achieving an av-erage RMS of ∼ − without self calibration,were mosaiced to produce a deep image. Stacking of indi-vidual snapshots reproduced the expected √ t increase insensitivity, where t is the total integration time, and im-plied the classical and sidelobe confusion limits were notreached. After a round of self calibration the final deepimage, at zenith, achieved an RMS of ∼ − with a (cid:46) restoring beam. Radio sources were cata-logued using the Background And Noise Estimation tool b a n e , to measure the direction-dependent noise acrossthe image, and a e g e a n to perform source finding and characterisation (Hancock et al., 2012, 2018) . Using theGLEAM catalogue, sources above 10 σ in MIDAS ES andGLEAM were used to correct the MIDAS ES flux-scale.For 887 such sources in GAMA 23, correction factorswere derived based on the integrated flux density ratiobetween MIDAS ES and GLEAM. The correction fac-tors followed a Gaussian distribution, and indicated anagreement with GLEAM within 5%. Given the 8% un-certainty in the GLEAM flux density scale (Sect. 2.1.1),we prescribe an 8% uncertainty in the MIDAS ES fluxdensity scale. In 2019 the ASKAP delivered the first observationsconducted with the full 36-antenna array configuration.The design of ASKAP opens up a unique parameterspace for studying the extra-galactic radio source pop-ulation. With a maximum baseline of 6 km, ASKAPproduces images with ∼ resolution at 887 MHz. Theshortest baseline is 22 m, recovering maximum angularscales of ∼ . ◦ . As part of the Evolutionary Map ofthe Universe (EMU; Norris, 2011) Early Science, theGAMA 23 field was observed at 887 MHz and madepublicly available on the CSIRO ASKAP Science DataArchive (CASDA ). Details of the reduction are brieflysummarized here. This data has been reduced by theASKAP collaboration using the A S K A P s o f t data-reduction pipeline (Whiting et al. in prep) on the Galaxysupercomputer hosted by the Pawsey SupercomputingCentre. PKS B1934-638 was used to perform bandpassand flux calibration for each of the unique 36 beams.Bandpass solutions were applied to the target fields.The final images are restored with a 10.55 × . aegean and bane publicly available on GitHub: https://github.com/PaulHancock/Aegean . This work made use of 2.0.2version of aegean https://data.csiro.au/collections/domain/casdaObservation/search/ emnant Radio Galaxies Discovered in a Multi-frequency Survey ◦ ) elliptical beam, and achieves an RMS of σ ≈ − µ Jy beam − . Henceforth we refer to theseobservations as EMU-ES. We use observations conducted by the National Ra-dio Astronomy Observatory (NRAO) using the VeryLarge Array (VLA; Thompson et al., 1980) to sam-ple an intermediate frequency range. The NRAO VLASky Survey (NVSS; Condon et al., 1998) surveysthe entire sky down to Dec= − ◦ at 1400 MHz. Ob-servations for NVSS were collected predominately inthe D configuration, however the DnC configurationwas used for Southern declinations. Final image prod-ucts were restored using a circular synthesized beamof 45 × . At 1400 MHz, NVSS achieves an RMS of σ = 0 .
45 mJy beam − . The GAMA Legacy ATCA Sky Survey(GLASS; Huynh et al. in prep) offers simultane-ous, high-frequency (5.5, 9.5 GHz) observations ofGAMA 23 observed by the ATCA. Observations forGLASS were conducted over seven semesters between2016 – 2020 (PI: M. Huynh, project code: C3132).Data at each frequency were acquired with a 2 GHzbandwidth, made possible by the Compact ArrayBroadband Backend (Wilson et al., 2011), with thecorrelator set to a 1 MHz spectral resolution. Observa-tions for GLASS were conducted in two separate ATCAarray configurations, the 6A and 1.5C configurations,contributing 69% and 31% towards the total awardedtime, respectively. The shortest interferometer spacingis 77 m in the 1.5C configuration, providing a largestrecoverable angular scale of 146 at 5.5 GHz. As partof the observing strategy, GLASS was divided into six8.31 deg regions (regions A − F), for which region D(RA=345 ◦ , Dec=-33.75 ◦ ) was observed, reduced andimaged first. For this reason, this paper focuses only onregion D.Processing and data reduction were conducted usingthe Multichannel Image Reconstruction, Image Analysisand Display ( M I R I A D ) software package (Sault et al.,1995), similar to the method outlined by Huynh et al.(2015, 2020). The 1435 region D pointings are restoredwith the Gaussian fit of the dirty beam, and convolvedto a common beam of 4 × (BPA=0 ◦ ) at 5.5 GHz, andachieves an RMS of ∼ µ Jy beam − . A similar processat 9.5 GHz results in a 3 . × . (BPA=0 ◦ ) synthe-sized beam and achieves an RMS of ∼ µ Jy beam − .Although the same theoretical sensitivity is expectedat 5.5 GHz and 9.5 GHz, the sparse overlap in adjacentpointings, larger phase calibration errors, and increased NVSS catalogue publicly available on VizieR: https://vizier.u-strasbg.fr/viz-bin/VizieR?-source=%20NVSS radio frequency interference (RFI) all result in a drop insensitivity. Henceforth, we refer to GLASS observationsconducted at 5.5 GHz and 9.5 GHz as GLASS 5.5 andGLASS 9.5 respectively.
As part of the GLASS legacy survey (Sect. 2.1.5),uGMRT has observed GAMA 23 in band-3 (250–500 MHz) centered at 399 MHz. The project 32_060(PI: Ishwara Chandra) was awarded 33 hours to cover 50contiguous pointings spanning a 50 square-degree region.Observations were conducted in a semi-snapshot mode,with ≈
30 minutes per pointing distributed through three10 minute scans. In band-3, the wide-band correlator col-lects a bandwidth of 200 MHz divided into 4,000 finechannels. Data reduction was conducted using a Com-mon Astronomical Software Application (CASA; Mc-Mullin et al., 2007) pipeline . Data reduction followedthe standard data reduction practices such as data flag-ging, bandpass and gain calibration, application of solu-tions to target scans, imaging and self-calibration (seeIshwara-Chandra et al. 2020 for details). The image isrestored by a 15.8 × . (BPA= 9.5 ◦ ) beam, andachieves a best RMS of ∼ µ Jy beam − . Note thatseveral bright sources throughout the field adverselyimpact the data reduction, resulting in large spatialvariations in the RMS. Due to their power-law spectral energy distributions, theintegrated luminosities of radio galaxy lobes decreasewith increasing frequency (Sect. 1). Given that rem-nants can display ultra-steep radio spectra, this presentsa sensitivity challenge associated with their detection.For a survey such as GLASS which has high θ ∼ resolution (Sect. 2.1.5), resolution bias further exacer-bates this problem: diffuse low-surface-brightness regionsescape detection with greater ease, resulting in an under-estimation of the integrated flux density. To combatthe resolution bias suffered by GLASS, we carried outlow-resolution observations with ATCA.The project C3335 (PI: B. Quici) was awarded 14hours to conduct low-resolution observations of eachremnant identified for this work, at 2.1, 5.5 and 9 GHz.Observations at 2.1 GHz ( LS -band) were conducted inthe H168 configuration; observations at 5.5 and 9 GHz( CX -band) were conducted in the H75 configuration. Theminimum and maximum baselines achieved within eachconfiguration are 61 m and 192 m for H168, 31 m and89 m for H75 respectively (excluding baselines formedwith antenna CA06). Bandpass, gain and flux calibra-tion was performed using PKS B1934-638. Due to the Pipeline can be found at , and makes use of CASA version 5.1.2-4
Quici et al.
Table 2
Details of the additional ATCA data collected here under project code C3335, PI: B. Quici. Each column, inascending order, details the configuration used to conduct observations, the date observations were conducted, the centralfrequency of the receiver band, the bandwidth available within each band, the approximate time spent per-source in eachconfiguration, the secondary calibrator observed, the average noise per pointing, and the shape of the restoring beam. Note,primary calibrator PKS B1934-638 was observed for all observations.
Config. Date Frequency Bandwidth Duration Secondary Noise Beam shapeobserved [GHz] [GHz] [min] calibrator [mJy beam − ] Bmaj[ ], Bmin[ ], PA[ ◦ ] H168 18/10/19 2.1 2 22
PKS B2259-375
PKS B2254-367 uv coverage, each targetwas observed on a rotating block of length ∼
30 minutes,with approximately C A S A soft-ware package , and followed the standard data reductionpractices. As part of preliminary flagging, the data are‘clipped’, ‘shadowed’ and ‘quacked’ using the flagdata task, in order to flag for zero values, shadowed antennasand the first five seconds of the scans, respectively. Fortyedge channels of the original 2049 are also flagged dueto bandpass roll-off (Wilson et al., 2011). Again usingthe flagdata task, the uncalibrated data are then auto-matically flagged for RFI using mode=‘tfcrop’ , whichcalculates flags based on a time and frequency window.The data are manually inspected in an amplitude ver-sus channel space, to ensure RFI is adequately flagged.Observations conducted in the LS -band and CX -bandwere split into four and eight sub-bands, respectively.Calibration was performed per sub-band per pointing.To perform calibration, the complex gains and band-pass are solved for first using the primary calibrator.Complex gains and leakages were solved for next us-ing the secondary calibrator. After applying a flux-scalecorrection based on PKS B1934-638, the calibration so-lutions were copied individually to each target scan. Asecondary round of automatic RFI flagging is performedwith flagdata where mode=‘rflag’ , which is used forcalibrated data. The primary and secondary calibrator,as well as each target scan are flagged in this way. In total,approximately ≈
52% and ≈
15% of the available band-width was flagged due to RFI present in the LS and CX bands, respectively. Within the hybrid configurationsthe first five antennas, CA01–CA05, provide the densepacking of short (10–100 m) spacings. Antenna CA06is fixed and provides much larger spacings of ∼ uv coverage, Time per source was varied slightly to accommodate for thefainter/brighter sources. This work makes use of CASA version 5.1.2-4 all baselines formed with antenna CA06 are excludedto achieve a well-behaved point-spread-function. Theaverage noise properties and restoring beams at 2.1, 5.5and 9 GHz are respectively: σ ∼ . − and θ = 104 × , σ ∼ .
22 mJy beam − and 88 × ,and σ ∼ .
14 mJy beam − and θ = 54 × . We useany unresolved GLASS sources present within the targetscans to evaluate a calibration uncertainty. At 5.5 GHz,the ratio of the integrated flux density between thelow-resolution ATCA observations and GLASS was con-sistently within 3%. We use this value as the absoluteflux-scale uncertainty. Details of these observations aresummarized in Table 2. A comparison of these observa-tions and GLASS, at 5.5 GHz, is presented in Figure 2.The reader should note, due to persistent RFI in the LS -band, we were unable to make a detection of mostof our targets at this frequency. Observed with the Visible and Infrared Survey Tele-scope for Astronomy (VISTA), the VISTA Kilo-degreeINfrared Galaxy survey (VIKING) provides medium-deep observations over ∼ across the Z , Y , J , H & K s bands, each achieving a 5 σ AB magnitude limitof 23.1, 22.3, 22.1, 21.5 and 21.2 respectively. VIKINGimaging is used to perform both an automated andmanual host galaxy identification (see Sect. 3).
The GAMA 23 photometry catalogue (Bellstedt et al.,2020) contains measured properties and classificationsof each object catalogued by the
P ro Fo u n d (Robotham et al., 2018) source finding routine, whichuses a stacked r + Z image to perform initial source find-ing. Approximately 48,000 objects have spectroscopicredshifts provided by the Anglo-Australian Telescope(AAT), and the survey is ∼
95% spectroscopically com-plete up to an i band magnitude of 19.5 (Liske et al.,2015). For objects without spectroscopic redshifts, weuse the near-UV to far-IR photometry (available in the Publicly available on GitHub: https://github.com/asgr/ProFound emnant Radio Galaxies Discovered in a Multi-frequency Survey h m s s s s -34°46'47'48'49'50' RA (J2000) D e c ( J ) Figure 2.
A 5.5 GHz view of the radio source MI-DAS J225337 − × σ , where σ = 210 µ Jy beam − and is the local RMS.photometry catalogue) to obtain photometric redshiftsusing a public photometric redshift code (EAZY; Bram-mer et al., 2008) . Our methodology for compiling a sample of radio galax-ies, and classifying their activity status, is presentedbelow. We begin our selection at low frequencies wheresteep-spectrum radio lobes are naturally brighter. Notonly does MIDAS ES provide access to such low frequen-cies, but also its relatively low spatial resolution resultsin a sensitivity to low-surface-brightness emission thatis required to recover emission from diffuse, extendedradio sources.Genuine remnant radio galaxies will not display ra-dio emission from the core associated with the AGN jetactivity at the centre of the host galaxy. As such, we clas-sify any radio galaxy as ‘active’ if positive radio emissionis observed from the radio core. Similarly, we prescribea ‘candidate remnant’ status to any radio galaxy thatdemonstrates an absence of radio emission from the core.True emission from the radio core will be unresolved evenon parsec scales, meaning observations with high spatial EAZY publicly available on GitHub: https://github.com/gbrammer/eazy-photoz resolution are ideal for their detection. This also ensuresthat the emission from the radio core will not experienceblending by the backflow of the radio lobes. As demon-strated by Mahatma et al. (2018), sensitive observationsare equally important to enable the detection of radiocores. GLASS addresses both of these considerations byproviding sensitive ( ∼ µ Jy beam − ), high resolution(4 × ) radio observations at 5.5 GHz. Our sample isconstructed by following the steps outlined below (seeTable 3), and is presented in full as a supplementaryelectronic table (see Table 9 for column descriptions).1. Limit the search footprint within GLASS region D
The footprint, within which radio galaxies are se-lected, is constrained to 343 ◦ ≤ RA ≤ ◦ , − ◦ ≤ Dec ≤ − . ◦ . This excludes the outerregions of higher noise from the GLASS mosaicthus ensuring the GLASS noise levels range be-tween 25 − µ Jy beam − at 5.5 GHz. By maintain-ing almost uniform noise levels across the field, thisreduces the bias of selecting brighter radio coresin higher-noise regions. By applying this footprintto the MIDAS ES component catalogue, 676 com-ponents are selected. The resulting sky coveragewithin this footprint is 8.31 deg .2. Flux-density cut
A 10 mJy flux density cut at 216 MHz is imposedon the sample. Given the low angular resolution,this ensures the MIDAS ES detections are robust,e.g. greater than 10 σ for an unresolved source. Weidentify 446 radio sources brighter than 10 mJy at216 MHz.3. Angular size cut
Our decision to impose a minimum size constraintwas motivated by two factors: firstly to minimiseblending of the radio core with the radio lobes, andsecondly to allow for an interpretation of the radiosource morphology. By imposing a minimum 25 angular size constraint, we ensured a minimum ofsix GLASS 5.5 synthesized beams spread out acrossthe source. For each of the 446 radio sources, weproduce 2 × cutouts centered at the cataloguedMIDAS ES source position. EMU-ES, GLASS 5.5,GLASS 9.5 and VIKING K s − band cutouts aregenerated, and contours of the radio emission areoverlaid onto the VIKING K s − band image. Hence-forth, we refer to these as image overlays. WhileGLASS 5.5 has the advantage in spatial resolu-tion, EMU-ES has a significantly better brightness-temperature sensitivity. This consideration is impor-tant since faint radio lobes, while seen in EMU-ES,may become undetected by GLASS 5.5.As such, a first pass is conducted by identifyingany radio source with an angular size greater than Image cutouts are generated using
Astropy module
Cutout2D
Quici et al.
Table 3
Summary of each sample criteria discussed in Sect. 3. Steps 1–4 describe the radio galaxy sample selection. Step 5describes the active/remnant classification. Step 6 describes host galaxy association. † Step 3 is broken into two parts, denoted by steps 3.1 and 3.2, for which the resulting sample size is the sum total.
No. Sample step Criteria Sample size1 Limit footprint within GLASS region D 343 ◦ ≤ RA ≤ ◦ ◦ ≤ Dec ≤ -35 ◦ S >
10 mJy 4463 Angular size cut † θ ≥ θ GLASS ≥ (82)3.2 θ GLASS < & θ EMU − ES ≥ (27) ( z s ) ( z p ) ( no z ) Host identification (Remnant) See Sect. 4 3 ( z s ) ( z p ) in GLASS 5.5 (e.g. θ GLASS 5 . ≥ ). Due tothe manageable size of the sample, we do this stepmanually by visually matching the correct compo-nents of each radio source, and measure the linearangular extent across each radio source. We iden-tify 82 radio sources this way. For radio sourceswith θ GLASS 5 . < θ EMU − ES ≥ (e.g. see Figure 3). An additional27 radio sources are identified this way, giving atotal of 109 radio sources greater than 25 . Forconsistency, the angular size of each radio sourceis re-measured using EMU-ES, by considering thelargest angular size subtended within the footprintof radio emission above 5 σ .4. AGN dominated
As a result of their sensitivity to low-surface-brightness emission, both MIDAS ES and EMU-ESare able to detect radio emission from a typical face-on spiral galaxy. Radio emission from these objectsis not driven by a radio-loud AGN, and thereforethese sources need to be removed from the sam-ple. While the radio emission of virtually all radiogalaxies extends well beyond the host galaxy, radioemission from spiral galaxies is associated only withthe optical component of the galaxy. Thus usingthe aforementioned image overlays, we remove threeradio sources that trace the optical/near-infraredcomponent of the host galaxy, as revealed withVIKING K s − band imaging.We provide an examplein Figure 4. The remaining sample contains 106 extended radio galaxies, forming the parent samplefor this analysis.5. Activity status
To constrain the nuclear activity associated withan AGN, we use GLASS 5.5 imaging to searchfor evidence of a radio core. For a successful ra-dio core detection, we require a compact objectwith a peak flux density greater than 3 σ . We use b a n e to produce an RMS image associated witheach GLASS 5.5 image cutout. Only pixel valuesabove 3 σ are considered. We use the orientationand morphology of the radio lobes as a rough con-straint on the potential position of the radio core.Following this method, we classify 94 radio galaxiesas active, and a further 11 as candidate remnantradio galaxies. We emphasise that this method onlyselects candidate remnant radio galaxies, since theexistence of a faint, undetected radio core is stillpossible. For each remnant candidate we place a3 σ upper limit on the peak flux density of the core.Here, σ is measured by drawing a circle equivalentto four GLASS synthesized beams at the position ofthe presumed host galaxy and measuring the RMSwithin this region.6. Host identification
For radio sources with a core, we use a 1 searchradius to cross match the position of the radio corewith the GAMA 23 photometry catalog. Out of94 such sources, the hosts of 80 radio galaxies areidentified this way, of which 26 and 54 have spec-troscopic and photometric redshifts, respectively.The hosts of the remaining 14 sources are eitherextremely faint in K s -band, or remain completelyundetected in VIKING, potentially due to lying at emnant Radio Galaxies Discovered in a Multi-frequency Survey h m s s s s -32°31'40"32'00"20"40"33'00" Right Ascension D e c li n a t i o n Figure 3.
Example of the radio source MIDAS J230304-323228satisfying the criterion: θ GLASS < & θ EMU − ES ≥ . Thelow-surface-brightness lobes are escaping detection in GLASS,resulting in an incomplete morphology. The contours representEMU-ES (navy blue), GLASS 5.5 (cyan) and GLASS 9.5 (ma-genta), with levels set at [3,4,5,7,10,15,25,100] × σ , where σ is thelocal RMS of 43, 26 and 40 µ Jy beam − respectively. Contoursare overlaid on a linear stretch VIKING K s -band image. Theseemingly absent hotspots would imply these are remnant lobes,however the presence of a radio core means this source is classifiedas ‘active’. The true nature of this source may be a restarted radiogalaxy, however the lack of any resolved structure around the coreis puzzling. higher redshift. Host identification for remnant can-didates is discussed on a per-source basis in Sect. 4,as this is a complicated and often ambiguous pro-cedure. For each radio source we also use WISE(Wright et al., 2010) 3.4 µ m and 4.6 µ m images todetermine if any potential hosts were not present inthe VIKING imaging, however this did not revealany new candidates. For each of the 104 radio galaxies, integrated fluxdensities are compiled from the data described inSect. 2.1. To compile the integrated flux densities at119, 154 and 186 MHz (GLEAM SGP), 216 MHz (MI-DAS ES), and 1400 MHz (NVSS), we use their appro-priate source catalogues described in their relevant datasections. As a result of the high spatial resolution at399 MHz (uGMRT observations), 887 MHz (EMU-ES)and 5.5 GHz (GLASS 5.5), sources are often decomposedinto multiple components. To ensure the integrated fluxdensities are measured consistently across these surveys, h m s s s m s -33°43'30"44'00"30"45'00"30" Right Ascension D e c li n a t i o n Figure 4.
Example of a non AGN-dominated radio source, MI-DAS J225802-334432, excluded from the sample. Analysis of theradio morphology shows that the radio emission traces the opticalcomponent of the host galaxy. The contours represent EMU-ES(navy blue), GLASS 5.5 (cyan) and GLASS 9.5 (magenta), withlevels set at [3,4,5,7,10,15,25,100] × σ , where σ is the local RMS of45, 28 and 41 µ Jy beam − respectively. Contours are overlaid ona linear stretch VIKING K s -band image. The radio emission ishosted by IC 5271 (ESO 406-G34). we convolve their image cutouts of each source with a54 circular resolution (e.g. the major axis of the MI-DAS synthesized beam). Integrated flux densities arethen extracted using a e g e a n by fitting a Gaussianto radio emission. For each remnant candidate observedwith ATCA at low resolution at 2.1, 5.5 and 9 GHz, weuse a e g e a n to measure their integrated flux densities.The 5.5 GHz integrated flux density reported for eachremnant candidate is exclusively taken from these lowresolution ATCA observations, not GLASS. Finally, foreach integrated flux density measurement, uncertaintiesare calculated as the quadrature sum of the measurementuncertainty and the absolute flux-scale uncertainty. To better understand their energetics, we model the in-tegrated radio spectrum of each remnant candidate. Weuse a standard power-law model of the radio continuumspectrum (Eqn 1), where the spectral index, α , and theflux normalization S are constrained by the fitting, and ν is the frequency at which S is evaluated: S ν = S ( ν/ν ) α (1)Given we can expect to see evidence of a curvaturein their spectra, especially over such a large frequency0 Quici et al. range, we also fit a generic curved power-law model(Eqn 2). Here, q offers a parameterization of the curva-ture in the spectrum, where q < q typically ranges within − . ≤ q ≤
0. Al-though q is not physically motivated, Duffy & Blundell(2012) show that it can be related to physical quantitiesof the plasma lobes such as the energy and magneticfield strength. S ν = S ν α e q (ln ν ) (2)Fitting of each model is performed in Python using the c u rv e _ f i t module; fitted models are presented inFigures 5a − − BIC which suggests a preference towards the second modelfor ∆BIC > < < | ∆BIC| <
2, whereas amodel is strongly preferred if | ∆BIC| >
6. In Table 4, wecalculate ∆BIC = BIC power − law − BIC curved − power − law . We present and discuss each of the 11 candidate rem-nant radio galaxies below. Seven are found to displayhotspots in GLASS. Image overlays and the radio contin-uum spectrum are presented in Figure 5. General radioproperties are presented in Table 4.
Figure 5a shows extremely relaxedlobes and an amorphous radio morphology. No com-pact structures that would indicate hotspots are ob-served. The average 154 MHz surface brightness is ∼
32 mJy arcmin − , satisfying the low SB criterion(SB <
50 mJy arcmin − ) employed by Brienza et al.(2017). The diffuse radio emission is undetected by theuGMRT observations, NVSS, GLASS, as well as the2.1 GHz and 9 GHz ATCA follow-up observations. Un-surprisingly, we find that the source spectrum appearsultra-steep at low frequencies, and demonstrates a curva-ture ( q = − .
11) across the observed range of frequencies.The radio properties point towards an aged remnant.
Host galaxy.
Identification of the host galaxy is ratherchallenging here as the amorphous radio morphologyprovides little constraints on the host position. No clearhost galaxy is seen along the centre of the radio emission,however this can easily be explained if the radio lobeshave drifted. We approximate the central position of theradio emission by taking the centre of an ellipse drawn to best describe the radio source. G1 ( z p = 0 . from the radio center, corresponding to a 61 kpcoffset. G2 ( z p = 0 . from the radiocenter, corresponding to a 80 kpc offset. G3 ( z p = 0 . from the radio center, corresponding to a102 kpc offset. Without any additional information, wetake G1 as the likely host galaxy. We note that G4 ( z p =0 .
41) shows compact radio emission at 887 MHz, howeverit is unclear whether this is related to the extendedstructure. We include the radio spectrum arising from G4in Figure 5a, and note that it contributes approximately5% to the total radio flux density at 887 MHz. If G4 isunrelated, its radio spectrum should be subtracted fromthe integrated spectrum of MIDAS J225522-341807.
Figure 5b shows a pair of relaxedradio lobes, with a diffuse bridge of emission connect-ing each lobe along the jet axis. The 154 MHz aver-age surface brightness is calculated as ∼ − , satisfying the low surface brightness criterion.The edge brightened regions likely represent the ex-panded hotspots of the previously active jet, similarto what is observed in B2 0924+30 (Shulevski et al.,2017). The source is undetected by the uGMRT observa-tions, GLASS, as well as the ATCA follow-up at 2.1 GHzand 9 GHz. Curvature is evident in the spectrum, whichbecomes ultra-steep above 1.4 GHz. Host galaxy.
Along the projected centre of the jet axis,a collection of three potential host galaxies exist withina ∼ aperture. The redshift of each host, G1 ( z s =0 . z p = 0 . z s = 0 . Figure 5c shows two relaxed, low sur-face brightness lobes that are asymmetrical in shape.The flattened ‘pancake’-like morphology of the North-ern lobe can be explained by the buoyant rising of thelobes (Churazov et al., 2001). The surface brightness ofeach lobes is approximately 43 mJy arcmin − , satisfyingthe low SB criterion employed by Brienza et al. (2017).The source is undetected by the follow-up 2.1 GHz and9 GHz ATCA observations. The spectrum seems con-sistent with a single power law ( α = − . Host galaxy.
G1 ( z p = 0 .
57) lies 4 . away from theradio center, corresponding to a 32 kpc offset. G2 ( z p =0 . away from the radio center, correspond-ing to a 61 kpc offset. This assumes that the lobes areequidistant from the host, which is not always the case. emnant Radio Galaxies Discovered in a Multi-frequency Survey Table 4
Summarized radio properties of the selected remnant candidates. S gives the 216 MHz integrated flux density.LAS gives the largest angular size measured from EMU-ES. S core gives the 5.5 GHz upper limit placed on the radio core peakflux density using GLASS. α fit denotes the spectral index fitted by each model. The curvature term modelled by the curvedpower-law model is represented by q . As per Sect. 3.3, the ∆BIC is calculated between each model and presented in the finalcolumn. A reduced chi-squared ( χ ) is also evaluated for each model. MIDAS Name Fig. S
LAS S core
Power-law Curved power-law ∆BIC (mJy) ( ) ( µ Jy beam − ) α fit χ α fit q χ J225522 −
5a 24.7 ± < -1.40 ± ± ± −
5b 18.3 ± < -1.10 ± ± ± −
5c 14 . ± . < -0.86 ± ± ± −
5d 170 . ± . < -0.92 ± ± ± −
5e 192 . ± . < -0.87 ± ± ± −
5f 36 . ± . < -0.86 ± ± ± −
5g 113 . ± . < -0.73 ± ± ± −
5h 55 . ± . < -0.72 ± ± ± −
5i 153 . ± . < -0.86 ± ± ± −
5j 198 . ± . < -1.00 ± ± ± However, we retain G1 as the likely host galaxy.
Figure 5d shows a typical low-resolution FR-II radio galaxy as evidenced by the edge-brightened morphology. The average 154 MHz surfacebrightness is 160 mJy arcmin − . The source is firmlydetected by the ATCA follow-up at all frequencies, re-vealing that the spectrum is highly curved ( q = − . ν (cid:38) . ± .
17, suggestingthe lobes are remnant. The properties of the spectrumstrongly suggest a lack of energy supply to the lobes,however, GLASS 5.5 reveals compact emitting regionsat the edges of each lobe that may suggest recent energyinjection. We divert a detailed analysis of this source toSect. 5.3.
Host galaxy.
The radio lobes are unambiguously asso-ciated with galaxy G1 ( z s = 0 . Figure 5e demonstrates an elongated,‘pencil-thin’, radio galaxy with an edge-brightened FR-II morphology. GLASS detects only the brightest andmost compact emitting regions, and misses the lowersurface brightness emission seen at 887 MHz. The radiosource is detected in all but the 2.1 GHz ATCA follow-upobservations. The radio spectrum is well modelled bya power-law ( α = − . Host galaxy.
G1 ( z p = 1 . z p = 1 . z p = 1 . Figure 5f demonstrates a pair of lobeswith compact emitting regions seen by GLASS. Theradio spectrum is well approximated as a power-law( α = − .
86) with no evidence of a spectral curvatureover the observed range of frequencies.
Host galaxy.
G1 ( z p = 0 . Figure 5g shows a peculiar radio mor-phology; while the western lobe shows bright emittingregions in GLASS, the counter lobe is completely dif-fuse and does not show a hotspot. It is unclear what iscausing this.
Host galaxy.
The galaxies G1 ( z p = 0 .
32) and G2 ( z s =0 . Figure 5h shows a typical FR-II radiogalaxy, as implied by the edge-brightened morphology.The source exhibits clear hotspots in each lobe, as seenby GLASS 5.5 and GLASS 9.5. The source is detected inall but the 2.1 GHz ATCA follow-up. Modelling the radiocontinuum spectrum gives a spectral index of approxi-2
Quici et al. mately α = − .
7, revealing no significant energy losses.The ∆BIC offers tentative evidence for some spectralcurvature, however this may just be a result of a poorlyconstrained spectrum at low ( ≤
215 MHz) frequencies.
Host galaxy.
The radio source is unambiguously as-sociated with G1 ( z s = 0 . Figure 5i shows two distinct radio lobeswith compact emitting regions observed at 5.5 GHz and9.5 GHz. The spectral index is approximately α = − . q ∼ − .
02) and does not nec-essarily require an absence of energy injection.
Hostgalaxy.
Three likely host galaxies are identified by theiralignment along the projected jet axis; G1 ( z p = 1 . z p = 0 . z p = 0 . Figure 5j . Both lobes are detectedin GLASS 5.5, however, only the southern lobe showsemission in GLASS 9.5. It is unclear what is causingthis; both components are resolved by GLASS and showsimilar morphologies, so it is unlikely they are unrelated.The radio spectrum is well approximated by a power-law ( α = − Host galaxy.
G1 ( z p = 0 .
9) is a favourable host galaxycandidate, due to its small angular separation from theprojected centre between the lobes. G2 ( z p = 0 . To understand the limitations imposed by our selectioncriteria, we investigate the core prominence distributionacross our sample. We define the core prominence(CP) as the ratio of core to total flux density, i.e.CP = S core /S total . The total integrated flux densityis measured at 216 MHz. We take the GLASS 5.5measurement of the radio core flux density, and re-scaleto 216 MHz assuming the spectrum of the radio core is aflat spectrum ( α = 0) (e.g; Hardcastle & Looney, 2008).For the selected remnant candidates, we present upperlimits on their CP by using the 3 σ upper limits on theircore peak flux density. Our results are presented inFigure 6. Sources with radio cores show a wide distributionin their CP, varying within the range 10 − – 10 − .The median CP of the sample is ∼ × − , almosttwo orders of magnitude larger than the median CPreported by Mullin et al. (2008) for the 3CRR sample(e.g. ∼ × − ). This can be expected given themost powerful radio galaxies are preferentially selectedby 3CRR. Instead, comparing our CP range to theLOFAR-selected sample compiled by Mahatma et al.(2018) – e.g. see their Figure 4 – we find the ranges areconsistent. Although the reader should note Mahatmaet al. (2018) compute their CP at 150 MHz, meaning a∆ ν = 66 MHz frequency shift should be accounted for ifa direct comparison is made.As discussed in Sect. 1, the ‘absent radio core’criterion only selects remnant candidates. Genuineremnant radio galaxies will not display a radio core,meaning their CP should approach null. In fact, the CPin such sources should be lower than any active radiosource in which a radio core is present. This impliesthat a clean separation should exist between active andremnant radio galaxies, however, this is not what wesee in Figure 6. Instead, we find that the CP upperlimits imposed on the remnant candidates overlap withcore-detected radio galaxies. This comes as a result ofour sample criteria. Given the GLASS detection limit( ∼ µ Jy beam − ), only remnant candidates brighterthan ∼
500 mJy will show CP upper limits belowwhat is observed for core-detected radio galaxies (e.g.log(CP) (cid:46) − . (cid:46) − ) CPcriterion is used. This observation echoes the resultsof (Brienza et al., 2017), who show that none of theirultra-steep spectrum remnants are selected by low CP(e.g. < . emnant Radio Galaxies Discovered in a Multi-frequency Survey (a) MIDAS J225522 − . EMU-ES contour levels: [3,4,5,7,10] × σ . GLASS 5.5 contour levels: [3,4,5] × σ . GLASS 9.5 contours are notpresented due to an absence of radio emission above 3 σ . Compact component at RA=22 h m s , Dec= -34 ◦ is unrelated. The radiospectrum of the compact radio component, G4, is demonstrated by the blue markers. Radio emission from G4 is undetected by GLASS 5.5, wethus present a 3 σ upper limit. h m s s s s -34°17'30"18'00"30"19'00" Right Ascension D e c li n a t i o n G1G2G3G4 F l u x D e n s i t y [ m J y ] = 1.4 ± 0.09 q = 0.11 ± 0.08 (b) MIDAS J225607 − . EMU-ES contour levels: [3,4,5,7,10,12,15,20] × σ . GLASS 5.5 contour levels: [3,4,5] × σ . GLASS 9.5 contoursare not presented due to an absence of radio emission above 3 σ . Compact component at RA=22 h m s , Dec= -34 ◦ is unrelated. h m s s s s s -34°31'30"32'00"30"33'00" Right Ascension D e c li n a t i o n G1G3G2 F l u x D e n s i t y [ m J y ] = 1.1 ± 0.04 q = 0.073 ± 0.009 Figure 5. Left:
Plotted are the remnant candidates presented in Sect. 4. Background image is a VIKING K s band cutout set on alinear stretch. Three sets of contours are overlaid, representing the radio emission as seen by EMU-ES (black) , GLASS 5.5 (orange) andGLASS 9.5 (blue) . Red markers are overlaid on the positions of potential host galaxies. Right:
The radio continuum spectrum between119 MHz and 9 GHz. The integrated flux densities at 5.5 GHz come from the low-resolution ATCA observations (Sect. 2.1.7) not thelower resolution GLASS images. A simple power-law (Eqn 1) and curved power-law (Eqn 2) model are fit to the spectrum, indicated bythe purple and blue models, respectively. Quici et al. (c) MIDAS J225608 − . EMU-ES contour levels: [3,4,5,7,10] × σ . GLASS 5.5 contour levels: [3,4,5] × σ . GLASS 9.5 contoursare not presented due to an absence of radio emission above 3 σ . h m s s s s -34°18'20"40"19'00"20"40" Right Ascension D e c li n a t i o n G1G2G3 F l u x D e n s i t y [ m J y ] = 0.86 ± 0.05 q = 0.04 ± 0.07 (d) MIDAS J225337 − . EMU-ES contour levels: [4,5,10,30,50,70] × σ , GLASS 5.5 contour levels: [3,4,5,6] × σ . GLASS 9.5contours are not presented due to an absence of radio emission above 3 σ . h m s s s s -34°47'00"30"48'00"30" Right Ascension D e c li n a t i o n G1 F l u x D e n s i t y [ m J y ] = 0.92 ± 0.05 q = 0.12 ± 0.01 (e) MIDAS J225543 − . EMU-ES contour levels: [3,4,5,7,15,30,100] × σ , GLASS 5.5 contour levels: [3,5,10,20] × σ .GLASS 9.5 contour levels: [3,5,10,20] × σ h m s s s s s -34°40'00"30"41'00"30" Right Ascension D e c li n a t i o n G1G2G3
Frequency [GHz] F l u x D e n s i t y [ m J y ] = 0.87 ± 0.01 q = 0.0033 ± 0.01 Figure 5. – continued. emnant Radio Galaxies Discovered in a Multi-frequency Survey (f) MIDAS J225919-331159 . EMU-ES contour levels: [5,10,20,40,60] × σ , GLASS 5.5 contour levels: [3,4,5,6,10,20] × σ .GLASS 9.5 contour levels: [3,4,5,6] × σ h m s s s s -33°11'20"40"12'00"20"40" Right Ascension D e c li n a t i o n G1 F l u x D e n s i t y [ m J y ] = 0.86 ± 0.02 q = 0.035 ± 0.01 (g) MIDAS J230054 − . EMU-ES contour levels: [3,4,5,7,15,30,100,300] × σ , GLASS 5.5 contour levels: [3,5,10,20,30] × σ .GLASS 9.5 contour levels: [3,5,10,20] × σ h m s m s s s -34°00'30"01'00"30"02'00" Right Ascension D e c li n a t i o n G2G1G3 F l u x D e n s i t y [ m J y ] = 0.73 ± 0.03 q = 0.027 ± 0.02 (h) MIDAS J230104-334939 . EMU-ES contour levels: [5,8,15,35,50] × σ , GLASS 5.5 contour levels: [3,5,7,9,11] × σ . GLASS 9.5contour levels: [3,4,5,6] × σ h m s s s s s -33°49'15"30"45" Right Ascension D e c li n a t i o n G1 Frequency [GHz] F l u x D e n s i t y [ m J y ] = 0.72 ± 0.03 q = 0.023 ± 0.03 Figure 5. – continued. Quici et al. (i) MIDAS J230321 − . EMU-ES contour levels: [5,10,30,100,300] × σ , GLASS 5.5 contour levels: [3,5,10,20,30,40,50] × σ . GLASS 9.5contour levels: [3,5,10,20] × σ h m s s s s s -32°53'00"30"54'00"30" Right Ascension D e c li n a t i o n G1G2G3 F l u x D e n s i t y [ m J y ] = 0.86 ± 0.01 q = 0.019 ± 0.009 (j) MIDAS J230442 − . EMU-ES contour levels: [5,10,30,100,300] × σ , GLASS 5.5 contour levels: [3,5,10,20,30,40,50] × σ . GLASS 9.5contour levels: [3,5,10,20] × σ h m s s s s -34°13'00"20"40"14'00"20" Right Ascension D e c li n a t i o n G1G2 F l u x D e n s i t y [ m J y ] = 1 ± 0.02 q = 0.0079 ± 0.02 Figure 5. – continued. emnant Radio Galaxies Discovered in a Multi-frequency Survey CP S / m J y Core detectedRemnant cand. (with hotspot) Remnant cand. (no hotspot)
Figure 6.
216 MHz CP distribution of radio sources (seeSect. 5.1.1). Core-detected radio galaxies are represented by theblue markers. 3 σ upper limits are placed on the remnant CP,denoted by the left-pointing arrows. Orange and red colored ar-rows are used to indicate remnant candidates with and withouthotspots, respectively. The solid black line gives the value of theCP above which we are complete, given the 10 mJy integratedflux density threshold and the 75 µ Jy beam − average GLASS 5.5detection limit. The orange line traces the lowest CP that can berecovered at the corresponding total flux density. Uncertaintieson the CP are propagated from the uncertainties on the total andcore flux density. A histogram of CP is presented in the top panel. = 1.2 = 1.2 =0.5Core detectedRemnant cand. (with hotspot)Remnant cand. (no hotspot) L A S Figure 7.
The high-frequency spectral index α is plottedagainst the low-frequency spectral index α . A third color-baraxis is over-plotted to show the largest angular size in arc-seconds.Solid black line represents a constant spectral index across bothfrequency ranges. Dashed black line represents a spectral curvatureof SPC = 0 .
5. The red dotted and dot-dashed lines represent a α = − . The integrated spectral properties of our sample areexplored over two frequency ranges. The integratedflux densities at 119, 154, 186, 216 and 399 MHzare used to develop a low-frequency spectral index, α . To compute correct fitting uncertainties, we fitpower-law models to the data in linear space. A high-frequency spectral index, α , is computed using α = log ( S /S )log (887 / , and an associated uncertainty∆ α = / q(cid:0) ∆ S S (cid:1) + (cid:0) ∆ S S (cid:1) . We populateour results onto an α − α plot, e.g. Figure 7, and sum-marize our results in Table 5.Despite the access to high frequencies ( ν = 5.5GHz), we find that the selected remnant candidateswith hotspots show similar spectral properties to radiogalaxies with an active radio core. At low frequencies,remnant candidates with hotspots display spectralindices that are consistent with continuous injection.The high frequency spectral index does appear to besteeper than for the bulk of remnant candidates withhotspots, however, this is also observed for core-detectedradio sources and simply reflects the preferential ageingof higher-energy electrons. No remnant candidates withhotspots display an ultra-steep low-frequency spectralindex, and only one such source (e.g. Sect. 4.2.1)demonstrates a high-frequency ultra-steep spectralindex. Regarding these remnant candidates with nonultra-steep spectra, their position on the α − α plotcan be explained if these are young, recently switchedoff remnants. However, their spectral properties canjust as easily be explained if these are active radiogalaxies in which the radio core is below the GLASSdetection limit. We can not rule out either of thesepossibilities based on their spectra alone, and as suchthey must remain as remnant candidates. We note thatMahatma et al. (2018) also report on a large overlap inthe observed spectral index (150 MHz–1.4 GHz) betweentheir active and candidate remnant radio galaxies. Onlyone of their remnant candidates display hotspots at6 GHz with the Very Large Array telescope, meaningit is not necessarily only the remnant candidates withhotspots which display similar spectral indices as activeradio galaxies. A great example of this is the remnantradio galaxy Blob 1 identified by Brienza et al. (2016).We also find a small fraction (3/94) of core-detectedradio sources which demonstrate an ultra-steep spectralindex. The angular size of these sources is well belowthe largest angular scale that GLASS can recover at5.5 GHz, suggesting the curvature is genuine. This tellsus that ultra-steep selection will not only select radiogalaxies in which the AGN has switched off, however itis interesting to note these sources also do not displayGLASS hotspots. It is possible these sources representrestarted radio galaxies in which the ‘core’ represents a8 Quici et al.
Table 5
Spectral index statistics calculated based on data represented in Figure 7. The median and mean spectral index,indicated by med and mean subscripts, are presented for the low α and high α frequency ranges. f US , low and f US , high represent the low- and high-frequency ultra-steep fractions, respectively. † A range is given here, as it is unclear whether MIDAS J225608 − Sample α , med α , mean α , med α , mean f US , low f US , high Core-detected -0.60 -0.63 -0.66 -0.67 0/94 3/94Remnant cand. (with hotspot) -0.69 -0.69 -0.84 -0.88 0/7 1/7Remnant cand. (without hotspot) -1.27 -1.29 -1.27 -1.38 1/3 (2 − † / We investigate the sample distribution in redshift,total radio power, and largest linear size, given thehost galaxy identifications. We stress that many of theselected remnant candidates have uncertain host galaxyassociations, presenting a major challenge in analysingtheir rest-frame properties. An additional uncertaintycomes from the photometric redshifts, which make up60 /
104 (57%) of the sample. In addition, 14 activeradio galaxies do not have an optical identification,meaning photometric redshift estimates can not beattained. If we assume their host galaxies are at least10 . M (cid:12) , e.g. the lowest stellar mass reported byBest et al. (2005) to host a radio-loud AGN, we canapply the K − z relation (e.g. see Longair & Lilly 1984,Rocca-Volmerange et al. 2004) to estimate the lowestredshift for which a 10 . M (cid:12) galaxy will be undetectedbelow the VIKING K s -band AB magnitude limit(Sect. 2.2.1). This suggests that the host galaxies ofthe 14 unidentified radio galaxies must be above z = 1.The caveat here is that Best et al. (2005) investigatelocal ( z ≤ .
3) radio-loud AGN samples, whereas herewe are assuming higher redshift. Figure 1 of Smolčićet al. (2017) shows a hint of a decline in stellar masswith redshift, however, our assumption of the minimumstellar mass seems valid up to z ∼
1. For each radiosource, we calculate the total radio power following: P = 4 πSD L (1 + z ) − α − , where D L is the luminositydistance. For the 14 radio galaxies without opticalidentifications, the radio power is calculated assuming z ≥
1. Our results are presented in Table 6 and Figure 8.
100 1000Physical size / kpc2425262728 l o g ( P M H z / W H z ) Figure 8.
216 MHz radio power against the largest linear size.Core-detected radio sources (blue markers), remnant candidateswith hotspots (orange markers) and remnant candidates withouthotspots (red markers) are displayed. Circular and square mark-ers are used to denote spectroscopic and photometric redshifts,respectively. Lower limits on the 14 radio sources without hostidentifications are denoted by green arrows. Plotted also are thelargest linear sizes that would result in a 5 σ detection at 216 MHzat z = 0 . z = 1 (red). Limits are calculated assuminga uniform brightness ellipse, and a lobe axis ratio of 2.5 (solidline) and 1.5 (dashed line). Aged remnants often display low axisratios, e.g. MIDAS J225522 − emnant Radio Galaxies Discovered in a Multi-frequency Survey Table 6
Derived distribution averages from Sect. 5.1.3. The number of radio sources included in each category are denotedby N . The redshift, z , radio power, P , and largest linear size (LLS) are presented. The subscripts med and mean refer to themedian and mean values. In the upper half of the table, we consider the entire sample of 104 radio sources. In the lower half,we consider only those with spectroscopic redshifts . † Including the 14 core-detected radio galaxies with z ≥ N z med z mean P med P mean LLS med
LLS mean log (W Hz − ) log (W Hz − ) kpc kpc Full sample.
Core-detected 80 0.519 0.59 25.2 25.2 277 379Core-detected †
94 0.549 0.651 25.3 25.3 294 388Remnant candidates 10 0.504 0.619 25.5 25.8 435 512
Spectroscopic redshifts.
Core-detected 25 0.303 0.290 25.1 25.2 171 322Remnant candidates 3 0.312 0.279 25.2 25.1 351 299Assuming radio lobes continue expanding once thejets switch off, which at least appears to be the casefor FR-II radio galaxies as shown by Godfrey et al.(2017), an expectation of this is for remnants to displaylarger physical sizes with respect to their active radiogalaxy progenitors. Interestingly, our results suggestthat the largest linear sizes of core-detected radiogalaxies and remnant candidates are similar. This maybe explained by the observational bias against remnantradio galaxies of large linear size; such sources willpreferentially fall below a fixed detection limit due totheir lower surface brightness profiles. Since the linearsize and age of a radio galaxy are correlated, for a fixedjet power and environment, it is not unreasonable tosuggest that these ‘missing’ remnants also correspondto older remnant radio galaxies, which in turn wouldimply our sample predominately comprises of youngremnants (see also Mahatma et al. 2018). If so, thisresult is consistent with the low fraction of ultra-steepremnants discussed in in Sect. 5.1.2. We do howevercaution this analysis since seven remnant candidateshave ambiguous host galaxy associations, as well asthe uncertainties surrounding the photometric redshiftestimates.We instead consider the sample of 28 spectroscopically-confirmed radio galaxies to see whether we can draw thesame conclusions as above. Within this ‘spectroscopicsample’, we can be confident not only of the sampleredshifts, but also of the three remnant candidates(e.g. see Sect. 4.1.2, 4.2.1, and 4.2.5) which haveunambiguous host galaxy associations, meaning wecan be confident of their positions on the power-sizediagram. As demonstrated in Figure 8, the absenceof remnant radio galaxies of large linear size becomesquite clear in the ‘spectroscopic sample’, and isconsistent with the previously-discussed conclusions.Our limiting factor here is the small sample size, however this will be addressed in future work wherewe can expect a factor ∼ The fraction of remnant radio galaxy candidates iden-tified in this work provides an upper limit to the gen-uine fraction of remnant radio galaxies, f rem , presentwithin this sample. Of 104 radio galaxies, 10 are identi-fied as remnant radio galaxy candidates, resulting in a f rem ≈
10% upper limit on the remnant fraction. Sari-palli et al. (2012), Brienza et al. (2017) and Mahatmaet al. (2018) each constrain a remnant fraction from ra-dio observation and their results are presented in Table 7.At face value, the remnant fraction obtained in this workis consistent with that of Brienza et al. (2017) and Ma-hatma et al. (2018), and shows a considerable increaseover the fraction constrained by Saripalli et al. (2012).The apparent inconsistency with Saripalli et al. (2012)may very well be a result of their selection. Their samplewas selected at 1.4 GHz, where ultra-steep remnants maypotentially be missed, and they also excluded sourceswithout a radio core but hotspots still present within thelobes. It is interesting to note that despite the differencein the flux limit and angular size cut compared to thesamples complied by Mahatma et al. (2018), Brienzaet al. (2017), and Jurlin et al. (2020), the upper lim-its on the remnant fraction appear consistent. Shabalaet al. (2020) show that the remnant fraction predictedby constant-age models, e.g. those in which the jets are0
Quici et al.
Table 7
Remnant fractions constrained by previous authors. Each column, in ascending order, represents the cited study,the sky coverage over which the sample is compiled, the flux limit across the sample (or the faintest source in the sample),the frequency at which the flux cut is made, the angular size cut of the sample, the number of radio galaxies within thesample, and the resulting remnant fraction.
References. (1) Saripalli et al. (2012), (2) Brienza et al. (2017) and Jurlin et al. (2020), (3) Mahatma et al. (2018),(4) This work.
Ref. Sky coverage Flux limit Sample frequency θ cut Sample size f rem (deg ) (mJy) (MHz) ( )1 7.52 1 1400 30 119 < 4%2 35 40 150 60 158 < 11%3 140 80 150 60 127 < 9%4 8.31 10 216 25 104 4 (cid:46) f rem (cid:46) Table 8
Summarized properties of MIDAS J225337 − χ )is provided to assess the quality of fit. The injection index α inj ,observed-frame break frequency ν b and quiescent fraction T are presented for the fitted continuous injection (CI) andcontinuous injection-off CI-off models. We quote a ∆BICcalculated between the two models. Model χ α inj ν b T ∆BICfitted GHz CI 3.43 -0.608 1.60 ± − ± ± active for a constant duration, is highly sensitive to ob-servable constraints, e.g. the flux and angular size limit.On the other hand the remnant fraction predicted bypower-law age models, e.g. those in which the durationof the active phase is power-law distributed, shows littledependence on observable parameters. It is thereforepossible that the similarity in remnant fractions impliesa preference towards power-law age models that describeAGN jet activity, although this needs to be pursued inmore detailed future work. As presented in Sect. 4, the lobes of seven remnant can-didates display compact emitting regions in GLASS, po-tentially indicating a hotspot formed by an active jet. Asdiscussed in Sections 5.1.1 and 5.1.2, these sources couldbe interpreted either as recently switched off remnants,or, active radio galaxies with unidentified radio cores.We thus propose a lower limit to the remnant fraction,by considering the limiting case where each remnantcandidate with a hotspot is an active radio galaxy. Thiswould suggest a f rem = 4 /
104 ( ≈ A particularly interesting source to examine is MI-DAS J225337 − α < − .
52) spectrum ofthe lobes, suggesting energy losses consistent with anaged remnant radio galaxy, and the compact 5.5 GHzemitting regions in GLASS, which may in turn suggestcurrent or recent energy injection. It is unclear whetherthese are genuine hotspots; the north-eastern lobe is de-tected just above the noise level and thus demonstratesmany ‘hotspot-like’ features. A singular, bright ‘hotspot’is evident in the south-western lobe, however, the emit-ting region is clearly resolved at a physical resolutionof 7 ×
14 kpc (the physical resolution of GLASS 5.5 at z = 0 . Radio AGN inSemi-analytic Environments (RAiSE) code which con-siders, amongst other things: (i) the evolution of themagnetic field strength; (ii) the adiabatic losses con-tributing to the radio luminosity evolution of the lobes;and (iii) the Rayleigh-Taylor mixing of the plasma con-tained within the remnant lobes. These processes aredescribed by the dynamical (Turner & Shabala, 2015)and synchrotron emissivity (Turner et al., 2018b) modelsfor lobes FR-II radio lobes, which follow the continuousinjection model (CI model; Kardashev, 1962; Pachol- emnant Radio Galaxies Discovered in a Multi-frequency Survey F l u x den s i t y [ Jy ] Frequency [GHz] Model CI α inj = 0.608 ν b = 1.60 GHz0.0010.010.11 0.1 1 10 F l u x den s i t y [ Jy ] Frequency [GHz] (a)
Continuous injection model (CI) F l u x den s i t y [ Jy ] Frequency [GHz] Model CI (KGJP) α inj = 0.550 ν b = 2.29 GHz τ /t on = 1.380.0010.010.11 0.1 1 10 F l u x den s i t y [ Jy ] Frequency [GHz] (b)
Continuous injection with ‘off’ component model (CI off)
Figure 9.
Modelled integrated spectrum of MIDAS J225337 − σ uncertainty envelope is represented by the violet shaded region. As discussed in Sect. 5.3, the modeluncertainties take into account only the uncertainties on the flux density measurements, and do not reflect the underlying uncertaintiesdue to an inhomogeneous magnetic field. For reference, a best-fit to the data using single power-law model is represented by a blue line. czyk, 1970; Jaffe & Perola, 1973). The remnant spec-trum follows the continuous injection ‘off’ model (CI-off;Komissarov & Gubanov, 1994) alternatively known asthe KGJP model, which Turner (2018) parameterisewith two break frequencies. A jet power Q = 10 . W,magnetic field strength B = 1 .
08 nT, and total sourceage τ = 71 Myr are constrained by the method of Turneret al. (2018b). The synchrotron spectrum is modelledby the method of Turner (2018) which, provided suffi-cient spectral sampling, allows the injection index α inj ,break frequency ν b and quiescent fraction T to beuniquely constrained. Our results are shown in Fig-ure 9. Both models constrain an injection index thatis consistent within their typically observed range of − . < α inj < − . ( χ ) for each model, and a ∆ BIC= 2 . CI − BIC CI − off ) which demonstrates a preference towardsthe CI-off model (see Table 8). Although the distinc-tion between the two models is ultimately driven bythe 9 GHz measurement, CI-off provides a better modelof the observed spectrum suggesting these are remnantlobes. Modelling the spectrum as a remnant suggeststhe source spends approximately 21 Myr in a remnantphase.Turner et al. (2018b) showed that the CI model pro-vides a statistically significant fit to the broad frequency The quiescent fraction is defined as T = t off /τ , where t off isthe duration of the remnant phase, and τ is the total age of theradio source. radio spectra of over 86 per cent of FR-IIs in the Mullinet al. (2008) sample of 3C sources (typically 12 measure-ments between 0.01 and 10 GHz from Laing & Peacock1980; but see Harwood 2017). Further, despite the non-physical CI model assumption of time-invariant magneticfields, Turner et al. (2018b) find that CI model is anexcellent fit to the simulated spectra of lobed FR-IIwhich do consider magnetic field evolution, and are ableto recover the source dynamical age.As a final sanity check on the duration of the remnantphase fitted above (by the integrated CI-off model),we constrain the age of the most recently acceleratedelectron populations which we assume are located nearthe ‘hotspots’. We convolve the radio observations at399 MHz, 887 MHz and 5.5 GHz images to a common11 circular resolution and measure the integrated fluxdensity within an aperture centered at the southern‘hotspot’. We fit the spectrum with a Tribble JP modelassuming the magnetic field derived previously in ourRAiSE modelling. We arrive at a remnant age of t off =13 +8 − Myr, consistent with the t off ≈
21 Myr derivedfrom the integrated CI-off model. We stress that theobservations are not properly (u,v) matched, and thusare not necessarily seeing the same radio emission; e.g.at 5.5 GHz, the flux density is likely underestimated dueto resolving out the extended radio emission. We leavea more detailed analysis of this source to future work.Next, we model the hotspot to better under-stand its typical fading timescale. We model a ‘MI-DAS J225337 − Quici et al.
50 55 60 65 70
Source age (Myr) − − − − P e a k fl u x ( J y / b e a m ) LobeHotspot, B HS ≈ × B L Hotspot, B HS ≈ × B L σ limit (GLASS 5.5) Figure 10.
A ‘MIDAS J225337 − Q = 10 . W, aninjection energy index of s = 2 .
1, an equipartition factor of
B/B eq = 0 .
22, and a total source age of 71 Myr of which 50 Myris spent in an active phase, and a further 21 Myr is spent as aremnant. The shaded blue bar corresponds to the time duringwhich the source is active, after which the jets are switched off andthe hotspots/lobes begin to fade. The evolution of the synchrotronemission from the lobes (solid black tracks) and the hotspots(dashed tracks) are shown as a function of the total source age.The assumption that the hotspot magnetic field strength is afactor five greater than the lobes (colored in orange) comes fromCygnus A, however we also assume a factor ten increase in thehotspot magnetic field strength (colored in red) to consider shorterfading timescales. We explore this in terms of the peak flux density,as this ultimately decides whether the emitting regions are de-tected in observations. The vertical drop in the flux density tracksreflects the depletion of electrons capable of producing emissionat 5.5 GHz. As expected, the synchrotron emission evolves fasterin the hotspot, however, their fading timescale is non-negligible incomparison to that of the lobes. (without the hotspot) and convolve the output mapwith the GLASS 5.5 synthesized beam. We consider theGLASS 5.5 beam with the largest integrated flux (e.g.the maximum peak flux density), as this ultimately de-termines whether the youngest emitting regions of thelobes are detected. The hotspot is modelled assuming aJP spectrum (Jaffe & Perola, 1973) for the same initialproperties as the lobe, but for an increased magneticfield strength. Cygnus A displays a factor five increasein the hotspot magnetic field strength, in comparison tothat of the lobes (Carilli & Barthel, 1996). We make thesame assumption for MIDAS J225337 − not be visible inMIDAS J225337 − t off ≈
21 Myr rem-nant phase. At the implied age of the remnant, themodelled emitting regions of the lobes range between100–150 kpc in size. This is consistent with the observedemitting regions in GLASS, which demonstrate a pro-jected linear size of ∼
96 kpc. This indicates that thebright features in GLASS are entirely consistent withthe youngest plasma regions, and are not necessarily thehotspots of an active jet.As such, the interpretation of this source is challeng-ing for the following reasons. Modelling the spectrumsuggests these are the lobes of a remnant radio galaxy.While hotspots will remain visible for a non-negligibleperiod of time after injection is switched off, they shouldnot be visible in this source assuming the remnant age iscorrect. The two pieces of evidence are consistent withone another if the bright GLASS features are just theyoungest emitting regions of the lobes. If we alternativelyassume MIDAS J225337 − τ = 100 Myr) to the remnant case but asubstantially lower jet power of Q = 10 . W is requiredto match the relatively low observed flux density. Thisjet power is typically associated with an FR-I morphol-ogy for the known host galaxy mass (e.g. Ledlow, 1994; emnant Radio Galaxies Discovered in a Multi-frequency Survey f rem ≈ characteristic active lifetime of ∼
100 Myr, ourobserved remnant fraction would thus suggest an observ-able remnant phase lasting ∼
10 Myr. The modellingof a ‘MIDAS J225337 − Within a sub-region of 8.31 deg of the GAMA 23 field,we have compiled a sample consisting of 104 extended,low-frequency selected (216 MHz) radio galaxies. Usingthe 5.5 and 9.5-GHz GLASS survey, we have adopted the‘absent radio core’ criterion to search for remnant radiogalaxy candidates. Our conclusions are summarized asfollows:• We identify 10 new remnant radio galaxy candi-dates, thereby constraining an f rem ≤
10% upperlimit on the fraction of radio galaxies with quies-cent AGN. Our upper limit is consistent with thatproposed by previous authors, and suggests thatremnants must have a short observable lifetime.• Seven remnant candidates show compact emittingregions in GLASS, an observation that can only beexplained if the jets have recently switched off. Amuch smaller fraction (3/10) show relaxed, hotspot-less lobes, and only one displays an ultra-steepspectrum across the entire frequency range. Thisimplies remnants are detected soon after switchingoff, suggesting a rapid fading during the remnantphase.• The small fraction of ultra-steep ( α < − .
2) rem-nants is likely a result of the oldest remnant lobesescaping detection due to their expansion.• At present, the upper limits placed on the remnantcore prominence are too weak to confidently ruleout the presence of AGN activity. Those with com-pact hotspots and a non ultra-steep spectrum musttherefore retain their remnant candidate classifi-cation. Considering the limiting case in which all these are active radio galaxies, we would expect a f rem ≈
4% remnant fraction.• MIDAS J225337 − − −
10 Myrat 5.5 GHz after the jets switch off. This would im-ply that the presence of a hotspot in radio mapsmay not necessarily reflect an active jet, and byextension we can expect an appreciable fraction ofgenuine remnants to still display hotspots.
BQ acknowledges a Doctoral Scholarship and an Aus-tralian Government Research Training Programme schol-arship administered through Curtin University of West-ern Australia. NHW is supported by an AustralianResearch Council Future Fellowship (project numberFT190100231) funded by the Australian Government.This scientific work makes use of the Murchison Radio-astronomy Observatory, operated by CSIRO. We ac-knowledge the Wajarri Yamatji people as the traditionalowners of the Observatory site. Support for the opera-tion of the MWA is provided by the Australian Govern-ment (NCRIS), under a contract to Curtin Universityadministered by Astronomy Australia Limited. We ac-knowledge the Pawsey Supercomputing Centre whichis supported by the Western Australian and AustralianGovernments. The Australian SKA Pathfinder is partof the Australia Telescope National Facility which ismanaged by CSIRO. Operation of ASKAP is fundedby the Australian Government with support from theNational Collaborative Research Infrastructure Strat-egy. The Australia Telescope Compact Array is partof the Australia Telescope National Facility which isfunded by the Australian Government for operationas a National Facility managed by CSIRO. We thankthe staff of the GMRT that made these observationspossible. GMRT is run by the National Centre for Ra-dio Astrophysics of the Tata Institute of FundamentalResearch. CHIC acknowledges the support of the De-partment of Atomic Energy, Government of India, underthe project 12-R&D-TFR-5.02-0700. SW acknowledgesthe financial assistance of the South African Radio As-tronomy Observatory (SARAO) towards this researchis hereby acknowledged ( ). IP acknowl-edges support from INAF under the SKA/CTA PRIN"FORECaST" and the PRIN MAIN STEAM "SAuROS"projects. The National Radio Astronomy Observatory isa facility of the National Science Foundation operatedunder cooperative agreement by Associated Universities,Inc. H.A. benefited from grant CIIC 90/2020 of Univer-sidad de Guanajuato, Mexico. We acknowledge the workand support of the developers of the following followingpython packages: Astropy (Astropy Collaboration et al.,4
Quici et al. and Topcat(Taylor, 2005). We thank an anonymous referee for theirinsightful comments that have improved the manuscript.This work was compiled in the very useful free onlineL A TEX editor Overleaf.
REFERENCES
Astropy Collaboration et al., 2013, A&A, 558, A33Beardsley A. P., et al., 2020, PASA, 37, e014Bellstedt S., et al., 2020, MNRAS, 496, 3235Best P. N., 2009, Astronomische Nachrichten, 330, 184Best P. N., Kauffmann G., Heckman T. M., Brinch-mann J., Charlot S., Ivezić Ž., White S. D. M., 2005,MNRAS, 362, 25Brammer G. B., van Dokkum P. G., Coppi P., 2008,ApJ, 686, 1503Brienza M., et al., 2016, A&A, 585, A29Brienza M., et al., 2017, A&A, 606, A98Brienza M., et al., 2018, A&A, 618, A45Brienza M., et al., 2020, arXiv e-prints, p.arXiv:2003.13476Briggs D. S., 1995, in American Astronomical SocietyMeeting Abstracts. p. 112.02Carilli C. L., Barthel P. D., 1996, A&A Rev., 7, 1Churazov E., Brüggen M., Kaiser C. R., Böhringer H.,Forman W., 2001, ApJ, 554, 261Condon J. J., Cotton W. D., Greisen E. W., Yin Q. F.,Perley R. A., Taylor G. B., Broderick J. J., 1998, AJ,115, 1693Cordey R. A., 1987, MNRAS, 227, 695Driver S. P., et al., 2011, MNRAS, 413, 971Duffy P., Blundell K. M., 2012, MNRAS, 421, 108Fanaroff B. L., Riley J. M., 1974, MNRAS, 167, 31PFeretti L., Gioia I. M., Giovannini G., Gregorini L.,Padrielli L., 1984, A&A, 139, 50Giovannini G., Feretti L., Gregorini L., Parma P., 1988,A&A, 199, 73Godfrey L. E. H., Morganti R., Brienza M., 2017, MN-RAS, 471, 891Hancock P. J., Murphy T., Gaensler B. M., Hopkins A.,Curran J. R., 2012, MNRAS, 422, 1812Hancock P. J., Trott C. M., Hurley-Walker N., 2018,PASA, 35, e011Hardcastle M. J., 2018, MNRAS, 475, 2768Hardcastle M. J., Looney L. W., 2008, MNRAS, 388,176Harwood J. J., 2017, MNRAS, 466, 2888Harwood J. J., Hardcastle M. J., Croston J. H., 2015,MNRAS, 454, 3403 http://ds9.si.edu/site/Home.html Hurley-Walker N., et al., 2015, MNRAS, 447, 2468Hurley-Walker N., et al., 2017, MNRAS, 464, 1146Huynh M. T., Bell M. E., Hopkins A. M., Norris R. P.,Seymour N., 2015, MNRAS, 454, 952Huynh M. T., Seymour N., Norris R. P., Galvin T., 2020,MNRAS, 491, 3395Ishwara-Chandra C. H., Taylor A. R., Green D. A., StilJ. M., Vaccari M., Ocran E. F., 2020, MNRAS, 497,5383Jaffe W. J., Perola G. C., 1973, A&A, 26, 423Johnston S., et al., 2007, PASA, 24, 174Jurlin N., et al., 2020, arXiv e-prints, p.arXiv:2004.09118Kardashev N. S., 1962, Soviet Ast., 6, 317Komissarov S. S., Gubanov A. G., 1994, A&A, 285, 27Laing R. A., Peacock J. A., 1980, MNRAS, 190, 903Ledlow M. J., 1994, in American Astronomical SocietyMeeting Abstracts emnant Radio Galaxies Discovered in a Multi-frequency Survey arXiv:astro-ph/0612759 )Scheuer P. A. G., 1974, MNRAS, 166, 513Schoenmakers A. P., de Bruyn A. G., Röttgering H. J. A.,van der Laan H., Kaiser C. R., 2000, MNRAS, 315,371Shabala S. S., Jurlin N., Morganti R., Brienza M., Hard-castle M. J., Godfrey L. E. H., Krause M. G. H.,Turner R. J., 2020, MNRAS,Shulevski A., Morganti R., Oosterloo T., Struve C., 2012,A&A, 545, A91Shulevski A., et al., 2017, A&A, 600, A65Smolčić V., et al., 2017, A&A, 602, A6Taylor M. B., 2005, in Shopbell P., Britton M., Ebert R.,eds, Astronomical Society of the Pacific ConferenceSeries Vol. 347, Astronomical Data Analysis Softwareand Systems XIV. p. 29Thompson A. R., Clark B. G., Wade C. M., Napier P. J.,1980, ApJS, 44, 151Tingay S. J., et al., 2013, PASA, 30, e007Turner R. J., 2018, MNRAS, 476, 2522Turner R. J., Shabala S. S., 2015, ApJ, 806, 59Turner R. J., Rogers J. G., Shabala S. S., Krause M.G. H., 2018a, MNRAS, 473, 4179Turner R. J., Shabala S. S., Krause M. G. H., 2018b,MNRAS, 474, 3361Wayth R. B., et al., 2015, PASA, 32, e025Wayth R. B., et al., 2018, PASA, 35, 33Wilson W. E., et al., 2011, MNRAS, 416, 832Wright E. L., et al., 2010, AJ, 140, 1868van der Walt S., Colbert S. C., Varoquaux G., 2011,Computing in Science Engineering, 13, 226
Quici et al.
Table 9
Column descriptions corresponding to the supplementary electronic table. The MIDAS_Name is derived using theMIDAS_RA and MIDAS_Dec columns. Entries without a GAMA_IAUID, mag_Ks, or NED_Object_Name are assigned avalue of ’ − No. Column name Unit Description1 MIDAS_name J hh:mm:ss-dd:mm:ss Name of radio source in J2000 format.2 MIDAS_RA deg Right Ascension of MIDAS source.3 MIDAS_Dec deg Declination of MIDAS source.4 AGN_status – Activity state of AGN.5 FR_classification – Fanaroff & Riley Classification: FR-I, FR-II.6 LAS arc-seconds Largest angular size measured from EMU-ES.7 Hotspots (L1) – Number of GLASS hotspots present in L1.8 Hotspots (L2) – Number of GLASS hotspots present in L2.9 peak_flux_core µ Jy/beam 5.5 GHz radio core peak flux density.10 err_peak_flux_core µ Jy/beam Error on 5.5 GHz radio core peak flux density.11 S mJy Integrated flux density at MHz.12 err_S mJy Error on integrated flux density at MHz.13 S mJy Integrated flux density at MHz.14 err_S mJy Error on integrated flux density at MHz.15 S mJy Integrated flux density at MHz.16 err_S mJy Error on integrated flux density at MHz.17 S mJy Integrated flux density at MHz.18 err_S mJy Error on integrated flux density at MHz.19 S mJy Integrated flux density at MHz.20 err_S mJy Error on integrated flux density at MHz.21 S mJy Integrated flux density at MHz.22 err_S mJy Error on integrated flux density at MHz.23 S mJy Integrated flux density at MHz.24 err_S mJy Error on integrated flux density at MHz.25 S mJy Integrated flux density at MHz.26 err_S mJy Error on integrated flux density at MHz.27 S mJy Integrated flux density at MHz.28 err_S mJy Error on integrated flux density at MHz.29 S mJy Integrated flux density at MHz.30 err_S mJy Error on integrated flux density at MHz.31 CP – 216 MHz core prominence.32 GAMA_IAUID – ID of host galaxy in GAMA photometry catalogue.33 Host_RA deg Right ascension of host galaxy.34 Host_Dec deg Declination of host galaxy.35 z – Redshift.36 z_type – Redshift type: Spectroscopic, Photometric, lower-limit.37 mag_K s – Host galaxy VIKING K s -band magnitude.38 NED_Object_Name – Name of NED object corresponding to the GAMA_IAUID.39 log10_L log10(W Hz −1