Powerful Radio Sources in the Southern Sky. II. A SWIFT X-Ray Perspective

We recently constructed the G4Jy-3CRE, a catalog of extragalactic radio sources based on the GLEAM 4-Jy (G4Jy) sample, with the aim of increasing the number of powerful radio galaxies and quasars with similar selection criteria to those of the revised release of the Third Cambridge catalog (3CR). The G4Jy-3CRE consists of a total of 264 radio sources mainly visible from the Southern Hemisphere. Here, we present an initial X-ray analysis of 89 G4Jy-3CRE radio sources with archival X- ray observations from the Neil Gehrels Swift Observatory. We reduced a total of 615 Swift observations, for about 0.89 Msec of integrated exposure time, we found X-ray counterparts for 61 radio sources belonging to the G4Jy-3CRE, 11 of them showing extended X-ray emission. The remaining 28 sources do not show any X-ray emission associated with their radio cores. Our analysis demonstrates that X-ray snapshot observations, even if lacking uniform exposure times, as those carried out with Swift, allow us to (i) verify and/or re ne the host galaxy identi cation; (ii) discover the extended X-ray emission around radio galaxies of the intracluster medium when harbored in galaxy clusters, as the case of G4Jy 1518 and G4Jy 1664, and (iii) detect X-ray radiation arising from their radio lobes, as for G4Jy 1863.


INTRODUCTION
The Third Cambridge catalog (3C; Edge et al. 1959) and its revised versions (3CR and 3CRR;Bennett 1962;Laing et al. 1983, respectively) is widely considered the gold standard amongst catalogs of powerful radio sources.Since their releases, they have enabled core investigations into the nature of radio-loud active galactic nuclei (i.e., radio galaxies and quasars Begelman et al. 1984;Urry & Padovani 1995;Harvaneck et al. 2001;Hardcastle & Croston 2020), their environments at all scales and feedback processes occurring therein (McNamara & Nulsen 2007, 2012;Fabian 2012;Morganti 2017).
However, despite its success, the 3C suffers from an artificial limitation since it is restricted to the Northern Hemisphere.About 2/3 of 3C sources are not visible from the Southern Hemisphere thus having limited access to modern astronomical facilities and instruments, as the Multi Unit Spectroscopic Explorer (MUSE; Bacon et al. 2010) and the Enhanced Resolution Imager and Spectrograph (ERIS; Kenworthy et al. 2018) mounted at the Very Large Telescope (VLT), the Atacama Large Millimeter/submillimeter Array (ALMA), the MeerKAT radio telescope (see e.g., Sejake et al. 2023) and the High Energy Stereoscopic System (HESS) at very high energies.In the near future, new facilities located in the Southern Hemisphere will include the Square Kilometre Array (SKA; McMullin et al. 2020), the Vera Rubin Observatory (a.k.a., Large Synoptic Survey Telescope -LSST; Ivezić et al. 2019), the Extremely Large Telescope2 (ELT) and the Cherenkov Telescope Array (CTA).
Several attempts were made to create the southern equivalent of the 3C.For instance, Best et al. (1999) used the Molonglo Reference Catalogue (MRC; Large et al. 1981) to select an equatorial sample of 178 radio sources with flux density above 5 Jy at 408 MHz, in the range of declination between -30 • and 10 • and having Galactic latitudes |b| ≥ 10 • .This sample is characterized by a high spectroscopic completeness and its footprint mitigates the 3C observability limitation.Burgess & Hunstead (2006a,2006b) created the Molonglo Southern 4 Jy sample (MS4), a sample of southern radio sources at 408 MHz with integrated flux densities S 408 > 4.0 Jy.However it was only until the Murchison Widefield Array (MWA Tingay et al. 2013) became operational in Western Australia that the southern counterpart of the 3C at ∼178 MHz was finally created.We recently built a sample of 264 extragalactic radio sources extracted from the GaLactic and Extragalactic All-sky MWA 4-Jy (G4Jy) catalog (White et al. 2020a,b), namely: the G4Jy-3CRE (Massaro et al. 2022, hereinafter paper I), that is the southern equivalent to the 3CR in terms of its nominal flux limit threshold of 9 Jy (Bennett 1962;Spinrad et al. 1985).
In paper I, we carried out the comparison between archival radio maps and optical images to search for G4Jy-3CRE host galaxies.This analysis was combined with an extensive literature search performed to identify those radio sourceswith a redshift, z, measurement.We found that 79% of the G4Jy-3CRE sources (i.e., 208 out of 264) have a clear optical counterpart of their radio cores.For 181 radio sources, the optical counterpart is also coincident with the mid-IR one reported in the G4Jy catalog (White et al. 2020a,b).Using both NASA Extragalactic Database (NED) 3 and the SIMBAD Astronomical Database 4 , we found spectroscopic z measurements for a total of 145 sources (out of 264), corresponding to ∼55% of the G4Jy-3CRE catalog.To be considered for this work, redshifts must have: (i) a published figure of the optical spectrum, or (ii) a description of such spectrum with emission and/or absorption lines clearly reported in a table and/or in the publication.These conservative criteria were already adopted in previous analyses and spectroscopic campaigns (see e.g., Massaro et al. 2016;Peña-Herazo et al. 2020, 2022;Kosiba et al. 2022).
In this second paper of the series, we present a first X-ray perspective of the G4Jy-3CRE catalog based on targeted X-ray observations.By searching the archive of the X-Ray Telescope (XRT; Burrows et al. 2000Burrows et al. , 2005) ) on board the Neil Gehrels Swift Observatory (Gehrels et al. 2004), we found that a total of 90 G4Jy-3CRE radio sources, out of 264, were already observed, with at least one observation having nominal exposure time, T exp , above 250 s.The main aims of the current X-ray analysis are to: (i) use the position of X-ray counterparts, when detected, to verify and eventually refine results of the previous optical analysis (see paper I for more details); (ii) test which sources show extended X-ray emission that could be a signature of emission from the intracluster medium (ICM) for those harbored in galaxy clusters and groups and (iii) obtain measurements of their X-ray count rate, useful to plan X-ray follow up observations.
We also compared Swift X-ray images with radio maps at different frequencies.Similarly to paper I, we mainly used those radio maps available in the databases of the Very Large Array (VLA) Low-Frequency Sky Survey Redux5 (VLSSr; Cohen et al. 2007), the Tata Institute of Fundamental Research (TIFR) Giant Metrewave Radio Telescope (GMRT) Sky Survey (TGSS; Intema et al. 2017), the Sydney University Molonglo Sky Survey (SUMSS; Mauch et al. 2003), the National Radio Astronomy Observatory (NRAO) VLA Sky Survey (NVSS; Condon et al. 1998), the Very Large Array (VLA) Sky Survey (VLASS; Lacy et al. 2020), corresponding to nominal frequencies of 74 MHz, 150 MHz, 843 MHz, 1.4 GHz and 3 GHz, respectively, and the NRAO VLA Archive Survey (NVAS)6 databases.This analysis is only devoted to X-ray observations and it does not include any optical and ultraviolet investigation feasible thanks to the observations collected with the Ultra-violet Optical Telescope (UVOT) instrument (Roming et al. 2005) on board Swift.A dedicated paper on broadband photometry for the G4Jy-3CRE sample, that includes also UVOT data, is currently in preparation (García-Pérez et al. 2022).
This paper is structured as follows: § 2 is dedicated to a brief overview of the G4Jy-3CRE catalog and the selection criteria underlying the Swift-XRT archival search.§ 3 is devoted to a description of data reduction and analysis procedures adopted here while § 4 is dedicated to the results we achieved.A comparison with previous X-ray analyses is also presented in § 5 while § 6 illustrates the search for mid-infrared and optical counterparts using X-ray images.Summary, conclusions and future perspectives are given in § 7. Finally, Appendix A reports all Swift X-ray images with radio contours overlaid, while Appendix B is devoted to a comparison between Chandra and Swift X-ray observations for four sources having unpublished Chandra datasets.
As in paper I, we adopt cgs units for numerical results and we assume a flat cosmology with H 0 = 69.6 km s −1 Mpc −1 , Ω M = 0.286 and Ω Λ = 0.714 (Bennett et al. 2014).Spectral indices, α, are defined by flux density, S ν ∝ ν −α .For optical images, we used those available in the archives of the Panoramic Survey Telescope & Rapid Response System (Pan-STARRS; Flewelling et al. 2020) and the Dark Energy Survey (DES; Abbott et al. 2018).These optical data were complemented by images available in the red filter of the Digital Sky Survey7 (DSS), for those sources outside the Pan-STARRS and DES footprints.Pan-STARRS magnitudes are reported in the AB system (Oke 1974;Oke & Gunn 1983) while DES magnitudes are given in optical filters similar to those of Pan-STARRS and the Sloan Digital Sky Survey (SDSS; Ahn et al. 2012).

SAMPLE SELECTION
The G4Jy-3CRE sample contains a total of 264 extragalactic radio sources (paper I).The parent sample of the G4Jy-3CRE catalog was the larger complete sample of extragalactic radio sources with flux density above the threshold of 4 Jy at 151 MHz, namely the G4Jy catalog (White et al. 2020a,b) based on the recent GaLactic and Extragalactic All-sky MWA survey8 (GLEAM Wayth et al. 2015;Hurley-Walker et al. 2017).It is worth noting that MWA is considered as the SKA precursor at low radio frequencies (Tingay et al. 2013).
All radio sources included in the G4Jy-3CRE catalog lie at Dec. < −5 • , also having Galactic latitudes |b| >10 • and are located outside the original footprint of the 3CR catalog.G4Jy-3CRE sources have flux density measurements at 174 MHz S 174 and at 181 MHz S 181 , integrated in the GLEAM sub-bands, above 8.13 Jy and 7.85 Jy, respectively, where these thresholds correspond to the 9 Jy limiting sensitivity, at ∼178 MHz, that is the nominal value adopted to create the 3CR.The flux density thresholds were computed assuming a power-law radio spectrum and adopting the spectral index reported in the G4Jy catalog.Thus the G4Jy-3CRE is equivalent, in terms of radio flux density selection, to the northern 3CR extragalactic sample (Bennett 1962;Spinrad et al. 1985).We highlight that relatively bright sources at Dec. < +30 • and Galactic latitudes |b| > 10 • , including the Orion Nebula, were all masked in the GLEAM extragalactic catalog and are not listed in the G4Jy sample (see White et al. 2020a,b, and references therein for a list of masked sources and additional details).
Here, we crossmatched the G4Jy-3CRE catalog with the Swift archive, adopting a search radius of 10 ′ , and then we selected only sources having, at least, one XRT observation with nominal T exp longer than 250 s.Our search was restricted only to X-ray observations performed in Photon Counting mode (Hill et al. 2004) to ensure that an X-ray image can be obtained and compared with those at other frequencies.Adopting these criteria, we found a total of 90 radio sources with Swift-XRT data available in the archive.We excluded G4Jy 1038 (a.k.a.3C 279), one of the most famous blazars (see e.g., Lynds et al. 1965;Stocke et al. 1998) because its Swift-XRT observations are already, extensively, discussed in the literature (see e.g.Collmar et al. 2010;Hayashida et al. 2015;Larionov et al. 2020).
We processed a total of 615 individual observations for 89 radio sources belonging to the G4Jy-3CRE catalog.All these observations were acquired between May 2005 and November 2022.Given that G4Jy 1748 and G4Jy 1749 lie in the same field by the Swift-XRT, within ∼4 ′ angular separation, we processed only the 9 archival observations centered on the former radio source.
The distribution of the integrated T exp for the 89 radio sources is shown in Figure 1.We remark that even when the integrated T exp appears relatively large (i.e., above ∼ 10 ksec), it is the result of merging several, relatively short, observations.Since the integrated T exp , for a large fraction of the sources (i.e., ∼90%), is smaller than ∼ 25 ksec, this prevents us from performing a uniform spectral analysis.Thus, we focused on searching for X-ray counterparts and refining the mid-IR and optical analysis.
Figure 1.Integrated exposure times Texp for all merged event files reduced in our analysis.The black histogram refers to radio sources having an X-ray detected counterpart while the red one represents those for which no X-ray source is spatially coincident with the radio core.A total of 627 observations were retrieved from the Swift-XRT archive and analyzed to investigate the X-ray emission of 89 G4Jy-3CRE radio sources.The two dashed vertical lines mark the threshold of 5 ksec and 12 ksec, respectively.We found an X-ray counterpart for all radio sources with integrated Texp larger than ∼ 12 ksec and for 85% of the sample (i.e., 35 out of 41) observed for more than 5 ksec.
During our archival search, we also found 4 radio sources, namely, G4Jy 171, G4Jy 260, G4Jy 411 and G4Jy 1613 with unpublished Chandra X-ray observations.In these cases, we also reduced these Chandra data and we compared them with the results obtained from the Swift-XRT images.A brief overview of the Chandra data reduction procedure adopted here, together with the comparison between X-ray images collected by the two satellites, is reported in Appendix B.
3. Swift X-RAY DATA REDUCTION AND ANALYSIS

X-ray source detection
We adopted the same data reduction procedures described in previous Swift-XRT analyses (see e.g.Massaro et al. 2008a,b;Paggi et al. 2013;Marchesini et al. 2019Marchesini et al. , 2020)), providing results in agreement with those of the Swift-XRT X-Ray point source catalogs (D'Elia et al. 2013;Evans et al. 2014Evans et al. , 2020)).Therefore, we report here only basic details about the data reduction while additional information can be obtained from the references previously listed.The entire X-ray analysis performed here and all X-ray images shown in the following sections are restricted to the 0.5-10 keV energy range, unless stated otherwise.
Raw Swift-XRT data were downloaded from the archive9 and reduced to obtain clean event files using the xrtpipeline task, which is part of the Swift X-Ray Telescope Data Analysis Software (XRTDAS Capalbi et al. 2005), distributed within the HEAsoft package (version 6.30.1).Event files were calibrated and cleaned with standard filtering criteria using the xrtpipeline task, combined with the latest calibration files available in the High Energy Astrophysics Science Archive Research Center (HEASARC) calibration database (CALDB) version (v.20220907)10 .In particular, using xselect, we excluded time intervals (i) with count rates higher than 40 photons per second and (ii) having the CCD temperature exceeding -50 • C in regions located at the edges of the XRT camera (D'Elia et al. 2013).Clean event files, for the same sources, were merged using the xselect task, while corresponding exposure maps were merged with the XIMAGE software11 .
We created images for all merged event files and a detection run was then performed using the sliding cell DETECT (i.e., det) algorithm in XIMAGE (Giommi et al. 1992).We chose a threshold of the signal-to-noise ratio (SN R) equal to 3 to claim an X-ray detection and we adopted the following X-ray detection flags (XDF) to characterize the detection and the identification: • x (X-ray counterpart): all G4Jy-3CRE sources having a unique X-ray source located at the position of the radio source, within the X-ray positional uncertainty region, e.g.G4Jy 85 shown in left panel of Figure 2; • e (extended X-ray source): radio sources for which the X-ray counterpart, detected in the Swift-XRT image, shows diffuse/extended X-ray emission, e.g.G4Jy 540 in the central panel of Figure 2; • u (undetected X-ray counterpart): those merged event files for which the det algorithm did not report an X-ray detection, above the chosen threshold, at the location of the radio source (see e.g.G4Jy 718 in the right panel of Figure 2).We also used this flag for 4 sources for which their X-ray emission is contaminated by a nearby extended objects, thus making it challenging to claim a detection (see following sections for more details).
For all sources that are undetected in X-rays with the det algorithm (labeled u in Table 2), we carried out a second detection run using the SOurce STAtistic (sosta) tool available in XIMAGE.The sosta algorithm uses a local background to determine the significance of a source detection and to estimate its count rate while detections performed using the det procedure are based on the global background.Thus, the former algorithm could be more accurate for observations with relatively low integrated T exp .To perform this detection run we were prompted to select the radio source position to compute background, source counts, count rate, and detection significance.Radio sources having a candidate X-ray counterpart detected using sosta have all SN R below 3, computed using the det task, and all the number of photons are reported without uncertainties in Table in Table 2 to indicate that they correspond to a 3σ upper limit.The sosta algorithm does not provide the X-ray source position and thus it is not reported in Table 2.
In all merged event files, where we detected an X-ray counterpart of G4Jy-3CRE radio sources, we measured: (i) the coordinates of the X-ray centroid (J2000 equinox) using the xrtcentroid task, when detected using the det algorithm; (ii) the number of photons n 90 , within a circular region of radius equal to 120.207 ′′ (i.e., 51 pixels) centered on the X-ray coordinates enclosing 90% of the Swift-XRT PSF; (iii) the expected number of background photons for the same area n b,90 ; (iii) the count rate, in photons/sec and the SN R of the X-ray detection when obtained with the det algorithm.If no statistical uncertainty is reported on the measured count rate, this corresponds to a 3σ upper limit obtained with the sosta algorithm.All these parameters, together with the total number of X-ray observations processed per source, the integrated T exp and the value of the XDF are reported in Table 2.The X-ray image, obtained from the merged event file of G4Jy 85, smoothed with a Gaussian kernel of 10 pixels (i.e., 23.57′′ ).The magenta circle is centered on the position of the X-ray counterpart associated with this radio sources and it has a radius of 120.207 ′′ (i.e., 51 pixels) enclosing 90% of the Swift-XRT PSF.Radio contours are drawn using the radio map at 150 MHz starting at level of 0.3 Jy/beam and increasing by a factor of 3.This is a clear example of a radio source for which the X-ray counterpart of its radio core is detected at the level of the SN R reported in the figure.Central panel) same image as the left panel but for G4Jy 540, an example of detected extended X-ray emission.Here radio contours were computed using the archival radio map at 150 MHz starting at level of 0.1 Jy/beam and increasing by a factor of 6.Right panel) same image as the left panel but for G4Jy 718, an example of a radio source lacking an X-ray counterpart.In this case the magenta circle is centered on the radio position reported in the G4Jy catalog.Radio contours overlaid to the X-ray image are drawn from the radio map at 843 MHz starting at level of 0.2 Jy/beam and increasing by a factor of 2.
Finally, in Appendix A, we show X-ray images for all merged event files with radio contours overlaid and a magenta dashed circle of 120.207 ′′ superimposed, as in Figure 2. The circular region is centered on either the location of the X-ray counterpart of each G4Jy-3CRE radio source, if detected in the Swift-XRT, or on its radio coordinates, as reported in the G4Jy catalog, if not.The frequencies of the radio maps used to draw the radio contours and the contour parameters are reported in Appendix A (see Table 16 for more details).

Extended X-ray emission
To determine whether an X-ray source is extended we adopted the following criterion.We assumed a Poissonian distribution for all measured numbers of photons and their uncertainties.We counted the difference δ n between the number of photons (n 10 ) within an annulus of inner radius equal to 40.069 ′′ (i.e., 17 pixels) and outer radius of 120.207 ′′ (i.e., 51 pixels) and that of background photons n b,10 expected in a region having the same area: δ n = n 10 − n b,10 .According to the latest model release of the Point Spread function (PSF) for Swift-XRT 12 a circular region of 40.069 ′′ and 120.207 ′′ encloses 80% and 90% of the PSF, respectively.Thus, we expect that the measured number of photons δ n, for a point-like source, corresponds to ∼10% of the total flux.
Consequently, we computed the 10% of the total number of expected photons n ex rescaling those (i.e., n 90 ) measured within a circular region of 120.207 ′′ and subtracting the average background photons (n b,90 ) expected for the same area: n ex = 1 9 (n 90 − n b,90 ).We considered as extended those X-ray sources for which the difference (δ n − n ex ) is larger than zero within a level of confidence of 3σ.The number of background photons was obtained by measuring counts within a circular region of 102 pixels in radius, masking other X-ray sources detected using the det algorithm with SN R > 3, in all merged event files, and then rescaled for the ratio of the areas.For extended X-ray sources, if the det algorithm detected more than a single source close to the X-ray intensity peak, we only reported in Table 2 the one with the highest SN R. 1.4 Note-Col.(1) the G4Jy name of the radio source, also adopted in the G4Jy-3CRE catalog; col.(2, 3) the Right Ascension and the Declination (Equinox J2000) of the brightness-weighted radio centroid as reported in the G4Jy catalog; col.(4, 5) the Right Ascension and the Declination (Equinox J2000) measured from the distribution of the X-ray photons for radio sources with XDF=x or XDF=e; col.( 6) the number of photons measured in the 0.5-10 keV energy range within a circular region enclosing 90% of the Swift-XRT PSF, together with the average number of photons expected in the background for the same area, as reported in parenthesis; col.( 7) the X-ray count rate measured in photons/sec with the 1σ uncertainty in parenthesis as obtained using det algorithm.For those sources having not statistical uncertainty reported, thus indicated with a dashed line, the count rate corresponds to a 3σ upper limit obtained using the sosta algorithm; col.( 8) the value of the SN R for all radio sources with an X-ray counterpart detected using the det algorithm.X-ray counterparts, at SN R < 3 correspond to those detected running sosta algorithm and in these cases no X-ray coordinates were computed; col.(9) the integrated Texp; col.( 10) the total number of sources reduced and analyzed; col.( 11) the XDF assigned in our analysis ("x " labels radio sources with a detected X-ray point-like counterpart; "e" stands for radio sources with extended X-ray emission around their radio cores and "u" for G4Jy-3CRE with radio cores undetected in the X-ray band); col.( 12 (ξ) Morton & Tritton (1982), (see also paper I for more details).

An X-ray overview of the whole sample
We found that all 25 G4Jy-3CRE sources with an integrated exposure time above T exp ≃ 9 ksec have an X-ray counterpart.The only exception is G4Jy 373, an FR II radio galaxy at z = 0.1126 (Carter et al. 1983) lying behind the Fornax Cluster at an angular separation of ∼5 ′ from NGC 1399 (see e.g., Killeen et al. 1988;Hilker et al. 1999), for which the extended X-ray emission of the nearby galaxy cluster prevents us from detecting its X-ray counterpart.When selecting a threshold of integrated T exp above 5 ksec, 88% of sources (i.e., 37 out of 41) have an X-ray counterpart, as shown in Figure 1.
There are 28 radio sources with no X-ray emission associated with their radio cores, flagged as undetected.This list also includes G4Jy 77 and G4Jy 1605 that are two extended radio sources, the former is a radio phoenix of the galaxy cluster Abell 85 (see e.g., Bagchi et al. 1998;Kempner et al. 2004;Ichinohe et al. 2015) while the latter is the radio relic of Abell 3667 (see e.g., Johnston-Hollitt et al. 2008;Owers et al. 2009).These sources are not expected to be associated with an X-ray point-like counterpart.
We also found 11 radio sources with extended X-ray emission around them.However, for three out of 11 sources, namely: G4Jy 416, G4Jy 651, G4Jy 1749, such diffuse emission is contaminated by that of a nearby X-ray object.These sources are all marked as undetected.Signatures of extended X-ray emission are also evident in G4Jy 1863, the only source for which we claim the X-ray detection of its radio lobes, as described below.The remaining 50 radio sources have a detected X-ray point-like counterpart associated with their radio core.

Extended X-ray emission around G4Jy 1518 and G4Jy 1664
We found clear signatures of extended X-ray emission around two radio sources in the Swift-XRT images: G4Jy 1518 and G4Jy 1664.
G4Jy 1518 belongs to the equatorial sample of Best et al. (1999) and it is classified as a radio galaxy lying at z=0.226.It is also listed in the Molonglo Reference Catalogue of radio sources (MRC 1912-269 Large et al. 1981).From a radio perspective, G4Jy 1518 appears to be a "restarted" radio galaxy (), as shown in the archival radio maps of Figure 3.The host-galaxy identification is consistent with detection of the radio core, as recently shown using MeerKAT observations (Sejake et al. 2023).
G4Jy 1664 is a radio galaxy at z=0. 15662 (a.k.a. 6dF J2056043-195635 Stickel & Kuehr 1994;Jones et al. 2009) with a classical FR II radio morphology, as shown in Figure 4.The redshift estimate was originally derived from 4 GHz and 3 GHz, respectively.The frequency of the radio map from which contours were drawn is reported together with the intensity of the first level.All radio contours increase in level by a binning factor as indicated.The radio morphology of G4Jy 1518 shows a classical double-double radio structure similar to "restarted" radio galaxies.
stellar absorption features, clearly present in its optical spectrum.The radio core emission appears to be similar to Giga-Peaked radio Sources (O'Dea et al. 1991).

X-ray counterparts of radio lobes in G4Jy 1863
An intriguing case is G4Jy 1863, a narrow-line hard X-ray selected giant radio galaxy (Cusumano et al. 2010;Oh et al. 2018;Bruni et al. 2020) at z=0.0959 (Tritton 1972;Danziger & Goss 1983;Ramos Almeida et al. 2011) lying in the direction of the galaxy cluster Abell 4067 (Abell 1958;Abell et al. 1989;Teague et al. 1990) and shown in Figure 5. G4Jy 1863 shows several signatures of a past merger in its optical image, for instance it has several irregular shells and two faint arcs (Ramos Almeida et al. 2011).
In Figure 5 the red circles mark the positions of detected X-ray sources close to the radio structure of G4Jy 1863, and in particular those labelled as S1 and S2.These two point-like sources have a clear optical and mid-infrared counterpart, as shown in the central and in the right panel of the same figure.The proximity of these two X-ray sources prevented us from claiming the presence of extended X-ray emission according to the procedure previously described.
However, considering the number of photons associated with both the northern and the southern radio lobes, in circular regions of 20 pixel radius, in comparison with those expected in the background for the same area, we claim a detection above 3σ level of confidence for the X-ray counterparts of its radio lobes.This could be due to inverse Compton scattering off seed photons arising from the Cosmic Microwave Background (see e.g., Hoyle 1965;Longair 1970;Harris & Grindlay 1979;Hardcastle et al. 2002;Harris & Krawczynski 2006;Croston et al. 2005;Worrall 2009).

COMPARISON WITH PREVIOUS X-RAY ANALYSES
Results obtained with our detection analysis are in agreement with those achievable when crossmatching radio positions, as well as mid-IR and optical ones (when available) with X-ray sources listed in the clean sample of the Second Swift-XRT Point Source Catalog13 (2SXPS Evans et al. 2020).The only differences we found are summarized below: • For four sources, namely: G4Jy 27, G4Jy 446, G4Jy 854 and G4Jy 1757, all marked with XDF=u, there is a counterpart listed in the 2SXPS catalog found within the X-ray positional uncertainty at 90% level of confidence and having the X-ray position consistent with their radio core.We also reported their X-ray detection based on the run carried out with the sosta algorithm having relatively low values of SN R, below the threshold adopted for the det algorithm.This difference is based on the fact that the 2SXPS catalog used a lower threshold on the SN R to claim a detection.Moreover the 2SXPS catalog is built reducing Swift-XRT observations with nominal T exp lower than the threshold of 250 sec adopted here.
• Two more sources, namely G4Jy 45 and G4Jy 1038 (a.k.a.3C 279) have an X-ray counterpart listed in the 2SXPS catalog, that we did not analyze here.G4Jy 45 has XRT observations beyond the threshold of 10 ′ separation set to select sources and thus was automatically excluded when retrieving datasets from the Swift archive.G4Jy 1038, as previously stated, was not selected as it is extensively discussed in the literature.
Figure 5.The X-ray image (left panel), smoothed with a Gaussian kernel of 42.426 ′′ , for the radio source G4Jy 1863.Radio contours are drawn using a radio map at 843 MHz starting at level of 90 mJy/beam and increasing by a factor of 3. Black circles, of 20 pixel radius, mark the location of point-like sources detected at SN R < 3 while blue arrows mark the location of both the northern (N) and the southern (S) radio lobe.The X-ray counterpart of the radio core (S1) and that of the nearby companion galaxy (S2) have a clear counterpart in the optical (r-band) image retrieved from DES r band (central panel) as well as at mid-infrared frequencies at 3.4µm obtained with the WISE All-sky survey (right panel).X-ray intensity contours, obtained from the Swift-XRT event files and drawn starting at a level of 0.01 photons, are overlaid to both the optical and the mid-infrared image to help identifying the location of the low energy counterparts for S1 and S2.The X-ray emission of both the northern and the southern radio lobes is detected, above 3σ level of confidence, comparing the number of photons measured within a circular region, of 20 pixel radius, centered on their radio intensity peak and that expected, on average, on a background region of the same area.
Finally, the only difference between our analysis and that of the 2SXPS catalog and/or the literature (i.e., Maselli et al. 2022) is for G4Jy 20 that is marginally detected in both analyses but undetected in our investigation.The main reason is due to the threshold on T exp chosen to retrieve Swift-XRT archival observations.Including all datasets with nominal T exp above 50 sec we also found a marginal detection, adopting the sosta algorithm, with SN R = 1.6 in agreement with previous analyses.

MID-INFRARED AND OPTICAL COUNTERPARTS IDENTIFIED WITH X-RAY OBSERVATIONS
We also compared the results of the X-ray analysis with those obtained from the search for optical counterparts, and all results of this comparison are summarized as follows.
• For the radio source G4Jy 672, with an assigned mid-IR counterpart reported in paper I, the X-ray counterpart lies within an angular separation of ∼2 ′′ from the optical host galaxy, thus confirming its previous association.
• There are two more sources, namely, G4Jy 249 and G4Jy 1432 (see Figure 6), for which available radio maps lack the angular resolution needed to precisely locate their radio cores and consequently their host galaxies in the optical band.However, using the Swift-XRT images it has been possible to determine the position of their optical counterparts using their X-ray emission.We are also able to confirm that X-ray detected sources is spatially coincident with their mid-IR counterparts associated in the G4Jy catalog.
• Despite the lack of a potential optical counterpart, the X-ray emission of G4Jy 1136 is spatially associated with its mid-IR counterpart (see also Figure 7) as previously stated in the G4Jy catalog (White et al. 2020a,b).
On the other hand, the comparison with mid-IR, optical, and X-ray data carried out for G4Jy 1192, revealed the location of its optical host galaxy, as shown in the right panel of Figure 8, in agreement with the mid-IR association described in paper I.

SUMMARY, CONCLUSIONS AND FUTURE PERSPECTIVES
In this second paper of the series, we present a first overview of the X-ray archival observations for G4Jy-3CRE radio sources, mainly focusing on those observed with the XRT instrument on board the Neil Gherels Swift Observatory.
The G4Jy-3CRE sample was recently extracted from the G4Jy catalog (White et al. 2020a,b) to obtain a list of powerful radio sources, selected at low radio frequencies, equivalent, in terms of flux density, to the 3C catalog but having all sources located in the Southern Hemisphere (paper I).The main advantage underlying the selection of this sample is the opportunity to study powerful radio sources and emission processes occurring in their large scale environments with a catalog observable with modern Southern Hemisphere telescopes and instruments such as MUSE and ERIS at VLT and/or ALMA and, in the near future, also SKA, LSST and ELT.
We retrieved, reduced and analyzed all archival Swift-XRT datasets collected between May 2005 and November 2022, having nominal T exp larger than 250 sec for a total of 615 observations processed for 89 sources.We found 28 radio sources with no X-ray counterpart detected above the chosen threshold of SN R = 3.We found 11 radio sources with extended/diffuse X-ray emission, estimated according to the criterion based on the Swift-XRT PSF model.In particular, G4Jy 1518 and G4Jy 1664 show clear extended X-ray emission, previously unknown, suggesting that they could be hosted in galaxy clusters, while for G4Jy 1863 we claim the X-ray counterpart detection of its radio lobes.All remaining 40 radio sources have a detected X-ray point-like counterpart of their radio core.The main reason underlying the non-detection of X-ray counterparts for all radio sources analyzed here is due to their relatively small integrated T exp .
Swift X-ray observations can provide first insights into the presence of diffuse/extended X-ray emission, used as the basis for future studies, and also to (i) confirm previous mid-IR and optical associations and to (ii) locate precisely the host galaxy position for those sources for which the angular resolution of the radio maps was not sufficient to reveal the position of the radio core.
We also obtained additional results that refine our previous radio-to-optical comparison (paper I): • For G4Jy 672, having no assigned mid-IR counterpart in the original G4Jy catalog, the X-ray counterpart lies within an angular separation of ∼2 ′′ from the optical host galaxy, thus confirming its optical association since it is smaller than the typcial Swift-XRT positional uncertainty for that T exp (see e.g.Evans et al. 2020).There is further confirmation of the host-galaxy position via detection of the radio core in a recent MeerKAT image (Sejake et al. 2023).
• Our previous radio-to-optical comparison did not allow us to firmly establish the host galaxies of G4Jy 249 and G4Jy 1432 due to the relatively poor angular resolution of the radio maps used to locate their radio cores.However, using merged Swift-XRT images we determined precisely the location of their optical counterpart from their X-ray emission.
• We also confirmed that the X-ray emission of G4Jy 1136 is spatially associated with its assigned mid-IR counterpart as listed in the original G4Jy catalog.
• The mid-IR, optical and X-ray comparison, carried out for G4Jy 1192, also revealed the location of its optical host galaxy.
Swift-XRT archival datasets, even if not homogeneous in terms of integrated T exp that span a range between 5 to 96 ksec, proved to be a powerful tool to refine the optical search for host galaxies of powerful radio sources and discover the presence of extended X-ray emission.In the case of G4Jy 1518 and G4Jy 1664, having snapshot XRT observations of ∼4 and ∼ 2 ksec integrated T exp , it has been possible to reveal the presence of the surrounding ICM.
Swift-XRT observations are certainly paving the path to identify those radio sources deserving X-ray follow up observations to perform detailed X-ray spectral analyses.This is crucial to complete optical spectroscopic observations The blue circle in the right panel is centered on the location of the X-ray counterpart to mark its position and that of the corresponding host galaxy.Radio contours are overlaid to both images drawn at the frequency indicated (i.e., 843 MHz obtained from the SUMSS archive) together with the intensity of the first contour level and increasing in level by a binning factor of 2. (Lower panels) same as upper panels for the radio source G4Jy 1432.Thanks to this optical-to-X-ray comparison we have been able to distinguish which source is the optical counterpart for these two radio sources having radio maps with insufficient angular resolution to locate precisely the radio core.Cyan cross marks the location of the radio source (i.e., brightness-weighted radio centroid as reported in the G4Jy catalog) while the red cross marks the mid-IR assigned counterpart, associated in the analysis of the G4Jy sample.2, smoothed with a Gaussian kernel of 10 pixels (i.e., 23.57′′ ) and with radio contours drawn from the TGSS radio map overlaid.These were computed starting at level of 0.1 Jy/beam and increasing in level by a binning factor of 4. Right panel) The same field but observed as part of the WISE All-sky survey at 12µm.The magenta dashed circle in the right panel is centered on the location of the X-ray counterpart to mark its position and that the mid-IR associated source (position indicated by the red cross in the right panel).According to our previous analysis (paper I) this radio source lacks an optical counterpart but the Swift-XRT analysis revealed that the mid-IR counterpart assigned in the G4Jy catalog is correct as it overlaps with its X-ray counterpart.It is worth highlighting the presence of a relatively bright star located in the north-western direction with respect to the position of G4Jy 1136.However it appears fainter in the 12µm WISE image, shown here, rather than in that at 3.4µm used in paper I.The cyan cross, if present, marks the location of the radio source (i.e., brightness-weighted radio centroid as reported in the G4Jy catalog) while the red one that of the assigned mid-IR counterpart associated in the analysis of the G4Jy sample.
necessary to (i) obtain the z measurements for the whole G4Jy-3CRE catalog and (ii) classify all radio sources listed therein from an optical perspective.
Additional X-ray datasets have been already requested and approved for the XMM-Newton satellite, as "filler programs" and as snapshot observations, to achieve a more complete high energy overview of the G4Jy-3CRE catalog while follow up X-ray observations will be also requested in the near future to deeply investigate the large-scale environment of those radio sources harbored in the galaxy clusters/groups.The X-ray image of G4Jy 1192 obtained from the merged XRT event file smoothed with a Gaussian kernel of 10 pixels (i.e., 23.57′′ ), as in Figure 2 and in Figure 7 with radio contours drawn from the TGSS radio map overlaid.These contours were computed starting at level of 0.1 Jy/beam and increase in level by a binning factor of 4. Middle and Right panels) The same field but observed as part of the WISE All-sky survey at 3.4µm and in the red filter collected from the DSS archive, respectively.The magenta dashed circle in the left panel is centered on the location of the X-ray counterpart to mark its position, the same is marked by the blue dashed circles in the other two panels, while the cyan cross marks the location of the radio source (i.e., brightness-weighted radio centroid) as reported in the G4Jy catalog.In our previous analysis (paper I) G4Jy 1192 lacks both mid-IR and optical counterparts.However, thanks to the Swift-XRT analysis, we have been able to identify these counterparts.X-ray contours obtained from the image in the left panel are overlaid to both mid-IR and optical images for comparison.G4Jy 1635, G4Jy 1640, G4Jy 1664, G4Jy 1671, G4Jy 1708, G4Jy 1717, G4Jy 1723, G4Jy 1748 & 1749

Figure 2 .
Figure2.Left panel) The X-ray image, obtained from the merged event file of G4Jy 85, smoothed with a Gaussian kernel of 10 pixels (i.e., 23.57′′ ).The magenta circle is centered on the position of the X-ray counterpart associated with this radio sources and it has a radius of 120.207 ′′ (i.e., 51 pixels) enclosing 90% of the Swift-XRT PSF.Radio contours are drawn using the radio map at 150 MHz starting at level of 0.3 Jy/beam and increasing by a factor of 3.This is a clear example of a radio source for which the X-ray counterpart of its radio core is detected at the level of the SN R reported in the figure.Central panel) same image as the left panel but for G4Jy 540, an example of detected extended X-ray emission.Here radio contours were computed using the archival radio map at 150 MHz starting at level of 0.1 Jy/beam and increasing by a factor of 6.Right panel) same image as the left panel but for G4Jy 718, an example of a radio source lacking an X-ray counterpart.In this case the magenta circle is centered on the radio position reported in the G4Jy catalog.Radio contours overlaid to the X-ray image are drawn from the radio map at 843 MHz starting at level of 0.2 Jy/beam and increasing by a factor of 2.

Figure 3 .
Figure3.Radio maps for G4Jy 1518 retrieved from the TGSS, NVSS and VLASS archives (left to right panels), at 150 MHz, 1.4 GHz and 3 GHz, respectively.The frequency of the radio map from which contours were drawn is reported together with the intensity of the first level.All radio contours increase in level by a binning factor as indicated.The radio morphology of G4Jy 1518 shows a classical double-double radio structure similar to "restarted" radio galaxies.

Figure 4 .
Figure 4. Same as Figure 3 for G4Jy 1664.The radio morphology of G4Jy 1664 is typical of FR II radio sources at 3 GHz while it appears a bit extended at lower frequencies where the double radio structure does not appear.

Figure 6 .
Figure6.(Upper panels) the left image is obtained from the merged XRT event files, smoothed with a Gaussian kernel of 10 pixels (i.e., 23.57′′ ), as in Figure2while the right one is the optical image retrieved from the DSS archive in the red filter.The blue circle in the right panel is centered on the location of the X-ray counterpart to mark its position and that of the corresponding host galaxy.Radio contours are overlaid to both images drawn at the frequency indicated (i.e., 843 MHz obtained from the SUMSS archive) together with the intensity of the first contour level and increasing in level by a binning factor of 2. (Lower panels) same as upper panels for the radio source G4Jy 1432.Thanks to this optical-to-X-ray comparison we have been able to distinguish which source is the optical counterpart for these two radio sources having radio maps with insufficient angular resolution to locate precisely the radio core.Cyan cross marks the location of the radio source (i.e., brightness-weighted radio centroid as reported in the G4Jy catalog) while the red cross marks the mid-IR assigned counterpart, associated in the analysis of the G4Jy sample.

Figure 7 .
Figure7.Left panel) The X-ray image of G4Jy 1136 obtained from the merged XRT event file as in Figure2, smoothed with a Gaussian kernel of 10 pixels (i.e., 23.57′′ ) and with radio contours drawn from the TGSS radio map overlaid.These were computed starting at level of 0.1 Jy/beam and increasing in level by a binning factor of 4. Right panel) The same field but observed as part of the WISE All-sky survey at 12µm.The magenta dashed circle in the right panel is centered on the location of the X-ray counterpart to mark its position and that the mid-IR associated source (position indicated by the red cross in the right panel).According to our previous analysis (paper I) this radio source lacks an optical counterpart but the Swift-XRT analysis revealed that the mid-IR counterpart assigned in the G4Jy catalog is correct as it overlaps with its X-ray counterpart.It is worth highlighting the presence of a relatively bright star located in the north-western direction with respect to the position of G4Jy 1136.However it appears fainter in the 12µm WISE image, shown here, rather than in that at 3.4µm used in paper I.The cyan cross, if present, marks the location of the radio source (i.e., brightness-weighted radio centroid as reported in the G4Jy catalog) while the red one that of the assigned mid-IR counterpart associated in the analysis of the G4Jy sample.

Figure 8 .
Figure8.Left panel) The X-ray image of G4Jy 1192 obtained from the merged XRT event file smoothed with a Gaussian kernel of 10 pixels (i.e., 23.57′′ ), as in Figure2and in Figure7with radio contours drawn from the TGSS radio map overlaid.These contours were computed starting at level of 0.1 Jy/beam and increase in level by a binning factor of 4. Middle and Right panels) The same field but observed as part of the WISE All-sky survey at 3.4µm and in the red filter collected from the DSS archive, respectively.The magenta dashed circle in the left panel is centered on the location of the X-ray counterpart to mark its position, the same is marked by the blue dashed circles in the other two panels, while the cyan cross marks the location of the radio source (i.e., brightness-weighted radio centroid) as reported in the G4Jy catalog.In our previous analysis (paper I) G4Jy 1192 lacks both mid-IR and optical counterparts.However, thanks to the Swift-XRT analysis, we have been able to identify these counterparts.X-ray contours obtained from the image in the left panel are overlaid to both mid-IR and optical images for comparison.

Table 1 .
Results for all G4Jy-3CRE sources with archival Swift-XRT observations

Table 2 .
Parameters for radio contours overlaid on X-ray images.
Table 2 continued on next page Table 2 (continued) Table 2 continued on next pageTable 2 (continued) Col. (1) the G4Jy name of the radio source, also adopted in the G4Jy-3CRE catalog; col.(2) frequency of the radio map used to draw contours in MHz; col.(3) minimum level of the radio contours lmin; col.(4) binning factor for the radio contours.