{"id":"302cd236-acf3-4e2f-b65b-8f57bfa8f35f","arxiv_id":"1908.08761","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A new sample of 249 southern radio galaxies with photometric redshifts shows powerful sources in massive hosts (10^11 to 10^12 solar masses), with low-mass blue galaxies likely being false identifications.","lead":"Astronomers built a catalogue of 249 bright southern radio galaxies by matching radio, infrared, and optical sky surveys, and measured the masses of their host galaxies. The most powerful sources sit in extremely massive galaxies, and the sample is designed for future hydrogen gas studies with the MeerKAT telescope.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central mass claim rests entirely on unverified DES photometric redshifts; a systematic z_phot error of ~0.1 would shift the quoted 10^11–10^12 M_sun range by ~0.5 dex.","rationale":"The reader's weakest assumption correctly identifies the lack of spectroscopic redshifts for the radio galaxy hosts as the most load-bearing issue. The central mass-range claim is produced by SED fitting at fixed BPZ photometric redshifts, and the paper itself flags both the absence of spec-z and the possibility that some photo-z are significantly wrong. Because a modest systematic photo-z error shifts stellar masses by several tenths of a dex, the same order as the width of the claimed 10^11–10^12 M_sun range, the headline statement is conditional on redshift accuracy. The paper does have independent support: the likelihood-ratio analysis is standard, the FAST stellar masses are robust to SPS/SFH/IMF choices at the 0.2 dex level, and the K-z consistency check is a useful sanity check. None of this, however, replaces direct redshift verification for this specific class of objects. Since the reader already assigned a conditional verdict with this limitation as the basis, no change to the verdict is needed; the concrete spectroscopic test would determine whether the condition is actually met.","tokens_in":13160,"tokens_out":9563,"duration_ms":112803,"concrete_test":"Obtain optical spectra for a representative subsample of roughly 30 of the 249 sources spanning the full z_phot and Ks range, or cross-match to existing OzDES, 2dF, or 6dF spectroscopy in the SPT-E field. Compare z_phot to z_spec and rerun FAST on the same grizJHK photometry with z_spec fixed for the matched sources. If the normalized median absolute deviation of Delta z / (1+z) exceeds 0.1, or if the median log M* of the high-z subsample shifts by more than 0.3 dex, the 10^11–10^12 M_sun claim is not yet supported by the data.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline result that the bulk of high-redshift hosts have M* in 10^11–10^12 M_sun is derived by running FAST at the DES SVA1 GOLD BPZ photometric redshifts, with no spectroscopic redshifts for any of the 249 sources. Section 4 explicitly states 'Spectroscopic redshifts are not available for our radio galaxy sample,' and Section 8 concedes 'their photometric redshifts could be significantly wrong.' The DES photo-z are calibrated on the general galaxy population, not on red, massive, possibly AGN-contaminated radio hosts. FAST treats z_phot as fixed, so any systematic photo-z error propagates directly into L_843, absolute magnitudes, and M*. At z ~ 1, a Delta z of about 0.1 changes the luminosity distance by roughly 10–20 per cent and the stellar mass by roughly 0.4–0.5 dex, comparable to the width of the claimed 10^11–10^12 M_sun range. Moreover, the 249 sources are the subset of 1,195 LR counterparts that survive a 'good photometric information' cut, so if BPZ preferentially fails or biases faint or red sources, the sample itself could be selected in a way that artificially produces the reported mass range. The K-z comparison in Fig. 5 is a consistency check, but it uses the same photo-z on the x-axis and a Ks-selected sample, so it cannot independently validate the redshifts. The central claim is therefore conditional on an unverified and potentially biasing measurement step.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper cross-matches the SUMSS 843 MHz radio catalogue with VHS near-infrared and DES optical data over 148 deg^2, using the likelihood-ratio technique to identify 1,195 reliable counterparts and then requiring 'good photometric information' from DES SVA1 GOLD BPZ photometric redshifts to define a final sample of 249 radio galaxies. The authors fit stellar population synthesis models with FAST to derive host stellar masses and star-formation rates, reporting that the high-redshift radio galaxy hosts typically have stellar masses between 10^11 and 10^12 solar masses, consistent with samples like 3CRR, 6C, and 7CRS. They also identify a population of low-mass blue galaxies, which they attribute to false positives in the cross-identification, and they present the sample as a target list for future HI absorption studies with MeerKAT.","tokens_in":13488,"tokens_out":3983,"duration_ms":43930,"significance":"If the reported mass range is correct, the paper extends host-galaxy mass measurements to a fainter radio-selected population at z ~ 0.5-1 in the southern sky, supporting the view that powerful radio AGN reside in extremely massive galaxies and providing a useful catalogue for upcoming HI absorption campaigns. The likelihood-ratio analysis is careful and includes a quantitative contamination estimate (6.8 per cent), and the comparison to the K-z relation is a useful sanity check. The photometric-redshift dependence, however, makes the central result conditional: the masses and redshifts come from the same unverified measurement chain, so the headline claim is not yet robust. The catalogue itself, with reliability estimates, is a valuable community resource.","major_comments":[{"comment":"The central claim that the bulk of high-redshift hosts have stellar masses in the range 10^11-10^12 Msun is derived by fixing the DES SVA1 GOLD BPZ photometric redshifts in FAST, and Section 4 states that no spectroscopic redshifts are available for the sample. As the authors concede in Section 8, 'their photometric redshifts could be significantly wrong.' A systematic photo-z error of ~0.1 at z ~ 1 changes the luminosity distance by 10-20 per cent and the derived stellar mass by roughly 0.4-0.5 dex, which is comparable to the width of the quoted mass range. The K-z comparison in Fig. 5 is not an independent validation because it uses the same photo-z on the x-axis. The paper needs either a spectroscopic subsample, an explicit propagation of the BPZ redshift PDFs through the FAST fitting, or a quantitative estimate of the allowed photo-z bias, before the mass range can be regarded as established.","section":"Section 4 and Section 6 (Figs 5 and 6)"},{"comment":"The selection step that reduces 1,195 LR counterparts with Rel>0.8 to 249 sources with 'good photometric information' is not characterized. If the DES photo-z pipeline preferentially fails or biases faint, red, or bright sources, the final sample's mass distribution could be a selection artifact. The authors should compare the 249-source sample to the parent 1,195-source set in terms of radio flux, Ks magnitude, colour, reliability, and, where possible, the DES photometric redshift quality flags. Such a comparison is needed to support the implicit claim that the 249 sources are representative of the reliable radio galaxy population in the survey area.","section":"Section 4 (sample selection from 1,195 to 249)"},{"comment":"The SED-derived star-formation rates appear to hit a boundary in the fitting grid: Table 1 lists log10(SFR) values of -2.06 and -2.15 for galaxies with log10(M*) of ~11.7, which are effectively zero and unphysical. This suggests that the FAST grid does not adequately sample low SFRs for passive galaxies. The paper states that varying the SPS model, SFH, and IMF gives a scatter of less than 0.2 dex, but does not show the basis for this claim or clarify whether it applies to SFR as well as mass. Since Section 7 uses the SFR comparison to argue that the low-mass blue galaxies are false positives, the reliability of the SFR estimates is load-bearing. The authors should display the SFR uncertainties and the variation across model choices, or temper the conclusion to acknowledge that the SFR measurements are not sufficiently reliable to distinguish false positives from genuine low-mass hosts.","section":"Section 6 and Table 1 (SFR estimates)"}],"minor_comments":[{"comment":"The word 'mid-identifications' in the abstract should be 'misidentifications'.","section":"Abstract"},{"comment":"The sentence containing 'This is in line with expectationsl' contains a typo: 'expectationsl' should be 'expectations'.","section":"Section 3, paragraph on Q0"},{"comment":"Several entries are missing J- and H-band magnitudes (e.g., IDs 4 and 6); the table notes or caption should state whether these are non-detections, upper limits, or data gaps in the VHS coverage.","section":"Table 1"},{"comment":"The statement 'we find no evidence for this in our sample' regarding the radio luminosity-host mass relation could be strengthened by providing a quantitative test, e.g., a Spearman rank test between L_843 and M* within the flux-limited sample, since the absence of evidence is otherwise difficult to interpret.","section":"Section 6, final paragraph"},{"comment":"The caption should specify which Delhaize et al. (2017) and Novak et al. (2017) relation is plotted (e.g., the SFR-L_1.4GHz relation for star-forming galaxies) and the assumed band and spectral index used to convert to 843 MHz.","section":"Figure 8 caption"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of MNRAS and the catalogue is likely to be used by the HI absorption community. The main risk is the unverified photometric redshifts: the authors themselves acknowledge the possibility of significant photo-z errors, yet the central mass range is derived directly from those redshifts. I would encourage the editor to request a concrete treatment of photo-z systematics, ideally with a small spectroscopic follow-up sample or a propagation of the full BPZ posteriors, and a characterization of the 1195-to-249 selection step. With those additions the paper would be a solid contribution; as it stands, the headline claim is somewhat ahead of the evidence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Darling,\n\nYou should know about this paper if you care about southern hemisphere radio galaxy samples for HI follow-up. It builds a new catalog of 249 powerful radio galaxies with photometric redshifts, and the central host-mass result (10^11-10^12 M_sun) is only as good as the DES BPZ photometric redshifts, which are unverified for this population. The catalog itself is genuinely useful; the physical claims are more fragile.\n\nWhat is new: the combination of SUMSS, VHS DR3, and DES SVA1 GOLD with a likelihood ratio match has not been published before. The authors do a careful job: reliability thresholds, contamination estimates (6.8%), a K-z relation consistency check, and FAST SED fits with BC03 and Maraston templates. They also honestly flag the low-mass blue galaxies as likely false positives, using the radio/SFR comparison to argue they are interlopers. For MeerKAT/ASKAP HI absorption target lists, this is exactly the kind of sample we want.\n\nWhere it wobbles: the photometric redshifts. Spectroscopic redshifts are not available for any of the 249 objects. The DES photo-z are calibrated on the general galaxy population, not on red, massive, AGN-contaminated radio hosts. FAST treats z_phot as fixed, so any systematic bias goes straight into rest-frame luminosities, stellar masses, and the K-z relation. A 0.1 error in z at z~1 shifts M* by ~0.4-0.5 dex, which is about the width of the mass range they quote. The paper's own Section 8 concedes their photometric redshifts could be significantly wrong. The drop from 1,195 LR counterparts to 249 sources with 'good photometric information' also needs a fuller characterization; if BPZ fails on faint or red sources, the surviving sample could be selected to produce the mass range they see. The K-z comparison is not independent, since the same photo-z are on the x-axis.\n\nThat said, none of this invalidates the catalog as a target list. The radio sources are real, and the cross-match reliability is quantified. The science about host galaxy masses and SFRs should be read as conditional on a photo-z validation that the paper does not provide. A referee could reasonably ask for a small spectroscopic campaign, or at least a robustness test with simulated photo-z offsets.\n\nMy take: send it to review. It deserves referee time, but a referee should push on the photo-z dependence and the selection function. I would cite this for the sample, not for the mass measurements.","headline":"A useful new southern radio galaxy catalog for HI follow-up whose headline host-mass range is conditional on unverified DES photometric redshifts.","tokens_in":14047,"tokens_out":3314,"would_cite":true,"duration_ms":29814,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Powerful radio galaxies in a new southern sample live in galaxies of 10^11 to 10^12 solar masses, matching the brighter classic samples.","keywords":["radio galaxies","host stellar masses","photometric redshifts","likelihood ratio matching","southern sky surveys","K-z relation","star-forming contaminants"],"falsifier":"Take a random subset of about 30 of the 249 hosts and measure their spectroscopic redshifts; if they disagree systematically with the DES photometric redshifts for the red, massive hosts, the rest-frame radio luminosities, stellar masses, and the K-z relation would shift, and the claimed consistency with the brighter 3CRR/6C/7CRS samples would be weakened.","tokens_in":13012,"feed_emoji":"📡","tokens_out":8418,"duration_ms":72453,"temperature":0.7,"pith_summary":"By matching the SUMSS radio catalogue at 843 MHz to near-infrared VHS galaxies and DES optical photometry, this paper constructs a new sample of 249 powerful southern radio sources over 148 square degrees. The authors fit spectral synthesis models to the optical and near-infrared light to derive host stellar masses and star-formation rates. They find that the high-redshift hosts cluster at stellar masses of $10^{11}$ to $10^{12}$ solar masses, the same range as the hosts of the much brighter 3CRR, 6C and 7CRS samples, and that the sample follows the established radio-galaxy K-z relation. A population of low-mass blue galaxies is interpreted as likely false identifications, because their radio luminosities exceed what their star-formation rates can explain. The resulting catalogue is intended as a southern-hemisphere target list for HI absorption studies with MeerKAT and similar facilities.","feed_headline":"Southern radio galaxy hosts weigh 10^11-10^12 Suns","feed_subtitle":"Cross-matching SUMSS, VHS and DES yields 249 powerful radio sources and a target list for MeerKAT hydrogen studies.","key_machinery":"The central machinery is the likelihood-ratio technique for identifying near-infrared counterparts to radio sources, applied in the Ks band using the VHS catalogue. It computes, for each candidate, the ratio of the probability of being a true counterpart to the probability of being a background galaxy, and keeps sources with reliability above 0.8. Photometric redshifts come from BPZ applied to DES SVA1 Gold photometry, and stellar masses and star-formation rates are derived by fitting Bruzual & Charlot (2003) and Maraston (2005) spectral synthesis templates with the FAST code. These quantities are then compared with the K-z relation and host masses of the 3CRR, 6C and 7CRS samples, and with the radio luminosity function of star-forming galaxies from Mauch & Sadler (2007) to test whether low-mass blue hosts are real or mismatches.","core_discovery":"The paper claims that powerful southern radio sources selected at 843 MHz, identified to VHS near-infrared galaxies with the likelihood-ratio technique and assigned DES photometric redshifts, reside in the same massive galaxy hosts as the historically complete northern samples. The bulk of the high-redshift hosts have stellar masses between $10^{11}$ and $10^{12}$ $M_\\odot$, corresponding to roughly 1 to 10 times the characteristic mass of the galaxy mass function. The sources follow the radio-galaxy K-z relation defined by 3CRR, 6C and 7CRS, and the authors find no evidence for a correlation between radio luminosity and host mass within their single flux-limited sample. The low-mass blue galaxies, which at face value look like star-forming radio hosts, are argued to be false positives because their radio powers are far too high for their SED-derived star-formation rates.","pith_inferences":["The radio-luminosity versus star-formation-rate comparison used here could serve as a generic screening tool for low-resolution radio surveys, flagging blue low-mass candidates as likely mismatches before costly spectroscopy.","If the photometric redshifts hold up, the same SUMSS-VHS-DES pipeline could be run on the deeper EMU survey when it arrives, extending the K-z relation and host-mass distribution to fainter sources without new spectroscopy.","The sample's effective redshift limit near z<1 is set by the 4000 Angstrom break entering the near-infrared; pushing to z>1 would require adding mid-infrared photometry to the SED fits."],"forward_implications":["The 249-source catalogue gives southern-hemisphere targets for HI absorption follow-up with MeerKAT and ASKAP.","Host masses this high at the 10 mJy flux level imply that the connection between powerful radio emission and very massive galaxies persists to lower radio luminosities than the classic bright samples probe.","The absence of a radio luminosity-host mass correlation within this sample, combined with wider and deeper samples, would allow a direct test of whether jet power depends on galaxy mass.","The expected false-positive fraction of about 6.8 percent bounds how much contamination remains in the final catalogue; higher-resolution radio imaging should shrink it."],"supporting_citations":[{"why":"Defines the likelihood-ratio technique used to match SUMSS radio sources to VHS near-infrared counterparts.","marker":"Sutherland & Saunders 1992"},{"why":"Supplies the method for estimating Q0, the fraction of radio sources with detectable counterparts, and the reliability calculation.","marker":"Fleuren et al. 2012"},{"why":"Provides the reliability threshold of 0.8 used to accept counterparts.","marker":"Smith et al. 2011"},{"why":"The BPZ code that produced the DES photometric redshifts used for the sample.","marker":"Benítez 2000"},{"why":"Characterizes the accuracy of the DES photometric redshifts on the general galaxy population.","marker":"Sánchez et al. 2014"},{"why":"The FAST code used to fit spectral synthesis models to the photometry.","marker":"Kriek et al. 2009"},{"why":"One of the two stellar population synthesis template sets used in the SED fitting.","marker":"Bruzual & Charlot 2003"},{"why":"The alternative template set used to check systematic effects on stellar masses.","marker":"Maraston 2005"},{"why":"Provides the radio luminosity function of star-forming galaxies used to interpret the low-mass blue sources.","marker":"Mauch & Sadler 2007"},{"why":"Gives the comparison host masses for powerful radio galaxies from Spitzer.","marker":"Seymour et al. 2007"}],"fun_headline_variants":["Southern radio galaxies: massive hosts, blue impostors flagged","249 southern radio hosts found; blue low-mass ones are false","Massive hosts confirmed for southern radio galaxies, blue ones rejected","Southern radio sample: 249 hosts, massive, blue are false positives"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central assumption is that the DES photometric redshifts, calibrated on ordinary galaxies, remain accurate for these red, massive, possibly active hosts, even though none of the 249 sources has a spectroscopic redshift to verify them.","fun_headline_variants_meta":{"raw":{"variants":["Southern radio galaxies: massive hosts, blue impostors flagged","249 southern radio hosts found; blue low-mass ones are false","Massive hosts confirmed for southern radio galaxies, blue ones rejected","Southern radio sample: 249 hosts, massive, blue are false positives"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000833,"raw_usage":{"total_tokens":3655,"prompt_tokens":982,"completion_tokens":2673,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":598,"completion_tokens_details":{"reasoning_tokens":2601}},"tokens_in":598,"tokens_out":2673,"duration_ms":18047,"temperature":1.0,"reasoning_tokens":2601,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:29:17.586312+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a random subset of about 30 of the 249 hosts and measure their spectroscopic redshifts; if they disagree systematically with the DES photometric redshifts for the red, massive hosts, the rest-frame radio luminosities, stellar masses, and the K-z relation would shift, and the claimed consistency with the brighter 3CRR/6C/7CRS samples would be weakened.","supporting_citations":[],"review_version":1}