{"id":"1cb4a36d-eb9a-4878-891f-4929025742bb","arxiv_id":"2507.14094","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"After placing local and distant galaxy abundances on a common scale, the authors find Milky Way high-alpha disc patterns at z~2-3, evidence for alpha-bimodality in M31, and place the MW and GSE on the z~3 mass-metallicity relation.","lead":"This paper puts oxygen and iron measurements from distant young galaxies and from old stars in the Milky Way and nearby galaxies onto one common scale, then compares them. It finds that Milky Way-like chemical patterns appear at Cosmic Noon and that the Andromeda galaxy likely has the same two-layer chemical structure as the Milky Way.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Fig. 7 MZR agreement is weakened because the Kurbatov masses are metallicity-truncated cumulative masses, so the plotted mass and [Fe/H] are not independent; reanalysis with independent z~3 masses is required.","rationale":"I share the reader's focus on the z~3 mass estimates but locate the problem one level deeper. In Sec. 6.3 the quoted masses are cumulative present-day masses of the surviving Aurora/GSE components below a metallicity cut, not total stellar masses of the MW/GSE at z~3, and the same GC-age metallicity estimates set both the plotted [Fe/H] and the integration limit used for the mass. Thus the two axes of each Fig. 7 point are not independent, and the comparison tests a cumulative metallicity distribution rather than the galaxy-scale MZR. This is more load-bearing than the absence of uncertainties alone: even with perfect Kurbatov masses, the plot would not establish the claimed consistency as formulated. The calibration dependence is real but the paper is transparent about it and it goes in both directions; the mass-definition issue is not addressed. I therefore keep the reader's CONDITIONAL verdict: the central claim needs reanalysis with independent, full stellar mass estimates at z~3, not just an error bar added to the Kurbatov masses. The paper still has value as a careful homogenization and comparison effort, but the headline MZR consistency claim should be treated as unverified until this reanalysis is done.","tokens_in":48691,"tokens_out":13909,"duration_ms":620343,"concrete_test":"Recompute the MW and GSE points in Fig. 7 using independent z~3 total stellar masses, e.g., integrate the SFHs of Sanders et al. (2021a) and Hasselquist et al. (2021) up to 11.7 Gyr, or use Kurbatov et al. masses integrated over all metallicities rather than truncated at [Fe/H]=-1 or -1.3; if either point moves off the NIRVANDELS MZR by more than its extrapolated 1-sigma scatter, the claimed consistency is an artifact of the metallicity-selected mass definition.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is in Sec. 6.3: the MW and GSE points in Fig. 7 are placed using masses from Kurbatov et al. (2024) that are not independent total stellar masses at z~3 but cumulative present-day stellar masses of the surviving components below a metallicity threshold — 4.52e8 M_sun for Aurora, 'integrating up until [Fe/H]=-1', and 1.05e8 M_sun for GSE, averaged over integrations to -1.3 and -1.0. The [Fe/H] coordinates of the same points (-1.04 and -1.20) are the GC-age estimates that motivated those integration limits. Hence the mass and metallicity axes of each point are derived from the same present-day metallicity distribution of the surviving component, not from an independent measurement of the total stellar mass at z~3. The NIRVANDELS MZR is a relation between total stellar mass and abundance; comparing it to a metallicity-selected sub-component mass conflates the cumulative metallicity distribution of one Galactic component with a galaxy-scale relation. The apparent 'striking consistency' can therefore arise partly by construction, independent of whether the Kurbatov masses are accurate. The absence of quoted mass uncertainties, noted in the text, compounds the problem: a factor 2-3 reduction in GSE mass, as the authors themselves allow via Sanders et al. (2021a), would move the point substantially, and no quantitative agreement metric is provided.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper assembles a homogenized comparison of oxygen and iron abundances between resolved stellar populations in the Milky Way and Local Group and unresolved star-forming galaxies at z~2-3. The authors place all measurements on the Kobayashi et al. (2020a) solar scale, apply shifts to GALAH and globular-cluster oxygen abundances based on Amarsi et al. (2019), add M31 planetary nebula values converted through GCE models, and compare the resulting [O/Fe] vs [Fe/H] trends. They report good agreement between z~2-3 galaxies and the MW high-alpha disc, an apparent alpha-bimodality in M31, and use globular-cluster ages to project the MW and GSE to z~3, claiming that their estimated masses and abundances are strikingly consistent with the NIRVANDELS mass-metallicity relation.","tokens_in":48991,"tokens_out":5592,"duration_ms":65831,"significance":"If the central result holds, the paper offers a valuable framework for connecting resolved Galactic archaeology with high-redshift galaxy surveys, and its explicit treatment of solar scales, NLTE corrections, dust depletion, and abundance-method systematics is a useful contribution. The compilation itself, including the table of Cosmic Noon abundances, is a resource for future work. The central qualitative claims are defensible in outline, but the most headline result--the z~3 MZR placement--rests on projected masses that are not independent of the metallicity coordinates, and the MW comparison scale is calibrated on a very small number of stars. The paper is transparent about several of these limitations, which is a strength, but the load-bearing claims require additional analysis before they can be considered established.","major_comments":[{"comment":"The 'striking consistency' claim is not yet supported because the two projected points are constructed rather than independently measured. The MW mass (4.52e8 M_sun) is obtained by integrating the Aurora density profile up to [Fe/H]=-1, and the GSE mass (1.05e8 M_sun) is an average of integrations to -1.3 and -1.0; these are precisely the metallicity limits motivated by the same GC-age estimates that set the [Fe/H] coordinates (-1.04 and -1.20) of the points. The NIRVANDELS MZR is a relation between total stellar mass and abundance, so comparing a metallicity-selected sub-component mass against it conflates the cumulative metallicity distribution of one Galactic component with a galaxy-scale relation. The absence of quoted mass uncertainties, which the text acknowledges, and the authors' own allowance that GSE may have been 2-3 times smaller at z~3 (via Sanders et al. 2021a) mean the plotted points could shift substantially, yet no quantitative agreement metric is provided. Please re-derive the comparison with independent total-mass estimates at z~3, or explicitly present the current points as upper limits with a sensitivity analysis.","section":"§6.3, Fig. 7"},{"comment":"The -0.25 dex offset applied to all GALAH [O/Fe] values is calibrated using only two stars in common with A19, as the text itself states ('Unfortunately, we cannot calibrate the offset directly'). The dwarf-only comparison reduces the offset to 0.15 dex, yet the full giant sample is shifted by 0.25 dex, so the choice of offset absorbs an unquantified systematic. Because this shift propagates directly into the MW component tracks used in the high-redshift comparison and into the z~3 [O/Fe] values, please provide an uncertainty on the shift, test the sensitivity of Figs 3, 5, 6, and 7 to shifting by 0.15 rather than 0.25 dex, and discuss the per-metallicity temperature de-trending slopes as additional free parameters.","section":"§3.5"},{"comment":"The M31 comparison is partly model-dependent: the binned planetary nebula [O/Fe] and [Fe/H] values are produced by the GCE models of Kobayashi et al. (2023), a paper with overlapping authorship, and those same model-transformed values are then used to support the existence of alpha-bimodality in M31. This does not make the MW/M31 agreement vacuous, but it weakens the claim of independent confirmation. Please either compare the PNe trends to empirical [O/Ar] ratios or to M31 RGB stellar abundances, or explicitly label the M31 points as GCE-transformed predictions and recast the 'support' claim as a consistency test of the model.","section":"§3.4 and §5.2"}],"minor_comments":[{"comment":"The caption states 'after applying a global shift in [O/Fe]K20' as part of the agreement assessment; this is a statement of the calibration choice, not an independent validation, and should be phrased as such.","section":"Fig. 3 caption"},{"comment":"There is a missing citation: the sentence beginning 'two disc-like components...' contains a stray 'f' before '(Hayden et al.'; please correct the typo and complete the citation.","section":"§1"},{"comment":"There are repeated words: 'we have have overestimated' and the later 'derive derive SFH'; both should be corrected.","section":"§6.3"},{"comment":"The caption contains 'calculated using the using the fiducial strong-line calibrated scheme'; please remove the duplicated phrase.","section":"Fig. 7 caption"},{"comment":"The cosmology section states the Hubble constant and matter density but does not explicitly state Omega_Lambda; since a flat model is assumed this is determined, but please state it for clarity.","section":"§6.1"}],"recommendation":"major_revision","confidential_remarks":"The paper has the right ambition and a useful compilation, but the central MZR claim is currently over-stated relative to the evidence. The authors should be asked to either obtain independent z~3 mass estimates for the MW and GSE or to reframe the result as a sensitivity test with explicit upper limits. The M31 comparison would also benefit from a clearer statement that the plotted values are GCE-transformed; the current text risks leaving readers with the impression of an entirely observational comparison."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the useful part: this is a careful compilation that puts a lot of diverse abundance data onto a common scale and spells out the systematics (solar scale, NLTE, dust depletion, temperature trends) in a way that future work can use. The MW component separation is standard but well executed, and the idea of using globular cluster ages to project the MW and GSE back to z~3 is genuinely new. The comparison between MW high/low-alpha discs and M31 old/young PNe is a nice cross-check, even if the broad claim of alpha-bimodality in M31 already exists in Kobayashi et al. (2023).\n\nThe soft spots are real. The -0.25 dex shift applied to all GALAH [O/Fe] values is calibrated with only two stars in common with Amarsi et al. (2019). The authors handle this honestly, but it means the zero point of every MW comparison has a single anchor. The M31 PNe points are GCE-transformed abundances from Kobayashi et al. (2023) (a co-author paper), so using them to 'support' alpha-bimodality is partly circular; the qualitative conclusion is probably right, but the evidence isn't independent.\n\nThe load-bearing problem is the z~3 MZR placement. The masses from Kurbatov et al. (2024) are not independent total stellar masses at z~3; they are cumulative masses of the surviving component below a metallicity threshold—Aurora integrated to [Fe/H]=-1, GSE averaged over thresholds bracketing the adopted [Fe/H]. The [Fe/H] coordinates of those points come from the same GC-age estimates that set the thresholds. So the mass and metallicity are derived from the same present-day metallicity distribution, and comparing them to a relation between total stellar mass and abundance can produce agreement by construction. The paper acknowledges the mass uncertainties are not quoted and that the GSE point is an upper limit, but this is still a central claim described as 'strikingly consistent.' I'd want a reanalysis with truly independent z~3 mass estimates before trusting it.\n\nNet: I'd send this to a referee. It's not a desk reject. The compilation and the systematics discussion are worth having, and the z~3 projection is an interesting hypothesis even if the current evidence is weaker than the headline suggests. The referee should focus on the Kurbatov masses and the M31 model dependence; both are fixable in a revision.","headline":"Useful homogenization exercise with a novel z~3 projection, but the MZR placement rests on non-independent masses and a thin calibration.","tokens_in":49629,"tokens_out":3076,"would_cite":true,"duration_ms":38323,"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":"Ancient Milky Way fits the z≈3 galaxy mass–metallicity relation","keywords":["stellar abundances","mass-metallicity relation","galactic chemical evolution","globular cluster ages","Gaia-Sausage-Enceladus","oxygen-to-iron ratio","Tinsley-Wallerstein diagram","Cosmic Noon"],"falsifier":"A direct test would be to recompute the GSE stellar mass at 11.7 Gyr from a galactic chemical evolution model with a full star-formation history; if that mass comes out more than a factor of two or three below $1.05\\times10^8\\,M_\\odot$, the GSE point moves off the low-mass extrapolation of the NIRVANDELS mass–metallicity relation and the paper's central comparison fails.","tokens_in":48464,"feed_emoji":"🌌","tokens_out":11346,"duration_ms":109032,"temperature":0.7,"pith_summary":"This paper seeks to establish that the chemical history of the Milky Way can be read on the same scale as galaxies seen ten billion years away. It puts oxygen and iron measurements from $z\\sim2$–$3$ star-forming galaxies, Milky Way field stars and globular clusters, Local Group dwarfs, and Andromeda planetary nebulae onto one abundance scale, and compares them in the $[\\mathrm{O/Fe}]$ versus $[\\mathrm{Fe/H}]$ plane. Using globular-cluster ages, the authors project the Milky Way and the Gaia–Sausage–Enceladus (GSE) progenitor back to $z\\sim3$ and find their inferred masses, oxygen abundances, and iron abundances land on the low-mass extension of the mass–metallicity relation of $z\\sim3$ star-forming galaxies. If that holds, the ancient Milky Way was an ordinary low-mass star-forming galaxy of its epoch, and Galactic archaeology and high-redshift surveys are describing the same population.","feed_headline":"Ancient Milky Way fits the z≈3 galaxy mass–metallicity relation","feed_subtitle":"Globular-cluster ages put the Milky Way and its ancient merger right on the oxygen and iron relations of young galaxies.","key_machinery":"The argument is carried by the Tinsley–Wallerstein diagram, the plane of $[\\mathrm{O/Fe}]$ versus $[\\mathrm{Fe/H}]$ that encodes the competition between core-collapse supernova enrichment (oxygen) and delayed Type Ia supernova enrichment (iron). To compare resolved stellar spectra with unresolved high-redshift galaxies, all abundances are placed on the K20 solar abundance scale, with the GALAH sample shifted by $-0.25$ dex in $[\\mathrm{O/Fe}]$ and globular clusters by $+0.15$ dex to match ground-truth abundances. The chrono-chemical projection uses isochronal ages for globular clusters to select clusters near the $z\\sim3$ look-back time; their mean $[\\mathrm{Fe/H}]$ and $[\\mathrm{O/Fe}]$ define the Milky Way and GSE positions, and adopted stellar masses place those positions on the mass–metallicity plane.","core_discovery":"The central claim is that the Milky Way at $z\\sim3$ was a modest, actively star-forming galaxy whose chemistry places it on the same mass–metallicity relation as the NIRVANDELS galaxies, and that the same is true of the GSE progenitor. Using isochronal ages of in-situ globular clusters (11–12 Gyr) and two GSE-tagged clusters (11.0 and 11.5 Gyr), the authors place the Milky Way at $([\\mathrm{Fe/H}], [\\mathrm{O/Fe}])=(-1.04\\pm0.33, 0.55\\pm0.18)$ and GSE at $(-1.20\\pm0.13, 0.45\\pm0.16)$ on the K20 abundance scale. With stellar masses for the inner halo (Aurora) and GSE taken from a recent density-profile study, these points fall on the low-mass extrapolation of both the oxygen and iron mass–metallicity relations measured by NIRVANDELS, and they do so independently of the strong-line calibration or stellar model used. The paper also reports that the high- and low-$\\alpha$ sequences of the Milky Way discs match old and young planetary nebulae in Andromeda, supporting an $\\alpha$-bimodality in Andromeda's inner disc.","pith_inferences":["A testable extension would be to apply the same globular-cluster projection to other Local Group galaxies with known cluster ages, converting each old cluster into a fossil redshift marker for its host galaxy.","If the Milky Way/GSE mass–metallicity match survives, it implies that strong-line calibrations used at cosmic noon can be cross-checked against resolved stellar abundances, since two independent methods converge on the same relation.","The size of the abundance-scale shifts needed to bring surveys together suggests that future large stellar surveys must publish NLTE oxygen zero-point anchors before Galactic and extragalactic samples can be combined at the 0.1 dex level."],"forward_implications":["If the Milky Way and GSE really sat on the $z\\sim3$ mass–metallicity relation, the Milky Way's ancestors belong to the same low-mass, actively star-forming population that JWST now resolves at $z>4$.","The match between the Milky Way high- and low-$\\alpha$ discs and the old and young Andromeda planetary nebulae implies that $\\alpha$-bimodality is not unique to the Milky Way and can be seen in integrated light.","The agreement of high-redshift galaxies with the Milky Way high-$\\alpha$ disc and inner halo in $[\\mathrm{O/Fe}]$ versus $[\\mathrm{Fe/H}]$ suggests oxygen-to-iron ratios act as a common enrichment clock across ten billion years.","Some NIRVANDELS galaxies fall below the Milky Way at $z\\sim3$ in $[\\mathrm{O/Fe}]$, indicating that their star formation histories included enough Type Ia enrichment to differ from a simple constant-star-formation model."],"supporting_citations":[{"why":"Provides the NIRVANDELS oxygen and iron mass–metallicity relations at $z\\sim3$ that the Milky Way and GSE positions are compared against.","marker":"Stanton et al. (2024)"},{"why":"Supplies the NIRVANDELS sample and the FUV stellar-metallicity fitting method used for iron abundances.","marker":"Cullen et al. (2021)"},{"why":"Provides the isochronal globular-cluster ages used to project the Milky Way and GSE to $z\\sim3$.","marker":"VandenBerg et al. (2013)"},{"why":"Supplies the stellar masses adopted for the Milky Way inner halo and GSE at $z\\sim3$.","marker":"Kurbatov et al. (2024)"},{"why":"Provides the globular-cluster [O/Fe] and [Fe/H] abundances used for the chrono-chemical projection.","marker":"Carretta et al. (2009)"},{"why":"Supplies the GALAH DR3 stellar abundances from which the Milky Way structural components are defined.","marker":"Buder et al. (2021)"},{"why":"Defines the ground-truth abundance scale used to apply global shifts to the GALAH and globular-cluster oxygen values.","marker":"Amarsi et al. (2019)"},{"why":"Assigns the globular clusters used to locate the GSE progenitor in the projection.","marker":"Myeong et al. (2019)"},{"why":"Defines the K20 solar abundance scale and the chemical evolution models used for interpretation.","marker":"Kobayashi et al. (2020a)"}],"fun_headline_variants":["Milky Way fits z≈3 galaxy mass–metallicity relation","Ancient Milky Way on Cosmic Noon's chemical scaling","GSE and Milky Way match z=3 abundance relations","Milky Way's past chemistry mirrors distant galaxies","Andromeda alpha-bimodality echoes Milky Way discs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the adopted $z\\sim3$ stellar masses for the Milky Way's inner halo ($4.52\\times10^8\\,M_\\odot$) and the GSE progenitor ($1.05\\times10^8\\,M_\\odot$) are approximately right, even though they come with no quoted uncertainties and the GSE mass could be two to three times too high if GSE was still growing at 11.25 Gyr.","fun_headline_variants_meta":{"raw":{"variants":["Milky Way fits z≈3 galaxy mass–metallicity relation","Ancient Milky Way on Cosmic Noon's chemical scaling","GSE and Milky Way match z=3 abundance relations","Milky Way's past chemistry mirrors distant galaxies","Andromeda alpha-bimodality echoes Milky Way discs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000285,"raw_usage":{"total_tokens":1787,"prompt_tokens":1162,"completion_tokens":625,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":778,"completion_tokens_details":{"reasoning_tokens":543}},"tokens_in":778,"tokens_out":625,"duration_ms":7408,"temperature":1.0,"reasoning_tokens":543,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T16:01:01.001606+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct test would be to recompute the GSE stellar mass at 11.7 Gyr from a galactic chemical evolution model with a full star-formation history; if that mass comes out more than a factor of two or three below $1.05\\times10^8\\,M_\\odot$, the GSE point moves off the low-mass extrapolation of the NIRVANDELS mass–metallicity relation and the paper's central comparison fails.","supporting_citations":[],"review_version":1}