{"id":"779cf501-e7b5-4800-9d12-126b4915249e","arxiv_id":"2506.02876","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A barred galaxy simulation reproduces the age (observed as metallicity) trends of the bulge velocity ellipses, supporting kinematic fractionation without an accreted classical bulge.","lead":"This paper compares the velocity ellipses of young and old stars in a simulated barred galaxy with those of metal-rich and metal-poor stars in the Milky Way bulge, using APOGEE DR16 and Gaia DR3 data. The authors find qualitative agreement and argue this supports a purely secular formation of the bulge via kinematic fractionation, with no need for a significant accreted classical bulge.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central inference depends on an untested metallicity-as-age proxy, and the simulation's own age-metallicity relation is weakened by missing metal diffusion; the [Fe/H]-split comparison may not test kinematic fractionation.","rationale":"The reader's weakest assumption bundles the isolated-evolution setup with the metallicity-as-age proxy. I agree that the proxy is a serious issue, and I think it is the single most load-bearing one because it is the logical bridge between the model's age-split predictions and the APOGEE [Fe/H]-split observations. The simulation's own age-metallicity relation is stated to be weakened (Section 3), so the authors cannot internally validate the mapping. This is not an ad hominem or a demand for a different model; it is a request for a robustness test that the manuscript itself suggests is missing. I also considered the absence of a quantified significance test for the APOGEE rho_rl-[Fe/H] trend; while real, it is secondary because even a significant trend would not establish kinematic fractionation unless the proxy holds. The isolated-simulation concern is also real but less directly tied to the internal logic, since the paper claims sufficiency ('without the need to invoke') rather than uniqueness. The concrete test I propose uses the existing simulation output and the same selection criteria, so it is feasible and would settle whether the age-metallicity mapping is doing essential work. If the metallicity-split model reproduces the APOGEE trend, the central claim is strengthened; if not, the comparison is not a valid test. Given this, the conditional verdict is appropriate, and I do not recommend changing it.","tokens_in":39408,"tokens_out":9047,"duration_ms":110949,"concrete_test":"Using the same simulation snapshot, assign each bulge star its tracked [Fe/H]; apply the APOGEE selection (|l|<11 deg, |b|<13 deg, R_GC<3.5 kpc, [Fe/H]>-1), split by the median [Fe/H] or the same [Fe/H] bins as Fig. 13, and recompute rho_rl vs [Fe/H] for 3.5<|b|<6.6 deg. Compare with both the age-split model and the APOGEE data. If the model's [Fe/H]-split rho_rl does not reproduce the APOGEE monotonic increase in |rho_rl|, then the observed trend is not explained by kinematic fractionation through the assumed age-metallicity mapping, and the central claim fails. If it does reproduce the trend, the proxy concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Section 9.3 iv, Fig. 13) equates the model's age trend in rho_rl with the observed [Fe/H] trend, which requires [Fe/H] to be a monotonic age proxy in the bulge. The paper states this assumption in Section 5 ('we use metallicity as a proxy for age, with the expectation that older stars are more metal-poor') but provides no evidence for it in the MW sample. More importantly, Section 3 explicitly states that the simulation lacks metal diffusion, producing 'an excess of low-metallicity stars forming at all ages, and a weakened relation between metallicity and age,' which is why the authors define populations by age instead of metallicity. Thus the model cannot validate the very mapping on which the comparison depends. If the MW's age-metallicity relation is broad or non-monotonic (e.g., due to rapid early enrichment or accreted populations), the observed monotonic rise in |rho_rl| with [Fe/H] could be produced by a mix of populations unrelated to the age and kinematic-fractionation sequence in the model. The paper does not quantify the significance of the [Fe/H] trend, nor does it test the proxy, so the claimed model-data agreement is not yet established as evidence for kinematic fractionation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a kinematic analysis of velocity ellipsoids in an N-body+SPH galaxy simulation from Debattista et al. (2017), comparing model predictions for two age-selected bulge populations (young, 4-7 Gyr; old, 9.5-10 Gyr) with APOGEE DR16+Gaia DR3 bulge stars split at the median [Fe/H]. The authors compute the anisotropy beta_ij, correlation rho_ij, and vertex deviation l_v in galactocentric and heliocentric frames, and compare vertical profiles along the bulge minor axis and full (l,b) maps. They find that rho_rl increases in amplitude with decreasing age in the model and with increasing [Fe/H] in the data, and interpret this as evidence that the Milky Way bar's amplitude varies continuously with [Fe/H] through kinematic fractionation, making a significant accreted classical bulge unnecessary. The paper also argues that vertex deviation is a blunt tracer of bar strength whereas the correlation is robust and promising.","tokens_in":39618,"tokens_out":5440,"duration_ms":64346,"significance":"If the central inference holds, the paper would strengthen the secular-origin scenario for the Milky Way bulge by connecting a specific observable (the v_r-v_l correlation) to a continuously varying bar strength across stellar populations. The paper's strengths include careful bootstrap error estimation with B=500, an explicit test of bootstrap validity in Appendix C showing that vertex deviation is biased for small samples while correlation is well-behaved, a Monte Carlo study of distance-error effects in Section 7, and a comparison with weaker-bar and oval models in Appendix B that supports the proposed interpretation of rho_rl as a bar-strength tracer. These methodological checks make the descriptive kinematic results trustworthy. However, the decisive model-data link depends on an untested metallicity-as-age proxy and on an unquantified trend, so the interpretative conclusion is not yet established.","major_comments":[{"comment":"The central claim that the APOGEE [Fe/H] trend in rho_rl is evidence for kinematic fractionation depends on [Fe/H] being a monotonic proxy for age, but the paper neither validates this proxy nor can its simulation validate it. Section 5 states only the expectation that older stars are more metal-poor, while Section 3 reports that the simulation lacks metal diffusion, producing an excess of low-metallicity stars forming at all ages and a weakened age-metallicity relation, which is why the model populations are defined by age rather than metallicity. A broad or non-monotonic bulge age-metallicity relation could produce the observed [Fe/H] trend from mixed populations unrelated to the kinematic-fractionation sequence. Please test the proxy directly (e.g., with known APOGEE ages or chemical clocks), and/or compare model metallicity-sorted populations (acknowledging the weakened relation) to see whether the predicted trend survives. If this cannot be done, the inference should be framed as a prediction conditional on a monotonic relation.","section":"Section 5; Section 9.3(iv), Fig. 13"},{"comment":"The paper describes the observed trend as a rise in |rho_rl| with [Fe/H] that \"appears to be\" present and may \"plateau or even decline\" within errors, but no statistical significance is quantified. With only three or four metallicity bins and bootstrap errors that appear comparable to the bin-to-bin variation, a constant rho_rl within the quoted uncertainties may be consistent with the data. Please provide a quantitative test of monotonicity (e.g., rank correlation with bootstrap significance, or a linear fit with uncertainty), and report the significance of the trend in the 3.5<|b|<6.6 degree bin where the claim is strongest.","section":"Section 9.1 and Fig. 13"}],"minor_comments":[{"comment":"Equation (5) appears to have a typographical issue in the summation notation, with an extra symbol inside the sum; please check and correct the displayed formula.","section":"Equation (5)"},{"comment":"The histogram in Fig. 13 is labeled only with 'N', and the inset values such as '[248 248 117]' are not defined in the caption; please state explicitly what these numbers represent.","section":"Fig. 13 caption"},{"comment":"The acronym 'VRT-LSST' is used without being defined; please expand it or provide a reference at first use.","section":"Section 9.2"},{"comment":"The central-oval model has not been presented before, but the appendix gives no description of how it was constructed or how its parameters compare to the fiducial model; a brief description would help the reader assess the comparison.","section":"Appendix B"},{"comment":"The statement that the APOGEE sample 'may be biased in distance for different metallicities' is noted but not investigated beyond a radial-cut test; please state explicitly whether the quoted rho_rl robustness checks included metallicity-dependent selection functions or whether this remains an open issue.","section":"Section 5.2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of the journal and the authors are transparent about reusing the same simulation as in their prior work. The descriptive kinematic results are valuable and well tested; the main revision should focus on validating or appropriately down-weighting the metallicity-as-age mapping and on quantifying the significance of the key [Fe/H] trend."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the short version: this is a solid, careful follow-up to Debattista et al. (2017), and the best thing in it is the case that rho_rl, the correlation between v_r and v_l, is a more useful tracer of bar strength than the vertex deviation. The maps and the bootstrap analysis in Appendix C are good work. But the paper's central inference—that the observed [Fe/H] trend in rho_rl mirrors the model's age trend and therefore supports kinematic fractionation—rests on a metallicity-as-age proxy that is asserted, not tested. That is a real soft spot, and it should keep the paper from claiming too much.\n\nWhat is actually new: the 2D maps of anisotropy, correlation, and vertex deviation across the (l,b) plane in both galactocentric and heliocentric frames, and the systematic comparison to APOGEE DR16+Gaia DR3. The demonstration that vertex deviation saturates near +/-45 degrees when the velocity ellipse is nearly isotropic, and that rho_rl responds monotonically to bar strength in the models (including the weak-bar and oval models in Appendix B), is a useful methodological contribution. They also test the bootstrap assumption in Appendix C, which is more than most papers do, and they explicitly quantify the effect of distance errors.\n\nThe soft spots: the metallicity-as-age proxy is the big one. The simulation has no metal diffusion, which weakens its own age-metallicity relation—the authors say this is why they define model populations by age. That means the model cannot validate the mapping on which the comparison depends. The paper does not test whether [Fe/H] is a monotonic age proxy in the MW bulge, and given the bulge's complex enrichment history, that assumption is not trivial. Also, the agreement between model and data is explicitly qualitative, never quantified, and some selection choices (e.g., [Fe/H] < -1 cut, median split, excluding young stars to avoid the nuclear disc) are post hoc rather than derived from the model. These are not fatal, and the authors are honest about most of them, but together they mean the evidence for 'continuously varying bar amplitude in the MW' is suggestive, not established.\n\nIf I were refereeing, I would ask for a quantitative comparison (or at least a stated metric) and some justification for the age-metallicity mapping, perhaps using a simulation with metal diffusion or an independent observational constraint. But this deserves peer review; it is a legitimate contribution to the bulge kinematics literature.","headline":"A careful, useful follow-up on bulge velocity ellipses with a genuinely handy new diagnostic (rho_rl), but the central claim rests on an untested metallicity-as-age proxy.","tokens_in":40250,"tokens_out":2480,"would_cite":true,"duration_ms":25701,"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":"A purely secularly evolved bar simulation reproduces the Milky Way bulge's velocity-ellipse trends, so a significant accreted classical bulge is not required.","keywords":["Galactic bulge","velocity ellipsoid","vertex deviation","kinematic fractionation","bar strength","APOGEE","Gaia DR3","N-body simulation"],"falsifier":"Measure $\\rho_{rl}$ along the minor axis for bulge stars split into narrow bins of individual asteroseismic age; if the correlation amplitude does not increase monotonically with decreasing age at $3.5^\\circ<|b|<6.6^\\circ$, the kinematic-fractionation explanation of the metallicity trend would be falsified.","tokens_in":39176,"feed_emoji":"🌌","tokens_out":14802,"duration_ms":124124,"temperature":0.7,"pith_summary":"This paper argues that the Milky Way bulge's velocity ellipses, quantified by the anisotropy $\\beta_{rl}$, correlation $\\rho_{rl}$, and vertex deviation $l_{\\rm v}$, are produced by a bar that forms and evolves purely secularly, with no significant accreted classical bulge. The central evidence is along the bulge minor axis, where the correlation between the heliocentric radial and longitudinal velocities becomes more negative with decreasing stellar age in the simulation and with increasing metallicity in APOGEE DR16 data matched to Gaia DR3 proper motions. That continuous trend is the signature of kinematic fractionation: a bar that separates cooler, younger, more metal-rich stars from hotter, older, more metal-poor ones by their initial velocity dispersions. If the argument holds, the bar's amplitude in the Milky Way varies continuously with population rather than being constant above a metallicity threshold, and the metal-poor bulge need not be an accreted, unbarred component.","feed_headline":"Bulge velocity-ellipse trends reproduced without an accreted bulge","feed_subtitle":"Correlation between radial and tangential velocities grows with metallicity, matching a purely secular bar simulation.","key_machinery":"The machinery is the velocity-dispersion tensor of a stellar population, compressed into three dimensionless quantities: in-plane anisotropy $\\beta_{ij} = 1-\\sigma_{jj}^2/\\sigma_{ii}^2$, correlation $\\rho_{ij} = \\sigma_{ij}^2/(\\sigma_i\\sigma_j)$, and vertex deviation $l_{\\rm v}$, with $\\tan(2l_{\\rm v}) = 2\\rho_{ij}\\sqrt{1-\\beta_{ij}}/|\\beta_{ij}|$ (Eqn. A10). The load-bearing mechanism is kinematic fractionation: the bar forms with a strength set by each age cohort's initial velocity dispersion, so younger, cooler stars become strongly barred and X-shaped while older, hotter stars remain weakly barred and boxy, and the same orbit families imprint age-dependent quadrupoles in $\\beta$, $\\rho$, and $l_{\\rm v}$ that project into observable minor-axis trends along $l=0^\\circ$.","core_discovery":"The paper's central claim is that $\\rho_{rl}$ between the heliocentric radial and longitudinal velocities is a clean, continuous tracer of bar strength, while the vertex deviation $l_{\\rm v}$ is a blunt one. In the isolated $N$-body$+$SPH simulation, young (4-7 Gyr) stars form a strong bar with a prominent X-shape and show much stronger negative $\\rho_{rl}$ than old (9.5-10 Gyr) boxy stars; APOGEE stars split at the median $[{\\rm Fe/H}] = -0.21$ dex follow the same separation. At fixed latitude, $\\rho_{rl}$ rises in amplitude with decreasing age in the model and with increasing $[{\\rm Fe/H}]$ in the data, which the paper reads as the first indication that the Milky Way bar's amplitude varies smoothly with metallicity, as kinematic fractionation predicts, rather than being constant above some metallicity. Along the way the paper shows that $l_{\\rm v}$ peaks near $-45^\\circ$ for both young and old populations whenever the velocity ellipse is nearly isotropic, so equal vertex-deviation peaks do not imply equal bar strengths, and that $\\rho_{rl}$ changes little under distance uncertainties up to 35%, making it the recommended statistic for future bulge surveys.","pith_inferences":["If the continuous $\\rho_{rl}$ trend survives larger samples, the Milky Way bulge would be a single age-stratified bar-disc system whose bimodal metallicity distribution can arise from kinematic fractionation plus a thick disc, pushing any classical bulge below roughly 2% of the stellar mass.","A sharper test would replace the $[{\\rm Fe/H}]$-age proxy with asteroseismic or Cepheid ages for bulge stars; the model predicts that $\\rho_{rl}$ amplitude orders populations by age alone, independent of metallicity.","The appendix result that bootstrap errors for $l_{\\rm v}$ are biased below a few hundred stars suggests that some small-sample historical vertex-deviation measurements in the bulge should be treated as upper limits on non-axisymmetry rather than detections of distinct components.","The same $\\rho_{rl}$ diagnostic could be applied to external barred galaxies with integral-field kinematics, where the model's $(l,b)$ maps give a concrete template for how projection mixes near- and far-side bar streaming."],"forward_implications":["Along the bulge minor axis, $l_{\\rm v}$ reaches nearly $-45^\\circ$ for both young and old populations at $|b|<6^\\circ$, so vertex deviation alone cannot be used to argue that metal-poor bulge stars belong to a separate, unbarred accreted component.","$\\rho_{rl}$ is the recommended bar-strength tracer: it grows monotonically with decreasing age and increasing $[{\\rm Fe/H}]$, is robust to radial cuts and to distance uncertainties as large as 35%, and remains unbiased at small sample sizes.","Distance errors up to about 20% for the young population leave the anisotropy, correlation, and vertex deviation essentially unchanged, so current APOGEE plus Gaia DR3 data are adequate for the qualitative comparison.","The latitude band $3^\\circ<|b|<6^\\circ$ along the minor axis is the most promising window for future surveys to test whether the Milky Way bar amplitude is a continuous function of stellar population.","The model predicts that bar signatures such as the X-shape, strong streaming motions, and forbidden velocities weaken smoothly with age, so bulge samples with reliable individual ages should show a continuum of bar strength rather than two discrete components."],"supporting_citations":[{"why":"Introduces the isolated N-body+SPH simulation and the kinematic fractionation mechanism that this paper uses to generate age-dependent bar strengths.","marker":"Debattista et al. (2017)"},{"why":"Provides the cleaned APOGEE bulge sample, metallicities, and spectro-photometric distances that define the observed comparison.","marker":"Rojas-Arriagada et al. (2020)"},{"why":"Supplies the DR3 proper motions used to compute the observed longitudinal velocities and velocity ellipses.","marker":"Gaia Collaboration et al. (2023)"},{"why":"Earlier study of the same simulation establishing the age-separated kinematics and forbidden-velocity signatures that this work extends.","marker":"Gough-Kelly et al. (2022)"},{"why":"Describes the simulation's initial conditions, star formation, bar formation, and the nuclear disc that motivates the age cuts.","marker":"Cole et al. (2014)"},{"why":"Provides the observed radial-velocity quadrupole in the Milky Way against which the model's young-population kinematic maps are checked.","marker":"Bovy et al. (2019)"},{"why":"Measured the vertex deviation of metal-rich and metal-poor bulge stars, the observational pattern this paper reinterprets via the correlation.","marker":"Soto et al. (2007)"},{"why":"Confirmed the metal-rich vertex deviation and metal-poor spheroidal kinematics that are compared with the model's young and old populations.","marker":"Babusiaux et al. (2010)"},{"why":"Supplies the 27-degree bar orientation angle used to project the model into the observed bulge coordinate frame.","marker":"Wegg & Gerhard (2013)"}],"fun_headline_variants":["Bar strength traced by velocity correlation, not vertex deviation","Milky Way bar strength scales with metallicity, model shows","Velocity correlation reveals bar strength, vertex deviation does not","Bar simulation matches bulge kinematics without accretion","Young stars show stronger bar correlation in bulge model"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The comparison rests on treating metallicity as a proxy for age in the observed bulge and on assuming an isolated, merger-free simulation captures the Milky Way's bulge assembly; if $[{\\rm Fe/H}]$ and age are scrambled in the real bulge, the agreement could be coincidental.","fun_headline_variants_meta":{"raw":{"variants":["Bar strength traced by velocity correlation, not vertex deviation","Milky Way bar strength scales with metallicity, model shows","Velocity correlation reveals bar strength, vertex deviation does not","Bar simulation matches bulge kinematics without accretion","Young stars show stronger bar correlation in bulge model"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000432,"raw_usage":{"total_tokens":2306,"prompt_tokens":1147,"completion_tokens":1159,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":763,"completion_tokens_details":{"reasoning_tokens":1085}},"tokens_in":763,"tokens_out":1159,"duration_ms":8308,"temperature":1.0,"reasoning_tokens":1085,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:13:14.406045+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure $\\rho_{rl}$ along the minor axis for bulge stars split into narrow bins of individual asteroseismic age; if the correlation amplitude does not increase monotonically with decreasing age at $3.5^\\circ<|b|<6.6^\\circ$, the kinematic-fractionation explanation of the metallicity trend would be falsified.","supporting_citations":[{"cited_title":"P., Clarkson W","cited_arxiv_id":null,"evidence_quote":"Earlier study of the same simulation establishing the age-separated kinematics and forbidden-velocity signatures that this work extends."},{"cited_title":"M., Kuijken K., 2007, @doi [ApJ https://doi.org/10.1086/521098] 10.1086/521098 , 665, L31","cited_arxiv_id":null,"evidence_quote":"Measured the vertex deviation of metal-rich and metal-poor bulge stars, the observational pattern this paper reinterprets via the correlation."}],"review_version":1}