{"id":"11789dc6-2e58-4e08-b480-0036fb3ba3ba","arxiv_id":"2412.01089","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Asteroseismic masses of red giants in M9 and M19 give integrated mass loss values of 0.16 and 0.33 solar masses, with Type II cluster M19 losing more mass than Type I clusters of similar metallicity.","lead":"Asteroseismic observations of two globular clusters, M9 and M19, reveal how much mass red giant stars lose as they age: about 0.16 and 0.33 solar masses respectively. The larger loss in M19, a rare 'Type II' cluster, hints that such clusters behave differently from ordinary clusters, which may reshape how we think about their origins.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Type II mass-loss claim rests on EAGB KDE modes built from 4–5 stars; a single reclassification or outlier could erase the M19 offset.","rationale":"The reader's weakest-assumption analysis identifies the small EAGB samples as the key vulnerability, and my independent reading converges on the same point. The M19 mass-loss value is the single observation that drives the Type II interpretation: if the 4-star EAGB mode is perturbed by one reclassification or one outlier, the difference from Type I clusters shrinks below significance. The paper is otherwise careful: the distance-modulus systematic is explicitly tested and shown to largely cancel in the mass difference, the sigma-clipping is described, and the authors hedge their conclusion. The publicly released TPFstitch and pyMON pipelines are positive independent support for the data-processing side. However, the statistical fragility of the EAGB mode is not cured by the quoted standard error, because the mode of a 4-point KDE is not normally distributed and the classification of individual stars is uncertain. This does not require rejecting the paper; it requires the authors to demonstrate robustness of the M19 mode under leave-one-out and bootstrap resampling before the Type II claim is treated as established. The current CONDITIONAL verdict is appropriate, and I would keep it unchanged.","tokens_in":30779,"tokens_out":4201,"duration_ms":39415,"concrete_test":"Perform a leave-one-out and bootstrap analysis of the EAGB mode for each cluster: recompute the KDE mode and ΔM_RGB-EAGB after removing each EAGB star, and generate 10^4 bootstrap resamples of the 4–5 EAGB masses. Report the distribution of ΔM, especially for M19. If the leave-one-out range exceeds the quoted ±0.03 random uncertainty, or if more than ~10% of bootstrap draws put M19's ΔM below 0.20 M⊙, the 'significantly larger' Type II claim is not supported by the current EAGB sample.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central Type I/Type II dichotomy rests on the EAGB mode masses, but these modes are built from only 5 (M9) and 4 (M19) stars after sigma clipping (Sec. 5.1, Tables A1/B1). For M19, the quoted EAGB mode 0.50 ± 0.02 M⊙ is close to the white-dwarf core mass, and the paper itself flags that this may reflect small sample size (Sec. 5.1). With four stars, the KDE mode is not a statistically stable estimator: the quoted standard error on the mean does not capture the sensitivity of the mode to dropping or reclassifying one star, nor the comparable effect of the ±110 K Teff and νmax uncertainties. The phase classification is photometric; the paper admits two M9 EAGB stars (M9AGB70, M9AGB119) could be RGB, and the M19 EAGB sample includes a star labelled M19RGB275 whose phase membership is not independently confirmed. If one M19 EAGB star were reclassified as RGB or its mass shifted by ~0.08–0.12 M⊙, the headline M19 mass loss would drop from 0.33 M⊙ to roughly the M9/M80 range, removing the claimed Type II difference. The distance-modulus test in Fig. 10 does not protect against this, because it only varies a single global parameter.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents the first asteroseismic detection of solar-like oscillations in two Galactic globular clusters, M9 and M19, using K2 Campaign 11 photometry. The authors measure nu_max with a new pipeline (pyMON), determine stellar parameters photometrically, obtain distance moduli from PARSEC and BaSTI isochrone fits, and compute masses with the Delta-nu-independent scaling relation (Eq. 2). Average masses are estimated as modes of KDEs for RGB and EAGB samples after iterative sigma clipping, and integrated mass loss is derived as Delta M_RGB-EAGB = 0.16 +/- 0.02 (rand) +/- 0.03 (sys) M_sun for M9 and 0.33 +/- 0.03 (rand) +0.09/-0.07 (sys) M_sun for M19. They compare these values with M4 and M80, propose a preliminary Type I mass-loss-metallicity relation, and argue that M19, a Type II cluster, has anomalously high mass loss. They also search for bimodal mass distributions indicative of multiple populations.","tokens_in":31065,"tokens_out":6949,"duration_ms":60587,"significance":"If the results hold, this is the first seismic mass-loss measurement in M9 and M19, and the first indication that Type II clusters may follow a different mass-loss-metallicity relation. The paper provides reproducible tools (TPFstitch, pyMON) and carefully separates random and distance-modulus systematics, and the M9 measurement is a useful addition to the seismic mass-loss sample. However, the Type II claim is built on EAGB modes from four stars in M19 (and five in M9), a sample size acknowledged in the text to be too small for robust mode estimation. The paper's own caveats in Sec. 5.1 therefore need to be elevated from a caveat to a central qualification of the headline claim.","major_comments":[{"comment":"The central claim that M19 has significantly larger integrated mass loss than Type I clusters rests on the EAGB KDE mode of 0.50 +/- 0.02 M_sun, built from four stars after sigma clipping, and the M9 EAGB mode is built from five. The text itself notes (Sec. 5.1) that the M19 EAGB average is close to the white-dwarf core mass and 'may be due to the small sample size of four stars'. The quoted random uncertainty is the standard error on the mean; it does not measure the sensitivity of the KDE mode to deleting, reclassifying, or shifting one star. In particular, M19RGB275 (0.49 M_sun) appears in the EAGB sample without an independent phase confirmation, and the M9 stars M9AGB70 and M9AGB119 are admitted to be possible RGB stars (Sec. 5.5). A leave-one-out or jackknife analysis of the EAGB mode, and a recomputation of Delta M for plausible phase reassignments and for +/-110 K Teff shifts, should be reported. Without such a stability test, the abstract's 'significantly larger' and the Type II interpretation are not supported by the present sample.","section":"Sec. 5.1, Table 3, Figs. 7-8"},{"comment":"The mass-loss-metallicity relation for Type I clusters is fit through three points (M4, M80, M9) with no reported uncertainties on the fit parameters, and the same three points are used to argue that M19 deviates. With three points, the slope of 0.24 is not a meaningful fit unless fit uncertainties and a goodness-of-fit or residual test are provided. The phrase 'first mass loss-metallicity trend that uses direct model-independent mass measurements' overstates the case because the masses still depend on the scaling relation (Eq. 2), on photometric temperatures, and on isochrone-fitted distance moduli; these are model-dependent inputs even though Delta-nu is not used. Please either provide a proper fit with parameter uncertainties and a scatter estimate, or present Eq. (4) as an illustrative line and soften the 'first' claim.","section":"Sec. 5.4, Eq. (4)"},{"comment":"The systematic uncertainties quoted in Table 3 are derived only from the distance-modulus variation (Sec. 5.2). The adopted +/-110 K Teff uncertainty and, for M19, the 117 +/- 16 K offset applied to photometric temperatures (Sec. 4.2.2) are not included in the quoted systematic errors for the masses or for Delta M. Since M is proportional to Teff^{-7/2}, a 110 K error at Teff about 4900 K changes individual masses by about 8%, which is comparable to the 0.16-0.33 M_sun mass-loss signal on a roughly 0.6-0.8 M_sun base. The authors should either propagate these Teff systematics into Table 3 or state explicitly why they are negligible for the mass-loss differences.","section":"Table 3, Sec. 5.2"}],"minor_comments":[{"comment":"The text contains typos: 'loose more mass' should be 'lose more mass', and 'spectropscopy' should be 'spectroscopy'.","section":"Sec. 6"},{"comment":"The isochrone fits adopt a Reimers mass-loss parameter eta_R (0.3-0.45) to fit the horizontal branch, and the same isochrones are used to validate the RGB masses in Fig. 9. This is not circular because the derived mass loss is not fed back, but the consistency check in Fig. 9 is partly with models that already assume a mass-loss law; state this limitation when interpreting the 2 sigma agreement.","section":"Sec. 4.1, Fig. 9"},{"comment":"The two-group splits at 0.65 M_sun (M9) and 0.75 M_sun (M19) are chosen after inspecting the data, and the reported uncertainties on the resulting mass differences do not account for the choice of split point. Please state explicitly that these splits are a posteriori and treat the mass-difference estimates as exploratory.","section":"Sec. 5.3"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the journal's scope and the data products are valuable. The main concern is that the headline Type II mass-loss claim is not statistically robust with the current EAGB sample sizes; this is fixable with a stability analysis and appropriately qualified language. I would not reject on the small-sample grounds alone, but the abstract and Sec. 5.4 should be brought into line with the caveats already present in Sec. 5.1."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know before you read this. First, it delivers the first asteroseismic detections and mass-loss measurements for M9 and M19, and the first seismic mass-loss estimate for a Type II globular cluster. Second, the paper is honest and careful, and the headline M19 result — more mass loss at fixed metallicity than Type I clusters — is real but not yet secure.\n\nThe strengths: they use a nu_max-only scaling relation consistently with their earlier M4 and M80 work, propagate random and systematic uncertainties properly, test the distance modulus dependence explicitly, check for blends with a similarity metric, calibrate M19 Teff against APOGEE, and release TPFstitch and pyMON. That is reproducible, citable work.\n\nThe soft spot is the EAGB sample: five stars in M9, four in M19. The paper says this openly and notes the 0.50 Msun M19 EAGB mode is close to the white-dwarf core mass. I disagree with the stress-test claim that one reclassification could erase the M19 offset. The four M19 candidates span 0.46-0.59 Msun; dropping the mislabeled M19RGB275 leaves three stars averaging ~0.53, which still gives a mass-loss of ~0.30, above M9 and M80. You would need a ~0.1 Msun systematic shift in the EAGB masses to bring M19 down to the Type I relation, which is outside their stated systematics. The fragile part is statistical: with four stars, the KDE mode is not a stable estimator, and the standard error on the mean understates the mode's uncertainty. The paper should be pushed to add jackknife tests and report the mean alongside the mode.\n\nThe mass-loss-metallicity relation is a three-point fit and is appropriately labeled preliminary. The bimodality section is speculative, but the authors flag that too. Citation pattern is fine; the self-citations are to their own earlier papers where the method was introduced.\n\nFor anyone working on stellar mass loss or GC multiple populations, this is a useful data point. Verdict: send to peer review. It deserves referee time. A referee should ask for robustness tests on the EAGB sample, not for new data.","headline":"First seismic mass loss for a Type II globular cluster, honestly and carefully done, but the EAGB sample is too small for the Type I/II dichotomy to be secure.","tokens_in":31648,"tokens_out":5561,"would_cite":true,"duration_ms":46615,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper reports the first asteroseismic mass-loss measurements for the globular clusters M9 and M19, and finds that M19—an iron-complex (Type II) cluster—loses significantly more mass than Type I clusters at comparable metallicity.","keywords":["asteroseismology","stars: mass-loss","stars: oscillations","globular clusters","red giants","solar-like oscillations","multiple populations","K2 photometry"],"falsifier":"Measure the early-AGB mass distribution of M19 with a sample of more than four stars, e.g., with the long-baseline space photometry of the kind the paper argues is needed; if the KDE mode shifts away from about $0.50\\,M_\\odot$ by more than the quoted uncertainty, the reported integrated mass loss of $0.33\\,M_\\odot$ is not representative.","tokens_in":30549,"feed_emoji":"🔭","tokens_out":11056,"duration_ms":89472,"temperature":0.7,"pith_summary":"Using K2 photometry, the paper detects solar-like oscillations in 55 red giants in M9 and 37 in M19, and converts them into stellar masses through a scaling relation that uses only the frequency of maximum oscillation power, $\\nu_{\\rm max}$, together with effective temperature and luminosity. It aims to establish that the difference between the average RGB and early-AGB masses gives an integrated mass loss of $0.16\\pm0.02\\,(\\mathrm{rand})\\pm0.03\\,(\\mathrm{sys})\\,M_\\odot$ for M9 and $0.33\\pm0.03\\,(\\mathrm{rand})^{+0.09}_{-0.07}\\,(\\mathrm{sys})\\,M_\\odot$ for M19. These are the first seismic mass-loss values for these clusters, and a sympathetic reader would care because they extend to four the number of globular clusters with asteroseismic mass-loss measurements and because the M19 value breaks the mass-loss-metallicity trend defined by the Type I clusters M4 and M80. The paper also argues tentatively that the EAGB mass distribution in M9 and the RGB distribution in M19 contain substructure consistent with a mass difference between stellar sub-populations, but does not claim a confirmed detection.","feed_headline":"M19's red giants shed twice the mass of M9's","feed_subtitle":"Asteroseismic masses for two more globular clusters hint that iron-complex clusters lose mass differently.","key_machinery":"The load-bearing object is the $\\Delta\\nu$-independent asteroseismic scaling relation $M/M_\\odot \\simeq (\\nu_{\\rm max}/\\nu_{\\rm max,\\odot})\\,(L/L_\\odot)\\,(T_{\\rm eff}/T_{\\rm eff,\\odot})^{-7/2}$, which yields stellar masses from the measured frequency of maximum oscillation power, bolometric luminosity, and effective temperature without needing a measured large frequency spacing $\\Delta\\nu$. The integrated mass loss is then defined as the difference between the modes of kernel-density estimates of the RGB and early-AGB mass distributions, with $\\sigma$-clipping used to keep outlying stars from biasing the modes. The distance modulus, the largest systematic in the individual masses, enters through the luminosity and largely cancels in the RGB-minus-EAGB difference, which is why the paper can quote mass-loss uncertainties much smaller than the systematic uncertainties on the masses themselves.","core_discovery":"The central discovery is that the integrated RGB-to-EAGB mass loss is $0.16\\pm0.02\\,(\\mathrm{rand})\\pm0.03\\,(\\mathrm{sys})\\,M_\\odot$ for M9 and $0.33\\pm0.03\\,(\\mathrm{rand})^{+0.09}_{-0.07}\\,(\\mathrm{sys})\\,M_\\odot$ for M19, measured by taking the modes of the seismic mass distributions of the two evolutionary phases. The M9 value is consistent with the mass-loss-metallicity trend set by M4 and M80, while the M19 value is significantly larger, leading the authors to propose that Type II (iron-complex) globular clusters follow a different mass-loss-metallicity trend than Type I clusters. The paper also claims that the mass distributions show no definitive bimodality, but contain tentative substructure: a possible two-component split in the M9 EAGB sample with a mass difference of $0.09\\pm0.04\\,M_\\odot$, and a possible split in the M19 RGB sample with a mass difference of $0.13\\pm0.03\\,M_\\odot$, both of which require spectroscopic abundance classification to interpret.","pith_inferences":["Editorial inference: if elevated mass loss is a genuine Type II signature, other Type II clusters with future seismic data (for example $\\omega$ Centauri or M22) should also fall above the Type I trend; this is a direct, testable consequence of the paper's interpretation.","Editorial inference: the M9 EAGB bimodality claim rests on five stars, so a natural extension is a Monte Carlo test of how often a five-star draw from a single-peaked distribution produces a KDE shoulder as large as the one seen here.","Editorial inference: the preliminary Type I relation predicts near-zero mass loss for very metal-poor clusters and substantial mass loss for near-solar-metallicity clusters; observing a cluster at either extreme would discriminate the steep seismic trend from the shallower model and dust trends plotted in the paper.","Editorial inference: the paper's use of KDE modes rather than means is well matched to small samples, but means would be more stable if future samples grow; a useful check would be to report both statistics once larger EAGB samples are available."],"forward_implications":["M9 and M19 become the third and fourth globular clusters with asteroseismic mass-loss measurements, giving a four-cluster sample from which the paper derives the first mass-loss-metallicity relation built on model-independent masses: $\\Delta M_{\\rm RGB-EAGB} = 0.24\\,[{\\rm Fe/H}] + 0.55$ (preliminary, Type I clusters only).","If the M19 result is correct, Type II globular clusters cannot be assembled onto the Type I mass-loss-metallicity trend: at a comparable metallicity they lose roughly twice as much mass, implying that cluster type and therefore formation history affect stellar mass loss.","Because the distance modulus largely cancels in the RGB-minus-EAGB difference, the integrated mass-loss values are more robust than the individual stellar masses, which carry a $\\sim$0.1–0.2 $M_\\odot$ systematic from the distance scale.","The tentative bimodal substructure in the M9 EAGB and M19 RGB distributions, if later confirmed by chemical abundances, would support models in which helium-enriched second-generation stars have lower masses and/or enhanced RGB mass loss."],"supporting_citations":[{"why":"Supplies the $\\Delta\\nu$-independent mass scaling method, the KDE-mode averaging, and the M4 mass-loss measurement against which the present results are compared.","marker":"HowellM4"},{"why":"Supplies the M80 seismic mass-loss measurement, the $\\nu_{\\rm max}$-magnitude prediction relation used for target selection, and the previous mass-loss-metallicity comparison.","marker":"HowellM80"},{"why":"First concrete detection of solar-like oscillations in globular-cluster red giants with K2, establishing the feasibility of the present study.","marker":"Miglio et al. (2016)"},{"why":"Provides the $\\Delta\\nu$–$\\nu_{\\rm max}$ scaling relation used to estimate a smoothing $\\Delta\\nu$ when measuring $\\nu_{\\rm max}$ from low signal-to-noise power spectra.","marker":"Stello et al. (2009)"},{"why":"Supplies the solar reference value $\\nu_{\\rm max,\\odot}=3090\\pm30\\,\\mu$Hz used in the mass scaling relation.","marker":"Huber et al. (2011)"},{"why":"Provides the theoretical RGB mass-loss and sub-population mass-difference models used as comparison in the mass-loss-metallicity figure.","marker":"Tailo et al. (2020)"},{"why":"One of the non-seismic RGB mass-loss-metallicity trends against which the seismic measurements are compared.","marker":"Gratton et al. (2010)"},{"why":"Provides the circumstellar-dust-based mass-loss-metallicity trend used as a comparison in the same figure.","marker":"Origlia et al. (2014)"},{"why":"Defines the iron populations and metallicity spread of M19 that the paper uses to interpret its Type II status.","marker":"Yong et al. (2016)"}],"fun_headline_variants":["M19 giants shed double M9's mass","Type II cluster M19 shows double mass loss vs Type I M9","Asteroseismic masses: M19 loses 0.33M☉, M9 0.16M☉","Iron-rich cluster M19 defies mass-loss trend"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The weakest premise is that the early-AGB mode masses in M9 (5 stars) and M19 (4 stars) faithfully represent those clusters' true early-AGB populations; if the tiny samples are biased or misclassified, the quoted integrated mass-loss values change.","fun_headline_variants_meta":{"raw":{"variants":["M19 giants shed double M9's mass","Type II cluster M19 shows double mass loss vs Type I M9","Asteroseismic masses: M19 loses 0.33M☉, M9 0.16M☉","Iron-rich cluster M19 defies mass-loss trend"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000784,"raw_usage":{"total_tokens":3539,"prompt_tokens":1104,"completion_tokens":2435,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":720,"completion_tokens_details":{"reasoning_tokens":2353}},"tokens_in":720,"tokens_out":2435,"duration_ms":20658,"temperature":1.0,"reasoning_tokens":2353,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T04:41:58.062542+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the early-AGB mass distribution of M19 with a sample of more than four stars, e.g., with the long-baseline space photometry of the kind the paper argues is needed; if the KDE mode shifts away from about $0.50\\,M_\\odot$ by more than the quoted uncertainty, the reported integrated mass loss of $0.33\\,M_\\odot$ is not representative.","supporting_citations":[{"cited_title":"G., Carretta E., Bragaglia A., Lucatello S., D'Orazi V., 2010, @doi [ ] 10.1051/0004-6361/200912572 , https://ui.adsabs.harvard.edu/abs/2010A&A...517A..81G 517, A81","cited_arxiv_id":null,"evidence_quote":"One of the non-seismic RGB mass-loss-metallicity trends against which the seismic measurements are compared."},{"cited_title":"R., Fabbri S., Fusi Pecci F., Dalessandro E., Rich R","cited_arxiv_id":null,"evidence_quote":"Provides the circumstellar-dust-based mass-loss-metallicity trend used as a comparison in the same figure."},{"cited_title":"S., Norris J","cited_arxiv_id":null,"evidence_quote":"Defines the iron populations and metallicity spread of M19 that the paper uses to interpret its Type II status."}],"review_version":1}