{"id":"429a0fec-213c-4ee4-9daf-34aee45c66e1","arxiv_id":"2505.08714","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Be stars in the Small Magellanic Cloud are more often in active disk-building phases and have fewer, longer outbursts than Large Magellanic Cloud Be stars, though mass differences between the samples may explain part of the effect.","lead":"Using 20 years of OGLE photometry, this study tracks disk build-up and dissipation in more than 3,000 Be stars in the Large and Small Magellanic Clouds. Be stars in the lower-metallicity Small Magellanic Cloud spend more time feeding their disks and keep detectable disks almost continuously, though the statistical origin of this difference is partly entangled with stellar mass.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The SMC-vs-LMC duty-cycle difference is confounded by mass: within each mass bin the DC distributions are statistically indistinguishable (Fig. 12), while the SMC mass distribution is biased toward higher masses, so the metallicity claim is not yet established.","rationale":"I selected mass confounding as the load-bearing concern because the paper's own numbers provide a direct falsification route: the headline DC difference is large (0.60 vs 0.44), but when mass is held fixed in Fig. 12 the DC difference disappears. The SMC sample is also smaller and biased toward high mass, and mass correlates strongly with all activity diagnostics in Sect. 6.3. Thus the central metallicity claim is underdetermined by the published analysis. I do not press the reader's model-grid/manual-classification concern as the primary attack, because even a perfect classification would not rescue the conclusion if the samples are not mass-comparable. The appropriate verdict is therefore CONDITIONAL, unchanged from the reader: the empirical distributions are useful, but the metallicity conclusion requires mass-matched analysis or explicit selection-bias correction. This is a partial agreement with the reader, who flagged the model grid as the weakest assumption but also noted the mass-matched KS tests in the rationale.","tokens_in":28791,"tokens_out":6517,"duration_ms":66568,"concrete_test":"Recompute the Fig. 8 and Fig. 9 comparisons using only the mass-matched subsample (Groups 2/3, N=468) and, separately, reweight the full Be-candidate sample by the inverse probability of having a mass estimate given I-band brightness and variability amplitude. If the mass/selection-adjusted median DC difference between LMC and SMC falls below the reported 0.16 or the KS p-value for DC rises above 0.05, the SMC-activity claim is a mass-selection artifact rather than a metallicity effect.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing threat to the central claim is not the inclination grid itself but the comparison base for DC and DDC. The headline median DC difference (0.60 SMC vs 0.44 LMC) is computed over the full Be-candidate sample, whose LMC and SMC mass distributions differ strongly (median 6.6 vs 9.5 Msun, Fig. 11) and whose SMC subsample is explicitly admitted to be biased toward high-mass, high-activity stars (Sect. 7.1). Since Sect. 6.3 shows DC, DDC, and Nout all increase with stellar mass, the population-level DC excess could be a selection artifact. The paper's own mass-matched KS tests (Fig. 12) find no significant LMC/SMC difference in DC in any mass bin (p = 0.36, 0.36, 0.21) or in DDC (p = 0.81, 0.47, 0.61); the only significant mass-matched difference is Nout in the M2 bin (p = 1.5e-4), not in the headline DC/DDC statistics. The authors acknowledge the activity bias and upper-limit status of DC (Sect. 7.2), but the abstract and conclusions assert the metallicity dependence. Unless the mass confound is removed by representative mass estimates for Group 1 stars or by explicit weighting, the central claim is not supported.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes ~2,144 LMC and ~989 SMC OGLE light curves of Be star candidates, manually classifying each light curve into phases (baseline, disk build-up, dissipation, plateau, isolated events) with guidance from viscous decretion disk models. From these classifications the authors define a duty cycle (DC), a disk duty cycle (DDC), an outburst rate (Nout), and, for a subsample with isolated events, build-up and dissipation durations. The central claim is that SMC Be stars are more active: they spend a larger fraction of time in mass-ejection phases (median DC 0.60 vs 0.44 in the LMC), have nearly permanent disks (median DDC 1.0 vs 0.99), and show longer but less frequent outbursts. The paper also reports that all activity diagnostics increase with stellar mass. The analysis is avowedly model-informed and the authors acknowledge several selection biases, including the exclusion of inactive Be stars and the under-representation of low-mass SMC stars.","tokens_in":29184,"tokens_out":4426,"duration_ms":46527,"significance":"If the central claim survives, this is the first population-level measurement of Be star mass-loss duty cycles in the Magellanic Clouds and a unique comparison of disk-activity diagnostics across two metallicities. The dataset is large, the light curves are real OGLE data, and the definitions of DC/DDC/Nout are straightforward, transparent, and easy to reproduce from the phase classifications. The mass determination via MCMC on photospheric models is a standard approach, and the use of KS tests is appropriate. The paper also provides a useful catalog of isolated disk events with durations, which will be valuable for future modeling. The main weakness is that the headline metallicity comparison is confounded by the very different mass distributions of the LMC and SMC samples, a concern that the authors themselves raise in Sect. 7.1 but do not resolve in the quantitative analysis.","major_comments":[{"comment":"The central claim that SMC stars are more active than LMC stars is not supported by the mass-matched comparison. The KS tests reported in Fig. 12 show that, within each mass bin, DC differences are not significant (p = 0.36, 0.36, 0.21) and DDC differences are not significant (p = 0.81, 0.47, 0.61). Because Sect. 6.3 demonstrates that DC and DDC increase with stellar mass, and Sect. 7.1 explicitly admits that the SMC subsample is biased toward high-mass stars (median 9.5 Msun vs 6.6 Msun for the LMC), the population-level median differences (0.60 vs 0.44 for DC; 1.0 vs 0.99 for DDC) are likely selection artifacts. The abstract and Sect. 8 currently assert a metallicity dependence without qualifying that the within-mass-bin analysis shows no significant difference; this should be corrected and the mass-matched result should be presented as the primary evidence, or the sample must be reweighted/expanded to remove the confound.","section":"Sect. 6.3 and Fig. 12; Sect. 7.4"},{"comment":"The mass-matched KS tests have limited statistical power because the SMC sample sizes in the mass bins are small (M1: 19, M2: 44, M3: 17), so the null result cannot be taken as evidence that the distributions are actually indistinguishable. The text in Sect. 6.3 says the distributions are 'statistically indistinguishable' for DC and DDC; a more cautious phrasing (e.g., 'no significant difference was detected') is needed, along with a discussion of the power or confidence intervals. Moreover, the mass-matched analysis is restricted to Groups 2 and 3, which constitute only 22% of the SMC Be candidates (Table 7), so the mass-matched subsample may itself not be representative of the full SMC population.","section":"Sect. 6.3, Fig. 12"},{"comment":"The manual classification of light curves into seven phase categories is the foundation of every duty cycle and outburst-rate measurement in the paper, yet the manuscript provides no validation of the classification's reliability. No inter-rater agreement test, repeatability check, or systematic comparison of the manual labels with the quantitative fit of Eqs. 5 and 6 (beyond the single example in Fig. 6) is reported. Given that the central demographic results rest entirely on these manual labels, the paper should either provide a reproducibility assessment (e.g., re-classifying a random subset by a second observer) or a sensitivity analysis showing that the headline median differences survive plausible classification uncertainties.","section":"Sect. 5 and Sect. 6.1"},{"comment":"The dynamical model grid used to classify stars as pole-on or edge-on and to identify build-up/dissipation phases is computed for a single 12 Msun, LMC-metallicity star with W=0.81 and alpha=1.0, while the sample spans masses from about 3 to 20 Msun and includes SMC (lower-metallicity) stars. The paper itself notes that the V-I loop orientation depends on base density (Sect. 4.1, bottom panel of Fig. 4) and that the intermediate-inclination range produces photometric signals below the detection threshold (Fig. 5), which biases the sample toward extreme inclinations. The authors should either run a small grid of models covering the mass/metallicity range and show that the phase classification is robust, or explicitly state which conclusions are insensitive to this modeling choice. This is a correctness-risk concern, not a circularity argument, but it is load-bearing because the phase labels determine all of DC, DDC, and Nout.","section":"Sect. 4.1, Table 2, Fig. 5"}],"minor_comments":[{"comment":"The abstract's phrase 'strong statistical differences' refers to the full-sample KS tests, but the mass-matched tests (Fig. 12) are the more relevant comparison for the metallicity claim; the abstract should be reworded to reflect the qualified nature of the result.","section":"Abstract and Sect. 8"},{"comment":"The KS test p-values are denoted by the symbol ρ (rho), which is unconventional and could be confused with a correlation coefficient; the paper should use 'p' or 'p-value' throughout.","section":"Sect. 6.1, Figs. 8 and 9"},{"comment":"The text says the mass was determined for 468 stars in Groups 2 and 3, but the earlier statement that Groups 2 and 3 contain 816 stars may confuse readers; a sentence clarifying that mass determination required two-band photometry (V and I, or B and I) and thus was possible for only 468 of the 816 would improve clarity.","section":"Sect. 6.2"},{"comment":"The sentence listing the mass-bin counts ('217, 151, 20 (LMC) 19, 44 and 17') is hard to parse; a table or a clearer format (e.g., 'LMC: 217, 151, 20; SMC: 19, 44, 17') would be an improvement.","section":"Sect. 6.3"},{"comment":"The discussion of DDC biases is honest and useful, but the conclusion that 'the disk is present for most of the observational period' (Sect. 8) should explicitly carry the caveat that this is an upper limit affected by the activity selection and a lower limit affected by the I-band detection of the inner disk only; the current text in Sect. 8 does not fully convey this tension.","section":"Sect. 7.2"}],"recommendation":"major_revision","confidential_remarks":"The dataset and the systematic phase classification are valuable contributions, and the paper is well within the scope of the journal. However, the central metallicity claim as stated in the abstract and conclusions is not supported by the paper's own mass-matched analysis, and the authors' admitted selection biases are not sufficiently accounted for. The major revisions I request (mass-matched analysis as primary or reweighting, validation of the manual classification, and a sensitivity test for the model grid) are, in my view, feasible within the manuscript's scope. I therefore recommend major revision rather than rejection, provided the authors either reframe the central claim to match the evidence or add the necessary analysis to justify it. I would also suggest that the editor ask the authors to make the classification data (phase labels per star and per time segment) publicly available, as it would greatly increase the reproducibility and long-term value of this work."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing in this paper is the duty-cycle (DC) and disk-duty-cycle (DDC) statistics for ~1,750 variable Be candidates in the LMC and SMC, plus the isolated disk-event parameters (tb, td, ΔI). No one has measured those on this scale before, and the OGLE light curves are real, with 20-year baselines. The manual classification is labor-intensive and, as far as I can tell, carefully done; the paper is also honest about its limitations in Sections 7.1 and 7.2.\n\nThe soft spot is the central claim. The abstract and conclusions say lower-metallicity SMC Be stars are more active, but the paper's own mass-binned KS tests (Fig. 12) show no significant LMC/SMC difference in DC or DDC within any mass bin. The only mass-matched difference is Nout in the M2 bin (p = 1.5e-4). The population-level DC difference (0.60 vs 0.44) is dominated by the fact that the SMC subsample is biased toward high-mass stars, and DC, DDC, and Nout all rise with mass. The authors acknowledge the SMC mass bias in Section 7.1, yet the conclusions still phrase the metallicity effect as established. That needs to be fixed.\n\nLesser concerns: the inclination classification relies on a single 12 Msun, alpha = 1.0, LMC-metallicity dynamical model; for a population study that is a coarse tool, though the paper uses it only for pole-on/edge-on binning and the authors flag the ambiguity. The manual phase labels are not released as a data product, which makes the DC/DDC numbers hard to audit; code release would help. The event-fit formulas come from Rimulo et al. (2018), so the self-citation is legitimate, but the new fit parameters here are not independently calibrated.\n\nNone of this kills the paper. The empirical distributions are valuable on their own; they are the first of their kind and will be useful for anyone modeling Be disk statistics. But the metallicity conclusion needs reframing. I would send this to a good referee, with a request that they push on the mass-confounding and the reproducibility issue. If the authors reweight the samples or acknowledge that the mass-matched comparison shows no metallicity effect on DC/DDC, the paper would be solid.","headline":"First population-level duty-cycle measurements for Magellanic Cloud Be stars, but the headline metallicity claim does not survive the paper's own mass-matched tests.","tokens_in":29740,"tokens_out":2274,"would_cite":true,"duration_ms":22940,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["97.30.Eh","97.10.Fy"],"model":"deepseek-v4-flash","headline":"First population-level measurement of Be star duty cycles: lower-metallicity SMC stars spend more time feeding disks, keep disks almost permanently, and show longer but rarer outbursts.","keywords":["Be stars","Magellanic Clouds","viscous decretion disk","duty cycle","disk duty cycle","OGLE light curves","stellar mass loss","metallicity"],"falsifier":"Take the 605 isolated disk events and refit them with dynamical models computed at each star's own mass, rotation, and metallicity instead of the single 12-solar-mass template; if the fitted disk densities and durations no longer reproduce the reported SMC-LMC amplitude and duration gaps, the metallicity conclusion would need revision. A cheaper check is observational: point H-alpha or mid-infrared observations at stars sitting at their I-band baseline, since residual disk emission there would falsify the diskless-baseline premise and push every disk duty cycle upward.","tokens_in":28631,"feed_emoji":"⭐","tokens_out":19040,"duration_ms":160204,"temperature":0.7,"pith_summary":"This paper analyzes about 20 years of OGLE photometry for more than 3,000 Be stars in the two Magellanic Clouds and classifies each light curve, with help from viscous-disk models, into phases of disk build-up, plateau, dissipation, and quiescence. Its central claim is the first population-level measurement of how much time Be stars actually spend ejecting mass (the duty cycle) and how often they harbor a detectable disk (the disk duty cycle). The result: stars in the lower-metallicity Small Magellanic Cloud are markedly more active than Large Magellanic Cloud stars, with median duty cycles of 0.60 versus 0.44, disks present essentially all the time (median disk duty cycle 1.0 versus 0.99), and outbursts that are longer but slightly rarer (median 0.26 versus 0.31 per year). If right, these numbers turn the erratic, individual behavior of Be stars into a demographic description that constrains how stellar mass loss depends on mass and metallicity.","feed_headline":"60% of the time, metal-poor Be stars feed their disks","feed_subtitle":"A 20-year survey of 3,000 light curves finds SMC Be stars build disks longer and more often than LMC stars.","key_machinery":"The load-bearing object is the V-I color-magnitude loop: in the viscous decretion disk (VDD) model, a disk event makes a pole-on star brighten and redden while an edge-on star dims with little color change, and the loop's orientation rotates with inclination. That diagnostic was recomputed here in the I band, using the singlebe hydrodynamic code coupled to the hdust radiative-transfer code for one fiducial 12-solar-mass star (rotation rate W = 0.81, viscosity parameter alpha = 1.0) across four feeding times and four base densities. A second grid of diskless, fast-rotating photospheric models at LMC and SMC metallicities converts a dereddened baseline position into an estimate of stellar mass. On top of these, an interactive manual classification assigns each segment of every light curve to one of seven categories (baseline, build-up, dissipation, plateau, isolated build-up, isolated dissipation, unclassified), from which the duty cycle, disk duty cycle, and outburst rate are computed, and published mass-reservoir fitting formulas are applied to isolated events to extract build-up and dissipation durations.","core_discovery":"The discovery, stated at population scale, is that Be star activity differs sharply between the two Clouds: among the 1,751 variable Be star candidates, SMC stars spend a median 60% of their time actively feeding a disk versus 44% for LMC stars, both populations keep a detectable disk for essentially the whole observing window (median disk duty cycles of 1.00 and 0.99), and SMC outbursts are longer but less frequent (median 0.26 per year versus 0.31). For 605 isolated disk events, SMC systems reach photometric amplitudes about three times larger and last notably longer (median total duration 645 days versus 407 days), which the authors read as evidence that low-metallicity disks are denser or fed for longer. Within both galaxies, every activity metric grows with stellar mass: the most massive stars show outburst rates about three times those of the least massive, and dissipation outlives build-up by a nearly constant ratio of about 1.5-1.6 that confirms the mass reservoir effect. The authors state that the median duty-cycle, disk duty-cycle, and outburst-rate values are upper limits, since quiet stars were intentionally excluded from the analyzed sample.","pith_inferences":["A consequence the paper leaves implicit: the duty-cycle statistics could be inverted into a completeness correction for Be star censuses, since a star is counted as 'Be' only when a disk is present and will be misclassified in a single-epoch survey with probability equal to the complement of its disk duty cycle; with disk duty cycles near unity, the already high SMC Be fraction may be close to the","A testable extension the paper does not perform: refitting its catalogued isolated events with per-star masses, inclinations, and metallicities could reveal whether the viscosity parameter differs between the Clouds, in which case part of the reported amplitude and duration gap would be a viscous effect rather than a difference in mass-ejection behavior.","The paper's own caveat that I-band emission traces only the inner disk implies its disk duty-cycle numbers are floor values; H-alpha or mid-infrared follow-up of stars at their photometric baseline would likely show residual outer-disk emission, pushing the effective disk duty cycle even closer to 1."],"forward_implications":["The Be state in these samples is near-permanent: a detectable disk exists essentially all of the time, even between mass-ejection episodes, so the disk is closer to a persistent condition than to a transient outburst for these objects.","Metallicity shapes the entire mass-loss phenomenology, not just the frequency of Be stars: at lower metallicity, stars spend more time feeding disks, build denser disks, and produce longer events.","Stellar mass drives activity within each galaxy: more massive stars have higher duty cycles, disk duty cycles, and outburst rates, with the outburst rate of the most massive bin about three times that of the least massive.","Because the sample deliberately excluded quiet stars, the reported medians are upper limits; adding the excluded inactive Be stars would lower the medians but leave the shape of the distributions above the first bin unchanged.","The near-constant dissipation-to-build-up ratio of about 1.5-1.6 in both galaxies gives population-level support for the mass reservoir effect, in which mass stored in the outer disk slows the fading of the inner disk."],"supporting_citations":[{"why":"Supplies the VDD model result that a disk event traces a loop in the V-I plane whose orientation encodes inclination, the basis for the pole-on/edge-on classification used throughout.","marker":"Haubois et al. (2012)"},{"why":"Supplies the mass reservoir effect and the analytic build-up/dissipation formulas (Eqs. 5 and 6) fitted to isolated disk events.","marker":"Rímulo et al. (2018)"},{"why":"Provides the observed range of disk base densities used to set the dynamical model grid.","marker":"Vieira et al. (2017)"},{"why":"Documents the mass reservoir effect in a well-studied star (omega CMa), the template for why dissipation outlasts build-up.","marker":"Ghoreyshi et al. (2018)"},{"why":"Sets the prior outburst-classification methodology for long-baseline photometry that this paper extends into a dynamical phase taxonomy.","marker":"Labadie-Bartz et al. (2017)"},{"why":"Establishes the SMC Be fraction and low-metallicity rotation behavior used to explain why SMC stars are more active.","marker":"Martayan et al. (2007)"},{"why":"Gives independent mid-infrared build-up (~474 d) and decay (~524 d) durations used as a cross-check on the Magellanic event timescales.","marker":"Jian et al. (2024)"},{"why":"Supplies the reddening maps that convert baseline magnitudes and colors into intrinsic values for mass estimation.","marker":"Skowron et al. (2021)"},{"why":"Provides Galactic outburst rates and their mass dependence against which the Magellanic Nout values are compared.","marker":"Bernhard et al. (2018)"}],"fun_headline_variants":["SMC Be stars feed disks 60% of the time, LMC only 44%","Low-metallicity Be stars keep disks longer and feed them more","20-year survey: SMC Be stars outfeed LMC in disk duty","Metal-poor Be stars: more disk feeding, longer bursts"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole analysis rests on the assumption that the brightness-and-color behavior predicted for a single 12-solar-mass model star reliably reveals each real star's orientation and whether it is building or losing a disk, and that a star sitting at its I-band baseline has no disk at all; if either premise fails across the diversity of real LMC and SMC stars, every duty-cycle and outburst statistic inherits the error.","fun_headline_variants_meta":{"raw":{"variants":["SMC Be stars feed disks 60% of the time, LMC only 44%","Low-metallicity Be stars keep disks longer and feed them more","20-year survey: SMC Be stars outfeed LMC in disk duty","Metal-poor Be stars: more disk feeding, longer bursts"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000846,"raw_usage":{"total_tokens":3756,"prompt_tokens":1095,"completion_tokens":2661,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":711,"completion_tokens_details":{"reasoning_tokens":2591}},"tokens_in":711,"tokens_out":2661,"duration_ms":17566,"temperature":1.0,"reasoning_tokens":2591,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:48:32.182085+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the 605 isolated disk events and refit them with dynamical models computed at each star's own mass, rotation, and metallicity instead of the single 12-solar-mass template; if the fitted disk densities and durations no longer reproduce the reported SMC-LMC amplitude and duration gaps, the metallicity conclusion would need revision. A cheaper check is observational: point H-alpha or mid-infrared observations at stars sitting at their I-band baseline, since residual disk emission there would falsify the diskless-baseline premise and push every disk duty cycle upward.","supporting_citations":[{"cited_title":"2024, , 682, A59, 10.1051/0004-6361/202347911","cited_arxiv_id":null,"evidence_quote":"Gives independent mid-infrared build-up (~474 d) and decay (~524 d) durations used as a cross-check on the Magellanic event timescales."}],"review_version":1}