{"id":"c3ed7202-9157-4991-bf74-a9f800cd7aa8","arxiv_id":"2504.15861","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Flickering in massive stars correlates with luminosity and matches sub-surface convection predictions, supporting that mechanism over internal waves or winds.","lead":"This paper measures the flickering (stochastic low-frequency variability) in light curves of 50 massive stars in the Cygnus OB associations and the Small Magellanic Cloud, using TESS satellite data. It finds that the strength and frequency of the flickering correlate with the stars' luminosity, and that the observed frequencies match predictions from sub-surface convection simulations.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Claimed sub-surface convection agreement is not quantitatively established; IGW models also cover the observed nu_char range, so the mechanism is not uniquely identified.","rationale":"The reader identified the mass-dependent luminosity conversion (Eq. 7) as the weakest assumption. That concern is real but secondary: even with perfect luminosities, the paper's central claim about sub-surface convection would still lack a quantitative basis and would still be degenerate with IGW predictions. The manuscript's own text in Sections 5.2 and 5.3 admits that the IGW simulations also overlap the observed parameter space and that more simulations are needed. Therefore the load-bearing weakness is the overinterpreted agreement with sub-surface convection, not the luminosity conversion. This concern warrants a conditional acceptance: the observational catalog and model-independent method are valuable, but the abstract and conclusion should be softened to state that the observations are consistent with sub-surface convection among other mechanisms, or a quantitative model comparison should be added to substantiate uniqueness.","tokens_in":43655,"tokens_out":7206,"duration_ms":72531,"concrete_test":"Quantitatively compare the observed nu_char values against the sub-surface convection and IGW model predictions in the nu_char-log L plane. For each model (T42L5.2, T35L5.0, M13TAMS from Schultz et al.; Anders et al. 2023; Edelmann et al. 2019 rescaled; Thompson et al. 2024), compute the weighted chi-square using the per-star average nu_char and the sector-to-sector scatter as the uncertainty, after converting model luminosities to spectroscopic luminosity via Eq. 7. If the reduced chi-square for the IGW models is comparable to or better than that for sub-surface convection, the data cannot discriminate between mechanisms, and the abstract should be revised to state consistency rather than agreement.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The abstract's strongest claim is that observed nu_char agrees with sub-surface convection predictions, thereby identifying the variability's physical origin. This claim rests on a qualitative visual comparison in Fig. 9 to only three simulation points (Schultz et al. 2022, 2023b): T42L5.2, T35L5.0, and M13TAMS. No goodness-of-fit or statistical test is provided. Section 5.2 itself acknowledges that 'the weakness of the sub-surface convection theory lies in the disappearance of such prominent zones at lower metallicities' and that more simulations are needed. More critically, Section 5.3 states that the Anders et al. (2023) core-convection IGW predictions 'do fall in regions in Fig. 9 that are covered by the observations,' and that the Edelmann et al. (2019) rescaled predictions 'approximately spans the observed range in nu_char at the given logL/L⊙.' Thus the observed nu_char versus spectroscopic luminosity relation is consistent with both sub-surface convection and IGWs. Since the paper's own text admits this degeneracy, the abstract's unqualified agreement with sub-surface convection overstates the evidence. The central claim of identifying the physical origin is therefore not uniquely supported by the data.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes stochastic low-frequency (SLF) variability in a new sample of 49 O- and B-type stars in six Cygnus OB associations plus the SMC star AV 232, reanalyzes 53 previously studied SLF variables from Bowman et al. (2020), and characterizes the variability by two methods: a Lorentzian-like fit to the power density spectrum (yielding alpha0, nu_char, gamma, C_W) and a model-independent approach (RMS, nu_50%, width w). The authors report Spearman correlations between these parameters and spectroscopic luminosity, place the stars in spectroscopic Hertzsprung-Russell diagrams, and compare the observed nu_char versus luminosity relation to simulations of sub-surface convection, internal gravity waves, and stellar winds. The central claim is that observed nu_char agrees with predictions from sub-surface convection, thereby identifying the physical origin of the SLF variability.","tokens_in":43861,"tokens_out":7228,"duration_ms":70514,"significance":"If the empirical characterization holds, this is a valuable homogeneous sample addition: it extends SLF variability studies to Cygnus OB associations, demonstrates that the new model-independent parameters (RMS, nu_50%, w) are less cadence-dependent than nu_char and gamma, and strengthens the evidence that SLF variability is common among massive stars and scales with spectroscopic luminosity. The paper is careful in several respects: it details the iterative prewhitening, uses nested-sampling Bayesian fits with a well-justified likelihood, explicitly tests cadence-dependent biases (Appendix C), reports Spearman coefficients with p-values, and makes residual light curves, tables, and figure-reconstruction data publicly available on Zenodo. The main weakness is that the physical-origin conclusion is supported only by a qualitative visual comparison to three simulation points, while the paper itself acknowledges that IGW predictions also fall within the observed parameter range.","major_comments":[{"comment":"The abstract's claim of 'good agreement between the observed nu_char of our sample and predictions from sub-surface convection' is not quantitatively established and is inconsistent with the paper's own discussion. The comparison in Fig. 9 rests on only three simulation points (T42L5.2, T35L5.0, M13TAMS) with no goodness-of-fit or statistical test. Section 5.3 states that the Anders et al. (2023) IGW predictions 'do fall in regions in Fig. 9 that are covered by the observations' and that the rescaled Edelmann et al. (2019) predictions 'approximately spans the observed range in nu_char', and Section 5.2 concedes that the sub-surface convection theory is weak at lower metallicities. The data are therefore degenerate between sub-surface convection and IGW mechanisms. The authors should either add a quantitative discriminator (e.g., a likelihood or residual-based comparison for each mechanism) or soften the abstract and conclusions from 'good agreement' to 'consistent with'.","section":"Abstract; Sect. 5.3; Fig. 9"},{"comment":"The abstract states that 'nu_char and nu_50% both decrease' for more evolved stars, but in the full sample the Spearman correlation between log L/L_sun and nu_char is r_s = 0.061 with p > 0.05, i.e., statistically insignificant. Only nu_50% shows a significant negative correlation with log L/L_sun (r_s = -0.381). The nu_char trend appears only as a qualitative impression from the colors in the spectroscopic HR diagram (Fig. 6). Please either quantify the nu_char trend separately with an appropriate significance statement or revise the abstract so that the luminosity-correlation claims match the reported statistics in Table 1.","section":"Table 1; Abstract"},{"comment":"The sector-averaged values of nu_char used in Figs. 8-9 are obtained by mixing 10-min FFI data and 2-min cadence data for different stars, while the B20 comparison sample is entirely 2-min data. The paper's own cadence test in Appendix C shows that resampling 2-min data to 10-min cadence changes the normalized nu_char by a factor 1.285 +/- 0.87 on average, with a standard deviation sigma(Delta nu_char) = 4.35 microHz (Table C1). This is comparable to the scatter seen in Fig. 9 and could introduce a systematic offset between the Cyg OB and B20 points in the simulation comparison. Please quantify the impact of this cadence mismatch on the reported nu_char values and on the conclusions drawn from Figs. 8-9, or restrict the simulation comparison to a homogeneous-cadence subset.","section":"Sect. 2.4; Appendix C; Tables C1-C2"}],"minor_comments":[{"comment":"The text 'Michulski Archive for Space Telescopes' should read 'Mikulski Archive for Space Telescopes'.","section":"Data Availability"},{"comment":"The caption contains 'he averages' and should read 'The averages'.","section":"Table B2 caption"},{"comment":"The caption contains the garbled expression 'w >= w12'; this should likely be 'w >= 12'.","section":"Fig. C3 caption"},{"comment":"The exclusion of 'one low-luminosity star... a clear outlier' is described only qualitatively; please state the quantitative criterion used for this exclusion and, ideally, show that the reported correlations are robust to including or excluding this star.","section":"Sect. 4, Fig. 5"},{"comment":"The notation in Eq. (7) is ambiguous: the left-hand side is a ratio of spectroscopic luminosities but is written with the same symbol L used for the bolometric luminosity. Please introduce an explicit symbol such as L_spec and state the solar constants explicitly, e.g., L_spec/L_spec,sun = (L/L_sun)(M/M_sun)^(-1).","section":"Eq. (7)"}],"recommendation":"major_revision","confidential_remarks":"The empirical sample, reduction procedures, and public data release are strengths, and I see no reason to doubt the measured correlations or the cadence-dependence analysis. My recommendation of major revision is driven solely by the mismatch between the strong physical-origin claim in the abstract and the degenerate, qualitative simulation comparison in Sect. 5. The authors can likely address this by either adding a quantitative treatment of the model comparison or appropriately weakening the claim; the scope and novelty of the observational analysis remain suitable for the journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my read of Pedersen & Bildsten. The empirical work is careful and genuinely useful; the headline interpretation is not supported as stated.\n\nTwo things to know up front. First, the paper gives us a new, well-characterized sample of 49 O/B SLF variables in Cygnus OB associations plus the SMC star AV 232, and it introduces a model-independent set of parameters (RMS, nu_50%, w) that are demonstrably less sensitive to TESS cadence than the usual Lorentzian parameters. That is a real methodological plus, and they back it with careful resampling tests and public data products. Second, the abstract's claim that nu_char agrees with sub-surface convection is stronger than the evidence in the paper. The comparison is visual, against three simulation points (Schultz et al.), and there is no statistical test. Later in Section 5.3 they concede that the Anders et al. IGW predictions 'do fall in regions in Fig. 9 that are covered by the observations' and that the rescaled Edelmann et al. predictions 'approximately span' the observed range. So the data are consistent with more than one mechanism, and the abstract overstates the case.\n\nWhat I think is solid: the data reduction is described in enough detail to be reproducible, the Bayesian fitting and detection threshold are sensible, and the Spearman correlations with p-values are reported rather than hidden. The new parameters are a useful addition to the field. The outlier exclusion is acknowledged, which is fair.\n\nThe soft spots, in rough order: (1) the sub-surface convection agreement is not quantitative; the authors themselves flag the low-metallicity problem and ask for more simulations. This should be fixed by either removing or heavily qualifying the mechanism claim in the abstract. (2) The spectroscopic luminosities use masses from Quintana & Wright and Bouret et al. via Eq. (7). If those masses are biased, the x-axis of the correlations shifts. That is a typical worry for this kind of sample, not a fatal one, but it deserves a sentence as a caveat. (3) Some of the nu_50% values scatter to hundreds of microhertz, which suggests sensitivity to the normalization choice at low S/N. They do show the cadence robustness, but I would like to see how the correlations fare if those extreme points are removed.\n\nWould I send this to a serious referee? Yes. The measurements are reproducible, the method discussion is worth engaging with, and the interpretation, once toned down, is an honest status report. I'd probably bring it to reading group to talk about the cadence test. If I were working on massive-star variability, I'd cite it for the empirical characterization, but not for the mechanism.","headline":"Solid measurements, overstated mechanism claim: the new Cygnus OB sample and cadence-robust parameters are worth having, but the data do not uniquely identify sub-surface convection.","tokens_in":44447,"tokens_out":2785,"would_cite":true,"duration_ms":27271,"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":"The paper finds that the stochastic low-frequency flicker of 103 O- and B-type stars is best explained by subsurface convection in the iron opacity zone, matching 3-D simulation predictions.","keywords":["stochastic low-frequency variability","massive stars","O-type stars","B-type stars","subsurface convection","iron opacity zone","TESS photometry","Cygnus OB associations"],"falsifier":"Take a subset of the Cygnus OB stars and measure their masses independently from binary orbits or asteroseismology, then re-derive the correlations; if the spectroscopic-luminosity trends vanish, the central comparison fails. Alternatively, run a 3-D envelope simulation at SMC metallicity for a star near 35 $M_\\odot$ and check whether its $\\nu_{\\rm char}$ falls on the observed relation, or test whether observed $\\nu_{\\rm char}$ tracks the thermal timescale of the iron-opacity convection zone across stars with and without such zones.","tokens_in":43392,"feed_emoji":"🌟","tokens_out":7662,"duration_ms":63029,"temperature":0.7,"pith_summary":"Massive O- and B-type stars show a common 'flicker' — stochastic low-frequency variability — in high-cadence space photometry, but its physical origin has been contested. This paper characterizes that flicker in 49 stars from six Cygnus OB associations plus the SMC star AV 232, and re-analyses 53 previously published stars, giving a combined sample of 103. It finds that the amplitude and slope of the variability, its RMS, the frequency containing 50% of the power, and the spectral width all correlate with spectroscopic luminosity, with amplitudes growing and characteristic frequencies falling as stars evolve. When the observed characteristic frequencies are compared with simulations, the paper finds good agreement with 3-D simulations of subsurface convection in the iron opacity zone, identifying that convection as the likely driver of the flicker. A model-independent characterization based on RMS, $\\nu_{50\\%}$, and $w$ is introduced and shown to be less affected by TESS observing cadence than the Lorentzian-fit parameters.","feed_headline":"Massive-star flicker points to subsurface convection","feed_subtitle":"TESS data on 103 O- and B-type stars match 3-D simulations of convection in the iron opacity zone.","key_machinery":"The argument is carried by two characterization tools and a comparison. The first is a Lorentzian-like fit to the power density spectrum, $M(\\nu) = \\eta(\\nu)\\,\\alpha_0 / (1 + (\\nu/\\nu_{\\rm char})^{\\gamma}) + C_W$, which yields the zero-frequency amplitude $\\alpha_0$, the characteristic frequency $\\nu_{\\rm char}$, and the slope $\\gamma$. The second is model-independent: the RMS of the residual light curve plus the cumulative integrated power $P_{\\rm int}(\\nu)$, from which the paper derives $\\nu_{20\\%}$, $\\nu_{50\\%}$, $\\nu_{80\\%}$ and the width $w = (\\nu_{80\\%}-\\nu_{20\\%})/\\nu_{50\\%}$. To compare stars across samples, bolometric luminosities are converted to spectroscopic luminosities via $L_{\\rm spec} = L\\,(M/M_\\odot)^{-1}$, which makes the stellar mass a load-bearing input. The decisive comparison is observed $\\nu_{\\rm char}$ against predictions from 3-D radiation-hydrodynamical simulations of subsurface convection in the iron opacity zone, along with predictions for internal gravity waves from convective cores.","core_discovery":"On the paper's own terms, the central discovery is that the stochastic low-frequency variability of massive main-sequence O- and B-type stars scales systematically with stellar luminosity and evolution, and that its characteristic frequency $\\nu_{\\rm char}$ matches the predictions of 3-D simulations of subsurface convection in the iron opacity zone. The paper reports significant correlations between spectroscopic luminosity and $\\alpha_0$, $\\gamma$, RMS, $\\nu_{50\\%}$, and $w$ across the full sample of 103 stars; $\\alpha_0$ and RMS increase for more evolved stars while $\\nu_{\\rm char}$ and $\\nu_{50\\%}$ decrease. Against the alternative explanations of surface granulation, internal gravity waves excited by the convective core, and stellar winds, the $\\nu_{\\rm char}$ comparison favours subsurface convection, with only partial overlap for internal gravity waves and no support from the granulation scaling relations.","pith_inferences":["If subsurface convection is the driver, then $\\nu_{\\rm char}$ should track the thermal timescale of the iron-opacity convection zone; this can be tested directly by computing that timescale from 1-D stellar models for each star in the sample.","Because about two-thirds of the stars show negative skewness but the skewness does not correlate with luminosity, photometric skewness alone is unlikely to isolate stellar winds; combining TESS light curves with UV spectroscopy could separate wind and convection contributions.","A natural extension is to apply the same RMS/$\\nu_{50\\%}$/$w$ analysis to the larger LMC and SMC samples with known metallicities, where the predicted absence of iron-opacity convection below certain masses would make the subsurface-convection hypothesis falsifiable."],"forward_implications":["Stochastic low-frequency variability is nearly universal among massive O- and B-type stars above $\\log L/L_\\odot \\geq 4$, with 49 of 54 such stars in the Cygnus OB sample showing the signal.","Because $\\alpha_0$ and RMS increase while $\\nu_{\\rm char}$ and $\\nu_{50\\%}$ decrease with evolution, SLF parameters can serve as coarse evolutionary-stage indicators for massive stars.","The RMS, $\\nu_{50\\%}$, and $w$ parameters are much less affected by TESS observing cadence than $\\alpha_0$ and $\\nu_{\\rm char}$, so future surveys with only 10-minute FFI data can still characterize SLF variability reliably.","If the $\\nu_{\\rm char}$ agreement with subsurface-convection simulations holds, TESS flicker observations become a direct probe of convection in the iron opacity zone of massive stars."],"supporting_citations":[{"why":"Supplies the 53 comparison O/B stars and the Lorentzian-based SLF characterization that this work redoes in the power density spectrum.","marker":"Bowman et al. (2020)"},{"why":"Provides the 3-D radiation-hydrodynamical simulations of 35 $M_\\odot$ envelopes whose $\\nu_{\\rm char}$ agrees with the observed sample.","marker":"Schultz et al. (2022)"},{"why":"Adds a 13 $M_\\odot$ terminal-age main-sequence simulation that extends the subsurface-convection comparison to lower masses.","marker":"Schultz et al. (2023b)"},{"why":"Predicts where iron-opacity subsurface convection zones exist, used to place AV 232 at the low-metallicity boundary.","marker":"Jermyn et al. (2022)"},{"why":"Gives observed granulation-frequency relations for lower-mass stars, used to rule out surface granulation as the SLF source.","marker":"Kallinger et al. (2014)"},{"why":"Wind simulations predicting negative skewness, compared to the observed skewness of the residual light curves.","marker":"Krtička & Feldmeier (2021)"},{"why":"3-D core-convection IGW simulations without core boosting, used as an alternative comparison for $\\nu_{\\rm char}$ and amplitudes.","marker":"Anders et al. (2023)"},{"why":"Source of the Cygnus OB sample and the stellar masses used to compute spectroscopic luminosities.","marker":"Quintana & Wright (2021)"},{"why":"Provides effective temperature, luminosity, and spectroscopic mass for the SMC star AV 232.","marker":"Bouret et al. (2021)"}],"fun_headline_variants":["Massive-star flicker traces subsurface convection","TESS data links stellar flicker to iron-zone convection","Flicker frequency of 103 stars matches convection model","Low-frequency shimmer in massive stars hints at convection","Stellar flicker scales with luminosity, matches convection"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire luminosity trend and the comparison to simulations assume that the stellar masses taken from published spectral analyses are accurate; if those masses are systematically biased, the spectroscopic luminosities shift and the reported correlations could weaken or disappear.","fun_headline_variants_meta":{"raw":{"variants":["Massive-star flicker traces subsurface convection","TESS data links stellar flicker to iron-zone convection","Flicker frequency of 103 stars matches convection model","Low-frequency shimmer in massive stars hints at convection","Stellar flicker scales with luminosity, matches convection"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000719,"raw_usage":{"total_tokens":3285,"prompt_tokens":1061,"completion_tokens":2224,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":677,"completion_tokens_details":{"reasoning_tokens":2148}},"tokens_in":677,"tokens_out":2224,"duration_ms":15378,"temperature":1.0,"reasoning_tokens":2148,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T11:15:38.176743+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a subset of the Cygnus OB stars and measure their masses independently from binary orbits or asteroseismology, then re-derive the correlations; if the spectroscopic-luminosity trends vanish, the central comparison fails. Alternatively, run a 3-D envelope simulation at SMC metallicity for a star near 35 $M_\\odot$ and check whether its $\\nu_{\\rm char}$ falls on the observed relation, or test whether observed $\\nu_{\\rm char}$ tracks the thermal timescale of the iron-opacity convection zone across stars with and without such zones.","supporting_citations":[],"review_version":1}