{"id":"116e3975-96cd-40b4-ab07-7b3745a0656e","arxiv_id":"2501.00240","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"TESS photometry of 37 LBVs shows no correlation between their stochastic variability parameters and stellar parameters, suggesting their microvariations are an extension of ordinary hot-supergiant (alpha Cygni) variability.","lead":"The authors measured the short-term brightness fluctuations of 37 luminous blue variable stars using TESS space photometry. They find no correlation between these fluctuations and stellar temperature, luminosity, or mass-loss, suggesting LBVs may not be a distinct variability class.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central null result assumes Fourier parameters are intercomparable across TESS baselines that differ by an order of magnitude; without a window-function or matched-baseline test, the absence of correlation could be a sampling artifact.","rationale":"The reader's weakest_assumption identifies the same issue I consider load-bearing: the paper compares Fourier parameters extracted from TESS light curves with 1-4 sectors (Galactic/SMC) versus 19-25 sectors (LMC) without correcting for or even quantifying the window-function effect. The central claim is a null result, and null results are only meaningful if the measurement would have been capable of detecting the sought correlation. Given that nu_char values as low as 0.0013 d^-1 are reported for LMC stars while the Galactic baseline of ~27 days has a frequency resolution of ~0.037 d^-1, many Galactic fits cannot constrain the low-frequency rollover that defines nu_char and alpha0. Without a matched-baseline analysis, the absence of correlation could simply reflect that the two populations are measuring different parts of the Fourier spectrum. I agree with the reader's conditional verdict: the paper's new data and validation against BRITE/ASAS-SN are real contributions, but the central interpretive claim needs the quantitative and comparability tests before acceptance. The concrete test I propose is a matched-baseline re-analysis of the LMC stars; if the parameters are robust to truncation, the concern is settled.","tokens_in":17287,"tokens_out":3940,"duration_ms":39557,"concrete_test":"Re-analyze the LMC/SMC stars in Table 1 using only their first 2-4 TESS sectors, matching the temporal baseline and cadence of the Galactic stars, and recompute alpha0, nu_char, and gamma with the identical Period04/lmfit pipeline. If the truncated-LMC parameters do not move systematically toward the Galactic values (e.g., if nu_char remains low and alpha0 remains high), the baseline mismatch is not responsible for the null result; if they do shift, the paper's central 'no correlation' conclusion is an artifact of comparing 1-month and 2-year light curves.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim (Abstract; Section 4) is that none of alpha0, nu_char, or gamma correlate with H-alpha strength, B-V, or g-magnitude, implying LBV microvariability is an extension of alpha Cygni variability. This is a null result, so it stands or falls on whether the measured Fourier parameters are truly comparable across the sample. Table 1 shows Galactic/SMC stars observed in 1-4 TESS sectors (~1-4 months) while LMC stars have 19-25 sectors (~2 years), with cadence changing from 30-min to 10-min between years 1-2 and 3-4. Section 3 acknowledges 'period04 does not handle gaps in the time-series, which does provide some issues with our Fourier properties, but these Fourier properties should be intercomparable for the population we are examining.' No test of this assertion is provided. The window function's spectral window has a width ~1/T; for a 1-month baseline, low-frequency power below ~0.03 d^-1 is effectively unresolved, while 2-year baselines resolve it. The fitted nu_char values in Table 2 range from 0.0013 to 0.297 d^-1, i.e., the characteristic timescales are 3 to 770 days. A substantial fraction of fitted nu_char values are near or below the frequency resolution of the short-baseline Galactic light curves, so those fits are not measuring the same quantity as the LMC fits. Similarly, alpha0 (zero-frequency amplitude) is strongly influenced by long-term S Dor-type trends, which are present in 2-year LMC light curves but almost entirely absent in 1-month Galactic light curves. If the Galactic sample is biased toward higher nu_char and lower alpha0 purely from window-function effects, while the LMC sample is not, then any true dependence of these parameters on stellar temperature/luminosity/mass-loss could be washed out or masked by population rather than physics. The conclusion therefore rests on an untested comparability assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a Fourier analysis of TESS light curves for 37 luminous blue variables (LBVs) and candidates drawn from the Richardson & Mehner (2018) census. Using the Harvey-style model of Eq. (1), the authors fit the amplitude spectrum of each star to obtain a red-noise amplitude alpha0, a characteristic frequency nu_char, a slope gamma, and a white-noise term Cw (Table 2). They then compare these Fourier parameters with H-alpha equivalent widths, B-V colors, and apparent g-band magnitudes (Figs. 4-7), concluding that none of the parameters correlate with these stellar properties. On this basis they argue that LBV microvariability is an extension of alpha Cygni variability in hot supergiants rather than a distinct class property, and that the LBV classification may not require a variability criterion.","tokens_in":17629,"tokens_out":3400,"duration_ms":34525,"significance":"If the null result is robust, the paper makes a useful observational contribution: it is the first systematic TESS-based Fourier characterization of a sample spanning Galactic, LMC, and SMC LBVs and candidates, more than tripling the sample of Nazé et al. (2021). The comparative light-curve extraction that preserves long-term S Dor-type variability, the validation against the independent BRITE analysis of P Cygni (Elliott et al. 2022), and the public availability of fit parameters in Table 2 are strengths. However, the central claim is a null result, so it stands or falls on whether the measured Fourier parameters are genuinely comparable across a sample that is extremely heterogeneous in time baseline, gap structure, and cadence, and on whether the absence of correlation is established quantitatively rather than by visual inspection. As presented, the analysis does not yet support the strength of the conclusions drawn.","major_comments":[{"comment":"The central claim of no correlation between Fourier parameters and H-alpha strength, B-V, or g-magnitude is based on visual inspection and unquantified curve fitting. No correlation coefficients, significance levels, or confidence intervals on the absence of trends are reported. For a null result that is the main finding, the paper should provide quantitative measures (e.g., Spearman or Pearson r with uncertainties from the MCMC posteriors, or an upper limit on any possible trend) to support 'confirm the absence of correlation' in the abstract.","section":"Section 4 and Figs. 4-7"},{"comment":"The intercomparability of Fourier parameters across the sample is load-bearing and is asserted rather than tested. Table 1 shows Galactic stars observed in 1-4 TESS sectors and LMC stars in 19-25 sectors, with cadence changing from 30-minute to 10-minute in years 3-4. The frequency resolution for a 1-month baseline is about 0.03 d^-1, while many fitted values of nu_char in Table 2 are below this (e.g., HD 269859 at 0.0013 d^-1, HD 269582 at 0.0027 d^-1, HD 269006 at 0.0028 d^-1, HD 34664 at 0.0015 d^-1). The paper's statement in Section 3 that 'period04 does not handle gaps in the time-series, which does provide some issues with our Fourier properties, but these Fourier properties should be intercomparable' is not sufficient. A matched-baseline test (e.g., fitting only the first four sectors of the LMC stars) or an explicit window-function analysis is needed to rule out the possibility that the apparent absence of correlation is a sampling artifact.","section":"Section 3 and Table 1"},{"comment":"Several of the Harvey-model fits are unphysical and appear to drive the dynamic range in the correlation plots. For example, HD 37836 has alpha0 = 1.31e6 +/- 1.9e3 ppt and Cw = 5603 +/- 34 ppt, and AG Carinae has alpha0 = 3264 +/- 99 ppt; these values are many orders of magnitude larger than typical red-noise amplitudes for the other stars and have formally tiny errors after the rescaling described in Section 3. HD 34664 has nu_char = 0.0015 +/- 0.0015 d^-1, i.e., consistent with zero. The paper should assess whether the Harvey model is appropriate for these stars, treat them as outliers or model the long-term S Dor component separately, and show that the null result is not driven by these extreme points.","section":"Table 2"},{"comment":"The comparison with H-alpha strength uses only a subset of the sample, since many stars in Table 2 have no W_lambda(H-alpha) measurement (e.g., HD 269687, HD 269006, HD 269582, and others). The paper does not state how many of the 37 stars have CHIRON H-alpha data, nor how missing values are handled in the correlation plots. The absence of a trend in H-alpha should be demonstrated on the subsample with actual measurements, and the sample size should be reported.","section":"Table 2 and Figs. 4-6"}],"minor_comments":[{"comment":"The axis labels contain typos: 'E uivalent Widths' and 'Characteristic Fre uency' should be 'Equivalent Widths' and 'Characteristic Frequency'.","section":"Figure 7"},{"comment":"The TESS sector list for HD 37974 reads '1-3, 5-6, 8-3', which is likely a typo for '8-13'.","section":"Table 1"},{"comment":"The error-rescaling formula described in the text (multiplying the emcee errors by chi2_{best fit}/(0.5*Ndata - 4)) is not standard and the factor of 0.5 is not justified; a reference or derivation would help the reader understand the reported uncertainties.","section":"Section 3"},{"comment":"The abstract states the paper 'confirm[s] the absence of correlation,' whereas Section 4 describes 'no discernible pattern' and 'no correlation found' through visual inspection and unquantified fitting; the language in the abstract is too strong given the analysis presented.","section":"Abstract and Section 4"},{"comment":"The y-axis label reads 'Log Amplitude (Normalized Flux)' while Table 2 lists alpha0 in ppt; the units should be made consistent or explicitly converted.","section":"Fig. 4 top panel"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a topic well within the scope of the journal and the sample is valuable, but the central null result is not yet supported by the quantitative analysis. The authors should be encouraged to add correlation statistics, a window-function or matched-baseline robustness test, and a treatment of the clearly unphysical fits. These are fixable within the manuscript's scope, so I do not recommend rejection, but the current version is not ready for acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis paper gives us the largest TESS sample of LBVs to date—37 stars, including LMC/SMC objects that had never been studied with high-precision space photometry—and it confirms the earlier Naze et al. result that the Fourier properties of LBV microvariability don't correlate with H-alpha strength, B-V, or luminosity. That null result is the headline, and it's an interesting one: it suggests LBVs might just be an extension of alpha Cygni variables. But the paper supports it mostly with visual inspection of plotted parameters; there are no correlation coefficients or significance tests reported anywhere.\n\nThe new dataset is genuinely useful. The authors validated their light-curve extraction against ASAS-SN and BRITE, and their P Cygni fit matches Elliott et al. That part is solid.\n\nThe soft spots are real but not fatal. Some of the Harvey-model fits are clearly unphysical—HD 37836 has alpha0 = 1.3e6 ppt, which is absurd—and the paper doesn't properly address why those fits should be trusted. More importantly, the sample has a built-in confound: Galactic and SMC stars were observed for 1–4 TESS sectors, while LMC stars got 19–25 sectors. The authors note that Period04 doesn't handle gaps and claim the Fourier properties are 'intercomparable,' but they don't test this. With a 1-month baseline, you can't resolve characteristic frequencies below ~0.03 d^-1, and a large fraction of the Galactic nu_char values are below that. The LMC fits can go much lower. If the baseline effects bias the Galactic values toward higher nu_char and lower alpha0, that alone could wash out any real correlation with stellar parameters. That's a testable worry—they could re-fit everything using a matched two-sector baseline, or at least simulate the window function—and the paper should have done it.\n\nThe conclusion about LBVs being an extension of alpha Cygni variables also lacks a direct comparison sample of non-LBV supergiants; it's inferred from the literature. That's fine for a suggestion, but it's presented more strongly than the analysis supports.\n\nWho should read this? People working on massive star variability will want the table and the light curves. I'd cite it for the dataset, not for the interpretation. It deserves peer review—the sample and null result are worth publishing—but I'd want to see a quantitative correlation analysis, a handling of the poor fits, and a baseline-comparison test before I'd trust the conclusion.\n\nRecommendation: send it to review with the expectation of substantial revision.","headline":"Useful new LBV dataset, but the central null result is under-analyzed and possibly biased by heterogeneous TESS baselines.","tokens_in":18264,"tokens_out":3085,"would_cite":true,"duration_ms":31390,"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":"LBV microvariability shows no link to temperature, luminosity, or mass loss.","keywords":["luminous blue variables","TESS","red noise","alpha Cygni variables","stellar variability","Fourier analysis","massive stars"],"falsifier":"Recompute $\\alpha_0$ and $\\nu_\\mathrm{char}$ for the same 37 stars using a gap-aware method such as Lomb-Scargle periodograms with window-function correction, or by injecting synthetic gaps into continuous light curves; if the fitted parameters shift systematically with the number of TESS sectors or with gap structure, the reported absence of correlation could be an artifact of the sampling, not a property of the stars.","tokens_in":1646,"feed_emoji":"🔭","tokens_out":1896,"duration_ms":41877,"temperature":0.7,"pith_summary":"This paper examines 37 luminous blue variable stars (LBVs) observed by TESS and fits the Fourier spectrum of each light curve to measure the amplitude of low-frequency red noise, a characteristic frequency, and the slope of the noise. It then compares those fitted parameters to H-alpha equivalent width (a proxy for mass-loss rate), B-V color (a proxy for temperature), and g-band magnitude (a proxy for luminosity). The central finding is that none of these comparisons shows a correlation: LBVs with very different temperatures, luminosities, and mass-loss rates have statistically indistinguishable short-term variability properties. The authors read this as evidence that LBVs are not a unique class with respect to their day-to-month variations but rather an extension of the alpha Cygni variables seen among hot supergiants. A sympathetic reader would care because this simplifies the classification of LBVs and points the search for the driving mechanism toward the same processes that govern supergiant variability.","feed_headline":"No link between LBV flickering and stellar properties","feed_subtitle":"37 TESS light curves show red noise independent of H-alpha, B-V, and g-magnitude.","key_machinery":"The central object is the Harvey-style semi-Lorentzian fit to the amplitude spectrum, $\\alpha_\\nu = \\alpha_0 / (1 + (\\nu/\\nu_\\mathrm{char})^\\gamma) + C_w$, where $\\alpha_0$ is the zero-frequency red noise amplitude, $\\nu_\\mathrm{char}$ is the characteristic frequency, $\\gamma$ is the logarithmic slope, and $C_w$ is white noise. The light curves were produced from TESS full-frame images with the eleanor package, Fourier transformed with Period04, and the parameters were estimated with lmfit nonlinear regression followed by an MCMC sampler (emcee). This machinery converts each star's time series into three numbers that can be compared across a heterogeneous sample, and it is what allows the paper to test whether variability properties track stellar parameters.","core_discovery":"The paper claims that the Fourier parameters of LBV light curves, specifically the red noise amplitude $\\alpha_0$, the characteristic frequency $\\nu_\\mathrm{char}$, and the logarithmic slope $\\gamma$, show no correlation with H-$\\alpha$ strength, B-V color, or average g-magnitude. This absence of correlation holds for a sample spanning Galactic, LMC, and SMC LBVs and candidate LBVs, with strong-active, weak-active, and dormant classifications interleaved without any systematic pattern. The authors interpret this as confirmation that the short-term microvariations of LBVs behave like the $\\alpha$ Cygni variability of hot supergiants, and they speculate that the long and short S Doradus cycles may be extensions of the same mechanism rather than a distinct phenomenon.","pith_inferences":["A testable extension would be to compare the fitted Fourier parameters of LBVs with those of non-LBV B-type and A-type supergiants in the same TESS fields, matched in luminosity and temperature, to see whether the distributions overlap completely.","Because the sample mixes very different time baselines and cadences, a gap-aware reanalysis with Lomb-Scargle periodograms or a window-function-corrected Fourier fit could reveal whether the reported null result is robust; this is an inference of the editor, not a claim of the paper.","If the extension hypothesis is right, then long-baseline monitoring of LBVs should show that their S Doradus cycles have the same Fourier shape as their microvariations, just at lower frequencies, which future decade-long TESS or ASAS-SN datasets could test.","The paper leaves open whether the red noise arises from surface convection zones or from internal gravity waves; comparing the phase behavior of photometric and spectroscopic variations for the same stars would help separate these two mechanisms."],"forward_implications":["If the absence of correlation is real, the short-term variability of LBVs does not depend on temperature, luminosity, or mass-loss rate, so those parameters cannot be used to predict how an LBV will flicker.","The lack of a clear boundary between strong-active, weak-active, and dormant/candidate LBVs suggests that these categories may reflect observing epoch and long-term phase rather than intrinsic differences in the microvariability mechanism.","The interpretation that LBVs are an extension of alpha Cygni supergiants implies that the same driving mechanism, whether internal gravity waves or sub-surface convection, could operate across the entire upper H-R diagram.","It follows from the paper's speculation that the long and short S Doradus cycles might be the same stochastic process seen on longer timescales, which would weaken the case for a distinct LBV variability criterion.","If the classification were simplified, future surveys could identify LBV-like stars by their position in the H-R diagram and their stochastic variability properties rather than by requiring observed eruptions."],"supporting_citations":[{"why":"Previous analysis of a smaller set of LBVs with TESS that first reported red noise in all eight stars; the present work extends that sample.","marker":"Nazé et al. (2021)"},{"why":"BRITE-Constellation analysis of P Cygni that the paper uses to validate its Fourier methodology and to compare its measured parameters.","marker":"Elliott et al. (2022)"},{"why":"Establishes the relationship between red noise parameters and luminosity/temperature in main-sequence OB stars, providing the interpretive comparison for the LBV results.","marker":"Bowman et al. (2020)"},{"why":"Provides the sub-surface convection model that is one of the two leading candidate driving mechanisms for the observed stochastic variability.","marker":"Jiang et al. (2018)"},{"why":"Defines the observational classification of LBVs into strong-active, weak-active, and dormant/candidate types used throughout the paper.","marker":"van Genderen (2001)"},{"why":"Supplies the Period04 software used to compute the Fourier transforms of the light curves.","marker":"Lenz & Breger (2005)"},{"why":"Provides the eleanor pipeline used to extract the TESS light curves from full-frame images.","marker":"Feinstein et al. (2019)"}],"fun_headline_variants":["LBV flicker defies stellar properties in TESS data","No trend in LBV variability across 37 TESS stars","TESS reveals LBVs mimic supergiant alpha Cygni behavior","LBV microvariations show no link to H-alpha, color, or mass"],"cache_read_input_tokens":20224,"weakest_assumption_plain":"The analysis treats Fourier parameters derived from light curves with very different lengths, gaps, and cadences as directly comparable, even though the software used does not account for gaps in the time series.","fun_headline_variants_meta":{"raw":{"variants":["LBV flicker defies stellar properties in TESS data","No trend in LBV variability across 37 TESS stars","TESS reveals LBVs mimic supergiant alpha Cygni behavior","LBV microvariations show no link to H-alpha, color, or mass"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000209,"raw_usage":{"total_tokens":1379,"prompt_tokens":891,"completion_tokens":488,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":507,"completion_tokens_details":{"reasoning_tokens":411}},"tokens_in":507,"tokens_out":488,"duration_ms":4697,"temperature":1.0,"reasoning_tokens":411,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:55:35.962115+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute $\\alpha_0$ and $\\nu_\\mathrm{char}$ for the same 37 stars using a gap-aware method such as Lomb-Scargle periodograms with window-function correction, or by injecting synthetic gaps into continuous light curves; if the fitted parameters shift systematically with the number of TESS sectors or with gap structure, the reported absence of correlation could be an artifact of the sampling, not a property of the stars.","supporting_citations":[{"cited_title":"D., Pablo, H., et al","cited_arxiv_id":null,"evidence_quote":"BRITE-Constellation analysis of P Cygni that the paper uses to validate its Fourier methodology and to compare its measured parameters."}],"review_version":1}