{"id":"a39ede20-6947-40f7-9f2d-47ceb919a413","arxiv_id":"2505.01291","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"In three bursts of 4U 1636-536, kHz QPOs are absent for 100-200 s after a Type-I burst and return around 200 s, suggesting temporary inner disk disruption and viscous restoration.","lead":"This paper finds that rapid X-ray flickering from the innermost accretion disk of a neutron star disappears for about 200 seconds after a thermonuclear burst and then returns. The timing matches viscous disk refilling, supporting the idea that bursts temporarily disrupt the inner disk.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Post-burst non-detection may be a broadened/damped QPO rather than true disappearance; the cited 10-20 keV check is not shown in §4.","rationale":"The reader's verdict of CONDITIONAL is appropriate. The paper does careful work: it uses AstroSat LAXPC data, checks trials-corrected null-hypothesis probabilities (Appendix A), reports upper limits, and is transparent about caveats (Section 5.1) and the excluded burst OB4. The pre-burst presence of kHz QPOs is well established (SNR 3.3-7.5, p<0.05 after trials), and the post-burst upper limits are below the pre-burst rms, which already rules out the simplest 'masked by noise' scenario because the count rates before and after are comparable. The remaining vulnerable step is the inference from 'not detected' to 'non-existent' and then to 'inner disk disrupted.' A QPO that is strongly broadened or frequency-drifting after the burst would be missed by both the fixed-parameter upper-limit fit and the narrow-bin blind search, so the data are consistent with a survived but damped oscillation. The paper's bullet about a 10-20 keV non-detection would have been the natural way to test for energy-dependent masking, but that analysis is not shown in Section 4; this is an omitted support. We therefore agree with the reader's core concern but sharpen it: the issue is not burst background noise per se, but the insensitivity of the detection tests to altered QPO morphology. The OB4 exclusion is disclosed and can be defended (no pre-burst QPO to track), so it does not change our assessment. The viscous timescale argument is explicitly order-of-magnitude and does not carry the load. Because the observational pattern is interesting and the statistical treatment is careful, the appropriate verdict remains CONDITIONAL: the disappearance pattern is likely real for narrow QPOs, but the physical interpretation of disk disruption needs either the 10-20 keV analysis or an injection-recovery test to close the broadening loophole.","tokens_in":16384,"tokens_out":9951,"duration_ms":96619,"concrete_test":"Simulate the post-burst 0-200 s segments for TNB-1/2/3 by injecting a synthetic Lorentzian QPO with the pre-burst frequency, rms, and centroid, but with widths broadened by factors of 2, 3, and 5, and in a second run with a linear frequency drift of ~100 Hz across the segment, into the real event files. Re-run the same detection pipeline (fixed-parameter fit of Table 1 and blind search of Table 2). If the injected broadened/drifting QPO is recovered at SNR ≥ 3, then the pipeline is sensitive to such signals and the observed non-detection would be evidence for true disappearance; if it is not recovered, the upper limits are too insensitive to rule out a damped survivor, and the 'non-existence' claim is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires that the absence of a kHz QPO in the first 100-200 s after each TNB is a real disappearance, not a detection-threshold artifact. The upper limits in Table 1 (e.g., Obs 1 rms <8.36% vs. pre-burst 20±3%, SNR 6.7 vs. 1.1) do rule out an unchanged QPO being hidden by Poisson noise, since the source count rates before and after are nearly equal (Section 4). However, the upper limits are derived by fixing the Lorentzian centroid and width to the pre-burst values, and the blind search in Table 2 is optimized over narrow frequency bins; neither test has meaningful sensitivity to a QPO whose width has broadened by a factor of several (e.g., due to burst-driven turbulence) or whose frequency drifts over the 200 s segment. Such a damped/broadened oscillation would spread power across many bins and evade both the fixed-parameter fit and the narrow-bin search, while still indicating that the oscillation mechanism survived the burst. The paper's own Section 5.1 concedes that non-detection 'may result not only from physical disruption but also from instrumental limitations.' Moreover, the bullet in Section 5 claiming 'non-detection of QPOs in 10-20 keV, where burst intensity is lower, ruling out energy band dependence' is not backed by any PDS analysis in Section 4, which reports only 3-20 keV power spectra. Without that energy-resolved check, the strongest rebuttal to the masking/damping alternative is missing. Thus the data establish suppression of the narrow QPO, but not the 'non-existence' asserted in the abstract, and the step from suppression to inner-disk disruption remains interpretive.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes three AstroSat LAXPC observations of the neutron-star low-mass X-ray binary 4U 1636-536, selecting one Type-I burst per observation (TNB-1, TNB-2, TNB-3) for which a lower kHz QPO is detected in the 100-200 s before burst onset. The authors fit power density spectra in the 3-20 keV band and report that the lower kHz QPO is not detected in the first 100-200 s after the burst (rms upper limits of ~4-8.4%), reappears after ~200 s, and that the fractional rms drops by ~5-6%. They interpret this as temporary disruption of the inner accretion disk by burst radiation, followed by viscous refilling on a ~200 s timescale, using t_visc = R_in^2/nu with R_in = 4e6 cm and nu = 1e11 cm2/s. Section 5.1 lists caveats including the small sample and possible instrumental masking of the post-burst signal.","tokens_in":16644,"tokens_out":5146,"duration_ms":51026,"significance":"If the disappearance is real, the paper provides one of the few systematic before/after characterizations of kHz QPO evolution across Type-I bursts in a single source, extending the earlier RXTE-based work of Peille et al. (2014). The pre-burst/post-burst count-rate matching and the null-hypothesis probability formalism in Appendix A, validated with simulated dead-time-affected event files, are careful strengths; Table 2 consistently shows p < 0.05 for segments with SNR > 3. The main weakness is that the post-burst non-detection is an upper-limit result whose physical interpretation depends on excluding masking or broadening of the oscillation, and one of the paper's stated supporting checks (the 10-20 keV band) is not actually shown.","major_comments":[{"comment":"The post-burst \"non-detection\" is established only through Lorentzian fits with centroid and width fixed to the pre-burst values (e.g., Obs 1 and Obs 3 post-burst rows with daggered parameters) and through a blind search that optimizes over narrow frequency bins (Table 2). Neither test has demonstrated sensitivity to a QPO that survives the burst but broadens or drifts, for example due to burst-driven turbulence or a changing inner disk radius. Since the central conclusion is that the oscillation disappears rather than is hidden, the authors should add a search over broader Lorentzian widths or an integrated excess-power statistic over 400-1200 Hz and report the corresponding upper limits for the 0-200 s post-burst segments.","section":"§4 and Table 1"},{"comment":"The bullet claims that non-detection of QPOs in the 10-20 keV band, \"where burst intensity is lower,\" rules out energy-band dependence, but Section 4 presents only 3-20 keV power spectra and no energy-resolved PDS analysis is shown. This check is the most direct rebuttal to the masking/broadening alternative and should either be presented explicitly or the claim should be removed.","section":"§5, first bullet"},{"comment":"The text states that the third burst in Observation 3 (OB4) is omitted from the analysis even though it \"meets the necessary conditions,\" because no QPO is detected in either the pre-burst or post-burst zone. Because this is a qualifying burst whose behavior differs from the three analyzed bursts, its exclusion weakens the claim of a systematic post-burst disappearance and ~200 s reappearance. The authors should either include OB4 in the analysis or justify its exclusion with explicit, pre-defined criteria.","section":"§3-4, OB4 in Observation 3"},{"comment":"The viscous-timescale agreement is not an independent test: R_in is fixed at 4e6 cm from a companion paper, the viscosity range 1e10-1e13 cm2/s is broad, and the value nu = 1e11 cm2/s that produces ~160-200 s is effectively chosen because it matches the observed delay. This should be presented as an order-of-magnitude consistency check with a free effective viscosity, not as a measurement of the viscosity, and the degeneracy between R_in and nu should be discussed.","section":"§5, Eqs. (2)-(4)"},{"comment":"The claimed \"drop of approximately 5-6%\" in fractional rms is not what Table 1 shows for the lower kHz QPO: TNB-1 drops from 20 +/- 3% to an upper limit of <8.36%, and TNB-3 drops from 11 +/- 3% to <5.0%, which are larger drops or upper limits only. The quantitative statement in the abstract and Section 4 should be reconciled with the tabulated values.","section":"Abstract, §4, and Table 1"}],"minor_comments":[{"comment":"The sentence describing Type-II TNBs says their duration can range from milliseconds to \"a few fours\"; this should read \"a few hours.\"","section":"§1"},{"comment":"The pre-burst interval is -200 to 0 s for Observations 1 and 3 but -100 to 0 s for Observation 2, while the text refers generically to \"100-200 sec before the burst\"; the interval definitions should be stated consistently in one place.","section":"§4 and Table 1"},{"comment":"The x-axis tick labels near \"108 109\" appear to be garbled time labels and should be fixed.","section":"Figure 2"},{"comment":"The phrase \"The kHz QPOs then re-emerges after approximately 200 sec\" has a subject-verb agreement error and should be corrected.","section":"Abstract"},{"comment":"The column header \"p (1-p)%\" is ambiguous; it should be made clear that the tabulated quantity is the confidence level 1 - p_N(<P_max).","section":"Table 2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of the journal and the empirical design is sound in its core before/after comparison. The requested additions are feasible within the manuscript's scope: a broader-signal search for the post-burst segments, the energy-resolved check, and a transparent discussion of the OB4 exclusion and the viscous-timescale degeneracy. I would not reject, because the matched count rates and the null-hypothesis formalism give real weight to the observed sequence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this if you care about kHz QPOs in neutron star LMXBs. It's a clean but incremental follow-up to Peille et al. (2014): three AstroSat bursts of 4U 1636-536, lower kHz QPO visible before, absent for 100-200 s after, back by ~200 s. The analysis is a step up from much of the field—trials-corrected null-hypothesis probabilities, simulated dead-time checks, and explicit upper limits. The authors also list honest caveats, including the possibility that burst photon flux masks the QPO.\n\nWhat's genuinely new: post-burst coverage out to 1200 s, which shows the QPO frequency and rms recovering; and a quantitative comparison of the recovery time with a viscous timescale, though that comparison is loose—a viscosity chosen from a 1e10-1e13 cm^2/s range to match the observed 200 s.\n\nThe soft spots are real but not disqualifying. The biggest is an unsupported bullet in Section 5: 'Non-detection of QPOs in 10–20 keV... ruling out energy band dependence.' Section 4 only shows 3-20 keV PDS. Either add that energy-resolved analysis or retract the claim. Second, the non-detection is only tested against a narrow Lorentzian. A broadened or drifting QPO would evade both the fixed-width upper limits and the blind search. The paper concedes this in §5.1, but the abstract still says 'non-existence,' which overstates the data. Third, the sample excludes one burst (OB4) that met the criteria but had no pre-burst QPO; that's a selection on pre-burst detection, and it makes the 'disappearance' claim partly tautological. Minor: the abstract's '5–6% rms drop' doesn't hold for TNB-1, where the limit goes from 20% to <8.4%.\n\nNet: the core observation is probably right, and the paper's own caveats are ahead of many in the field. It deserves a serious referee, but it needs the energy-band claim fixed and the language toned down. I'd send it out.","headline":"A careful but incremental AstroSat extension of Peille et al. showing kHz QPO suppression for ~200 s after three bursts in 4U 1636-536, with honest caveats but one unsupported energy-band claim.","tokens_in":17304,"tokens_out":4796,"would_cite":false,"duration_ms":42236,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A Type-I thermonuclear burst on the neutron star 4U 1636-536 makes the inner disk's kHz quasi-periodic oscillations disappear for about 200 seconds, and they return as the disk refills on a viscous timescale.","keywords":["kHz quasi-periodic oscillations","Type-I X-ray bursts","neutron star low-mass X-ray binaries","accretion disk disruption","viscous timescale","4U 1636-536","X-ray timing analysis"],"falsifier":"Reconstruct the post-burst power spectrum with the burst emission removed and with longer effective exposure, for instance by co-adding several bursts of the same source: a QPO appearing in the 0-200 s window at the pre-burst frequency with signal-to-noise ratio above 3 would refute the claimed disruption-and-refilling scenario.","tokens_in":16136,"feed_emoji":"💥","tokens_out":15062,"duration_ms":129504,"temperature":0.7,"pith_summary":"This paper tries to establish that a Type-I thermonuclear burst on the neutron star in 4U 1636-536 temporarily destroys the innermost part of its accretion disk, and that the disk rebuilds on a well-defined viscous time scale. The key evidence is the behavior of the lower kHz quasi-periodic oscillation: it is present in the 200 seconds before each of three bursts, is not detected in the first 100-200 seconds afterward, and reappears after roughly 200 seconds, with the fractional rms dropping by about 5-6% in the post-burst window. The recovery time matches the viscous refilling time $t_{\\mathrm{visc}} \\approx R_{\\mathrm{in}}^2/\\nu$ for a plausible viscosity, which is why the authors read the pattern as disk disruption followed by restoration. If true, this provides a time-resolved link between burst radiation and inner-disk dynamics in accreting neutron stars, and it supports the idea that kHz QPOs are produced in the innermost flow.","feed_headline":"Type-I burst erases inner-disk oscillations for about 200 seconds","feed_subtitle":"The 200-second gap matches the viscous time for the disk to refill after burst radiation pushes it back.","key_machinery":"The central object is the kHz quasi-periodic oscillation—a rapid, roughly 700-1100 Hz brightness wobble thought to trace the innermost accretion flow—used as a diagnostic of whether that flow is present. The identity that carries the argument is the viscous refilling time, $t_{\\mathrm{visc}} \\approx R_{\\mathrm{in}}^2/\\nu$, evaluated at the inner disk radius; with $R_{\\mathrm{in}} = 4\\times 10^6$ cm and $\\nu \\sim 10^{11}$ cm$^2$/s it gives about 160 s, close to the observed ~200 s gap. The paper also uses Lorentzian fits to the power density spectra, with signal-to-noise and null-hypothesis probabilities, to certify which intervals contain a QPO and which only support upper limits.","core_discovery":"The paper's central discovery is that the kHz QPO in 4U 1636-536 follows a reproducible disappear-and-reappear cycle around Type-I bursts. In each of the three analyzed bursts, the lower kHz QPO (centroid roughly 683-768 Hz before the burst) is detected in the 100-200 s before burst onset, is not detected in the first 100-200 s afterward with signal-to-noise ratio below 3 and rms upper limits of about 4-8.4%, and re-emerges around 200 s later. The fractional rms amplitude in the 3-20 keV band falls by about 5-6% in the immediate post-burst window. The authors interpret this as burst radiation pushing the inner accretion flow outward; once the burst ends, the inner disk refills on a viscous time scale, and the QPO returns. For an inner radius $R_{\\mathrm{in}} = 4\\times 10^6$ cm and kinematic viscosity $\\nu \\sim 10^{11}$ cm$^2$/s, the viscous time $t_{\\mathrm{visc}} \\approx R_{\\mathrm{in}}^2/\\nu \\approx 160$ s, matching the observed ~200 s restoration.","pith_inferences":["If radiation pressure is the cause, bursts closer to the Eddington limit should produce longer QPO-free gaps; a larger burst sample could test this correlation quantitatively.","The same analysis applied to other atoll neutron-star sources with frequent bursts could turn the ~200 s restoration into a general measure of inner-disk viscosity rather than a single-source result.","Because the first 200 s are upper limits rather than detections, co-adding many bursts in the same spectral state could push the rms limits low enough to reveal a weak residual QPO, which would discriminate between full disruption and partial suppression."],"forward_implications":["If the claim is right, the ~200 s QPO-free gap is a direct signature of inner-disk disruption: the same gap is seen after all three bursts regardless of spectral state or peak intensity.","The restoration time, matching $t_{\\mathrm{visc}} \\approx R_{\\mathrm{in}}^2/\\nu$ with $\\nu \\sim 10^{11}$ cm$^2$/s, turns burst-QPO timing into a probe of disk viscosity in this source.","Because kHz QPOs vanish and return with the inner flow, their recovery lets observers watch the inner disk rebuild in real time after a burst.","The upper kHz QPO can appear about 200 s after the burst even when it was undetectable before, suggesting the high-frequency part of the flow may recover ahead of the full oscillation pattern."],"supporting_citations":[{"why":"Prior report that kHz QPOs in 4U 1636-536 vanish and reappear after some bursts; this paper extends that result with longer time coverage.","marker":"Peille, P. et al. (2014)"},{"why":"Model placing kHz QPOs at characteristic radii of the innermost accretion flow, used to interpret the disappearance as inner-disk disruption.","marker":"Stella & Vietri 1998, 1999"},{"why":"Provides the signal-to-noise formula used to mark QPO detections and upper limits.","marker":"van der Klis (2004)"},{"why":"Provides the null-hypothesis probability approach used to confirm detections at better than 95% confidence.","marker":"Barret et al. (2008)"},{"why":"Source of the viscous time formula and the viscosity range used for the recovery-time estimate.","marker":"Frank et al. (1985)"},{"why":"Supplies the viscous disk refilling picture invoked for the ~200 s QPO recovery.","marker":"Spruit (1995)"},{"why":"Additional reference for the viscous time scale used in the comparison between the observed gap and the refilling time.","marker":"Ingram (2012)"}],"fun_headline_variants":["kHz QPO vanishes for 200 s after Type-I burst, then returns","200-second gap: disk disruption wipes out kHz QPOs after burst","Burst radiation pushes disk out; QPO re-emerges on viscous timescale","Type-I burst silences QPO for ~200 s, matching viscous refill time"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the missing kHz oscillation in the first 100-200 seconds after the burst is a real disappearance, not a signal hidden by burst-related noise or reduced sensitivity.","fun_headline_variants_meta":{"raw":{"variants":["kHz QPO vanishes for 200 s after Type-I burst, then returns","200-second gap: disk disruption wipes out kHz QPOs after burst","Burst radiation pushes disk out; QPO re-emerges on viscous timescale","Type-I burst silences QPO for ~200 s, matching viscous refill time"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000728,"raw_usage":{"total_tokens":3342,"prompt_tokens":1111,"completion_tokens":2231,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":727,"completion_tokens_details":{"reasoning_tokens":2144}},"tokens_in":727,"tokens_out":2231,"duration_ms":14154,"temperature":1.0,"reasoning_tokens":2144,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:22:04.230675+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Reconstruct the post-burst power spectrum with the burst emission removed and with longer effective exposure, for instance by co-adding several bursts of the same source: a QPO appearing in the 0-200 s window at the pre-burst frequency with signal-to-noise ratio above 3 would refute the claimed disruption-and-refilling scenario.","supporting_citations":[{"cited_title":"2014, A&A, 567, A80, doi: 10.1051/0004-6361/201423784","cited_arxiv_id":null,"evidence_quote":"Prior report that kHz QPOs in 4U 1636-536 vanish and reappear after some bursts; this paper extends that result with longer time coverage."},{"cited_title":"Accretion Disks","cited_arxiv_id":"astro-ph/9502098","evidence_quote":"Supplies the viscous disk refilling picture invoked for the ~200 s QPO recovery."},{"cited_title":"2012, PhD thesis, Durham University","cited_arxiv_id":null,"evidence_quote":"Additional reference for the viscous time scale used in the comparison between the observed gap and the refilling time."}],"review_version":1}