{"id":"0638d86a-7aca-4e37-91b6-ae539bb03312","arxiv_id":"1908.04906","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"In six sunspots mapped across five AIA temperature channels, five-minute oscillation power forms an expanding ring around the umbra that becomes disordered in the high corona, while three-minute power concentrates in the umbra and at coronal loop footpoints.","lead":"This paper maps where three- and five-minute oscillations appear inside six sunspots, from the temperature minimum to the corona, using SDO/AIA images. It finds five-minute oscillations form an expanding circle around the umbra that breaks down in the hot corona, while three-minute oscillations stay mostly in the umbra and reach coronal loop footpoints.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Per-pixel 95% chi-square tests without multiple-comparison correction leave the disordered high-corona five-minute power maps indistinguishable from false positives, so the suppression claim is not yet established.","rationale":"The reader's weakest assumption correctly identifies the per-pixel multiple-comparison problem as the load-bearing statistical issue. My reading of the paper confirms that the high-corona five-minute 'disappearance' is the central new inference, and it depends on interpreting sparse, spatially unstructured power in 211 and 131 angstrom maps as a physical suppression rather than as threshold noise. The paper's own wording ('almost hard to find the circular shape') shows that the high-corona conclusion is based on visual inspection of a null pattern. The proposed phase-randomization test would settle whether the observed pattern is statistically distinguishable from noise while preserving the red-noise properties of the data. The low-atmosphere ring and three-minute umbral/footpoint patterns are corroborated by earlier work and are not the main vulnerability. Therefore the CONDITIONAL verdict is appropriate, and no adjustment is needed.","tokens_in":7310,"tokens_out":6869,"duration_ms":76183,"concrete_test":"Generate phase-randomized surrogate light curves for each pixel in the 211 and 131 angstrom maps of all six sunspots: Fourier-transform the detrended time series, randomize the phases, inverse-transform to preserve the red-noise amplitude spectrum, and then run the Section 4 detection pipeline (linear fit in log-log, 95% chi-square threshold, sum of super-threshold power in the 3.3 +/- 0.5 mHz band). Build surrogate power maps and compare the fraction of nonzero pixels, summed power, and spatial clustering (e.g., Moran's I) of the observed 211/131 maps against the surrogate distribution. If the observed values fall within the surrogate envelope, the disordered high-corona five-minute pattern is consistent with pure noise and the suppression claim is unsupported; if they are outliers, the pattern is physically real.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4 constructs each power map by fitting each pixel's log-log spectrum, flagging bins above the 95% chi-square confidence level, and summing flagged power in a 1 mHz band. With 3600 s of data, the 3.3 mHz band contains about 3-4 independent Fourier bins, so a pure red-noise pixel has roughly an 18% chance of at least one flagged bin (1 - 0.95^4). The maps cover about 80x80 pixels, predicting roughly a thousand pixels flagged by chance in a signal-free channel. The observed 211 and 131 angstrom maps show exactly this sparse, spatially disordered pattern, and Section 4.1 relies on visual inspection (\"almost hard to find the circular shape\") before concluding that five-minute oscillations are suppressed in the high-temperature corona (abstract and Section 5). No false-discovery-rate or family-wise correction is applied, and no detection-sensitivity calibration for these channels is reported. The low-atmosphere expansion claim is not threatened because its spatial coherence is far above chance and it has prior support from Kolobov et al. (2016), but the high-corona suppression claim rests on an uncalibrated null result: absence of significant power is equated with physical suppression.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript analyzes one-hour SDO/AIA observations of six sunspots in five channels (1700, 304, 171, 211, and 131 Å) to map the spatial distribution of three- and five-minute oscillations. For each pixel it fits a power-law red-noise model to the log-log FFT spectrum, flags bins above a 95% chi-square confidence level, and sums the flagged power in 1 mHz bands centered at 3.3 and 5.6 mHz (Section 4). The authors report that five-minute power forms a circle near the umbra/penumbra boundary at low temperatures, expands outward to the lower corona, and becomes spatially disordered or disappears in 211 and 131 Å; they interpret this as suppression in the high-temperature corona. Three-minute power is mostly umbral at low temperatures and partly follows coronal fan structures with high power at loop footpoints.","tokens_in":7430,"tokens_out":4384,"duration_ms":43083,"significance":"The topic is timely and the use of six sunspots across five AIA channels is a useful observational extension of previous work, particularly the confirmation of Kolobov et al. (2016) for the low-atmosphere expansion. If the high-corona suppression of five-minute oscillations and the loop-footpoint concentration of three-minute oscillations can be placed on solid statistical footing, the results would be a valuable constraint on wave propagation models. The paper is clearly written and the data processing is reproducible in principle, but the statistical support for the central high-corona null result is currently incomplete.","major_comments":[{"comment":"The per-pixel 95% confidence test is applied independently to every pixel, but no multiple-comparison correction is reported. With one-hour 12 s cadence data, the 3.3 mHz band contains roughly three to four independent Fourier bins, so a pure red-noise pixel has about 1 - 0.95^4 ≈ 18% chance of containing at least one flagged bin; over an 80×80-pixel map this predicts of order 10^3 flagged pixels by chance. The sparse, disordered patterns seen in the 211 and 131 Å panels of Figure 3 are exactly what such false positives would look like. Therefore the conclusion in Sections 4.1 and 5 that five-minute oscillations are suppressed in the high-temperature corona is not established unless a false-discovery-rate or family-wise correction is applied and the spatial coherence of the detections is quantified.","section":"Section 4, step 3"},{"comment":"Treating the absence of significant power as physical suppression requires a sensitivity calibration for each channel. The paper does not provide detection limits for 211 and 131 Å, an injection-recovery test, or error bars on the power maps. Consequently the comparison between low- and high-corona maps cannot separate \"oscillations are suppressed\" from \"oscillations are present but below the detection threshold of this method.\" This concern is load-bearing because the high-corona suppression is a central claim repeated in the abstract and Section 5.","section":"Section 4.1"},{"comment":"The umbra/penumbra boundary is defined by \"an intensity threshold\" on the 1700 Å image, but the threshold value, the region over which it is applied, and the sensitivity of the boundaries to the threshold are not given. All spatial claims—circle at the umbral boundary, expansion to the penumbral boundary, and power within the umbra—depend on these boundaries, so the procedure should be specified and a robustness test should be included.","section":"Section 3"}],"minor_comments":[{"comment":"Only three of the six sunspots appear in the figures (NOAA 12638 in Figure 3, NOAA 11176 in Figure 4, and NOAA 11479 in Figure 5); the claims for the remaining spots rest on unreferenced textual statements.","section":"Figures 3-5"},{"comment":"The text refers to the blue shadow in Figure 1 as a period range of 2.5 to 5.5 minutes, but the power maps use 3.3 and 5.6 mHz centers; the relationship between these ranges should be clarified.","section":"Figure 1 and Section 4"},{"comment":"The cadence difference between 1700 Å (24 s) and the other channels (12 s) is not discussed in relation to frequency resolution and red-noise fitting; this could affect the comparability of the power maps.","section":"Section 2"},{"comment":"There are several grammatical issues, for example \"the circle-shape is disappeared\" and \"the powers with three-minute also appear to be a decrease trend\"; a language edit would improve readability.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"No concerns about scope or citation practice are raised here. The main barrier is statistical: the high-corona suppression claim needs either a multiple-comparison correction and sensitivity calibration, or a careful qualification of the claim as a tentative null result. A revision addressing this would be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nWhat should you know? This paper analyzes six sunspots in five AIA channels and produces power maps for 3- and 5-minute oscillations. The low-atmosphere results—expanding ring of 5-minute power, umbra-concentrated 3-minute power leaking to loop footpoints—are solid confirmations of Kolobov et al. (2016) and Reznikova & Shibasaki (2012). The authors say so explicitly, which is honest. The one candidate new result, that 5-minute oscillations become \"disordered\" and suppressed in 211 and 131 Å, does not survive a close look at the statistics.\n\nThe method is standard: detrend, FFT, fit a red-noise power law in log-log space, flag bins above a 95% chi-square confidence level. That's fine for a single spectrum. The problem is they apply it per pixel and build power maps from the flagged power. With ~3600 s of data, the 3.3 mHz band has ~3–4 independent bins, so a pure red-noise pixel has ~18% chance of at least one flag. With 80x80 pixels, that predicts about a thousand false positives per map. The sparse, scattered power in the 211/131 maps is exactly what that null model produces. So the \"disordered\" pattern and the suppression claim are arguably an artifact. The paper applies no multiple-comparison correction and gives no detection sensitivity for those channels.\n\nOther soft spots: no error bars on the power maps, the umbra/penumbra threshold is never specified, and only three of six sunspots are shown in detail for the corona. The \"randomly selected\" sample is fine, but the presentation makes it hard to assess how representative the omitted ones are.\n\nTo the paper's credit, the low-atmosphere expansion is spatially coherent and consistent with prior work, so that part holds. The three-minute loop-footpoint concentration also matches earlier findings. The analysis is reproducible in principle and the writing is clear, though the conclusion overstates what the high-corona maps can support.\n\nWho is this for? Someone doing sunspot seismology who wants a quick multi-event confirmation of known spatial distributions. It is not a new mechanism or a decisive test.\n\nRecommendation: send to peer review, but a serious referee should ask for a proper treatment of the per-pixel false-positive rate—either a false-discovery-rate correction, a detection threshold set by Monte Carlo, or analysis of the distribution of flagged pixels against a pure red-noise null. As it stands, the headline claim about high-corona suppression is not established. With that fixed, the paper would be a useful reference.","headline":"A competent confirmation study whose only genuinely new claim—suppression of five-minute oscillations in the high corona—is weakened by per-pixel significance testing without multiple-comparison correction.","tokens_in":8073,"tokens_out":2414,"would_cite":false,"duration_ms":23542,"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":"Sunspot five-minute oscillations form an expanding ring that disappears in the hotter corona, while three-minute oscillations gather in the umbra and at coronal loop footpoints.","keywords":["sunspot oscillations","five-minute oscillations","three-minute oscillations","power maps","SDO/AIA","solar atmosphere","coronal fan loops"],"falsifier":"Count false positives by re-running the same per-pixel test on shuffled light curves or by applying a false-discovery-rate correction; if the high-corona maps still show as many significant pixels with no spatial organization, the claimed five-minute suppression is not established and the disordered maps are plausibly pure noise.","tokens_in":6996,"feed_emoji":"🌞","tokens_out":6434,"duration_ms":62120,"temperature":0.7,"pith_summary":"Using one-hour image sequences of six sunspots from five SDO/AIA channels, this paper tries to establish how the two well-known sunspot oscillation periods, three minutes and five minutes, are arranged in space as the surrounding plasma temperature rises from about six thousand to ten million kelvin. The authors find that five-minute power forms a ring at the umbra-penumbra boundary, expands outward through the transition region and lower corona, and becomes disordered and largely absent in the higher corona; they read this as temperature-dependent suppression. Three-minute power stays mostly inside the umbra, but part of it follows coronal fan structures, with the strongest patches at loop footpoints. If the picture holds, it gives wave-propagation models a concrete, height-resolved geometry to reproduce and connects sunspot oscillations to energy transport into the corona.","feed_headline":"Sunspot five-minute oscillations vanish in the hot corona","feed_subtitle":"Five-minute power rings expand then vanish; three-minute power hugs loop footpoints.","key_machinery":"Per-pixel fast Fourier transform (FFT) power spectra normalized by intensity variance; each log-log spectrum is fitted with a linear red-noise model, and a chi-square 95% confidence level is used to select significant peaks. Power in two frequency bands, 1 mHz wide and centered at 3.3 mHz for five-minute oscillations and 5.6 mHz for three-minute oscillations, is summed to make power maps. The maps are overlaid on umbral and penumbral boundaries derived from 1700 Å intensity thresholds, allowing the paper to convert thousands of individual spectra into spatial maps of where each oscillation mode is significant at each temperature.","core_discovery":"Across six sunspots observed in five AIA channels spanning temperatures from about 0.6 × $10^{4}$ K to $10^{7}$ K, the paper finds that five-minute oscillation power forms a circle around the umbra-penumbra boundary in the temperature minimum, expands outward through the transition region and lower corona, and becomes disordered and largely absent in the higher corona, as seen in the 211 Å and 131 Å channels. The paper interprets this as temperature-dependent suppression of five-minute oscillations in high-temperature plasma. Three-minute oscillation power remains mostly inside the umbra up to the lower corona, with part of it appearing along coronal fan structures; the strongest patches sit near coronal loop footpoints. The paper concludes that three-minute oscillations can propagate along fan loops and that the two modes have distinct, temperature-stratified spatial distributions.","pith_inferences":["A testable extension is to overlay the power maps on co-temporal photospheric magnetograms: if the high-power coronal patches really mark loop footpoints, they should coincide with strong inclined-field footpoints rather than with random bright pixels.","The same per-pixel pipeline applied to pores and plage would separate a penumbra-specific ring mechanism from a generic temperature effect, since those structures lack a penumbra.","If the high-corona suppression is a genuine wave cutoff, it should appear as a frequency-dependent effect in Doppler data: five-minute power should fade before three-minute power as temperature rises, a prediction that simultaneous spectral observations could check."],"forward_implications":["Five-minute sunspot oscillations occupy a ring-shaped wave zone that expands with height and fades in million-degree plasma, so models must explain both the outward drift and the disappearance.","Three-minute oscillations can carry wave power from the umbra into coronal fan structures, with enhanced deposition near loop footpoints.","The umbra-penumbra split, with three-minute power under the umbra and five-minute power at the boundary, constrains where each mode is generated or converted.","The five AIA channels provide a temperature ladder against which future high-resolution observations can test how the ring radius and suppression height vary from sunspot to sunspot."],"supporting_citations":[{"why":"Supplies the red-noise fitting and 95% chi-square significance test used to identify true oscillation peaks in the FFT spectra.","marker":"Vaughan (2005)"},{"why":"Describes the SDO/AIA instrument and its channels, which provide the temperature-stratified image series used here.","marker":"Lemen et al. (2012)"},{"why":"Describes the SDO mission and the data characteristics underlying the 12- and 24-second cadence time series.","marker":"Pesnell et al. (2012)"},{"why":"Prior result that five-minute oscillations propagate along inclined magnetic field lines, which the expanding circle-shape confirms; the paper also shares one sunspot data set with them.","marker":"Kolobov et al. (2016)"},{"why":"Earlier finding that three-minute oscillations leak along coronal fan structures, which the loop-footpoint power maps support.","marker":"Reznikova & Shibasaki (2012)"},{"why":"Provides the contrasting umbra-confined distribution of three-minute oscillation power that this study compares and extends to more channels.","marker":"Kobanov et al. (2013)"},{"why":"States that three-minute oscillations are suppressed in the photosphere and enhanced in the chromosphere, a trend the multi-channel maps bear out.","marker":"Nagashima et al. (2007)"}],"fun_headline_variants":["Five-minute sunspot oscillations fade in hot corona","Sunspot 5-min waves suppressed in high corona","Three-minute sunspot waves cling to loop footpoints","Hot corona erases five-minute sunspot rings","Sunspot oscillations: five-minute fades, three-minute hugs loops"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a per-pixel 95% chi-square confidence test on red-noise-fitted power spectra identifies genuine oscillation modes even though thousands of pixels are tested without multiple-comparison correction.","fun_headline_variants_meta":{"raw":{"variants":["Five-minute sunspot oscillations fade in hot corona","Sunspot 5-min waves suppressed in high corona","Three-minute sunspot waves cling to loop footpoints","Hot corona erases five-minute sunspot rings","Sunspot oscillations: five-minute fades, three-minute hugs loops"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001611,"raw_usage":{"total_tokens":6393,"prompt_tokens":901,"completion_tokens":5492,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":517,"completion_tokens_details":{"reasoning_tokens":5415}},"tokens_in":517,"tokens_out":5492,"duration_ms":38880,"temperature":1.0,"reasoning_tokens":5415,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:28:53.122262+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Count false positives by re-running the same per-pixel test on shuffled light curves or by applying a false-discovery-rate correction; if the high-corona maps still show as many significant pixels with no spatial organization, the claimed five-minute suppression is not established and the disordered maps are plausibly pure noise.","supporting_citations":[{"cited_title":"R., Title, A","cited_arxiv_id":null,"evidence_quote":"Describes the SDO/AIA instrument and its channels, which provide the temperature-stratified image series used here."},{"cited_title":"D., Thompson, B","cited_arxiv_id":null,"evidence_quote":"Describes the SDO mission and the data characteristics underlying the 12- and 24-second cadence time series."},{"cited_title":"Y ., Chelpanov, A","cited_arxiv_id":null,"evidence_quote":"Prior result that five-minute oscillations propagate along inclined magnetic field lines, which the expanding circle-shape confirms; the paper also shares one sunspot data set with them."},{"cited_title":"E., & Shibasaki, K","cited_arxiv_id":null,"evidence_quote":"Earlier finding that three-minute oscillations leak along coronal fan structures, which the loop-footpoint power maps support."},{"cited_title":"I., Chelpanov, A","cited_arxiv_id":null,"evidence_quote":"Provides the contrasting umbra-confined distribution of three-minute oscillation power that this study compares and extends to more channels."},{"cited_title":"G., et al","cited_arxiv_id":null,"evidence_quote":"States that three-minute oscillations are suppressed in the photosphere and enhanced in the chromosphere, a trend the multi-channel maps bear out."}],"review_version":1}