{"id":"b3f36480-469e-4565-b41c-b9493695a26e","arxiv_id":"2506.07493","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"The dominant oscillation period in the quiet solar atmosphere decreases steadily from about 4.5 minutes in the photosphere to 2.8 minutes in the chromosphere, and two-fluid simulations reproduce the trend.","lead":"Using IRIS spectra of seven solar lines, the authors find that the dominant wave period shrinks from about 272 seconds in the lower photosphere to about 167 seconds in the chromosphere. The result gives direct evidence that the solar wave spectrum is filtered with height, a process tied to chromospheric heating.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Claimed period decrease is not quantitatively resolved: adjacent Table 2 periods differ by only 6–50 s, near the wavelet/frequency resolution of a 34-min series, and the Mg II height tables disagree.","rationale":"The paper's direction is probably right, but the supporting evidence for the quantitative height variation is not secure. The method uses a 34-min series; at 200 s that is roughly ten cycles, and the wavelet scale grid gives period bins about 18–24 s apart, so several Table 2 differences are below one bin. No standard errors are reported, and the internal height-table inconsistency makes the slope ambiguous. A synthetic recovery test directly settles whether the pipeline can resolve the claimed differences. I am not proposing rejection because the broad photospheric-to-chromospheric decrease is independently expected and the simulation broadly supports it; however, the paper should remain conditional on demonstrating this resolution.","tokens_in":12222,"tokens_out":15051,"duration_ms":175723,"concrete_test":"Run a recovery test on synthetic data: generate 34-min, 17-s-cadence Doppler series with known periods equal to the Table 2 sequence (or adjacent pairs like 240.5 s and 227.9 s), with amplitude and red-noise levels matched to the observed detrended DTS; apply the same wavelet global-spectrum peak finding and histogram/Gaussian fitting used in Section 3, and check whether the recovered mean periods are correctly ordered and separated by more than the bootstrap 95% uncertainty. If the pipeline cannot recover 6–13 s differences, the Table 2 trend is an artifact of frequency resolution. As a complementary check, split the real 34-min series into two 17-min halves and compare the recovered period sequences.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires the mean dominant periods in Table 2 to be ordered by formation height. But the period differences between adjacent heights are small: 6.4 s (234.10 vs 240.50), 5.3 s (227.90 vs 233.20), and 12.6 s (233.20 vs 204.60) are all less than or comparable to the wavelet scale spacing of a Morlet wavelet at P≈200–240 s (about 18–24 s with dj=0.125) and to the Fourier resolution 1/T≈0.5 mHz (≈25 s at 200 s) of the 34-min series. The paper reports only the histogram Gaussian sigma (23–54 s), not the standard error of the mean or any confidence interval; with per-location periods selected from a discrete wavelet grid, the histogram centers may not resolve the true inter-height differences. The numerical comparison in Fig. 5 is qualitative and cannot validate this. Compounding this, Table 1 and Table 2 give different Mg II formation heights (k2v/k2r swapped; k3 = 2.20 Mm vs 1.80 Mm), so the height axis of the trend is itself uncertain. The direction of the trend is probably correct—it is consistent with known 5-min photospheric and 3-min chromospheric oscillations—but the monotonic decrease and the reported slope of -53.36 s/Mm are not yet established at the stated precision.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports IRIS sit-and-stare observations of six photospheric absorption lines and the Mg II h&k line (k2v, k2r, k3 features) to derive Doppler velocity time series at nine heights from 0.17 Mm to 2.2 Mm. Wavelet analysis yields global power spectra and dominant periods at each height; histograms of the dominant period from ~700 spatial locations are fit with Gaussians to obtain mean periods. The authors find that the mean dominant period decreases monotonically with height from ~272 s at 0.17 Mm to ~167 s at 1.80 Mm, with a linear fit of slope -53.36 s/Mm and a claimed Pearson coefficient of 0.944. They supplement this with 2.5-D two-fluid numerical simulations using the JOANNA code and report qualitative agreement between the simulated filtered wave spectra and the observed periods. The central claim is that these observations demonstrate, for the first time, the height variation of the dominant wave period in the quiet Sun.","tokens_in":12500,"tokens_out":3844,"duration_ms":41817,"significance":"If the height ordering and the monotonic decrease are quantitatively established, this result would directly confirm theoretical predictions of wave-spectrum filtering in the solar atmosphere, linking the photospheric 5-min oscillations to the chromospheric 3-min oscillations. The paper combines a clever use of IRIS NUV diagnostics with state-of-the-art two-fluid simulations, and the observational data are independent of the simulations. The direction of the trend is broadly consistent with existing knowledge, but the stated precision (a linear slope of -53.36 s/Mm, R=0.944) is not yet supported by the presented analysis, primarily because the period differences between adjacent heights are near the resolution limits and because key statistical and tabular quantities are inconsistent.","major_comments":[{"comment":"The reported Pearson correlation coefficient is inconsistent: the text states 0.944, while Figure 3(d) shows R = -0.917. Since the period is expected to decrease with height, the sign in the figure is plausible, but the magnitudes disagree, and no uncertainty is given for either the coefficient or the fit parameters (slope -53.36 s/Mm, intercept 262.99). Please reconcile these values and provide error estimates, for example by bootstrapping the per-location dominant periods or by propagating the Gaussian-fit uncertainties of the histogram means.","section":"Sec. 3, Fig. 3(d) and text"},{"comment":"The formation heights of the Mg II features are inconsistent between Table 1 and Table 2. Table 1 lists Mg II k2v at 1.20 Mm, k2r at 1.55 Mm, and k3 at 2.20 Mm, while Table 2 lists k2r at 1.2 Mm, k2v at 1.55 Mm, and k3 at 1.80 Mm. The text in Section 2 likewise appears to swap k2v and k2r. Because the ordering of heights is the basis for the claimed monotonic decrease, these discrepancies must be resolved and the adopted heights justified from the cited literature (Vernazza et al. 1981; Leenaarts et al. 2013).","section":"Tables 1 and 2"},{"comment":"The differences in mean dominant period between adjacent heights (e.g., 6.4 s between Fe I 2792.327 and Fe I 2793.223; 5.3 s between Ni I 2799.347 and Fe I 2814.114; 12.6 s between the latter and Mn I 2801.907) are smaller than the wavelet scale spacing of a Morlet wavelet at periods 200–240 s with dj=0.125 (roughly 18–24 s) and are comparable to the Fourier resolution of the 34-minute time series (1/T ≈ 0.5 mHz, i.e., ≈25 s at 200 s). The paper reports only the Gaussian standard deviations (23–54 s) of the histograms, not the standard error of the mean or confidence intervals. Please demonstrate that the inter-height differences are statistically significant (e.g., via bootstrap or Monte Carlo) and explicitly address the wavelet/frequency resolution limits when claiming a monotonic decrease.","section":"Sec. 3, Table 2 and Fig. 3(c)"},{"comment":"The comparison between the observed dominant periods (dots) and the simulated Fourier power spectrum is described as “good agreement,” but no quantitative metric is provided. The simulation exhibits broad period bands (P≈300 s for y<0.5 Mm and 200<P<250 s higher up), while the observed dots are discrete; it is unclear how the dots are assigned to the simulation features and what tolerance is acceptable. Please define the comparison criterion (e.g., nearest-band assignment or chi-square statistic) and provide quantitative residuals between the observed periods and the simulation-derived periods at the corresponding heights.","section":"Sec. 4, Fig. 5"}],"minor_comments":[{"comment":"The abstract contains a duplicated word: “well well-established.”","section":"Abstract"},{"comment":"Several references are duplicated in the bibliography (e.g., Kayshap et al. 2018, Leenaarts et al. 2013, Wiśniewska et al. 2016). Please consolidate the list.","section":"References"},{"comment":"The caption states “Time-distance plots for horizontally Viy”; this appears incomplete and should read “horizontally averaged Viy.”","section":"Fig. 5 caption"},{"comment":"The table formatting has issues: the last row is numbered “7” again, the Mg II entries use “..” for missing uncertainties, and the text says seven spectral lines are used while the table lists nine features. Please clarify the numbering and the definition of “lines” vs. “features.”","section":"Table 1"},{"comment":"The sentence “Mgiik2r (Mgiik2r) form at a height of 1.2 Mm (1.55 Mm)” appears to have a typo in the parentheses; it likely should read “k2v (k2r).”","section":"Sec. 2"},{"comment":"The symbol “sit-n-stare” is nonstandard; the usual term is “sit-and-stare.”","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The core idea is attractive and the observational data are independent of the simulations, which is a strength. However, the quantitative claim of a monotonic decrease and a specific linear slope currently rests on statistically unresolved period differences and inconsistent height tables. These issues are fixable in a revision, so I do not recommend rejection. I also note a heavy self-citation pattern among the authors, but that does not by itself affect the scientific validity; the main need is a rigorous re-analysis with proper uncertainties."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I read the paper with care. The qualitative result is almost certainly correct: the dominant wave period in the quiet Sun drops from about 272 s in the photosphere to about 167 s in the chromosphere. That fits a well-established picture of 5-min and 3-min oscillations, and the paper's height-resolved IRIS data set is a useful new contribution. The authors used seven spectral lines (nine features from Mg II) to sample different heights, applied standard wavelet analysis, and compared with two-fluid simulations. The observational data are independent of the simulations, so there is no circularity there. The histograms across 700 slit positions are a reasonable way to get a mean period per line; I think the method is sound in spirit.\n\nThe soft spots are quantitative. The differences between adjacent photospheric periods are 5–13 s. For a 34-minute series and a Morlet wavelet with dj=0.125, the period resolution at these periods is roughly 18–24 s, so those adjacent values are not resolved. The Gaussian sigma values in Table 2 are 23–54 s, which also suggests the means are not statistically distinguishable. Yet the paper fits a linear trend with slope -53.36 s/Mm and reports a Pearson coefficient. That slope and the monotonic ordering of the means are not established at the stated precision. The coefficient is also inconsistent: the text says 0.944, the figure says -0.917, and a positive value makes no sense for a decreasing trend. Table 1 and Table 2 disagree on the Mg II formation heights (k2v and k2r are swapped, and k3 changes from 2.2 to 1.8 Mm), which undermines the height axis. The comparison with the JOANNA simulations in Figure 5 is qualitative; plotting the observed points on the simulation's power spectrum is not a validation. There are also minor copyediting issues, like duplicate row numbering in Table 1.\n\nThese are fixable. Add standard errors on the mean periods, show whether adjacent heights are actually resolved, correct the tables, and either provide error bars on the fit or soften the slope claim. The direction of the trend is not in doubt; the rate is.\n\nThis paper deserves peer review, not a desk reject. A referee should focus on the resolution analysis and the internal consistency. I would bring it to a reading group as an example of how wavelet resolution sets limits on period-height gradients. I would not cite it as a precise measurement until the revisions are made.\n\nRecommendation: send to review, request major revision.","headline":"The qualitative period decrease with height is almost certainly real, but the quantitative slope is not resolved by the data; fix the error analysis and the table inconsistencies before citing the gradient.","tokens_in":13050,"tokens_out":3653,"would_cite":false,"duration_ms":44139,"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":"Observed dominant period of solar oscillations decreases with height, from 272 seconds in the photosphere to 167 seconds in the chromosphere, matching two-fluid simulations.","keywords":["quiet Sun","solar oscillations","dominant wave period","photosphere","chromosphere","IRIS spectroscopy","wavelet analysis","two-fluid simulations"],"falsifier":"Re-observe the same quiet-Sun region with a multi-hour IRIS sit-and-stare sequence, recompute the global wavelet spectra for the same nine spectral features, and check whether the 20-50 s differences in dominant period between adjacent formation heights still exceed the width of the spectral peaks; if the differences fall inside the peaks, the claimed monotonic decrease is not resolved. A second, complementary check is to use lines with independently measured formation heights to see whether the same period-height slope emerges.","tokens_in":12012,"feed_emoji":"☀️","tokens_out":11987,"duration_ms":120079,"temperature":0.7,"pith_summary":"This paper tries to establish that solar wave oscillations do not keep a fixed dominant period as they travel upward: the dominant period falls from roughly 272 seconds in the low photosphere to about 167 seconds in the chromosphere. The evidence comes from IRIS spectra of six photospheric absorption lines and the Mg ii h&k chromospheric line, which yield Doppler-velocity time series at nine distinct formation heights. Wavelet analysis of those time series gives a dominant period at each height, and the nine values fall on a decreasing line with a Pearson correlation of -0.917. A two-fluid numerical simulation of wave propagation through the solar atmosphere reproduces the same height-dependent shortening. If correct, the result turns the broad textbook transition from 5-minute photospheric oscillations to 3-minute chromospheric oscillations into a continuously measured filtering curve, offering a direct test of wave-filtering theory on the Sun.","feed_headline":"Solar wave period shrinks from 272 to 167 seconds with height","feed_subtitle":"IRIS data show the dominant wave period falls from photosphere to chromosphere; two-fluid simulations reproduce the drop.","key_machinery":"The central machinery is a height ladder built from seven spectral lines with nine formation levels: six photospheric absorption lines (Ni i, Fe i, Mn i) between 0.17 and 0.83 Mm, plus the optically thick Mg ii h&k chromospheric line, whose k2v, k2r, and k3 features form at chromospheric heights. Doppler velocities from each feature, measured at about 700 positions along the IRIS slit over a 34-minute sequence, are detrended and transformed with Morlet wavelet analysis; the peak of the global wavelet spectrum at each height, gathered into a histogram and fitted with a Gaussian, defines the dominant period. On the simulation side, the paper uses a two-fluid ion-neutral numerical model with ionization and recombination, initiated with a semi-empirical temperature profile and a 5 G vertical plus 0.5 G transverse magnetic field, to compute Fourier power spectra of the vertical velocity; the positions of the spectral peaks as a function of height are the simulated counterpart to the observed dominant-period ladder.","core_discovery":"On its own terms, the paper claims that the wave spectrum is filtered continuously through the quiet-Sun atmosphere so that the dominant period decreases monotonically with height. The authors derive mean dominant periods of 272 s at 0.17 Mm, 234-240 s in the lower photosphere, 228-233 s in the mid-photosphere, 205 s at 0.83 Mm, and 184 s, 159 s, and 167 s for the three Mg ii h&k features (the two k2 peaks and the central k3 dip). A linear fit to mean period versus formation height gives a slope of -53.36 s/Mm with a Pearson coefficient of -0.917. The paper presents this as the first direct measurement of height-dependent dominant wave periods in the photosphere and chromosphere, and shows that two-fluid simulations including ionization and recombination reproduce the observed periods at the corresponding heights.","pith_inferences":["An untested extension is whether the same monotonic period decrease appears in network and plage regions; the present quiet-Sun result may not carry over where the magnetic field shapes the cutoff frequencies.","If the formation heights carry systematic errors, the measured period-height curve could be inverted to constrain those heights, making the trend a diagnostic rather than a conclusion.","A longer time series would reveal whether the decrease continues into the upper chromosphere and transition region or flattens near the acoustic cutoff, where the k3 feature already shows a slight uptick from 159 s to 167 s.","The two-fluid simulation's success suggests that synthetic IRIS line profiles could be generated from the same model and compared directly to observed Doppler shifts, turning the period-height comparison into a full line-profile validation."],"forward_implications":["The textbook 5-minute photospheric and 3-minute chromospheric oscillations are linked by a continuous, roughly linear drop in dominant period, not by a sharp transition at a single height.","The narrowing of the period distributions (Gaussian sigma from about 49-54 s in the mid-photosphere to 24-28 s for the upper Mg ii features) implies that the atmosphere acts as a bandpass filter that sharpens the wave spectrum as it propagates upward.","In the simulations, the 300 s waves are evanescent above the photosphere while shorter-period waves propagate into the chromosphere and low corona, identifying acoustic cutoff filtering as the physical mechanism behind the observed shortening.","Because the same filtering behavior was predicted in solar-type stars, the solar measurements provide a nearby height-resolved case that can anchor similar inferences for stellar chromospheres."],"supporting_citations":[{"why":"Supplies the IRIS data set and observing mode from which all Doppler-velocity time series are extracted.","marker":"De Pontieu et al. 2014"},{"why":"Provides the formation heights of the Mg ii h&k features and the model tabulations behind the nine-height ladder.","marker":"Leenaarts et al. 2013"},{"why":"Defines the extremum-finding routine used to convert Mg ii h&k profiles into k2v, k2r, and k3 Doppler velocities.","marker":"Pereira et al. 2013"},{"why":"Supplies the wavelet transform and global-spectrum method used to identify dominant periods.","marker":"Torrence and Compo 1998"},{"why":"Provides the semi-empirical temperature model used to initialize the numerical atmosphere.","marker":"Avrett and Loeser 2008"},{"why":"Describes the two-fluid numerical model and boundary setup whose simulated wave spectra are compared with the observations.","marker":"Murawski et al. 2022"},{"why":"Documents the two-fluid code used for the simulations and its treatment of waves in a partially ionized atmosphere.","marker":"Wójcik et al. 2020"},{"why":"Predicted that dominant periods decrease from photosphere to chromosphere in solar-type stars, the theoretical expectation this paper tests on the Sun.","marker":"Fawzy and Musielak 2016"}],"fun_headline_variants":["Solar wave period drops from 272 to 167 s with height","IRIS data: solar wave period falls with atmospheric height","Height shrinks solar wave period: 272 s to 167 s","Solar wave periods decrease with altitude, simulations match"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result stands on the model-based formation heights assigned to the seven spectral lines and on the 34-minute time series being long enough for the wavelet spectra to separate period differences of only 20-50 seconds between neighboring heights; if either assumption is wrong, the decreasing trend could be an artifact.","fun_headline_variants_meta":{"raw":{"variants":["Solar wave period drops from 272 to 167 s with height","IRIS data: solar wave period falls with atmospheric height","Height shrinks solar wave period: 272 s to 167 s","Solar wave periods decrease with altitude, simulations match"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000171,"raw_usage":{"total_tokens":1263,"prompt_tokens":930,"completion_tokens":333,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":546,"completion_tokens_details":{"reasoning_tokens":263}},"tokens_in":546,"tokens_out":333,"duration_ms":4416,"temperature":1.0,"reasoning_tokens":263,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:32:20.339008+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-observe the same quiet-Sun region with a multi-hour IRIS sit-and-stare sequence, recompute the global wavelet spectra for the same nine spectral features, and check whether the 20-50 s differences in dominant period between adjacent formation heights still exceed the width of the spectral peaks; if the differences fall inside the peaks, the claimed monotonic decrease is not resolved. A second, complementary check is to use lines with independently measured formation heights to see whether the same period-height slope emerges.","supporting_citations":[{"cited_title":"E., & Musielak, Z","cited_arxiv_id":null,"evidence_quote":"Predicted that dominant periods decrease from photosphere to chromosphere in solar-type stars, the theoretical expectation this paper tests on the Sun."}],"review_version":1}