{"id":"73ab07c8-1787-4dcd-9a73-897f81f73786","arxiv_id":"2411.18105","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Over 1989-2019, solar radio emission rotation periods increase from 22.49 days at 8800 MHz to 25.25 days at 245 MHz, indicating the corona rotates slower at higher altitudes.","lead":"Using 31 years of daily solar radio flux at seven frequencies, this paper measures how fast the Sun's corona rotates at different heights and how that changes over time. It finds that lower-frequency (higher-altitude) radio emission rotates more slowly, supporting the idea that the corona's rotation slows with altitude.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Frequency–altitude mapping conflates emission mechanism with height; RSTN bands sample different tracers, so the period–frequency trend may not imply radial differential rotation.","rationale":"The reader's CONDITIONAL verdict is appropriate. I agree with the reader's primary weakest assumption: the frequency–altitude mapping. I sharpen it: because RSTN noon flux at these seven frequencies samples different emission mechanisms (gyroresonance, free–free, noise storms), the period–frequency relation could be a property of the tracers rather than of the height of the emitting plasma. The paper's own discussion of emission mechanisms in Section 4 makes this confound explicit. The EEMD IMF selection and the absence of uncertainties are real but secondary; they affect whether the period differences are statistically significant, whereas the mapping concern affects the interpretation even if the period differences are perfectly significant. The proposed within-mechanism test would empirically separate the two explanations. If that test passes, the paper's conclusion would be strengthened; if it fails, the claim should be weakened to 'rotation periods of different radio-frequency emissions differ' rather than 'coronal rotation slows with altitude.' The reader's identification of the frequency–altitude mapping as the weakest point matches my reading, and the conditional verdict should stand until this confound is addressed.","tokens_in":13814,"tokens_out":5094,"duration_ms":50884,"concrete_test":"Run the same EEMD+wavelet pipeline on the three gyroresonance-dominated channels (2695, 4995, 8800 MHz) and on the three noise-storm-dominated channels (245, 410, 610 MHz) separately. If the monotonic decrease of rotation period with frequency persists within each mechanism group, the tracer-type confound is rejected and the radial-gradient interpretation is supported. If the trend appears only when comparing across groups, the period differences are more plausibly due to differences in the tracers themselves. Additionally, compute bootstrap confidence intervals for the Table 1 global-wavelet peak periods from red-noise surrogates and report whether adjacent-frequency differences exceed joint uncertainties.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that coronal rotation slows with altitude rests on a monotonic frequency–altitude mapping for RSTN noon flux (Section 4). The paper concedes that exact emission altitudes are hard to determine and that the 245 MHz height of ~1.3 R_sun is only a rough estimate. More importantly, the seven RSTN frequencies do not sample a single emission mechanism: 2695–8800 MHz are dominated by gyroresonant emission over sunspots, 1415 MHz includes free–free coronal loops and plage, and 245–610 MHz are dominated by noise-storm continua (Section 4). These tracers have different lifetimes, latitudes, and magnetic anchoring, so their apparent rotation periods can differ even if the local plasma rotates rigidly. The observed monotonic period–frequency trend is therefore not sufficient by itself to establish radial differential rotation; it may reflect a tracer-selection effect rather than a true height gradient. A secondary but related issue is that the EEMD IMF chosen as the solar rotation signal is selected visually, and no error bars accompany the Table 1 periods, so the significance of the 0.3–0.8 day adjacent-frequency differences is untested. The frequency–altitude mapping is the more load-bearing concern because without it the headline conclusion has no physical altitude axis.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes daily RSTN noon solar radio flux at seven frequencies (245–8800 MHz) from 1989 to 2019. Using Ensemble Empirical Mode Decomposition to isolate an IMF associated with the solar rotation cycle and continuous wavelet transform to measure its period, the authors report synodic rotation periods from 27.12 days at 245 MHz down to 23.96 days at 8800 MHz, which convert to sidereal periods of 25.25 down to 22.49 days. They interpret the monotonic period–frequency decrease as evidence that coronal rotation slows with increasing altitude from the low corona to about 1.3 R_sun, and that this radial gradient persists over nearly three solar cycles. They also analyze the time–frequency evolution of the rotation period and find that lower-frequency (higher-altitude) emission consistently shows longer periods than adjacent higher-frequency emission over most of the interval.","tokens_in":14062,"tokens_out":4050,"duration_ms":35986,"significance":"If the result holds, it would provide the longest continuous observational baseline to date for a radial gradient in coronal rotation, using public NOAA RSTN data and standard time–frequency methods. The paper also demonstrates that EEMD can extract a rotation-like signal from highly variable radio flux at 245 MHz and 8800 MHz, where traditional autocorrelation methods have failed. The analysis is largely reproducible in principle: the data are public, the EEMD and wavelet methods are standard, and the comparison with previous results (Vats et al. 2001; Bhatt et al. 2017; Singh et al. 2021) is explicitly discussed. However, the central physical conclusion rests on two assumptions that are not fully quantified: a monotonic frequency–height mapping across different emission mechanisms, and the statistical significance of the period differences.","major_comments":[{"comment":"The central claim that coronal rotation slows with altitude depends on a monotonic mapping from radio frequency to source height, but the seven RSTN bands do not sample a single emission mechanism. Section 4 states that 2695–8800 MHz are dominated by gyroresonant emission over sunspots, 1415 MHz by free–free loops and plage, and 245–610 MHz by noise-storm continua. These tracers have different source heights, lifetimes, latitudes, and magnetic anchoring, so their apparent rotation periods can differ even if the plasma at any given height rotates rigidly. The paper itself concedes that the 245 MHz altitude of ~1.3 R_sun is 'only a rough estimate' and that the true radiation altitude is 'significantly more complex.' The observed monotonic period–frequency trend is therefore not sufficient to establish radial differential rotation without additional assumptions. The authors should either restrict the comparison to frequencies within a single emission mechanism, provide independent imaging-based height estimates for each RSTN frequency, or explicitly reframe the conclusion as a frequency dependence of RSTN rotation periods rather than a radial altitude gradient.","section":"Section 4"},{"comment":"No uncertainties are reported for the seven global rotation periods, and the adjacent-frequency differences driving the conclusion are only 0.3–0.8 days in sidereal units. The global wavelet spectra in Figure 5 have finite peak widths, yet Table 1 lists periods to 0.01 days with no error bars. Additionally, the selection of which EEMD IMF represents the rotation signal is made visually (Section 3.1: 'only one IMF is presented for each radio frequency, which represents the intrinsic periodicity on the timescale of a solar rotation cycle'), and the EEMD parameters (ensemble size 500, white-noise standard deviation 0.2) are stated without sensitivity tests. The paper should quantify the uncertainty of each period (e.g., from wavelet peak width at the stated confidence level or from Monte Carlo resampling of the gap-filled series), provide an objective criterion for selecting the rotation IMF, and test the robustness of the Table 1 trend to EEMD parameter choices. Without these, the significance of the 0.3–0.8-day differences is untested, which is load-bearing for the 'consistently becomes gradually slower with altitude' claim.","section":"§3.1–3.2, Table 1"}],"minor_comments":[{"comment":"The sentence 'daily measurements of the disc-integrated solar radio flux, observed by RSTN in the frequency range of 245 MHz to 15.4 MHz' appears to contain a typo: the upper frequency should likely be 15.4 GHz (15,400 MHz), not 15.4 MHz.","section":"Section 2"},{"comment":"The description of data provenance is confusing: the text states that since 1988 data are derived from four stations and 'have not undergone quality control,' then says 'we only use the SGMR data in this study,' but later the analysis window is 1989–2019. Please clarify whether SGMR data after 1988 are quality controlled and how the SGMR subset is identified from the multi-station record.","section":"Section 2"},{"comment":"The EEMD ensemble size (500) and white-noise amplitude (0.2) are stated without justification or tests of sensitivity. Even if this is a minor issue for a well-established method, a brief discussion of why these choices are appropriate for 11,308-day solar flux series would aid reproducibility.","section":"Section 3.1"},{"comment":"The comparison with Singh et al. (2021) is qualitative ('in agreement with Singh et al. (2021)'). A quantitative statement, such as the average absolute difference in periods over the overlapping 405–2800 MHz range, would be more informative.","section":"Section 4"},{"comment":"There are several spelling errors, including 'centrimetric' for 'centimetric' and 'randomization' in contexts where 'randomness' or 'variability' is meant. These should be corrected during revision.","section":"Section 4"},{"comment":"The caption and text note that values near the start and end of the interval may suffer edge artifacts, but no quantitative indicator (e.g., shading the cone of influence in each panel) is provided. Adding such an indicator would help readers judge which parts of the temporal variation are robust.","section":"Figure 6"}],"recommendation":"major_revision","confidential_remarks":"The manuscript addresses a timely question and uses public data with standard methods, but the central altitude interpretation is vulnerable to the tracer-selection critique. The authors' response to the frequency–height mapping issue will determine whether the paper can be accepted in its current scope; if the mapping concern cannot be resolved, a reframing as a frequency–period trend study would still be publishable but with a weaker claim. The lack of error bars on Table 1 is fixable and should be a mandatory revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper's real novelty is the long baseline: seven RSTN frequencies from 1989 to 2019, nearly three solar cycles, with temporal evolution curves and EEMD extraction of rotation signals at 245 and 8800 MHz. That part deserves credit. The data are public, the methods are standard, and the authors engage directly with the contradictory Vats/Bhatt/Singh results. They also acknowledge edge effects and that the 1.3 solar-radius altitude for 245 MHz is a rough estimate.\n\nThe soft spots are real. The most load-bearing is the frequency-altitude mapping. The seven bands are not the same tracer seen at different heights: 2695-8800 MHz are dominated by gyroresonant emission over sunspots, 1415 MHz includes plage and free-free loops, and 245-610 MHz are noise-storm continua. These tracers have different lifetimes, latitudes, magnetic anchoring, and solar-cycle responses. A monotonic period-frequency trend can emerge from tracer selection even if the local plasma rotates rigidly. The paper itself says the rotation of radio emissions is \"significantly influenced by their dominant features and generation mechanisms,\" which undercuts the later claim that the trend is \"primarily determined by altitude.\"\n\nSecond, the Table 1 periods carry no uncertainties, and the adjacent-frequency differences driving the conclusion are only 0.3 to 0.8 days. Third, the selection of which EEMD IMF represents rotation is visual, with no quantitative selection rule and no sensitivity analysis for ensemble size, white-noise amplitude, the Morlet omega0, or the sliding window. These are not fatal to the empirical description, but they mean the headline conclusion is stated too strongly for the evidence.\n\nA revised version with error bars, a tracer-by-tracer breakdown, and a sensitivity analysis would be convincing. As it stands, this is a credible data product and a useful reference for the Vats/Bhatt/Singh disagreement, but the radial-gradient claim should be treated as suggestive rather than established. Send it to peer review; a serious referee can ask for the missing quantification.","headline":"Useful 30-year extension of the Vats/Bhatt/Singh radio-rotation work, but the radial-gradient conclusion is over-sold because the different frequency bands are different tracers, not just different heights.","tokens_in":14652,"tokens_out":2504,"would_cite":true,"duration_ms":24584,"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":"Daily radio flux from 1989 to 2019 shows the solar corona rotates more slowly with altitude, a radial differential rotation spanning nearly three solar cycles.","keywords":["Sun: rotation","Sun: corona","solar radio emission","radial differential rotation","coronal rotation","ensemble empirical mode decomposition","wavelet analysis","solar activity cycle"],"falsifier":"Track coronal bright points or noise-storm sources at independently measured heights, for example from EUV stereoscopy or interferometric imaging, over the same 1989–2019 interval and measure their rotation periods as a function of height; if the periods do not decrease monotonically with decreasing height, the frequency-based radial gradient would not be a genuine altitude effect.","tokens_in":13618,"feed_emoji":"☀️","tokens_out":9262,"duration_ms":76283,"temperature":0.7,"pith_summary":"This paper tries to settle a long-running dispute about whether the Sun's corona rotates at the same rate at all heights. Using 30 years of daily disc-integrated solar radio flux at seven frequencies from 245 to 8800 MHz, the authors extract the rotation-cycle signal with Ensemble Empirical Mode Decomposition and measure periods with a continuous wavelet transform. They find the sidereal rotation period falls from 25.25 days at 245 MHz to 22.49 days at 8800 MHz, meaning rotation is faster at higher frequencies and, because higher frequencies are emitted lower in the corona, faster at lower altitude. The claimed result is a persistent radial differential rotation from the low corona up to roughly 1.3 solar radii over nearly three solar cycles. If true, it provides a long-sought constraint on how angular momentum is distributed and transported in the solar atmosphere.","feed_headline":"Solar corona spins slower with height, 30-year radio record shows","feed_subtitle":"Radio flux at seven frequencies tracks a persistent radial rotation slowdown from the low corona to 1.3 solar radii.","key_machinery":"The argument rides on two linked pieces. The first is the frequency–height mapping for solar radio emission: centimetric radiation such as 8800 MHz originates in the transition region or low corona, while metric radiation at 245 MHz originates near or above 1.3 solar radii, so each observed frequency samples a different coronal height. The second is the signal-processing pipeline: Ensemble Empirical Mode Decomposition, which adds white-noise realizations and averages to isolate an intrinsic mode function at the roughly 27-day rotation timescale from otherwise noisy daily flux, followed by continuous wavelet analysis with a Morlet mother wavelet ($\\omega_0=12$) that determines both the global dominant period and its time-localized variation. The synodic-to-sidereal conversion $T_{\\rm sidereal} = \\frac{365.26\\,T_{\\rm synodic}}{365.26 + T_{\\rm synodic}}$ finishes the measurement chain.","core_discovery":"The central claim is that the solar corona exhibits a persistent radial differential rotation, with rotation slowing as altitude increases. The paper analyzes daily RSTN disc-integrated radio flux at 245, 410, 610, 1415, 2695, 4995, and 8800 MHz from 1989 to 2019. EEMD yields an intrinsic mode function at the rotation timescale for each frequency, and the global wavelet power spectrum of each IMF gives a dominant synodic period; after converting to sidereal values the periods are 25.25, 24.48, 24.15, 23.63, 23.07, 22.80, and 22.49 days, respectively. Because radio emission at higher frequencies is believed to originate lower in the corona, the monotonic decrease of period with frequency is read as a monotonic increase of rotation speed with decreasing height, from roughly 1.3 solar radii down to the transition region. The authors emphasize that the 245 MHz height is only a rough estimate, but the ordering of source heights across a wide wavelength range is secure. They further find that this radial gradient, with occasional cycle-modulated variations, holds throughout almost the entire 30-year interval.","pith_inferences":["Because the disc-integrated flux averages over all latitudes, part of the measured radial gradient could be contaminated by latitudinal differential rotation; resolved imaging of the same radio sources would separate the two effects.","The same EEMD-plus-wavelet pipeline could be applied to radio data at frequencies outside 245–8800 MHz, extending the inferred altitude range beyond the roughly 1.3 solar radii sampled here.","A persistent decrease of rotation speed with height, if coupled to the frozen-in magnetic field, implies that field lines between layers wind up over time; that winding could provide a measurable energy reservoir for coronal heating, though the paper does not quantify it."],"forward_implications":["The corona does not rotate as a solid body radially: the rotation period changes by about 2.8 days between 245 and 8800 MHz, so angular velocity varies continuously with height in the sampled layer.","Earlier conflicting results are resolved in favor of the studies that found faster rotation at lower altitude, but with a longer baseline and a wider frequency range.","Radio flux at the extreme frequencies, 245 and 8800 MHz, long considered too randomized for rotation studies, can be used to track rotation when EEMD is applied first.","The rotation-period difference between adjacent heights persists over nearly three solar cycles, with only modest solar-cycle modulation, so the radial gradient is a stable structural feature rather than a transient.","The faster-than-surface rotation of the corona inferred from these radio periods is consistent with small-scale magnetic fields, frozen into the plasma, dragging the upper atmosphere around faster than sunspots."],"supporting_citations":[{"why":"Supplies the initial frequency–altitude correspondence for coronal radio emission and the earlier claim that coronal rotation period increases with frequency; the paper's starting point and first contrast.","marker":"Vats et al. (2001)"},{"why":"Previous analysis of the same 1997–1999 radio data that found rotation period decreasing with frequency; the present result verifies this direction.","marker":"Bhatt et al. (2017)"},{"why":"Extended frequency-range study of 1994–1999 radio data concluding coronal rotation slows as altitude increases; the sidereal periods reported here are compared directly with its values.","marker":"Singh et al. (2021)"},{"why":"Electron density model used to estimate the 245 MHz source altitude at about 1.3 solar radii, anchoring the height scale.","marker":"Aschwanden and Benz (1995)"},{"why":"Observations of noise-storm altitudes at 432 and 150 MHz, about 1.20 and 1.35 solar radii, used to support the metric-wavelength altitude calibration.","marker":"Mercier et al. (2015)"},{"why":"Provides the radio emission mechanisms and source-region altitudes connecting each observed frequency to a coronal height.","marker":"Shibasaki et al. (2011)"},{"why":"Proposes the Ensemble Empirical Mode Decomposition used to isolate the rotation-cycle signal from noisy radio flux.","marker":"Wu and Huang (2009)"},{"why":"Supplies the continuous wavelet transform methodology and significance testing used to measure the rotation periods.","marker":"Torrence & Compo (1998)"},{"why":"Provides the wavelet-analysis code used for the continuous wavelet power spectra and global power spectra.","marker":"Grinsted et al. (2004)"},{"why":"One of the sources for the synodic-to-sidereal conversion formula used to turn observed periods into actual rotation periods.","marker":"Chandra and Vats (2011)"}],"fun_headline_variants":["Corona's spin slows with altitude, 30-year radio data confirm","Radio waves reveal corona's slow-spin gradient with height","Solar corona rotates slower higher up, 30-year study finds","Coronal rotation rate drops with height, radio record shows","Corona spins slower as altitude rises, radio data prove"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that higher radio frequencies really do originate lower in the corona and lower frequencies higher up, so that the monotonic frequency trend can be read as a height trend; the paper itself cautions that the 245 MHz height is only a rough estimate and true emission altitudes are more complex.","fun_headline_variants_meta":{"raw":{"variants":["Corona's spin slows with altitude, 30-year radio data confirm","Radio waves reveal corona's slow-spin gradient with height","Solar corona rotates slower higher up, 30-year study finds","Coronal rotation rate drops with height, radio record shows","Corona spins slower as altitude rises, radio data prove"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000215,"raw_usage":{"total_tokens":1502,"prompt_tokens":1093,"completion_tokens":409,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":709,"completion_tokens_details":{"reasoning_tokens":324}},"tokens_in":709,"tokens_out":409,"duration_ms":3670,"temperature":1.0,"reasoning_tokens":324,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:30:23.870708+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Track coronal bright points or noise-storm sources at independently measured heights, for example from EUV stereoscopy or interferometric imaging, over the same 1989–2019 interval and measure their rotation periods as a function of height; if the periods do not decrease monotonically with decreasing height, the frequency-based radial gradient would not be a genuine altitude effect.","supporting_citations":[],"review_version":1}