{"id":"4e90bc72-d6e9-41d0-85b7-228afa30f471","arxiv_id":"1908.11021","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Analysis of 54 years of OMNI2 data finds that high-speed solar wind rotates faster than low-speed wind, and the yearly rotation rate is negatively correlated with sunspot number at a 3-year lead.","lead":"This paper uses 54 years of hourly solar wind velocity data to study how the solar wind's rotation period changes with wind speed. It reports that faster solar wind rotates faster overall, with opposing trends within the high- and low-speed classes, and links the yearly rotation rate to sunspot number.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Uneven sampling and velocity binning may bias auto-correlation rotation periods; the central velocity-dependence claim needs a controlled check on the nearly complete post-1994 data.","rationale":"The reader identified the uneven sampling as the weakest assumption, and I agree that it is load-bearing. I extend the concern by noting that the velocity binning and transient handling are also undocumented in the available text, and that the non-monotonic velocity dependence could be an artifact of bin edges. The available manuscript text ends during Section 2.1, so the methods for handling missing data, computing yearly rotation rates, and testing statistical significance are not visible. The correct verdict remains UNVERDICTED because the paper may be sound but cannot be verified from the provided material. My agreement is partial because the reader focused on missing-data bias while I additionally flag the velocity-bin dependence as part of the same central-claim risk. The proposed concrete test targets both issues with a single reproducible check on the post-1994 data.","tokens_in":36,"tokens_out":2113,"duration_ms":81197,"concrete_test":"Recompute the auto-correlation rotation periods using only the nearly continuous post-1994 hourly OMNI2 data, stratified into the same velocity bins, and compare the inferred rotation periods with those from the full series. If the 'higher velocity rotates faster' relation disappears or reverses in the post-1994 subset, the headline claim is an artifact of pre-1994 gaps. As a cross-check, compute a Lomb-Scargle periodogram on the same bins to verify the peak positions without the missing-data assumptions inherent in lagged auto-correlation.","verdict_should_be":"UNVERDICTED","load_bearing_attack":"The central claim is that solar wind rotation period depends systematically on velocity, with higher-velocity wind rotating faster overall. Section 2.1 states the hourly OMNI2 series covers only 74.8% of possible records, with frequent gaps before 1994. Because the analysis uses auto-correlation on this uneven series, the standard lagged-product estimator uses only the pairs present at each lag, so the effective lag distribution is nonuniform and time-varying. Gaps that correlate with solar activity can shift the position of the 27-day peak or create the 'wave packet' of nearby peaks, directly affecting the inferred rotation period and its yearly values. The paper does not state whether missing values were interpolated, whether the auto-correlation was computed on the raw uneven series, or how the velocity stratification (v<450, 450-725, ≥725 km/s) and transient streams were treated. The abstract's non-monotonic statement, that rotation rate increases with velocity for high-velocity wind but decreases with velocity for low-velocity wind, could depend sensitively on the chosen bin edges and on the handling of transient streams. Without these details, the velocity-dependence result and the 3-year lead-lag with sunspot number are not adequately shielded from sampling artifacts.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes 54 years of hourly OMNI2 solar wind velocity data (27 Nov 1963 to 31 Dec 2017) using autocorrelation analysis to infer solar wind rotation periods. The central claims are: (1) high-velocity solar wind rotates faster than low-velocity wind; (2) for high-velocity wind the rotation rate increases with increasing velocity, whereas for low-velocity wind it decreases with increasing velocity; (3) for the entire solar wind, higher-velocity wind statistically has a faster rotation rate; and (4) the yearly rotation rate of solar wind velocity does not follow the Schwabe cycle but is significantly negatively correlated with yearly sunspot number when it leads by 3 years. Physical explanations for these findings are also proposed.","tokens_in":71825,"tokens_out":4635,"duration_ms":46723,"significance":"If the empirical claims are substantiated, the velocity dependence of solar wind rotation would provide a new observational constraint on coronal source structure and solar wind mapping, and the 3-year lead-lag with sunspot number would be relevant to solar cycle coupling studies. The paper has clear strengths: it uses the long, relatively homogeneous OMNI2 hourly record and applies a simple, direct autocorrelation approach, and the claims are falsifiable in principle. However, the manuscript as presented does not provide the statistical backbone needed to support these claims: there are no error bars on the reported rotation periods, no significance tests for the correlations, no details of how rotation periods are extracted from autocorrelation peaks, and no sensitivity analysis for the velocity binning. No code or extracted peak tables are provided, so the analysis cannot be independently reproduced from the manuscript.","major_comments":[{"comment":"The data completeness and sampling should be addressed directly. The manuscript reports that only 74.8% of the 474204 hourly records are present, and that before 1994 records are frequently missing. It does not state whether missing values were interpolated, how the autocorrelation was computed on the uneven series, or what minimum number of pairs was required at each lag. The standard lagged-product autocorrelation estimator uses only the pairs available at each lag, so the effective sample changes with lag and with solar cycle phase; this can shift the positions of autocorrelation peaks and distort the 'wave packet' structure from which rotation periods are read. The central velocity-dependence claim therefore needs a controlled check: repeat the analysis on the nearly complete post-1994 subset, and on unevenly resampled synthetic data with known periods, and show that the inferred rotation periods are stable.","section":"Section 2.1"},{"comment":"The claimed non-monotonic velocity dependence is the paper's main empirical result, but the manuscript does not report the extracted rotation period (or rotation rate) for each velocity bin, the associated uncertainties, or the exact bin edges used in the analysis beyond the Section 1 categories v<450, 450≤v<725, and v≥725 km/s. Shifting these boundaries, or treating the v≥725 km/s class differently as transient streams, could change the sign of the within-bin slopes. Please provide a table of the extracted rotation period for each velocity bin with confidence intervals, and a sensitivity analysis in which the bin edges are varied and transient intervals are removed.","section":"Section 1 and abstract"},{"comment":"The claim that the yearly rotation rate is 'significantly negatively correlated' with yearly sunspot number when it leads by 3 years is reported without a correlation coefficient, a p-value, or any statement of how serial correlation in the yearly solar wind and sunspot series was handled. With roughly 54 annual points and a 3-year lead applied, the effective number of independent samples is small, so naively computed significance may be overstated. The authors should show the full cross-correlation function over a range of lags, state the direction of the lead explicitly, and report significance levels that account for autocorrelation in both series.","section":"Abstract and yearly correlation analysis"}],"minor_comments":[{"comment":"The solar cycle is consistently misspelled as 'Schwable cycle'; it should be 'Schwabe cycle'.","section":"Abstract and text"},{"comment":"The affiliations contain stray spaces: 'Y unnan Observatories' and 'Unive rsity' should be corrected to 'Yunnan Observatories' and 'University'.","section":"Author affiliations"},{"comment":"The unit is written as 'kms−1' in several places; the standard form 'km s−1' should be used consistently.","section":"Units and notation"},{"comment":"Figure 1 is nearly illegible because the '+' markers are heavily overplotted in the dense post-1994 portion. The authors should consider showing a binned or decimated version, or a separate inset, to make the full 54-year time series readable.","section":"Figure 1"},{"comment":"The term 'wave packet' is used without a precise definition of how a set of nearby autocorrelation peaks is converted into a single 'rotation period'; this should be defined explicitly in the methods discussion.","section":"Section 2.1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript addresses an interesting observational question and uses a long dataset, but the two headline claims rest on an autocorrelation analysis of a substantially gappy time series. The absence of any description of missing-data handling, peak extraction, uncertainty quantification, or significance testing makes it impossible to audit the central results in the current form. I believe the claims are potentially fixable by adding the requested control analyses and statistical details, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Best thing to know: this is a short empirical paper claiming that solar wind rotation rate depends on flow speed—fast wind rotates faster overall, but within the low-velocity class the trend reverses—and that yearly rotation rate is negatively correlated with sunspot number at a 3-year lead. If true, that is a genuinely new characterization, extending the known ~27.5-day and harmonic periods in solar wind speed.\n\nWhat it does well: the dataset is appropriate (54 years of hourly OMNI2), the velocity stratification follows conventional categories, and the introduction is properly anchored in the older rotation literature. The lead-lag with solar activity is the kind of correlation that could motivate physical follow-up, and the authors say they propose explanations.\n\nWhere it gets soft: the visible text ends after the data description—no results, no error bars, no significance tests, no details on how the auto-correlation peaks were converted to rotation periods, and no statement on how the 74.8%-complete, gap-heavy series was handled. The stress-test concern is fair: the lagged-product estimator on an uneven series produces a nonuniform lag distribution, and gaps that correlate with solar activity can shift the 27-day peak or create the 'wave packet' of nearby peaks. The velocity-binned differences (v<450, 450–725, ≥725 km/s) could depend on bin edges and on how transient streams are treated. Without a controlled check on the nearly complete post-1994 data, or an explicit missing-data treatment, the central claim is not shielded from sampling artifacts. The 3-year lead-lag is also the kind of result that needs a multiple-comparison sanity check.\n\nBottom line: the paper is not obviously wrong, and the authors are not being sloppy in the visible text—they flag the unevenness themselves. But the central claim is under-specified in what I can see. The right move is to send to peer review with a demand for the full methods, error bars, and a post-1994 robustness check. If the methods section already exists in the complete manuscript and I just cannot see it, many of these concerns may dissolve.\n\nRecommendation: send it out. It is a modest but legitimate observational question, and a good referee can sort out the sampling issue quickly.","headline":"A plausible but under-verified claim that solar wind rotation rate depends on flow speed, from a 54-year OMNI2 auto-correlation analysis; the visible text lacks methods and safeguards against gappy-sampling bias.","tokens_in":72301,"tokens_out":2565,"would_cite":false,"duration_ms":28012,"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":"Fast solar wind rotates faster than slow solar wind","keywords":["solar wind","rotation period","auto-correlation","solar wind velocity","sunspot number","Schwabe cycle","coronal holes","OMNI2 data"],"falsifier":"Compute the autocorrelation rotation period separately for the nearly complete post-1994 hourly data and for the sparse pre-1994 data; if the fast-wind rotation advantage does not appear in both subsets, the full-sample result is not robust. A complementary check is to generate synthetic time series with one fixed rotation period, impose the actual 74.8% sampling pattern, and see whether inferred periods shift with simulated velocity.","tokens_in":71411,"feed_emoji":"☀️","tokens_out":8428,"duration_ms":78250,"temperature":0.7,"pith_summary":"Using 54 years of hourly solar-wind velocity measurements, this paper asks whether the solar wind rotates as a single rigid pattern or at a rate that depends on speed. Autocorrelation analysis of the 1963–2017 record shows that faster wind rotates faster: high-velocity wind has a shorter inferred rotation period than low-velocity wind, and within the fast population the rotation rate rises with velocity, while within the slow population it falls as velocity rises. The paper also reports that the yearly rotation rate does not track the sunspot cycle, but is significantly negatively correlated with sunspot number when the rotation rate leads by three years. If these results hold, solar-wind rotation is a speed-dependent, activity-delayed property rather than a fixed 27-day recurrence.","feed_headline":"Fast solar wind rotates faster than slow solar wind","feed_subtitle":"54 years of hourly data show rotation rate rises with wind speed and leads sunspots by 3 years.","key_machinery":"The central object is the autocorrelation function of the hourly solar-wind velocity series, which measures how well the series matches itself at increasing time lags; its peaks mark periodicities. The rotation period is read from the position of the main peak near the expected ~27.5-day rotation, alongside harmonic peaks near 13.7 and 9.1 days. A 'wave packet' of closely spaced significant peaks around the main period is attributed to several impulse streams that persist for a varying number of solar rotations, and it is the shift of these peaks with wind speed that carries the argument for velocity-dependent rotation.","core_discovery":"The central discovery is that the rotation period of the solar wind, extracted from autocorrelation peaks in hourly velocity, is not a single value. Higher-velocity wind statistically rotates faster than lower-velocity wind; the rotation rate increases with speed among high-speed wind and decreases with speed among low-speed wind. The paper further claims that the yearly rotation rate is not a simple Schwabe-cycle tracer: it is significantly negatively correlated with yearly sunspot number, with the rotation rate leading by three years. Physical explanations are proposed for these findings.","pith_inferences":["A testable extension is to split the 54-year record by source type rather than speed alone: if the velocity trend disappears within coronal-hole wind or within streamer wind, the effect is source-dependent rather than intrinsically speed-dependent.","If the 3-year lead is physical, the same delayed negative correlation should appear between sunspot number and other large-scale wind parameters, such as coronal-hole area or open magnetic flux; that would be an independent check.","The opposite slopes for fast and slow wind suggest two populations with different origins; binning by composition tracers such as helium abundance or charge state could separate the regimes more cleanly than velocity bins alone.","A direct sampling check is to run the same autocorrelation pipeline on synthetic data with a known single rotation period but the real 74.8% observation pattern; any apparent speed-dependent shift would indict the gaps rather than the Sun."],"forward_implications":["Reported solar-wind rotation periods should be quoted with the velocity range that produced them, because the autocorrelation peak shifts with speed.","Predictions of high-speed stream recurrence that assume a fixed 27-day rotation will misplace coronal-hole source longitudes by an amount that grows with wind speed.","The 3-year negative lead-lag between yearly rotation rate and sunspot number, if real, gives a testable time offset connecting solar activity to the large-scale rotation of the wind.","Future autocorrelation studies need to separate fast and slow wind before estimating a single rotation period, since the two populations have opposite velocity trends."],"supporting_citations":[{"why":"Describes the OMNI data set from which the 54-year hourly solar-wind velocity series is taken.","marker":"King & Papitashvili 2005"},{"why":"Establishes the ~27.5-day rotation period in solar wind speed that this paper measures through autocorrelation.","marker":"Svalgaard & Wilcox 1975"},{"why":"Supplies earlier determination of the rotation period and its harmonic periods in solar wind speed.","marker":"Clua de Gonzalez et al. 1993"},{"why":"Provides further measurements of the 27-day and harmonic periodicities used as comparison baselines.","marker":"Nayar et al. 2001"},{"why":"Documents the rotation and harmonic peaks and the wave-packet structure near the main period.","marker":"Katsavrias, Preka-Papadema & Moussas 2012"},{"why":"Attributes the 13.7-day harmonic to two streams per rotation and explains the wave-packet peaks by impulse streams.","marker":"Fenimore et al. 1978"},{"why":"Supports the two-stream interpretation of the 1/2 harmonic period in solar wind speed.","marker":"Mursula & Zieger 1996"},{"why":"Recent analysis of the solar-wind rotation wave packet that this paper extends by examining velocity dependence.","marker":"Li, Zhang, & Feng 2017"}],"fun_headline_variants":["Solar wind rotation rate scales with stream velocity","Fast wind spins faster: solar wind rotation study","Solar wind rotation leads sunspot cycle by 3 years","54 years of solar wind: fast rotation for fast streams","Sunspot numbers lag solar wind spin by three years"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the roughly 25% of missing hourly records before 1994, and the uneven sampling overall, do not shift the autocorrelation peaks differently for fast and slow wind, and that the velocity-dependent differences are not an artifact of how the data are binned.","fun_headline_variants_meta":{"raw":{"variants":["Solar wind rotation rate scales with stream velocity","Fast wind spins faster: solar wind rotation study","Solar wind rotation leads sunspot cycle by 3 years","54 years of solar wind: fast rotation for fast streams","Sunspot numbers lag solar wind spin by three years"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00044,"raw_usage":{"total_tokens":2138,"prompt_tokens":759,"completion_tokens":1379,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":375,"completion_tokens_details":{"reasoning_tokens":1304}},"tokens_in":375,"tokens_out":1379,"duration_ms":13029,"temperature":1.0,"reasoning_tokens":1304,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:25:54.590258+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the autocorrelation rotation period separately for the nearly complete post-1994 hourly data and for the sparse pre-1994 data; if the fast-wind rotation advantage does not appear in both subsets, the full-sample result is not robust. A complementary check is to generate synthetic time series with one fixed rotation period, impose the actual 74.8% sampling pattern, and see whether inferred periods shift with simulated velocity.","supporting_citations":[],"review_version":1}