{"id":"4acc2fcd-b434-4ca0-bb6c-6e4e3166e62e","arxiv_id":"2501.14558","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Blob-like plasma filaments in the Alcator C-Mod scrape-off layer move radially outward at 400 to 1000 m/s with similar sizes and speeds in ohmic and H-mode plasmas.","lead":"This paper measures the speed and size of blob-like plasma structures that carry particles outward in the edge of the Alcator C-Mod tokamak, finding speeds of 400 to 1000 meters per second. It shows these blob properties are similar in ohmic and high-confinement plasmas, supporting a universal picture of edge turbulence in fusion devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"TDE selection criteria may bias velocity estimates differently across confinement modes, so the reported similar velocities could be an artifact of conditional averaging rather than a physical result.","rationale":"The reader's weakest assumption was that TDE estimates are valid representations of blob motion, and my concern is a sharper version of that: the selection criteria in Section 4, combined with strongly differing intermittency across modes (gamma = 1.4 vs 6.2), could produce mode-dependent bias in the velocity estimates. This is the most load-bearing step because the paper's headline result is the similarity of velocities, and if the bias differs by mode, the similarity could be manufactured by the analysis. The proposed synthetic test is a direct, feasible check using the same stochastic model the authors already employ and the TDE code they developed. I do not claim the paper is wrong; the stochastic model fits are independent support, and the conditional averaging and PSD results are consistent. But the velocity comparison lacks the validation needed to separate physics from selection effects. The reader's CONDITIONAL verdict already captures the need for uncertainty quantification, so I do not change the verdict; the synthetic test would provide the missing validation.","tokens_in":9830,"tokens_out":9877,"duration_ms":100748,"concrete_test":"","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim of comparable blob velocities across confinement modes rests entirely on the TDE velocity estimates in Fig. 8, but these estimates are conditional on all five criteria in Section 4, including criterion (ii) requiring a unimodal cross-conditionally averaged pulse and criterion (iii) requiring a resolved time lag. The manuscript does not report the fraction of diode positions or events that pass these criteria, nor does it show velocity distributions or any uncertainty quantification for Fig. 9. This matters because the fluctuation statistics differ strongly between the modes: the Gamma shape parameter is 6.2 for ohmic, 1.4 for ELM-free, and 4.2 for EDA H-mode (Section 3). In the high-gamma ohmic case, overlapping pulses can shift the delay maximum, while in the low-gamma ELM-free case, isolated pulses may yield cleaner but possibly different delays. If the selection efficiency or the delay bias depends on gamma, the observed similarity of mean velocities (632-703 m/s) could be an artifact of the analysis rather than a shared physical property. The absence of error bars in Fig. 9 makes it impossible to distinguish a genuine common velocity range from a mode-dependent selection bias.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents gas puff imaging (GPI) measurements from three Alcator C-Mod discharges (ohmic, ELM-free H-mode, and EDA H-mode) and applies an improved time-delay estimation (TDE) method to infer blob velocities. The authors report radially outward blob velocities of 400–1000 m/s in all three confinement modes, similar vertical blob sizes from conditional averaging, and similar conditional burst shapes and power spectra. They interpret the results within a stochastic pulse model and conclude that blob properties are similar across ohmic and H-mode plasmas, supporting universality of blob-driven transport in the scrape-off layer.","tokens_in":10008,"tokens_out":3771,"duration_ms":35959,"significance":"If the central claim holds, the results extend earlier statistical comparisons of PDFs and PSDs to direct measurements of blob velocities and sizes, strengthening the case for universal blob-driven transport in tokamak scrape-off layers. The paper's strengths are the detailed description of the TDE method, the use of cross-conditional averaging with a reproducibility metric, and the connection to an established stochastic model. However, the evidence base is limited to three single discharges and no uncertainty quantification is provided for the headline velocity and size comparisons, so the strength of the claim currently exceeds the support.","major_comments":[{"comment":"The central claim of similar velocities across modes rests on average radial velocity profiles presented without error bars, confidence intervals, or any measure of scatter (for example, the standard deviation over the vertical diode positions or over independent time segments). Since each confinement mode is represented by a single discharge (Table 1), the reader cannot determine whether the differences among the SOL-averaged velocities (703, 644, and 632 m/s) are significant or whether the 400–1000 m/s range is a robust feature. The manuscript should provide uncertainty estimates, ideally from multiple time windows or multiple discharges, to support the cross-mode comparison.","section":"Section 4, Fig. 9"},{"comment":"The TDE velocities are assigned only when all five criteria are met, but the paper does not report the fraction of diode positions or events satisfying these criteria, nor does it quantify the conditional reproducibility values beyond the color scale in Fig. 8. If the acceptance rate or the delay bias differs systematically between modes—for example because the Gamma shape parameter differs strongly (6.2 for ohmic, 1.4 for ELM-free H-mode, and 4.2 for EDA H-mode, as reported in Section 3)—the observed similarity of mean velocities could be a selection artifact rather than a shared physical property. Please report the acceptance statistics and, if possible, the distribution of velocity estimates for each mode.","section":"Section 4, criteria (i)–(v), Fig. 8"},{"comment":"The vertical blob sizes are reported as single FWHM values (0.62, 0.81, and 0.69 cm for the three modes) without uncertainties, and the finite spatial resolution (3.8 mm spot size) and possible line-of-sight integration effects are not discussed. Since the transit-time argument in Section 5 uses the product of size and velocity to infer similar radial transit times, the absence of uncertainty on the sizes undermines the quantitative comparison across modes. At minimum, include an estimate of the measurement uncertainty on each FWHM and discuss the resolution correction.","section":"Section 3, Fig. 7, and Section 5"},{"comment":"The TDE method assumes that large-amplitude fluctuations correspond to coherent structures whose brightness pattern is approximately preserved between neighboring view points, as stated implicitly by criteria (i)–(v). The manuscript does not test this assumption directly for the present data, nor does it compare with an independent blob-tracking or synthetic-data test. A systematic validation (for example, using the three-point estimator on synthetic GPI data as in Ref. [38]) would substantially strengthen the claim that the derived velocities are unbiased measures of blob motion.","section":"Section 4"}],"minor_comments":[{"comment":"There are two typos in the text: 'confiment' should be 'confinement' and 'Lorentizan' should be 'Lorentzian'.","section":"Section 5"},{"comment":"The phrase 'gas pu ff' appears in the body text and in some references; it should be 'gas puff'.","section":"Section 2"},{"comment":"The caption states 'The color coding gives conditional reproducibility' but the quantity shown is 1 − CV; please define CV (conditional variance) either in the caption or in the text near the figure.","section":"Figure 8 caption"},{"comment":"The introduction says 'an improved time-delay estimation method' but does not specify what is improved relative to the previous method in Refs. [37,38]; please clarify whether the improvement is in the application, the algorithm, or the analysis procedure.","section":"Section 1"},{"comment":"Reference [38] is listed as 'submitted'; if it has been accepted by the time of publication, please update the reference with the journal and DOI.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of the journal and the authors are well known for their work on stochastic modeling of scrape-off layer fluctuations. The main issue is the lack of uncertainty quantification and the limited statistical basis (one discharge per mode) for the cross-mode universality claim. This is fixable with additional analysis (error bars, acceptance statistics, or a synthetic-data validation) and does not, in my view, require rejection. I recommend major revision rather than rejection because the central observation is plausible and the existing data are valuable, but the current manuscript does not yet support the strength of the conclusions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new thing here is the first time-delay-estimation-based radial blob velocity maps for Alcator C-Mod comparing ohmic, ELM-free H-mode, and EDA H-mode. The TDE method itself comes from the authors' prior work, but applying it to this dataset gives genuinely new numbers: radial velocities of 400–1000 m/s and vertical blob sizes of 0.6–0.8 cm, with similar values across the three confinement modes. If you work on SOL turbulence, this is a useful data point.\n\nThe paper does several things well. The TDE criteria are laid out explicitly, the analysis is described in enough detail to be followed, and the central conclusion of cross-mode similarity is backed by more than just the TDE velocities. The conditionally averaged burst shapes, the PSDs, the Gamma fits, and the conditional-averaging blob sizes all point the same way. The stochastic model is used for interpretation, not to produce the velocity data, so there is no real circularity. The citation pattern is also honest: the authors build directly on their own earlier method papers and on the prior fluctuation-statistics papers for these same discharges, and they say so.\n\nThe soft spots are real but not fatal. First, there are no error bars or uncertainty intervals anywhere in the velocity or size results, and each confinement mode is represented by a single discharge. The SOL-averaged velocities (632, 644, 703 m/s) are close, but without shot-to-shot variability or bootstrap intervals we cannot tell whether the residual spread is physics or just noise. Second, the TDE velocity estimates are conditional on all five criteria in Section 4, and the paper does not report how many diode positions or events passed the full set. Since the Gamma shape parameter differs strongly between modes (1.4 for ELM-free, 4.2 for EDA, 6.2 for ohmic), the stress-test concern that pulse overlap biases delays differently across modes is plausible. That concern does not overturn the paper's qualitative message, because the same conclusion is supported by independent observables, but it does mean Fig. 9 should be treated as indicative rather than quantitative. Third, there is no data or code availability statement, which is a practical barrier for anyone who wants to re-analyze the raw GPI data.\n\nWho is this for? Experimental and modeling people studying blob transport, SOL intermittency, and wall-loading estimates. A serious referee can ask for error bars, selection-efficiency diagnostics, and ideally more than one discharge per mode. I would send it to peer review and let the revision process tighten the claims.\n\nRecommendation: engage with it. The measurements are useful, the analysis is mostly sound, and the missing uncertainty quantification is fixable.","headline":"A careful TDE study of C-Mod blob velocities that plausibly shows similar blob motion across ohmic and H-mode, but one discharge per mode and missing error bars keep the similarity from being quantitative.","tokens_in":10591,"tokens_out":1829,"would_cite":true,"duration_ms":19157,"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":"Blob filaments in Alcator C-Mod move radially outward at 400-1000 m/s in ohmic and H-mode plasmas alike.","keywords":["Alcator C-Mod","scrape-off layer","blob filaments","turbulent transport","gas puff imaging","time-delay estimation","stochastic model","H-mode plasmas"],"falsifier":"A synthetic imaging benchmark would settle the method: generate mock gas-puff-imaging data with blobs of known velocities and controlled shape evolution, run the time-delay estimation, and check whether the recovered velocities match the inputs; if mild distortion produces large velocity errors, the reported 400-1000 m/s range would not be robust.","tokens_in":9607,"feed_emoji":"⚛️","tokens_out":8296,"duration_ms":65455,"temperature":0.7,"pith_summary":"This paper sets out to determine whether the blob-like filaments that dominate cross-field transport in the tokamak scrape-off layer move and scale the same way across different confinement regimes. Using an improved time-delay-estimation technique on gas-puff imaging data from Alcator C-Mod, it reports that in an ohmic plasma, an ELM-free H-mode, and an EDA H-mode, blob filaments move radially outward at 400-1000 m/s, with similar speeds and vertical sizes in all three cases. If correct, the result supports a universal picture of scrape-off-layer transport and reinforces predictions of how plasma will interact with reactor walls. The observations are consistent with a stochastic model in which the fluctuations are a superposition of uncorrelated pulses.","feed_headline":"Blobs race outward at 400-1000 m/s in all C-Mod modes","feed_subtitle":"Same speeds and sizes across ohmic and two H-mode states point to universal scrape-off-layer transport.","key_machinery":"The key machinery is an improved time-delay estimation (TDE) method for coarse-grained two-dimensional imaging data, which uses three non-aligned measurement points to recover both radial and poloidal components of a blob's velocity from cross-conditional averaged time delays. Each accepted velocity vector must satisfy five criteria concerning data availability, unimodal pulse shape, resolvable time lag, absence of strong electric field interference, and consistency with blob-like fluctuation statistics. The interpretation is anchored by a stochastic model describing the fluctuations as a superposition of uncorrelated two-sided-exponential pulses, which yields Gamma-distributed amplitudes, Lorentzian frequency spectra, and an exponential mean radial profile whose scale length is the product of blob velocity and parallel transit time.","core_discovery":"The paper's central claim is that the scrape-off layer of Alcator C-Mod is dominated by large-amplitude, blob-like filaments moving radially outward, and that their velocities and sizes are essentially the same in ohmic, ELM-free H-mode, and EDA H-mode plasmas. The radial velocity component increases from about 400 m/s near the separatrix to about 1000 m/s in the limiter shadow, with scrape-off-layer-averaged values of 703 m/s (ohmic), 644 m/s (ELM-free H-mode), and 632 m/s (EDA H-mode). Vertical blob sizes are comparable across the modes, with full widths at half maximum between 0.62 and 0.81 cm. These similarities line up with the close agreement of conditionally averaged burst shapes and frequency power spectra, and are consistent with the stochastic model of a superposition of uncorrelated pulses, strengthening the case that blob-driven cross-field transport is a universal feature of tokamak scrape-off layers.","pith_inferences":["A natural extension is to measure the Gamma shape parameter across a wider parameter space; if blob speed and size stay fixed while the waiting time changes, the shape parameter alone would track the confinement-mode differences.","The TDE method could be ported to synthetic turbulence simulations, where the exact blob velocities are known, to test whether the 400-1000 m/s range and its radial growth are genuine or an artifact of the estimation geometry.","If these velocities hold for future high-duty-cycle reactors, wall erosion in the far scrape-off layer may be largely independent of the operating regime, making blob generation rate the key control knob.","A direct measurement of the poloidal velocity component would test the assumption used here that radial motion dominates, which underlies the vertical size estimates."],"forward_implications":["Blob velocities and sizes are similar across ohmic, ELM-free H-mode, and EDA H-mode, so the radial transit time of blobs, $\\tau_\\perp = \\ell/v$, is also similar in these regimes.","The scrape-off-layer-averaged radial blob velocities (703, 644, and 632 m/s) imply that confinement mode mainly changes how often blobs are born, not how fast they travel.","The stochastic model's prediction of an exponential density profile with e-folding length given by blob velocity times parallel transit time is supported by the measured profiles.","The measured similarities across modes add to the evidence that blob-driven cross-field transport is universal in tokamaks, matching observations on other devices.","The radial velocity profile rises from about 400 m/s at the separatrix to about 1000 m/s in the limiter shadow, although the authors note this need not imply that individual blobs accelerate."],"supporting_citations":[{"why":"supplies the three-point time-delay estimation method applied to the GPI data.","marker":"[37]"},{"why":"develops the time-delay estimation for a superposition of localized pulses, validating the method.","marker":"[38]"},{"why":"analyzed the same H-mode data sets and established the burst and spectral statistics these findings build on.","marker":"[11]"},{"why":"demonstrated universality of the fluctuation statistics for these plasmas, which this paper extends to velocities.","marker":"[12]"},{"why":"introduced the stochastic model of uncorrelated pulses used to interpret the fluctuation statistics.","marker":"[19]"},{"why":"derives the exponential profile prediction linking blob velocity, parallel transit time, and density scale length.","marker":"[21]"},{"why":"provides blob velocity measurements from another device used as a comparison for universality.","marker":"[27]"}],"fun_headline_variants":["Blob transport: same speeds across C-Mod modes","C-Mod blobs: universal scrape-off-layer speeds","Blob velocities match across ohmic and H-mode","Similar blob speeds in all C-Mod plasmas","Universal blob transport in C-Mod scrape-off layer"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that each measured velocity truly represents a coherent blob: time delays between neighboring imaging channels are treated as rigid advection of a structure whose brightness pattern is preserved, and if blobs distort, decorrelate, or are mis-sampled by the viewing geometry, the estimated 400-1000 m/s velocities could be biased.","fun_headline_variants_meta":{"raw":{"variants":["Blob transport: same speeds across C-Mod modes","C-Mod blobs: universal scrape-off-layer speeds","Blob velocities match across ohmic and H-mode","Similar blob speeds in all C-Mod plasmas","Universal blob transport in C-Mod scrape-off layer"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000219,"raw_usage":{"total_tokens":1421,"prompt_tokens":902,"completion_tokens":519,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":518,"completion_tokens_details":{"reasoning_tokens":455}},"tokens_in":518,"tokens_out":519,"duration_ms":4824,"temperature":1.0,"reasoning_tokens":455,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T15:01:32.309194+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A synthetic imaging benchmark would settle the method: generate mock gas-puff-imaging data with blobs of known velocities and controlled shape evolution, run the time-delay estimation, and check whether the recovered velocities match the inputs; if mild distortion produces large velocity errors, the reported 400-1000 m/s range would not be robust.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the three-point time-delay estimation method applied to the GPI data."},{"cited_title":"Time delay velocity estimation from a superposition of localized and uncorrelated pulses","cited_arxiv_id":"2411.06544","evidence_quote":"develops the time-delay estimation for a superposition of localized pulses, validating the method."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"analyzed the same H-mode data sets and established the burst and spectral statistics these findings build on."},{"cited_title":"Theodorsen, O","cited_arxiv_id":null,"evidence_quote":"demonstrated universality of the fluctuation statistics for these plasmas, which this paper extends to velocities."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"introduced the stochastic model of uncorrelated pulses used to interpret the fluctuation statistics."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"derives the exponential profile prediction linking blob velocity, parallel transit time, and density scale length."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides blob velocity measurements from another device used as a comparison for universality."}],"review_version":1}