{"id":"fe06d515-9423-445b-b932-c510a4a66dca","arxiv_id":"2501.16686","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":2,"one_line_summary":"Single-drop breakup experiments and a two-parameter gamma model reproduce the Marshall-Palmer raindrop size distribution and roughly match in-situ flight data.","lead":"Laboratory experiments on single water drops breaking in an upward airstream produce drop size distributions that match the classic Marshall-Palmer raindrop formula and airborne rain measurements. The paper ties microscopic breakup physics to rainfall and radar predictions.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq (14) is unsupported: the eleven characteristic sizes, gamma parameters, and integral I are never reported, and Eq (4)'s parameter definitions are dimensionally inconsistent as written.","rationale":"The reader's weakest assumption is that the two-parameter gamma distribution derived from the mean and standard deviation of eleven characteristic sizes captures the full child-drop size distribution, and that these values are not reported. My concern overlaps with that, but I identify a sharper technical gap: even if the values were reported, Eq. (4) as written is ambiguous or dimensionally inconsistent unless the mean and standard deviation are explicitly normalized, and the route from the Weber-dependent gamma parameters to the single power law Eq. (14) is not shown. Both issues attach directly to the central claim, because Eq. (14) is the quantitative bridge to Marshall-Palmer and to the in-situ comparison. The paper has real strengths: the shadowgraphy and machine-learning holography experiments are described in detail, and the in-situ CAIPEEX data are a valuable independent comparison. However, the central theoretical claim cannot be evaluated without the missing numerical inputs and derivation. Since the issues are addressable by data release and a re-derivation, a conditional verdict is appropriate, not a rejection. I therefore keep the reader's conditional verdict unchanged, while noting that the specific concern I raise is more technical and somewhat different from the reader's general reproducibility concern.","tokens_in":11387,"tokens_out":5540,"duration_ms":56104,"concrete_test":"Ask the authors to report, for each Weber number, the eleven characteristic sizes, the normalized mean and standard deviation, alpha, beta, and the integral I = integral x^(7/2) P(x) dx; then verify that P(x) is normalized and has mean 1 in x, and recompute Eq. (14) from Eq. (2) to check whether Lambda = 40 and a = 0.21 follow.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative result is Eq. (14), d_avg^-1 = 40 R^-0.21, obtained by evaluating Eq. (2) with P(x) from Eq. (4). This result cannot be checked from the preprint: the eleven characteristic sizes, the resulting mean and standard deviation, the values of alpha and beta, and the value of the integral I = integral x^(7/2) P(x) dx are never reported. Without these, the claimed match to Marshall-Palmer and the CAIPEEX comparison are assertions. More specifically, Eq. (4) defines P(x) with x = d/d_avg dimensionless, but the text says alpha = (xbar/s)^2 and beta = s^2/xbar, where xbar and s are the mean and standard deviation of the characteristic sizes. If xbar and s are physical lengths (micrometers), then beta has units of length, the gamma density is not correctly normalized in dimensionless x, and the mean of x under P(x) is not 1; if they are meant to be normalized by d_avg, that is not stated, and because d_avg is set equal to the arithmetic mean of the characteristic sizes, the normalized mean is automatically 1, leaving only one free parameter. Either way the parameterization needs clarification. The derivation of Eq. (14) also requires specifying how the three Weber-number cases (We = 9.38, 16.9, 18.9) are combined into a single power law with fixed constants; no fitting or averaging procedure is given. These omissions leave the central claim unverifiable.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports laboratory experiments on the fragmentation of single water drops in an upward airstream at three Weber numbers (We = 9.38, 16.9, and 18.9), corresponding to bag, bag-stamen, and dual-bag breakup morphologies. The child-drop size distributions are measured using shadowgraphy and machine-learning-assisted digital in-line holography. The authors then construct a two-parameter gamma distribution model, Eq. (4), whose parameters are derived from the mean and standard deviation of eleven characteristic breakup sizes taken from the bag, rim, and node fragmentation modes. They claim that this model reproduces the experimental distributions, and that evaluating the integral in Eq. (2) yields the Marshall-Palmer-like relation d_avg^{-1} = 40 R^{-0.21} (Eq. (14)). The paper also compares the resulting drop size distributions with in-situ raindrop measurements from the CAIPEEX research flights. The central claims are that a lab-derived, breakup-based model can predict the steady-rain drop size distribution and that the model matches aerial observations.","tokens_in":11701,"tokens_out":5504,"duration_ms":53495,"significance":"If the central derivation and comparisons are correct, the paper offers a potentially important bridge between single-drop fragmentation physics and operational rainfall parameterization: a model that estimates davg from measurable breakup mechanisms without treating it as an unknown input. The experimental data set, which combines controlled lab breakup with in-situ aircraft measurements, is a valuable resource, and the use of holography with a U-Net segmentation pipeline is a methodological strength. However, the main quantitative result, Eq. (14), is not verifiable from the preprint as written, and the parameterization in Eq. (4) contains a dimensional ambiguity. The significance of the paper in its current form is therefore conditional on a revised and fully documented derivation.","major_comments":[{"comment":"The computation of the integral in Eq. (2) that leads to Eq. (14) is not reported. The eleven characteristic sizes, the resulting mean and standard deviation, the values of alpha and beta for each Weber number, and the numerical value of the integral are all absent. The coefficient 40 in Eq. (14) therefore cannot be checked. The manuscript must specify how the three Weber-number cases (We = 9.38, 16.9, and 18.9), which have distinct gamma parameters, are combined into a single power law with fixed constants; if one representative case is used, that choice must be justified.","section":"Section IV, Eq. (14)"},{"comment":"The definitions alpha = (xbar/s)^2 and beta = s^2/xbar are dimensionally inconsistent as written. The text states that xbar and s are the mean and standard deviation of the eleven characteristic sizes, which are measured in micrometers, but the variable x = d/davg in Eq. (4) is dimensionless. If xbar and s are meant to be normalized by davg, that normalization is not stated. Moreover, because davg is set equal to the arithmetic mean of the characteristic sizes, the normalized mean is identically 1, which reduces the gamma distribution to a single-parameter form; this reduction needs to be acknowledged and the subsequent parameter counting made explicit.","section":"Section IV, Eq. (4)"},{"comment":"The value n0 = 0.08 cm^-4 is taken from the Marshall-Palmer empirical fit and inserted into Eq. (2). Since Eq. (14) is then compared with the Marshall-Palmer relation, the comparison is partly anchored by the chosen n0. The authors should either derive n0 from the fragmentation model or clearly label it as an external input, and they should report the sensitivity of Eq. (14) to a realistic range of n0 values.","section":"Section IV, Eqs. (2) and (14)"},{"comment":"The conversion of the CAIPEEX in-situ droplet concentration data C(d) into the normalized distribution P(x) plotted in Figure 6 is not described. In particular, the authors do not state how davg is estimated from the measured size-resolved concentrations or how the normalization is performed. Without this information, the claimed 'satisfactory alignment' between the laboratory model and the aerial data cannot be quantitatively assessed.","section":"Section V, Figure 6"}],"minor_comments":[{"comment":"The phrase 'we compute the integral R x^{7/2} P(x) x. in Eq. (2)' is malformed; it should read 'the integral \\int x^{7/2} P(x) dx in Eq. (2)'.","section":"Section IV"},{"comment":"The initial droplet diameter is stated as 3.6 ± 0.08 mm in the main text but as 3.6 ± 0.06 mm in the caption of Figure 3; please reconcile the uncertainty.","section":"Section II and Figure 3 caption"},{"comment":"The sentence fragment '7 also emphasized the importance...' should be rephrased, for example as 'Villermaux and Bossa also emphasized the importance...'.","section":"Section VI"},{"comment":"The description 'n0 is the average spatial density of drops (measured in cm^-4)' is ambiguous; n0 is the intercept of the number concentration per unit size interval per unit volume, and the text should use standard terminology.","section":"Section II"},{"comment":"The statement 'All characteristic sizes are measured in µm' should be accompanied by explicit unit conversions, since davg in Eq. (14) is in centimetres and R in mm/hr.","section":"Section IV"},{"comment":"The shaded envelope in Figure 6 is not defined in the caption; specify whether it covers the full range of scattered data points, a percentile interval, or the combined range of all in-situ and laboratory data.","section":"Section V, Figure 6"}],"recommendation":"major_revision","confidential_remarks":"The paper's main quantitative claim rests on an unverifiable computation: the intermediate values and the final integral are not reported, and Eq. (4) has a dimensional inconsistency. These issues are fixable within the scope of the manuscript if the authors provide the full derivation and numerical values. I therefore recommend major revision rather than rejection. The editor may also wish to check whether the comparison with the Marshall-Palmer relation is sufficiently independent given the use of n0 = 0.08 cm^-4 from Marshall-Palmer itself."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here is my read. The genuinely new piece is the lab data: counter-current single-drop breakup at We = 9.38, 16.9, and 18.9, covering bag, bag-stamen, and dual-bag modes, with drop size distributions from machine-learning-assisted in-line holography, plus comparison to CAIPEEX in-situ flight data. That is real experimental work, five repetitions per case, and the holography pipeline is described well enough to be credible, building on Ade et al. 2023/2024. The theoretical model itself is not new—two-parameter gamma from Jackiw and Ashgriz, integral relation from Villermaux and Bossa—but applying it to counter-current breakup and comparing with Marshall-Palmer and CAIPEEX is a reasonable contribution.\n\nThe problem is that the central quantitative claim, Eq. (14) d_avg^{-1} = 40 R^{-0.21}, cannot be checked from the preprint. The eleven characteristic sizes, the mean and standard deviation, alpha, beta, and the value of the integral in Eq. (2) are never reported. So the claim that the model accurately predicts measurements in Fig. 5, and the deviation percentages quoted in the text, are assertions. I also think the stress-test note lands on Eq. (4): if xbar and s are physical lengths, beta has units of length, which is not consistent with the normalized gamma density in dimensionless x. If they are meant to be normalized by d_avg, the text should say so. Since d_avg is defined as the arithmetic mean of the characteristic sizes, the normalized mean is automatically 1, leaving only one free parameter; the paper does not discuss that. The way the three Weber cases are combined into one power law with fixed constants is also unspecified. And the Marshall-Palmer comparison is partly anchored by taking n0 = 0.08 cm^-4 from Marshall-Palmer, so the match is not fully independent.\n\nAll of this is fixable. Report the characteristic sizes and fitted gamma parameters for each Weber case, show the integral calculation, state the normalization convention, and quantify the in-situ comparison with a proper metric instead of a shaded envelope. The laboratory measurements and the CAIPEEX comparison are worth keeping. I would send it to peer review—the experimental contribution is solid enough to merit referee time—but the current version should not be accepted until the numbers behind Eq. (14) are on the page or in a supplement. For my own work, I would wait for that revision before citing the relation.","headline":"New lab DSD data and a plausible rain-model application, but Eq. (14) is currently an assertion until the characteristic sizes and gamma parameters are reported.","tokens_in":12301,"tokens_out":2678,"would_cite":false,"duration_ms":27073,"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":"Laboratory drop fragmentation, modelled with a two-parameter gamma distribution, recovers the Marshall-Palmer raindrop-size law and matches in-situ aerial data.","keywords":["raindrop size distribution","drop breakup","bag breakup","gamma distribution","Marshall-Palmer relation","digital in-line holography","Weber number","aerial in-situ measurement"],"falsifier":"Measure the full child-drop size distribution for the three Weber numbers (9.38, 16.9, 18.9) with the same holography setup, recompute the eleven characteristic sizes from the cited formulas, and check whether the gamma distribution with $\\alpha=(\\bar{x}/s)^2$ and $\\beta=s^2/\\bar{x}$ reproduces the measured tail at $d/d_{\\rm avg}\\approx 7.3$ within a few percent; if the tail deviates by more than the claimed 7\\%, the Marshall-Palmer match is not established.","tokens_in":11204,"feed_emoji":"🌧️","tokens_out":7989,"duration_ms":65944,"temperature":0.7,"pith_summary":"Raindrop size distributions determine how weather radar echoes are interpreted, since reflectivity scales with the sixth power of drop diameter. This paper asks whether the distribution observed in steady rain can be produced from the physics of a single drop breaking up in an upward airstream. It reports laboratory experiments on drops of 3.6 mm diameter in Weber-number regimes that produce bag, bag-stamen, and dual-bag breakup, together with a two-parameter gamma model whose parameters come from eleven characteristic fragment sizes. The model reproduces the measured distributions and yields $d_{\\rm avg}^{-1} = 40R^{-0.21}$, close to the Marshall-Palmer relation for steady rain ($d_{\\rm avg}^{-1} = 41R^{-0.21}$), and it aligns with airborne in-situ measurements from the CAIPEEX campaign.","feed_headline":"Lab drop breakups reproduce the classic Marshall-Palmer rain law","feed_subtitle":"Eleven fragment sizes from bag, rim, and node breakup reproduce the standard steady-rain drop-size law.","key_machinery":"The central object is the two-parameter gamma distribution $P(x = d/d_{\\rm avg}) = x^{\\alpha-1}e^{-x/\\beta}/(\\beta^{\\alpha}\\Gamma(\\alpha))$, with shape $\\alpha = (\\bar{x}/s)^2$ and rate $\\beta = s^2/\\bar{x}$, where $\\bar{x}$ and $s$ are the mean and standard deviation of eleven characteristic fragment sizes. Four sizes come from bag rupture (bag thickness, receding-rim thickness, the Rayleigh-Plateau ligament size, and satellite droplets), four from rim fragmentation (rim-thickness size, receding-rim collision size, and two satellite corrections), and three from node breakup at minimum, mean, and maximum node volume fractions. This replaces the single-parameter gamma model in which $\\alpha = \\beta$, and it turns the previously assumed average drop diameter into a computed output. The distribution is then fed through the rainfall-rate integral (Eq. 2) to reach the Marshall-Palmer-type law.","core_discovery":"The central claim is that all breakup modes contribute characteristic sizes that, combined in a two-parameter gamma distribution, predict the full child-drop size distribution, and that this distribution matches both laboratory measurements and the standard steady-rain relation. Computing the integral in Eq. (2) with the model gives $d_{\\rm avg}^{-1} = 40R^{-0.21}$, so the average drop diameter tracks rainfall rate with the same exponent (0.21) as Marshall-Palmer and a coefficient of 40 versus 41. At large normalized sizes ($d/d_{\\rm avg} = 7.3$), the model deviates from Marshall-Palmer by 6.6--7.6\\%, while the single-parameter gamma model deviates by 24--27\\%. The same distribution is then compared with in-situ DSD measurements from three research flights, with which it shows satisfactory alignment.","pith_inferences":["If the eleven characteristic sizes can be predicted across the Weber-number range rather than measured case by case, the model could extend to unsteady or intense rain without in-situ calibration.","The comparison with flight data is visual and qualitative; a quantitative, altitude-resolved test against the CAIPEEX spectra would be a sharper check of the model's tail behaviour.","The model's fidelity at large $d/d_{\\rm avg}$ could be tested directly by computing the sixth moment $\\int d^6 P(d)\\,dd$ from the holography data and comparing it with radar reflectivity--rain-rate relations.","A natural extension is to test whether the same gamma parameters also describe DSDs altered by collisional breakup or mixed-phase microphysics, processes the paper itself lists as unaddressed."],"forward_implications":["The model yields $d_{\\rm avg}^{-1}=40R^{-0.21}$, effectively recovering Marshall-Palmer's $41R^{-0.21}$ for steady rain.","At the large-drop end ($d/d_{\\rm avg}\\approx 7.3$), the model stays within about 7\\% of Marshall-Palmer across all three breakup modes, whereas the single-parameter model deviates by 24--27\\%.","The same two-parameter distribution describes bag, bag-stamen, and dual-bag breakup, suggesting the eleven characteristic sizes capture the breakup physics relevant to rain.","In-situ drop spectra from the three CAIPEEX flights fall within the envelope of the laboratory and theoretical distributions, supporting laboratory-scale experiments for rainfall modelling.","Radar-relevant quantities such as the sixth moment of the drop size distribution can be computed from the model rather than assumed as input."],"supporting_citations":[{"why":"Supplies the baseline Marshall-Palmer relation for steady rain that the model reproduces.","marker":"[5]"},{"why":"Provides the single-parameter gamma model and Eq. (2) that the present model extends.","marker":"[7]"},{"why":"Gives the characteristic-size formulas and the two-parameter gamma framework used here.","marker":"[17]"},{"why":"Establishes the experimental size-distribution method at moderate Weber numbers used as the starting point.","marker":"[18]"},{"why":"Provides the machine-learning in-line holography pipeline used to measure child-drop sizes.","marker":"[22]"},{"why":"Supplies the CAIPEEX campaign context for the in-situ aerial data.","marker":"[31]"},{"why":"Provides the particle-size spectra from the three research flights used for comparison in Figure 6.","marker":"[32]"}],"fun_headline_variants":["Lab drop breaks match Marshall-Palmer rain law","Single-drop fragmentation reproduces rain drop-size law","Lab experiments validate Marshall-Palmer drop-size relation","Fragmentation lab data confirm classic rain drop-size model","Laboratory drops mirror steady-rain drop-size distribution"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"A two-parameter gamma distribution whose shape and rate are fixed by the mean and standard deviation of eleven theoretically estimated characteristic sizes captures the full measured child-drop distribution, including the large-drop tail used for the Marshall-Palmer comparison.","fun_headline_variants_meta":{"raw":{"variants":["Lab drop breaks match Marshall-Palmer rain law","Single-drop fragmentation reproduces rain drop-size law","Lab experiments validate Marshall-Palmer drop-size relation","Fragmentation lab data confirm classic rain drop-size model","Laboratory drops mirror steady-rain drop-size distribution"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000213,"raw_usage":{"total_tokens":1389,"prompt_tokens":878,"completion_tokens":511,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":494,"completion_tokens_details":{"reasoning_tokens":437}},"tokens_in":494,"tokens_out":511,"duration_ms":5636,"temperature":1.0,"reasoning_tokens":437,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T11:27:08.150671+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the full child-drop size distribution for the three Weber numbers (9.38, 16.9, 18.9) with the same holography setup, recompute the eleven characteristic sizes from the cited formulas, and check whether the gamma distribution with $\\alpha=(\\bar{x}/s)^2$ and $\\beta=s^2/\\bar{x}$ reproduces the measured tail at $d/d_{\\rm avg}\\approx 7.3$ within a few percent; if the tail deviates by more than the claimed 7\\%, the Marshall-Palmer match is not established.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the baseline Marshall-Palmer relation for steady rain that the model reproduces."},{"cited_title":"Villermaux \\ and\\ author B","cited_arxiv_id":null,"evidence_quote":"Provides the single-parameter gamma model and Eq. (2) that the present model extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the characteristic-size formulas and the two-parameter gamma framework used here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the experimental size-distribution method at moderate Weber numbers used as the starting point."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the machine-learning in-line holography pipeline used to measure child-drop sizes."},{"cited_title":"Prabhakaran , author P","cited_arxiv_id":null,"evidence_quote":"Supplies the CAIPEEX campaign context for the in-situ aerial data."},{"cited_title":"Patade , author T","cited_arxiv_id":null,"evidence_quote":"Provides the particle-size spectra from the three research flights used for comparison in Figure 6."}],"review_version":1}