{"id":"a3846ccf-279a-4c20-a311-be63c12e3b95","arxiv_id":"2511.07688","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In simulated atmospheric turbulence, individual speckles can survive and stay localized over several Rayleigh lengths, and their lifetime distribution sharpens with turbulence strength.","lead":"This paper follows individual bright spots (speckles) as a simulated laser beam travels through turbulent air and measures how long each one survives. It finds that a beam shatters as its wavefront decorrelates, and that some speckles stay bright and narrow for several times their natural spreading length — useful for turbulence diagnostics and free-space laser links.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Speckle-lifetime scalings depend on an unspecified detection threshold/filter; without sensitivity analysis the persistence claim is unverified.","rationale":"The reader's weakest assumption is exactly where the argument is most exposed. The paper's novel contribution is not ensemble intensity statistics but object-level lifetimes; those lifetimes are generated by a detection algorithm whose two key parameters (η and the radius threshold) are either fixed with no robustness check or not specified at all. Because the same thresholds also set the normalization z0,j, the persistence ratio Δz/z0 is not independent of the detector. The 500-run statistics give confidence that the trends are reproducible under one parameter set, but they do not establish that the trends survive threshold variation. I do not see an internal inconsistency or an obviously false physical claim; the issue is unverified parameter sensitivity. A sensitivity sweep and a shuffled-label control would settle it. Since the reader's conditional verdict already captures this, no change is needed.","tokens_in":8712,"tokens_out":3986,"duration_ms":42663,"concrete_test":"Using the same 500-run dataset, vary η over {0.001, 0.005, 0.01, 0.02, 0.05} and vary the radius filter over multiples of the Fried parameter r0 (e.g., 0.5, 1, 2 r0). Recompute N_speckles, mean Δz/z0, and the 10%/2% survival fractions for each (C_n^2, w0, ℓ0) setting. If the 7×/4× scaling or the qualitative weak-to-strong ranking changes by more than 20%, the persistence claim is threshold-dominated. Also run a shuffled-z control: at each z-plane, reassign speckle labels to random nearest peaks from the previous plane using identical thresholds; if the resulting Δz/z0 distribution approximates the reported one, the tracking does not identify persistent objects. Report both results in the paper.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claims—7× speckle count increase and 4× shorter average lifetimes from weak to strong turbulence, plus the 10%/2% survival fractions in Sec. III.C—are all computed from a speckle population defined by two thresholds. Sec. II.A fixes the intensity threshold at η=0.01 Imax (source-plane maximum) but leaves the radius threshold for removing adjacent/collocated speckles completely unspecified. Sec. II.C defines persistence Δz_j as the distance between creation and disappearance events “identified by the intensity and radius thresholds,” so lifetime itself is a property of the detector, not solely of the field. The normalization z0,j=πw0,j^2/λ uses the speckle width at creation, which is also threshold-dependent because the creation time and initial width change when η changes. No sensitivity analysis, no matching tolerance, no code, and no data are provided. Consequently, the reported scaling of lifetime distributions with C_n^2, and especially the claim of “persistence ... multiple diffraction lengths,” may be a threshold artifact rather than a robust property of turbulent transport. This is the load-bearing weak point because every headline number in the abstract and conclusion traces back to this detection/tracking definition.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a numerical study of individual speckle dynamics during Gaussian-beam propagation through atmospheric turbulence. Using split-step propagation with a modified von Kármán spectrum, the authors detect and track speckles via local intensity maxima above ηI_max (η=0.01) and an unspecified radius threshold. They report that beam fragmentation, defined as a rapid increase in speckle number, coincides with the decay of a magnitude-squared coherence metric C(z). They further characterize speckle confinement (width evolution) and persistence (normalized lifetime Δz_j/z0,j) over 500 realizations, concluding that speckles are spatially localized and persist over distances significantly longer than their associated Rayleigh length, with statistics that scale with turbulence strength (e.g., a 7× increase in speckle count and 4× decrease in average lifetime from weak to strong turbulence). The paper suggests these persistence statistics could serve as a diagnostic for the refractive-index structure parameter.","tokens_in":9015,"tokens_out":3346,"duration_ms":37353,"significance":"If the central claim holds, the paper offers a genuinely new object-level perspective on speckle dynamics, complementing the usual ensemble statistics of scintillation and coherence. The methodological strengths are the use of a standard propagation model (split-step with modified von Kármán spectrum), ensemble statistics over 500 runs, and the introduction of four complementary diagnostics (mode field radius, spatial correlation C(z), speckle confinement, and persistence). The finding that speckle lifetimes scale with turbulence strength, if robust, could have practical value for free-space optical sensing and communication. However, the quantitative conclusions are currently conditioned on detection and tracking thresholds that are neither fully specified nor subjected to sensitivity analysis. This is a load-bearing weakness because the headline numbers in the abstract and conclusion—'persistence multiple diffraction lengths,' '7× more speckles,' '4× shorter lifetimes,' '10%/2% survival fractions'—are all computed from a speckle population defined by those thresholds. The paper's significance therefore rests on an as-yet-unverified robustness property.","major_comments":[{"comment":"The speckle detection algorithm is incomplete: the intensity threshold is given as η=0.01 I_max, but the 'radius threshold' for removing adjacent or collocated speckles is never specified. Every count, confinement, and persistence statistic in Sec. III.C (including the 7× count increase, 4× lifetime decrease, and the 10%/2% survival fractions) is conditional on this unspecified filter. Without stating the radius threshold and providing a sensitivity analysis (e.g., varying η and the radius over reasonable ranges), the reported scalings could be detector artifacts rather than properties of the turbulent field.","section":"Sec. II.A"},{"comment":"The normalized persistence Δz_j/z0,j is defined using z0,j = π w0,j^2/λ, where w0,j is the initial width of the j-th speckle at the time of creation. Both the creation time and the initial width are determined by the intensity and radius thresholds. Consequently, the lifetime histograms in Fig. 4 (insets) are not threshold-invariant observables; changing η shifts the birth/death events and also rescales z0,j. The paper should either demonstrate that the qualitative conclusions are unchanged over a range of thresholds, or define a threshold-independent measure of speckle identity.","section":"Sec. II.C"},{"comment":"The tracking between adjacent propagation steps is said to be based on 'nearest neighbor conditions' (Fig. 1 caption), but no matching tolerance is given—e.g., the maximum allowed transverse displacement between steps, or the criterion for distinguishing a reappearing speckle from a new one. Without this, a speckle that moves more than the tolerance between two closely spaced planes would be counted as a death and a later birth, inflating the number of short-lived speckles and biasing the persistence distribution. The matching algorithm must be specified precisely, and the sensitivity of the results to the matching tolerance should be tested.","section":"Fig. 1 and Sec. II.A"},{"comment":"The paper states that 'strong turbulence across 500 iterations' yields ≈4738 speckles, but it does not specify which C_n^2 value corresponds to 'strong' in this count. From Fig. 3, one infers C_n^2=10^-13 m^{-2/3}, but this should be stated explicitly in the text. More importantly, the 10%/2% survival fractions are quoted without error bars or run-to-run variability; given that only 500 realizations are used, the statistical uncertainty in these tail fractions may be substantial and should be reported.","section":"Sec. III.C"}],"minor_comments":[{"comment":"The text says 'Both quantities are plot as functions' — 'plot' should be 'plotted'.","section":"Sec. III.B"},{"comment":"The phrase 'which can be seen in the histograms (Fig. 4, inset)' should be plural: 'insets'.","section":"Sec. III.C"},{"comment":"The integrals for the mode field radius lack explicit integration limits; add 'over the transverse plane' or write ∫∫ over x,y to avoid ambiguity.","section":"Eq. (5)"},{"comment":"Reference [34] is incomplete ('D. G. Voelz, (No Title)'); it should cite the full book: D. G. Voelz, Computational Fourier Optics: A MATLAB Tutorial (SPIE Press, 2011). Reference [46] also lacks publisher location/details.","section":"References"},{"comment":"The quantity C(z) in Eq. (6) is called a 'spatial coherence' metric, but it is actually the overlap between the vacuum-propagated and turbulence-propagated fields, not the mutual coherence function of the field. This terminology could confuse readers; consider calling it the 'field-overlap correlation' or 'vacuum-field fidelity' to distinguish it from standard spatial coherence.","section":"Sec. II.B"},{"comment":"The right panels (a1–a4, b1–b4, c1–c4) are described in the caption as 'transverse profiles,' but the axis labels and color scale are not fully explained; it would help to define the color scale (e.g., intensity normalized to I_max) in the caption.","section":"Fig. 2"}],"recommendation":"major_revision","confidential_remarks":"The paper's topic is suitable for the journal and the object-level tracking approach is interesting. The main concern is that the quantitative claims are not yet robust because the detection/tracking thresholds are underspecified. I would encourage the editor to ask for a sensitivity analysis and a precise statement of the radius threshold and matching tolerance. If the authors can show that the scaling and persistence conclusions are insensitive to these choices, the paper could be acceptable after that revision. I did not find evidence of circular parameter fitting or invented entities; the issue is purely one of threshold definiteness."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is the first paper I know that tries to track individual speckles as objects through turbulent propagation rather than treating them as samples of a stationary random field. That framing is genuinely new, and the persistence/fragmentation distributions are worth taking seriously. The propagation setup is standard — split-step Fresnel with a modified von Kármán spectrum — and 500 realizations per parameter set is decent for this kind of simulation. The global diagnostics (mode-field radius and the normalized overlap with the vacuum-propagated field) are clearly defined, and the result that larger beams decorrelate and fragment closer to the source is consistent with the physics and comes out cleanly.\n\nThe main soft spot is not the turbulence model; it is the object definition. Speckles are found with η = 0.01 Imax and “a radius threshold” that removes adjacent or collocated peaks, but that radius threshold is never specified. Every central number — the 7× increase in speckle count, the 4× shorter average lifetimes, the 2% vs 66% survival fractions — counts objects defined by that filter. Persistence is the distance between creation and disappearance events “identified by the intensity and radius thresholds,” and the normalization z0,j uses the speckle width at creation, which is itself threshold-dependent because creation time and initial width both change with η and the filter. There is no sensitivity analysis, no matching tolerance, no code, and no data. So at this stage the quantitative scalings are properties of the detector/tracker as much as of the turbulent field. The qualitative story — more and shorter-lived speckles as Cn2 increases — is probably robust, but the specific numbers cannot be independently checked.\n\nTwo smaller things. First, the metric called “spatial coherence” C(z) is actually an overlap between the turbulence-propagated field and the vacuum-propagated field. That is a useful measure of remaining global order, but it is not the transverse spatial coherence of the field, and “complete spatial decorrelation” is stronger than what C(z) = 0.01 shows. Second, the histograms in Fig. 4 are presented without error bars even though 500 runs are available; error bars would help the reader assess the long tails.\n\nOverall: the central idea is good and the paper deserves a serious referee. I would send it to review with a request for a sensitivity analysis over η and the radius filter, and ideally released code. If the scalings survive, this is a solid new diagnostic for effective Cn2.","headline":"New object-level speckle tracking with a genuinely useful persistence diagnostic, but the headline scalings sit on an unspecified radius threshold and no sensitivity analysis; worth refereeing with a request for robustness checks.","tokens_in":9425,"tokens_out":2860,"would_cite":true,"duration_ms":41372,"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":"Turbulence produces individual speckles that persist as localized objects over distances several times their Rayleigh length, and their lifetimes encode turbulence strength.","keywords":["speckle tracking","atmospheric turbulence","optical propagation","speckle persistence","beam fragmentation","scintillation","refractive index structure parameter","partial localization"],"falsifier":"Vary the detection threshold η from, say, 0.001 to 0.05 Imax in the same simulated propagations. If the 7× increase in speckle count and 4× decrease in average lifetime between weak and strong turbulence disappear, or if no speckle survives beyond one diffraction length under strong turbulence, the central claim of turbulence-encoded persistence would be falsified. Alternatively, a field experiment with a high-speed camera at two or more observation planes that finds no persistent localized intensity features under strong turbulence would contradict the result.","tokens_in":8595,"feed_emoji":"🌀","tokens_out":5110,"duration_ms":47330,"temperature":0.7,"pith_summary":"This paper argues that speckles in turbulent propagation are not just anonymous members of a random intensity field but discrete, trackable objects with well-defined creation and death events. Using simulations of the paraxial wave equation with modified von Kármán turbulence, the authors detect speckles by an intensity threshold and track them plane by plane. They find that speckles are spatially confined and persist over distances significantly longer than their associated Rayleigh length. The persistence statistics shift from a bimodal to a near-exponential distribution as turbulence strengthens, with the speckle count increasing sevenfold and mean lifetime dropping fourfold from weak to strong turbulence. If correct, this means that short-exposure multi-plane measurements of speckle lifetimes could be used to estimate the atmospheric refractive-index structure parameter C_n^2, complementing the usual ensemble intensity statistics.","feed_headline":"Speckles survive many diffraction lengths in turbulent beams","feed_subtitle":"Lifetime statistics of individual speckles scale with turbulence strength, offering a new basis for atmospheric sensing.","key_machinery":"The key machinery is an intensity-based speckle detection and tracking algorithm. Each speckle is identified as a local maximum above η=0.01 Imax, with adjacent or collocated peaks removed by a radius threshold. Persistence Δz_j is the propagation distance between creation and disappearance events, normalized by the speckle's diffraction length z_0,j; confinement r̃_j(z) is the average radius of the 1/e^2 intensity contour. A magnitude-squared coherence C(z) between vacuum- and turbulence-propagated fields defines the decorrelation distance at which fragmentation begins. These definitions convert a continuous random field into discrete object-level statistics.","core_discovery":"The central claim is that individual speckles can be tracked as persistent, localized objects, and that their persistence and confinement are not fixed properties but scale with turbulence strength. The authors report that beam fragmentation coincides with complete spatial decorrelation as measured by magnitude-squared coherence, and that larger beams decohere and fragment closer to the source, indicating that smaller beams retain coherence longer. Past the decorrelation distance, the beam behaves as a collection of statistically independent speckles, some of which survive multiple diffraction lengths: under strong turbulence about 10% of detected speckles propagate at least one diffraction","pith_inferences":["If the persistence statistics are governed by the low-frequency phase structure rather than small-scale intensity fluctuations, then a simpler predictive relation between coherence radius and lifetime distribution may exist; the paper does not test this.","The reported 7× count increase and 4× lifetime reduction may be sensitive to the chosen thresholds; a sweep over η and the radius cutoff would show whether the scaling is a property of turbulent transport or of the detector definition.","The same tracking framework could be applied to experimental data from a high-speed camera at several observation planes in a real atmospheric path; the paper's simulation-based scalings give concrete predictions to compare against.","The connection to localization suggests that speckle persistence might be used to infer not only C_n^2 but also inner scale ℓ_0, since the paper finds ℓ_0 affects the theoretical scintillation index but not the correlation metric."],"forward_implications":["Speckle lifetime distributions can serve as a diagnostic for the effective refractive-index structure parameter C_n^2, using short-exposure multi-plane measurements.","Smaller beams retain partial coherence over longer distances than larger beams, which is directly relevant to beam-size selection for free-space optical links.","Beyond the decorrelation distance, the beam's dynamics become largely independent of initial beam parameters, simplifying models of deep-turbulence propagation.","Persistence and confinement together imply a partial localization of optical energy in substructures of the beam, analogous to localization in disordered dielectrics, which may inform imaging through obscurants."],"fun_headline_variants":["Speckles survive multiple diffraction lengths in turbulence","Smaller beams resist decoherence, speckles persist further","Localized speckles outlast Rayleigh length in turbulent beams","Speckle persistence scales with turbulence strength"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The entire population of tracked speckles is defined by a fixed intensity threshold of 1% of the initial maximum and an unspecified radius threshold that merges nearby peaks; if that definition is changed, the counts, lifetimes, and scaling may not hold.","fun_headline_variants_meta":{"raw":{"variants":["Speckles survive multiple diffraction lengths in turbulence","Smaller beams resist decoherence, speckles persist further","Localized speckles outlast Rayleigh length in turbulent beams","Speckle persistence scales with turbulence strength"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000562,"raw_usage":{"total_tokens":2434,"prompt_tokens":600,"completion_tokens":1834,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":344,"completion_tokens_details":{"reasoning_tokens":1771}},"tokens_in":344,"tokens_out":1834,"duration_ms":16184,"temperature":1.0,"reasoning_tokens":1771,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T22:59:20.488076+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Vary the detection threshold η from, say, 0.001 to 0.05 Imax in the same simulated propagations. If the 7× increase in speckle count and 4× decrease in average lifetime between weak and strong turbulence disappear, or if no speckle survives beyond one diffraction length under strong turbulence, the central claim of turbulence-encoded persistence would be falsified. Alternatively, a field experiment with a high-speed camera at two or more observation planes that finds no persistent localized intensity features under strong turbulence would contradict the result.","supporting_citations":[],"review_version":1}