{"id":"5d8eabaa-51ac-4b2e-97c6-711175f92503","arxiv_id":"2607.20672","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":2.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A review proposing that barchan-barchan and barchan-obstacle interaction outcomes are controlled by Shields and Stokes numbers, transverse position, and size ratio, based on subaqueous grain-scale experiments and simulations.","lead":"This review synthesizes grain-scale experiments and simulations of how crescent-shaped sand dunes (barchans) interact with each other and with obstacles. It proposes that a few dimensionless numbers — Shields and Stokes numbers, transverse position, and size ratio — control the interaction outcomes, with implications for dune fields on Earth and Mars.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central dimensionless-control claim is under-supported: Fig. 4 omits Stokes number for barchan-barchan, Fig. 7e rests on sparse data, and Eq. 3.3 collapses durations only to within a factor of 50; no out-of-sample validation is shown.","rationale":"The reader's CONDITIONAL verdict is appropriate. My stress-test identifies a more fundamental concern than aeolian transferability. The abstract's central claim lists Shields and Stokes numbers, transverse position, and size ratio as the basic controls. However, the evidence for barchan-barchan interactions is the authors' ad hoc θ–ξN maps (Fig. 4), which omit Stokes number entirely; the only map using a Stokes-like parameter is the obstacle map (Fig. 7e), with sparse data and no obstacle-shape coordinate. The collision timescale (Eq. 3.3) collapses data only to a factor of 50 (0.04–2), so it is not a tight scaling. No independent or out-of-sample test of the maps is presented; the authors themselves state that field validation is still necessary (Sec. 5). Because the central claim's predictive power is unestablished even within the subaqueous regime, the aeolian extrapolation is a secondary risk. A decisive test would be new CFD-DEM or experiments at untried (θ, ξN, St) combinations. If the maps fail there, the central claim needs revision; if they pass, the framework gains support. Therefore I agree with the CONDITIONAL verdict, and my concern does not change it.","tokens_in":23225,"tokens_out":8879,"duration_ms":72268,"concrete_test":"Use the CFD-DEM framework of Lima et al. (2024) to run a small grid of aligned barchan-barchan simulations that varies θ and ξN across the Fig. 4 boundaries and also varies St (via grain diameter/density ratio) at fixed θ and ξN. If simulated outcomes at new (θ, ξN) points fall outside the predicted regimes, or if outcomes are insensitive to St, the claimed control by Shields and Stokes numbers is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's central proposal — that barchan-barchan and barchan-obstacle interactions 'depend basically on the Shields and Stokes numbers ... transverse position ... and size ratio' — is not matched by the evidence presented. For barchan-barchan interactions, the only quantitative organizers are the 'ad hoc classification maps' (Sec. 3(a), Fig. 4), whose axes are the Shields number θ and the grain-number ratio ξN, with transverse offset σ as a symbol code; Stokes number does not appear on either axis, and the review reports no experiment or simulation that varies St while holding θ, ξN, and σ fixed. Thus the abstract's inclusion of St as a basic control for barchan-barchan interactions is unsupported. For barchan-obstacle interactions, Fig. 7e does use a modified Stokes number St·H_obst/W_obst, but the map is built from one group's experiments plus three simulation points, and the obstacle shape (cylinder vs block, Fig. 7a-d) is not a coordinate; no evidence shows shape is irrelevant. The only quantitative timescale, Eq. 3.3, is said to normalize measured collision durations to values between 0.04 and 2 (Sec. 3(a)(ii)) — a factor of 50 spread that is too broad for a predictive scaling and that the text does not explain. No independent data set is used to test the maps; the conclusions themselves call for 'systematic field-based validation of predictive maps derived from subaqueous experiments.' If the maps do not predict outcomes for new subaqueous parameter combinations, the central claim fails even before the aeolian extrapolation is considered.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This review synthesizes experimental and numerical work on binary barchan-barchan and barchan-obstacle interactions, focusing on grain-scale subaqueous studies (water flumes, annular channels, CFD-DEM) and placing them in the context of remote sensing and aeolian dune fields. The central proposal is that interaction outcomes are controlled by a small set of dimensionless parameters: Shields number, Stokes number, transverse offset, and size ratio. The paper presents ad hoc classification maps (Fig. 4 for barchan-barchan; Fig. 7e for barchan-obstacle), a collision timescale (Eq. 3.3), and detailed discussions of flow structures, grain trajectories, and forces. It ends with caveats about extrapolating subaqueous results to aeolian environments and with a call for systematic field-based validation.","tokens_in":23653,"tokens_out":7576,"duration_ms":58966,"significance":"The synthesis is useful: it gathers recent grain-scale experiments and CFD-DEM simulations, documents mechanisms (wake-induced repulsion, vortex-driven trapping, force distributions), and carefully states limitations, e.g. 'should be interpreted with caution' (Section 3(a)(i)) and 'extrapolations ... must be carried out with caution' (Section 3(b)). If the proposed dimensionless organization were established, it would provide practical, falsifiable tools for predicting collision and obstacle-interaction outcomes. However, the manuscript's central claim is stronger than the evidence it assembles: no independent validation is shown, the barchan-barchan maps omit Stokes number, the obstacle map rests on sparse data, and the timescale collapse spans a factor of 50. These are not mere presentation issues; they bear directly on the abstract's proposal. The review's contribution as a summary of mechanisms is significant, but its predictive framework is not yet supported.","major_comments":[{"comment":"The abstract states that barchan-barchan and barchan-obstacle interactions 'depend basically on the Shields and Stokes numbers ... transverse position ... and size ratio.' For the barchan-barchan case, however, the only quantitative organizers shown are the θ–ξ_N maps in Fig. 4, with the transverse offset σ encoded by symbols; Stokes number appears nowhere and no experiment or simulation is cited that varies St while holding θ, ξ_N, and σ fixed. Thus the role of St in barchan-barchan interactions is unsupported by the review's own evidence. This is load-bearing for the central proposal; either supply such evidence or narrow the claim.","section":"Abstract; Sec. 3(a)(i), Fig. 4"},{"comment":"The barchan-obstacle map is built from the experiments of Assis et al. [11] plus three CFD-DEM points from Lima et al. [70]. It uses the modified Stokes grouping St H_obst/W_obst and the size ratio H_obst/W, but obstacle shape (cylinder vs block; Fig. 7a-d) is not a coordinate and no test of shape irrelevance is reported. With this data density and an ad hoc dimensionless group, the statement in Sec. 4 that these maps 'provide a predictive framework' is not supported. The later call for 'systematic field-based validation' (Sec. 5) is appropriate but underscores the gap.","section":"Sec. 4, Eq. (4.1), Fig. 7e"},{"comment":"The timescale t_s is proposed as a scaling framework for collision duration, yet the measured durations divided by t_s vary between 0.04 and 2—a factor of 50. A collapse spanning two orders of magnitude does not validate a scaling law unless the residual is shown to be random and small relative to the variations in the input parameters. The text neither explains nor analyzes this spread, so the assertion that t_s allows comparison 'across different sediment compositions' is not yet demonstrated.","section":"Sec. 3(a)(ii), Eq. (3.3)"},{"comment":"Neither classification map nor the timescale is tested against an independent data set; each is constructed from the same data it organizes. Given that the proposed controls are partly ad hoc and that the conclusions explicitly call for validation, the abstract's 'we propose' should be framed as a hypothesis to be tested rather than as an established result. This is fixable by reframing, or by adding an explicit re-analysis of an independent subaqueous or field data set.","section":"Sec. 5; Overall"}],"minor_comments":[{"comment":"Typo: 'layouts ot the four devices' should be 'layouts of the four devices'.","section":"Sec. 2(b), Fig. 1 caption"},{"comment":"Typo: 'apace' should be 'space'.","section":"Fig. 7e caption"},{"comment":"The symbols D_t and D_i are used but never defined; \\bar d is the mean grain diameter, so D likely denotes dune length/width. Please define them.","section":"Eq. (3.3)"},{"comment":"The column heading 'Duran et al. [34]' seems mismatched: the row entries correspond to Durán et al. 2005 (ref. [33]) on breeding and solitary-wave behavior; ref. [34] is the 2011 size-distribution paper. Please verify.","section":"Table 3"},{"comment":"The degree of diffusivity \\Phi_B is introduced but not defined or used; either define it or remove it.","section":"Sec. 3(a)(i), Fig. 4f"},{"comment":"Typo: 'most of the the results' should be 'most of the results'.","section":"Sec. 3(b)"},{"comment":"Reference [1] formatting: 'Copernicus Browserhttps://...' is missing a space.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The central framework is derived largely from the authors' own prior publications (Assis & Franklin 2020, 2021; Assis et al. 2022, 2023; Lima et al. 2024, 2025). This is not a problem per se, but it makes the absence of independent validation more consequential. A revised version that explicitly positions the dimensionless-control claim as a hypothesis, with a quantitative assessment of the maps' predictive power, would be appropriate for this journal. The paper's scope fits a review venue; the issue is the gap between the abstract's claim and the evidence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a review, not a new-results paper, and it is a good review in many ways. It synthesizes a decade of grain-scale subaqueous experiments and CFD-DEM simulations on barchan-barchan and barchan-obstacle interactions, mostly from the authors' own group, and it is explicit about the limits of transferring those results to aeolian conditions. The central proposal—that interaction outcomes organize around the Shields number, a Stokes-type number, transverse offset, and size ratio—is a reasonable synthesis, but it is not tested in this paper. The classification maps are the authors' earlier ad hoc maps, and the timescale collapse only brings measured durations into a factor-of-fifty band. If you read it as a review with a suggestive framework, it is useful; if you read the framework as established, it overreaches.\n\nWhat the paper does well: it gives a clear taxonomy of barchan-barchan outcomes (merging, exchange, fragmentation-exchange, fragmentation-chasing, chasing) and includes a nomenclature table that maps terms across studies—that table alone is worth the price. The barchan-obstacle section is also solid, bringing together Assis et al. 2023 and Lima et al. 2025 CFD-DEM work, including force and grain-trajectory data. The authors are disciplined with caveats: they call the classification maps ad hoc, flag the quasi-2D stability result as needing caution, and twice warn that subaqueous-to-aeolian extrapolation must be made with caution. That honesty is genuine and should be credited.\n\nSoft spots, in order of importance. First, the abstract's claim that interactions 'depend basically on the Shields and Stokes numbers' is stronger than the evidence inside the review. For barchan-barchan, the maps use θ and grain-number ratio; Stokes number does not appear as a coordinate, and no experiment or simulation varies St while holding the others fixed. For barchan-obstacle, the map does use a modified Stokes number, but the data are one group's experiments plus three simulation points, and obstacle shape is not a coordinate. So the dimensionless framework is plausible but under-supported. Second, Eq. (3.3) is presented as a timescale collapse but normalized durations span 0.04 to 2—a factor of fifty. The text does not explain this spread or show it is a scaling success. Third, the aeolian transfer problem is real: bedload and saltation produce different grain-following behavior, and the authors themselves say direct extrapolation is unsafe. The review is honest about this but does not resolve it.\n\nWho this is for: students and researchers new to barchan morphodynamics who want a compact, well-referenced overview of grain-scale experiments and simulations, plus the group's own synthesis. Not the place to find a validated predictive law.\n\nRecommendation: yes, send it to peer review. A serious referee will press on the missing Stokes-number evidence and the broad timescale scatter, but the review is organized, honest, and worth airing.","headline":"A competent, honest review of the authors' own grain-scale work; the proposed dimensionless framework is suggestive but not yet validated.","tokens_in":24196,"tokens_out":3099,"would_cite":true,"duration_ms":26348,"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":"This paper argues that barchan interactions are governed by Shields and Stokes numbers, transverse position, and size ratio, and that grain-scale flow mechanics explains each outcome.","keywords":["barchan dunes","dune-dune interactions","dune-obstacle interactions","Shields number","Stokes number","grain-scale experiments","CFD-DEM simulations","dune migration"],"falsifier":"A satellite-based census of aeolian barchan collisions (e.g., in the Taklimakan or Bodélé deserts) that plots outcomes on the θ–ξN map and finds regime boundaries that differ systematically from the subaqueous map would falsify the claim that this small parameter set governs interactions across environments.","tokens_in":23082,"feed_emoji":"🏜️","tokens_out":5643,"duration_ms":52159,"temperature":0.7,"pith_summary":"Barchan dunes—crescent-shaped sand dunes found on Earth, Mars, and beyond—rarely act alone; they collide, chase, merge, split, and meet obstacles such as buildings, bridge piers, and crater rims. This review synthesizes grain-scale water experiments and grain-resolved CFD-DEM simulations to argue that these varied outcomes are controlled by a small set of parameters: the Shields number (flow entrainment relative to grain resistance), the Stokes number (how closely grains follow the flow), the transverse offset of the interacting dunes, and the size ratio between the interacting objects. It provides classification maps in these dimensionless spaces for both dune-dune and dune-obstacle encounters, plus a timescale for collisions, and traces each outcome to concrete flow mechanisms such as wake-induced erosion and obstacle-dune vortices. The authors are careful to warn that these subaqueous findings should be extrapolated to aeolian dunes only with caution, because wind-driven saltation differs fundamentally from bedload transport.","feed_headline":"Four numbers predict how barchan dunes interact","feed_subtitle":"Water-tank tests map collision and obstacle outcomes onto dimensionless charts, offering a basis for desert and Mars dune forecasting.","key_machinery":"The load-bearing tools are the two ad hoc classification maps, built from water-channel experiments and four-way CFD-DEM simulations that track individual grains. For dune-dune interactions, the map plots the Shields number θ against the dimensionless grain-number difference ξN (Eq. 3.1), with off-center collisions additionally tagged by the offset parameter σ. For dune-obstacle interactions, the map plots a modified Stokes number St·H_obst/W_obst against the obstacle-to-dune height ratio H_obst/W. The paper also derives a collision timescale ts (Eq. 3.3) that scales interaction duration with dune separation, relative celerity, grain size, and density. The mechanism that explains chasing and","core_discovery":"On the paper's own terms, the central claim is that barchan-barchan and barchan-obstacle interactions depend essentially on the Shields and Stokes numbers, the transverse position of bedforms, and the size ratio between the interacting objects. For binary barchan encounters, the authors identify five collision patterns—merging, exchange, fragmentation-exchange, fragmentation-chasing, and chasing—organizing them in maps of the Shields number θ versus the grain-number difference ratio ξN. For obstacles, three regimes (pass-over, bypass, trapped) are organized in a map of a modified Stokes number versus the obstacle-to-dune height ratio. The paper's grain-scale experiments and simulations show","pith_inferences":["If the framework extends to aeolian dunes, the much higher grain inertia (Stokes number roughly a thousand times larger) implies that vortex-driven bypass and trapped void regions will be far weaker; saltating grains should impact obstacles more directly, so the subaqueous maps likely bound rather than pinpoint aeolian behavior.","A concrete test: plot collision outcomes from multidecadal satellite image sequences of desert barchans on the θ–ξN maps to see whether the same regime boundaries emerge despite the transport-mode difference.","The parameter-sparse description suggests agent-based dune-field models could adopt these grain-scale-informed rules directly, making large-scale simulations consistent with grain-scale physics."],"forward_implications":["Interaction outcomes become predictable from a few measurable flow and geometry parameters, replacing case-by-case empiricism.","The obstacle classification map gives engineers a first-pass tool for whether a migrating dune will pass over, bypass, or be trapped by infrastructure such as bridge piers or buildings.","The collision timescale equation lets interaction durations be compared across different grain-size mixtures, a step toward interpreting aeolian dune-field images.","Repeated collisions should drive dune corridors toward chasing-dominated, size-selected configurations, consistent with field observations.","The framework gives a concrete way to extrapolate to Mars, where crater rims act as obstacles, provided the aeolian caution is respected."],"fun_headline_variants":["Five collision modes emerge from two dune numbers","Shields and Stokes numbers map dune collision types","Grain-scale physics predicts barchan dune encounters","Water-tank dunes reveal collision and obstacle rules","How barchans interact: a grain-scale guide"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing assumption is that results from subaqueous bedload experiments and simulations transfer to aeolian saltation-driven barchans; the paper itself states that this extrapolation must be made with caution.","fun_headline_variants_meta":{"raw":{"variants":["Five collision modes emerge from two dune numbers","Shields and Stokes numbers map dune collision types","Grain-scale physics predicts barchan dune encounters","Water-tank dunes reveal collision and obstacle rules","How barchans interact: a grain-scale guide"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000245,"raw_usage":{"total_tokens":1363,"prompt_tokens":724,"completion_tokens":639,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":468,"completion_tokens_details":{"reasoning_tokens":564}},"tokens_in":468,"tokens_out":639,"duration_ms":5973,"temperature":1.0,"reasoning_tokens":564,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T09:40:28.952648+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A satellite-based census of aeolian barchan collisions (e.g., in the Taklimakan or Bodélé deserts) that plots outcomes on the θ–ξN map and finds regime boundaries that differ systematically from the subaqueous map would falsify the claim that this small parameter set governs interactions across environments.","supporting_citations":[],"review_version":1}