{"id":"1f8f28f8-8789-4540-9eee-856c0d906e35","arxiv_id":"2411.17813","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A new 'planarity' metric shows the Milky Way's satellite positions are unusually flat but its velocities are inconclusive, and Lambda-CDM simulations routinely produce such kinematically supported planes.","lead":"This paper introduces a new way to measure how flat the arrangement of a galaxy's small companion galaxies is, and applies it to the Milky Way and to a cosmological simulation. It finds the Milky Way's satellites are positionally flat, but current velocity data cannot confirm the plane is rotating; simulated Milky Way-like galaxies show such flat, rotating planes are common, so the observations do not contradict standard cosmology.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Kinematic coherence in NewHorizon is asserted from aggregate Gini percentiles, not from same-host overlap of position and velocity plane spaces; the central LCDM-comparison claim therefore lacks its stated quantitative support.","rationale":"I focused on the kinematic-coherence claim because it is the load-bearing pillar that connects the simulation comparison to the abstract's conclusion, and because it is internally testable from the published code and NewHorizon outputs. The MW positional planarity result is strong (100th percentile against 1000 isotropic replications) and the velocity null is honestly stated. The novel plane-space metric is well motivated and the code is released. The concern is not that NewHorizon is the wrong simulation or that a single simulation cannot be used; it is that the specific statistic presented in Figure 6 does not measure what the paper says it measures. High Gini percentiles for positions and velocities in separate, aggregate samples do not establish that the same galaxies have coherent position-velocity plane structure. The paper's own Section 5 criterion for kinematic support is spatial correspondence of the plane-space maps, and no such comparison is reported for the simulation. This is an addressable gap: a same-host cross-tabulation of plane-space overlap, or a quantitative comparison of peak-bin locations, would settle it. If the overlap is high, the conditional verdict can be lifted; if not, the abstract's kinematic claim would need to be weakened. This does not change the reader's CONDITIONAL verdict, because the gap is real but fixable and the underlying MW-versus-LCDM positional comparison may still survive. The forward-modeling issue raised by the reader is also relevant but secondary: even a perfect forward model would not establish kinematic support unless the same-host plane-space correspondence is demonstrated for the simulated galaxies.","tokens_in":15303,"tokens_out":12931,"duration_ms":125173,"concrete_test":"For each NewHorizon host in the Section 7.1 sample and each snapshot, compute the full m by m plane-space histograms for positions and for velocities. Define a same-host coherence score, e.g., the cosine similarity or Spearman correlation between the flattened position and velocity histograms. Build a null distribution by shuffling velocities among satellites at fixed positions and recomputing the score. If the median coherence score of simulated hosts is not significantly above the shuffled-null median, or if fewer than about 50 percent of hosts show a score above the 95th percentile of the null, the claim that kinematically supported planes are common in NewHorizon is not supported. Also report the fraction of hosts that simultaneously exceed the 100th percentile in position, velocity, and angular momentum, to check whether the aggregate fractions in Figure 6 come from the same hosts.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Section 5 the paper defines kinematic support as correspondence between the position and velocity plane-space maps for the same satellite system. The simulation analysis in Section 7.1, however, does not measure this correspondence. Figure 6 and the surrounding text report only the fraction of hosts whose position, velocity, and angular-momentum Gini coefficients individually exceed the 100th percentile of the isotropic distribution. These are marginal, aggregate statistics. A host can have a high velocity Gini percentile while its velocity plane-space concentration lies in a different region from its positional concentration, and the reported 90% and 80% fractions need not be realized by the same hosts. The statement that 'positions, velocities and angular momenta are highly correlated, demonstrating kinematic coherence' is not backed by any same-host cross-tabulation, correlation statistic, or comparison of the actual 2D plane-space maps. Since the MW velocity plane space is itself not significantly non-random (Section 6), the paper's key supporting claim that MW-like LCDM galaxies have kinematically supported planes rests on this unsupported inference. If the apparent agreement of the curves in Figure 6 is due to different hosts, or if position and velocity plane-space peaks do not coincide for typical hosts, the conclusion that the MW's observed planarity is consistent with LCDM loses the kinematic pillar on which the abstract explicitly relies.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a new 'plane space' representation for galaxy satellite systems: each pair of satellite vectors (position, velocity, or angular momentum) defines a plane through the host, the planes are binned into an m x m histogram in spherical coordinates, and the Gini coefficient of the histogram is used as a scalar 'planarity' measure. The authors apply this to 46 Milky Way (MW) satellites using Gaia EDR3 data from Li et al. (2021), propagating measurement uncertainties with 1000 stochastic replications. They find that MW positional planarity is highly significant (Gini in the 100th percentile of an isotropic comparison), while velocity planarity is not statistically significant. They then apply the same method to MW-like hosts in the NewHorizon cosmological simulation, reporting that at least 90 per cent of hosts (or 80 per cent when tracked across time) exhibit positional planarity comparable to or greater than the MW, and they claim that positions, velocities, and angular momenta are 'highly correlated, demonstrating kinematic coherence.' The paper concludes that the MW satellite plane structure is not in tension with the standard ΛCDM paradigm.","tokens_in":15562,"tokens_out":6410,"duration_ms":58846,"significance":"If substantiated, the plane-space method would be a useful new tool for satellite-galaxy studies, and the MW positional planarity result would reinforce the growing body of work showing that prominent satellite planes are not rare in ΛCDM. The stochastic treatment of Gaia errors is careful, and the authors are appropriately cautious about the MW velocity non-detection, attributing it to large measurement errors. The central weakness is that the kinematic-coherence claim for the simulation rests on aggregate Gini percentiles rather than a same-host comparison of position and velocity plane spaces, and no forward-modeling of Gaia-like velocity errors is applied to simulated satellites. These gaps are fixable and do not undermine the well-supported positional-planarity conclusion.","major_comments":[{"comment":"The text states that 'positions, velocities and angular momenta are highly correlated, demonstrating kinematic coherence,' but the analysis only reports the fraction of hosts whose position, velocity, and angular-momentum Gini coefficients individually exceed the 100th percentile of the isotropic distribution. These are marginal statistics; they do not establish that the same hosts are planar in all three spaces, nor that the plane-space peaks coincide. A host can have a high velocity Gini percentile while its velocity concentration lies in a different region of plane space from its positional concentration. The authors should provide a same-host cross-tabulation (e.g., the fraction of hosts satisfying position and velocity thresholds jointly) or a direct comparison of the 2D plane-space maps (e.g., bin-wise correlation or mutual information). Without this, the kinematic-coherence claim, which the abstract and conclusions explicitly rely on, is not quantitatively supported.","section":"Section 7.1, Figures 6 and 7"},{"comment":"The comparison between the MW and NewHorizon treats simulated velocities as noiseless, while the MW velocity vectors have substantial Gaia EDR3 uncertainties as modelled in Section 3. The conclusion that 'as MW satellite galaxy velocity measurements improve, kinematic support may be confirmed' requires that the simulated kinematic signal survive realistic velocity errors. The paper does not degrade NewHorizon velocities with typical Gaia-like uncertainties, so it is not established that the kinematic coherence seen in simulations would be observable in current or near-future MW data. Please add a forward-modeling test in which simulated velocity vectors are perturbed with error distributions comparable to those in Li et al. (2021), and report the resulting Gini percentiles for the simulation hosts.","section":"Section 7.1 and Section 3"},{"comment":"The free parameter m (the histogram resolution) is set to 25 and described as 'calibrated by eye,' with the claim that 'the overall analysis is not sensitive to the precise value of m (for example, 20 < m < 30)' being asserted without quantitative support. Since the Gini coefficient and its isotropic percentile depend on the binning, the paper should report the key results (e.g., the MW position Gini percentile and the 90% and 80% host fractions in Figures 6 and 7) for several values of m, ideally in a small table or figure. Without this, the reader cannot assess whether the headline conclusions are an artifact of the chosen resolution.","section":"Section 5 (Plane space approach)"}],"minor_comments":[{"comment":"The statement that 'around 35 per cent of the as-is distribution overlaps with the shuffled scenario' should specify how the overlap is measured (e.g., fraction of as-is replications within the shuffled 95 per cent interval, or area overlap of the two kernel density estimates).","section":"Section 4, Figure 2"},{"comment":"The text says that isotropic percentiles are established 'for any given number of satellite galaxies,' but Section 7 does not explicitly state that each NewHorizon host's Gini percentile is computed with an isotropic distribution matching that host's satellite number. Please clarify this in the simulation analysis, especially because host sample sizes vary with time in Figures 6 and 7.","section":"Section 5.1"},{"comment":"The text quotes thresholds such as 'more than 99 per cent of the position vector replications' and 'only 20 per cent of the velocity vector replications' relative to the 95th percentile of the isotropic distribution; marking the 95th percentile on Figure 5 would make these comparisons directly readable.","section":"Section 6, Figure 5"},{"comment":"The authors acknowledge that planes are assumed to pass through the host centre and that this may lower Gini values for off-centre planes; a simple quantitative test in the simulation (e.g., fitting planes with a free offset and comparing the resulting planarity statistics) would strengthen this methodological choice.","section":"Section 5 (plane-through-origin assumption)"},{"comment":"The reference list contains duplicate entries: Samuel et al. 2021a and 2021b both cite MNRAS, 504, 1379, and Sawala et al. 2023a and 2023b both cite Nature Astronomy, 7, 481; these should be corrected or merged.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for MNRAS and the positional-planarity result is solid and honestly presented. The main gap is the kinematic-coherence claim in the simulation, which currently rests on marginal rather than joint statistics, and the absence of error forward-modeling for simulated velocities. Both are fixable within the manuscript's scope. I would be happy to review a revised version that adds a same-host cross-tabulation or map-overlap statistic and a velocity-error degradation test."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the plane space / Gini metric is a real addition to the toolbox, and the MW positional result is solid. The simulation side needs more work before the 'kinematically supported planes are common' claim is convincing.\n\nWhat's actually new: the plane space construction from all satellite pairs, summarized by a Gini coefficient against an isotropic null, is not in the cited prior work. It's a useful way to measure aggregate planarity without committing to a single plane, and positions, velocities, and angular momenta can be treated separately. The paper ships code and uses the Li et al. EDR3 data in a careful way (1000 replications, skew-normal errors).\n\nThe MW positional planarity is the strongest part: 99%+ of replications have Gini above the 95th percentile of isotropic, the median image is concentrated, and the authors do not overclaim the velocity null. They say plainly that kinematic coherence cannot be confirmed from current data. That is honest.\n\nSoft spots: the simulation comparison is where it gets shaky. The 'kinematic coherence' claim in Sec 7.1 is based on the fraction of hosts whose position, velocity, and angular momentum Gini percentiles individually exceed the 100th percentile threshold. That is an aggregate, marginal statistic. It does not show that the same host has concentrated position and velocity plane spaces in the same region of the plane space, which is what 'kinematically supported' means by the paper's own definition. Figure 8 shows four example hosts with qualitatively tracking curves, but that is anecdotal. Without a same-host cross-tabulation or a correlation statistic between position and velocity Gini percentiles (or better, between the plane-space maps), the claim that planes in NewHorizon are commonly kinematically supported is not demonstrated. This matters because the abstract leans on it.\n\nAlso, the simulated hosts are noiseless, while the MW velocities have large errors. The paper acknowledges this and argues that improved measurements may confirm kinematic support, but that's a hypothesis, not a test. Forward-modeling the Gaia errors onto the simulation would be the natural fix.\n\nMinor: m=25 is calibrated by eye, though the authors say the analysis is insensitive to 20