{"id":"cdbdae45-f83a-4417-ac51-82a5a5b8b079","arxiv_id":"2509.02546","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":10,"one_line_summary":"Boötes I shows a shallow 4σ intrinsic velocity gradient aligned with its orbit, a negative metallicity gradient, and a dark matter profile that weakly favors a cusp over a core.","lead":"This paper combines new and archived spectra of the ultra-faint dwarf galaxy Boötes I to build the largest kinematic and metallicity sample in such a galaxy, with 148 member stars. It reports a 4σ velocity gradient along the galaxy's orbit after removing perspective rotation, plus updated dark matter and chemical evolution constraints.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Headline 4σ significance not supported by reported uncertainties; intrinsic-gradient detection needs a proper null test against the perspective template.","rationale":"The paper's central contribution is the first perspective-rotation-corrected velocity gradient in a UFD, claimed at 4σ. I read §5.1 as a subtraction of a predicted rigid-body perspective field from a GMM-fitted heliocentric gradient. The reader's conditional verdict is reasonable: the combined dataset is carefully assembled, the individual datasets are internally consistent, and the N-body mock provides some validation of the perspective template. However, the statistical claim in the abstract is not derivable from the reported posterior: the intrinsic gradient components in Table 3 imply only ~2.8σ against the no-intrinsic-gradient null if the perspective template is fixed, and no model comparison is reported. The 99% bootstrap statement in §5.1 concerns the observed gradient under perspective rotation, not the significance of the residual after subtraction. Additionally, the correction's uncertainty budget omits systematic differences among published proper-motion measurements (e.g., Li et al. 2021) and possible spatial variations in membership, selection, and binary masking. These issues do not prove the detection false, but they mean the headline overstates the evidence. A proper null test and a robustness check against an independent proper motion would settle whether the detection is 4σ or ~3σ. The Jeans and GCE results are explicitly assumption-limited by the authors (§6.1, Appendix D), so they do not independently strengthen the headline. The conditional verdict remains appropriate; no adjustment is needed.","tokens_in":47755,"tokens_out":18257,"duration_ms":173880,"concrete_test":"From the GMM posterior samples, compute the Savage-Dickey Bayes factor (or a likelihood-ratio p-value) for the model with a free intrinsic gradient versus the model with only the perspective-rotation template, with the perspective amplitude and direction sampled from priors based on the adopted Gaia proper motion and on the independent Li et al. (2021) proper motion. Report the resulting significance; if it is not ≥4σ in both cases, the '4σ' headline should be downgraded.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's '4σ' detection is not traceable to the numbers reported in §4.4/§5.1. The intrinsic velocity gradient is a vector with components Δv1'=−0.09±0.03 and Δv2'=−0.08±0.05 km/s/arcmin (Table 3). Against the null that only the rigid-body perspective template is present (predicted vector ~0.105 along the proper-motion axis), the observed residual vector has χ²≈(0.085/0.03)²+(0.08/0.05)²≈10.6 for 2 dof, i.e., p≈0.005 (~2.8σ), not 4σ, absent an unreported covariance matrix. The '1.2±0.3' in the abstract is a ratio of a positive-definite magnitude to one error bar, not a formal significance. No likelihood-ratio or Bayes-factor test for the zero-intrinsic-gradient model is presented. The bootstrap test quoted in §5.1 calibrates only how often perspective rotation produces a nonzero observed gradient, which is expected 99% of the time under the null, not the significance of the residual after subtraction. In addition, the subtraction is conditional on the adopted Gaia proper motion and distance; the quoted correction uncertainty (0.004) propagates random errors only and does not include the documented discrepancy with Li et al. (2021) or possible spatial variations in membership, selection, and binary masking. If the true significance is ~3σ and/or the residual shifts under an independent proper motion, the central claim is overstated.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper combines new S5 AAT spectroscopy with archival AAT, MMT, and VLT data to build a 148-member sample of the ultra-faint dwarf galaxy Boötes I, identifies 15 new binary candidates from a 16-year baseline, and fits systemic properties with a Gaussian mixture model. The headline results are an intrinsic line-of-sight velocity gradient of 1.2±0.3 km/s/rh after correcting for solid-body perspective rotation, a resolved metallicity gradient of −0.10±0.02 dex/rh, a dark matter inner slope γ=1.0+0.5/−0.6 from axisymmetric Jeans modeling, and one-zone chemical evolution parameters favoring rapid (τSFH≈0.2 Gyr), inefficient star formation with a large mass-loading factor (η≈200). The paper explicitly discusses tensions with the previous steep velocity gradient of Longeard et al. (2022), compares the observed gradient with an N-body simulation, and candidly lists several modeling limitations.","tokens_in":48184,"tokens_out":4562,"duration_ms":46541,"significance":"If correct, the paper would provide the first perspective-rotation-corrected velocity gradient in a Milky Way ultra-faint dwarf, based on the largest kinematic and metallicity sample in any such galaxy, and would add meaningful constraints on Boötes I's dark matter profile and chemical evolution. The strengths include the homogeneous S5 data reduction, careful binary identification, explicit cross-checks across four independent datasets, an N-body mock with a perspective-rotation baseline, and a publicly available combined catalog. These are substantial contributions. However, the headline 4σ detection is not supported by the reported component uncertainties, the significance test for the intrinsic gradient is not properly formulated, and some of the quoted chemical evolution constraints are explicitly prior-dominated rather than independently derived from the new MDF. These issues are fixable but require revision, so the paper should not be accepted in its current form.","major_comments":[{"comment":"The claimed \"4σ\" detection of the intrinsic velocity gradient is not traceable to the reported component uncertainties. With Δv'_1=−0.09±0.03 and Δv'_2=−0.08±0.05 km/s/arcmin, the residual vector against the null of zero intrinsic gradient has χ²≈(0.09/0.03)²+(0.08/0.05)²≈11.6 for 2 degrees of freedom before accounting for any correlation, i.e. p≈0.003, roughly a 3σ signal rather than 4σ. Quoting the ratio of the positive-definite magnitude 0.12 km/s/arcmin to one error bar is not a valid significance for a two-component vector. The authors should report the full covariance matrix of the inferred gradient components and present a likelihood-ratio or Bayes-factor test comparing the zero-intrinsic-gradient model with the model including an intrinsic gradient.","section":"Abstract; §5.1 and Table 3"},{"comment":"The bootstrap test quoted in §5.1 calibrates the distribution of the observed gradient under the null that only perspective rotation is present, finding that a non-zero observed gradient is recovered 99% of the time; this is expected under that null and does not test whether the residual after subtracting the rigid-body template is significant. The paper needs a null test defined on the residual: e.g. posterior predictive draws of the perspective-only model compared to the observed residual vector. In addition, the quoted correction uncertainty of 0.004 km/s/arcmin appears to propagate only random errors in the adopted proper motion and distance; the documented discrepancy with Li et al. (2021) noted in §4.4 should be propagated or tested, since the intrinsic gradient is defined relative to the adopted rigid-body template.","section":"§5.1, bottom panels of Figure 6"},{"comment":"The prior listed for log10ρ0 in Table 4 is U(−5,−5), which is a delta function rather than the broad prior described in the text; if taken literally, this prior would make the reported Jeans constraints on ρ0 and γ impossible. This appears to be a typographical error, but it must be corrected and the intended bounds stated explicitly, because the dark matter inference in Section 6 depends directly on this prior.","section":"Table 4, Section 6"},{"comment":"The paper explicitly states in §7 that the inference of τSFH=0.2±0.1 Gyr is \"dominated by the tight priors set by CMD-based SFH\" from Durbin et al. (2025). The abstract and summary nevertheless present the rapid star formation timescale as a new constraint from the chemical evolution analysis. This overstates the independent information in the MDF; the authors should either refit with a broader prior and report the sensitivity, or rephrase the abstract and summary to state that the MDF is consistent with, but does not independently determine, the short star formation timescale.","section":"§7 and Abstract"}],"minor_comments":[{"comment":"There is a duplicated word in the sentence \"consistent consistent with expectations from cold dark matter cosmology\" that should be corrected.","section":"§6"},{"comment":"The text reports the proper-motion discrepancy with Li et al. (2021) but does not quantify how that alternate proper motion would shift the perspective-rotation template of 0.105 km/s/arcmin in §5.1; adding this one-number sensitivity estimate would greatly strengthen the paper.","section":"§4.4 and Table 3"},{"comment":"The discussion of tidal versus rotational origin is appropriately cautious, but the sentence \"To confidently discrimination between rotational ... and tidal origins\" contains a grammatical error that should be fixed.","section":"§5.1.3"}],"recommendation":"major_revision","confidential_remarks":"The abstract's 4σ claim is likely to be the most visible result, and as it stands it is not supported by the quoted component errors or by a proper significance test. I would require the significance analysis to be redone before publication. The paper otherwise contains valuable data and careful multi-dataset cross-checks, and the issues appear addressable within the scope of a revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this is a solid observational paper with a valuable dataset, and the central claim about Boötes I's velocity gradient is probably in the right direction. The steep gradient reported by Longeard et al. becomes shallow once perspective rotation is removed; that qualitative revision is credible. Second thing: the abstract's \"4σ\" is not backed by the numbers in the text. The stress-test note mostly lands. The residual intrinsic-gradient components from Table 3 are −0.09 ± 0.03 and −0.08 ± 0.05 km s−1 arcmin−1; with independent errors that is roughly 2.8σ, and no likelihood-ratio or Bayes factor against the zero-intrinsic-gradient model is presented. The quoted bootstrap—that perspective rotation alone produces a nonzero observed gradient 99% of the time—does not calibrate the significance of the residual after subtraction. So the headline detection, as advertised, is not yet established, even though the observed gradient being much shallower than Longeard et al. is credible.\n\nWhat is solid: the combined sample (148 members, 115 clean velocities, 92 [Fe/H], 15 new binaries) is the largest kinematic and metallicity sample in any UFD and will be a reference dataset. Re-reducing archival AAT data with the S5 pipeline and publishing the combined catalog is concrete, reproducible progress. The metallicity gradient (−0.10 ± 0.03 dex per half-light radius) is a clean confirmation of earlier work. The paper is also honest about limitations: it admits τSFH is prior-dominated, the dark matter profile is weakly constrained, and the tidal-versus-rotational origin of the gradient is unsettled. The N-body mock is a useful sanity check, though the fact that their fiducial simulation does not reproduce the claimed intrinsic gradient deserves more discussion than it gets.\n\nWhere it is soft, in proportion: the missing significance test is the main issue. The perspective-rotation correction depends on the adopted Gaia proper motion and distance and on a rigid-body model; the quoted correction uncertainty (0.004 km s−1 arcmin−1) propagates random errors only, not the documented proper-motion discrepancy with Li et al. (2021) or possible selection and binary-masking asymmetries. That matters because the claimed residual is only around 3σ. The Jeans γ ∼ 1 result is weak and the paper says so; it should not be oversold. The GCE mass-loading factor η ∼ 200 is model-dependent and strongly tied to adopted yields, as the text notes. Citation patterns look fair; the direct engagement with the Longeard discrepancy is testable and useful.\n\nWho this is for: UFD observers and modelers, especially those working on cusp-core tests and tidal disruption. It deserves peer review—an editor should send it out. The referees should require a proper null test for the intrinsic gradient, the full covariance matrix of the gradient components, propagation of proper-motion and distance systematics, and an abstract that matches the actual significance. If the significance settles at ~3σ, it is still a valuable paper; it just is not a 4σ detection.","headline":"A genuinely useful dataset and a likely real revision of Boötes I's velocity gradient, but the headline 4σ is not supported by the reported statistics and needs a proper null test before it can be quoted.","tokens_in":48773,"tokens_out":2746,"would_cite":true,"duration_ms":29271,"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":"This paper claims that after removing the perspective-rotation signature of a moving rigid body, Boötes I retains a 4σ, orbit-aligned line-of-sight velocity gradient, along with a resolved metallicity gradient and a dark matter profile…","keywords":["Boötes I","ultra-faint dwarf galaxy","velocity gradient","perspective rotation","dark matter density profile","galactic chemical evolution","stellar kinematics","metallicity gradient"],"falsifier":"Re-fit the same 148-star catalog with an independent proper motion measurement for Boötes I and with membership probabilities left free; if the claimed 1.2 km s$^{-1} r_h^{-1}$ residual gradient does not survive in magnitude and direction, it was an artifact of the rigid-body perspective correction. A companion test is to search for extra-tidal members beyond roughly 8–13 half-light radii, the tidal radius estimated in the paper, whose presence or absence would distinguish tidal stripping from internal rotation.","tokens_in":47577,"feed_emoji":"🌌","tokens_out":8328,"duration_ms":72733,"temperature":0.7,"pith_summary":"Boötes I is one of the closest and best-studied ultra-faint dwarf galaxies, yet its dynamics and chemistry have been read from samples that barely reach one half-light radius. This paper merges about 15 years of archival spectra with new S5 observations to build the largest sample of member stars in any Milky Way ultra-faint dwarf: 148 members extending to roughly 7 half-light radii. Its central claim is that once the geometric 'perspective rotation' of a moving rigid body is subtracted — something never before done for an ultra-faint dwarf — Boötes I still shows a real line-of-sight velocity gradient of 1.2 km/s per half-light radius, aligned with its orbit. If true, the gradient is a fossil of the galaxy's dynamical history, and earlier claims of a much steeper gradient would be wrong. The same dataset tightens the dark matter inner slope ($\\gamma=1.0^{+0.5}_{-0.6}$, weakly cusped) and constrains chemical evolution to rapid, inefficient star formation with outflows that eject roughly 200 solar masses of gas per solar mass of stars.","feed_headline":"Boötes I keeps a 4σ velocity gradient after perspective correction","feed_subtitle":"New 148-star sample pins an orbit-aligned 1.2 km/s gradient and a cusp-leaning dark halo.","key_machinery":"The load-bearing object is a coordinate-and-velocity model: the galaxy is treated as an extended rigid body moving at the systemic proper motion, and the line-of-sight velocity field expected from that motion alone — perspective rotation — is computed from the adopted distance, proper motion, and on-sky position of each star. Subtracting that predicted field leaves the 'intrinsic' gradient. Around this sit three other mechanisms: a Gaussian mixture model that assigns membership from velocity, metallicity, and proper motion; an axisymmetric Jeans model with a generalized Hernquist dark halo (inner slope $\\gamma$, outer slope $\\beta$, flattening $Q$, velocity anisotropy $\\beta_z$) fitted to the unbinned line-of-sight velocity dispersion; and the one-zone galactic chemical evolution model of Weinberg et al. (2017) fitted to the metallicity distribution function, with parameters for star-formation timescale, truncation time, star-formation efficiency, and mass-loading factor.","core_discovery":"The paper's headline discovery is kinematical. Modeling Boötes I as a solid body moving with its measured Gaia proper motion and distance predicts a perspective-rotation velocity gradient of $0.105\\pm0.004$ km s$^{-1}$ arcmin$^{-1}$ across the observed field; the paper subtracts this and finds a residual intrinsic gradient of $0.12^{+0.04}_{-0.03}$ km s$^{-1}$ arcmin$^{-1}$ ($1.2^{+0.4}_{-0.3}$ km s$^{-1} r_h^{-1}$), at about 4$\\sigma$ significance, oriented along Boötes I's orbit. This is much shallower and differently oriented than the $0.40\\pm0.10$ km s$^{-1}$ arcmin$^{-1}$ gradient reported previously, which the paper argues is not reproducible from the data. The paper also reports a resolved radial metallicity gradient of $-0.010\\pm0.003$ dex arcmin$^{-1}$ ($-0.10\\pm0.03$ dex $r_h^{-1}$), a dark matter inner slope $\\gamma=1.0^{+0.5}_{-0.6}$ that weakly favors a cusp while not ruling out a core, and chemical evolution parameters — star-formation timescale $\\tau_{\\rm SFH}=0.2\\pm0.1$ Gyr, star-formation efficiency $\\mathrm{SFE}=0.07^{+0.04}_{-0.03}$ Gyr$^{-1}$, mass-loading factor $\\eta=203^{+27}_{-36}$ — describing a short, inefficient burst of star formation with strong outflows.","pith_inferences":["If the same rigid-body perspective correction were applied to other spatially extended ultra-faint dwarfs, some previously reported 'tidal' velocity gradients might shrink or reorient; the paper's own bootstrap shows that without correction a spurious gradient is recovered 99% of the time, so this is a caution for the field.","A decisive test of the tidal versus rotation interpretation is to measure [$\\alpha$/Fe] or neutron-capture abundances for stars beyond $3r_h$: chemically distinct populations would point to an accreted or major-merger origin, while smooth gradients would favor internal or tidal processes.","The inferred mass-loading factor of $\\sim200$ suggests that gas removal in ultra-faint dwarfs is dominated by feedback-driven outflows; if this holds generally, chemical evolution models of other ultra-faint dwarfs with fixed low $\\eta$ may systematically underestimate outflow strength."],"forward_implications":["The previously reported steep ($0.40\\pm0.10$ km s$^{-1}$ arcmin$^{-1}$) gradient in Boötes I is not reproduced; the intrinsic gradient is about 1.2 km s$^{-1} r_h^{-1}$, so dynamical interpretations built on the steep value need revision.","Because the corrected gradient is aligned with the orbit but the estimated tidal radius ($\\sim8$–$13r_h$) lies beyond the sampled stars, tides alone may not explain it; the origin is left open between tidal deformation, rotation, and a past merger.","With 115 clean velocities, the inferred velocity dispersion drops to $4.0^{+0.4}_{-0.3}$ km s$^{-1}$, and the dark matter inner slope $\\gamma=1.0^{+0.5}_{-0.6}$ weakly favors the cuspy halos predicted by cold dark matter, though cores remain allowed.","The large, homogeneous 92-star metallicity sample and the resolved metallicity gradient place Boötes I on the Local Group mass–metallicity relation and support self-enrichment as the source of its metallicity spread, without requiring a merger.","The chemical evolution fits imply $\\eta\\sim200$: for every solar mass of stars formed, about 200 solar masses of gas are ejected, so strong feedback, not necessarily reionization truncation, shaped Boötes I's chemical evolution."],"supporting_citations":[{"why":"Supplies the previous steep velocity gradient claim whose catalog this paper reanalyzes and contradicts.","marker":"L22"},{"why":"Provides the archival VLT sample of velocities, metallicities, and binary candidates that forms the core of the combined dataset.","marker":"J21"},{"why":"Provides the MMT Hectochelle sample extending to about 3 half-light radii with velocities and metallicities.","marker":"W23"},{"why":"Supplies the original VLT FLAMES data and binary identifications that J21 and this paper build upon.","marker":"K11"},{"why":"Supplies the axisymmetric Jeans modeling machinery and the previous VLT-only dark matter constraints that this paper extends.","marker":"Hayashi et al. (2023)"},{"why":"Supplies the one-zone galactic chemical evolution model used to fit the metallicity distribution function.","marker":"WAF17"},{"why":"Supplies the adopted center, half-light radius, ellipticity, and position angle used in the geometric and kinematic modeling.","marker":"Muñoz et al. (2018)"},{"why":"Supplies membership probabilities, proper motions, and orbit parameters used for target selection and interpretation.","marker":"Pace et al. (2022)"},{"why":"Supplies the S5 survey design, target selection, and spectral reduction and calibration used for the new observations.","marker":"Li et al. (2019)"},{"why":"Supplies the solid-body perspective rotation formalism that the central velocity-gradient subtraction relies on.","marker":"Kaplinghat & Strigari 2008"}],"fun_headline_variants":["Boötes I: 4σ orbit-aligned gradient after perspective fix","Boötes I's true gradient: 4σ, orbit-aligned, cusp halo","Boötes I: perspective-corrected 4σ gradient, cusp dark matter","Boötes I: 4σ gradient after perspective removal, cusp-leaning halo","Boötes I: revised 4σ velocity gradient, cusp dark halo"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire 'intrinsic' gradient rests on treating Boötes I as a rigid solid body with the adopted Gaia proper motion and RR Lyrae distance, and on assuming that membership cuts, binary masking, and sample selection cannot produce a velocity pattern that mimics the residual gradient.","fun_headline_variants_meta":{"raw":{"variants":["Boötes I: 4σ orbit-aligned gradient after perspective fix","Boötes I's true gradient: 4σ, orbit-aligned, cusp halo","Boötes I: perspective-corrected 4σ gradient, cusp dark matter","Boötes I: 4σ gradient after perspective removal, cusp-leaning halo","Boötes I: revised 4σ velocity gradient, cusp dark halo"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00136,"raw_usage":{"total_tokens":5655,"prompt_tokens":1218,"completion_tokens":4437,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":834,"completion_tokens_details":{"reasoning_tokens":4325}},"tokens_in":834,"tokens_out":4437,"duration_ms":28782,"temperature":1.0,"reasoning_tokens":4325,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:37:11.450436+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-fit the same 148-star catalog with an independent proper motion measurement for Boötes I and with membership probabilities left free; if the claimed 1.2 km s$^{-1} r_h^{-1}$ residual gradient does not survive in magnitude and direction, it was an artifact of the rigid-body perspective correction. A companion test is to search for extra-tidal members beyond roughly 8–13 half-light radii, the tidal radius estimated in the paper, whose presence or absence would distinguish tidal stripping from internal rotation.","supporting_citations":[],"review_version":2}