{"id":"cd0115db-ca6e-4b97-8466-9370b5f2d2f8","arxiv_id":"2507.10544","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Cosmological surface brightness dimming and resolution loss make cosmic noon rotation curves look artificially symmetric and compact, suggesting that declining rotation curves at high redshift may be an observational artifact.","lead":"This paper artificially moves 19 local, disturbed galaxies to redshift 2.2 to see what their rotation curves would look like at cosmic noon. It finds the distant versions look smaller, smoother, and more symmetric than they really are, which warns against trusting cosmic noon rotation curves for dark matter estimates.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Asymmetry reduction may be an artifact of comparing curves of different radial extent; the paper does not match radii when comparing low-z and mocked cosmic noon rotation curves.","rationale":"The reader's weakest_assumption identifies mock fidelity (Gonçalves et al. 2010) as the key risk. I agree that mock fidelity matters, but the more load-bearing and internally addressable concern is that the asymmetry comparison is confounded by different radial coverage. The paper explicitly shows asymmetry grows with radius in low-z galaxies, and explicitly states the mocked OSIRIS curves are 1.5 times shorter. A metric like Δχ²_red or mean vasym will therefore be lower in the mocks even if the intrinsic velocity field at matched radii is perfectly reproduced. This is not an external calibration issue; it threatens the interpretation of the paper's own Figures 3 and 10. The conclusion may still survive a matched-radius test, because SINFONI's beam smearing could genuinely erase small-scale kinematic structure, and the combined effect of truncation plus smoothing still makes the total observable information unreliable for dynamical modeling. But the quantitative claim that cosmic noon galaxies appear 'more regular' is not yet established as distinct from 'less extensively sampled.' A matched-radius reanalysis would settle this, so the conditional verdict is appropriate rather than a rejection. This is a precise, addressable weakness, and the paper's overall cautionary message is valuable if confirmed.","tokens_in":28240,"tokens_out":3705,"duration_ms":48788,"concrete_test":"For each galaxy, truncate the original low-z rotation curve and velocity map at the maximum radius reachable in the corresponding mock observation (OSIRIS and SINFONI separately, using the same physical or R/Reff cutoff). Recompute Δχ²_red and vasym on these truncated low-z data and compare pairwise with the mock values. If the truncated low-z asymmetries become statistically indistinguishable from the mocks, the reported smoothing is fully explained by radial truncation; if the mocks remain significantly less asymmetric at matched radii, beam smearing and dimming genuinely reduce apparent asymmetry. Report paired differences with uncertainties, not just group means.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central evidence that cosmic noon rotation curves appear more regular is the reduction in asymmetry metrics: mean Δχ²_red drops from 158.05 at low-z to 2.81 (mocked OSIRIS) and 33.04 (mocked SINFONI), and vasym is lower in both mocks (Figures 3 and 10). However, the mocked OSIRIS curves are, by the authors' own statement, on average 1.5 times shorter (Figure 8), and the paper shows that low-z asymmetry increases with radius (Figure 6, Spearman 84%). Any asymmetry measure that averages over the sampled radial range will therefore be lower in the mocks simply because the high-asymmetry outer regions are missing or obscured. Δχ²_red is also not directly comparable across different radial baselines and point counts: a quadratic fit over a shorter, inner segment generally yields smaller residuals even with identical intrinsic kinematics. The paper does not present a matched-radius comparison, such as truncating the low-z curves at the maximum radius of each mock or evaluating asymmetry in common radial bins. Without this, the claim that the mocks are intrinsically smoother is not cleanly separated from the trivial statement that we see less of the galaxy at cosmic noon.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"Using 19 Lyman Break Analogs at z ~ 0.2 observed with OSIRIS, the authors construct artificial z ~ 2.2 observations mimicking OSIRIS with adaptive optics and SINFONI without adaptive optics, then extract rotation curves and quantify asymmetry via a reduced chi-squared metric (following Übler et al. 2021) and a kinemetry-based vasym metric (Shapiro et al. 2008). They find that the mocked cosmic noon curves are shorter, smoother, and less asymmetric than the original low-z curves, that asymmetry in low-z LBAs increases with radius, and that the second velocity moment Vrms is not conserved under the mock transformations. They conclude that cosmic noon rotation curves may be biased by cosmological dimming and resolution loss and may not reliably trace the gravitational potential.","tokens_in":28495,"tokens_out":4855,"duration_ms":53489,"significance":"If the radial-matching concern is resolved, the paper provides a useful cautionary demonstration using a well-motivated local analog population. Its strengths include using two independent asymmetry diagnostics, connecting the measured asymmetry values to the Übler et al. (2021) sample, and making velocity maps and rotation curves available. The analysis is not circular: it does not fit parameters to produce its conclusion, and it makes falsifiable predictions about how asymmetry metrics should change under realistic redshift degradation. The main limitations are that the mock observations are inherited from Gonçalves et al. (2010) without independent validation in this work, and that the current asymmetry comparison mixes radial truncation with intrinsic smoothing effects, which is the critical issue for the paper's central claim.","major_comments":[{"comment":"The comparison of Δχ²_red between low-z and mocked curves is not matched in radial extent. Mock OSIRIS curves are on average 1.5 times shorter than the low-z curves (Fig. 8), and Fig. 6 shows that low-z asymmetry increases with radius (Spearman ρ ≈ 0.84). Since Δχ²_red is computed over the sampled radial range and depends on the number of data points and the baseline, the lower mock values may simply reflect the fact that the high-asymmetry outer regions are not sampled. The paper needs a matched-radius comparison, such as truncating the low-z curves at the maximum radius of each mock or evaluating Δχ²_red in common radial bins, to separate the claim that the mocks are intrinsically smoother from the trivial statement that we see less of the galaxy at cosmic noon.","section":"§3.2, Figs. 3, 6, 8"},{"comment":"The mocked SINFONI observations do not include cosmological surface brightness dimming; the text states that the dimming is expected to be offset by the instrument's higher sensitivity, so \"no extra corrections for cosmological dimming were applied.\" This means the SINFONI mock isolates resolution and beam-smearing effects only, whereas the OSIRIS mock includes dimming. Consequently, the comparison between the two mocks (e.g., Figs. 10 and 11) conflates instrument sensitivity with the inclusion of dimming, and the summary statement about the \"combined effects of cosmological dimming and low resolution\" is not directly supported for the SINFONI branch. The authors should either include dimming in the SINFONI mock with an appropriate sensitivity model or explicitly frame the SINFONI result as testing only resolution effects.","section":"§2.4"},{"comment":"The reported Δχ²_red mean values (158.05, 2.81, 33.04) are quoted without per-galaxy uncertainties, bootstrap intervals, or a demonstration that the means are not driven by one or two outliers. Given the small sample size (15 low-z galaxies, fewer for mocked OSIRIS after the exclusions listed in §3.1), a paired or bootstrap analysis on the matched subsample is needed to assess the robustness of the central asymmetry reduction, especially because the radial-extent mismatch flagged above may affect individual galaxies to different degrees.","section":"§3.2, Eq. (3)"}],"minor_comments":[{"comment":"The sentence \"mergers and interactions causes more disturbance in galaxy's gravitational field\" should be \"mergers and interactions cause more disturbance in the galaxy's gravitational field.\"","section":"Abstract"},{"comment":"The text says \"we find the χ2_red from eq. 2\" but the reduced chi-squared is defined in Eq. (3); the cross-reference should be corrected.","section":"§3.2"},{"comment":"K is described only as \"the degrees of freedom\"; it would be clearer to define K explicitly (e.g., N minus the number of fitted parameters) so that the reduced nature of the statistic is unambiguous.","section":"Eq. (3)"},{"comment":"Footnotes a and b both point to the same SDSS/DR7 URL; if the stellar mass and SFR catalogs are identical, the two footnotes should be collapsed into one.","section":"Table 1"},{"comment":"The citation \"Tiley 2020\" in the introduction refers to a Nature news item listed in the references as \"Galaxy disk observed to have formed shortly after the Big Bang\"; this is not an appropriate citation for the physical claim about disk evolution through mergers and accretion, and it should be replaced with the relevant primary literature.","section":"§1, References"},{"comment":"In the discussion, \"LBAS\" should be \"LBAs,\" and in Figure 3 the caption \"The blue dots refers\" should be \"The blue dots refer.\"","section":"§4, Figure captions"}],"recommendation":"major_revision","confidential_remarks":"The paper's scope is appropriate for an astro-ph.GA journal, and I see no problematic citation pattern beyond the minor Tiley issue. The main risk is that the core asymmetry comparison currently conflates radial truncation with smoothing; however, this is fixable with a matched-radius analysis, so a major revision is appropriate rather than a rejection. The reliance on Gonçalves et al. (2010) mocks should also be stated more explicitly as an assumption in the revised version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know about arXiv:2507.10544. First, it is a new quantitative test of a familiar worry: that cosmic noon rotation curves look cleaner than they really are. Second, the test is confounded by radial truncation, so the headline numbers are not as clean as the abstract suggests.\n\nThe genuinely new content is taking 19 LBAs, artificially redshifting them to z~2.2 with two mock IFU setups (OSIRIS with AO, SINFONI natural seeing), and measuring rotation curve asymmetry with two independent metrics (Übler et al.'s Δχ²_red and Shapiro et al.'s vasym). Both metrics decrease in the mocks. Using real galaxies rather than only simulations is a step forward, and the two mocks usefully separate cosmological dimming (OSIRIS) from resolution blurring (SINFONI). The paper is honest about the limitations of 2D kinematic fitting and engages with the contradictory literature, including Lelli et al. (2023) and Tiley et al. (2019). That is solid work to build on.\n\nThe main soft spot is exactly what the stress-test note flags: the comparison is not matched in radius. The authors themselves show asymmetry increases outward (Spearman 84%) and that mock OSIRIS curves are on average 1.5 times shorter. Then they compare whole-curve asymmetry values. A quadratic fit over a shorter inner segment gives lower residuals even with identical underlying kinematics. The drop from Δχ²_red ~158 to ~2.8 is dramatic, but until the low-z curves are truncated to the same radial range as the mock, you cannot separate \"the galaxy genuinely appears smoother\" from \"we only see the ordered inner part.\" The SINFONI mocks extend further, so radial coverage is less of an issue there, but resolution smoothing dominates; a radial-bin comparison would help. There is also no significance testing: the paper reports mean values without showing the scatter or a test that the distributions differ. That is a minor issue if the effect is as large as shown, but it is missing.\n\nTwo smaller concerns. First, the two mocks are not directly comparable because only OSIRIS includes cosmological dimming; the authors state this, but it weakens the \"combined effects\" language. Second, the mock observations are inherited from Gonçalves et al. (2010) without independent validation here; that is not fatal, but the evidence chain depends on a prior paper's simulation assumptions.\n\nOverall: the qualitative conclusion is plausible and worth publishing, but the paper overclaims when it says rotation curves \"might not be reliable enough\" for dynamical modeling and dark matter estimates. It demonstrates a bias exists; it does not quantify the impact on derived dark matter fractions. That is a caution, not a verdict.\n\nWho should read it: anyone interpreting IFU rotation curves at z~1-3, and anyone citing Genzel or Lang results. It deserves a serious referee; the natural requests would be a matched-radius control and error estimates on the asymmetry means.","headline":"A credible cautionary result that cosmic noon rotation curves look smoother than they are, weakened by a missing matched-radius control that muddies the quantitative claim.","tokens_in":29008,"tokens_out":3397,"would_cite":true,"duration_ms":42981,"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":"The paper argues that cosmological surface brightness dimming and loss of resolution make cosmic-noon rotation curves look smaller, smoother, and more symmetric than the true kinematics, so these curves may not be reliable for dynamical…","keywords":["rotation curves","cosmic noon","Lyman Break Analogs","cosmological surface brightness dimming","integral field spectroscopy","beam smearing","galaxy kinematics","dark matter"],"falsifier":"A decisive test would be to obtain high-resolution, deep IFU data of the same galaxy, or a statistical twin, at z≈2.2 that resolves the outer disk and reaches the faint surface brightness levels of the low-redshift data; if the rotation curve then remains smooth and declining and the asymmetry stays as low as in the mocks, the paper's attribution of those features to bias would be overturned, whereas recovery of the asymmetric, extended rotation curve would confirm it.","tokens_in":28117,"feed_emoji":"🔭","tokens_out":6075,"duration_ms":71616,"temperature":0.7,"pith_summary":"The paper asks whether the declining rotation curves reported for star-forming galaxies at cosmic noon are genuine or an observational mirage. It takes 19 local Lyman Break Analogs, which resemble the clumpy, interacting galaxies seen at z~2, and artificially pushes them to z≈2.2 through mock integral-field observations that reproduce surface brightness dimming, adaptive-optics point-spread functions, and sensitivity. Comparing the real low-redshift kinematics with the mocked cosmic-noon kinematics, it finds the mocked galaxies look smaller, smoother, and substantially more symmetric than their true progenitors, and their second velocity moment no longer traces the same gravitational potential. The conclusion is that cosmic-noon rotation curves may appear well behaved simply because of observational bias, so they should be used with caution when estimating dark matter.","feed_headline":"Cosmic-noon rotation curves look smooth due to bias","feed_subtitle":"Surface brightness dimming and resolution loss make distant galaxies appear simpler, threatening dark matter estimates.","key_machinery":"The carrying object is a mock-observation pipeline that relocates actual IFU data of 19 Lyman Break Analogs from $z\\approx0.2$ to $z\\approx2.2$. It rebins flux maps to the angular diameter distance at the new redshift, applies the $(1+z)^{-4}$ cosmological dimming, convolves with Gaussian cores and halos representing adaptive-optics and seeing profiles, adds sky background and noise, and matches OSIRIS and SINFONI spaxel scales and integration times. Two asymmetry diagnostics then quantify the effect: a quadratic-polynomial fit to each side of the rotation curve that yields $\\Delta\\chi^2_{\\mathrm{red}}$ following the method used in the cosmic-noon studies being tested, and a Fourier harmonic expansion of the velocity field via Kinemetry, whose higher-order terms normalized by the dominant coefficient $B_1$ give $v_{\\mathrm{asym}}$. A final check compares $V_{\\mathrm{rms}}=\\sqrt{V_{\\mathrm{rot}}^2+\\sigma^2}$ between the real and mocked data as a proxy for the gravitational potential traced by the kinematics.","core_discovery":"On its own terms, the paper establishes that the combined effects of cosmological surface brightness dimming (surface brightness scaling as $(1+z)^{-4}$) and loss of spatial resolution imprint a systematic bias on rotation curves at cosmic noon. For the same galaxies, the asymmetry quantified by the quadratic-fit $\\Delta\\chi^2_{\\mathrm{red}}$ drops from a mean of 158.05 at $z\\approx0.2$ to 2.81 in the AO-assisted mocked observations and 33.04 in the seeing-limited mocked observations; the Kinemetry-based asymmetry $v_{\\mathrm{asym}}$ is also smaller at $z\\approx2.2$. The AO-mocked curves reach only about 1.5 times shorter radii than the real ones, and the estimated inclinations appear more face-on. The second velocity moment $V_{\\mathrm{rms}}=\\sqrt{V_{\\mathrm{rot}}^2+\\sigma^2}$ of the mocked galaxies does not follow a 1:1 relation with the low-redshift data, indicating that the observed kinematics trace a biased version of the gravitational potential. The paper therefore concludes that rotation curves of distant galaxies might not be reliable enough for dynamical modeling and estimating dark matter properties.","pith_inferences":["Editorial extension: the same mock-observation apparatus could be applied to other high-redshift kinematic tracers, such as CO or [CII] emission, to see whether the bias is tracer-dependent or universal.","Editorial extension: if the bias is as strong as reported, part of the scatter in derived V/σ and dark matter fractions across cosmic-noon surveys may reflect differences in resolution and depth rather than differences in galaxy physics.","Editorial extension: a quantitative prediction implied by the paper is that recovering the true outer rotation curve of a z≈2 galaxy requires reaching surface brightnesses below what current AO-assisted IFUs achieve; this can be tested with planned deeper observations."],"forward_implications":["Reported declining rotation curves at cosmic noon cannot, by themselves, be read as evidence for a small dark matter fraction in distant halos.","Galaxies that are actually interacting or merging will often be classified as smooth, symmetric, rotation-dominated systems when viewed at z≈2, biasing samples toward cold-disk interpretations.","Dynamical modeling of cosmic-noon galaxies needs to incorporate the bias, for example through forward-modelling of realistic mock observations, rather than correcting only for beam smearing.","Because asymmetry grows toward galaxy outskirts, stacked rotation curves from co-added faint data may not recover the true outer profile, as the faint asymmetric signal is partly random.","Future high-resolution IFU observations with ELT-class instruments are needed to separate the intrinsic shape of the cosmic-noon rotation curve from the observational bias."],"supporting_citations":[{"why":"Supplies the 19 Lyman Break Analog observations and the artificial redshift to z≈2.2 that form the core comparison.","marker":"Gonçalves et al. (2010)"},{"why":"Provides the IFU observation simulation prescription that the mock observations replicate.","marker":"Law et al. (2006)"},{"why":"Defines the LBA sample and shows merger signatures vanish at z~2, motivating the use of LBAs as cosmic-noon analogs.","marker":"Overzier et al. (2010)"},{"why":"Establishes that LBAs resemble cosmic-noon star-forming galaxies in key properties, justifying the sample choice.","marker":"Basu-Zych et al. (2009)"},{"why":"Supplies the quadratic-fit Δχ²_red asymmetry metric and the comparison values for simulated and observed cosmic-noon galaxies.","marker":"Übler et al. (2021)"},{"why":"Supplies the Kinemetry-based vasym asymmetry method used as the second diagnostic.","marker":"Shapiro et al. (2008)"},{"why":"Supplies Kinemetry and PaFit, used to extract kinematic axes, centers, and Fourier harmonic terms.","marker":"Krajnović et al. (2006)"},{"why":"Reports declining rotation curves and small dark matter fractions at cosmic noon, the high-redshift claim the paper argues is biased.","marker":"Genzel et al. (2017)"},{"why":"Reports averaged declining rotation curves at cosmic noon; the paper argues these observations suffer from the same bias.","marker":"Lang et al. (2017)"}],"fun_headline_variants":["Cosmic-noon rotation curves: bias masquerades as simplicity","Dimming and resolution hide galaxy complexity at cosmic noon","Rotation curves at cosmic noon are misleadingly smooth","Bias at cosmic noon: rotation curves don't trace dark matter","Redshift dimming skews rotation curves, warps dark matter clues"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the mock observations faithfully reproduce how real OSIRIS and SINFONI data at z≈2.2 would look, including surface brightness dimming, point spread, and sensitivity, so the reduced asymmetry in the mocked data is a real observational bias and not an artifact of the simulation.","fun_headline_variants_meta":{"raw":{"variants":["Cosmic-noon rotation curves: bias masquerades as simplicity","Dimming and resolution hide galaxy complexity at cosmic noon","Rotation curves at cosmic noon are misleadingly smooth","Bias at cosmic noon: rotation curves don't trace dark matter","Redshift dimming skews rotation curves, warps dark matter clues"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000202,"raw_usage":{"total_tokens":1430,"prompt_tokens":1044,"completion_tokens":386,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":660,"completion_tokens_details":{"reasoning_tokens":301}},"tokens_in":660,"tokens_out":386,"duration_ms":5164,"temperature":1.0,"reasoning_tokens":301,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:28:16.159434+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test would be to obtain high-resolution, deep IFU data of the same galaxy, or a statistical twin, at z≈2.2 that resolves the outer disk and reaches the faint surface brightness levels of the low-redshift data; if the rotation curve then remains smooth and declining and the asymmetry stays as low as in the mocks, the paper's attribution of those features to bias would be overturned, whereas recovery of the asymmetric, extended rotation curve would confirm it.","supporting_citations":[{"cited_title":"R., Steidel, C","cited_arxiv_id":null,"evidence_quote":"Provides the IFU observation simulation prescription that the mock observations replicate."},{"cited_title":"A., Heckman, T., Schiminovich, D., et al","cited_arxiv_id":null,"evidence_quote":"Defines the LBA sample and shows merger signatures vanish at z~2, motivating the use of LBAs as cosmic-noon analogs."}],"review_version":1}