{"id":"081818af-d063-4c24-9907-53bbc1999647","arxiv_id":"2504.12403","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Strong S-shaped stellar disk warps occur in about 50% of edge-on disk galaxies at z≈2 but only 10-15% at z≈0, implying strong redshift evolution.","lead":"This study measures how often distant disk galaxies are visibly warped, using about 1,000 edge-on galaxies from HST and JWST images. It reports that strong S-shaped warps were about four times more common at early cosmic times than today, a pattern linked to galaxy interactions and mergers.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The redshift trend may be a measurement-selection artifact: ψe is measured at the outermost detectable radius, which shrinks and gets noisier with z, and no full completeness/false-positive calibration is supplied.","rationale":"The central claim is plausible and the paper is unusually candid about its limitations. The sample is far larger than earlier work, the skeleton method is described in detail, and the finding of a monotonic rise in strong S-shaped warp fraction is a potentially important result. The paper's own text, however, contains the ingredients of the concern: ψe is measured at the outermost detectable point (Sec. 2), high-z disks are traced less far and are noisier (Sec. 3.2), and the JWST sample is incomplete and visually selected (Sec. 2). The artificial-redshift check on 'a few bright galaxies' and the removal of objects with extent ≤3h address only a narrow slice of the selection function; neither provides a quantitative completeness or false-positive rate. Since the central astrophysical conclusion—that warp frequency tracks the merger/interaction rate—depends on the measured fraction rising with z, a z-dependent measurement bias of the kind the authors themselves describe would be sufficient to explain the trend. The reported correlation p-values treat all measurements as equally reliable and do not fold in this systematic. A synthetic-injection completeness test would settle the matter. I therefore do not move the reader's verdict: CONDITIONAL remains appropriate pending such a test and, ideally, a published catalog including per-object radial extent and measured profiles. The minor reference mismatch (in-text Reshetnikov 1995 vs. reference-list Reshetnikov 1998) is not load-bearing.","tokens_in":7821,"tokens_out":4695,"duration_ms":54777,"concrete_test":"Inject synthetic edge-on disk models with known warp amplitudes (0°, 2°, 4°, 8°) and known S/U geometry into real HST/JWST noise maps at z ≈ 0.3, 0.8, 1.5, and 2.0, matching the sample's stellar masses, scale lengths, PSF, and surface-brightness dimming. Run the same skeleton pipeline and visual-selection protocol blindly, and build a completeness/false-positive matrix for ψe > 4° per redshift. Recompute the strong-warp fraction corrected by this matrix. If the corrected fraction at z ≈ 2 is no longer ~50% or the trend is non-monotonic, the claimed evolution is an artifact.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central observable is ψe, defined in Sec. 2 as measured at 'the location of the most distant ΔH(r) data point from the center that the method could capture.' The authors themselves note in Sec. 3.2 that high-redshift disks 'can typically be traced only to a greater surface brightness' and that skeleton measurements 'sometimes become unreliable if the extent of a galaxy's skeleton is too small,' leading to 'the appearance of a \"false\" warp due to isophotes becoming too influenced by noise.' Because the radial baseline on which ψe is measured shrinks and becomes noisier with redshift, the measured angle is a redshift-dependent mixture of true warp amplitude and noise-induced slope error; the reported Pearson p-values are computed on the raw measurements and do not incorporate this systematic. The completeness problem is compounded by the acknowledged incompleteness and visual selection of the JWST sample (Sec. 2: 'the second sample is inherently incomplete and does not encompass all edge-on galaxies detectable in the JWST observations'). The two internal checks—artificial redshifting of 'a few bright galaxies' and removing objects with extent ≤3h—are too limited to quantify a detection-efficiency or false-positive rate as a function of z. Therefore the claimed 10–15% to ~50% rise in strong S-shaped warp fraction cannot yet be distinguished from a z-dependent measurement bias.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper measures warp properties of roughly 1,000 edge-on galaxies selected from HST/COSMOS and JWST/DJA fields, using skeletonization of isophotes to derive a warp angle psi_e at the outermost detectable radius. It reports that the fraction of galaxies with strong S-shaped warps (psi_e > 4 degrees) increases from about 10-15% at z=0 to about 50% at z=2 (Fig. 4), and that the mean psi_e increases with redshift (Pearson r=0.19, p about 1e-5; Fig. 5). The authors interpret this as evidence that interactions and mergers produce more warps at earlier epochs. They include several checks: a stellar mass cut, exclusion of galaxies with small extent, and an artificial-redshift test on a few bright galaxies.","tokens_in":7941,"tokens_out":6604,"duration_ms":66318,"significance":"If the trend is real, it is a valuable falsifiable observable: it ties the frequency of strong stellar warps to the rising interaction and merger rate up to cosmic noon and can help constrain warp excitation mechanisms. The result is presented as a raw observed fraction rather than a model-dependent fit, and no fitted free parameters enter the redshift evolution. The paper is also careful to separate S-shaped and U-shaped warps and to check the mass-redshift selection. The principal weakness is that the central observable psi_e is measured at a redshift-dependent outermost radius in samples that are incomplete and visually selected, so the trend could in part be a measurement artifact. The reported correlation and fractions are not yet supported by a full completeness or false-positive calibration, despite the authors' useful but limited internal checks.","major_comments":[{"comment":"The central observable is not measured at a fixed physical radius; it is measured at the outermost detectable Delta H(r) point, and the text acknowledges that psi_e 'inherently depends on the data's spatial extent.' Because the detectable spatial extent shrinks and the noise increases with redshift, as the authors themselves note in Sec. 3.2, the raw psi_e is a redshift-dependent mixture of true warp amplitude and noise-induced slope error. The Pearson correlation in Fig. 5 and the binned fractions in Fig. 4 are therefore not yet established as intrinsic trends. A quantitative forward-model or injection-recovery calibration as a function of redshift, and ideally a re-measurement at a fixed rest-frame radius such as 4h, is needed before the claimed evolution can be accepted.","section":"Section 2, definition of psi_e; Figs. 4-5"},{"comment":"The artificial-redshift check uses 'a few bright galaxies' and only concludes that measurements 'sometimes become unreliable' when the skeleton extent is too small. This does not provide a detection-efficiency or false-positive rate as a function of redshift. Removing galaxies with extent at or below 3h is a useful robustness check, but for the remaining galaxies there is no demonstration that the probability of measuring psi_e > 4 degrees is redshift-independent. The sentence 'the measured warp angle at the outermost point ... is still higher for these high-redshift galaxies' is an observed raw comparison, not a bias correction.","section":"Section 3.2, artificial-redshift test"},{"comment":"The paper states that the JWST sample 'is inherently incomplete and does not encompass all edge-on galaxies detectable,' and the final selection involved visual inspection by at least three authors. Without a completeness function, or at least a selection function in terms of apparent size, surface brightness, and warp amplitude, the 50% fraction at z approximately 2 in Fig. 4 cannot be distinguished from a selection bias toward visually conspicuous bent disks. The mass cut M* > 10^9 solar masses addresses the mass-redshift degeneracy shown in Fig. 1 but does not address this morphological selection.","section":"Section 2, JWST sample selection; Fig. 4"}],"minor_comments":[{"comment":"The text cites Reshetnikov (1995), but the reference list contains Reshetnikov (1998); please correct this inconsistency.","section":"Section 1 and References"},{"comment":"The vertical error bars are described as 95% confidence intervals, but the method used to compute them (e.g., Wilson or Clopper-Pearson binomial intervals) is not stated; please specify.","section":"Figure 4"},{"comment":"The statement that no significant correlation between b/a and redshift is found is not accompanied by a statistic or a figure; please provide the correlation coefficient and p-value or remove the claim.","section":"Section 3.1"},{"comment":"The p-values for Fig. 5 are given for the S-shaped and U-shaped subsamples but not for the entire sample, although the text quotes the correlation for the entire sample first; please make the sample definitions consistent.","section":"Section 3.2"},{"comment":"The inference that 'nearly all spiral galaxies at z approximately 2 may be affected by substantial warping' relies on the conversion of observed to true warp fractions from Sanchez-Saavedra et al. (1990), but the conversion is not described; please add a brief explanation or soften the statement.","section":"Section 3.1"},{"comment":"The abstract describes the evolution as 'potential,' while the conclusions state it more firmly as a result; please harmonize the wording.","section":"Abstract and Section 4"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a timely question and assembles the largest sample of distant edge-on galaxies used for warp statistics to date. However, the central trend is threatened by a selection and measurement bias that the current tests do not quantify. I recommend major revision with additional simulations, such as injection-recovery or full forward modeling of the skeleton measurement, and either a completeness function or a redefinition of psi_e at a fixed physical scale. The Letter format may be too short for these additions; a short paper with an appendix or supplementary material would be more appropriate."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is the biggest dataset of distant edge-on disk galaxies assembled so far, and it shows a clean monotonic rise in the fraction of strong S-shaped warps, from about 10–15% locally to roughly 50% at z≈2. That trend is plausible and worth taking seriously, but it is not yet nailed down.\n\nWhat the paper does well: the sample is genuinely new. Roughly 1,000 objects spanning z≈0–2.5, combining HST/COSMOS and JWST data, is a real step beyond the 45 galaxies in Reshetnikov et al. (2002) and the recent small JWST samples. The measurement pipeline is clearly described, and the authors are honest about the limits of their data. They report a check on wavelength dependence, a stellar-mass cut to reduce selection bias, exclusion of objects with tiny radial extent, and a few artificial-redshift experiments. They also compare their local fractions with published values and get sensible agreement. The interpretation that the trend tracks the interaction and merger rate is reasonable and consistent with independent merger-rate evolution studies.\n\nThe soft spot is exactly where the stress-test puts it: completeness. The primary warp angle ψe is measured at the outermost detectable isophote, which shrinks and gets noisier with redshift. The authors themselves note that high-z disks are traced only to higher surface brightness and that small skeletons can produce false warps. The internal checks are too limited to quantify detection efficiency or false-positive rate as a function of z. A few bright artificial redshift tests and a 3h cut do not substitute for a full completeness simulation. The JWST sample is explicitly incomplete and visually selected. So the 50% fraction could be inflated by a selection or noise effect. That said, the trend is monotonic, the authors use cautious wording throughout, and the local anchor is well-established. I do not think the concern falsifies the result, but it is load-bearing.\n\nMinor point: the text cites Reshetnikov 1995 while the reference list has Reshetnikov 1998. Citation hygiene aside, the self-citations to prior methodology are fine.\n\nWho this is for: anyone working on disk galaxy structure, warp formation, or high-z morphology. The paper deserves a serious referee. My recommendation: send it to review with a request for a published catalog and a proper completeness/false-positive analysis. If a referee pushes for those, the result could become a solid measurement; as it stands, it is a well-posed observational claim that needs more work before being taken as final.","headline":"The largest sample yet of high-z edge-on disks shows a plausible rise in strong S-shaped warp fraction with redshift, but selection effects keep it from being a firm measurement.","tokens_in":8700,"tokens_out":1597,"would_cite":true,"duration_ms":19526,"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":"The observed fraction of strong S-shaped stellar disk warps rises from about 10–15% locally to about 50% at z≈2, a trend the authors attribute to a higher galaxy interaction and merger rate in the early universe.","keywords":["galactic warps","disk warps","edge-on galaxies","high-redshift galaxies","galaxy evolution","galaxy interactions","JWST","HST"],"falsifier":"Measure warp angles in a sample selected identically at every redshift using deep rest-frame optical images that reach a fixed number of disk scale lengths, and check whether the fraction of S-shaped warps with psi_e > 4 degrees still climbs to about 50% at z≈2; if the corrected fraction stays near the local 10–15%, the redshift trend is a selection artifact. A second check is to run the skeletonization pipeline on artificially redshifted local galaxies with known warp angles across the full redshift range and require that the recovered psi_e distribution be unbiased before trusting the trend.","tokens_in":7503,"feed_emoji":"🌌","tokens_out":6585,"duration_ms":63825,"temperature":0.7,"pith_summary":"About half of all nearby disk galaxies have warped stellar disks, but warp statistics at earlier cosmic epochs were almost unknown. This paper measures vertical bends in roughly 1,000 edge-on galaxies out to z≈2 using HST and JWST images and finds that the observed fraction of strong S-shaped warps (amplitude greater than 4 degrees) rises from about 10–15% near z≈0 to about 50% at z≈2. The mean warp angle also increases with redshift, with a weak but statistically significant correlation. The authors interpret the trend as evidence that warps are driven by gravitational interactions and mergers, which were more common in the past. If the trend is real, warp statistics become a redshift-dependent probe of how often disk galaxies were tidally perturbed across cosmic time.","feed_headline":"At cosmic noon, half of disk galaxies are strongly warped","feed_subtitle":"HST and JWST images of ~1,000 edge-on galaxies show strong S-shaped warps rising from about 10% at z≈0 to 50% at z≈2.","key_machinery":"The central machinery is isophote skeletonization: for each edge-on galaxy, multiple isophotes are traced down to the noise level, each enclosed region is reduced to a one-pixel-wide skeleton line, and the skeletons are averaged into a disk center line whose vertical offset ΔH(r) from a straight line is fit with a third-order polynomial. The warp angle psi_e is measured at the outermost reliable point of this fit, with psi_4h at four disk scale lengths as a secondary measure. S-shaped warps, where the disk bends in opposite directions on the two sides, are separated from U-shaped warps, and only S-shaped warps are used for the main statistics because U-shapes can arise from projection effects. This pipeline turns archive images into a redshift-dependent warp statistic.","core_discovery":"At redshifts from z≈0 to z≈2 the observed fraction of galaxies with strong S-shaped stellar disk warps grows from 10–15% to about 50%, where a strong warp is defined as psi_e > 4 degrees, measured at the outermost reliably traced isophote of the disk. Across the sample the warp angle weakly but significantly increases with redshift (Pearson r≈0.19, p roughly 1e-5), and the distribution of warp angles broadens from an exponential scale of 2.1 degrees at z<0.35 to 3.5 degrees at z≥0.8. The onset of warping also moves inward in units of the exponential scale length at higher redshift. The authors argue that this redshift evolution mirrors the rising galaxy merger and interaction rate and the higher gas content of high-redshift disks, so warps trace the dynamical history of disk galaxies.","pith_inferences":["If confirmed, the strong-warp fraction could be used as a cheap morphological tracer of the galaxy merger and interaction rate at z≈1–2, complementing close-pair and asymmetry statistics because warps are visible in single-band imaging without spectroscopy.","Measuring psi_e at the outermost detectable isophote could bias the redshift trend in either direction; a definitive test would require deep rest-frame optical images that trace all disks to a fixed number of scale lengths.","Cosmological simulations of disk galaxies should be able to reproduce the redshift-dependent warp fraction; if they cannot reach roughly 50% at z≈2, mechanisms beyond discrete mergers, such as gas accretion and torques from the cosmic web, may need to carry more of the burden.","A direct prediction of the interaction-driven picture is that high-redshift strong warps should correlate with tidal debris, companions, or asymmetric morphologies, which deep JWST imaging could test."],"forward_implications":["If the trend is real, nearly all spiral galaxies at z≈2 may be substantially warped once the relation between observed and true warp fractions is accounted for, not just the observed 50%.","The inward shift of warp onset with redshift means high-redshift disks are perturbed closer to their centers, consistent with stronger or more frequent tidal forcing.","The exponential warp-angle distribution grows from a scale of 2.1 degrees at z<0.35 to 3.5 degrees at z≥0.8, implying that the whole population shifts toward larger bends, not just a rare tail.","The observed low-redshift strong-warp fraction of 10–15% matches previous local estimates, placing the claimed evolution on top of a consistent z≈0 baseline.","The authors connect the trend to independently measured merger-rate evolution and to the higher gas content of distant galaxies, so warps would reflect the more interaction-rich assembly history of disks at cosmic noon."],"supporting_citations":[{"why":"Supplies the earlier small-sample result that stellar warps at z≲1 are common and larger than local ones, which this study scales up.","marker":"Reshetnikov et al. (2002)"},{"why":"Provides the HST/COSMOS edge-on galaxy sample selection and database used for the lower-redshift portion of the study.","marker":"Usachev, Reshetnikov & Savchenko (2024)"},{"why":"Describes the DAWN JWST Archive from which the high-redshift sample, redshifts, and stellar masses were drawn.","marker":"Valentino et al. 2023"},{"why":"Supplies additional high-redshift edge-on galaxies included in the sample and serves as a selection comparison.","marker":"Tsukui et al. (2024)"},{"why":"Provides the skeletonization measurement method, the warp-angle definition, and nearby-galaxy onset statistics used for comparison.","marker":"Reshetnikov et al. (2016)"},{"why":"Defines the S-shaped versus U-shaped warp classification and reports the local strong-warp fraction of about 11%.","marker":"Reshetnikov & Combes (1998)"},{"why":"Gives the local strong-warp fraction of about 15% and the environment dependence used as a low-redshift baseline.","marker":"Ann & Park (2006)"},{"why":"Provides local warp statistics and the relation between observed and true warp fractions used to argue that nearly all z≈2 spirals may be warped.","marker":"Sánchez-Saavedra et al. (1990)"},{"why":"Supplies the measured rise of galaxy merger rates with redshift that links warp frequency to interaction history.","marker":"Duncan et al. 2019"},{"why":"Describes the artificial redshifting procedure used to check whether measurement effects could bias the warp-angle trend.","marker":"Chugunov et al. (2025)"}],"fun_headline_variants":["Half of distant disk galaxies show strong warps","Galaxy warps spike at cosmic noon, drop by today","From 10% to 50%: warps in disks surge with redshift","Mergers explain why early disks are more warped"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the rise in the detected fraction of strong S-shaped warps with redshift is real and not produced by selection effects: higher-redshift galaxies have shorter and noisier skeletons, psi_e is measured at the outermost detectable isophote rather than at a fixed physical radius, and the JWST half of the sample is acknowledged to be incomplete and visually selected.","fun_headline_variants_meta":{"raw":{"variants":["Half of distant disk galaxies show strong warps","Galaxy warps spike at cosmic noon, drop by today","From 10% to 50%: warps in disks surge with redshift","Mergers explain why early disks are more warped"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000524,"raw_usage":{"total_tokens":2555,"prompt_tokens":993,"completion_tokens":1562,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":609,"completion_tokens_details":{"reasoning_tokens":1491}},"tokens_in":609,"tokens_out":1562,"duration_ms":11827,"temperature":1.0,"reasoning_tokens":1491,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T12:32:22.994329+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure warp angles in a sample selected identically at every redshift using deep rest-frame optical images that reach a fixed number of disk scale lengths, and check whether the fraction of S-shaped warps with psi_e > 4 degrees still climbs to about 50% at z≈2; if the corrected fraction stays near the local 10–15%, the redshift trend is a selection artifact. A second check is to run the skeletonization pipeline on artificially redshifted local galaxies with known warp angles across the full redshift range and require that the recovered psi_e distribution be unbiased before trusting the trend.","supporting_citations":[{"cited_title":"2002, A&A, 382, 513","cited_arxiv_id":null,"evidence_quote":"Supplies the earlier small-sample result that stellar warps at z≲1 are common and larger than local ones, which this study scales up."},{"cited_title":"2024, MNRAS, 529, L78","cited_arxiv_id":null,"evidence_quote":"Provides the HST/COSMOS edge-on galaxy sample selection and database used for the lower-redshift portion of the study."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes the DAWN JWST Archive from which the high-redshift sample, redshifts, and stellar masses were drawn."},{"cited_title":"2016, MNRAS, 461, 4233","cited_arxiv_id":null,"evidence_quote":"Provides the skeletonization measurement method, the warp-angle definition, and nearby-galaxy onset statistics used for comparison."},{"cited_title":"2006, New Astron., 11, 293","cited_arxiv_id":null,"evidence_quote":"Gives the local strong-warp fraction of about 15% and the environment dependence used as a low-redshift baseline."},{"cited_title":"2019, ApJ, 876, 110","cited_arxiv_id":null,"evidence_quote":"Supplies the measured rise of galaxy merger rates with redshift that links warp frequency to interaction history."}],"review_version":1}