{"id":"a41c57b0-3c14-4eff-8115-b060d3866bbb","arxiv_id":"2411.19426","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"TNG50 simulations show the fraction and amplitude of lopsided disk galaxies decrease from about 60% at redshift 2 to about 40% today, with interactions and gas accretion more effective at high redshift.","lead":"Using the IllustrisTNG galaxy simulation, this paper shows that disk galaxies were much more lopsided, with one side more extended, at early cosmic times (redshift 2) than today, and that galaxy interactions and gas inflow were more effective at making them so. The result helps explain why JWST sees more lopsided galaxies in the early universe and connects a puzzling observation to physical mechanisms.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Causal mechanism claim is the weakest link: interaction/accretion comparisons are not matched and cannot separate cause from effect.","rationale":"The reader's weakest-assumption concern about using z=0 disk-selection criteria and a fixed A1 threshold at all redshifts is valid and worth testing. However, I judge the more load-bearing weakness to lie in the causal mechanism analysis, because the paper's stated aim and main conclusion are about why lopsidedness is more widespread at high redshift. The selection-bias issue primarily affects whether the redshift trend is quantitatively robust, while the uncontrolled comparisons in Sec. 4.4 directly undermine the 'more effective at high-redshift' claim even if the trend is real. The paper itself acknowledges the causal ambiguity in Sec. 5.4, where it suggests lopsidedness could funnel gas inward, but does not resolve it. A matched-control and time-ordering test using the same simulations would settle whether the proposed mechanisms actually precede and drive the lopsidedness. This concern does not require rejecting the paper, because the qualitative redshift trend and the structural correlations are valuable; it does require softening the causal language and adding such an analysis before the central interpretation can be accepted. Hence the conditional verdict remains appropriate.","tokens_in":24137,"tokens_out":7940,"duration_ms":75599,"concrete_test":"Using the public TNG50 snapshots, build matched control samples: for each lopsided galaxy at each redshift, select symmetric galaxies in the same bins of stellar mass, R90, central stellar mass density, and local environment; recompute the excess neighbor fraction (Fig. 6) and net accretion-rate difference (Fig. 7). Then track the matched pairs 0.5-1 Gyr forward and backward to test whether a neighbor encounter or accretion episode precedes A1 crossing the 0.1 threshold. If the matched differences vanish, or if accretion peaks after lopsidedness develops, the claim that these mechanisms are 'significantly more effective' at high redshift should be withdrawn or substantially softened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central interpretation (Abstract; Secs. 4.4.1-4.4.2 and 6) that tidal interactions and gas accretion are significantly more effective at high redshift rests on two uncontrolled comparisons. First, Figs. 6-7 compare lopsided and symmetric galaxies without matching the internal properties that Sec. 4.3 and Fig. 5 show are strongly correlated with A1 at every redshift (lower central stellar mass density, larger R90/Rh). Because the high-redshift samples have different size and density distributions, the larger neighbor fractions and net accretion rates of lopsided galaxies could reflect these structural differences rather than a redshift-dependent efficiency of the proposed mechanisms. Second, the net accretion rate is measured over the preceding 1 Gyr (Eq. 8), and Sec. 5.4 explicitly entertains the reverse channel: lopsidedness may funnel gas inward via internal torques, so high accretion could be a consequence rather than a trigger. Without a matched control sample or time-resolved event ordering, the claim that these mechanisms are 'significantly more effective at high-redshift' is an overstatement; the data support a correlation, not a causal mechanism.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper uses the TNG50 cosmological simulation to study stellar lopsidedness (the m=1 Fourier amplitude A1) in disk-like galaxies at ten redshifts between z=0 and z=2. At each redshift the authors independently select rotation-dominated, flattened, extended galaxies (Secs. 2.2-2.5), measure A1 in the radial range Rh<r<1.4R90, and classify galaxies with A1>0.1 as lopsided (Secs. 3-4.1). They report a lopsided fraction of ~60% at 1.5<z<2, a decrease of fraction and amplitude toward z=0, no strong dependence on local environmental density, and a strong correlation of A1 with central stellar mass density and disk size at all redshifts (Secs. 4.2-4.3, Figs. 3-5). Using a neighbor-count interaction proxy and a 1 Gyr net accretion rate (Secs. 4.4.1-4.4.2, Figs. 6-7), they argue that tidal interactions and gas accretion with subsequent star formation are significantly more effective at high redshift. The final sections compare the simulation sample with the JWST sample of Le Bail et al. (2023) using a four-type classification based on central and disk sSFR (Sec. 5).","tokens_in":24348,"tokens_out":7374,"duration_ms":63710,"significance":"This is a timely and useful paper. The main quantitative results—the high lopsided fraction at 1.5<z<2 and the redshift decline—provide the first direct cosmological-simulation comparison to the JWST lopsidedness excess of Le Bail et al. (2023), and the structural correlation (low central stellar mass density, large disk size) is a robust, physically plausible finding that connects high-redshift and local studies. The paper is transparent about several limitations, including that it does not directly measure asymmetric gas accretion and that lopsidedness itself may funnel gas inward (Secs. 4.4.2 and 5.4). The strengths are the systematic Fourier analysis applied uniformly over redshift, the use of a well-established simulation with public data, and the explicit discussion of differences from observations. The main weakness is that the causal interpretation—higher efficiency of interactions and accretion at high redshift—rests on uncontrolled comparisons and a snapshot-based interaction proxy, so the abstract's strongest claim is not yet supported.","major_comments":[{"comment":"The sample selection and the lopsided threshold are both inherited from z=0 analyses and are not tested for redshift validity. The lambda_R>0.31*sqrt(epsilon), epsilon>0.4, and R90>3 kpc criteria are calibrated at z=0, and the A1=0.1 threshold is fixed at all redshifts; yet the manuscript itself notes that high-redshift disks are strongly perturbed (Sec. 2.2) and that the z=0 outer-radius definition is not appropriate at z>0.5 (Sec. 3). Because the headline claim is the redshift evolution of the lopsided fraction, the authors need to show the trend is not an artifact of these choices—for example by varying the lambda_R/R90 cuts and the A1 threshold, or by calibrating A1 to the noise level at each redshift.","section":"Sec. 2.5 and Sec. 4.1"},{"comment":"The interaction proxy counts neighbors with stellar mass ratio >1:10 within R200 at the output snapshot; this is not a measurement of recent tidal interactions. Stable neighbors can be counted while recent fly-bys are missed, as the text itself acknowledges: 'we are not considering the role of recent fly-bys.' Furthermore, Fig. 6 excludes satellite galaxies by construction, although satellites are the population whose environments differ most strongly with redshift. This proxy can support a statement about the presence of potential perturbers, but not the conclusion that 'close tidal interactions are significantly more effective at high-redshift' (Abstract). Merger-tree-based encounter histories or time-resolved orbit classification are needed.","section":"Sec. 4.4.1 and Fig. 6"},{"comment":"The lopsided and symmetric samples are not matched in the structural properties that Sec. 4.3 (Fig. 5) shows are strongly correlated with A1. The bottom panels of Fig. 5 show that the redshift-dependent samples have different mu*, Rh, and R90 distributions, so the larger net accretion rates and neighbor fractions of lopsided galaxies could reflect these structural differences rather than a redshift-dependent efficiency of the proposed mechanisms. In addition, the net accretion rate in Eq. (8) is measured over the preceding 1 Gyr, and Sec. 5.4 explicitly entertains the reverse channel, in which lopsidedness funnels gas inward. Without matched control samples or time-resolved event ordering (accretion/interaction preceding the onset of lopsidedness), the statement that gas accretion and interactions are 'significantly more effective at high-redshift' is an overstatement; the data as presented support a correlation, not a causal mechanism.","section":"Sec. 4.4.2, Eq. 8, Fig. 7, Sec. 5.4"},{"comment":"Because A1 is averaged over Rh<r<1.4R90 and R90 decreases with redshift (Fig. 5), the measurement region changes systematically with redshift. The paper states in Sec. 3 that the trends are not significantly affected by the radial range, but no supporting figure or test is shown. A robustness test fixing the radial range in physical units (or in units of Rh) should be provided, since the radial range is known to change the fraction of lopsided galaxies.","section":"Sec. 3 and Fig. 3"}],"minor_comments":[{"comment":"In the first bullet list, 'Fig. 4.4.1' should be 'Fig. 6'.","section":"Sec. 6"},{"comment":"The phrase 'as previously discussed in Sec. 5.4' is a forward reference; the gas-accretion discussion appears in the following subsection.","section":"Sec. 5.3"},{"comment":"The abstract and Sec. 4.4.2 use 'net accretion rate' for Eq. (8), which also includes stars formed in the disk; the text should clarify that this is a net gas input plus newly formed stellar mass, not a direct inflow measurement.","section":"Eq. 8"},{"comment":"The right panel would benefit from error bars or confidence intervals; the TNG50 volume is small and the sample sizes at z=2 are limited, so the ~60% fraction should be accompanied by an estimate of cosmic variance.","section":"Fig. 3"}],"recommendation":"major_revision","confidential_remarks":"The paper is suitable for A&A after the causal claims are either supported by matched/temporal analysis or toned down. The redshift-dependent sample selection and fixed A1 threshold require sensitivity tests, and the quantitative comparison with Le Bail et al. should be framed as qualitative given the differences in methods and redshift coverage."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth your time. This is the first systematic characterization of lopsidedness in TNG50 across 0<z<2, and it directly engages with the JWST result of Le Bail et al. The headline trend—roughly 60% lopsided at z~1.5–2 declining to ~40% at z=0—is plausible and likely robust. The Fourier method is standard, the simulation is public, and the paper is honest about several analysis choices. The correlation between lopsided amplitude and internal properties (low central stellar mass density, large radius) holding at every redshift is a clean, useful result.\n\nThe main soft spot is the interpretive claim that tidal interactions and gas accretion are 'significantly more effective at high-redshift.' The evidence for that is correlations: lopsided galaxies have more massive neighbors within R200 and higher net accretion rates than symmetric galaxies, and these differences shrink toward z=0. But the lopsided and symmetric samples are not matched for the internal properties that the paper itself shows correlate with A1. High-z galaxies have different size and density distributions, so the higher neighbor fractions and accretion rates could partly reflect structural differences rather than a redshift-dependent efficiency of the mechanisms. On top of that, net accretion is measured over the preceding 1 Gyr, and the paper explicitly discusses the reverse channel—lopsidedness funneling gas inward via torques. So the data support a correlation, not a causal mechanism. The abstract and conclusions go a bit beyond that. I would not call this fatal; the central trend stands, but the causal language needs tempering.\n\nMinor issues: the z=0 selection criteria (lambda_R threshold, A1=0.1) are applied at all redshifts without calibration, and there are no formal significance tests—just IQRs. Neither of these changes my overall read. The single TNG50 volume is a real but standard limitation.\n\nThis paper deserves a serious referee. It is a solid, useful contribution for anyone working on galaxy asymmetries, disk structure, or high-z JWST morphology. I would send it to review with a request to soften the mechanism claims and ideally add a matched control sample or time-resolved event ordering.","headline":"Useful TNG50 result on the redshift evolution of lopsidedness, but the 'more effective mechanisms at high-z' claim is overstated: the comparisons are not matched and reverse causality remains a live possibility.","tokens_in":24932,"tokens_out":1837,"would_cite":true,"duration_ms":18477,"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 the high fraction of lopsided galaxies observed at 1.5<z<2 is reproduced by the TNG50 simulations, that lopsidedness declines toward z=0, and that the decline reflects tidal interactions and gas accretion being more…","keywords":["galaxies: high-redshift","galaxies: interactions","galaxies: structure","galaxies: star formation","lopsidedness","disk galaxies","IllustrisTNG","Fourier decomposition"],"falsifier":"A concrete check: measure the lopsided fraction of a JWST sample at 1.5<z<2 using the same first-mode amplitude metric on mock images of TNG50 galaxies; if the recovered fraction is well below the stated ~60% because the selected disks are systematically different at high z, the redshift trend would be shown to be a selection artifact. Alternatively, running the same selection on a larger TNG volume and finding the high-redshift fraction consistent with ~30% rather than ~60% would falsify the claim.","tokens_in":23948,"feed_emoji":"🌀","tokens_out":4782,"duration_ms":36671,"temperature":0.7,"pith_summary":"This paper uses the TNG50 cosmological simulation to ask whether the high fraction of lopsided disk galaxies seen by JWST at 1.5<z<3 is real and physical, rather than an observational artifact. Selecting disk-like galaxies independently at each redshift from z=0 to z=2, it finds that about 60% are lopsided at 1.5<z<2, decreasing to roughly 40-45% at z=0. Lopsidedness does not depend on local environment, but it does correlate strongly with internal structure: galaxies with low central stellar mass density and extended disks are the most susceptible. The paper argues that the same triggers, recent tidal interactions with mass ratio >1:10 and gas accretion followed by star formation, work at all redshifts, but that both are significantly more effective at high redshift. If correct, this explains the JWST excess as genuinely more efficient perturbation mechanisms in the early universe.","feed_headline":"Simulations explain why early disk galaxies were so lopsided","feed_subtitle":"TNG50 simulations show the same trigger mechanisms were simply more efficient when the universe was young.","key_machinery":"The measurement engine is the azimuthal Fourier decomposition of the face-on stellar mass distribution of TNG50 disk galaxies. The first-mode amplitude, $A_1$, is computed in concentric annuli of width 0.1 kpc and averaged over $R_h < r < 1.4 R_{90}$, inside $|h_z| < 2 h_{90}$, with $A_1 > 0.1$ defining a lopsided galaxy. The sample is built at each redshift with the $\\lambda_R$-$\\epsilon$ diagram ($\\lambda_R > 0.31\\sqrt{\\epsilon}$, $\\epsilon > 0.4$, $R_{90} > 3$ kpc), so each redshift is an independently selected population of disk-like galaxies. The redshift-resolved sample is what allows the authors to separate internal susceptibility from the efficiency of external triggers.","core_discovery":"On the paper's own terms, the central discovery is that the redshift trend in lopsidedness is driven by the efficiency of perturbation mechanisms, not by sample selection. In TNG50, the fraction of lopsided disk galaxies (first Fourier mode amplitude $A_1$ averaged over $R_h < r < 1.4 R_{90}$ exceeding 0.1) reaches ~60% at 1.5<z<2, matching the JWST fraction of 64% reported for 22 dusty star-forming galaxies, and falls toward z=0. The same internal-property correlation seen locally holds at every redshift: low central stellar mass density and larger disk size make a galaxy more prone to strong lopsidedness. Yet since high-redshift disks are actually denser and more compact, the observed decrease in lopsidedness toward low redshift cannot come from internal properties alone; it implies the triggering mechanisms are different or more efficient at high redshift. The paper identifies recent close tidal interactions (mass ratio >1:10 within $R_{200}$) and net gas accretion with subsequent star formation as the two mechanisms, finding both act at all redshifts but with greater effect at $z \\gtrsim 1$.","pith_inferences":["Editorial: if the redshift dependence of $A_1$ is as strong as TNG50 suggests, lopsidedness could serve as a cheap, high-redshift proxy for the merger and accretion rate, applicable to large JWST samples without spectroscopy.","Editorial: the claim that gas accretion with subsequent star formation triggers lopsidedness could be tested directly in the simulation by checking whether the accreted gas has an asymmetric angular momentum distribution and whether the $m=1$ mode grows right after accretion events, a step the authors leave to follow-up work.","Editorial: the fixed $A_1=0.1$ threshold may mix genuinely different perturbation states across redshift; re-deriving the threshold from the simulated amplitude distribution at each z would show whether part of the trend is threshold-driven.","Editorial: applying the same analysis to a larger-volume, lower-resolution TNG box would check whether cosmic variance in the small TNG50 volume affects the claimed ~60% high-redshift fraction."],"forward_implications":["The ~60% lopsided fraction at 1.5<z<2 in simulations supports the interpretation that JWST's 64% lopsided galaxies are not mostly observational bias.","Because high-redshift disks are denser and more compact, the strong internal-property correlation would predict less lopsidedness at high z if triggers were equally efficient; the opposite trend is evidence for more efficient early-universe perturbation mechanisms.","Local environment (10th-nearest-neighbor density) does not set lopsidedness; at all redshifts the susceptibility is set by internal structure, so environment acts only through the frequency of encounters.","Gas accretion followed by star formation is a viable lopsidedness trigger at all redshifts, and at z>1 it is strong enough to also explain the elevated lopsidedness of central galaxies, not just satellites.","The outside-in quenching scenario, in which lopsidedness funnels gas to the center and quenches the disk, is consistent with TypeII galaxies' star-forming cores embedded in quenched lopsided disks at high redshift."],"supporting_citations":[{"why":"Provides the JWST sample of 22 dusty star-forming galaxies whose 64% lopsided fraction is the observational benchmark the simulations must reproduce.","marker":"Le Bail et al. (2023)"},{"why":"Sets the local-universe lopsided fraction (~30%) and the classic observational definition of lopsidedness used for comparison.","marker":"Zaritsky & Rix (1997)"},{"why":"Establishes the TNG50 lopsidedness measurement method and the $A_1=0.1$ threshold at $z=0$ that this paper extends to higher redshift.","marker":"Varela-Lavin et al. (2023)"},{"why":"Previous TNG50 $z=0$ study of lopsidedness versus environment and internal properties that this work extends across cosmic time.","marker":"Dolfi et al. (2023)"},{"why":"The TNG50 simulation release paper, supplying the simulation data and its resolution.","marker":"Nelson et al. (2019)"},{"why":"Simulated asymmetric gas accretion and showed it can produce strong lopsided perturbations, the mechanism the authors invoke for gas accretion.","marker":"Bournaud et al. (2005)"},{"why":"Defines the $\\lambda_R$-$\\epsilon$ threshold used to select rotation-dominated disk galaxies.","marker":"Emsellem et al. (2011)"},{"why":"Provides the net accretion rate and disk height definitions used to quantify gas accretion.","marker":"Iza et al. (2022)"}],"fun_headline_variants":["Simulations reveal why early disk galaxies were so lopsided","Why early galaxies were lopsided: TNG50 shows triggers more efficient","Early galaxies lopsided due to more efficient interactions and gas inflow","TNG50: young disk galaxies lopsided because triggers were stronger"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis assumes the local-universe disk-selection criteria ($\\lambda_R > 0.31\\sqrt{\\epsilon}$, $\\epsilon > 0.4$, $R_{90} > 3$ kpc) and the fixed lopsided threshold $A_1=0.1$ remain valid at all redshifts up to $z=2$, so that the independently selected samples are comparable in what they represent.","fun_headline_variants_meta":{"raw":{"variants":["Simulations reveal why early disk galaxies were so lopsided","Why early galaxies were lopsided: TNG50 shows triggers more efficient","Early galaxies lopsided due to more efficient interactions and gas inflow","TNG50: young disk galaxies lopsided because triggers were stronger"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00023,"raw_usage":{"total_tokens":1590,"prompt_tokens":1159,"completion_tokens":431,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":775,"completion_tokens_details":{"reasoning_tokens":354}},"tokens_in":775,"tokens_out":431,"duration_ms":4261,"temperature":1.0,"reasoning_tokens":354,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T10:11:36.580014+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check: measure the lopsided fraction of a JWST sample at 1.5<z<2 using the same first-mode amplitude metric on mock images of TNG50 galaxies; if the recovered fraction is well below the stated ~60% because the selected disks are systematically different at high z, the redshift trend would be shown to be a selection artifact. Alternatively, running the same selection on a larger TNG volume and finding the high-redshift fraction consistent with ~30% rather than ~60% would falsify the claim.","supporting_citations":[{"cited_title":"& Rix, H.-W","cited_arxiv_id":null,"evidence_quote":"Sets the local-universe lopsided fraction (~30%) and the classic observational definition of lopsidedness used for comparison."},{"cited_title":"2011, Monthly Notices of the Royal Astronomical Society, 414, 888","cited_arxiv_id":null,"evidence_quote":"Defines the $\\lambda_R$-$\\epsilon$ threshold used to select rotation-dominated disk galaxies."},{"cited_title":"G., Scannapieco, C., Nuza, S","cited_arxiv_id":null,"evidence_quote":"Provides the net accretion rate and disk height definitions used to quantify gas accretion."}],"review_version":1}