{"id":"2411c6d6-b42f-4897-b7cb-df941950c515","arxiv_id":"1908.00547","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"In TNG100, strong bars form in galaxies with early, prominent discs and are associated with rapid central star formation quenching, unlike unbarred galaxies.","lead":"This paper studies how strongly barred galaxies form and stop forming stars in the TNG100 cosmological simulation. It finds that strong bars appear in galaxies with early-built discs and are tied to rapid quenching of the galaxy's central region.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The causal bar-quenching claim is under-identified: the control sample is matched only by z=0 stellar mass (§2.5), and at tnorm=0 barred galaxies already differ in mass, bulge, and BH growth, so the post-bar SFR drop may trace early assembly/AGN feedback rather than the bar.","rationale":"Good-faith reading: the paper's descriptive findings are careful and largely supported by public TNG100 data; the evolutionary alignment via tnorm is a genuine step forward, and the paper is appropriately cautious in places (§5.2–5.3). The load-bearing weakness is the causal interpretation of the post-bar SFR decline. Matching only by z=0 stellar mass does not remove the fact that barred galaxies are already different before the bar forms: higher stellar mass, lower B/T, and more massive black holes. Since AGN feedback at fixed mass is known to quench galaxies in TNG, the bar-quenching signal may be an assembly-history artefact rather than a bar effect. The survivor selection acknowledged in §5.3 compounds this by excluding bars that form and later dissolve, which could be preferentially linked to the quenching process. My recommendation is unchanged from the reader's: conditional acceptance, with the causal language revisited after a control matched at bar formation and preferably with TNG50 or other higher-resolution checks.","tokens_in":26963,"tokens_out":7109,"duration_ms":77597,"concrete_test":"Recompute the §4.2–4.4 evolution analysis with a control sample matched at tnorm=0 rather than at z=0: for each strong-bar galaxy, select unbarred galaxies with similar stellar mass, halo mass, BH mass, and D/T at the same cosmic time at which the bar forms, and then compare the subsequent central SFR and SFEdr/SFEgal tracks. If the differential drop disappears (falls within the 20–80th percentile scatter), the bar-quenching claim is not supported; if it persists, the causal case is materially strengthened. As a secondary check, repeat the analysis including bars that form and dissolve before z=0, as acknowledged in §5.3, to test survivor bias.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—bar-driven quenching, possibly assisted by nuclear feedback—rests on comparing z=0 strong-bar galaxies with unbarred controls matched only by z=0 stellar mass (§2.5). At tnorm=0, the epoch of bar formation, the strong-barred sample is already offset from the control set: up to ~0.4 dex more stellar mass, lower bulge fraction, and more massive SMBHs (§4.3, Figs 12 and 15). Each of these pre-existing differences is itself associated with earlier quenching in TNG (e.g., kinetic-mode AGN feedback; Weinberger et al. 2018). The observed decline in central SFR and SFEdr/SFEgal after tnorm=0 is therefore equally compatible with a scenario in which bars form preferentially in galaxies that are already on a path to early quenching, with the bar being a correlate rather than the cause. The within-bar versus outside-bar contrast in Fig. 13 is suggestive, but the same aperture in unbarred controls also declines, so the differential timing is not by itself decisive. The paper's own §5.3 flags an additional selection effect: only bars surviving to z=0 are analysed, and a large population of early bars that dissolve is expected; if destruction is linked to the quenching process, survivor bias further weakens the causal reading. Descriptive claims (bar fraction, mass trend, lower gas content in barred galaxies) are not in question; the causal attribution is under-identified.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes strong bars in 270 z=0 disc galaxies from TNG100 with stellar masses 10^10.4-10^11 Msun. It identifies 107 barred galaxies (59 strong, 48 weak) via Fourier decomposition and finds that the barred fraction increases with stellar mass, barred galaxies have lower gas-to-stellar mass ratios and lower sSFRs than unbarred ones, and strong bars form mostly at 0.5<z<1.5 in galaxies with early-built discs. Tracking progenitor histories, the authors compare strong-barred galaxies with unbarred controls matched by z=0 stellar mass and report a rapid drop in central star formation after bar formation, which they interpret as evidence for bar quenching, possibly assisted by nuclear feedback.","tokens_in":27225,"tokens_out":5391,"duration_ms":57734,"significance":"If the causal interpretation holds, this is an important cosmological confirmation of bar-driven quenching and a demonstration that early disc assembly and nuclear feedback jointly set the bar population in a full hydrodynamical simulation. The paper is valuable for its descriptive measurements: bar fractions, bar lengths, mass trends, gas content, and comparison with observed local samples, all in a state-of-the-art cosmological run. The bar-age definition using stability conditions and the explicit treatment of merger histories are careful features. However, the central causal claim about bar quenching is not uniquely supported by the current comparison setup, as detailed below.","major_comments":[{"comment":"The control sample of unbarred galaxies is matched to strong-barred galaxies only by stellar mass at z=0 (§2.5). At tnorm=0, the epoch of bar formation, the strong-barred sample is already offset from the controls by up to ~0.4 dex in stellar mass and has lower bulge-to-total ratio and more massive SMBHs (§4.3, Figs 12 and 15). Because each of these pre-existing differences is independently associated with earlier quenching in TNG, the observed post-bar decline in central SFR and SFEdr/SFEgal (Figs 11 and 13) is equally compatible with strong bars forming preferentially in galaxies already on a path to early quenching. The within-bar versus outside-bar contrast in Fig. 13 is suggestive, but the same aperture in unbarred controls also declines, so the differential timing is not by itself decisive. Please add a control matched at tbar on stellar mass, B/T, and BH mass (or a propensity-score/conditional analysis) and show that the quenching signal survives.","section":"§2.5 and §4.3 (Figs 12, 15)"},{"comment":"Only bars that survive to z=0 enter the sample, and the paper itself notes in §5.3 that a large population of early bars that are subsequently destroyed is expected. Bar destruction through gas inflows, buckling, or interactions is plausibly correlated with the same central gas consumption that the paper attributes to the bar, so survivor bias can exaggerate the apparent post-bar quenching signal. Since this selection effect is acknowledged but not quantified, the central causal reading (\"indicative of bar quenching\") is weakened; at minimum the authors should test how many bars form and dissolve in TNG100 and whether their host properties differ, or restrict the causal claim to the survivor population.","section":"§5.3"},{"comment":"Fig. 15 shows that strongly barred galaxies have systematically higher SMBH masses and higher kinetic-mode energy rates than unbarred controls at all times, and the specific BH accretion rate and central sSFR follow similar evolutions. This is as consistent with early BH growth and feedback creating the conditions for bar formation and quenching as with the bar being the primary quenching agent. The statement that AGN feedback plays a \"secondary, or at least ancillary\" role is therefore not established by the presented evidence; it is a speculation, as the authors themselves note. Please either supply a quantitative decomposition (e.g., gas depletion timescales, torque estimates, or a comparison with a TNG variant without BH feedback) or explicitly present the causal hierarchy as an open question rather than a conclusion.","section":"§5.2 and Fig. 15"},{"comment":"The SFE comparison relies on median tracks from 59 strong bars divided into three mass bins, but the paper does not report the number of galaxies contributing to each median bin or any confidence interval for the difference between barred and unbarred samples. With roughly 20 objects per bin, the apparent rapid drop after tnorm=0 could be driven by individual outliers. Please report per-bin sample sizes and uncertainty estimates (e.g., bootstrap confidence intervals or a significance test for the differential drop in SFEdr/SFEgal).","section":"§4.4, Fig. 13"}],"minor_comments":[{"comment":"The typeset definition of tnorm is ambiguous; if the intended definition is (tbar - tlookback)/tbar, please write it explicitly with parentheses.","section":"§2.4, Eq. (4)"},{"comment":"The phase-constancy condition \"Φ< rbar const\" needs a precise tolerance; currently it is unclear how strictly the phase must be constant within rbar.","section":"§2.3"},{"comment":"The axis label of the inset panel is missing; the inset shows logarithmic SFR but the reader must infer the quantity.","section":"Fig. 11"},{"comment":"The caption states \"sSFR∼10^12 yr^-1\" for the passive galaxy; the exponent should presumably be negative (10^-12 yr^-1).","section":"Fig. 9 caption"},{"comment":"Several inline ranges such as \"M∗ 10^10.4−11M⊙\" are missing the exponent notation and are hard to read; please format these consistently as 10^10.4−10^11 M⊙.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The descriptive results are solid and the paper is generally well written. My main concern is that the abstract and Section 5.2 go beyond what the comparison setup can support: the control sample matches only z=0 stellar mass, and the authors themselves note the survivor bias and the speculative nature of the AGN role. If the authors add a control sample matched at the epoch of bar formation or systematically soften the causal language, the paper would be acceptable. I would not reject, because the measurements of the bar population and its evolution are useful and the limitations are partly acknowledged."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is the first population-level study of strong bars in TNG100, and the descriptive results are solid. The causal claim about bar-driven quenching is under-identified, but the paper is honest about its limitations and deserves a serious referee.\n\nWhat's new is the normalized time axis. By aligning galaxies on bar-formation time rather than redshift, they show that barred galaxies already have earlier-assembled discs, more stellar mass, and more massive black holes at the epoch of bar formation, and that their central star-formation efficiency drops sharply afterward. That is a real step beyond z=0 snapshots and prior simulation studies.\n\nThe soft spot is the control sample. Unbarred galaxies are matched only by z=0 stellar mass. At tnorm=0, the barred sample is already offset by up to 0.4 dex in mass, lower bulge fraction, and higher BH mass—each a known correlate of quenching in TNG. So the post-bar SFR decline could be the continuation of an earlier trajectory that also helped produce the bar, making the bar a correlate rather than cause. The within-bar versus outside-bar SFE contrast is suggestive, but the same central aperture in unbarred controls also declines, so the differential timing is not decisive.\n\nSmaller issues: 59 strong bars split into three mass bins is thin, the A2 thresholds (0.2, 0.3) are hand-picked, and only bars surviving to z=0 are analyzed. The authors flag the survivor bias in §5.3 and use appropriately careful language (\"indicative,\" \"possibly\"). That helps, but the gap between data and causal claim is real.\n\nFor a revision, I'd want a control sample matched on halo mass or assembly history (or a demonstration that the results hold with such matching), a robustness check on the A2 thresholds, and quantification of the destroyed-bar population. TNG50 would be the natural place.\n\nWho this is for: people working on bars in cosmological simulations, and anyone who wants a clean example of the control-sample pitfalls in secular evolution studies. The descriptive parts will be cited.","headline":"First population-level look at strong bars in TNG100; descriptive results hold, but the bar-quenching causal claim is under-identified by the single-variable control sample.","tokens_in":27884,"tokens_out":2865,"would_cite":true,"duration_ms":26495,"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":"In the TNG100 simulation, strong stellar bars form in early-assembled discs and quench their galaxies' centres shortly after appearing.","keywords":["galaxy bars","bar quenching","star formation quenching","disc galaxy evolution","cosmological hydrodynamical simulation","AGN feedback","TNG100 simulation","stellar mass assembly"],"falsifier":"A decisive run would switch off black-hole feedback in the same simulation: if strong bars still quench their central regions shortly after formation, the bar alone suffices; if the central gas keeps forming stars, the quenching attributed to the bar is actually driven by the black hole. Observational confirmation would come from resolved maps of central star formation in galaxies caught just after bar formation at redshift around 0.5 to 1, which should show the star-formation efficiency dropping first inside the bar radius.","tokens_in":26708,"feed_emoji":"🌌","tokens_out":15079,"duration_ms":135266,"temperature":0.7,"pith_summary":"The paper uses the TNG100 cosmological simulation to argue that strong stellar bars are not randomly placed features: they grow in disc galaxies that assembled their dominant disc early, and their appearance is followed by a rapid shutdown of central star formation. Among 270 massive disc galaxies at redshift zero, 40 per cent are barred and 22 per cent host strong bars, with the strong-bar fraction rising with stellar mass and with older bars found in more massive galaxies. The strong bars form between redshift 0.5 and 1.5, during a phase of enhanced star formation and black-hole growth, and after the bar settles the star-formation efficiency inside its radius drops steeply while mass-matched unbarred galaxies keep forming stars. The paper reads this as evidence for bar quenching, possibly helped by nuclear feedback from an early-grown black hole. The result connects an observable internal structure, the bar, to a possible route for shutting down star formation in massive galaxies.","feed_headline":"Bars form early, then switch off galaxy centers","feed_subtitle":"Simulated barred galaxies go quiet at their centers soon after bar formation; unbarred counterparts stay active.","key_machinery":"The central tool is the second Fourier harmonic amplitude $A_2$ of the face-on stellar surface density, whose peak sets the bar strength and whose peak radius sets the bar length; a strong bar is a stable structure with $A_2$ at least 0.3 and a constant phase inside that radius. The comparison clock is the normalized time since bar formation, $t_{\\rm norm}$, which is zero when the bar becomes stable and one at $z=0$; this allows the histories of barred and unbarred galaxies to be aligned at the same stage of bar evolution. A control sample of unbarred galaxies matched by $z=0$ stellar mass supplies the counterfactual baseline.","core_discovery":"Within the TNG100 cosmological simulation, strong stellar bars are the product of a specific assembly history: the galaxy must build a dominant, dynamically cool disc while its bulge is still small, and it must do so early, between redshift 1.5 and 0.5. These galaxies show enhanced star formation and black-hole accretion while the bar is being built, then a rapid drop in star-formation efficiency inside the bar radius once the bar settles. The nuclear region becomes gas-poor and quenched well before $z=0$, while unbarred galaxies matched in $z=0$ stellar mass continue forming stars on the main sequence. The paper interprets this sequence as bar quenching: the bar torques gas inward, exhausts the central cold gas, and shuts off star formation, with early black-hole feedback acting as a supporting rather than the primary cause.","pith_inferences":["If the control sample were matched on halo mass, environment, and early assembly history as well as stellar mass, the causal attribution to the bar could be cleaner; until then, pre-existing differences between the populations remain a candidate explanation for the quenching.","A direct model-variation experiment is implied: simulations without black-hole feedback should produce fewer or weaker strong bars if early AGN feedback is what keeps bulges small, and barred galaxies in those runs would reveal whether the bar alone can quench.","The same bar-detection method could be applied to every snapshot to produce a predicted cosmic bar fraction curve, giving high-redshift surveys a quantitative target to confirm or refute the early-formation scenario.","If bars are later destroyed by mergers or buckling, the model predicts that central star formation can re-ignite only if fresh gas is available, which connects this work to studies of bar lifetimes."],"forward_implications":["The cosmic bar fraction should be a clock for when massive discs became dynamically cold: more massive galaxies should host older bars and quench earlier, as the paper finds.","A sharp drop in star-formation efficiency inside the bar after formation implies that bars are a plausible internal quenching channel even without changes in large-scale environment.","At fixed final stellar mass, the barred and unbarred populations encode different assembly histories: barred galaxies built their stars and black holes early, unbarred galaxies grew later and more gradually.","Since merger histories are similar for the two populations, interactions are unlikely to be the main trigger of strong bars in this mass range; the early disc and small bulge are the decisive conditions.","The parallel early growth of black holes and later central quenching suggests AGN feedback and bar formation may work together, with the bar establishing the conditions for its own nuclear shutoff."],"supporting_citations":[{"why":"Supplies the Fourier decomposition method used to measure bar strength and length.","marker":"Athanassoula & Misiriotis 2002"},{"why":"Provides the kinematic decomposition used to classify galaxies as disc-dominated.","marker":"Genel et al. 2015"},{"why":"Sets the observed redshift trend of declining bar fraction that the early formation epoch must match.","marker":"Sheth et al. 2012"},{"why":"Offers the observational link between strong bars and low specific star formation that this paper tests.","marker":"Gavazzi et al. 2015"},{"why":"Gives the observed increase of bar fraction with stellar mass that the simulation reproduces.","marker":"Cervantes-Sodi et al. 2015"},{"why":"Shows in a zoom-in simulation that early AGN feedback can keep bulges small and allow a strong bar to form.","marker":"Bonoli et al. 2016"},{"why":"Traces bar-driven gas inflows and central gas removal, the mechanism invoked for nuclear quenching.","marker":"Spinoso et al. 2017"},{"why":"Demonstrates in idealised simulations that gas-rich galaxies form later and weaker bars, supporting the gas-poor condition.","marker":"Athanassoula et al. 2013"},{"why":"Shows isolated barred galaxies experience a rapid drop in star formation efficiency, supporting bar quenching without AGN.","marker":"Khoperskov et al. 2018"},{"why":"Establishes AGN feedback as the TNG quenching mechanism for massive haloes, used to weigh its role against the bar.","marker":"Weinberger et al. 2018"}],"fun_headline_variants":["Strong bars form early, then switch off galaxy centers","Early bars quench galactic nuclei in simulations","Bar formation precedes central quenching in TNG100","Strong bars emerge by z=1, then starve their centers"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The comparison treats unbarred galaxies matched only by stellar mass at $z=0$ as the control for what barred galaxies would look like without a bar, so if the two populations already differ in halo mass, environment, bulge size, or early black-hole growth, the later differences in gas and star formation cannot be attributed to the bar alone.","fun_headline_variants_meta":{"raw":{"variants":["Strong bars form early, then switch off galaxy centers","Early bars quench galactic nuclei in simulations","Bar formation precedes central quenching in TNG100","Strong bars emerge by z=1, then starve their centers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000173,"raw_usage":{"total_tokens":1329,"prompt_tokens":1044,"completion_tokens":285,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":660,"completion_tokens_details":{"reasoning_tokens":222}},"tokens_in":660,"tokens_out":285,"duration_ms":3612,"temperature":1.0,"reasoning_tokens":222,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:47:30.767818+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive run would switch off black-hole feedback in the same simulation: if strong bars still quench their central regions shortly after formation, the bar alone suffices; if the central gas keeps forming stars, the quenching attributed to the bar is actually driven by the black hole. Observational confirmation would come from resolved maps of central star formation in galaxies caught just after bar formation at redshift around 0.5 to 1, which should show the star-formation efficiency dropping first inside the bar radius.","supporting_citations":[{"cited_title":"M., Hernquist, L., et al., 2015, , 804, L40","cited_arxiv_id":null,"evidence_quote":"Provides the kinematic decomposition used to classify galaxies as disc-dominated."},{"cited_title":"D., Combes F., 2018, A&A, 609, A60","cited_arxiv_id":null,"evidence_quote":"Shows isolated barred galaxies experience a rapid drop in star formation efficiency, supporting bar quenching without AGN."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes AGN feedback as the TNG quenching mechanism for massive haloes, used to weigh its role against the bar."}],"review_version":1}