{"id":"5aa7822c-264d-418b-9d0e-b64c51d573d7","arxiv_id":"2607.08824","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"In Thesan-Zoom, bursty star formation prevents prolonged survival of large dust reservoirs (M_dust/M_star ≳ 10^{-3}), so it can explain z≳10 UV-bright galaxies only if it settles by z∼8.","lead":"Radiation-hydrodynamical zoom simulations show that bursty star formation rapidly destroys and ejects dust, so large dust reservoirs cannot persist for long. This limits burstiness as an explanation for UV-bright galaxies at z≳10 unless the bursts settle by z∼8, where dust is observed.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"The incompatibility claim rests on dust being co-spatial with feedback and fully exposed; missing dense-clump shielding is the softest link.","rationale":"The paper is internally consistent and the time-variable dust cycle is well documented (Figs. 3, 7, 8, 15). The strongest claim is explicitly model-dependent and the authors already state that missing shielding physics or a rapid settling of burstiness by z ∼ 8 would be required for consistency with observed dust reservoirs. The reader’s weakest_assumption is exactly the load-bearing soft spot; no deeper internal contradiction appears. Therefore the CONDITIONAL verdict and medium correctness risk remain appropriate; no upgrade or downgrade is warranted. The concrete test above is the minimal higher-resolution / multiphase check that would decide whether the concern actually moves the claim.","tokens_in":45963,"tokens_out":642,"duration_ms":6087,"concrete_test":"Re-run one intermediate-mass target (e.g. m11.5z4 or m12.2z4) with an explicit multiphase sub-grid: treat all gas with n_H > 0.1 cm^{-3} and T < 10^4 K as containing an unresolved cold phase (n ≥ 10 cm^{-3}, T ≤ 100 K) that boosts accretion (as in Aoyama et al. 2018 / Trayford et al. 2026) and is only partially coupled to the SN energy kernel. Measure the fraction of cosmic time with M_dust/M_star ≥ 10^{-3} and the median IR-bright duration. If either quantity rises by ≳ factor 2 relative to the fiducial run, the incompatibility claim weakens.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (bursty SF prevents prolonged survival of M_dust/M_star ≳ 10^{-3}) is driven by the cycle in Sec. 5.1 and Fig. 15: dust grows with dense gas, then is preferentially destroyed/ejected once stars form because it is co-spatial with the feedback sites. The paper itself flags that unresolved cold dense clumps (n ≳ 10 cm^{-3}, T ≲ 100 K) that could shield dust are missing (Sec. 2.2, Sec. 5.4, Appendix B). If such clumps exist and survive the SMUGGLE injection kernel, both the destruction efficiency and the post-burst recovery timescale change, so the duty cycle of dust-poor phases (and therefore the implication for the z ≳ 10 UVLF) is no longer guaranteed. The single-grain-size a_eff = 0.1 μm and accretion-only-in-star-forming-gas assumptions further lock the model into this co-spatial destruction picture. The reader correctly identified this as the weakest assumption; it is load-bearing for the incompatibility statement.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper analyses cosmic dust in the Thesan-Zoom radiation-hydrodynamical zoom-in suite, which couples an on-the-fly dust model (adapted from McKinnon et al. 2016, 2017) to multi-phase ISM physics, SMUGGLE feedback and radiative transfer at z ≥ 3. Simulated galaxies reproduce observed DGR and DMR trends with metallicity and broadly match dust temperatures and UV–IR offsets, but show a dust deficit at high sSFR and systematically low A1500 even after TODDLERS post-processing of unresolved birth-cloud dust. The central claim is that the emergent bursty star-formation cycle (illustrated in Sec. 5.1 and Fig. 15) produces short-lived IR-bright phases (median ~20 Myr) and longer dust-poor phases, so that large reservoirs (M_dust/M_star ≳ 10^{-3}) cannot survive for a significant fraction of cosmic time; therefore bursty SF can explain the z ≳ 10 UVLF excess only if it settles by z ~ 8, or the model is missing shielding physics.","tokens_in":46275,"tokens_out":1231,"duration_ms":10324,"significance":"If the result holds, it supplies a concrete, falsifiable constraint linking two major high-z puzzles: the JWST UVLF excess and the presence of substantial dust reservoirs by z ~ 8. The work is valuable because it is one of the few suites that self-consistently couples live dust, multi-phase ISM and radiation, produces quantitative predictions (IR-phase duty cycle, T_dust–ΔMS correlation, UV–IR offsets) that can be tested with ALMA/NOEMA and JWST, and explicitly flags the model limitations that would reverse the conclusion. The forward-modelling of UV–IR offsets and dust surface-density distributions, and the transparent comparison to multiple observational samples and other simulations, are clear strengths.","major_comments":[{"comment":"Sec. 5.1 and Fig. 15 establish the co-spatial growth–destruction cycle that underpins the incompatibility claim, yet Sec. 2.2, Sec. 5.4 and Appendix B acknowledge that unresolved cold dense clumps (n ≳ 10 cm^{-3}, T ≲ 100 K) capable of shielding dust from the SMUGGLE injection kernel are missing. Because the paper itself identifies this as the softest link, the central statement that bursty SF is incompatible with prolonged survival of M_dust/M_star ≳ 10^{-3} should be explicitly conditioned on the absence of such shielding, and a quantitative estimate (or at least a clear statement of the required change in destruction efficiency or recovery timescale) should be added so that the UVLF implication remains falsifiable rather than model-contingent.","section":null},{"comment":"Sec. 4.3 and Fig. 14 show that A1500 remains low compared with observations even after TODDLERS birth-cloud dust is included. The abstract and Sec. 5 correctly note that a shielding mechanism is therefore required, but the paper does not quantify how much additional shielded dust (or what change in feedback topology) would be needed to reach the observed A1500 while preserving the bursty UVLF success at z ≳ 10. Without that estimate the claim that the model is already close enough for the duty-cycle argument to be robust is under-supported.","section":null},{"comment":"The single effective grain size a_eff = 0.1 μm and the restriction of accretion to star-forming gas (Eq. 1 and Sec. 2.2) lock dust production to the same dense gas that later hosts feedback. Sec. 5.4 mentions grain-size evolution only in passing. A short test or literature-based estimate of how a multi-bin size distribution (or accretion outside star-forming gas) would alter the post-burst recovery timescale would strengthen the claim that the incompatibility is robust to dust-physics uncertainties.","section":null}],"minor_comments":[{"comment":"Fig. 2 caption and text: the vertical stripes of constant M_star are attributed to rapid dust growth; a brief note that they could also reflect temporary dust ejection followed by re-accretion would avoid over-interpretation.","section":null},{"comment":"Sec. 3.6: the statement that dust temperatures >100 K 'should be easily observable but have not been detected' needs a short caveat on the limited volume and mass range of the zoom sample.","section":null},{"comment":"Eq. (1): the ad-hoc assignment of T = 10^4 K to star-forming gas for the accretion timescale is tested in Appendix B, but the main text should flag this assumption more prominently when the growth rate is first introduced.","section":null},{"comment":"Fig. 12: the assumed Gaussian PSF with σ = 0.5 pkpc is reasonable for a population comparison, but a sentence noting the range of actual ALMA beams in the REBELS/ALPINE samples would help readers judge the comparison.","section":null},{"comment":"Throughout: a few typographical inconsistencies remain (e.g., 'thesan-zoom' capitalisation, occasional missing spaces around Å). A light copy-edit pass would improve readability.","section":null}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a solid contribution from a high-profile simulation suite and is appropriate for a major astrophysics journal. The central claim is interesting and well-motivated, but it is currently overstated relative to the acknowledged resolution and dust-physics limitations. A major-revision decision that forces the authors to condition the incompatibility statement and to quantify the shielding requirement will produce a more durable paper without requiring new simulations. I see no citation or novelty issues."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The punchline is simple: inside Thesan-Zoom, bursty star formation and prolonged dust survival (M_dust/M_star ≳ 10^{-3}) do not coexist. Dust grows with dense gas, then is preferentially destroyed or ejected once stars form because it sits co-spatial with the feedback sites. IR-bright phases last only ~20 Myr median; high-sSFR systems are systematically dust-poor. That forces a clean implication for the JWST UVLF excess: if burstiness is the explanation at z ≳ 10, it has to settle by z ~ 8 where large dust reservoirs are already seen. The authors state this without overclaiming.\n\nWhat is new is the systematic dust census on this suite—DGR/DMR vs Z, dust mass vs M_star and sSFR, T_dust vs ΔMS, UV–IR offsets, surface-density distribution, and A_1500 with TODDLERS post-processing—plus the explicit duty-cycle numbers and the incompatibility statement. The paper does the comparisons carefully against ALMA/JWST/DLA samples and other simulations, and the time-evolution panels (especially Fig. 15) make the cycle transparent. Forward modelling of offsets and quasar-sightline Σ_dust is done properly. Citation pattern is normal for a Thesan companion; no circularity in the central claim.\n\nThe soft spot is real and load-bearing, but the authors flag it themselves. The single effective grain size, accretion only in star-forming gas, and SMUGGLE injection kernel leave no room for unresolved cold dense clumps that could shield dust. If those clumps exist and survive, both destruction efficiency and post-burst recovery change, so the dust-poor duty cycle (and therefore the UVLF implication) is no longer guaranteed. Attenuation stays too low even with TODDLERS. That is a genuine model limitation, not a hidden flaw; the paper is honest about it.\n\nThis is for people working on high-z dust, feedback, or the UVLF excess. It deserves a serious referee. I would cite the duty-cycle numbers and the incompatibility statement when discussing bursty models. Engage with it; the result is useful even if later higher-resolution or multi-grain runs soften the conclusion.","headline":"Solid dust analysis of Thesan-Zoom that cleanly shows, inside this model, bursty SF cannot keep large dust reservoirs; the UVLF implication is real but rests on missing dense-clump shielding.","tokens_in":46921,"tokens_out":561,"would_cite":true,"duration_ms":6743,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Bursty star formation prevents large dust reservoirs from lasting long in early galaxies.","keywords":["cosmic dust","bursty star formation","high-redshift galaxies","dust survival","UV attenuation","thesan-zoom","ISM feedback","dust temperature"],"falsifier":"Deep continuum or absorption measurements showing large dust reservoirs (M_dust/M_star greater than or equal to 10^{-3}) coexisting with high specific star-formation rates in a statistical sample of z greater than or equal to 10 UV-bright galaxies would falsify the claimed incompatibility.","tokens_in":46866,"feed_emoji":"🌌","tokens_out":829,"duration_ms":7609,"temperature":0.7,"pith_summary":"This paper argues that, in a self-consistent radiation-hydrodynamical model of early galaxies, short intense star-formation bursts repeatedly destroy and eject dust so thoroughly that large dust reservoirs cannot persist for most of cosmic time. The simulated systems match several observed dust-to-gas and dust-to-metal trends and produce realistic dust temperatures and UV–IR spatial offsets during brief IR-bright windows, yet they under-produce UV attenuation and dust mass at high specific star-formation rates. The central claim is therefore that burstiness and prolonged dust survival are incompatible inside this framework. That incompatibility matters because bursty star formation is frequently invoked to explain the excess of UV-bright galaxies seen at redshift greater than or equal to 10: if the same burstiness also erases dust, the bursts must settle by redshift approximately 8, where substantial dust is already observed, or else missing physics must protect dust from feedback.","feed_headline":"Bursty star formation erases lasting dust in early galaxies","feed_subtitle":"Large dust reservoirs cannot survive long; the high-z UV excess may require bursts to settle by z~8","key_machinery":"The coupled on-the-fly dust model (formation via stellar yields, metallicity-dependent accretion, thermal and supernova sputtering) fully linked to multi-phase ISM and SMUGGLE stellar feedback; the machinery shows that feedback co-spatial with newly formed dust preferentially destroys and ejects it after each burst.","core_discovery":"In the thesan-zoom model, bursty star formation prevents large dust reservoirs (M_dust/M_star greater than or equal to 10^{-3}) from surviving over a significant fraction of cosmic time; short-lived IR-bright phases (median duration roughly 20 Myr) alternate with longer dust-poor phases, so that observed dust properties and the high-redshift UV luminosity function can be reconciled only if burstiness declines rapidly by z approximately 8 or if additional shielding physics is present.","pith_inferences":["If unresolved cold clumps systematically protect dust, models that currently under-produce attenuation may still be viable once resolution or sub-grid shielding improves.","A statistical sample of dust masses in UV-selected z greater than 10 galaxies would cleanly discriminate between pure burstiness and models that retain dust while still producing UV variability.","The same cycle that erases dust after each burst may also explain why some high-redshift samples appear split into dust-rich and dust-poor populations without requiring two distinct formation channels."],"forward_implications":["Bursty star formation can explain the z greater than or equal to 10 UV-bright excess only if it rapidly settles by z approximately 8, where large dust reservoirs are already detected.","Observed IR-bright high-redshift galaxies are short-lived post-burst phases lasting only tens of Myr and are therefore biased relative to the bulk population.","Dust-poor post-starburst phases should preferentially allow high escape fractions of ionizing photons.","A mechanism that shields dust from strong feedback is required if the same galaxy-formation model is to match the higher observed UV attenuation at lower redshifts."],"fun_headline_variants":["Bursty SF stops large dust reservoirs lasting in early galaxies","Short IR phases of ~20 Myr leave early galaxies dust-poor most of the time","Thesan-Zoom: bursty stars wipe out lasting dust at high redshift","Dust survival fails under bursty SF unless shielding or early settling","High-z UV excess needs bursty SF to calm by z~8 for dust survival"],"cache_read_input_tokens":32896,"weakest_assumption_plain":"The assumption that the adopted sub-grid dust physics and feedback scheme correctly capture how dust survives, even though unresolved dense clumps that could shield dust are missing.","fun_headline_variants_meta":{"raw":{"variants":["Bursty SF stops large dust reservoirs lasting in early galaxies","Short IR phases of ~20 Myr leave early galaxies dust-poor most of the time","Thesan-Zoom: bursty stars wipe out lasting dust at high redshift","Dust survival fails under bursty SF unless shielding or early settling","High-z UV excess needs bursty SF to calm by z~8 for dust survival"]},"model":"grok-4.5","effort":"low","cost_usd":0.00318,"raw_usage":{"total_tokens":1262,"prompt_tokens":954,"num_sources_used":0,"completion_tokens":101,"cost_in_usd_ticks":31800000,"prompt_tokens_details":{"text_tokens":954,"audio_tokens":0,"image_tokens":0,"cached_tokens":384},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":207,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":954,"tokens_out":101,"duration_ms":2963,"temperature":1.0,"reasoning_tokens":207,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-13T06:25:02.633862+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Deep continuum or absorption measurements showing large dust reservoirs (M_dust/M_star greater than or equal to 10^{-3}) coexisting with high specific star-formation rates in a statistical sample of z greater than or equal to 10 UV-bright galaxies would falsify the claimed incompatibility.","supporting_citations":[],"review_version":1}