{"id":"c14c1bf7-956a-4e07-bd10-0f46b23f72a5","arxiv_id":"2606.04740","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"FIRE-2 simulations show per-galaxy tidal disruption rates peak near z=2.5 at 4e-4 per year, correlate with SFR and central density, and remain high in satellite galaxies at early times.","lead":"This paper uses FIRE-2 cosmological simulations to estimate how often stars are torn apart by black holes in galaxies from redshift 1 to 10. The results link disruption rates to galaxy star formation and density, offering predictions for future high-redshift observations.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.3","headline":"FIRE-2 resolution limits on central stellar densities at z>1 are the load-bearing assumption for the reported TDR evolution","rationale":"The reader's weakest assumption directly identifies the resolution/central-density step that controls the entire TDR(z) result. Because the manuscript was reviewed from the abstract alone, the UNVERDICTED verdict already reflects this uncertainty; the concrete test above would decide whether the assumption holds without requiring any change to the current verdict label.","tokens_in":1749,"tokens_out":432,"duration_ms":15024,"concrete_test":"Extract the central stellar density profiles (or the proxy used for TDR) from the 10 most massive FIRE-2 galaxies at z=2.5; recompute TDR after replacing those densities with values drawn from a higher-resolution (e.g., 10x particle mass) re-simulation or from observed nuclear profiles of z~2.5 galaxies; if the median TDR changes by >3x, the reported peak value and its redshift location are not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline TDR(z) curve (peak ~4e-4 yr^{-1} at z~2.5) is obtained by applying a TDR estimator to the simulated M_BH, M_gal, SFR, and central stellar density in each FIRE-2 zoom-in galaxy. TDR formulas are exponentially sensitive to the stellar density within the black-hole influence radius (typically <<100 pc). FIRE-2 zoom-ins have baryonic particle masses ~10^4-10^5 M_sun and softening lengths of tens of pc at z>1; they therefore cannot directly resolve the nuclear star cluster or loss-cone population. Any sub-grid extrapolation or scaling relation used to infer the unresolved density enters the final TDR at leading order. If that mapping is inaccurate by even a factor of a few, both the normalization and the redshift of the peak shift. The correlation with SFR and central density is tautological once the estimator is applied, so it does not independently validate the result.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper uses FIRE-2 cosmological zoom-in simulations to compute per-galaxy tidal disruption rates (TDR) from z=1 to 10 across IMBH to SMBH masses. It reports that the averaged TDR rises from early times, peaks at ~4×10^{-4} yr^{-1} near z~2.5, and falls to ~10^{-5} yr^{-1} at z=1. The TDR is stated to correlate strongly with host SFR and central stellar density at all redshifts; trends with M_BH and M_gal are qualitatively similar to the local universe; satellite galaxies contribute an increasing fraction of the TDR at high z.","tokens_in":1948,"tokens_out":639,"duration_ms":22060,"significance":"If robust, the work supplies the first simulation-derived TDR(z) evolution and links it to galaxy properties, offering testable predictions for high-redshift TDE searches. The approach of extracting rates directly from cosmological zoom-ins is novel for this problem. Credit is due for covering a wide redshift range and including satellites. However, the result's significance hinges on whether the unresolved central densities can be reliably mapped to TDR.","major_comments":[{"comment":"Abstract and methods (TDR extraction): the headline TDR(z) curve is obtained by applying an estimator to simulated M_BH, M_gal, SFR, and central stellar density. No derivation, functional form, or sub-grid extrapolation for the loss-cone density is provided; the exponential sensitivity of TDR to density within the BH influence radius means any inaccuracy in the unresolved nuclear density shifts both normalization and peak redshift by factors of several.","section":"Abstract / Methods"},{"comment":"Abstract and results (validation): no comparison of the computed local (z~0) TDR to observed rates is shown, nor are error bars or resolution convergence tests reported. Without this anchor, the claimed evolution from z=10 to z=1 cannot be assessed for systematic bias.","section":"Abstract / Results"},{"comment":"Abstract (correlations): the reported strong correlation between TDR and central stellar density is expected by construction once the estimator is applied to that density; it therefore does not constitute independent validation of the redshift trend or the underlying density field.","section":"Abstract"}],"minor_comments":[{"comment":"Abstract: include a one-sentence description of the TDR estimator and the mass range of black holes considered.","section":"Abstract"},{"comment":"Figure clarity: ensure any TDR(z) plots show individual galaxy tracks or scatter in addition to the average to allow assessment of sample variance.","section":"Figures"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a good fit for astro-ph.HE but the resolution concern is load-bearing; if the authors cannot demonstrate that the central-density mapping is stable under resolution changes or calibrated to local data, the central claim may not survive."},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for their detailed and constructive report. We address each major comment below and indicate where revisions will be made to strengthen the manuscript.","responses":[{"response":"We agree that the methods section would benefit from an explicit derivation of the TDR estimator, including the functional form relating loss-cone refilling to central stellar density. The estimator follows the standard loss-cone formalism (e.g., as in Merritt & Wang 2005 and subsequent works), applied to the resolved central densities in FIRE-2. We will add this derivation, the precise functional form, and a brief discussion of sub-grid assumptions in a revised Methods section. This addresses the concern about transparency while noting that the simulations themselves provide the density evolution.","revision_made":"yes","referee_comment":"[Abstract / Methods] Abstract and methods (TDR extraction): the headline TDR(z) curve is obtained by applying an estimator to simulated M_BH, M_gal, SFR, and central stellar density. No derivation, functional form, or sub-grid extrapolation for the loss-cone density is provided; the exponential sensitivity of TDR to density within the BH influence radius means any inaccuracy in the unresolved nuclear density shifts both normalization and peak redshift by factors of several."},{"response":"We acknowledge the value of anchoring the results to local observations. In the revised manuscript we will include a direct comparison of our z≈0 TDR values to the observed local TDE rate range (∼10^{-5}–10^{-4} yr^{-1} per galaxy) from the literature, along with sample variance error bars derived from the zoom-in suite. Resolution convergence is limited by the fixed FIRE-2 resolution; we will add a brief discussion of this limitation and note that higher-resolution follow-up simulations would be needed for full convergence tests.","revision_made":"partial","referee_comment":"[Abstract / Results] Abstract and results (validation): no comparison of the computed local (z~0) TDR to observed rates is shown, nor are error bars or resolution convergence tests reported. Without this anchor, the claimed evolution from z=10 to z=1 cannot be assessed for systematic bias."},{"response":"We agree that the TDR–central-density correlation is expected by construction of the estimator. However, the independent correlation with SFR (which is not an input to the estimator) and the persistence of M_BH and M_gal trends across redshift provide additional support for the physical trends. We will revise the abstract and results text to clarify this distinction and to emphasize that the redshift evolution arises from the simulated evolution of galaxy properties rather than from the estimator alone.","revision_made":"yes","referee_comment":"[Abstract] Abstract (correlations): the reported strong correlation between TDR and central stellar density is expected by construction once the estimator is applied to that density; it therefore does not constitute independent validation of the redshift trend or the underlying density field."}],"tokens_in":1511,"tokens_out":639,"duration_ms":17164,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The main thing here is a new calculation of average tidal disruption rates from z=1 to 10 pulled directly from FIRE-2 zoom-ins. They report a rise to a peak of roughly 4e-4 per year near z=2.5, then a drop, plus a growing role for satellites at earlier times. That redshift range and the satellite piece are not in the earlier local-universe work they cite.\n\nThey also show the rates track star-formation rate and central stellar density at every redshift, and that the M_BH and M_gal trends look similar to what is seen nearby. Running the same estimator across the simulation snapshots is a straightforward way to get a first cosmic-evolution curve.\n\nThe soft spot is the resolution. FIRE-2 baryonic particles are 10^4-10^5 solar masses with softening lengths of tens of parsecs at z>1, so the nuclear densities inside the black-hole influence radius are not resolved. TDR formulas are exponentially sensitive to those densities, which means any sub-grid scaling or extrapolation they apply sets both the normalization and the location of the peak. The abstract gives no derivation, no local-universe validation, and no error bars, so it is hard to judge how much the result moves if the density mapping shifts by a factor of a few. The reported correlation with central density is expected once the estimator is applied, so it does not test the underlying assumption.\n\nThis is for people who need simulation priors on high-redshift transients or black-hole demographics. The approach is worth referee time to check the sub-grid step and see whether the full methods section supplies the missing validation.","headline":"FIRE-2 supplies the first simulation-based TDR(z) to z=10 but the central densities sit below the resolution limit, so the numbers rest on an untested extrapolation.","tokens_in":2425,"tokens_out":418,"would_cite":false,"duration_ms":20685,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"Simulations find the average tidal disruption rate per galaxy peaks near redshift 2.5 then falls sharply by redshift 1.","keywords":["tidal disruption events","cosmological simulations","black hole growth","galaxy evolution","star formation rate","central stellar density","satellite galaxies","redshift evolution"],"falsifier":"A direct measurement or statistical sample of tidal disruption events at redshift approximately 2.5 that yields an average per-galaxy rate far below or far above 4 times 10 to the minus 4 per year.","tokens_in":2664,"feed_emoji":"🌌","tokens_out":798,"duration_ms":19649,"temperature":0.7,"pith_summary":"This paper applies cosmological zoom-in simulations to track how often stars are torn apart by black holes inside galaxies from redshift 10 down to redshift 1. The calculation shows the typical rate per galaxy climbs from early times, reaches roughly 4 times 10 to the minus 4 events per year near redshift 2.5, and drops to about 10 to the minus 5 by redshift 1. The rates track closely with each galaxy's star-formation activity and the density of stars near its center at every epoch examined. The same mass trends seen locally also appear at high redshift, while satellite galaxies supply a rising fraction of the events as redshift increases.","feed_headline":"TDE rate per galaxy peaks near z=2.5 at 4e-4 per year","feed_subtitle":"FIRE-2 runs show the average disruption rate climbs then drops by more than an order of magnitude while tracking star formation and central","key_machinery":"FIRE-2 cosmological zoom-in simulations that model central stellar densities, black-hole populations, and dynamical conditions to compute per-galaxy tidal disruption rates across redshift.","core_discovery":"Using the FIRE-2 cosmological zoom-in simulations, the per-galaxy tidal disruption rate is computed over redshifts 1 to 10 for black holes ranging from intermediate-mass to supermassive. The averaged rate rises from the early universe, peaks at approximately 4 times 10 to the minus 4 per year near redshift 2.5, and declines to about 10 to the minus 5 per year at redshift 1. This rate correlates strongly with host-galaxy star-formation rate and central stellar density at all redshifts. The dependence on black-hole mass and galaxy mass remains qualitatively similar from high redshift to the local universe, and satellite galaxies show comparably high rates whose fractional contribution grows at","pith_inferences":["Confirmation of the redshift-2.5 peak would link the era of maximum star formation directly to the era of maximum central stellar densities that drive disruptions.","Higher rates at earlier times could alter estimates of how much black-hole growth occurs through stellar capture rather than gas accretion.","If satellite contributions are as large as modeled, wide-field high-redshift transient searches could preferentially detect events in merging or assembling galaxies."],"forward_implications":["Tidal disruption rates track star-formation rate and central density, so galaxies with elevated star formation should produce more events at any redshift.","Satellite galaxies maintain high rates whose share increases at high redshift, making them useful targets for finding intermediate-mass black holes.","The black-hole to galaxy mass trends stay consistent from high redshift to today, supporting similar scaling relations across cosmic time.","Cosmological simulations can now supply predictions for the cosmic evolution of tidal disruption rates that future surveys can test directly."],"fun_headline_variants":["TDE rate peaks at 4e-4 yr-1 near z=2.5 per galaxy","Per-galaxy TDR evolves peaking at z=2.5 in simulations","TDR correlates with SFR and galaxy density at all z","Satellites show rising TDR contribution at high redshifts"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The FIRE-2 simulations accurately capture the central stellar densities, black-hole populations, and dynamical conditions needed to compute tidal disruption rates at redshifts above 1.","fun_headline_variants_meta":{"raw":{"variants":["TDE rate peaks at 4e-4 yr-1 near z=2.5 per galaxy","Per-galaxy TDR evolves peaking at z=2.5 in simulations","TDR correlates with SFR and galaxy density at all z","Satellites show rising TDR contribution at high redshifts"]},"model":"grok-4.3","cost_usd":0.008758,"raw_usage":{"total_tokens":3910,"prompt_tokens":760,"num_sources_used":0,"completion_tokens":79,"cost_in_usd_ticks":87578000,"prompt_tokens_details":{"text_tokens":760,"audio_tokens":0,"image_tokens":0,"cached_tokens":64},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":3071,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":760,"tokens_out":79,"duration_ms":24943,"temperature":1.0,"reasoning_tokens":3071,"cache_read_input_tokens":64,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-28T05:25:49.901988+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A direct measurement or statistical sample of tidal disruption events at redshift approximately 2.5 that yields an average per-galaxy rate far below or far above 4 times 10 to the minus 4 per year.","supporting_citations":[],"review_version":1}