{"id":"55d07cd0-2afc-48fe-9bdf-120099884731","arxiv_id":"2501.18687","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"In TNG50, the resolved star-forming main sequence has slope 0.30, much shallower than observed, while the resolved mass-metallicity relation matches data; both can be described by a leaky-box model that prefers low net outflows.","lead":"This paper maps star formation, gas metallicity, and stellar mass onto 1 kpc squares in the TNG50 cosmological simulation and compares the resulting resolved scaling relations to observations. It finds that TNG50 reproduces the resolved mass-metallicity relation well but predicts a too-shallow resolved star-forming main sequence, likely because its AGN feedback is too strong, and it explains both relations with an extended leaky-box toy model.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Predicted rSFMS slope from SK + gas relation is 0.20 vs measured 0.30; the unresolved 33% gap and unmodeled radial SK variation undermine the central origin claim.","rationale":"The paper's core empirical contributions—the TNG50 rSFMS/rMZR measurements, scatter decomposition, host-mass dependence, and the rFMR null result—are solid and reproducible, with public data and scripts. The most load-bearing weakness is not the leaky-box inflow assumption (the reader's weakest_assumption, Eq. 11), which concerns a clearly secondary 'Finally' section, but the paper's central explanation of the rSFMS origin. The derivation in Section 4.1 (Equations 5-7) predicts alpha = nk ~ 0.20, yet the measured slope is 0.302, a 33% shortfall the paper itself concedes without closure. The suggested molecular-gas explanation is plausible but untested, and Figure 10 shows the SK normalization varies with galactocentric radius, meaning the simple power-law product omits a degree of freedom that Section 4.1.1 later uses to explain host-mass dependence. The AGN-feedback argument for the observed-vs-simulated discrepancy cannot rescue this internal inconsistency: low-mass hosts (little AGN) have steeper slopes (~0.67), even farther from 0.20. A quantitative test using molecular-gas tracers could settle whether the missing factor is the gas phase. Since the measurements stand but the central theoretical claim is not fully demonstrated, the reader's CONDITIONAL verdict is appropriate; no adjustment is needed.","tokens_in":32369,"tokens_out":6433,"duration_ms":62626,"concrete_test":"Use the Diemer et al. (2018) molecular-gas post-processing on the same TNG50 spaxels to construct Sigma_H2 maps, fit the H2-based SK relation (Sigma_SFR vs Sigma_H2) and the Sigma_H2-Sigma_* relation, and predict alpha = n_H2 * k_H2. If the predicted slope matches the measured 0.30 (and the low-mass-host value ~0.67), the proposed molecular-gas resolution holds; if not, the rSFMS origin claim should be revised from quantitative to qualitative.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4.1 combines Equations 5 and 6 to argue that the rSFMS follows from the Schmidt-Kennicutt relation and the local gas-stellar mass relation, predicting alpha = n*k = 0.20. The directly measured rSFMS slope is 0.302 (Section 3.1), a 33% shortfall that the paper itself concedes and does not quantitatively close. The paper suggests molecular gas as a possible resolution, but this is not tested. Moreover, Figure 10 shows that the SK normalization depends on galactocentric radius, so the single-power-law product in Equation 7 omits a radial term that Section 4.1.1 later invokes to explain host-mass dependence. This internal mismatch cannot be attributed to the AGN-feedback explanation for the observed-vs-simulated discrepancy, because low-mass hosts, where AGN feedback is weak, have rSFMS slopes (~0.67) even further from the predicted 0.20. Thus the central assertion that the rSFMS arises as an indirect combination of these two relations is quantitatively unsupported.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript uses TNG50-1 to construct 1 kpc spaxel maps for 2,734 galaxies with Mstar > 1e9 Msun at z=0, and measures the resolved star-forming main sequence (rSFMS), the resolved mass-metallicity relation (rMZR), their host-mass dependence, and the resolved Schmidt-Kennicutt (SK) and gas-mass relations. It reports an rSFMS slope alpha=0.302, compares it with CALIFA, MaNGA, ALMaQUEST, PHANGS, and EAGLE, and argues that the rSFMS is an indirect consequence of the SK law combined with the local gas-stellar mass relation (predicted alpha=nk=0.20). The paper also extends the Zhu et al. (2017) resolved leaky-box model with explicit inflow and outflow terms, and interprets the TNG scaling relations as preferring a low net outflow rate.","tokens_in":32593,"tokens_out":6430,"duration_ms":60046,"significance":"The paper provides a useful TNG50 benchmark for kpc-scale scaling relations, with careful spaxel construction and explicit robustness tests in Appendices A and B, and it makes the code and data products publicly available. Its transparency about the prescribed nature of the SK law in TNG is a strength. The rMZR comparison and the host-mass decomposition of the rSFMS are valuable. If the quantitative gap in the origin argument and the ad hoc sub-kpc inflow assumption were addressed, the proposed interpretive framework would be a meaningful step toward connecting resolved observations with simulation subgrid physics.","major_comments":[{"comment":"The claimed origin of the rSFMS is only qualitative: combining the fitted SK law (epsilon=5.89e-14, k=1.52) with the gas-mass relation n=0.137 gives alpha=nk=0.20, while the rSFMS slope measured in Section 3.1 is alpha=0.302. The 33% shortfall is acknowledged but not closed, and the suggested molecular-gas explanation is not tested. Because this is the quantitative basis for conclusion (iii), the central claim should either be reframed as a qualitative tendency or be supported by a molecular-gas or radial-inclusive calculation.","section":"Section 4.1, Eq. (7)"},{"comment":"Equation (7) assumes a single-power-law SK normalization, but Figure 10 shows that the SK normalization depends on galactocentric radius, and Section 4.1.1 invokes exactly this radial dependence to explain the host-mass dependence of the rSFMS. Omitting the radial term from the derivation while making it central to the mass-dependence argument is internally inconsistent: the same mechanism cannot be negligible in the origin claim and dominant in the mass-dependence explanation. A derivation that includes the radial dependence of epsilon, or an explicit justification for averaging it out, is needed.","section":"Section 4.1.1, Figure 10"},{"comment":"The host-mass-resolved slopes in Table 1 range from alpha=0.671 in the lowest-mass bin to alpha=0.25-0.33 at high mass. The low-mass slope is more than three times the predicted 0.20 from Section 4.1, and this is the regime where AGN feedback in TNG is weakest. The attribution of the simulated-versus-observed slope difference to over-strong AGN feedback in massive hosts (Section 3.1.1) therefore does not address the failure of the origin model at low host masses.","section":"Section 3.1.1, Table 1"},{"comment":"The leaky-box model's conclusions about net outflow rates depend on the assumption dSigma_gas,in/dt = eta_in Sigma_SFR at 1 kpc scales, which the paper itself labels as '(crude)'. Equations (12)-(14) and the Section 4.2.2 preference for a low net outflow rate all follow from this proportionality. The assumption should be checked directly in TNG, for example by measuring gas fluxes across spaxel boundaries; without such a check, the outflow/inflow conclusion is not established.","section":"Section 4.2, Eq. (11)"},{"comment":"The conclusion that the rMZR is insensitive to eta_out (Section 4.2.2) follows from the assumption in Equation (20) that outflowing gas has exactly the spaxel's current metallicity. This assumption is not checked against TNG, where wind metal loading can differ from the local ISM value. If outflows are preferentially metal-enriched or metal-poor, the rMZR would depend on eta_out and the contrast with 'net outflow' models would weaken.","section":"Section 4.2, Eqs. (20)-(22)"}],"minor_comments":[{"comment":"The text refers to the 'Kennicut-Schmidt relation'; this should be 'Kennicutt-Schmidt relation'.","section":"Section 3, introductory paragraph"},{"comment":"The PHANGS comparison is described as a 'solid purple line' in the text and as a 'solid black line' in the caption; please make the line descriptors consistent.","section":"Figure 2 caption"},{"comment":"The text says Figures 14 and 15 show predictions for the rSFMS and rMZR, but Figure 14 shows the gas-mass main sequence; please correct the figure reference.","section":"Section 4.2.2"},{"comment":"The word 'IllutrisTNG' is a typo and should read 'IllustrisTNG'.","section":"Section 5, item (ii)"},{"comment":"The 'five representative spaxels' are mentioned repeatedly but their initial Sigma_gas values are not defined in the text or figures; please identify these tracks explicitly.","section":"Section 4.2.1, Figures 12-13"}],"recommendation":"major_revision","confidential_remarks":"The measurements and robustness tests are solid and the authors are appropriately transparent about limitations. The requested revision is centered on the quantitative support for the origin claim (the 0.20 vs 0.302 slope gap) and on testing the inflow assumption that anchors the leaky-box conclusions. These are fixable within the manuscript's scope, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid, useful paper for anyone working on resolved scaling relations or using TNG as a benchmark. The measurements of the rSFMS and rMZR at 1 kpc in TNG50 are carefully done, robust to kernel choice and pixel cuts, and the rMZR agreement with IFS surveys is genuinely good. The leaky-box extension with separate inflow and outflow terms is a nice addition, and the authors are honest about the crude assumption behind it. The public GitHub scripts are a plus.\n\nThe main soft spot is the derived slope. Combining the SK law and gas-stellar mass relation predicts alpha = 0.20, about 33% below the measured 0.30. The paper concedes this and suggests molecular gas as a possible resolution, but does not test it. The radial dependence of the SK normalization (Fig. 10) means the single power-law product in Eq. 7 is incomplete; the host-mass dependence analysis in Section 4.1.1 actually leans on that radial effect, so the origin argument is more qualitative than the text sometimes implies. The stress-test note is right that low-mass hosts have slopes near 0.67, even further from 0.20; the radial argument might rescue that, but that same radial argument means Eq. 7 is missing a term.\n\nThe leaky-box model has its own caveats. Equation 11 assumes inflow proportional to SFR on sub-galactic scales, which is crude. The 'low net outflow' conclusion is calibrated against the same TNG data from which the model parameters are drawn, so it is a description rather than an independent prediction. The authors acknowledge much of this, which I credit.\n\nNone of this sinks the paper. The measurements are the core value and will be a useful reference for simulation-observation comparisons. The rMZR agreement and the host-mass decomposition are worth having. A referee should push for a quantitative treatment of the slope gap or a clearer statement that the combination is not a full derivation.\n\nWho this is for: anyone comparing simulations to IFU surveys, or building resolved gas-regulator models. It deserves a serious referee; I'd send it out, likely with a request for a major revision focused on the slope discrepancy.","headline":"Useful TNG50 measurements with an honest but unclosed 33% gap in the origin-of-rSFMS derivation.","tokens_in":33199,"tokens_out":2878,"would_cite":true,"duration_ms":26224,"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":"The paper argues that TNG50's resolved star-forming main sequence is a by-product of the Schmidt-Kennicutt law and a local gas-to-stellar mass relation, with its shallow slope coming from AGN feedback in massive hosts.","keywords":["resolved star-forming main sequence","resolved mass-metallicity relation","IllustrisTNG","Schmidt-Kennicutt law","leaky-box model","AGN feedback","integral field spectroscopy","galaxy evolution"],"falsifier":"Measure the resolved star-forming main sequence in galaxies with $M_\\star<10^{10}\\,M_\\odot$ using an integral-field survey; if the slope is near $0.3$ rather than the $\\sim0.66$ that TNG50 produces for such hosts, the claim that AGN feedback causes the shallow stacked slope fails.","tokens_in":32105,"feed_emoji":"🌌","tokens_out":12099,"duration_ms":104407,"temperature":0.7,"pith_summary":"This paper asks whether a cosmological galaxy-formation simulation reproduces the kpc-scale relations that integral-field surveys now measure, and what the answer says about the physics. It shows that TNG50's resolved star-forming main sequence (the power-law relation between star-formation rate surface density and stellar mass surface density on 1-kpc patches) is too shallow compared with observed galaxies, and it argues the shortfall is not missing star-formation physics but AGN feedback acting too strongly in massive hosts. The same spaxel-by-spaxel analysis finds that the resolved mass-metallicity relation matches the observed shape and that the local star-formation rate does not drive its scatter; host stellar mass does. The paper then derives the rSFMS as a composition of the resolved Schmidt-Kennicutt law and a local gas-to-stellar mass relation, and shows that a generalized leaky-box model reproduces the simulated gas and metallicity relations best when the net outflow rate is low. This matters because it turns resolved scaling relations into concrete tests of feedback implementations rather than just descriptions of the interstellar medium.","feed_headline":"TNG50's too-shallow star-forming main sequence traced to AGN","feed_subtitle":"The simulation reproduces kpc-scale metal patterns but over-quenches massive galaxies, pointing to AGN feedback strength.","key_machinery":"The central machinery is a sample of 1-kpc $\\times$ 1-kpc spaxel maps built from TNG50 galaxies with $M_\\star>10^9\\,M_\\odot$, treated as individual parcels that can gain and lose gas. Three elements carry the argument: the Springel & Hernquist (2003) volumetric star-formation prescription, which makes the resolved Schmidt-Kennicutt relation a written-in rule rather than an emergent result; the composition identity $\\Sigma_{\\rm SFR}\\propto\\Sigma_{\\rm gas}^k\\propto\\Sigma_\\star^{nk}$, which converts the two host-independent gas relations into the rSFMS; and the generalized leaky-box equations (8), (13), and (22), which evolve stellar, gas, and metal surface densities with explicit inflow and outflow mass-loading factors $\\eta_{\\rm in}$ and $\\eta_{\\rm out}$.","core_discovery":"On 1-kpc spaxels at $z=0$, TNG50 produces a resolved star-forming main sequence $\\Sigma_{\\rm SFR}\\propto\\Sigma_\\star^\\alpha$ with a stacked slope $\\alpha=0.30$, shallower than the $0.7$ to $1.1$ slopes reported by integral-field surveys. The paper's central claim is that this relation is not independently fundamental: it is the composition of the resolved Schmidt-Kennicutt law $\\Sigma_{\\rm SFR}\\propto\\Sigma_{\\rm gas}^k$ (best-fit $k=1.52$) and the local gas-stellar mass relation $\\Sigma_{\\rm gas}\\propto\\Sigma_\\star^n$ (best-fit $n=0.137$), which together predict $\\alpha=nk\\simeq0.20$. The remaining difference from the measured slope, and the flattening of the rSFMS with host mass, is attributed to AGN feedback: per-galaxy slopes fall from $0.66$ in the lowest-mass hosts to $0.30$ or inverted in the highest-mass hosts, and the high-mass systems dominate the stacked sample. The resolved mass-metallicity relation, by contrast, matches the observed shape, with its scatter governed by host stellar mass; and a generalized leaky-box model with explicit inflow and outflow terms describes the simulated gas and metallicity relations, preferring a near-zero net outflow rate.","pith_inferences":["If the mass dependence is real, integral-field surveys that deliberately include low-mass star-forming galaxies should measure a steeper rSFMS than surveys dominated by $M_\\star>10^{10.5}\\,M_\\odot$ systems; the mass-slope trend itself could become a calibrator for AGN feedback strength.","The compositional derivation is quantitatively incomplete (predicted $\\alpha=0.20$ versus measured $0.30$); redoing the same derivation with a molecular-gas version of the local gas-stellar relation could close the gap in TNG50, just as molecular gas does for the observed rSFMS.","The paper's leaky-box conclusions rest on the proportionality between inflow and SFR at kpc scales; measuring the actual gas flux across spaxel boundaries in TNG50 would test Equation 11 directly and would either validate or revise the low-net-outflow preference.","The absence of a resolved fundamental metallicity relation in TNG50, in tension with some observational analyses, may reflect AGN-driven central enrichment; forward-modeling synthetic integral-field observations from the simulation could determine whether the discrepancy is physical or diagnostic."],"forward_implications":["The rSFMS slope in TNG50 is not set by the star-formation law alone; it is the product of the Schmidt-Kennicutt index and the local gas-to-stellar mass index, so matching observed slopes means getting both gas relations right.","Because low-mass TNG50 hosts already produce per-galaxy slopes near the observed range, the simulation's overall mismatch is concentrated above roughly $10^{10.5}\\,M_\\odot$, identifying AGN feedback as the physics to adjust.","The good resolved mass-metallicity agreement indicates that metal production, return, and redistribution in TNG50 are consistent with integral-field observations even where star-formation feedback appears too aggressive.","In the leaky-box model, the gas-stellar mass relation is controlled by the net outflow rate while the metallicity relation is controlled mainly by inflow dilution, so TNG50's resolved relations imply a low net outflow rate.","Resolved scaling relations therefore provide a comparative test of feedback implementations: a simulation tuned to reproduce global galaxy properties can still fail or succeed on 1-kpc scales, and these relations show where to look."],"supporting_citations":[{"why":"Defines the resolved leaky-box model that this paper generalizes to include explicit inflow and outflow terms.","marker":"Zhu et al. (2017)"},{"why":"Provides the effective equation of state and volumetric Schmidt-Kennicutt prescription that makes the simulated resolved SK relation non-emergent.","marker":"Springel & Hernquist (2003)"},{"why":"Sets the canonical form of the Schmidt-Kennicutt relation used in Equation 5 and as a comparison for its slope.","marker":"Kennicutt (1998)"},{"why":"Describes the TNG AGN feedback model that the paper argues over-suppresses star formation in massive hosts.","marker":"Weinberger et al. (2017)"},{"why":"Provides the EAGLE comparison and the method of mapping resolved relations from simulations onto observed spaxel scales.","marker":"Trayford & Schaye (2019)"},{"why":"Provides a CALIFA observed rSFMS baseline that is steeper than TNG's stacked slope.","marker":"Cano-Díaz et al. (2016)"},{"why":"Provides a MaNGA observed rSFMS baseline and sensitivity limits used in spaxel selection.","marker":"Hsieh et al. (2017)"},{"why":"Provides the PHANGS observed rSFMS baseline and reinforces the slope discrepancy.","marker":"Pessa et al. (2021)"},{"why":"Provides the MaNGA observed rMZR that TNG's shape is compared against.","marker":"Barrera-Ballesteros et al. (2016)"},{"why":"Supplies TNG gas-phase decomposition that explains why the molecular-gas SK index differs from the total-gas index.","marker":"Diemer et al. (2018)"}],"fun_headline_variants":["TNG50's flat star-forming sequence linked to AGN feedback","AGN overkill flattens TNG50's resolved star-forming main sequence","TNG50 runs shallow on star-forming slopes, AGN blamed","Why TNG50's kpc-scale star formation is too shallow: AGN","TNG50's SFMS slope too low; AGN feedback the culprit"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole leaky-box argument rests on the assumption that the gas inflow rate into a 1-kpc patch is proportional to that patch's star formation rate, an assumption the paper itself calls crude and untested at sub-galactic scales.","fun_headline_variants_meta":{"raw":{"variants":["TNG50's flat star-forming sequence linked to AGN feedback","AGN overkill flattens TNG50's resolved star-forming main sequence","TNG50 runs shallow on star-forming slopes, AGN blamed","Why TNG50's kpc-scale star formation is too shallow: AGN","TNG50's SFMS slope too low; AGN feedback the culprit"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00146,"raw_usage":{"total_tokens":5996,"prompt_tokens":1190,"completion_tokens":4806,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":806,"completion_tokens_details":{"reasoning_tokens":4707}},"tokens_in":806,"tokens_out":4806,"duration_ms":30952,"temperature":1.0,"reasoning_tokens":4707,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T22:50:20.843526+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the resolved star-forming main sequence in galaxies with $M_\\star<10^{10}\\,M_\\odot$ using an integral-field survey; if the slope is near $0.3$ rather than the $\\sim0.66$ that TNG50 produces for such hosts, the claim that AGN feedback causes the shallow stacked slope fails.","supporting_citations":[{"cited_title":"B., Barrera-Ballesteros, J","cited_arxiv_id":null,"evidence_quote":"Defines the resolved leaky-box model that this paper generalizes to include explicit inflow and outflow terms."},{"cited_title":"W., & Schaye, J","cited_arxiv_id":null,"evidence_quote":"Provides the EAGLE comparison and the method of mapping resolved relations from simulations onto observed spaxel scales."},{"cited_title":"C., Lin, L., Lin, J","cited_arxiv_id":null,"evidence_quote":"Provides a MaNGA observed rSFMS baseline and sensitivity limits used in spaxel selection."}],"review_version":1}