{"id":"a019cea5-c0f0-4e77-bcf8-16229c3fdd28","arxiv_id":"2412.06594","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Using stellar abundances of C, O, Mg, Si, and Fe, the authors find that super-Jupiter hosts have at least as much disk metal content as Jupiter hosts, but the difference is not robust to propagated uncertainties.","lead":"This paper compares the metal content of protoplanetary disks around stars hosting Jupiter-mass planets (1 to 4 Jupiter masses) and super-Jupiters (4 to 13 Jupiter masses), finding the super-Jupiter hosts have equal or slightly higher metal content. The authors interpret this as evidence that super-Jupiters can form by core accretion, but the statistical support weakens once measurement uncertainties are propagated.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The abstract's 'strong evidence' for core accretion is contradicted by the authors' own Monte Carlo propagation, in which fewer than 50% of 10,000 uncertainty realizations yield p<0.05 for the Jupiter versus super-Jupiter metal-content comparison.","rationale":"The paper compiles a useful sample and the direction of the effect is stable, but its statistical analysis does not support the strength of the claim. The reader identified the present-day versus primordial abundance assumption as the weakest link; I agree that is an external-validity concern, especially because it leans on an unpublished self-cited paper. However, the more decisive problem is internal: the authors' own Monte Carlo test in Section 4 shows that the apparent difference between Jupiters and super-Jupiters is not robust to measurement uncertainties. This is a correctness problem, not just an interpretation problem, because it directly contradicts the abstract's 'consistently demonstrate' and 'strong evidence.' Section 5 even concedes that some statistical tests did not support the hypothesis of distinct distributions, yet the conclusion still states the results consistently demonstrate the claim. I would keep the rejection, but re-ground it in the Monte Carlo result rather than the abundance assumption. A bootstrap or posterior analysis on the same sample would settle whether the directional difference is real or largely an artifact of ignoring uncertainties.","tokens_in":10915,"tokens_out":4846,"duration_ms":51195,"concrete_test":"Re-run the Jupiter/super-Jupiter comparison as a bootstrap on the published Ztotal values (10,000 resamples, split at 4 Mjup) and report the 95% confidence interval for the mean difference as well as the fraction of resamples in which the super-Jupiter mean exceeds the Jupiter mean. If the confidence interval includes zero or the exceedance fraction is below about 0.95, the abstract's 'consistently demonstrate' and 'strong evidence' are not supported; if the interval is strictly positive, the directional claim survives and the remaining issue is overstatement rather than an invalid conclusion.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim depends on the assertion that super-Jupiters form in disks with at least as much metal content as Jupiters, often slightly higher, and that this is 'strong evidence' for core accretion. Section 4 reports raw t-test p=0.023 and KS/AD p-values of about 0.03-0.04 for Ztotal, but the subsequent Monte Carlo propagation of uncertainties in Z, stellar mass, and planetary mass gives p<0.05 in fewer than 50% of 10,000 realizations. The paper itself states that this test does not support the hypothesis that the two samples have statistically distinct distributions or mean values. Despite this, Section 5 repeats that the results 'consistently demonstrate' the conclusion. The directional statement that super-Jupiters have at least as much metal content is robust, since only 1 in 10,000 realizations gave a lower mean for super-Jupiters, but a non-significant difference cannot justify 'strong evidence.' This internal contradiction is independent of the stellar-abundance assumption and is the primary reason the headline conclusion overreaches the data.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper compares the total metal content available in protoplanetary disks around stars hosting 'Jupiters' (1–4 Mjup) and 'super-Jupiters' (4–13 Mjup). Using C, O, Mg, Si, and Fe abundances from the Hypatia Catalog and the stoichiometric model of Santos et al. (2017a), the authors compute a metal mass fraction Z and a total metal content Ztotal under the assumption that disk mass scales linearly with stellar mass. They report raw t-test, KS, and AD p-values of 0.023–0.042 suggesting a slightly higher Z and Ztotal for super-Jupiters, but their own Monte Carlo propagation shows that these differences are not robust: p<0.05 occurs in fewer than 50% of realizations for Ztotal and in only about 6% for Z*total. The paper nevertheless concludes in the abstract and discussion that the results are 'strong evidence' that super-Jupiters form via core accretion.","tokens_in":11306,"tokens_out":6930,"duration_ms":70410,"significance":"If the directional result were robust, it would be a useful observational constraint: the metal budget in protoplanetary disks would not appear to be the reason super-Jupiters are rare, which would mildly favor core accretion over gravitational instability. The paper's strength is that it moves beyond [Fe/H] alone by including rock-forming elements and it explicitly reports the unfavorable Monte Carlo result. However, because the headline significance collapses under the paper's own error propagation, the central claim as stated is not supported. The more defensible directional statement (super-Jupiters do not have less disk metal than Jupiters) could still be of interest to planet formation modelers.","major_comments":[{"comment":"The headline claim is contradicted by the paper's own Monte Carlo test in Section 4. The raw t-test/KS/AD p-values (0.023–0.042) are reported as statistically significant, but the MC propagation over 10,000 realizations yields p<0.05 in fewer than 50% of cases for Ztotal and in only about 6% of cases for Z*total (over 94% with p>0.05). The text explicitly states that 'this test does not support the hypothesis that the two samples have statistically distinct distributions or mean values of Ztotal.' Nevertheless, the abstract and Section 5 describe the results as 'strong evidence' and 'consistently demonstrate' that super-Jupiters form in disks with at least as much metal content. Because the significance disappears once measurement uncertainties are propagated, the 'strong evidence' claim is not supported. The directional statement is more defensible (only 1 in 10,000 realizations gave a lower mean for super-Jupiters), but the paper must either provide a statistical argument for why the raw p-values are the relevant ones or substantially weaken the conclusion.","section":"Abstract; Section 4; Section 5"},{"comment":"The computation of Z and Ztotal relies on the assumption that present-day stellar atmospheric abundances reflect the primordial composition of the protoplanetary disk. The paper states that 'the abundance ratios of rock-forming elements remain relatively constant' over stellar evolution and supports this with Adibekyan et al. (2024, A&A, submitted), a self-cited manuscript that is not yet published and not available for verification. Since this assumption is load-bearing for the central claim, the authors should either provide the supporting analysis in the present paper or cite a published, independent source; at minimum, the results should be flagged as conditional on this assumption.","section":"Section 3"},{"comment":"The definition of Z*total is unclear and appears internally inconsistent. The text defines Z*total = log(Z × Mstar) × σdisk, where σdisk is a dispersion assumed to be 0.8 dex, but then states that 'σdisk was sampled from a log-normal distribution centered at 1 with a standard deviation of 0.8 dex.' Multiplying a log quantity by a dispersion does not generate scatter; it rescales the values. Please provide the exact error model used, justify the factor of 0.8, and ensure the text matches the analysis. This matters because Figure 3 and the corresponding MC results for Z*total are used to argue that the two samples have similar distributions.","section":"Section 3, Eq. (1)"}],"minor_comments":[{"comment":"The paragraph beginning 'The observational evidence that Jupiter-mass planets tend to form around stars with high metallicity...' is duplicated verbatim and should be removed in one place.","section":"Section 5"},{"comment":"The 'proto-solar disk level' is used in the text and in Figure 2 but is never quantitatively defined; please state explicitly that it corresponds to Z = 1.25% and Ztotal = 1.25 under the adopted solar reference.","section":"Section 3"},{"comment":"The paper reports 428 massive planets orbiting 396 stars after the SWEET-Cat cross-match and 265 planets in the final sample, but it does not state how many targets remained after the Hypatia cross-match; please provide the intermediate sample size.","section":"Section 2"},{"comment":"The statement that excluding the 10 systems with planets in both the Jupiter and super-Jupiter groups 'does not affect the outcomes of the tests' would be more convincing if the resulting p-values were reported.","section":"Section 4"},{"comment":"The [X/H] < 0.6 cutoff is justified only as excluding 'potentially inaccurate abundances' and removes four targets; please add a sensitivity test showing that the conclusions are unchanged if these targets are retained or if a different threshold is used.","section":"Section 2"}],"recommendation":"major_revision","confidential_remarks":"The paper's main conclusion is overstated, but the underlying comparison and the honest reporting of the negative Monte Carlo result make the manuscript salvageable. I recommend major revision rather than rejection. The editor may wish to emphasize that a non-significant difference cannot be described as 'strong evidence,' and that the submitted self-citation for the constancy of rock-forming element ratios should be replaced by a published reference or an inline analysis."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The useful thing here is the negative result: across 10,000 uncertainty realizations, super-Jupiters almost never have lower total metal content than Jupiters. That is a genuine, robust finding, and it does push back on earlier claims that super-Jupiters prefer lower-metallicity environments. The larger sample and the move from [Fe/H] to Z (using C, O, Mg, Si, Fe) are also worthwhile extensions. Give the authors credit for doing the Monte Carlo propagation in the first place, and for reporting the unflattering result in Section 4.\n\nThe problem is the gap between that result and the abstract. The raw p-values (0.023–0.042) are marginal, and the paper's own MC test finds p<0.05 in fewer than half of realizations for Ztotal and essentially never for Z*_total. Section 4 even says, plainly, that the test 'does not support the hypothesis that the two samples have statistically distinct distributions or mean values.' Yet the abstract and Discussion call this 'strong evidence' for core accretion. That is an internal contradiction, and it is the main reason the paper cannot be published as is.\n\nA second soft spot: the assumption that present-day stellar abundances reflect primordial disk composition, and that rock-forming element ratios stay constant over stellar evolution, is supported only by a self-cited, submitted manuscript. It may be true, but it is not yet publicly verifiable. That is a minor-to-moderate weakness, not a fatal one.\n\nThere are also small editorial issues: the same paragraph appears twice in the Discussion, and the paper could be tighter.\n\nWho is this for? People working on the planet–metallicity correlation and formation mechanisms, and anyone teaching how to interpret marginal p-values. The dataset and the Z computations are real and reproducible, and the directional claim is worth testing with future samples. But the headline conclusion needs to be reframed: the paper shows that super-Jupiters are not metal-poor relative to Jupiters, which weakens the case for GI-driven formation, rather than proving that core accretion produces them.\n\nI would send it to peer review, but only with a request for major revision. The authors should remove 'strong evidence', limit the conclusion to the robust directional statement, and either support the abundance-ratio assumption with published work or add sensitivity tests. After that, it could be a solid contribution to the observational literature.","headline":"The paper's useful result is the robust directional claim that super-Jupiters do not form in metal-poor disks; the 'strong evidence' for core accretion is not supported by the paper's own Monte Carlo tests.","tokens_in":11696,"tokens_out":1924,"would_cite":false,"duration_ms":22155,"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":"Super-Jupiters can form by core accretion: their host disks carry at least as much total metal as Jupiter-forming disks.","keywords":["planet formation","core accretion","gravitational instability","super-Jupiters","host star metallicity","chemical abundances","disk metal content","exoplanet formation"],"falsifier":"Rebuild the analysis using only abundances from one homogeneous high-resolution survey for every host star and compare the mean Ztotal of the 4–13 Mjup group with the 1–4 Mjup group; if it is then lower, the paper's statistical claim is overturned. Separately, measure C, O, Mg, Si, and Fe in stars of fixed initial composition but different evolutionary states (for example, open-cluster stars of different masses); if the rock-forming element ratios change with evolution, the present-day-abundance premise fails.","tokens_in":10717,"feed_emoji":"🪐","tokens_out":11446,"duration_ms":108290,"temperature":0.7,"pith_summary":"This paper asks whether super-Jupiters, planets roughly 4–13 times Jupiter's mass, can be built by the same core-accretion process that makes ordinary Jupiters, or whether their formation demands gravitational instability. The authors assemble 172 Jupiter-mass planets and 93 super-Jupiters, use the host stars' C, O, Mg, Si, and Fe abundances to estimate the total metal content of each protoplanetary disk, and compare the two groups. They find that super-Jupiter hosts have at least as much disk metal as Jupiter hosts and often slightly more, with mean total metal content above the proto-solar disk value. They take this as evidence that the metal budget does not rule out core accretion for the most massive giant planets, and they encourage modelers to test that conclusion.","feed_headline":"Super-Jupiters form in disks at least as metal-rich as Jupiters","feed_subtitle":"A 265-planet sample finds 4–13 Jupiter-mass hosts carry equal or slightly more disk metal, favoring core accretion.","key_machinery":"The central machinery is the stoichiometric model that turns stellar atmospheric abundances of C, O, Mg, Si, and Fe into Z, the mass fraction of heavy elements in planet-building solids, anchored to a modern solar composition reference. From Z the paper constructs Ztotal = Z × Mstar under a linear disk-mass–stellar-mass scaling, plus a conservative variant Z*_total = log(Z × Mstar) × σdisk that propagates the observed ~0.8 dex scatter in disk mass. The comparison between Jupiter and super-Jupiter groups is carried by two-sample t-tests, Kolmogorov–Smirnov and Anderson–Darling tests, and Monte Carlo draws that fold in abundance, stellar-mass, and disk-mass uncertainties.","core_discovery":"The paper's central claim is that stars hosting planets with masses above 4 Jupiter masses do not have metal-poor protoplanetary disks relative to stars hosting 1–4 Jupiter-mass planets. Using a stoichiometric model that converts atmospheric C, O, Mg, Si, and Fe abundances into Z, the metal mass fraction of planet-building material, and multiplying by stellar mass to approximate the disk's total metal content Ztotal, the authors report mean Z = 1.79 ± 0.58% for super-Jupiter hosts versus 1.63 ± 0.58% for Jupiter hosts, and mean Ztotal = 2.21 ± 0.84 versus 1.99 ± 0.69. A two-sample t-test on Ztotal gives p = 0.023, with Kolmogorov–Smirnov and Anderson–Darling tests also favoring distinct distributions, although Monte Carlo realizations that include the uncertainties do not always reach significance. The authors conclude that super-massive planets can plausibly form via core accretion; they note that 12 of 93 super-Jupiters have sub-solar disk metal content, where gravitational instability may still be needed.","pith_inferences":["A direct test of the paper's premise would be to compare present-day surface abundances with the compositions of intact protoplanetary disks around young stars; any systematic offset in C/O or Mg/Si ratios would propagate directly into Z.","If the constancy assumption holds, the paper implies that [Fe/H] alone systematically understates the metal content of low-metallicity, α-enhanced stars, which could explain why earlier work saw super-Jupiter hosts as metal-poor.","The p-values near 0.02–0.05 suggest the effect is real but small; a larger homogeneous abundance survey of super-Jupiter hosts would show whether the ~0.22 excess in mean Ztotal persists.","The 12 sub-solar super-Jupiters are a natural place to look for signs of gravitational instability, such as wide orbits or different companion populations, because core accretion is hardest to invoke there."],"forward_implications":["The metal content of protoplanetary disks is not the bottleneck that rules out core accretion for planets above 4 Jupiter masses.","Earlier hints that super-Jupiter hosts are metal-poor likely came from using iron abundance alone; including C, O, Mg, and Si shifts the picture.","About 87% of super-Jupiter hosts in the sample have at least proto-solar total disk metal content, so core accretion can plausibly build most of these planets; the remaining sub-solar cases remain a separate puzzle.","Planet-formation models should treat super-Jupiters as a viable core-accretion outcome under these observed Ztotal values rather than assuming a sharp switch to gravitational instability at 4–10 Jupiter masses."],"supporting_citations":[{"why":"Lays out the core-accretion model whose viability for super-Jupiters the paper defends.","marker":"Pollack et al, 1996"},{"why":"Introduces gravitational instability, the alternative formation channel the paper argues against for most super-Jupiters.","marker":"Boss, 1997"},{"why":"Established the planet–metallicity correlation that motivates the core-accretion interpretation.","marker":"Santos et al, 2001"},{"why":"Earlier comparison of Jupiter and super-Jupiter host metallicities that this work updates with a larger sample.","marker":"Adibekyan et al, 2013"},{"why":"Supplies the stoichiometric model used to convert C, O, Mg, Si, and Fe abundances into the metal fraction Z.","marker":"Santos et al, 2017a"},{"why":"Provides the solar reference abundances used to put the measured abundances on an absolute scale.","marker":"Asplund et al, 2021"},{"why":"Compiles the stellar abundance data used to derive Z for the sample hosts.","marker":"Hinkel et al, 2014"},{"why":"Source of the disk-mass dispersion σdisk used in the conservative Z*_total estimate.","marker":"Manara et al, 2023"},{"why":"Asserts that rock-forming element abundance ratios stay constant over FGK-star evolution, the assumption that lets present-day abundances stand in for disk composition.","marker":"Adibekyan et al. (2024, A&A, submitted)"}],"fun_headline_variants":["Metal-rich disks spawn super-Jupiters","Super-Jupiters born in metal-rich disks","Disk metals favor core accretion for super-Jupiters","Super-Jupiters not metal-poor: disk metals high","Core accretion viable for super-Jupiters via disk metals"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the C, O, Mg, Si, and Fe measured in a host star's atmosphere today have the same relative amounts as the metals in its protoplanetary disk at the time the planets formed, so that present-day abundance ratios can stand in for the disk's primordial mix; if stellar evolution has changed those ratios, the estimated metal contents are unreliable and the comparison between the two planet groups collapses.","fun_headline_variants_meta":{"raw":{"variants":["Metal-rich disks spawn super-Jupiters","Super-Jupiters born in metal-rich disks","Disk metals favor core accretion for super-Jupiters","Super-Jupiters not metal-poor: disk metals high","Core accretion viable for super-Jupiters via disk metals"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001217,"raw_usage":{"total_tokens":5025,"prompt_tokens":982,"completion_tokens":4043,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":598,"completion_tokens_details":{"reasoning_tokens":3967}},"tokens_in":598,"tokens_out":4043,"duration_ms":34252,"temperature":1.0,"reasoning_tokens":3967,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T19:29:00.384332+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Rebuild the analysis using only abundances from one homogeneous high-resolution survey for every host star and compare the mean Ztotal of the 4–13 Mjup group with the 1–4 Mjup group; if it is then lower, the paper's statistical claim is overturned. Separately, measure C, O, Mg, Si, and Fe in stars of fixed initial composition but different evolutionary states (for example, open-cluster stars of different masses); if the rock-forming element ratios change with evolution, the present-day-abundance premise fails.","supporting_citations":[{"cited_title":"Orbital and physical properties of planets and their hosts: new insights on planet formation and evolution","cited_arxiv_id":"1311.2417","evidence_quote":"Earlier comparison of Jupiter and super-Jupiter host metallicities that this work updates with a larger sample."},{"cited_title":"arXiv e-prints arXiv:2409.11965","cited_arxiv_id":null,"evidence_quote":"Asserts that rock-forming element abundance ratios stay constant over FGK-star evolution, the assumption that lets present-day abundances stand in for disk composition."}],"review_version":1}