{"id":"3702db74-c6e3-4124-8884-70dad3ca5ce0","arxiv_id":"2412.06137","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Giant planets around M dwarfs are preferentially found around metal-rich stars, with no significant metallicity difference between hot and warm Jupiters.","lead":"Using near-infrared spectra of 22 M dwarfs with known giant planets and 746 field M dwarfs, the authors find that giant planets orbit metal-rich M dwarfs at 4-5 sigma significance. The result extends the well-known planet-metallicity correlation from Sun-like stars to M dwarfs and suggests a shared formation channel.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proper-motion selection of the field sample (Sec. 2.1) biases it toward kinematically hot, older, metal-poor M dwarfs; the 4-5 sigma metallicity excess (Sec. 5.1) may be a selection artifact unless the field sample is shown to be representative.","rationale":"The paper is carefully executed: homogeneous SpeX metallicities, uniform methodology, honest discussion of selection functions, and a contamination simulation. The central claim is statistically strong conditional on the field sample. I agree with the reader that the weakest assumption is field representativeness. Without validation of the proper-motion-selected field sample against an unbiased volume-limited population, the 4-5 sigma significance could be an artifact of comparing a kinematically biased (old, metal-poor) field to a broader planet-host sample. No other concern is as load-bearing: calibration extrapolation above [Fe/H]=0.56 affects only a few stars and would weaken rather than reverse the claim; the hot/warm Jupiter null is acknowledged as underpowered; and the two-planet systems are checked to not alter results. A kinematic matching test using Gaia astrometry is the decisive check, and it is feasible with existing data.","tokens_in":26977,"tokens_out":7887,"duration_ms":78071,"concrete_test":"Compute transverse velocities from Gaia DR3 proper motions and parallaxes for both the 746 field stars and the 22 planet hosts. If the field sample's transverse-velocity distribution is significantly higher (K-S p<0.05), apply a kinematic-matched control: redo the Section 5.1 K-S/A-D tests on a randomly drawn subset of field stars matched to the planet hosts' transverse-velocity distribution (e.g., rejection sampling in 1 km/s bins). If the reported p-value inflates above 0.003 (3 sigma), the metallicity excess is a selection artifact; if it remains below 0.003, the claim is robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Sec. 5.1: 4-5 sigma metallicity excess of 22 M dwarf giant planet hosts over 746 field M dwarfs) rests on the field sample being representative of the parent population from which the planet hosts are drawn. The field sample comes from Terrien et al. (2012), whose targets were selected primarily from the Lépine & Shara (2005) and Lépine & Gaidos (2011) proper-motion catalogs. Proper-motion selection enriches for high transverse-velocity, kinematically old stars; in the solar neighborhood, such stars are systematically metal-poor. The planet host sample is assembled from TESS, HATS, NGTS, and RV surveys that do not apply a proper-motion cut. If the field sample is shifted low in metallicity relative to the true volume-complete M dwarf population, the K-S/A-D p-values in Table 5 overstate the significance; the observed excess could reflect age/kinematics rather than a planet formation metallicity dependence. The paper simulates contamination by undetected giant planets but does not test field-sample representativeness against a volume-limited sample or compare kinematic properties of the two samples. This is the weakest load-bearing assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper compares the stellar metallicity distributions of 22 M dwarfs hosting 27 confirmed giant planets with 746 field M dwarfs without known giant planets. All metallicities are derived homogeneously from IRTF/SpeX near-infrared spectra using the same methodology and calibrations. The authors report that giant planets favor metal-rich M dwarfs at 4–5 sigma significance, that hot (a/R* <= 20) and warm (a/R* > 20) Jupiters have indistinguishable metallicity distributions, and that there is no significant correlation between stellar metallicity and planet mass. They also examine multi-giant-planet systems and the dependence of the metallicity signal on stellar mass.","tokens_in":27280,"tokens_out":6829,"duration_ms":67651,"significance":"If the central claim holds, this is an important result: it would extend the well-established giant-planet metallicity correlation from FGK stars to M dwarfs, with direct implications for core-accretion formation models. The paper's strengths include the use of a single instrument and homogeneous analysis for all planet hosts, public machine-readable catalogs, permutation-based Anderson-Darling tests with Monte Carlo propagation of metallicity uncertainties, and an explicit contamination simulation for undetected giant planets in the field sample. The authors are also transparent about known limitations, including calibration extrapolation beyond the validated parameter range and the heterogeneity of the planet-survey selection functions. The main risk to the conclusion is the kinematic selection of the field sample, which is not tested against a volume-limited benchmark.","major_comments":[{"comment":"The field M dwarf sample is drawn from Terrien et al. (2012), whose targets were selected primarily from the Lépine & Shara (2005) and Lépine & Gaidos (2011) proper-motion catalogs, whereas the 22 planet hosts come from TESS, HATS, NGTS, and RV surveys that do not apply a proper-motion cut. High proper-motion M dwarfs in the solar neighborhood are kinematically older and systematically more metal-poor on average, so this selection mismatch could lower the field metallicity distribution relative to the parent population from which the planet hosts are drawn, inflating the 4–5 sigma excess reported in Table 5. The paper simulates contamination by undetected giant planets but does not test the representativeness of the field sample against a volume-limited sample, nor does it compare the kinematic properties (e.g., tangential velocity or Gaia-based U,V,W) of the two samples. I request two additional robustness tests: (a) compare the field sample metallicity distribution to a volume-complete M dwarf sample (e.g., a Gaia-selected volume-limited sample or the 10 pc RECONS sample), and (b) repeat the K-S and A-D tests after restricting both samples to a common proper-motion or tangential-velocity range. Until these tests are provided, the significance level of the central metallicity-excess claim is not fully established.","section":"Section 2.1 and Section 5.1, Table 5"},{"comment":"The paper acknowledges that the Mann et al. (2013) and Rojas-Ayala et al. (2012) metallicity calibrations are untested for M dwarfs beyond M5 and for [Fe/H] > 0.56 dex, yet the planet sample contains stars outside this range, including GJ 3512 (M5V), HIP 79431 ([Fe/H]M13,K = 0.72), and TOI-5205 ([Fe/H]M13,K = 0.69). Because these are among the most metal-rich hosts, they could influence the significance of the K-S and A-D results. The authors should perform a sensitivity test that excludes these extrapolated points and report the resulting p-values, or otherwise demonstrate quantitatively that the extrapolation does not drive the conclusion.","section":"Section 4.1 and Table 3"}],"minor_comments":[{"comment":"The author name 'Sharon X. W ang' appears to contain a spacing typo; it should likely read 'Sharon X. Wang'.","section":"Title page"},{"comment":"The contamination simulation sentence is ambiguous: 'We randomly choose 22 stars in the field M dwarf sample, assuming they host giant planets and have the highest metallicity (0.7 dex) as in our planet sample' could be read as selecting the highest-metallicity stars rather than assigning the highest planet-sample metallicity to randomly chosen stars. Please rephrase for clarity.","section":"Section 2.1"},{"comment":"The phrase 'we do not induce the constraint on the scaled semi-major axis a/R* from the light curve' should likely be 'we do not include the constraint'.","section":"Section 4.2"},{"comment":"The right panel of Figure 1 appears to lack an explicit color-bar label for the metallicity color coding, and the left panel caption could clarify what the color scale represents; please add the missing labels.","section":"Figure 1"},{"comment":"The cold Jupiter boundary at a/R* = 200 is described as 'somewhat arbitrarily' chosen, and the comparison between hot and cold Jupiters yields p-values between 4.4e-3 and 2.6e-1. Given the multiple testing inherent in choosing the boundary, the authors should state explicitly that these p-values are not corrected for the boundary choice and that the trend is tentative, as they do in the text.","section":"Section 5.2, Figure 6"},{"comment":"For the K-band field vs. M+WJ comparison, the K-S fraction of resampled trials with p <= 0.003 is only 24.4%, and the A-D fraction is 49.2%; this indicates that the warm-Jupiter subgroup result is not robust to metallicity uncertainties in the K band. The text in Section 5.1 discusses the combined sample but does not mention this limited robustness for the subgroup; adding a sentence would be helpful.","section":"Table 5"}],"recommendation":"major_revision","confidential_remarks":"The paper is well-written and the central claim is plausible, but the proper-motion selection of the field sample is a genuine threat to the claimed significance. The requested kinematic comparisons and volume-limited validation are feasible with the data already in hand (Gaia DR3 measurements are in Table 1) and should be done before publication. The calibration-extrapolation sensitivity test is also straightforward. I do not see circularity or internal inconsistency; the issue is robustness of the statistical comparison. The paper is within scope for this journal and, once the additional tests are provided, would be a valuable contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick read: this is a careful, incremental but real step. The genuinely new pieces are the homogeneous SpeX metallicities for 22 M-dwarf giant planet hosts (27 planets) and the hot/warm Jupiter null result, which goes against the tentative difference the same group reported in 2023. The metal-rich preference itself is not new—Johnson & Apps and Rojas-Ayala already suggested it—but 4-5 sigma with a single instrument and pipeline is a real improvement.\n\nThe methods are transparent. They use established calibrations (Mann et al. 2013, Rojas-Ayala et al. 2012), treat H- and K-band separately, Monte Carlo the metallicity errors, and simulate contamination by undetected giants in the field sample. Tables are complete and the spectra are available. I believe the central claim: giant planets around M dwarfs do prefer metal-rich hosts, and the FGK correlation extends to low-mass stars.\n\nThe soft spot is exactly what the reader flagged. The 'field' sample comes from Terrien et al. (2012), whose targets were drawn from the Lépine & Shara (2005) and Lépine & Gaidos (2011) proper-motion catalogs. Proper-motion selection enriches for high tangential velocity, likely older and metal-poorer stars than a volume-limited M-dwarf population. The planet hosts come from TESS/HATS/NGTS/RV surveys with no such cut. If the field baseline is shifted low, the observed excess is inflated. The paper acknowledges the planet sample's heterogeneous selection, but it never tests whether the field sample is representative. That is the weakest load-bearing assumption, and it is fixable: show the Gaia kinematics of both samples, compare with a volume-limited sample (e.g., within 15 pc or the Cifuentes catalog), or redo with a matched sample. The paper already has the Gaia data.\n\nOther issues are minor. The K-band results are weaker than H-band, especially for warm Jupiters alone (p ~ 1e-2). A few high-metallicity hosts lie beyond the calibration range and the authors flag that themselves. The hot/warm null is underpowered, but they phrase it correctly as a non-rejection. The a/R* = 200 cold Jupiter boundary is admittedly arbitrary, but they don't lean on it.\n\nRecommendation: send it to referees. Ask for a robustness test of the field sample before accepting. This is not a desk reject.","headline":"Homogeneous SpeX sample gives the cleanest evidence yet that M-dwarf giant planets prefer metal-rich hosts, but the proper-motion-selected field baseline needs validation before I'd trust the 4-5 sigma.","tokens_in":27788,"tokens_out":4998,"would_cite":true,"duration_ms":46704,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper finds that M dwarfs hosting confirmed giant planets are systematically more metal-rich than field M dwarfs, at 4–5 sigma significance, and that hot and warm Jupiters around M dwarfs show the same host-metallicity preference.","keywords":["M dwarfs","giant planets","stellar metallicity","planet-metallicity correlation","hot Jupiters","warm Jupiters","near-infrared spectroscopy","SpeX"],"falsifier":"A decisive check would be an unbiased, volume-limited sample of M dwarfs with metallicities measured by an independent method, such as high-resolution optical spectroscopy, combined with a well-characterized giant planet survey: if the 4–5 sigma excess does not appear, the proper-motion or survey selection is the cause. A second decisive check is to show directly that the proper-motion-selected field sample's metallicity distribution differs from a Gaia-based volume-limited sample in a way that accounts for the excess.","tokens_in":26769,"feed_emoji":"🪐","tokens_out":7479,"duration_ms":61074,"temperature":0.7,"pith_summary":"This paper asks whether M dwarfs that host giant planets are systematically more metal-rich than the general M dwarf population, in the way Sun-like stars are known to be. It compares 746 field M dwarfs without known giant planets to 22 M dwarfs hosting 27 confirmed giant planets, with all metallicities measured in one homogeneous way from near-infrared SpeX spectra. The authors find that giant planets favor metal-rich M dwarfs at the 4–5 sigma level and that hot and warm Jupiter hosts have statistically indistinguishable metallicity distributions. If correct, this extends the well-established giant planet–metallicity correlation down to the most common stars in the solar neighborhood and suggests a shared formation channel across stellar types.","feed_headline":"Giant planets prefer metal-rich M dwarfs at 4–5 sigma","feed_subtitle":"Homogeneous spectra of 22 planet hosts show the same metallicity bias seen around Sun-like stars.","key_machinery":"The load-bearing tool is a homogeneous metallicity scale applied to both samples: near-infrared H-band and K-band spectra from the single instrument SpeX, reduced with the same pipeline and converted to [Fe/H] and [M/H] using empirical calibrations anchored to wide binaries with Sun-like primaries. The field comparison sample comes from a SpeX survey of proper-motion-selected nearby M dwarfs, and the planet hosts were observed in the same mode with the same instrument. The statistical argument then rests on two-sample Kolmogorov–Smirnov and Anderson–Darling tests, with permutation and resampling procedures that fold in the metallicity measurement uncertainties, run separately on H-band and K-band outputs because the two calibrations show a small offset.","core_discovery":"On the paper's own terms, the central discovery is that the iron abundance distributions of field M dwarfs and M dwarfs with confirmed giant planets cannot be drawn from the same parent distribution: Kolmogorov–Smirnov and Anderson–Darling tests give p-values between $4.9 \\times 10^{-7}$ and $6.3 \\times 10^{-5}$ depending on the band, corresponding to 4–5 $\\sigma$ significance, and the result survives resampling of the metallicity uncertainties. Hot Jupiters ($a/R_* \\le 20$) and warm Jupiters ($a/R_* > 20$) show indistinguishable metallicity distributions, and the paper finds no significant correlation between host metallicity and planet mass, no special metallicity excess for multi-planet hosts, and no apparent preference for higher metallicity among mid-to-late M dwarfs compared with early-M hosts. The authors interpret the metal-rich preference as evidence that giant planet formation around M dwarfs proceeds through a channel shared with FGK stars, likely core accretion.","pith_inferences":["Beyond the paper: if the correlation is real, the giant planet occurrence rate around M dwarfs should rise steeply with [Fe/H], and future uniform surveys could measure that slope and compare it with the FGK power law.","Beyond the paper: the tentative weaker metallicity preference of cold Jupiters with $a/R_* > 200$ hints that wide-orbit giants may form through a different pathway, such as disk gravitational instability; this is a testable prediction for long-term radial velocity and astrometry surveys.","Beyond the paper: because M dwarfs are the most common stars in the solar neighborhood, a robust planet–metallicity relation here strengthens the case for prioritizing metal-rich M dwarfs in future transit and RV searches for giant planets."],"forward_implications":["The giant planet–metallicity correlation, previously well established for FGK stars, extends to M dwarfs, so any formation theory must explain the same metal-rich preference across a wide range of stellar mass.","Hot and warm Jupiters around M dwarfs appear to share a common origin, since their host metallicity distributions are statistically indistinguishable.","Contamination of the field sample by undetected giant planets is negligible: injecting up to 22 high-metallicity hosts into the field sample shifts its median metallicity by less than 1 sigma.","Within the current sample, stellar iron abundance does not set the final giant planet mass, as no significant host metallicity–planet mass correlation is found."],"supporting_citations":[{"why":"Supplies the field M dwarf catalog and the H-band and K-band [Fe/H] and [M/H] measurements used as the comparison population.","marker":"Terrien et al. (2015)"},{"why":"Defines the SpeX survey and the H-band calibration relation that both the field sample and planet sample share.","marker":"Terrien et al. (2012)"},{"why":"Provides the metal algorithm and wide-binary calibrations that convert the SpeX spectra into metallicities.","marker":"Mann et al. (2013)"},{"why":"Supplies the K-band metallicity calibration used alongside the H-band one.","marker":"Rojas-Ayala et al. (2012)"},{"why":"Earlier photometry-based evidence that M dwarfs with planets are metal-rich, which this work compares against.","marker":"Johnson & Apps (2009)"},{"why":"The FGK giant planet–metallicity correlation that this paper extends to M dwarfs.","marker":"Fischer & Valenti (2005)"},{"why":"Previous heterogeneous hint that hot Jupiters prefer more metal-rich M stars, revisited here with homogeneous data.","marker":"Gan et al. (2023b)"}],"fun_headline_variants":["Metal-rich M dwarfs host more giant planets, 4–5 sigma","M dwarf giants prefer metal-rich stars, 5 sigma","Hot and warm Jupiters share metallicity around M dwarfs","Giant planets around M dwarfs: metal-rich hosts win","M dwarf giants: no metallicity gap between hot and warm Jupiters"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the 746 proper-motion-selected field M dwarfs fairly represent the parent population from which the 22 giant-planet hosts are drawn; if that field sample is biased toward older, metal-poor stars, the measured metallicity excess could be a selection artifact.","fun_headline_variants_meta":{"raw":{"variants":["Metal-rich M dwarfs host more giant planets, 4–5 sigma","M dwarf giants prefer metal-rich stars, 5 sigma","Hot and warm Jupiters share metallicity around M dwarfs","Giant planets around M dwarfs: metal-rich hosts win","M dwarf giants: no metallicity gap between hot and warm Jupiters"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000252,"raw_usage":{"total_tokens":1604,"prompt_tokens":1032,"completion_tokens":572,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":648,"completion_tokens_details":{"reasoning_tokens":480}},"tokens_in":648,"tokens_out":572,"duration_ms":5343,"temperature":1.0,"reasoning_tokens":480,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T19:59:18.995640+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check would be an unbiased, volume-limited sample of M dwarfs with metallicities measured by an independent method, such as high-resolution optical spectroscopy, combined with a well-characterized giant planet survey: if the 4–5 sigma excess does not appear, the proper-motion or survey selection is the cause. A second decisive check is to show directly that the proper-motion-selected field sample's metallicity distribution differs from a Gaia-based volume-limited sample in a way that accounts for the excess.","supporting_citations":[],"review_version":1}