{"id":"3c0bfdc3-0db5-4878-93a1-7aa22af10ca7","arxiv_id":"2411.13011","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Galactic lithium abundance rose, dipped at stellar ages of 4-5 Gyr, then rose rapidly; young lithium-rich stars migrated inward from the outer disc.","lead":"Using 22,034 stars with precise ages and 3D NLTE lithium abundances, this paper maps how lithium changed across the Milky Way disc over 14 billion years. It finds young lithium-rich stars were born in the outer disc and migrated inward, and dates the lithium dip to ages of 4 to 5 billion years.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Birth-radius calibration in Eq. 7 is the load-bearing step: unquantified systematics in the assumed time-dependent [Fe/H] gradient can shift Rbirth by several kpc and could reverse the outer-origin and gradient-period claims.","rationale":"The reader's weakest_assumption is exactly the load-bearing point: Eq. 7, the Lu et al. (2022) calibration, and the absence of uncertainty propagation. I agree, and I would not change the CONDITIONAL verdict. The directly observed temporal trend (rise, dip, rise) and the 4-5 Gyr Li-dip age identification are supported by the data themselves and by the independent Kiel-diagram check in Fig. 6, so they are not the weakest link. The spatial claims, by contrast, are entirely downstream of Rbirth; if the calibration is wrong or even just not robust at the kpc level, the 'outer disc origin and inward migration' narrative and the three gradient periods lose their evidential support. The proposed test is deliberately calibration-focused: recomputing Rbirth with two alternative plausible gradient prescriptions would immediately show whether the 70% outer-disc fraction and the sign of the 1-3 Gyr gradient are stable. A Monte Carlo propagation adds the necessary uncertainty quantification that the paper currently omits. No ad hominem or theatrical framing is intended; this is a standard robustness check for a method whose output feeds every spatial conclusion in the paper.","tokens_in":14125,"tokens_out":10901,"duration_ms":110978,"concrete_test":"Replace the Lu et al. (2022) gradient table in Eq. 7 with (i) a time-independent gradient fitted to the present-day GALAH [Fe/H]-Rguide relation and (ii) the Minchev et al. (2018) time-dependent calibration, recomputing Rbirth for all 22,034 stars. Then re-measure the fraction of A(Li)>2.7, age<4 Gyr stars with Rbirth>10 kpc and the slope of A(Li) vs Rbirth in the 1-3 Gyr bin. If the fraction shifts by more than 20 percentage points or the slope sign flips, the outer-origin/migration and three-period conclusions are calibration-dependent. As a secondary check, Monte Carlo propagate the quoted age and [Fe/H] errors plus the two Lu et al. calibrations to place confidence intervals on those quantities.","verdict_should_be":"UNCHANGED","load_bearing_attack":"All novel spatial conclusions—the outer-disc origin of young Li-rich stars, their inward migration, and the three A(Li) gradient periods—are computed from Rbirth via Eq. 7. Eq. 7 assumes a linear, time-dependent ISM [Fe/H] gradient calibrated with Lu et al. (2022), and the paper propagates neither stellar age uncertainties (median ~10%), [Fe/H] uncertainties, nor the spread between the two Lu et al. simulation calibrations through the formula. With a present-day gradient near -0.05 dex/kpc, a 0.1 dex systematic in [Fe/H] or in the zero-point moves Rbirth by ~2 kpc; the '~70% have Rbirth>10 kpc' claim and the second-period sign change are therefore precision-dependent. No independent validation of the birth radii is provided (e.g., against open clusters or an alternative estimator such as Minchev et al. 2018), so the migration and gradient-period claims cannot be distinguished from calibration artifacts. The paper also does not check sign consistency: if young stars increase in A(Li) with [Fe/H] as stated for Fig. 2, Eq. 7's negative gradient predicts a negative A(Li)-Rbirth slope for ages 1-4 Gyr, opposite to the positive gradient shown in Fig. 7.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript uses 22,034 GALAH DR3 main-sequence turn-off and subgiant stars with ages from Sun et al. (2023a), 3D NLTE Li abundances from Wang et al. (2024), and birth radii computed from the Lu et al. (2022) method to study temporal and spatial variations of Li in the Milky Way disc. It reports a non-monotonic age--A(Li) relation: a gradual rise from 14 to 6 Gyr, a decline to about 4.5 Gyr, and a rapid rise thereafter. It further claims that most young Li-rich stars (age < 4 Gyr, A(Li) > 2.7 dex) have large birth radii and migrated inward, that Li-dip stars occupy ages 4--5 Gyr, and that the radial A(Li) gradient with respect to birth radius has three distinct epochs (14--6, 6--4, and 4--1 Gyr ago).","tokens_in":14417,"tokens_out":7155,"duration_ms":67704,"significance":"If the results hold, this would be one of the first large-sample studies connecting precise MSTO/subgiant ages, 3D NLTE Li abundances, and birth radii, providing new constraints on Li enrichment and radial migration in the Galactic disc. A particular strength is the use of external catalogs for ages and Li abundances, which limits circularity. However, the current support for the main claims is incomplete: the central temporal trend is shown without uncertainty bands, the birth-radius calibration is not error-propagated or independently validated, and there is an internal sign inconsistency between the [Fe/H]--A(Li) and Rbirth--A(Li) relations for young stars. These issues must be addressed before the paper's quantitative conclusions can be accepted.","major_comments":[{"comment":"The central temporal trend is derived from LOESS smoothing with frac=0.15 and is plotted without confidence intervals, so the reported decline from about 2.2 dex at 6 Gyr to about 2.1 dex at 4.5 Gyr is not shown to be statistically significant relative to the large scatter in the age--A(Li) plane. Because the median age uncertainty is about 10%, a 1--1.5 Gyr feature can be smeared by age errors. Please provide bootstrap confidence bands for the LOESS fits in Figs. 1 and 4, demonstrate sensitivity to the smoothing fraction used (0.15 vs 0.4), and report a quantitative significance estimate for the 6--4.5 Gyr decline.","section":"§3.1, Fig. 1(a)"},{"comment":"Birth radii are derived from a linear ISM metallicity-gradient calibration without propagating age and [Fe/H] uncertainties and without treating the systematic difference between the two Lu et al. (2022) simulation calibrations. With a gradient near -0.05 dex/kpc, a 0.1 dex systematic in [Fe/H] or in the gradient zero-point shifts Rbirth by about 2 kpc, which is large enough to affect the '~70% have Rbirth > 10 kpc' statement and the inferred gradient signs. Please add error propagation, a sensitivity analysis over the Lu et al. calibrations, and an external validation of the derived birth radii against open clusters or an alternative birth-radius estimator.","section":"§2.2, Eq. (7)"},{"comment":"There is an internal sign inconsistency between the [Fe/H]--A(Li) and Rbirth--A(Li) relations for young stars. Equation (7) with a negative ISM metallicity gradient makes Rbirth decrease with [Fe/H] at fixed age, yet Fig. 2(a) shows A(Li) increasing with [Fe/H] for age < 4 Gyr while Fig. 7 shows a positive A(Li)--Rbirth gradient in the 1--3 and 3--4 Gyr bins. These two relations cannot both hold if Eq. (7) maps [Fe/H] to Rbirth monotonically. Please check the sign of the adopted gradient, reconcile the two figures, or explicitly explain why the sign reversal is physically expected.","section":"§3.1 and §3.3, Figs. 2 and 7"}],"minor_comments":[{"comment":"The text says the decreasing trend from ~6 to ~4 Gyr appears in 'nearly all' subsamples except Fig. 4(a), but then states the Li-dip age range is 'irrespective of positions in the Milky Way'; these statements should be reconciled.","section":"§3.2, text near Fig. 4"},{"comment":"There is a typo: 'subginat' should be 'subgiant'.","section":"§3.2"},{"comment":"The caption phrase 'no weighting function adjustment (delta = 0)' is unclear in the context of the statsmodels LOESS implementation; please rephrase to describe what delta=0 means.","section":"Fig. 1 caption"},{"comment":"The data availability statement only offers data 'on reasonable request'; making the sample table and derived Rbirth values publicly available would improve reproducibility.","section":"Data Availability"},{"comment":"The Lu et al. (2022) reference is cited as an arXiv e-print; please update to the published version if available.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The internal sign inconsistency between the [Fe/H]--A(Li) relation and the Rbirth--A(Li) gradient is the main risk: if it is not resolved, the spatial conclusions and the three-period gradient claim may not be salvageable. The paper also relies heavily on the authors' own Paper II for ages; the editor may wish to verify the publication status of Paper II. The novelty is moderate and the manuscript is concise, but the missing uncertainty propagation is a substantial gap given the strength of the claims."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a look for the temporal result, but the spatial half of the paper doesn't survive contact with its own assumptions.\n\nWhat's new: they combine precise ages from their own earlier program with the external GALAH 3D NLTE Li catalogue and simulation-based birth radii for 22,000 MSTO/subgiant stars. That's a clean setup. The age–A(Li) trend—rise to ~6 Gyr, slight dip around 4.5 Gyr, then steep rise—is an independent empirical constraint, and the 4–5 Gyr age range for the Li dip is supported by a Kiel-diagram check using the Wang et al. catalogue. That part is worth taking seriously, and it does not depend on the birth-radius machinery.\n\nThe soft spot is exactly where the stress-test points: the birth radius. Equation 7 is a linear, lookback-time-dependent [Fe/H] gradient taken from Lu et al., and none of the relevant uncertainties are propagated. That would be a minor complaint if the spatial trends were robust to plausible shifts. But there is a sharper problem. For stars younger than 4 Gyr, the paper says A(Li) increases with [Fe/H]. With a negative ISM gradient, Eq. 7 makes Rbirth a strictly decreasing function of [Fe/H]. Therefore A(Li) should decrease with Rbirth in any given young age bin. The paper's Fig. 7 instead shows a positive A(Li)–Rbirth gradient for 1–3 Gyr stars. That is a direct sign contradiction, not a matter of error bars. It suggests either the mapping is misapplied or the interpretation of the radial profile is flawed. The '~70% of Li-rich stars have Rbirth>10 kpc' claim is also fragile: a 0.1 dex zero-point shift moves Rbirth by ~2 kpc, and there is no validation against open clusters or an alternative estimator.\n\nThe authors owe the reader a resolution of that sign issue, at minimum. If the birth radii are wrong, the outer-origin and migration narrative collapses, and the three gradient epochs become an artifact of the assumed calibration. The temporal claims can be salvaged, but the spatial claims need either a corrected analysis or explicit retraction.\n\nI'd send this to a serious referee—the sample and the temporal result justify the time—but the referee should demand an explanation of the sign inconsistency and a proper validation of the birth radii before any spatial conclusions are accepted. As it stands, the paper is not coherent on its own terms in the spatial half.","headline":"The age–lithium relation is a real empirical step forward, but the spatial story has an internal sign contradiction that undermines the migration and gradient claims.","tokens_in":14989,"tokens_out":5843,"would_cite":true,"duration_ms":55822,"reading_group":"yes","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that lithium in the Milky Way's disc rose from 14 to 6 billion years ago, dipped to a minimum near 4.5 billion years, then rose again sharply, with the youngest lithium-rich stars born in the outer disc and migrating…","keywords":["lithium abundance","Galactic disc","radial migration","birth radius","stellar ages","Li dip","GALAH survey","main-sequence turn-off stars"],"falsifier":"Take the same 22,034 stars and compute $R_{\\rm birth}$ from a forward chemical-evolution model with explicit radial migration and a non-linear, time-dependent metallicity gradient, rather than the linear lookback-time gradient of Eq. (7). If the young Li-rich stars no longer cluster at $R_{\\rm birth} > 10$ kpc, or the 6-4 Gyr gradient no longer turns negative and broken, the three-epoch lithium history is an artifact of the assumed linear gradient. Independently, measure lithium in young stars whose orbital actions indicate outer-disc birth and check whether their $A(\\mathrm{Li})$ exceeds 2.7 dex.","tokens_in":13925,"feed_emoji":"🌌","tokens_out":10965,"duration_ms":91890,"temperature":0.7,"pith_summary":"This paper uses 22,034 main-sequence turn-off and subgiant stars with precise ages, 3D non-local thermodynamic equilibrium (NLTE) lithium abundances, and reconstructed birth radii to argue that lithium in the Milky Way's disc did not grow smoothly. It finds three phases in $A(\\mathrm{Li})$—the logarithmic lithium abundance relative to hydrogen: a gradual rise from 14 to 6 billion years ago, a decline to a minimum near 4.5 billion years, and a rapid rise after that, with stars younger than 4 billion years reaching values above the primordial prediction. The paper also argues that young lithium-rich stars ($A(\\mathrm{Li}) > 2.7$ dex, age $<4$ Gyr) were mostly born in the outer disc and migrated radially inward, and that the radial lithium gradient flipped from positive to negative/broken to positive across these eras. A reader should care because this dates the long-known lithium dip to 4-5 Gyr and ties chemical enrichment to inside-out disc formation and radial migration.","feed_headline":"Galactic lithium dipped 6-4.5 billion years ago, then surged","feed_subtitle":"Young lithium-rich stars point to the outer disc, revealing how migration shaped the Galaxy's chemical past.","key_machinery":"The load-bearing object is the birth radius $R_{\\rm birth}$, assigned through Eq. (7): $R_{\\rm birth} = ([\\mathrm{Fe/H}] - [\\mathrm{Fe/H}](0,\\tau)) / \\nabla[\\mathrm{Fe/H}](\\tau)$, where $\\nabla[\\mathrm{Fe/H}](\\tau)$ is the interstellar metallicity gradient at lookback time $\\tau$, taken from a linear relation calibrated by two sets of cosmological simulations and interpolated from a published table. This converts a present-day snapshot of stellar lithium into a time-resolved radial map of enrichment, and it is what lets the paper attribute young Li-rich stars to the outer disc. Supporting machinery includes the Bayesian stellar ages from matching Gaia luminosities with GALAH spectroscopic parameters, the 3D NLTE lithium catalogue, and LOESS local non-parametric regression used to extract trends from the age- and radius-binned scatter.","core_discovery":"On its own terms, the central discovery is that the temporal evolution of disc lithium is non-monotonic and spatially structured. In the age-$A(\\mathrm{Li})$ plane, the sample shows $A(\\mathrm{Li})$ rising from about 1.5 dex at 14 Gyr to roughly 2.2 dex at 6 Gyr, slipping to about 2.1 dex at 4.5 Gyr, and then climbing past 2.7 dex by 2 Gyr. The downturn marks the Li dip, and the paper pins its age range to 4-5 Gyr, spanning main-sequence turn-off stars through subgiants, with most of these stars having $T_{\\rm eff}$ below 6200 K. Binning by guiding-center radius and $z_{\\rm max}$, the paper finds that young Li-rich stars (age $<4$ Gyr, $A(\\mathrm{Li}) > 2.7$ dex) have birth radii mostly above 10 kpc in every spatial bin, indicating that they formed in the outer disc and migrated to the local and inner discs; stars born in the inner disc instead show a rapid lithium enrichment between 8 and 6 Gyr ago. The radial profile of lithium with respect to birth radius exhibits three epochs: a positive gradient at 14-6 Gyr ago, a negative and broken gradient at 6-4 Gyr ago driven by Li-dip stars, and a return to a positive gradient at 4-1 Gyr ago.","pith_inferences":["The paper leaves implicit that the outer-disc origin claim can be checked kinematically: these young Li-rich stars should have orbital actions indicating inward migration, and their current guiding radii should be systematically smaller than their birth radii.","If the 4-5 Gyr age range is robust, it links the Li dip to a narrow mass range at the turn-off, providing an empirical mass-age-metallicity constraint that models of diffusion, rotation, and internal gravity waves must reproduce.","A natural extension would be to look for abundance ratios tied to specific lithium production sites (novae versus AGB stars) in stars formed during the inner-disc 8-6 Gyr burst versus the outer-disc post-4 Gyr burst, testing whether the two enrichment episodes have different nucleosynthetic origins.","The softest point is the linear-gradient birth radius; re-deriving $R_{\\rm birth}$ with a non-linear or directly simulated metallicity-radius relation would show how much of the three-epoch gradient story survives a change in that assumption."],"forward_implications":["The Li dip becomes a dated phenomenon: field stars aged 4-5 Gyr at the turn-off should show depleted $A(\\mathrm{Li})$, giving an independent age indicator for intermediate-age stellar populations.","Young Li-rich stars in the solar neighbourhood should not be read as local enrichment; they trace inward radial migration from the outer disc, so models of the local disc must include migration to reproduce the upper envelope of the $A(\\mathrm{Li})$–metallicity relation.","The inner disc had a distinct lithium enrichment burst at 8-6 Gyr, separate from the outer disc's post-4 Gyr burst, implying that the dominant lithium sources or their efficiencies differed between regions and epochs.","Galactic chemical evolution models that predict $A(\\mathrm{Li})$ as a function of radius and time should reproduce the positive–negative/broken–positive sequence of gradients; failure to do so would point to missing lithium sources or wrong migration prescriptions."],"supporting_citations":[{"why":"Supplies the linear ISM metallicity-gradient calibration and interpolated table used in Eq. (7) to assign birth radii; the outer-disc and gradient claims stand on this.","marker":"Lu et al. (2022)"},{"why":"Provides the 3D NLTE lithium abundances for GALAH DR3 that define all $A(\\mathrm{Li})$ values and the Li-dip morphology.","marker":"Wang et al. (2024)"},{"why":"Paper II: constructs the 22,034-star MSTO/subgiant sample and the Bayesian ages with median 9.8% uncertainty used for every age bin.","marker":"Sun et al. (2023a)"},{"why":"Releases GALAH DR3 spectroscopic parameters, stellar parameters, and the value-added catalogue from which the sample, $z_{\\rm max}$, and kinematics are drawn.","marker":"Buder et al. (2021)"},{"why":"Provides the age-based thin/thick disc classification and the contrasting Li-metallicity behaviour that the paper compares with for old stars.","marker":"Bensby & Lind (2018)"},{"why":"Introduces the original birth-radius approach via time-dependent metallicity gradients that Lu et al. (2022) enhance and Eq. (7) adopts.","marker":"Minchev et al. (2018)"}],"fun_headline_variants":["Galactic lithium dips at 4-5 Gyr, then rebounds","Lithium-rich young stars trace outer disc migration","Milky Way lithium gradient flips across three epochs","From dip to surge: disc lithium's 14-Gyr story","Outer disc births lithium-rich stars that drift inward"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a star's birth radius can be recovered from its current [Fe/H] and age by assuming the interstellar metallicity gradient was always linear and is known from simulation calibration; if that linear relation or calibration is wrong, the claimed outer-disc origin of young Li-rich stars and the three gradient periods would be systematically shifted.","fun_headline_variants_meta":{"raw":{"variants":["Galactic lithium dips at 4-5 Gyr, then rebounds","Lithium-rich young stars trace outer disc migration","Milky Way lithium gradient flips across three epochs","From dip to surge: disc lithium's 14-Gyr story","Outer disc births lithium-rich stars that drift inward"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000448,"raw_usage":{"total_tokens":2374,"prompt_tokens":1172,"completion_tokens":1202,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":788,"completion_tokens_details":{"reasoning_tokens":1120}},"tokens_in":788,"tokens_out":1202,"duration_ms":11823,"temperature":1.0,"reasoning_tokens":1120,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:55:59.058220+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the same 22,034 stars and compute $R_{\\rm birth}$ from a forward chemical-evolution model with explicit radial migration and a non-linear, time-dependent metallicity gradient, rather than the linear lookback-time gradient of Eq. (7). If the young Li-rich stars no longer cluster at $R_{\\rm birth} > 10$ kpc, or the 6-4 Gyr gradient no longer turns negative and broken, the three-epoch lithium history is an artifact of the assumed linear gradient. Independently, measure lithium in young stars whose orbital actions indicate outer-disc birth and check whether their $A(\\mathrm{Li})$ exceeds 2.7 dex.","supporting_citations":[],"review_version":1}