{"id":"a6a2d38e-080b-44e7-93d7-688302645f25","arxiv_id":"2412.14743","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Age-resolved maps of Milky Way disk stars show the thin disk thickens with stellar age and flares at large radius, while the thick disk's scale height is nearly age-independent.","lead":"Using 138,667 red clump stars from the LAMOST and Gaia surveys, the authors split the Milky Way's disk into five age groups and measured how thick and how long each group's stellar component is. They find the thin disk thickens with age and flares outward, while the thick disk's height stays nearly constant with age, which constrains how the Galaxy formed and evolved.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The flaring claim hinges on a selection-function correction whose completeness assumptions are weakest in the outer-disk, high-|Z| bins where the flaring is reported largest; an end-to-end mock recovery test is needed before the trend can be trusted.","rationale":"Among the several correctable issues in the manuscript, including the ambiguous 9,176 versus 138,667 sample statement, the hand-set Rpeak = 8 kpc in Section 4.2, and the BIC values in Table 2 that sometimes prefer a single exponential at R > 12 kpc, the selection-function correction is the most load-bearing because it is the step that turns raw counts into the density maps used by every downstream fit. If this correction is biased, the vertical scale heights, flaring slopes, and radial surface densities are all affected. The paper does include a mock test for age uncertainties in Appendix A, which is genuine supporting evidence for the age-ordering of hZ1, but no equivalent test is provided for the selection function. The reader's diagnosis matches the weakest point: the Castro-Ginard et al. completeness assumptions are least secure at faint magnitudes and in crowded or low-latitude fields, which is precisely the regime of the outer-disk, high-|Z| bins where the largest flaring is reported. The quoted uncertainties are MCMC-only and do not include this systematic. I therefore agree with the CONDITIONAL verdict rather than escalating: the inner-disk structure and the age trend of hZ1 are plausible and consistent with prior work, but the outer-disk flaring claim should not be accepted until the selection function is validated end-to-end.","tokens_in":16559,"tokens_out":14238,"duration_ms":125003,"concrete_test":"Build an end-to-end mock: draw a synthetic red-clump population from a two-component disk with chosen hZ1(R) and hZ2(R), including one no-flaring input and one flaring input matching the claimed trend. Place the mock through the real LAMOST DR8 targeting footprint with realistic G-magnitude and color limits plus a Gaia DR3 completeness mask, then apply the same Castro-Ginard et al. selection-function correction and the same MCMC vertical-profile fitting used in Section 4.1. If the recovered hZ1(R) and hZ2(R) match the input flaring slopes, and recover flat trends for a flat input, the selection-function concern is resolved; if the pipeline produces flaring from a flat input, the central flaring claim is likely contaminated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central flaring result depends on the sample-selection correction used to build the density maps. Section 3.2 adopts the Castro-Ginard et al. (2023) method, which estimates the LAMOST selection function by comparing the observed subsample to a Gaia DR3 parent in cells of HEALPix, G, and G-GRP. This is unbiased only if the Gaia parent is complete in every cell used. The outer-disk bins that carry the flaring claim (R > 10 kpc, |Z| > 1-2 kpc) contain faint RC stars at photometric distances of 12-16 kpc, where the LAMOST faint limit and, in crowded or low-latitude cells, Gaia DR3 incompleteness become important. The RC classification completeness of 90-95% from Wang et al. (2023) is not propagated either; if it depends on magnitude or color it acts as an unmodeled selection term. These corrected counts enter Eq. 3 as N(di), so residual incompleteness at high |Z| propagates directly into the fitted hZ1 and hZ2. The quoted error bars are MCMC fitting uncertainties only and do not include the selection-function systematic. The reported increase of hZ with radius is largest exactly where the correction weights are largest and least constrained, so the flaring could be an artifact of the correction rather than a property of the disk.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes a sample of 138,667 primary red clump stars from LAMOST DR8 and Gaia DR3, dividing them into five mono-age bins, constructing stellar number density maps in the R-Z plane, and fitting vertical profiles with a double-exponential disk model (Eq. 5) and radial surface density profiles with a broken exponential (Eq. 7). The central results are that both components of the vertical density profile flare in the outer disk; that the first (thin-disk-like) component's scale height increases with age at fixed radius while the second (thick-disk-like) component's scale height is roughly age-independent; and that the radial surface density profiles of both components peak near 7.5-8.5 kpc and decline outward. The paper also compares the mono-age results to earlier mono-abundance studies and discusses implications for radial migration and disk evolution.","tokens_in":16777,"tokens_out":7275,"duration_ms":59275,"significance":"If the results hold, the paper provides an age-resolved structural map of the Milky Way disk, with quantitative scale heights, flaring rates, and scale lengths for mono-age populations that can constrain models of disk heating, radial migration, and flaring mechanisms. The analysis uses a large, carefully classified RC sample from Wang et al. (2023), a modern selection-function correction via GaiaUnlimited (Castro-Ginard et al. 2023), and includes a mock test in Appendix A demonstrating that age uncertainties do not erase the age trends. These are genuine strengths. However, the strength of the central claims is limited by unpropagated systematic errors in the distance calibration and selection function, and by the fact that the radial break radius is assumed rather than measured.","major_comments":[{"comment":"The text reports 'The resultant dataset comprised 9176 RCs' after applying the selection criteria in Section 3.1, whereas the abstract and conclusion state that the analysis uses a sample of 138,667 primary red clump stars. The relationship between the 9176-star subsample and the 138,667-star main sample is never explained. This is load-bearing: it is unclear whether the density maps in Section 3.3 are constructed from 138,667 stars or from 9,176 stars, and whether the distance-calibration relation of Eq. (1) is derived from the same sample used for the structure fits. Please clarify explicitly that 9176 is the high-latitude, low-reddening calibration subsample used to derive Eq. (1), and that the subsequent analysis uses the full 138,667-star sample, or correct the numbers so they are consistent throughout.","section":"3.1"},{"comment":"The radial profile fitting does not actually measure the peak radius. The text of Section 4.2 says the fitting is restricted to R > Rpeak, and the caption of Fig. 6 states 'Rpeak (set to 8 kpc)'. Thus the break radius is assumed, not fitted. The abstract, Section 5.2, and the conclusion nevertheless claim that the radial surface density profiles 'predominantly peak within a radial range of 7.5-8.5 kpc.' That range is not a result of any fit presented in the paper; it is an input assumption. Please either fit Rpeak as a free parameter (even with a coarse grid, given the limited number of inner bins) or revise the claims to state explicitly that Rpeak was fixed to 8 kpc and that the data are consistent with a peak in that region without having measured it.","section":"4.2"},{"comment":"The selection-function correction adopted from Castro-Ginard et al. (2023) assumes that the Gaia DR3 parent catalog is complete in every HEALPix, G-band, and color cell used. In the outer-disk, high-|Z| bins where the flaring signal is strongest (approximately R > 10 kpc, |Z| > 1-2 kpc, see Fig. 3 and Table 2), the RC stars are at photometric distances of 12-16 kpc and are near both the Gaia magnitude limit and the LAMOST faint limit, so residual incompleteness is likely largest exactly where the correction weights are largest. The quoted uncertainties in Table 2 are MCMC statistical errors only and do not include this systematic. Please add an end-to-end mock recovery test that injects a known stellar density distribution into the selection-function framework, applies the full pipeline (including the Gaia/LAMOST completeness limits), and checks whether the input flaring trends are recovered; or, failing that, provide a quantitative estimate of the residual selection bias in the outer-disk bins and propagate it into hZ1 and hZ2.","section":"3.2"},{"comment":"The distance calibration carries a systematic uncertainty that is not propagated into the structural results. Section 3.1 mentions a 3-5% systematic uncertainty in distance, and Eq. (1) is a polynomial fit whose coefficients have uncertainties; additionally, the reddening correction via Eq. (2) and the adopted extinction ratios introduce further systematics. These uncertainties affect the density maps through Eq. (3) and hence every fitted scale height and scale length. None of these are included in the error bars of Table 2 or Table 3. Please propagate the distance systematics, for example by re-running the entire fitting procedure with distances shifted by ±3-5% and by ±1-sigma variations of the Eq. (1) coefficients, and report the resulting systematic contributions to hZ1, hZ2, and hR.","section":"3.1"},{"comment":"The paper's claim that the vertical profiles are 'best described by a dual-component disk model' is not supported in several bins. For example, in the 7-9 Gyr population at 11-12 kpc the BIC for the double-exponential model is +3.431 while that for the single-exponential is -51.112, and the 12-14 kpc bins for the 0-3 and 3-5 Gyr populations also slightly favor the single-exponential model. The text acknowledges this but still draws the universal two-component conclusion. Please quantify the number of radial bins (out of the 35 total) in which the double-exponential is preferred by, say, ΔBIC > 10, and discuss whether the exceptional bins indicate a real breakdown of the two-component description in the outer disk or simply a loss of statistical power due to sparse counts.","section":"Table 2"}],"minor_comments":[{"comment":"The title contains a typo: 'Rrevealed' should be 'Revealed'.","section":"Title"},{"comment":"The author name 'Chun W ang' appears with an erroneous space; this is likely a typesetting issue that should be corrected.","section":"Author list"},{"comment":"Equation (8) writes Σ(R) = 2ρ1 hZ(R), but for the double-exponential model the vertical integral is 2ρ1(hZ1 + f hZ2). Please define the effective hZ(R) used in the radial surface density computation, or write the integrated form explicitly, so that the radial fitting in Section 4.2 is unambiguous.","section":"Eq. (8)"},{"comment":"The MCMC description gives 500 walkers and 40,000 steps but does not state the burn-in length or convergence criteria; please add these details so that the fit quality and the effective number of independent samples can be assessed.","section":"4.1"},{"comment":"The mock test in Appendix A only investigates the effect of age uncertainties; it does not include distance errors, selection-function residuals, or reddening systematics. Please state this limitation explicitly in the appendix, since the text currently says the methodology is 'robust' without qualifying the scope of the test.","section":"Appendix A"},{"comment":"The statement that hZ1 'typically reduces up to a radius of R = 7.5 kpc' is based on 1-kpc wide radial bins, so the minimum near R=7.5 kpc is a bin-scale feature rather than a resolved measurement; please phrase the description accordingly.","section":"5.1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of the journal and addresses a topic of current interest, but the central claims need additional support. The sample-size confusion and the fixed-Rpeak issue can be fixed with clarifications, but the selection-function systematic is a deeper problem: the flaring result is strongest in the exact region where the correction is most uncertain, and the paper provides no end-to-end test. I would advise the editor to request a major revision in which the authors either add a realistic mock recovery test for the selection function or substantially soften the flaring claims, and to propagate distance systematics. The novelty relative to Yu et al. (2021) and Lian et al. (2022) is moderate, but the mono-age decomposition is a useful addition if the systematics are addressed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a competent, incremental paper that applies mono-age decomposition to a large LAMOST RC sample and recovers the expected story—two-component vertical structure, age-dependent thin-disk scale height, flaring in the outer disk. The main results are probably right, but the paper presents the radial break as measured when it is actually fixed, and it does not fully address the selection-function systematic that matters most for the flaring claim. It deserves referee time, but it needs revision.\n\nThe genuinely useful pieces: the sample is the largest mono-age RC sample used for this kind of structural fit (138,667 stars), and the Appendix A mock test on age uncertainties is a good check that bin mixing does not drive the age trends. The finding that hZ2 (thick-disk component) is roughly age-independent while hZ1 grows with age is a clean, interesting result, consistent with prior work but worth having from a larger sample. The reference list is appropriate and the paper engages honestly with earlier results.\n\nThe soft spots. First, the abstract and discussion say the radial density profiles peak at 7.5–8.5 kpc, but the fitting in §4.2 sets Rpeak = 8 kpc and only fits R > Rpeak. The break radius is not actually fitted, so the 'measured' peak range is not supported by the analysis. That is an overstatement and easy to fix by either fitting Rpeak as a free parameter or reframing the claim.\n\nSecond, the flaring result depends on the selection-function correction from Gaia DR3, and the correction is least reliable in the outer-disk, high-|Z| bins where flaring is largest. The paper gives MCMC-only error bars and doesn't propagate distance-calibration or selection-function systematics. The stress-test concern is fair: an end-to-end mock test that injects a known density model and recovers it through the selection-function correction would tell us how much of the outer-disk flaring is real. The paper has a mock test for age errors but not for selection.\n\nThird, the BIC table actually favors a single exponential in several of the outermost, oldest bins—the paper notes this, but still uses the double-exponential parameters there. That's a minor issue, but it undercuts the claim that both components flare at the largest radii.\n\nThe 9,176 vs 138,667 numbers: the 9,176 is the calibration subset, but the text doesn't spell that out. That's a clarity problem, not a scientific one.\n\nBottom line: the qualitative conclusions are consistent with previous work (Yu et al. 2021; Xiang et al. 2024), so this is an incremental confirmation with a larger sample, not a breakthrough. But it is competently done, honest about some limitations, and the flaws are fixable. I'd send it to peer review; a good referee will push for a real fit of the break radius, systematic error bars, and a selection-function robustness test.","headline":"Solid incremental mono-age structural study of the Milky Way disk with a large LAMOST RC sample; the main trends are plausible, but the radial break is fixed rather than measured and selection-function systematics are not fully addressed.","tokens_in":17447,"tokens_out":5948,"would_cite":false,"duration_ms":45290,"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 Milky Way's disk has two vertical components: the thin one thickens with stellar age, the thick one stays constant, and both flare outward.","keywords":["Galaxy disk structure","red clump stars","mono-age populations","disk flaring","scale height","scale length","LAMOST","Gaia DR3"],"falsifier":"Take a photometrically complete sample of red clump stars selected without spectroscopic targeting, measure the same vertical density profiles in the outer disk beyond 10 kiloparsecs and more than 2 kiloparsecs above the plane, and check whether $h_{Z1}$ and $h_{Z2}$ still rise with radius; if the flaring disappears, the trend was produced by missing faint stars, not by the disk.","tokens_in":1721,"feed_emoji":"🌌","tokens_out":2722,"duration_ms":67028,"temperature":0.7,"pith_summary":"This paper maps the Milky Way's stellar disk in age slices, using 138,667 primary red clump stars from LAMOST and Gaia. It claims that the vertical distribution of each mono-age population is best described by two exponential components: a thin disk whose scale height grows with stellar age at every radius, and a thick disk whose scale height stays nearly the same across all ages. Both components flare outward beyond roughly 8 kpc. If this age-resolved structure is real, it provides a direct view of how the disk has heated and thickened over cosmic time and imposes constraints on formation mechanisms such as radial migration. The paper also finds that the radial surface density of both components peaks near the solar circle, between 7.5 and 8.5 kpc.","feed_headline":"Galactic disk's thin component thickens with age","feed_subtitle":"Mono-age red clump stars show the thin disk flares and grows with age while the thick disk stays constant.","key_machinery":"The paper's central tool is the double-exponential vertical density model, $\\rho = \\rho_1[\\exp(-|Z-Z_0|/h_{Z1}) + f\\exp(-|Z-Z_0|/h_{Z2})]$, fit with an MCMC sampler to radial bins across five mono-age populations. The two scale heights, $h_{Z1}$ and $h_{Z2}$, separate the morphologically thin and thick disks, and their dependence on radius and age carries the entire argument: a fit is judged by whether $h_{Z1}$ tracks age and whether both scale heights rise with radius. The radial surface-density model is a broken exponential with break radius $R_{\\mathrm{peak}}$, which lets the paper claim a common peak near 7.5-8.5 kpc. Selection effects are handled by a Bayesian per-star weight derived from the Gaia DR3 parent catalog.","core_discovery":"The central claim is that the vertical density profile of each mono-age red clump population follows a double exponential rather than a single one, with a compact Component 1 ($h_{Z1}$ roughly 0.15-0.5 kpc) and a diffuse Component 2 ($h_{Z2}$ roughly 0.8-2.5 kpc). At fixed Galactocentric radius, $h_{Z1}$ rises with age across the five age bins, while $h_{Z2}$ is essentially the same for all age groups. Both scale heights increase with radius beyond the solar neighborhood, so the disk flares, and the flaring of the first component is somewhat stronger in older populations. The radial surface density profiles follow a broken exponential that peaks at $R_{\\mathrm{peak}}$ between 7.5 and 8.5 kpc for both components, with the morphological thick disk having a larger scale length than the thin disk. The paper interprets this as evidence that the thin disk has been continually heated and has undergone radial migration, while the thick disk formed early and remained structurally stable.","pith_inferences":["Because age uncertainties blur adjacent bins, the reported rise of $h_{Z1}$ with age is likely a lower bound on the true heating trend; a cleaner measurement with asteroseismic ages should steepen it.","The near-solar break at 7.5-8.5 kpc coincides with the outer Lindblad resonance of the Galactic bar; testing whether the break radius moves with the bar pattern speed would directly connect disk structure to the bar potential.","If the flaring of Component 2 is truly age-invariant, then the thick disk's flaring was set at birth or by a single early merger event; looking for the flare amplitude to be constant across the oldest mono-age bins in independent surveys would discriminate between secular heating and a one-time event."],"forward_implications":["If the age-scale-height trend holds, the thin disk has been heating continuously over its lifetime, so present-day thin-disk stars of different ages must have different vertical velocity dispersions at the same radius.","The age-independent thick disk scale height implies the thick disk's structure was set early, so chemodynamical models should not let the thick disk evolve much after the first few billion years.","The common break at 7.5-8.5 kpc in both components points to a global structural transition near the solar radius, which any Galactic disk model must reproduce.","The larger scale length of the morphological thick disk compared with the thin disk, if physical, conflicts with chemical-based definitions that put the thick disk shorter; reconciling the two will require treating age and chemistry separately."],"supporting_citations":[{"why":"Supplies the red clump sample with ages, masses, and the RGB/RC classification used to build mono-age populations.","marker":"Wang et al. 2023"},{"why":"Provides the Bayesian method to estimate the LAMOST selection function against the complete Gaia DR3 catalog.","marker":"Castro-Ginard et al. 2023"},{"why":"Provides the Ks-band absolute magnitude calibration and selection criteria used to compute red clump distances.","marker":"Huang et al. 2020"},{"why":"Established the double-exponential vertical fit and broken-exponential radial fit that this paper extends to mono-age populations.","marker":"Bovy et al. 2016"},{"why":"Earlier LAMOST red clump analysis whose flaring trends and scale-height minimum near the solar radius this paper corroborates with a different selection correction.","marker":"Yu et al. 2021"},{"why":"Reported flaring trends and discussed age-uncertainty mixing, serving as the comparison baseline for the age-dependent flaring result.","marker":"Ted Mackereth et al. 2017"}],"fun_headline_variants":["Disk's thin component grows thicker with star age","Aging stars reveal Galactic disk's thin layer flaring","Old stars show thin Galactic disk puffs up over time","Thin disk thickens, thick disk stays put as stars age","LAMOST red clump stars reveal age-dependent disk flaring"],"cache_read_input_tokens":19328,"weakest_assumption_plain":"The analysis assumes that the correction for LAMOST's incomplete sky targeting, derived from the complete Gaia catalog, leaves no remaining bias, particularly in the outer disk and at large heights where faint red clump stars fall below detection limits.","fun_headline_variants_meta":{"raw":{"variants":["Disk's thin component grows thicker with star age","Aging stars reveal Galactic disk's thin layer flaring","Old stars show thin Galactic disk puffs up over time","Thin disk thickens, thick disk stays put as stars age","LAMOST red clump stars reveal age-dependent disk flaring"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000522,"raw_usage":{"total_tokens":2547,"prompt_tokens":989,"completion_tokens":1558,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":605,"completion_tokens_details":{"reasoning_tokens":1475}},"tokens_in":605,"tokens_out":1558,"duration_ms":8511,"temperature":1.0,"reasoning_tokens":1475,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:57:31.498201+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a photometrically complete sample of red clump stars selected without spectroscopic targeting, measure the same vertical density profiles in the outer disk beyond 10 kiloparsecs and more than 2 kiloparsecs above the plane, and check whether $h_{Z1}$ and $h_{Z2}$ still rise with radius; if the flaring disappears, the trend was produced by missing faint stars, not by the disk.","supporting_citations":[{"cited_title":"G., Kostrzewa-Rutkowska, Z., et al","cited_arxiv_id":null,"evidence_quote":"Provides the Bayesian method to estimate the LAMOST selection function against the complete Gaia DR3 catalog."},{"cited_title":"2020, The Astrophysical Journal Supplement Series, 249, 29 15 Juri´ c, M., Ivezi´ c,ˇZ., Brooks, A., et al","cited_arxiv_id":null,"evidence_quote":"Provides the Ks-band absolute magnitude calibration and selection criteria used to compute red clump distances."},{"cited_title":"F., et al","cited_arxiv_id":null,"evidence_quote":"Established the double-exponential vertical fit and broken-exponential radial fit that this paper extends to mono-age populations."}],"review_version":1}