{"id":"54782753-7b7f-40e5-8405-bec89b023c6a","arxiv_id":"1908.01788","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"Accretion of metal-poor gas from a circumbinary disc can reproduce the observed depletion patterns of 58 post-AGB stars, but only if initial accretion rates exceed about 3e-7 solar masses per year and disc masses approach 1e-2 solar masses.","lead":"A stellar evolution modelling study tests whether re-accretion of dusty, metal-poor gas from a circumbinary disc explains the missing heavy elements in post-AGB star atmospheres. It finds that high initial accretion rates and massive discs are needed, and that accretion can stretch post-AGB evolution times by a factor of two to five.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed minimum initial accretion rate (~3e-7 Msun/yr) is not robust to the adopted post-AGB wind prescription: Fig. 16 shows that a 10x weaker wind lets a 5e-8 Msun/yr 0.55-Msun model become depleted at 5000-6000 K, so the quantitative central claim is conditional on an uncalibrated wind law.","rationale":"The reader's CONDITIONAL verdict and weakest assumption focus on the universal accreted-gas composition. That is a real limitation and the paper states it clearly. I do not think it is the single most load-bearing point for the quantitative headline, because the dilution needed to reach a given [Zn/Ti] is dominated by the ratio M_conv/Mdot once the gas is strongly depleted; for [Zn/Ti] values below saturation the required accreted mass is only weakly dependent on the exact [Zn/Ti] of the gas. By contrast, the wind sensitivity shown in the paper's own Fig. 16 changes the minimum accretion rate by more than an order of magnitude and does so precisely in the mass and Teff range that sets the ≳3e-7 bound. The adopted wind law is not just an outside-consensus choice; the paper admits the mechanism is not applicable to post-AGB stars and that there is no calibrated post-AGB alternative. Thus the quantitative central claim is underdetermined by an uncalibrated input. The paper deserves credit for the MESA grid, the explicit parameter exploration, and the honest discussion of winds; this keeps the verdict CONDITIONAL rather than a rejection. The concrete check would settle whether the quantitative part survives.","tokens_in":30303,"tokens_out":9452,"duration_ms":110420,"concrete_test":"Re-run the 0.55 and 0.65 Msun grids from Sect. 5.2 with the Schröder-Cuntz wind (Eq. 1) reduced by factors of 3, 10, and 30, and ideally with an alternative physically motivated cool-wind prescription such as no chromospheric wind below log Teff=3.7, keeping all other parameters identical. Then re-derive the minimum Mdot(0) and Md needed for the model tracks to lie on or above the observed Teff-[Zn/Ti] envelope in Figs. 9-10. If the minimum Mdot(0) remains at or above 3e-7 Msun/yr in the 0.55 Msun bin, the headline constraint survives; if it drops to ~1e-7 or below, the abstract's quantitative claim must be weakened to state that high accretion rates are required only under the adopted wind prescription.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's central quantitative claim is that initial accretion rates ≳3e-7 Msun/yr and disc masses ~1e-2 Msun are needed to reproduce depleted post-AGB stars. This lower bound is derived using the Schröder & Cuntz (2005) wind prescription (Eq. 1), which the paper itself notes is a chromospheric-activity wind not observed in post-AGB stars and which Cranmer & Saar (2011) argue cannot describe post-AGB winds (Sect. 4.2.1). The sensitivity test in Fig. 16 directly undercuts the bound: for a 0.55 Msun model at Mdot(0)=5e-8 Msun/yr, reducing the wind by a factor of 10 is sufficient to produce depleted photospheres at 5000-6000 K, i.e. six times below the headline threshold. Because the evolutionary speed that accretion must overcome is set by wind plus nuclear burning, and the cool-wind regime is empirically unconstrained in exactly the Teff range where depleted post-AGB stars sit, the 'needed' accretion rates and disc masses are not uniquely determined by the data. The qualitative mechanism (re-accretion depletes the photosphere) is supported, but the precise quantitative constraints in the abstract and conclusions are conditional on an uncalibrated mass-loss law.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper models the post-AGB and post-RGB evolution of stars that accrete refractory-depleted gas from a circumbinary disc, using MESA tracks for core masses 0.40-0.65 Msun and a viscous disc model for the time-dependent accretion rate. The accreted-gas abundance pattern is calibrated from the most depleted observed post-AGB stars, and the model tracks are compared in the Teff-[Zn/Ti] plane with a sample of 58 Galactic disc-type post-AGB stars whose luminosities are estimated from Gaia DR2 distances. The main claims are that high initial accretion rates (≳3e-7 Msun/yr) and large initial disc masses (~1e-2 Msun) are needed to reproduce depleted post-AGB stars, that accretion can extend the post-AGB evolution timescale by factors of 2-5, that unsaturated plateau patterns should be more common among post-RGB stars, and that post-RGB stars can become depleted at lower effective temperatures than post-AGB stars.","tokens_in":30620,"tokens_out":5157,"duration_ms":59562,"significance":"If the quantitative conclusions hold, this is a useful step in connecting circumbinary discs to the depletion phenomenon and to post-AGB lifetimes and planetary-nebula formation. The paper has genuine strengths: it is the first to my knowledge to treat the time-dependent accretion from a viscous disc self-consistently in detailed MESA post-AGB models; it uses a homogeneous 58-star observational sample; it compares tracks in the observable Teff-[Zn/Ti] plane; and it includes explicit sensitivity tests for mixing depth (Fig. 15) and wind strength (Fig. 16). The qualitative mechanism, re-accretion of dust-free gas producing photospheric depletion, is supported by the models. The quantitative headline constraints, however, are conditional on two assumptions that the paper itself identifies as uncertain: the universal adopted composition of the accreted gas, and the adopted cool-wind mass-loss law. Because both assumptions affect the inferred accretion-rate and disc-mass thresholds, the significance of the stated constraints is lower than the abstract implies.","major_comments":[{"comment":"","section":"Sect. 3.2 and Fig. 2; Sect. 4.3"},{"comment":"","section":"Sect. 4.2.1, Eq. (1), and Fig. 16"},{"comment":"","section":"Sect. 4.1.1-4.1.2 and Tables A.1-A.2"}],"minor_comments":[{"comment":"","section":"Tables A.2 and A.1"},{"comment":"","section":"Fig. 16 caption and Sect. 4.2.1"},{"comment":"","section":"Sect. 4.2.2"},{"comment":"","section":"Sect. 4.2 and marketplace statement"}],"recommendation":"major_revision","confidential_remarks":"The paper is honest about its main uncertainties and already contains the sensitivity tests that expose the two biggest limitations, which is a credit to the authors. My view is that the qualitative mechanism is sound and publishable, but the abstract and conclusions should be reworded so that the quantitative thresholds do not outrun the calibration of the accretion composition and the wind law. I would not request new physics beyond the sensitivity analysis and reframing. The stress-test concern about the wind prescription lands: Fig. 16 directly undercuts the 3e-7 Msun/yr threshold as a robust number."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nHere's my read of Oomen et al. The qualitative mechanism is probably right; the quantitative headline is not as firm as the abstract suggests.\n\nThe genuinely new piece is the first set of MESA post-AGB/post-RGB models that include time-dependent accretion from a viscously evolving circumbinary disc, with a nuclear network extended to cover the refractory elements that trace depletion. The parameter study is thorough: five core masses, five initial accretion rates, three disc masses, four accretion start temperatures, plus explicit sensitivity runs for mixing depth and wind strength. Comparing the models to a 58-star observed sample with Gaia-based luminosities is a real step forward. The result that accretion can stretch post-AGB evolution timescales by factors of 2-5 is interesting and likely robust in direction.\n\nThe soft spots are in the quantitative claims. The abstract states that initial accretion rates >=3e-7 Msun/yr and disc masses ~1e-2 Msun are needed. But the wind law used for cool post-AGB stars (Schroder & Cuntz) is not calibrated for these objects; the authors themselves note that it is a chromospheric-activity wind and that Cranmer & Saar find it unsuitable. Their own Fig. 16 shows that a 10x weaker wind lets a 5e-8 Msun/yr model become depleted at 5000-6000 K. That is a factor of 6 below the headline threshold. Since the boundary between depleted and non-depleted stars at low Teff is what constrains the accretion rate, the lower bound is conditional on an unknown wind law. The authors flag this in the discussion, but the abstract and conclusions do not carry the caveat.\n\nSecond, the composition of the accreted gas is calibrated from the most depleted observed stars (Sect. 3.2, Fig. 2). The models then reproduce the saturated abundance patterns by construction. They do acknowledge the diversity in turn-off temperatures and that one composition cannot fit all stars, and they present the models as upper limits on depletion. Still, the phrase 'successfully accounts for the chemical peculiarity' overstates the independence of the check.\n\nMinor issues: no hash or direct link to the MESA inlists (only a promise), and the model-observation comparisons are visual, without propagated errors. These are easy to fix.\n\nOverall, this is a serious piece of work for the post-AGB/binary/PN community. It deserves peer review. The authors should be asked to either calibrate the cool wind or explicitly present the accretion-rate constraints as functions of wind strength, and to soften the abstract accordingly. The circularity in the gas composition should be made more prominent. With those revisions, the paper would be a solid contribution.","headline":"Solid, useful MESA study of depletion by re-accretion, but the headline accretion-rate constraint is a factor of ~6 too strong once the uncalibrated wind law is varied, so the abstract oversells the lower bound.","tokens_in":31172,"tokens_out":5672,"would_cite":true,"duration_ms":50916,"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":"Re-accretion of metal-poor gas from a circumbinary disc can reproduce the observed depletion of post-AGB stars, provided the disc begins with a mass near $10^{-2}\\,M_\\odot$ and initial accretion rates of at least about…","keywords":["post-AGB stars","circumbinary discs","chemical depletion","refractory element abundances","gas accretion","post-RGB stars","stellar evolution models","abundance patterns"],"falsifier":"A decisive test would be to measure the disc mass of a strongly depleted star with $T_\\mathrm{eff}\\lesssim5000$ K and luminosity above $7500\\,L_\\odot$: the model requires $\\dot{M}(0)\\gtrsim3\\times10^{-7}\\,M_\\odot\\,\\mathrm{yr}^{-1}$ and $M_d\\sim10^{-2}\\,M_\\odot$ for such an object, so a disc found below $10^{-3}\\,M_\\odot$ would falsify the required parameter range. A second check is to compare turn-off temperatures of depletion patterns with independently measured orbital separations; the model's single accreted composition predicts they should not vary, while observed variation would show the input-composition assumption is wrong.","tokens_in":30062,"feed_emoji":"⭐","tokens_out":8798,"duration_ms":83082,"temperature":0.7,"pith_summary":"Many binary post-AGB stars show photospheric underabundances of elements that condense into dust at high temperatures, a pattern called depletion. The paper argues that this happens because metal-poor gas re-accretes from a circumbinary disc while dust is pushed away, and it tests the idea with stellar-evolution models that add such gas to stars and compare them with 58 observed disc-type post-AGB stars. To match the observed depletion, especially in cool stars, the models require initial accretion rates of at least about $3\\times10^{-7}\\,M_\\odot\\,\\mathrm{yr}^{-1}$ and initial disc masses near $10^{-2}\\,M_\\odot$. At these rates accretion also lengthens post-AGB evolution by a factor of two to five, and the models reproduce the observed mix of saturated and plateau-shaped depletion patterns, with plateau patterns expected mainly among slowly evolving post-RGB stars.","feed_headline":"Depleted post-AGB stars trace gas falling back from a disc","feed_subtitle":"Matching 58 observed stars needs initial accretion above 3e-7 solar masses per year and disc masses near 1e-2 solar masses.","key_machinery":"The load-bearing mechanism is gas dilution of the outer convective envelope by a time-dependent accretion flow. The disc drains viscously with rate $\\dot{M}(t)=\\dot{M}(0)(1+4\\dot{M}(0)t/M_d)^{-3/2}$, where $\\dot{M}(0)$ is the initial accretion rate and $M_d$ the initial disc mass; half the inflowing gas goes to the post-AGB star and half to the companion. The accreted gas has refractory-element abundances suppressed by up to $-4$ dex relative to the star's initial composition (calibrated from the most depleted observed objects), and as it mixes into the shrinking convective envelope it drives down photospheric $[\\mathrm{Zn/Ti}]$ at a rate set by the ratio of the depletion timescale to the evolutionary timescale. The comparison is made in the $T_\\mathrm{eff}$ versus $[\\mathrm{Zn/Ti}]$ plane, where an observed star must lie on or below a model curve if some choice of accreted composition is to explain it.","core_discovery":"The paper's central claim is that the chemical peculiarity called depletion in disc-type post-AGB stars is produced by re-accretion of metal-poor gas from a circumbinary disc, and that this mechanism works without fine tuning only when the disc starts massive and feeds the binary fast. Using models that add a refractory-element-poor gas to evolving post-AGB envelopes and comparing them with 58 observed stars, the paper finds that initial accretion rates below about $5\\times10^{-8}\\,M_\\odot\\,\\mathrm{yr}^{-1}$ rarely produce depleted photospheres before the star becomes hot, while initial rates of $\\gtrsim3\\times10^{-7}\\,M_\\odot\\,\\mathrm{yr}^{-1}$ together with disc masses near $10^{-2}\\,M_\\odot$ reproduce the observed depleted stars, including cool, mildly depleted post-RGB objects. The same accretion delays the stellar evolution track, extending the post-AGB phase by a factor of two to five. The paper also argues that the diversity of depletion patterns follows from how far dilution has proceeded: plateau patterns are partially diluted envelopes, saturated patterns have converged to the composition of the accreted gas, and the turn-off temperature itself reflects the composition of the gas supplied by the disc.","pith_inferences":["If the required disc masses are correct, then non-depleted disc stars are probably not missing the accretion mechanism but started with smaller discs or began accreting later; the paper's own dilution logic implies their accreted gas may also be less depleted.","A testable consequence the paper leaves implicit is that the observed spread in turn-off temperatures (800 to 1500 K) should correlate with orbital separation, because the inner gas-disc radius is set by binary properties while dust sublimation is set by stellar luminosity; interferometric disc radii could check this.","Because accretion extends post-AGB lifetimes by factors of two to five, counts of post-AGB stars and planetary nebulae may imply lower birth rates than previously estimated, and population synthesis using these lifetimes would revise inferred binary-interaction rates.","A direct extension would apply the same accretion prescription to post-RGB binaries with measured orbital periods, since the paper predicts lower accretion rates suffice there; a sample with pulsation-based luminosities could test the luminosity-bin classification."],"forward_implications":["Observed depletion becomes a signpost of a massive infant disc: stars that are depleted at low effective temperature must have started with $M_d\\sim10^{-2}\\,M_\\odot$ and $\\dot{M}(0)\\gtrsim3\\times10^{-7}\\,M_\\odot\\,\\mathrm{yr}^{-1}$.","Post-AGB lifetimes are not set by nuclear burning and winds alone; accretion can stretch them by a factor of two for low-mass post-RGB stars and up to five for more massive post-AGB stars, so planetary-nebula formation and dispersal calculations should include this effect.","Saturated abundance patterns should be common because massive post-AGB stars dilute their envelopes quickly; plateau patterns should be rarer and preferentially found in the slower-evolving post-RGB systems.","Post-RGB stars can become depleted below about 5000 K, while $0.65\\,M_\\odot$ post-AGB stars only deplete above roughly 6000 K, cleanly separating the two populations in the $T_\\mathrm{eff}$ versus $[\\mathrm{Zn/Ti}]$ plane."],"supporting_citations":[{"why":"Supplies the viscous circumbinary-disc evolution law whose time-dependent accretion rate is adopted in Eq. 5.","marker":"Rafikov 2016b"},{"why":"Proposes the mechanism that gas accretes while dust is radiation-pressure-repelled, the basis of the depletion scenario.","marker":"Waters et al. 1992"},{"why":"Shows only hot stars in wide orbits are depleted, used to argue that low-disc-mass models starting at high effective temperature do not match the sample.","marker":"Oomen et al. 2018"},{"why":"Earlier suggestion that accretion can extend post-AGB evolution, quantified here with detailed models.","marker":"Izzard & Jermyn 2018"},{"why":"One of the most depleted stars, IRAS 11472-0800, sets the assumed accreted-gas composition in Sect. 3.2.","marker":"Van Winckel et al. 2012"},{"why":"Hydrodynamical binary simulations showing roughly half the gas entering the cavity goes to each star, giving the factor two in the disc drain rate.","marker":"Farris et al. 2014"},{"why":"Provides the hot-wind mass-loss prescription and the post-AGB timescale context used in the models.","marker":"Miller Bertolami 2016"}],"fun_headline_variants":["Re-accretion from a disc explains depleted post-AGB stars","Gas falling back from a disc depletes post-AGB stars","Massive disc accretion drives depletion in post-AGB binaries","Re-accreting disc gas depletes post-AGB photospheres","Disc infall slows post-AGB evolution and drains metals"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The models assume one universal chemical composition for the accreted gas, fixed by the most depleted stars in the sample; if the real accreted gas is less metal-poor or differs from star to star, the predicted depletion curves, minimum accretion rates, and required disc masses all shift.","fun_headline_variants_meta":{"raw":{"variants":["Re-accretion from a disc explains depleted post-AGB stars","Gas falling back from a disc depletes post-AGB stars","Massive disc accretion drives depletion in post-AGB binaries","Re-accreting disc gas depletes post-AGB photospheres","Disc infall slows post-AGB evolution and drains metals"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000568,"raw_usage":{"total_tokens":2793,"prompt_tokens":1150,"completion_tokens":1643,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":766,"completion_tokens_details":{"reasoning_tokens":1550}},"tokens_in":766,"tokens_out":1643,"duration_ms":12330,"temperature":1.0,"reasoning_tokens":1550,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:04:07.028462+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test would be to measure the disc mass of a strongly depleted star with $T_\\mathrm{eff}\\lesssim5000$ K and luminosity above $7500\\,L_\\odot$: the model requires $\\dot{M}(0)\\gtrsim3\\times10^{-7}\\,M_\\odot\\,\\mathrm{yr}^{-1}$ and $M_d\\sim10^{-2}\\,M_\\odot$ for such an object, so a disc found below $10^{-3}\\,M_\\odot$ would falsify the required parameter range. A second check is to compare turn-off temperatures of depletion patterns with independently measured orbital separations; the model's single accreted composition predicts they should not vary, while observed variation would show the input-composition assumption is wrong.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proposes the mechanism that gas accretes while dust is radiation-pressure-repelled, the basis of the depletion scenario."},{"cited_title":"D., Duffell, P., MacFadyen, A","cited_arxiv_id":null,"evidence_quote":"Hydrodynamical binary simulations showing roughly half the gas entering the cavity goes to each star, giving the factor two in the disc drain rate."}],"review_version":1}