{"id":"531385b7-c15f-41f9-b5b4-f8f4894947f0","arxiv_id":"2501.09058","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A uniform reanalysis of 74 resolved millimeter-wave debris belts shows belt dust masses fall with age fastest at small radii, most belts are broad, and vertical thickness shows no age trend.","lead":"Astronomers mapped 74 dusty planetesimal belts around nearby stars with millimeter-wave telescopes, creating the largest uniform sample of their sizes, widths, masses, and thicknesses. The data show small belts lose dust faster, most belts are broad rather than narrow rings, and belt thickness does not increase with age, which constrains how planetary systems evolve.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 2 dex young-old mass offset in Fig. 8 may reflect birth differences rather than depletion, because the model in Sect. 5.3 assumes initial mm-grain mass is independent of stellar and belt properties.","rationale":"The paper's strongest and most novel claim is the evolutionary mass-radius trend; the data release and uniform modelling are robust and valuable. The weakest link is the conversion of a cross-sectional age comparison into a depletion law. The model in Sect. 5.3 contains a degeneracy between M0 and D and a strong assumption that M0 is universal. Since the young and old samples differ in selection and distance, the observed offset could be a birth-property effect. Existing bias maps are built from the observed sample and cannot distinguish between evolution and initial-condition differences. A hierarchical fit with free initial mass distribution and forward-modelled selection is the direct falsifier. This does not impugn the data or the qualitative claims (e.g., belts are broad; vertical aspect ratios), but it means the quantitative epsilon~1 result should be presented as conditional on the assumption. I agree with the reader's weakest_assumption.","tokens_in":49736,"tokens_out":4175,"duration_ms":45808,"concrete_test":"Fit a hierarchical Bayesian model to the REASONS [Mass1.33mm, R, age] data with initial mass M0 drawn from a log-normal hyperdistribution with mean mu0 and scatter sigma0 (allowing a possible L* dependence), with free D, delta, eps, and a forward-modeled selection function that applies the actual SONS and Lieman-Sifry criteria to simulated belts. Then compare the posterior for mu0, sigma0 between the Sco-Cen and field subsamples. If a single mu0, sigma0 can reproduce both age groups after collisional evolution, the evolution claim stands; if the data require mu0 to differ by more than 0.5 dex between groups, the observed offset is not prima facie evidence for depletion. A lighter check would be to evolve the 23 Sco-Cen belts with the best-fit D and eps to 1 Gyr and ask whether their predicted Mass1.33mm at R matches the field star distribution at the same R and L*.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 5.3's evolutionary interpretation treats the ~2 dex vertical offset between young Sco-Cen and old field belts in Fig. 8 as depletion, but the model's isochrones have a built-in degeneracy: Mass1.33mm(t) = M0/[1 + (t-t0)eps/tc] with tc=(D/M0)R^delta is algebraically invariant under simultaneous rescaling of M0 and D along the equilibrium branch. The claim that the offset measures eps=1 via Eq. 4 assumes M0 is identical for Sco-Cen and field stars. This is not merely a detail: the two subsamples are drawn with different selection functions (Lieman-Sifry IR excess threshold for Sco-Cen vs SONS 850 um flux limit for the field) and at different median distances (127 pc vs 24 pc), so an initial-mass offset between the populations would produce exactly the observed trend. The bias maps in Fig. 7 cannot break this degeneracy because they are constructed by resampling the observed N-2 parameter distribution, implicitly assuming unobserved belts share the observed population's properties. A forward model with free initial-mass hyperparameters is needed before the linear-depletion claim can be accepted.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents the REASONS sample: 74 planetesimal belts around nearby stars observed with ALMA and SMA, uniformly modelled in the visibility domain with a Gaussian ring prescription. The main scientific claims are (1) a dearth of small belts, (2) most belts are broad rather than narrow rings, with widths inconsistent with protoplanetary ring widths, (3) vertical aspect ratios implying orbital inclinations of ~1–20 deg with no age trend, and (4) a mass–radius trend in which older field belts are less massive and larger than young Sco-Cen belts, interpreted as radius-dependent collisional evolution with mass depletion roughly linear in time (epsilon ~ 1 in Eq. 4). The paper also releases the full data products and tabulated parameters to the community.","tokens_in":50017,"tokens_out":4085,"duration_ms":45141,"significance":"If the evolutionary interpretation holds, this is the first large resolved sample to show directly that the observable planetesimal belt population is a collisional outcome rather than a set of static birth properties, which would be a major step for debris disc population modelling. The uniform reduction of 74 belts, with public release of calibrated visibilities, images, and fitted parameters, is a substantial community resource. The modelling is careful: the visibility fitting uses MCMC with residual checks, and the authors state key caveats about the Gaussian ring assumption, fixed aspect ratio when unresolved, and the selection-bias construction. The population inferences, however, rest on simplifying assumptions that are partly acknowledged, and the most load-bearing of these—the common initial mass in the collisional evolution model—is not tested against alternatives.","major_comments":[{"comment":"The inference that the young-to-old mass offset measures a linear depletion rate (epsilon ~ 1) assumes Mass_1.33mm(t0), the initial mm-grain belt mass, is identical for Sco-Cen and field stars (stated in the text as 'assumed to be independent of stellar and belt properties'). On the equilibrium branch of the model, Mass_1.33mm(t) reduces to (D R^delta)/((t-t0)^epsilon), which is algebraically invariant under a simultaneous rescaling of M0 and D. The two subsamples differ strongly in median distance (127 pc vs 24 pc) and selection functions (Lieman-Sifry 70 um excess criterion vs SONS 850 um flux limit), so a systematic birth-mass difference between them would produce exactly the vertical offset seen in Fig. 8. The bias maps of Fig. 7 cannot break this degeneracy because they are built by resampling the observed N-2 parameter distribution, implicitly assuming unobserved belts have the same properties as observed ones (Sec. 5.1). A forward model with free initial-mass hyperparameters (or a direct comparison controlling for stellar type and distance) is required before the linear-depletion claim can be accepted.","section":"Sec. 5.3, Eq. (4)"},{"comment":"The numerical value of epsilon depends sensitively on the assumed collisional-age baseline. The text says the offset of ~2 dex in mass occurs over '~2 dex of collisional age (between ~10 Myr and ~1 Gyr)', but the old-field isochrone shown in Fig. 8 is for 5 Gyr, and many field stars in Table A.2 have ages of several Gyr. Using t_young = 15 Myr and t_field = 5 Gyr, the denominator log(t_field - t0) - log(t_young - t0) is about 3 dex, which would give epsilon ~ 2/3 rather than ~1 for a 2 dex mass offset. Because the field-star ages carry large uncertainties (e.g., HD38858 has age 5 +/- 5 Gyr, and several systems list only broad ranges), the paper should either propagate age uncertainties into Eq. (4) or state the epsilon conclusion as sensitive to the adopted age baseline.","section":"Sec. 5.3, Eq. (4)"},{"comment":"The bias maps are constructed by drawing the N-2 non-plotted parameters randomly from the 74 observed belts, so they answer only 'what fraction of belts at this [X,Y] location would have been detected if they had the same distribution of other parameters as the observed population'. This is an acknowledged assumption, but it directly limits the strength of the conclusion drawn in Sec. 5.1 that the dearth of small belts cannot be explained by selection effects: a population in which small belts are intrinsically rarer but have different (e.g., narrower or warmer) properties would not be represented. The qualitative conclusion may survive, but the quantitative claim should be softened or supplemented with a sensitivity test that varies the N-2 distribution.","section":"Sec. 5.1, bias maps"}],"minor_comments":[{"comment":"There is a typo in the text: 'tables available on Z enodo' should read 'Zenodo'.","section":"Sec. 3.1"},{"comment":"In the bullet on aspect ratios, the phrase 'stirred very earlyA diversity' is missing a space; it should read 'stirred very early. A diversity'.","section":"Sec. 6"},{"comment":"The notation '50+34−34th percentile' in the description of Table 1 is confusing; it appears to be a LaTeX artifact and should be written as '50th percentile with 34th and 16th percentiles' or similar.","section":"Sec. 4.2.4"},{"comment":"The statement that relative velocities of 0.1–4 km/s follow from the derived inclinations uses Eq. 10 of Matrà et al. (2019b), but this equation is not reproduced; a brief restatement of the assumed mean eccentricity or the vrel-to-inclination relation would improve reproducibility.","section":"Sec. 5.4"}],"recommendation":"major_revision","confidential_remarks":"The core data products—uniform visibilities, images, and derived parameters for 74 belts—are valuable and appear carefully produced. The evolutionary claim in Sec. 5.3, however, relies on an untested assumption of universal initial mass, and the epsilon measurement is sensitive to the adopted age baseline. The authors already acknowledge the assumption, but the manuscript would be substantially strengthened by a simple forward model that lets initial mass vary between populations, or by a test that restricts the comparison to a common stellar-type subsample. I would not reject the paper on this basis, but the central claim should be reframed as provisional until such a test is done."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The REASONS paper is the real thing: 74 belts uniformly modelled from visibilities, 15 new ALMA/SMA targets, and all data products and code on Zenodo/GitHub. That part is rock solid. The width distribution (median fractional width 0.71, ~70% broad) and the vertical aspect-ratio sample for 24 belts are new population-level results that will be cited for years. The comparison with protoplanetary ring widths is careful, and the K-S caveat is honestly stated.\n\nThe soft spot is exactly where the stress-test aimed: Section 5.3. The ~2 dex offset between young and old belts in [Mass-R] space is interpreted as evidence for linear, radius-dependent collisional depletion. But the model assumes Mass_1.33mm(t0) is independent of stellar and belt properties, with t0=10 Myr and a single constant D. The isochrone equation has a built-in degeneracy between M0 and D on the equilibrium branch, and the Sco-Cen vs field subsamples differ in selection (Lieman-Sifry 70 um excess vs SONS 850 um flux) and median distance (127 pc vs 24 pc). A birth-mass offset between the populations would produce exactly the observed trend. The bias maps in Fig. 7 are constructed by drawing the N-2 parameters from the observed sample, so they cannot break this degeneracy—they answer \"what fraction of observed-like belts would be detected\", not \"would a different birth population be detectable\". The authors do explicitly state the assumption, but then conclude \"strong evidence\" for linear depletion. That overstates what the data and model can support.\n\nThe aspect ratio result is cleaner: no age trend, consistent with early stirring or a diversity of pathways. The lower limits on stirrer sizes are model-dependent but the data are real. Fixing h=0.03 for unresolved belts is a reasonable default and clearly stated. Excluding HD36546 from the width analysis is honest.\n\nWho is this for? Debris disk theorists and observers planning ALMA follow-up. The data release alone makes it worth publishing. I would send it to a serious referee, with the request that the evolutionary claims be softened to \"consistent with\" or supplemented by a forward model with free initial-mass hyperparameters. The measurement paper is a strong contribution; the evolutionary conclusion is the part that needs revision.","headline":"A landmark data release and careful reanalysis; the collisional evolution claim is plausible but rests on an assumption the bias maps cannot test.","tokens_in":50620,"tokens_out":2524,"would_cite":true,"duration_ms":28371,"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":"The REASONS survey of 74 resolved planetesimal belts finds that belt dust mass is depleted with age, more quickly in smaller belts, matching collisional evolution.","keywords":["planetary systems","submillimeter: planetary systems","circumstellar matter","surveys","techniques: interferometric","collisional evolution","debris discs","exo-Kuiper belts"],"falsifier":"Find, after applying the paper's own selection corrections, an old field belt (age >200 Myr) with a mass and radius in the top-left region of the mass-radius diagram typical of the 10–30 Myr Sco-Cen belts, such as roughly $1\\,M_\\oplus$ of millimetre grains at $R<50$ au; its presence would break the claimed depletion pattern, as would a coeval old population showing the same mass-radius distribution as the young associations.","tokens_in":49561,"feed_emoji":"🪐","tokens_out":8876,"duration_ms":87333,"temperature":0.7,"pith_summary":"Planetesimal belts are rings of debris left over from planet formation, and this paper uses a sample of 74 of them, resolved at millimetre wavelengths with ALMA and the SMA, to ask what happens to that debris over time. The central claim is that the belts are not static: their dust is ground away by collisions, and the depletion is fastest in the smallest belts. The evidence is a uniform visibility-modelling analysis that places young Sco-Cen belts and old field-star belts on a mass-radius diagram, where the old population sits below the young one by about two orders of magnitude in dust mass. That offset aligns with the collisional-evolution locus and implies mass loss roughly proportional to time after the belt reaches collisional equilibrium. If correct, a resolved belt's mass and radius carry information about its age and its collisional history, turning debris discs into evolutionary probes.","feed_headline":"74 resolved belts reveal collisional aging of exo-Kuiper belts","feed_subtitle":"Uniform ALMA/SMA census shows belt dust mass falls roughly linearly with time, fastest at small radii","key_machinery":"The argument is carried by the mass-radius-age diagram built from uniform interferometric visibility modelling: each belt is modelled as an axisymmetric radially Gaussian ring to recover the centroid radius $R$ and fractional width $\\Delta R/R$, and a millimetre dust mass is derived from the 1.33 mm flux density assuming large grains radiate as blackbodies. The evolutionary interpretation rests on collisional isochrones from the analytical model ${\\rm Mass}_{1.33{\\rm mm}}(t)={\\rm Mass}_{1.33{\\rm mm}}(t_0)/(1+[(t-t_0)\\epsilon]/t_c)$ with collisional timescale $t_c=(D/{\\rm Mass}_{1.33{\\rm mm}}(t_0))R^{\\delta}$; taking $\\epsilon=1$ and $\\delta=13/3$ predicts a diagonal locus ${\\rm Mass}_{1.33{\\rm mm}}\\propto R^{13/3}$ for belts that have reached collisional equilibrium. The match of the old field population to this diagonal, together with the young population sitting above it, converts a scatter plot into evidence for radius-dependent collisional depletion.","core_discovery":"In $[{\\rm Mass}_{1.33{\\rm mm}}{-}R]$ space, the REASONS sample shows a clear trend: belts around old field stars are on average significantly less massive and at the same time larger than belts around the youngest moving group, and the older population lies along a diagonal locus consistent with ${\\rm Mass}_{1.33{\\rm mm}}\\propto R^{13/3}$, the expectation for belts in collisional equilibrium. The vertical offset of roughly 2 dex in mass over roughly 2 dex in collisional age implies $\\epsilon\\sim 1$, i.e. mass depletes as $(t-t_0)^{-1}$, consistent with the simple collisional cascade model and disfavouring shallower $t^{-0.4}$ evolution. The same dataset shows that most belts are broad discs rather than narrow rings, with a median fractional width of 0.71 and about 70% of well-measured belts having $\\Delta R/R>0.5$, a distribution inconsistent with protoplanetary ring widths, and that the 24 belts with constrained vertical aspect ratios have heights implying inclinations of 1–20 degrees with no age trend, which the authors interpret as early stirring by bodies of at least about 140 km in size.","pith_inferences":["One testable extension of the $t^{-1}$ result: if the initial-mass assumption holds, belt masses from resolved mm surveys could serve as rough age indicators for old systems, though calibrating the zero-point would require coeval populations at several ages.","A direct way to separate natal from evolutionary origin is to map mass-radius in several young associations at 10–30 Myr; if initial mass anticorrelates with radius within a single age cohort, part of the old-young offset is set at birth rather than by collisions.","The vertical-structure result predicts that gas-bearing debris discs should not show vertically settled mm grains; measuring the aspect ratio in a larger sample of CO-rich belts would test whether gas densities there are truly too low to affect mm-grain dynamics.","If broad belts hide unresolved gaps, then higher-resolution imaging should resolve a substantial fraction of the wide belts into two or more narrow rings, softening the reported width contrast with protoplanetary discs."],"forward_implications":["Small, massive belts are rare around old stars because they deplete fastest, not because they never formed.","Selection-corrected mass-radius diagrams become a standard diagnostic for debris-disc evolution.","The majority of resolved belts are broad rather than narrow rings, so narrow rings like Fomalhaut's are not representative of the detectable belt population.","Vertical aspect ratios imply stirring by bodies at least about 140 km in size happens within roughly 10–50 Myr, so heights trace excitation state rather than age.","For belts in collisional equilibrium, the expected ${\\rm Mass}_{1.33{\\rm mm}}\\propto R^{13/3}$ locus gives a direct way to identify which belts are still primordial versus collisionally processed."],"supporting_citations":[{"why":"Supplies the analytical collisional evolution model whose isochrones, with epsilon=1 and delta=13/3, define the expected mass-radius-age locus and the t^{-1} depletion rate.","marker":"Wyatt et al. (2007a)"},{"why":"Defines the SONS 850 micron flux-limited sample and the detectability threshold that shapes the REASONS selection function.","marker":"Holland et al. (2017)"},{"why":"Supplies the ALMA-resolved Sco-Cen belts and the 70 micron excess selection that sets the young sample's detectability floor.","marker":"Lieman-Sifry et al. (2016)"},{"why":"Provides the earlier resolved sample, the radius-luminosity trend, and the selection-effect methodology that the REASONS bias maps extend.","marker":"Matrà et al. (2018)"},{"why":"Provides moving group membership and ages used to place Sco-Cen and young moving-group belts on the collisional age axis.","marker":"Gagné et al. (2018)"},{"why":"Provides the slower t^{-0.4} collisional evolution model that the mass-radius-age comparison disfavours.","marker":"Löhne et al. (2008)"}],"fun_headline_variants":["Smaller exo-Kuiper belts lose dust faster, ALMA finds","Most belts around stars are broad, not narrow, survey says","Exo-belt inclinations of 1-20° suggest early stirring","REASONS maps 74 belts, reveals collisional aging"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The inference that the young-to-old mass offset is depletion rather than birth differences rests on the assumption that the initial millimetre-grain belt mass is independent of stellar and belt properties, with a common start time of 10 Myr and a single collisional constant; if young and old belts were born with different masses or collisional histories, the observed offset would not measure the depletion rate.","fun_headline_variants_meta":{"raw":{"variants":["Smaller exo-Kuiper belts lose dust faster, ALMA finds","Most belts around stars are broad, not narrow, survey says","Exo-belt inclinations of 1-20° suggest early stirring","REASONS maps 74 belts, reveals collisional aging"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001914,"raw_usage":{"total_tokens":7582,"prompt_tokens":1119,"completion_tokens":6463,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":735,"completion_tokens_details":{"reasoning_tokens":6386}},"tokens_in":735,"tokens_out":6463,"duration_ms":43806,"temperature":1.0,"reasoning_tokens":6386,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:11:15.437312+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find, after applying the paper's own selection corrections, an old field belt (age >200 Myr) with a mass and radius in the top-left region of the mass-radius diagram typical of the 10–30 Myr Sco-Cen belts, such as roughly $1\\,M_\\oplus$ of millimetre grains at $R<50$ au; its presence would break the claimed depletion pattern, as would a coeval old population showing the same mass-radius distribution as the young associations.","supporting_citations":[],"review_version":1}