{"id":"67efe607-3f7d-453d-9df3-73cdd977ef11","arxiv_id":"1908.10514","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Across eleven giant molecular clouds, about 63% of 835 ammonia-traced dense clumps are gravitationally bound, and clumps on and off filaments show no significant difference in virial state.","lead":"Astronomers mapped the dense ammonia gas in eleven giant molecular clouds and identified 856 clumps, finding that about 63% are heavy enough for gravity to bind them on their own. The survey provides the most uniform large dataset of such clumps in the Milky Way, a key step toward understanding how massive stars and star clusters are born.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 63% bound fraction rests on upper-limit leaf masses; using the paper's own clipped masses (median ~5x lower) would raise alpha_vir ~5x and likely move most of the 523 clumps above alpha=2.","rationale":"In good faith, the paper is a data release plus an empirical virial census; the central claim is not formally derivable and rests on a reasonable but unvalidated mass definition. The paper itself flags the limitation in §3.4, which is creditworthy, but the load-bearing statistic is presented without a corresponding systematic band. The reader's weakest_assumption identified the same issue, and the reader's conditional verdict is appropriate. The public data products and reduction code are strong independent support for reproducibility, and there is no evidence of circular reasoning or fabrication. The proposed check is straightforward because Mclip is already tabulated and the full virial analysis is reproducible. The verdict should remain CONDITIONAL: the paper should be published after the mass-estimator systematic is either propagated into the headline bound fraction or explicitly folded into a reported range.","tokens_in":55572,"tokens_out":3136,"duration_ms":38695,"concrete_test":"Recompute alpha_vir for all 835 catalogued leaves using Mclip from the published online catalog instead of Mobs, keeping sigma, R, and T unchanged; report the all-cloud bound fraction, per-cloud bound fractions, and the on-/off-filament alpha distributions. If the all-cloud bound fraction falls below ~50%, or if W48/M17 drop below ~0.7, the abstract's '523 (~63%)' must be reported as an upper-limit-dependent range rather than a single value.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Summary item 2, that 523/835 (~63%) of ammonia-identified clumps are gravitationally bound by gravity alone, depends directly on the leaf mass estimator chosen in §3.4. The paper measures Mobs by summing all H2 column density inside each dendrogram mask and explicitly states this is 'likely an upper limit'; the alternative 'clipped' masses are 'typically a factor of ~5 (median) lower.' Since alpha_vir = Mvir/Mobs (Eq. 1 and §3.5), a factor-of-five upward correction to alpha_vir would move most leaves currently in the 0.4 < alpha < 2 range above the alpha = 2 threshold. The published W3-west catalog (Table 5) illustrates the sensitivity: nearly all listed alpha values are 0.67-1.82, so multiplying by ~5 leaves essentially none bound. The paper presents only the integrated masses in the headline and does not propagate the clipped-mass systematic into the 63% statistic or the per-cloud bound fractions. The no-difference claim (item 4) is also affected because it compares alpha distributions computed with the same upper-limit masses. The acknowledged single-component NH3 fitting in M17 and W48 (§3.1) broadens line widths and inflates Mvir, which acts in the opposite direction; it does not rescue the mass systematic. This is not an internal inconsistency, but it is an unquantified systematic in the central quantitative claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents the first data release from the KEYSTONE survey, mapping NH3 (1,1) and (2,2) emission with the GBT KFPA across eleven Galactic giant molecular clouds. The authors fit the ammonia lines to produce temperature, centroid velocity, velocity dispersion, and column-density maps, then use dendrograms on the integrated intensity maps to define 856 clumps (leaves). For 835 clumps with Herschel-based masses, they compute virial parameters and report that ~63% (523) have alpha_vir < 2 and are therefore 'bound by gravity alone.' They compare clump properties on and off dust-continuum filaments, finding no significant differences, and identify a population of massive hubs/ridges that coincide with water masers, multiple 70 micron protostars, and the Kauffmann-Pillai massive-star-formation threshold. The paper also estimates cloud-weight and turbulent pressure contributions to virial stability for subsets of the sample.","tokens_in":55887,"tokens_out":3321,"duration_ms":37721,"significance":"If the central quantitative claims hold, this is one of the largest homogeneous samples of virial parameters for dense ammonia clumps in high-mass star-forming regions, and it provides useful constraints on how clump stability depends on filament association and environment. The survey data and catalogs are made public, the analysis uses established and well-described tools, and the appendix on distance-dependent resolution bias is a genuine strength. However, the headline 63% bound fraction is directly tied to the choice of mass estimator, and the paper itself supplies the evidence that this choice is the dominant systematic: the clipped masses are typically a factor of ~5 lower, which would raise alpha_vir by a comparable factor and likely move most of the 523 'bound' clumps above the alpha_vir = 2 threshold. The central claim therefore needs additional work before it can be considered supported.","major_comments":[{"comment":"The headline statistic that 523 of 835 clumps (~63%) are gravitationally bound is not robust to the mass-estimator uncertainty that the authors themselves describe. Section 3.4 states that summing all column density within leaf boundaries is 'likely an upper limit' on the mass, and that the alternative clipped masses are 'typically a factor of ~5 (median) lower.' Because alpha_vir = Mvir/Mobs directly, a factor-of-five lower Mobs would raise alpha_vir by approximately a factor of five. Most of the 523 leaves currently in the 0.4 < alpha_vir < 2 range would then fall above alpha_vir = 2. The W3-west catalog in Table 5 illustrates the point: the listed alpha values are mostly 0.67-1.82, so multiplying by ~5 leaves essentially none bound. The paper adopts the integrated masses throughout and does not propagate the clipped-mass systematic into the global 63% fraction, the per-cloud bound fractions in Table 6, or the correlation between bound fraction and protostellar surface density in §3.7. The authors should present the virial-parameter distributions and bound fractions for both mass estimators, or otherwise justify why the upper-limit masses should be preferred for this specific statistic.","section":"§3.4, §3.5, Eq. (1)"},{"comment":"The single-component NH3 fitting approximation is acknowledged to broaden fitted line widths in W48 and M17, where the highest bound fractions (0.89 and 0.87, respectively) are reported. Broadened line widths inflate Mvir and therefore alpha_vir, which means the true bound fractions for those clouds may be even higher than reported; however, this effect does not offset the mass-estimator systematic, and it introduces a separate, unquantified bias into the per-cloud and global bound fractions. The paper defers multiple-velocity-component fitting to a future paper, but a quantitative estimate of how many leaves in W48 and M17 are affected is needed before those clouds can be used to support the headline bound-fraction claim.","section":"§3.1, Table 6"},{"comment":"The distance-bias analysis in Appendix A shows that degrading the closest clouds to the W48 linear resolution causes essentially all identified structures to appear bound, and the text correctly notes that this may inflate the bound fraction in W48 and M17. This is an important caveat, but it is not carried into the abstract or Summary item 2, which state the 63% figure without mentioning the resolution dependence or the mass-estimator dependence. The summary should be revised to state the bound fraction as conditional on the adopted upper-limit masses and on the native resolution, or to report the range obtained under the alternative mass and resolution choices.","section":"Appendix A, §3.5, Summary item 2"}],"minor_comments":[{"comment":"The sentence 'converted into a turbulent pressure using Equations 5 and 9' appears to contain a reference error: the turbulent pressure P_T is defined in Equation 8, not Equation 9. Please check the equation numbering.","section":"§4.5, after Eq. (8)"},{"comment":"The header 'Bad N(H2) Pixels' is described in the footnote as the 'fraction of pixels in the leaf that were saturated in the H2 column density map,' but the column appears to contain integer-like values in Table 5. Please clarify the units and formatting of this column.","section":"Table 4, column 14"},{"comment":"The abstract says '40-100%' of clumps are aligned with filaments, while Table 6 and §4.1 report on-filament fractions from 0.35 in Cygnus X South to 1.0 in W3-west. The abstract should use the actual range (35-100%) or state the rounding convention.","section":"Abstract and §4.1"},{"comment":"The power-law fit to the mass-radius relation is described as using an MCMC sampler with 'orthogonal least-squares likelihood function,' but the exact likelihood and treatment of measurement errors are not specified. A brief description or reference would improve reproducibility.","section":"§3.4, Figure 30"}],"recommendation":"major_revision","confidential_remarks":"This is a solid survey paper with valuable public data, and the main issues are concentrated in how the headline bound fraction is presented and supported. The mass-estimator systematic is acknowledged in the text but not propagated into the central statistic, and I believe the authors can address this with a re-analysis or a bounding calculation. The paper is within scope for the journal and does not require a reject recommendation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Jared, here is my read of Keown et al. (1908.10514). The headline result is 523 of 835 clumps (63%) with alpha_vir < 2, but that number has a load-bearing asterisk. The masses used are the integrated column-density sums, which the authors themselves call upper limits. Their own clipped masses are typically five times lower; if those are closer to the truth, alpha_vir = Mvir/Mobs goes up by roughly a factor of five and most of those 523 clumps would sit above the alpha=2 line. The paper discloses this in Section 3.4 but does not propagate it into the abstract or the summary. The honest phrasing is \"up to 63% may be bound,\" not \"63% are bound.\"\n\nWhat the paper does well: it is a substantial data release. First homogeneous NH3 (1,1)/(2,2) maps and clump catalogs for eleven GMCs, with public cubes, catalogs, and code. The virial analysis uses independent Herschel column densities, no fitted parameters are recycled into the headline, and the distance-bias appendix is a good-faith attempt to quantify a real resolution effect. The hub/ridge identification is checked against water masers and 70 micron sources, which gives it some independent support.\n\nSoft spots, in proportion. The mass systematic is the main one; it is central and unquantified in the headline. Second, the single-component NH3 fits in M17 and W48 broaden line widths and inflate Mvir, which pushes alpha up. That acts in the opposite direction from the mass bias, so it does not cancel; it just means those two high-bound-fraction clouds have competing systematics. Third, the \"no significant difference\" claim between on- and off-filament clumps is based on overlapping histograms, not a formal test; a KS test would be cheap. Fourth, the distance-resolution bias is only tested on the three nearest clouds, which is fine as a caveat but limits the quantitative reach.\n\nWho benefits: anyone working on high-mass star formation, ammonia surveys, or virial analysis will reuse this catalog. The paper deserves peer review — the data alone justify that — but the quantitative claims need tightening before publication, and the abstract should state the mass assumption explicitly. I would send it to a competent referee and ask for the clipped-mass version of the bound fraction to be reported alongside the headline.","headline":"The survey is a genuine data release with honest caveats, but the headline 63% bound fraction is an upper limit in disguise because it rests on upper-limit masses.","tokens_in":56556,"tokens_out":2356,"would_cite":true,"duration_ms":25219,"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 KEYSTONE survey of ammonia in eleven galactic giant molecular clouds finds that about 63% of the 835 dense clumps with mass estimates are gravitationally bound (virial parameter below two), and that clumps on and off dust filaments…","keywords":["giant molecular clouds","ammonia","virial parameter","dense clumps","filaments","dendrograms","high-mass star formation","Herschel"],"falsifier":"Recompute the virial parameters using the clipped (background-subtracted) leaf masses instead of the integrated column-density masses. If the median factor-of-five mass reduction is applied, most of the 523 clumps with $\\alpha_{\\rm vir} < 2$ would move above $\\alpha_{\\rm vir} = 2$, directly testing whether the 63% bound fraction is an artifact of the mass definition.","tokens_in":55369,"feed_emoji":"🌌","tokens_out":5560,"duration_ms":55114,"temperature":0.7,"pith_summary":"The KEYSTONE survey maps ammonia (1,1) and (2,2) emission across eleven giant molecular clouds between 0.9 and 3.0 kpc and identifies 856 dense gas clumps with a dendrogram analysis. The paper's central claim is that 523 of the 835 clumps with mass estimates (about 63%) have virial parameters below two, meaning gravity alone is strong enough to bind them, and that this bound fraction does not depend on whether a clump sits on a dust-continuum filament. The result matters because it extends core-scale virial statistics from nearby low-mass clouds to the clouds that dominate Galactic star formation, and it suggests that filaments are not a prerequisite for forming bound dense clumps in these environments. The paper also identifies massive, low-virial 'hubs' and 'ridges' that sit at filament intersections or within single filaments and meet empirical criteria for future massive-star or cluster formation.","feed_headline":"Survey finds 63% of dense gas clumps are gravitationally bound","feed_subtitle":"Ammonia mapping of eleven galactic giant molecular clouds shows gravity dominates, whether or not clumps sit on filaments.","key_machinery":"The load-bearing object is the virial parameter $\\alpha_{\\rm vir} = M_{\\rm vir}/M_{\\rm obs}$, with $M_{\\rm vir} = 5\\sigma^2 R/(aG)$, where $\\sigma$ is the total (thermal plus nonthermal) velocity dispersion, $R$ is the effective radius, and $a$ accounts for the radial power-law density profile ($\\rho \\propto r^{-1.5}$); structures with $\\alpha_{\\rm vir} < 2$ are called gravitationally bound. Masses come from summing Herschel H$_2$ column density inside dendrogram 'leaf' masks, and line widths come from single-velocity-component LTE fits to ammonia hyperfine spectra. The virial comparison is what turns maps of temperature, velocity dispersion, and column density into a statement about which structures can collapse.","core_discovery":"The paper reports that the majority of ammonia-identified clumps in its sample are gravitationally bound: 523 out of 835 (about 63%) have virial parameter $\\alpha_{\\rm vir} < 2$, using masses from integrated Herschel H$_2$ column density and line widths from NH$_3$ (1,1)/(2,2) fits. The bound fraction is nearly identical for on-filament clumps (about 65%) and off-filament clumps, and the mass, radius, temperature, and velocity dispersion distributions of the two populations are indistinguishable. In addition, a subset of unusually massive clumps ('hubs' and 'ridges') have virial parameters 0.2--0.5 and lie above the empirical massive-star threshold, typically hosting water masers and multiple 70 $\\mu$m protostars.","pith_inferences":["If the clipped masses are closer to the true clump masses, the headline bound fraction likely flips below 50%; this is an untested consequence of the paper's own mass uncertainty range.","The on/off-filament null result suggests that the 'mass flow along filaments' picture, while supported by hub observations, may not be required for clump-scale gravitational binding in massive GMCs; a direct test would be measuring whether off-filament clumps have different accretion rates than on-filament clumps.","The distance-resolution bias implies that unified comparisons of star formation efficiency across clouds require either common linear resolution or a resolution-dependent correction; upcoming high-resolution arrays could test whether W48-like clouds really contain more bound clumps.","Magnetic field support, not included in $\\alpha_{\\rm vir}$, would push the true bound fraction below 63% if fields are dynamically important in these environments."],"forward_implications":["If gravity alone binds most dense ammonia clumps, magnetic fields and external pressure only push them further toward collapse, so the default expectation in these GMCs is continuing gravitational contraction rather than dispersal.","The near-identical virial parameters on and off filaments imply that filament membership is not a controlling variable for dense-clump stability in high-mass GMCs, weakening the case that star formation there is exclusively a filament-channeled process.","Hubs and ridges with $\\alpha_{\\rm vir} \\approx 0.2$--$0.5$ and masses above the empirical $M(r) > 870\\,M_\\odot\\,(r/{\\rm pc})^{1.33}$ threshold are the most plausible precursors of massive stars and clusters in the sample.","Because distance-adjusted tests in the appendix show that lower resolution makes structures appear more bound, the cloud-to-cloud spread in bound fraction (0.3 to 0.9) is partly a resolution effect, not purely an environmental difference.","Most clumps (about 69%) are sub-virial once cloud weight pressure is included, so the dense clumps in GMCs are typically pressure-confined structures that may later become gravity-dominated as they accrete."],"supporting_citations":[{"why":"Supplies the virial-analysis equations, including the virial mass formula and the energy-density treatment used for cloud weight and turbulent pressure.","marker":"Keown et al. (2017)"},{"why":"Provides the ammonia line-fitting pipeline and single-velocity-component LTE modeling approach applied to the NH$_3$ (1,1) and (2,2) spectra.","marker":"Friesen et al. (2017)"},{"why":"Defines the dendrogram structure-identification methodology and the 'clipping' technique used to estimate lower-limit leaf masses.","marker":"Rosolowsky et al. (2008)"},{"why":"Establishes the virial parameter definition and the density-profile correction factor $a$ that the paper adopts.","marker":"Bertoldi & McKee (1992)"},{"why":"Provides the empirical mass-radius threshold that the paper uses to classify hubs and ridges as massive-star or cluster precursors.","marker":"Kauffmann & Pillai (2010)"},{"why":"Supplies the radius uncertainty method and the alternative mass-clipping approach that brackets the leaf mass estimates.","marker":"Chen et al. (2019a)"},{"why":"Offers the cloud weight pressure formulation that the paper adds to the virial balance beyond gravity and kinetic energy.","marker":"Kirk et al. (2017)"}],"fun_headline_variants":["Survey: 63% of dense gas clumps are gravitationally bound","Ammonia map reveals gravity-bound majority in GMCs","Filaments don't matter: dense clumps equally bound","Massive hubs identified in ammonia survey of GMCs","GMC census: 523 clumps bound, hubs for massive stars"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis assumes that summing all Herschel column density inside each dendrogram mask gives a mass close to the true clump mass, although the paper notes this is likely an upper limit and that 'clipped' masses are typically about five times lower; because $\\alpha_{\\rm vir} = M_{\\rm vir}/M_{\\rm obs}$, a factor-of-five smaller mass would raise the virial parameter by about a factor of five and could move most of the 523 'bound' clumps above the $\\alpha_{\\rm vir} = 2$ threshold.","fun_headline_variants_meta":{"raw":{"variants":["Survey: 63% of dense gas clumps are gravitationally bound","Ammonia map reveals gravity-bound majority in GMCs","Filaments don't matter: dense clumps equally bound","Massive hubs identified in ammonia survey of GMCs","GMC census: 523 clumps bound, hubs for massive stars"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000665,"raw_usage":{"total_tokens":3117,"prompt_tokens":1107,"completion_tokens":2010,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":723,"completion_tokens_details":{"reasoning_tokens":1923}},"tokens_in":723,"tokens_out":2010,"duration_ms":14396,"temperature":1.0,"reasoning_tokens":1923,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:41:32.293754+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the virial parameters using the clipped (background-subtracted) leaf masses instead of the integrated column-density masses. If the median factor-of-five mass reduction is applied, most of the 523 clumps with $\\alpha_{\\rm vir} < 2$ would move above $\\alpha_{\\rm vir} = 2$, directly testing whether the 63% bound fraction is an artifact of the mass definition.","supporting_citations":[{"cited_title":"W., Pineda, J","cited_arxiv_id":null,"evidence_quote":"Defines the dendrogram structure-identification methodology and the 'clipping' technique used to estimate lower-limit leaf masses."}],"review_version":1}