{"id":"eed3e92f-979d-4e0d-ae11-1d06f25d5b0a","arxiv_id":"2509.07670","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A new all-sky catalog of 3,345 molecular clouds, derived from 3D dust reddening data, measures cloud distances from 90 pc to 4.3 kpc and traces the Local Bubble and Galactic structures.","lead":"Astronomers used a 3D dust map to identify 3,345 molecular clouds across the full sky, measuring each cloud's distance and basic properties. The catalog traces the Local Bubble's shell and large-scale structures in the Milky Way, giving a new resource for studying where stars form.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The catalog's core claim is unvalidated: dendrogram thresholds were tuned empirically with no completeness or contamination check, so the 3,345 clouds and the structures they trace may be parameter artifacts.","rationale":"The paper's central product is the catalog itself, and every downstream claim, from physical property distributions to the Local Bubble shell and large-scale cavities, depends on the dendrogram leaves actually corresponding to molecular clouds. The reader's weakest-assumption analysis identifies exactly this point: the Table 1 thresholds are empirically tuned and no completeness or contamination estimate is provided. I agree with that assessment. The W25 parametric model, which already includes a Local Bubble component, adds a specific reason why a threshold artifact could masquerade as a physical structure, but the primary unresolved issue remains the lack of validation of the clustering output. The paper deserves credit for a clearly described pipeline and for comparing with earlier catalogs, and the W25 map and dustmaps3d code are publicly available, so the proposed injection-recovery test is feasible. Since the reader's CONDITIONAL verdict already requires such validation, my stress-test does not move the verdict; it reinforces it.","tokens_in":15400,"tokens_out":7988,"duration_ms":79451,"concrete_test":"Construct a synthetic all-sky dust map by taking the smooth diffuse baseline of W25 (or an equivalent smooth model) and injecting thousands of synthetic clouds with known distance, size, and E(B-V) spanning the claimed catalog range. Run the identical pipeline, including Gaussian smoothing, HEALPix gridding, the Eq. 4 distance binning, and astrodendro with the Table 1 parameters, then measure the recovery fraction and false-positive rate per distance bin. If recovery is below 80% or contamination exceeds 10% in any bin, the 3,345-cloud count and the derived mass and radius distributions are not robust; if the input population is accurately recovered, the threshold concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing assumption is that each dendrogram leaf in the W25 extinction-gradient cubes is a real molecular cloud and that the Table 1 thresholds (min_value=0.05 mag/kpc; min_delta=0.5/0.4/0.3/0.1 mag/kpc; min_npix=2000/500/200/50) separate true clouds from noise. This assumption is unsupported. Section 3.2 justifies the thresholds only by 'extensive empirical testing'; Figure S1 shows how property histograms shift when one parameter is changed, but it does not report how the number or identity of clouds changes, nor does it quantify false positives or false negatives. Because the headline results, including the 3,345-cloud count, the Local Bubble shell, the Giant Oval and Carina cavities, and the Lower Sagittarius-Carina Spur, are all derived from the same tuned pipeline, they are not protected against a systematic threshold bias. The issue is compounded by the input map: W25 is a parametric model that already includes a Local Bubble component and 'various molecular cloud components,' so some of the gradient peaks and the shell/cavity morphology in the extracted catalog may be imprinted by the model rather than independently detected. Without an injection-recovery test or an external tracer comparison, the catalog's completeness, purity, and the structural conclusions remain conditional on the tuning choices.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents an all-sky catalog of 3,345 molecular clouds identified by applying the astrodendro algorithm to the W25 three-dimensional dust extinction map. The method converts the HEALPix-based map into twelve data cubes with adaptive distance binning, computes extinction gradients, and runs dendrogram clustering with distance-dependent thresholds (Table 1). For each dendrogram leaf, the paper derives Galactic coordinates, distance, size, mass, surface density, and average dust density, and flags clouds associated with the Local Bubble shell. The catalog is used to trace the Local Bubble, high-latitude clouds, and large-scale Galactic structures including the Giant Oval Cavity, the Carina cavity, and the Lower Sagittarius-Carina Spur. The authors compare cloud property distributions with earlier CO- and dust-based catalogs and discuss systematic differences in mass and radius measurements.","tokens_in":15565,"tokens_out":3270,"duration_ms":32401,"significance":"If the catalog is reliable, it would be a valuable community resource: it provides all-sky coverage with heliocentric distances from 90 pc to 4.3 kpc, is based on dust extinction rather than CO emission, and therefore can include CO-dark clouds; the authors also make the input map and access software publicly available and compare their results with several previous catalogs. The paper is transparent about the algorithmic steps and provides a machine-readable table of cloud properties. However, the scientific value is conditional on the stability and validity of the dendrogram decomposition, because the headline number of 3,345 clouds, the Local Bubble shell morphology, and the reported large-scale cavities are all products of tuned clustering thresholds applied to a parametric dust map. The manuscript currently offers only limited quantitative validation of those products.","major_comments":[{"comment":"The dendrogram thresholds (min_value = 0.05 mag/kpc; min_delta = 0.5/0.4/0.3/0.1 mag/kpc; min_npix = 2000/500/200/50) are justified only by 'extensive empirical testing,' and Figure S1 shows histograms of derived properties under single-parameter variations but does not report how the number or identity of the 3,345 clouds changes, nor any false-positive or false-negative rate. Because the central claim of the paper is the catalog itself, the absence of a completeness and contamination analysis (for example, injection-recovery tests into the W25 map, or a comparison against CO-selected clouds with known distances) leaves open the possibility that a substantial fraction of the leaves are noise fluctuations or parameter artifacts. This is a load-bearing issue for the catalog and for all structural conclusions drawn from it.","section":"Section 3.2, Table 1"},{"comment":"The input W25 map is a parametric model that already includes a Local Bubble component and 'various molecular cloud components,' so the later identification of a quasi-spherical shell of clouds around the Local Bubble in Figure 9 may partly reflect the imprinted model rather than an independent detection. This circularity concern should be addressed by a concrete test, such as repeating the cloud extraction on an independent 3D dust map (e.g., Vergely et al. 2022) or by comparing the Local Bubble shell clouds with CO or HI observations that were not used to construct W25.","section":"Section 2 and Section 4.1"},{"comment":"The classification of roughly 650 clouds as Local Bubble shell members relies on adding 'a conservative offset of 70 pc' to the boundary distance derived from the W25 extinction profile, but no justification, derivation, or sensitivity analysis for this offset is given. Since the Local Bubble shell count and the associated structural interpretation depend directly on this offset, the paper should show how the number of flagged clouds and the shell morphology change as the offset is varied over a plausible range.","section":"Section 4.1, Figure 9"},{"comment":"The catalog entries in Table 2 list distances, radii, masses, surface densities, and dust densities without any uncertainties. For a catalog paper whose central product is Table 2, missing error bars are a serious omission: the paper quotes a typical reddening precision of 0.01–0.05 mag, but does not propagate these uncertainties, the distance-bin width, or the smoothing scale into the derived physical parameters. The authors should provide at least representative uncertainties, and preferably per-cloud uncertainties, for the tabulated quantities.","section":"Table 2 and Section 3.3"},{"comment":"The distance-adaptive min_delta and min_npix values are chosen to maintain a roughly constant physical size selection, yet Figure 10 and the text attribute part of the observed differences between Local Bubble and non-Local Bubble clouds to these distance-dependent parameters. The paper does not separate the intrinsic physical trends from the selection effects introduced by the parameter choices. A quantitative statement of the selection function of the algorithm as a function of distance, density, and angular size is needed before the mass-radius relation in Figure 12 or the physical-property comparisons in Section 5.1 can be interpreted.","section":"Section 3.2 and Figure 10"}],"minor_comments":[{"comment":"The sentence 'Although each parameter set is run on the full cube, only structures whose centroids fall within the corresponding distance range are retained' should be reconciled with Table 1, which lists four parameter sets and four distance ranges; the text does not explain how a structure with a centroid outside the chosen range is handled when multiple sets overlap at the bin boundaries.","section":"Section 3.2"},{"comment":"The paper refers to 'dust density' while the voxel values are extinction gradients in mag/kpc; this terminology is potentially confusing, and a brief clarification of the relation between rho_dust and the tabulated quantity would help readers.","section":"Section 4.2"},{"comment":"The reference to Wang et al. (2025) is listed as 'Submitted' in the reference list but described as 'ApJS, in press' in Section 1; this inconsistency should be corrected.","section":"Section 2"},{"comment":"There are several typographical errors in the supplementary captions, including 'regio' in Figure S3, 'Gaiacoverage' in Figure S3, and 'Similarto' in Figure S8; these should be fixed in the final version.","section":"Supplementary Material captions"},{"comment":"The distance binning starts at 0.05 kpc 'to exclude the Local Bubble,' but the Local Bubble is a three-dimensional structure, not merely a distance cutoff; the motivation for this choice should be stated more precisely.","section":"Section 3.1, Equation (4)"},{"comment":"The claim that voxel-based mass estimation 'yields a more physically meaningful measure of the gravitationally bound component' is a strong interpretive statement that goes beyond the data presented; I suggest softening it or supporting it with a virial or CO-line comparison.","section":"Section 5.1"}],"recommendation":"major_revision","confidential_remarks":"The paper is a candidate for publication in a journal that welcomes large catalogs, and the authors have been transparent about the methodological chain. My main concern is that the central product, Table 2, is presented without uncertainty estimates and without a quantitative validation of the clustering thresholds. I would be comfortable with a major revision that adds an injection-recovery or external-comparison validation, per-cloud or representative uncertainties, and a sensitivity analysis for the Local Bubble offset. I do not see a need to reject the manuscript on the basis of the current evidence, but the catalog should not be released for community use without those additions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a useful paper, and I'd send it to review, but the current version oversells what the method can support. The all-sky catalog of 3,345 molecular clouds is genuinely new: previous dust-based dendrogram catalogs covered part of the sky or far fewer objects, and the W25 map gives it a reach from 90 pc to ~4.3 kpc. The distance-adaptive parameters and the careful remapping of HEALPix to a flat grid are sensible engineering choices. The comparisons with Chen, Guo, Xie, and Miville-Deschênes are honest and show that the mass and radius distributions are not crazy. As a data product, this will be used.\n\nThe soft spots are real, though. The dendrogram thresholds in Table 1 were picked by \"extensive empirical testing,\" and Figure S1 only shows histograms of properties under changed parameters. It does not show how the number or identity of clouds changes, and there is no injection-recovery or completeness test. That means the headline count, the Local Bubble shell, the cavities, and spurs are all conditional on the tuning. Table 2 lists no uncertainties on distance, mass, or size. The 70 pc offset for the Local Bubble flag is arbitrary. And W25's parametric model already includes a Local Bubble plus cloud components, so the later 'detection' of a shell is partly baked into the input map. That is moderate circularity, not fatal, but the authors should say it louder.\n\nThe stress-test note says the core claim is unvalidated. I'd put it slightly more charitably: the catalog is probably right at a rough level, but the quantitative claims and structural interpretations carry no error budget. The paper would be strengthened by a real validation section — inject synthetic clouds into the W25 cubes and run the full pipeline, report completeness and contamination as a function of distance and contrast, and give uncertainties on the catalog entries. Also, release the full machine-readable catalog (Table 2 looks truncated in the arXiv version).\n\nWho is this for? Anyone needing a distance-tagged sample of local molecular clouds, especially for studies of the Local Bubble and high-latitude clouds. I'd cite it if I did that work, but I'd be cautious about using the individual masses and flags without checking them.\n\nMy recommendation: send to a serious referee. The right outcome is major revision, not rejection. The catalog is worth the referee time, but the paper should not appear in its current form without completeness tests and uncertainty estimates.","headline":"Useful all-sky molecular cloud catalog from 3D dust, but the tuning and missing uncertainties make it a resource rather than a definitive measurement.","tokens_in":16270,"tokens_out":2300,"would_cite":true,"duration_ms":20682,"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":"This paper establishes an all-sky catalog of 3,345 molecular clouds by applying distance-adaptive dendrogram clustering to a three-dimensional dust extinction map, showing that these clouds trace the Local Bubble shell and Galactic-scale…","keywords":["molecular clouds","interstellar medium","three-dimensional dust extinction","dendrogram clustering","Local Bubble","Galactic structure","dust reddening","astronomical catalogs"],"falsifier":"Compare the catalog with an independent census built from CO emission with well-characterized distances: if a large fraction of the dust-selected clouds lack any CO counterpart and cannot be attributed to known CO-dark gas, the dust-only selection may be counting structures that are not molecular clouds. Alternatively, inject artificial clouds of known size and contrast into the W25 map and measure the recovery fraction at each distance bin to test the completeness of the dendrogram parameters.","tokens_in":15087,"feed_emoji":"☁️","tokens_out":5746,"duration_ms":47435,"temperature":0.7,"pith_summary":"The paper sets out to show that an all-sky three-dimensional dust extinction map, rather than CO emission alone, can serve as a complete basis for finding molecular clouds and measuring their distances. By applying a dendrogram decomposition with distance-adaptive thresholds to the W25 reddening map, it identifies 3,345 clouds from roughly 90 pc to 4.3 kpc and assigns each a position, radius, mass, surface density, and dust density. The resulting spatial distribution traces the Local Bubble shell, a population of high-latitude clouds, and large-scale cavities and spurs in the Galactic disk. If correct, the catalog extends cloud studies to CO-dark and nearby diffuse clouds and provides a distance-resolved view of the local interstellar medium.","feed_headline":"3,345 molecular clouds cataloged from 3D dust map across the sky","feed_subtitle":"Cloud census spans 90 pc to 4.3 kpc and traces the Local Bubble shell, cavities, and spurs.","key_machinery":"The load-bearing mechanism is the dendrogram, a hierarchical clustering structure built from iso-density contours of an extinction-gradient data cube, pruned by minimum value, minimum contrast, and minimum pixel count. The distance-adaptive twist is that these thresholds change with heliocentric distance so that angular resolution and contrast effects do not systematically bias selection; each cloud is then a dendrogram leaf whose voxels yield a dust-weighted centroid, projected area, and total mass. This permits uniform physical-size selection across a wide distance range.","core_discovery":"The central claim is that hierarchical clustering of the W25 three-dimensional dust reddening map, with clustering thresholds that adapt to distance, recovers an all-sky population of 3,345 molecular clouds whose spatial distribution directly reflects the structure of the local interstellar medium. About 650 of these clouds lie on the boundary of the Local Bubble, and about 740 sit at high Galactic latitudes. The cloud positions outline the Local Bubble shell, a chimney and tunnel feature, the Giant Oval Cavity, the Carina cavity, and a continuous Lower Sagittarius-Carina Spur. The paper also argues that mass estimation based on the resolved dust voxels gives a more physically meaningful measure of the bound component than line-of-sight integrated extinction, and that clouds beyond 1 kpc follow a power-law mass-radius relation while nearby Local Bubble clouds scatter about it.","pith_inferences":["Because dust extinction traces all dusty gas, the catalog likely includes many transient, unbound clouds; a comparison with CO emission at matched distances could test what fraction are genuinely molecular versus dark dust features.","The distance-adaptive thresholds mean completeness varies with distance; injecting synthetic clouds of known size and contrast into the W25 map and measuring recovery rates would directly quantify this selection effect.","The apparent large-scale cavities and spurs seen in the cloud distribution could be tested by checking whether they persist in independent tracers such as HI emission or other 3D dust maps beyond 4.3 kpc.","The scatter in the mass-radius relation for nearby clouds may indicate an evolutionary sequence from pressure-confined to self-gravitating clouds; kinematic measurements would separate those families observationally."],"forward_implications":["Molecular clouds can be cataloged all-sky from dust extinction alone, including CO-dark clouds and nearby diffuse clouds that CO surveys miss.","Cloud distances come directly from the 3D map, avoiding kinematic distance ambiguities for clouds without associated stars.","The catalog gives a distance-resolved census of the local interstellar medium, including the Local Bubble shell, high-latitude clouds, and large-scale cavities and spurs.","Comparisons with CO-based catalogs become possible, showing that this method selects compact dense substructures and yields lower masses and smaller radii.","The cloud sample provides a well-characterized target list for follow-up studies of star formation, cloud evolution, and ISM dynamics."],"supporting_citations":[{"why":"Supplies the all-sky three-dimensional dust reddening map from which the cloud catalog is extracted.","marker":"T. Wang et al. (2025, hereafter W25)"},{"why":"Provides the dendrogram algorithm used to decompose the extinction cube into hierarchical cloud structures.","marker":"E. W. Rosolowsky et al. (2008)"},{"why":"Compiled CO survey data that serve as the standard baseline this dust-based catalog complements and compares against.","marker":"T. M. Dame et al. (2001)"},{"why":"CO-based catalog of 8,107 clouds used for comparing radius, mass, and surface-density distributions.","marker":"M.-A. Miville-Deschênes et al. (2017)"},{"why":"Previous dust-based dendrogram identification of 567 clouds that this paper extends to all sky with distance-adaptive thresholds.","marker":"B. Q. Chen et al. (2020)"},{"why":"Southern-sky dust-based cloud catalog used for comparison and validation of the approach.","marker":"H. L. Guo et al. (2022)"},{"why":"Supplies Gaia XP-spectrum reddening measurements that enter the W25 map construction.","marker":"X. Zhang et al. (2023)"},{"why":"Provides the mean molecular weight and conversion factors used to turn extinction into gas mass.","marker":"M. Lombardi et al. (2011)"},{"why":"Independent 3D dust map whose 'Giant Oval Cavity' is used to cross-check a large-scale void found in the cloud distribution.","marker":"J. L. Vergely et al. (2022)"},{"why":"Nearby dendrogram-based cloud sample whose tight mass-radius relation is overplotted for comparison.","marker":"S. Cahlon et al. (2024)"}],"fun_headline_variants":["All-sky 3D dust map catalogs 3,345 molecular clouds","3,345 clouds from 3D dust: all-sky catalog","Cloud census via 3D extinction: 3,345 objects","3,345 molecular clouds mapped through dust","Dust-based 3D map yields 3,345 cloud catalog"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the fixed, distance-adaptive dendrogram thresholds separate genuine molecular clouds from noise fluctuations in the dust map without systematically splitting or merging real structures; the paper selects these thresholds by extensive empirical testing and provides no independent completeness or contamination estimate.","fun_headline_variants_meta":{"raw":{"variants":["All-sky 3D dust map catalogs 3,345 molecular clouds","3,345 clouds from 3D dust: all-sky catalog","Cloud census via 3D extinction: 3,345 objects","3,345 molecular clouds mapped through dust","Dust-based 3D map yields 3,345 cloud catalog"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000274,"raw_usage":{"total_tokens":1644,"prompt_tokens":951,"completion_tokens":693,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":567,"completion_tokens_details":{"reasoning_tokens":602}},"tokens_in":567,"tokens_out":693,"duration_ms":6095,"temperature":1.0,"reasoning_tokens":602,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:11:43.807327+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the catalog with an independent census built from CO emission with well-characterized distances: if a large fraction of the dust-selected clouds lack any CO counterpart and cannot be attributed to known CO-dark gas, the dust-only selection may be counting structures that are not molecular clouds. Alternatively, inject artificial clouds of known size and contrast into the W25 map and measure the recovery fraction at each distance bin to test the completeness of the dendrogram parameters.","supporting_citations":[],"review_version":2}