{"id":"f7dec4b9-4ade-43e0-9039-fd0003f57d90","arxiv_id":"2506.16252","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Gamma-ray emissivities of local gas agree with directly measured cosmic-ray spectra after correcting for optically thick HI, with a ~10% gradient toward the inner Galaxy and CO-dark H2 dominating small clouds.","lead":"This paper uses 15 years of Fermi gamma-ray data to separate five nearby molecular cloud regions into atomic, molecular, and 'dark' gas components, and then measures the cosmic-ray intensity in each. It finds that gamma-ray emissivities line up with directly measured cosmic-ray spectra and that small clouds hide far more dark molecular gas than bright CO gas.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Per-region IC/isotropic normalizations vary from 0 to 4.3 across the five ROIs; this unquantified degeneracy can bias the broad-HI emissivity that anchors the claimed CR-model agreement.","rationale":"The paper has real strengths: 15 years of LAT data, a transparent component decomposition, and an explicit emissivity model based on measured CR spectra. The central claim is plausible, and the paper is honest about several caveats. However, the per-region freedom in IC/iso is not a minor detail: the fitted normalization range (0 to 4.3) is far outside the ~10% precision claimed for the CR comparison, and no cross-check is provided. The D_em,res linearity concern raised by the reader is real but primarily affects the CO-dark H2 results and is explicitly acknowledged in Section 2.2.2; the IC/iso degeneracy bears directly on the headline CR claim. A joint-fit test would settle whether the agreement survives when the smooth backgrounds are treated as physical global components, and should be a condition for accepting the CR-intensity result.","tokens_in":25341,"tokens_out":7918,"duration_ms":97579,"concrete_test":"Re-fit the five ROIs jointly, keeping C_IC and C_iso as global parameters common to all regions (one value per energy band), with the same GALPROP IC model, while gas coefficients remain per-region. Then compare the best-fit broad-HI emissivities C_HI,2 with the CR-model prediction. If the global-fit C_IC and C_iso are inconsistent with 1 and the gas emissivities shift by more than ~5% from the per-region-fit values, the claimed agreement is not robust; if the shifts stay within ~3%, the concern is resolved.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Equation (1) fits C_IC and C_iso independently in each of the five ROIs and energy bands. The best-fit values are physically implausible as a set: Cep/Pol requires C_IC ≈ 3.4–4.3 with C_iso = 0 (Table 4), Orion requires the opposite extreme C_IC = 0 with C_iso ≈ 1.3–1.7 (Table 5), while MBM/Pegasus and R CrA sit in between (C_IC ≈ 0.9–1.7). A single Galactic IC component should not change normalization by a factor of four simply because the ROI changes; the large excursion means the smooth IC/iso templates are absorbing spatial structure that is not modeled, possibly including gas. Because broad HI is the anchor for the CR-intensity result, any such absorption biases C_HI,2 and therefore the claimed agreement with the directly-measured CR model and the ~10% gradient. The paper's stated ~4% systematic error covers only the residual-template construction and statistics, not this IC/iso degeneracy. The reader's concern about D_em,res linearity is valid but mainly targets the CO-dark H2 results; the IC/iso degeneracy bears directly on the central CR claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper analyzes 15 years of Fermi-LAT gamma-ray data (0.1–25.6 GeV) toward five nearby molecular cloud regions (MBM/Pegasus, R CrA, Chamaeleon, Cep/Pol, Orion). The authors construct ISM templates from a Gaussian decomposition of the HI 21-cm line into non-local, narrow (optically thick), and broad (optically thin) components, from CO line emission, and from a residual Planck dust-emission template interpreted as CO-dark H2. A binned maximum-likelihood fit of Eq. (1) decomposes the gamma-ray intensity into gas components (each normalized by C_HI,i, C_CO, C_dust), an IC template, an isotropic background, point sources, and, for R CrA, a Fermi-bubble template. The main claims are: (i) narrow HI is optically thick; (ii) the gamma-ray emissivity per H atom of broad HI agrees with the model based on directly measured CR spectra at Earth, with a possible ~10% decrease toward the inner Galaxy over ~500 pc; and (iii) the CO-dark-to-CO-bright H2 ratio anti-correlates with cloud mass, reaching 5–10 for ~1000 M_sun clouds. The abstract and Section 5 present these as the principal results.","tokens_in":25628,"tokens_out":4417,"duration_ms":50596,"significance":"If the central claim holds, the paper resolves a long-standing discrepancy in which uniform-T_s analyses inferred excess gamma-ray emissivity relative to directly measured CR spectra, and it provides a new, consistent way to separate optically thin HI, optically thick HI, and CO-dark H2 over a range of cloud masses. The analysis is thorough in several respects: it uses 15 years of LAT data with the latest response functions, tests three dust templates and three IC models, applies iterative template construction with outlier rejection, and explicitly cross-checks spin-temperature corrections. The acknowledgment of limitations (e.g., the proportionality assumption for D_em,res, and the possible contribution of non-local HI to the Cep/Pol dip) is welcome. However, the key CR-intensity result rests on the broad-HI normalization being unbiased by the unconstrained IC/isotropic templates, and the secondary ISM claims depend on a dust-to-gas scaling that is acknowledged to be not granted. These points need quantitative treatment before the paper can be accepted.","major_comments":[{"comment":"The fit treats C_IC and C_iso as independent free parameters in each ROI and each energy band. The best-fit values vary dramatically across the five ROIs: C_IC = 0 in Orion (Table 5) versus 3.4–4.3 in Cep/Pol (Table 4), with intermediate values in MBM/Pegasus, R CrA, and Chamaeleon; C_iso is fixed to 0 in Cep/Pol but reaches ~1.6–1.7 in Orion. Because the IC and isotropic templates are smooth, they can absorb large-scale spatial structure not captured by the gas templates, including gas correlated with broad HI. Since broad HI is the anchor for the claimed CR-model agreement and the ~10% gradient, the quoted ~4% systematic uncertainty (Section 4.1) does not cover this degeneracy. I ask the authors to quantify the impact: for example, by profiling C_IC and C_iso with physically motivated priors, by joint fitting with a single IC normalization shared across ROIs, or by explicitly propagating the observed C_IC/C_iso excursions into the derived emissivities and their errors.","section":"§2.2.4, Eq. (1); Tables 4–5; §4.1"},{"comment":"The residual gas template D_em,res is assumed to be proportional to the CO-dark H2 column density, and the paper explicitly states that this is 'not granted.' The nonlinearity of the tau_353-to-column ratio with density (Remy et al. 2017) can directly bias the CO-dark H2 column densities in Table 6 and the anti-correlation in Fig. 13, which is a secondary but load-bearing result. The authors should estimate the size of this bias, for example by using a density-dependent dust-to-gas factor or by cross-checking against an independent tracer such as reddening or extinction in a subset of regions.","section":"§2.2.2; Table 6; Fig. 13"},{"comment":"The claimed ~10% gradient of the CR intensity with Galactocentric radius is based on five emissivity points, one of which (Cep/Pol) is a marked outlier attributed by the authors to non-local HI contamination. No statistical significance is given for the trend, and the error bars do not include the IC/iso degeneracy discussed above. The authors should either present a fit to the trend with a significance estimate that includes systematic uncertainties, or explicitly label the gradient as tentative and model-dependent.","section":"§4.1; Fig. 12"},{"comment":"C_iso is fixed to 0 for the Cep/Pol region and C_IC is fixed to 0 for the Orion region because the best-fit values were 'very small.' Setting parameters to zero on the basis of an unconstrained fit can bias the remaining coefficients, particularly C_HI,2. Please show that the unconstrained best-fit values are consistent with zero within uncertainties and demonstrate that fixing them does not change C_HI,2 by more than the quoted statistical errors, or treat this choice as an additional systematic uncertainty.","section":"Tables 4 and 5 captions"}],"minor_comments":[{"comment":"The velocity-boundary description contains unit errors: '|v_LSR| is decreased above 15° to -100 K km s^-1' and 'increased by 30 K km s^-1' should read km s^-1, not K km s^-1. Also, the sentence beginning 'Similarly, v_LSR is decreased...' appears to be missing a subject for the region name.","section":"§2.2.1"},{"comment":"In the Orion row, the broad-HI entry reads '57,2' instead of '57.2'; this typo should be corrected.","section":"Table 6"},{"comment":"The spelling 'Camaeleon' appears in the ROI definition and in the abstract/body inconsistently with 'Chamaeleon' used elsewhere; please harmonize.","section":"Abstract and §2.1"},{"comment":"The reference 'Panooulou, G. V.' should be 'Panopoulou, G. V.'; also, 'diredtly' in Section 1 and 'devided' in Section 2.3 are typos.","section":"References"},{"comment":"The figure caption states that the Orion point is shifted horizontally for clarity, but it is not obvious in the figure how the shift is indicated; please add a note or an explicit offset marker.","section":"Fig. 12"}],"recommendation":"major_revision","confidential_remarks":"The IC/iso degeneracy is, in my reading, the most important issue: it directly affects the central CR-intensity claim and is not covered by the quoted systematic errors. The paper's own limitation statements (e.g., on D_em,res proportionality) are candid but need to be turned into quantitative systematics. The work is otherwise careful and well within the scope of PASJ; with the requested additions, it could be publishable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a careful, mostly convincing analysis that resolves a long-standing embarrassment—why gamma-ray emissivities of local gas exceeded predictions from directly measured CR spectra. The fix is using HI linewidth to separate optically thick narrow HI from optically thin broad HI rather than assuming a single spin temperature. The extension from the authors' MBM study to four more regions with a consistent 15-year Fermi-LAT reanalysis is real new work, and the central result—broad HI emissivity matching the CR model within ~10%—holds up under the checks they ran.\n\nI agree with the reader's conditional verdict. The paper is transparent about its main assumptions (D_em,res linearity; velocity boundaries; Ts caveats), and I don't see a circularity problem: the emissivity model comes from direct CR measurements, independent of the Fermi data.\n\nThe soft spots are real but not fatal. The stress-test note about IC/isotropic normalizations is worth taking seriously: C_IC ranges from 0 to ~4.3 across ROIs, and C_iso from 0 to ~1.7. That's a large excursion for what should be roughly the same physical components. The per-region choice of IC template (Std-SA0 vs Std-SA100) and the freedom to fix IC or iso to zero add unquantified model freedom. The stated ~4% systematic error covers the residual template construction and statistics, not this degeneracy. If smooth IC/iso templates are absorbing gas-correlated structure, the broad-HI emissivity—the anchor for the CR claim—could be biased. That said, the fact that emissivities line up with the model across five independent regions, despite very different fit configurations, argues against a serious systematic that would break the main conclusion. The ~10% gradient is softer than the abstract suggests; the paper itself notes the Cep/Pol dip may be due to non-local HI contamination, and the model comparison is qualitative.\n\nThe CO-dark H2 results are interesting and consistent with Remy et al. (2018), but rest on the unverified linear dust-to-gas assumption. The authors flag this clearly.\n\nWho is this for? People working on diffuse gamma-ray emission, HI optically thick corrections, and CR gradients in the local ISM. It deserves a serious referee; I'd send it out, with the request that the authors quantify the IC/iso degeneracy—either by jointly fitting or by reporting systematic errors that span the template choices.\n\nMy recommendation: engage with it. The central claim is probably right, and the weaknesses are manageable.","headline":"A careful, mostly convincing analysis that likely resolves the long-standing local gamma-ray emissivity discrepancy, but the unquantified IC/isotropic degeneracy needs attention before the CR gradient claim is fully trusted.","tokens_in":26214,"tokens_out":2248,"would_cite":true,"duration_ms":24271,"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":"Decomposing the 21 cm HI line into narrow and broad components removes the apparent excess of local cosmic-ray emissivity and brings gamma-ray results in line with direct measurements.","keywords":["cosmic rays","interstellar medium","gamma-ray emissivity","HI line decomposition","CO-dark molecular hydrogen","Fermi-LAT","local interstellar gas","Galactic cosmic-ray gradient"],"falsifier":"A direct test would be a 21 cm absorption measurement of the broad HI components in one of the five regions: if their optical depth exceeds roughly 0.1, the optically thin column is underestimated and the apparent agreement with directly measured cosmic rays would be an artifact. A second check is to repeat the Cep/Pol fit with the non-local HI velocity boundary shifted by a few km/s; if the claimed ~10% inner-Galaxy gradient moves or disappears, that gradient is not robust.","tokens_in":25117,"feed_emoji":"🌌","tokens_out":10078,"duration_ms":106176,"temperature":0.7,"pith_summary":"Most previous gamma-ray studies of the local interstellar medium assumed a single spin temperature when converting 21 cm HI emission to gas column density, and they inferred a cosmic-ray intensity higher than what direct measurements at Earth predict. This paper tests the alternative that the line width carries information: narrow HI components are optically thick, while broad HI components are optically thin and can anchor the cosmic-ray normalization. Fitting Fermi-LAT data for five nearby molecular cloud regions with this decomposition yields emissivities per H atom that are 5-30% lower than the earlier uniform-$T_s$ results and agree with the model built from directly measured cosmic-ray spectra. The same fit produces a ~10% increase in emissivity toward the inner Galaxy over about 500 pc, and shows that CO-dark H2 dominates dark gas and that its ratio to CO-bright H2 reaches 5-10 in clouds of about 1000 solar masses.","feed_headline":"Local cosmic-ray intensity now matches direct measurements","feed_subtitle":"Five nearby clouds yield emissivities 5-30% lower than older estimates and a 10% inner-Galaxy rise over 500 pc.","key_machinery":"The central object is the template-fit model of Equation (1), in which the observed gamma-ray intensity is a sum of gas-phase templates (non-local, narrow, and broad HI; CO; residual dust) multiplied by a common energy-dependent emissivity per H atom, plus inverse-Compton, isotropic, and point-source components. The decisive mechanism is the HI line-width separation: components with Doppler temperature below 1000 K are treated as optically thick and corrected by a fitted scale factor, while broader components are treated as optically thin and set the cosmic-ray normalization. The residual dust-emission map, formed by fitting Planck dust templates with HI and CO templates and keeping the leftover, is interpreted as CO-dark H2; its proportionality to gas column is assumed and then tested against gamma-rays by comparing $\tau_{353}$, radiance, and revised $\tau_{353}$ versions.","core_discovery":"The central discovery is that the gamma-ray emissivity per H atom of the broad, optically thin HI component matches the prediction from directly measured cosmic-ray spectra at Earth in all five regions once narrow HI is allowed to be optically thick. This removes the discrepancy that had supported either enhanced local CR intensities or systematically underestimated gas columns. The paper also reports a ~10% emissivity gradient across ~500 pc toward the inner Galaxy, and finds that the ratio of CO-dark H2 to CO-bright H2 anticorrelates with cloud mass, with low-mass clouds around $10^3$ solar masses having ratios of 5-10. These results are obtained by decomposing the ISM into five phases: non-local HI, narrow-line optically thick HI, broad-line optically thin HI, CO-bright H2, and CO-dark H2 traced by residual dust emission.","pith_inferences":["A natural extension is to apply the same narrow/broad HI decomposition to other nearby high-latitude clouds; if the emissivity stays flat and the CO-dark ratio continues to rise toward low masses, the pattern is a general ISM property rather than a property of these five regions.","The results imply that the local cosmic-ray excess reported by some uniform-$T_s$ analyses may not be a real spectral feature, which would weaken claims of a local CR source or gradient based on those analyses.","If the CO-dark H2/CO-bright H2 anti-correlation holds in lower-metallicity or more diffuse clouds, it could connect to extragalactic CO conversion factors, where small, weakly shielded clouds dominate.","The assumed proportionality of the residual dust map could be tested directly by comparing the gamma-ray-derived CO-dark column with independent dust extinction maps in the same regions; a linear correlation would confirm the template."],"forward_implications":["If the paper is right, earlier gamma-ray estimates of the local cosmic-ray intensity and gas column built on uniform-$T_s$ HI templates are biased high by 5-30%.","The agreement with directly measured CR spectra means gamma-ray emissivity of optically thin HI can serve as a cross-check for cosmic-ray intensity independent of propagation models.","The ~10% inner-Galaxy rise over ~500 pc is compatible with standard CR propagation, so nearby clouds sample a modest CR gradient rather than a flat local intensity.","The CO-dark-to-CO-bright H2 ratio of 5-10 in small clouds means CO luminosity seriously undercounts molecular mass in low-mass clouds.","Studies of cosmic-ray sources that use CO-traced cloud masses must include CO-dark H2 or they will overestimate the energy budget of the accelerator."],"supporting_citations":[{"why":"Supplies the spatially coherent Gaussian decomposition of the HI4PI data used to separate HI line components.","marker":"Kalberla & Haud (2018)"},{"why":"Establishes that narrow HI with $T_D \\leq 1000$ K traces optically thick or dark gas and broad HI traces translucent gas, the basis for the two HI templates.","marker":"Kalberla et al. (2020)"},{"why":"Provides the previous MBM/Pegasus analysis and the gamma-ray emissivity model (direct CR spectra plus hadronic and bremsstrahlung yields) reused here.","marker":"Mizuno et al. (2022)"},{"why":"Gives directly measured cosmic-ray spectra at Earth that set the predicted emissivity.","marker":"Maurin et al. (2014)"},{"why":"Provides the AAfrag hadronic interaction model for pion-decay gamma-ray production.","marker":"Kachelriess et al. (2019)"},{"why":"Supplies the electron bremsstrahlung model contributing to the expected gamma-ray yield.","marker":"Orlando (2018)"},{"why":"Summarizes the previous uniform-$T_s$ emissivity results for the same clouds, the baseline for the downward revision.","marker":"Planck Collaboration XI (2015)"},{"why":"Provides the revised $\tau_{353}$ dust map with better zero-level calibration, selected as the residual-gas template base.","marker":"Casandjian et al. (2022)"},{"why":"Documents the density-dependent rise of the $\tau_{353}$-to-column ratio, the caveat that makes the residual-gas proportionality not granted.","marker":"Remy et al. (2017)"},{"why":"Provides the CO survey map used for the CO-bright H2 template and the $X_{CO}$ estimates.","marker":"Dame et al. (2001)"}],"fun_headline_variants":["Gamma rays match cosmic-ray spectra at Earth in nearby clouds","Cosmic-ray intensity matches direct measurements in five clouds","Dark-to-bright H2 ratio anticorrelates with cloud mass in gamma rays","New HI decomposition resolves local cosmic-ray gamma discrepancy","Inner-Galaxy CR gradient seen in Fermi gamma-ray data"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the residual dust-emission map is proportional to the amount of CO-dark H2, and that the broad HI gas is transparent enough for its gamma-ray glow to measure the cosmic-ray intensity; the paper itself notes the first proportionality is not granted.","fun_headline_variants_meta":{"raw":{"variants":["Gamma rays match cosmic-ray spectra at Earth in nearby clouds","Cosmic-ray intensity matches direct measurements in five clouds","Dark-to-bright H2 ratio anticorrelates with cloud mass in gamma rays","New HI decomposition resolves local cosmic-ray gamma discrepancy","Inner-Galaxy CR gradient seen in Fermi gamma-ray data"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000234,"raw_usage":{"total_tokens":1540,"prompt_tokens":1034,"completion_tokens":506,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":650,"completion_tokens_details":{"reasoning_tokens":421}},"tokens_in":650,"tokens_out":506,"duration_ms":6543,"temperature":1.0,"reasoning_tokens":421,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T23:45:32.828444+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct test would be a 21 cm absorption measurement of the broad HI components in one of the five regions: if their optical depth exceeds roughly 0.1, the optically thin column is underestimated and the apparent agreement with directly measured cosmic rays would be an artifact. A second check is to repeat the Cep/Pol fit with the non-local HI velocity boundary shifted by a few km/s; if the claimed ~10% inner-Galaxy gradient moves or disappears, that gradient is not robust.","supporting_citations":[{"cited_title":"W., & Haud, U","cited_arxiv_id":null,"evidence_quote":"Supplies the spatially coherent Gaussian decomposition of the HI4PI data used to separate HI line components."},{"cited_title":"W., Kerp, J., & Haud, U","cited_arxiv_id":null,"evidence_quote":"Establishes that narrow HI with $T_D \\leq 1000$ K traces optically thick or dark gas and broad HI traces translucent gas, the basis for the two HI templates."},{"cited_title":"2022, ApJ, 935, 97","cited_arxiv_id":null,"evidence_quote":"Provides the previous MBM/Pegasus analysis and the gamma-ray emissivity model (direct CR spectra plus hadronic and bremsstrahlung yields) reused here."},{"cited_title":"2014, A&A, 569, 32","cited_arxiv_id":null,"evidence_quote":"Gives directly measured cosmic-ray spectra at Earth that set the predicted emissivity."},{"cited_title":"V ., & Ostapchenko, S","cited_arxiv_id":null,"evidence_quote":"Provides the AAfrag hadronic interaction model for pion-decay gamma-ray production."},{"cited_title":"2018, MNRAS, 475, 2724","cited_arxiv_id":null,"evidence_quote":"Supplies the electron bremsstrahlung model contributing to the expected gamma-ray yield."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the revised $\tau_{353}$ dust map with better zero-level calibration, selected as the residual-gas template base."},{"cited_title":"2017, A&A, 601, 78","cited_arxiv_id":null,"evidence_quote":"Documents the density-dependent rise of the $\tau_{353}$-to-column ratio, the caveat that makes the residual-gas proportionality not granted."}],"review_version":1}