{"id":"fe23e36e-e367-44bb-a845-568ec6132a9a","arxiv_id":"2506.09125","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"With well-calibrated ALMA data, the galaxy-to-galaxy scatter in the CO(2-1)/CO(1-0) ratio is only about 0.05 dex, much lower than earlier single-dish estimates.","lead":"This study uses new ALMA observations of carbon monoxide (CO) in 14 nearby galaxies to measure how the ratio of two CO lines varies across and between galaxies. The results show that this variation is smaller than previously thought, which matters for how astronomers estimate molecular gas masses.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 0.05 dex galaxy-to-galaxy scatter is only interpretable as intrinsic if per-line calibration error is <~7%; at the 10% level the paper calls negligible, the ratio error alone is ~0.057 dex, equal to the observed scatter, leaving intrinsic variation unconstrained.","rationale":"The reader's weakest_assumption correctly identifies the relative flux calibration as the load-bearing element. My stress test sharpens this into a quantitative threshold: the reported 0.05 dex scatter is comparable to the ratio error produced by a 10% per-line calibration uncertainty (0.057 dex), so the central interpretation as intrinsic scatter is only safe if the per-line error is closer to 5%. The paper's statement that conclusions survive a 10% uncertainty is not consistent with error propagation. The proposed test, using the actual amplitude calibration solutions from the Cycle 9 Band 3 observations, would directly determine the per-line calibration scatter and settle whether 0.05 dex is an astrophysical measurement or an upper limit. Because the data products, sample construction, and correlation analysis are valuable regardless, a conditional acceptance is appropriate: the headline claim should be either supported by a direct calibration check or reframed as an upper limit.","tokens_in":36066,"tokens_out":14576,"duration_ms":146984,"concrete_test":"For the 12 Cycle 9 Band 3 executions, retrieve the ALMA pipeline amplitude calibration solutions and measure the run-to-run scatter in the flux scaling factor. If the per-line calibration error derived from this scatter exceeds ~8.5%, propagate it into the R21 uncertainties and recompute the galaxy-to-galaxy scatter; if the inferred intrinsic scatter is then consistent with zero, the headline '0.05 dex intrinsic scatter' should be reframed as an upper limit.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central quantitative claim requires that per-line ALMA flux calibration errors be near 5%. With 5% per line, the ratio uncertainty is 0.030 dex, leaving roughly 0.04 dex intrinsic scatter. At 10% per line, the ratio uncertainty becomes 0.057 dex, matching the reported 0.05 dex scatter, so the intrinsic scatter would be formally indistinguishable from zero. Section 2.2.3 states that the conclusions still hold even assuming ~10% higher uncertainty, but this is not supported by quadrature: a 10% per-line error injects ratio scatter comparable to the headline value. The 5% estimate is adopted from ALMA technical documentation and prior PHANGS TP repeatability, not directly demonstrated for the new Cycle 9 Band 3 7m+TP CO(1-0) observations used here. Since the observed scatter has the same magnitude as a plausible 10% calibration floor, the 0.05 dex value currently cannot be separated from calibration noise, and the claim that intrinsic galaxy-to-galaxy variation is much smaller than previous estimates rests entirely on the unverified 5% assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents new ALMA ACA 7m+TP CO(1-0) maps of 12 nearby galaxies, combines them with two archival ALMA CO(1-0) maps and PHANGS-ALMA CO(2-1) data for 14 galaxies, and measures the resolved CO(2-1)/CO(1-0) ratio R21 at a common 1.7 kpc resolution. It reports a sample-wide median R21 of about 0.64, a galaxy-to-galaxy scatter of 0.05 dex that is lower than the >0.1 dex scatter of earlier single-dish studies, an internal within-galaxy scatter of about 0.1 dex with enhanced central values near 0.75, and correlations of R21 with galactocentric radius, SFR surface density, specific SFR, and metallicity. The paper concludes that assuming a fixed R21 = 0.64 does not significantly bias kpc-scale molecular gas mass estimates for main-sequence galaxy samples, and that calibration uncertainties and the CO-to-H2 conversion factor dominate systematic scatter.","tokens_in":36341,"tokens_out":4850,"duration_ms":56630,"significance":"If the low galaxy-to-galaxy scatter and the reported scaling relations hold, this is a valuable calibration-quality benchmark for CO-derived molecular gas measurements and provides practical advice for studies using CO(2-1). The paper benefits from new ALMA data, a careful homogenization to a common resolution and grid, multiple averaging approaches, completeness checks, and a systematic cross-survey permutation analysis. The main quantitative claim, however, depends on the relative flux calibration accuracy between the new CO(1-0) and PHANGS CO(2-1) data; that dependence is the load-bearing point of the paper and needs explicit, quantitative support.","major_comments":[{"comment":"The headline claim that the galaxy-to-galaxy scatter is 0.05 dex is not separable from plausible per-line flux calibration errors at the level the paper itself discusses. If each line carries an independent 10% flux error, the ratio inherits sqrt(0.1^2 + 0.1^2) ≈ 0.141 dex fractional error, i.e., about 0.057 dex in log10, which equals the observed 0.05 dex scatter; at the adopted 5% per line, the injected scatter is about 0.030 dex and the residual intrinsic scatter is only about 0.04 dex after quadrature. The statement in §2.2.3 that the conclusions still hold at ~10% uncertainty is therefore not supported by the error budget as written, unless the 10% errors are strongly correlated between the two lines. Please provide a quantitative calibration-error treatment: either report direct repeatability or band-to-band transfer tests for the new Cycle 9 Band 3 observations, or explicitly reframe the 0.05 dex galaxy-to-galaxy scatter as an upper limit that includes plausible uncorrelated calibration noise.","section":"§2.2.3, §4.1.1, Fig. 4"},{"comment":"The size of the headline galaxy-to-galaxy scatter is quoted inconsistently: the Abstract and §4.1.1 give 0.05 dex, while point 1 of the Conclusions gives 0.06 dex. Since the comparison to the 0.11 dex single-dish scatter is central, please state one consistent number and specify exactly which quantity it describes (residual scatter around the fitted line, scatter of the galaxy medians, or scatter after subtracting a calibration term).","section":"§4.1.1 and Conclusions"}],"minor_comments":[{"comment":"The text contains a typo: 'resect' should be 'respect' in the description of the ratio-of-sums average.","section":"§3.2"},{"comment":"In the paragraph comparing to Leroy et al. (2022), the phrase 'in jibthe sample' appears to be an editing artifact and should be corrected.","section":"§4.4"},{"comment":"The caption contains the typo 'tht' for 'that' in the opening sentence describing the sample selection.","section":"Figure 1 caption"},{"comment":"For the log-log fits, the intercept q is reported in dex but the units are not stated; adding the units would avoid ambiguity for readers applying Eq. (5) to their own data.","section":"Table 3"},{"comment":"The sentence 'When we perform a liner regression' should read 'linear regression.'","section":"§4.2"}],"recommendation":"major_revision","confidential_remarks":"The skeptical concern about calibration is well-founded and targets the central quantitative claim. The paper is otherwise carefully executed with transparent methodology and appropriate hedging, so a major revision that adds a quantitative calibration-error budget or reframes the scatter as an upper limit should be sufficient. The inconsistency between the 0.05 dex and 0.06 dex scatter values should also be resolved."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this one. It is the first resolved, same-facility ALMA comparison of CO(2-1)/CO(1-0) across a small but meaningful sample, and it shows that a fixed R21 = 0.64 is good enough for kpc-scale molecular gas mass estimates in main-sequence galaxies. That conclusion is robust.\n\nWhat’s new: twelve new Cycle 9 ALMA 7m+TP CO(1-0) maps, combined with PHANGS CO(2-1), all processed through the same pipeline, convolved to 1.7 kpc. The calibration discussion is more careful than most. They include completeness checks, multiple averaging methods (median, ratio-of-sums, spectral stacking), and a cross-survey permutation analysis that makes the single-dish calibration problem concrete. The within-galaxy scatter of about 0.1 dex, and the fact that it dominates the galaxy-to-galaxy scatter, is solid.\n\nThe soft spot is the headline galaxy-to-galaxy scatter of 0.05 dex. That number only means something if the per-line absolute flux calibration error is near 5%. At 5% per line, the ratio uncertainty is about 0.03 dex and you recover an intrinsic scatter of roughly 0.04 dex. At 10% per line, the ratio uncertainty is about 0.057 dex, which is the same as the observed scatter. The paper says in Section 2.2.3 that the conclusions still hold at 10%, but that statement doesn’t survive quadrature. The 5% calibration estimate is adopted from ALMA documentation and prior repeatability work, not directly demonstrated for these new Band 3 observations. So the stronger claim—that previous single-dish surveys were dominated by calibration noise—is credible but not yet airtight. The weaker claim, that within-galaxy variation is larger than galaxy-to-galaxy variation, does not depend on this.\n\nMinor: the abstract says 0.05 dex and the conclusions say 0.06 dex; pick one. Sample is only 14 galaxies and the SFR surface density range is limited, which the authors acknowledge.\n\nWho should read this: anyone using CO(2-1) to estimate molecular gas masses, and people building the next generation of line-ratio predictors. It deserves peer review. I’d ask the authors to add a direct estimate of the calibration floor’s effect on the inferred intrinsic scatter, and to soften the galaxy-to-galaxy claim accordingly if they can’t verify 5% on these data.","headline":"A careful, genuinely useful ALMA line-ratio study; the headline 0.05 dex scatter is real but needs a tighter calibration argument.","tokens_in":36835,"tokens_out":3979,"would_cite":true,"duration_ms":37494,"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":"Using well-calibrated ALMA maps of 14 nearby galaxies, this paper finds the galaxy-to-galaxy scatter in the CO line ratio R21 is only about 0.05 dex, roughly half the previously reported values, and argues the earlier larger scatter was…","keywords":["CO line ratios","R21","molecular gas","nearby galaxies","ALMA","flux calibration","star formation rate surface density","radio lines: galaxies"],"falsifier":"Re-observe the same 14 galaxies' CO(1-0) and CO(2-1) emission with two independent, absolutely calibrated facilities and recompute the galaxy-to-galaxy scatter; if the scatter rises back above 0.1 dex, the low ALMA value is a calibration artifact, not an intrinsic property.","tokens_in":35939,"feed_emoji":"📡","tokens_out":8876,"duration_ms":84321,"temperature":0.7,"pith_summary":"This paper tries to establish how much the CO(2-1)/CO(1-0) line ratio, R21, really varies between and within nearby star-forming galaxies once calibration systematics are controlled. Using new ALMA 7m+TP CO(1-0) maps of 12 galaxies plus two archival maps, matched to ALMA CO(2-1) maps of the same galaxies on a common 1.7 kpc grid, the authors find a galaxy-to-galaxy scatter of only 0.05 dex. That is about half or less the scatter reported by single-dish surveys, and the paper argues most of the old scatter was flux calibration noise rather than astrophysics. Within galaxies, R21 varies about 0.1 dex and rises toward centers, with the local star-formation surface density the best predictor. If correct, a fixed R21 = 0.64 is adequate for kpc-scale molecular gas masses across main-sequence galaxy samples, with calibration and the CO-to-H2 conversion factor dominating the systematics.","feed_headline":"Galaxy-to-galaxy CO ratio scatter drops to 0.05 dex","feed_subtitle":"Matched ALMA maps show prior R21 scatter was mostly single-dish calibration error, not astrophysical variation.","key_machinery":"The load-bearing object is the resolved line ratio $R_{21} = W_{\\mathrm{CO(2-1)}}/W_{\\mathrm{CO(1-0)}}$ measured from integrated-intensity maps at a common physical resolution of 1.7 kpc. The argument is carried by three matched elements: new ALMA 7m+TP CO(1-0) observations processed through the survey imaging pipeline to recover short-spacing flux; the companion ALMA CO(2-1) maps of the same galaxies; and a comparison of survey permutations that isolates the 10–20% systematic offsets single-dish flux calibration injects into $R_{21}$. Three averaging schemes—the area-weighted median, ratio-of-sums, and spectral stacking—cross-check the ratio estimates, and the stated ~5% ALMA relative calibration uncertainty is the yardstick against which the 0.05 dex scatter is judged.","core_discovery":"On the paper's own terms, the central discovery is that the intrinsic galaxy-to-galaxy variation of R21 among nearby main-sequence star-forming galaxies is about 0.05 dex, smaller than the >0.1 dex quoted in earlier work, and that this smaller scatter is the astrophysical signal once flux calibration is controlled by using ALMA for both transitions. The sample-wide area-weighted mean ratio is 0.64, with central regions (r<2 kpc) rising to about 0.75, and typical internal scatter of 0.1 dex. The paper further finds that R21 correlates with SFR surface density with a power-law slope of 0.10–0.12 depending on binning, anticorrelates with galactocentric radius and metallicity, and that the previously reported large scatter largely disappears when both lines come from the same, well-calibrated facility. As a practical consequence, the authors conclude that a fixed $R_{21}=0.64$ does not significantly bias kpc-scale molecular gas mass estimates derived from CO(2-1) for large samples, while flux calibration and the CO-to-H2 conversion factor contribute more systematic scatter.","pith_inferences":["If the 0.05 dex inter-galaxy scatter is real, then CO excitation is a nearly universal property of star-forming disks at kpc scales, and reported environmental trends from mixed-telescope studies should be re-examined with same-facility data before being interpreted physically.","The survey-permutation offsets of 10–20% imply that future multi-line mapping campaigns should either secure both lines with the same array or tie single-dish data to absolute calibration standards, otherwise spurious trends of order the full astrophysical range can be manufactured.","A testable extension: because $R_{21}$ rises with SFR surface density with slope ~0.1, high-resolution (GMC-scale) maps should show steep local enhancements of $R_{21}$ around star-forming regions, a prediction that can be checked with existing high-resolution ALMA CO data.","The metallicity trend rests on only 8 galaxies spanning ~0.25 dex; extending the same matched-calibration analysis to low-metallicity dwarfs would determine whether the anti-correlation steepens, which would matter for high-redshift CO mass estimates."],"forward_implications":["A fixed $R_{21}=0.64$ can be used to convert kpc-scale CO(2-1) maps to molecular gas surface densities across main-sequence galaxy samples without introducing more than about 10 percent systematic bias.","Galaxy-to-galaxy comparisons of molecular gas excitation should focus on within-galaxy environmental drivers (centers, bars, arms) rather than global galaxy properties, since internal scatter is about 0.1 dex and exceeds the 0.05 dex inter-galaxy scatter.","The log-log slope of about 0.10–0.12 between $R_{21}$ and SFR surface density offers an empirical correction for resolved studies that want to improve on the fixed ratio.","Stellar mass provides a weaker but useful galaxy-wide predictor, with $\\log_{10}\\langle R_{21}\\rangle_{\\mathrm{gal}} \\approx -0.07\\, \\log_{10}(M_\\star/M_\\odot) + 0.57$ over $10^9$–$10^{11}\\,M_\\odot$.","Switching from CO(1-0) to CO(2-1) with fixed $R_{21}$ shifts the molecular Kennicutt-Schmidt index by about 2% sample-wide, up to roughly 10–20% for individual galaxies, which is smaller than conversion-factor uncertainties."],"supporting_citations":[{"why":"Supplies the CO(2-1) maps for the sample galaxies.","marker":"Leroy et al. 2021a"},{"why":"Documents the imaging pipeline and total-power flux repeatability that anchor the calibration assumption.","marker":"Leroy et al. 2021b"},{"why":"Single-dish-based comparison sample that shows the larger scatter the paper explains.","marker":"Leroy et al. 2022"},{"why":"Previous line-ratio study whose calibration uncertainties motivated the ALMA approach.","marker":"den Brok et al. 2021"},{"why":"Single-dish CO(1-0) survey whose quoted 25% calibration uncertainty illustrates the problem.","marker":"Yajima et al. 2021"},{"why":"Unresolved single-pointing sample used to check the global median ratio.","marker":"Keenan et al. 2024"},{"why":"Simulation work showing 7m+TP observations recover >90% of flux, supporting the flux-recovery claim.","marker":"Neumann et al. 2023a"},{"why":"Framework for CO-to-H2 conversion used to turn ratios into gas masses.","marker":"Bolatto et al. 2013"}],"fun_headline_variants":["ALMA-matched CO maps reveal 0.05 dex intergalactic R21 scatter","Fixed R21=0.64 safe for kpc-scale gas masses, ALMA data show","CO ratio scatter shrinks to 0.05 dex with ALMA-only calibration","Real R21 variation is 0.05 dex; previous scatter was calibration","Well-calibrated ALMA trims R21 scatter to 0.05 dex"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The relative flux calibration between the new CO(1-0) and the CO(2-1) maps is stable to about 5 percent across all 14 galaxies, so the measured 0.05 dex galaxy-to-galaxy scatter reflects intrinsic astrophysical variation rather than leftover calibration noise.","fun_headline_variants_meta":{"raw":{"variants":["ALMA-matched CO maps reveal 0.05 dex intergalactic R21 scatter","Fixed R21=0.64 safe for kpc-scale gas masses, ALMA data show","CO ratio scatter shrinks to 0.05 dex with ALMA-only calibration","Real R21 variation is 0.05 dex; previous scatter was calibration","Well-calibrated ALMA trims R21 scatter to 0.05 dex"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001077,"raw_usage":{"total_tokens":4621,"prompt_tokens":1174,"completion_tokens":3447,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":790,"completion_tokens_details":{"reasoning_tokens":3347}},"tokens_in":790,"tokens_out":3447,"duration_ms":27263,"temperature":1.0,"reasoning_tokens":3347,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T04:57:06.440065+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-observe the same 14 galaxies' CO(1-0) and CO(2-1) emission with two independent, absolutely calibrated facilities and recompute the galaxy-to-galaxy scatter; if the scatter rises back above 0.1 dex, the low ALMA value is a calibration artifact, not an intrinsic property.","supporting_citations":[],"review_version":1}