{"id":"5423c098-c57b-402e-b50e-03ee70b21901","arxiv_id":"1908.03756","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"For 95 star-forming complexes in three galaxies, UV, infrared, atomic hydrogen, and H-alpha emissions all correlate, while CO brightness grows with velocity scatter only up to about 70 km/s and appears to fall beyond that.","lead":"The authors used archived telescope images of three nearby galaxies to compare how much gas, dust, and starlight each star-forming region emits. They found that molecular gas traced by CO behaves differently from other emissions and report tentative evidence that very disturbed complexes glow faintly in CO despite their large internal motions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed inverse CO branch rests on about five high-ΔV, low-CO-flux complexes whose ΔV uncertainties the authors themselves state are comparable to ΔV; without noise-based error bars the two-regime result is not established.","rationale":"The reader's weakest_assumption correctly identifies the unquantified ΔV uncertainties and the resulting fragility of the two-regime claim. My reading of the paper fully supports this: the paper itself acknowledges in Section III.B that ΔV uncertainty cannot be formally assessed and is comparable to ΔV for low-CO-flux complexes, which is exactly the regime of the high-ΔV points. The number of complexes defining the decreasing branch is very small, and the 3σ-endpoint method is particularly sensitive to noise when emission is weak and multi-peaked. This is the load-bearing weakness because the claimed inverse branch is the only genuinely new result; the correlations of Hα with UV, IR, and HI are unsurprising and the PAH interpretation is explicitly acknowledged as model-dependent. A Monte Carlo noise test on the few high-ΔV spectra would directly settle whether the inverse branch is real or an artifact. The authors' own cautious language and the explicit caveat in Section V that the results require further verification justify the CONDITIONAL verdict, so no adjustment is needed. The paper also deserves credit for using archival homogeneous data and for clearly distinguishing ΔV from a Gaussian-fitted velocity dispersion, but these strengths do not remove the need for uncertainty quantification on the pivotal kinematic points.","tokens_in":15152,"tokens_out":2926,"duration_ms":33097,"concrete_test":"For each of the five high-ΔV complexes (NGC 628 #3 and #61; NGC 3351 #8, #19, and #23), take the original HERACLES spectrum, add Gaussian noise at the measured per-channel rms in 1000 Monte Carlo realizations, and recompute ΔV with the same 3σ-endpoint procedure. If any substantial fraction (e.g., more than 5%) of realizations give ΔV < 70 km/s, the inverse branch is not robust. Report the full ΔV distributions and refit the F_CO–ΔV relation with these propagated errors to see whether the negative slope survives.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in Section V is the two-regime relation: CO luminosity rises with ΔV below roughly 70 km/s and falls above it. The high-ΔV branch is supported by only a handful of complexes in Table 1 with ΔV > 70 km/s: NGC 628 #3 (ΔV ≈ 290 km/s) and #61 (ΔV ≈ 123 km/s), and NGC 3351 #8 (ΔV ≈ 223 km/s), #19 (ΔV ≈ 173 km/s), and #23 (ΔV ≈ 107 km/s). These are precisely the low-CO-flux, multi-peak objects for which Section III.B states that the uncertainty in ΔV cannot be accurately assessed and assumes it is comparable to ΔV itself. The 3σ-endpoint definition of ΔV is fragile for weak, multi-peak emission: a single noise excursion above 3σ in a distant channel expands ΔV dramatically, so the large ΔV values of faint complexes may be noise-driven rather than kinematic. Because the decreasing branch is defined by these few points, and because their abscissa errors are comparable to the position of the 70 km/s break, the existence of a distinct feedback-dominated regime is not statistically supported by the data as presented. The authors' cautious 'preliminary' framing is appropriate, but the abstract's statement that 'the increase in the velocity scatter is accompanied by a decrease in the CO line luminosity' goes beyond what the current measurements can bear.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"Smirnova and Wiebe analyze archival multi-wavelength observations (GALEX UV, Spitzer 3.6–24 µm, Herschel 70–160 µm, THINGS HI, HERACLES CO, and ground-based Hα) of star-forming complexes in NGC 628, NGC 2976, and NGC 3351, using apertures defined in their earlier study and convolved to a common resolution. They report that UV, 8 µm, and 24 µm fluxes and surface brightnesses correlate with Hα; that the F8/F24 flux ratio anticorrelates with Hα flux; that HI correlates with Hα while CO correlates more weakly; and that the CO velocity scatter ΔV, defined as the full 3σ width of the CO spectrum, is claimed to relate to the CO flux in two regimes—rising below about 70 km/s and falling above it. The high-ΔV branch is described in the body as preliminary, and the authors explicitly state that the uncertainty in ΔV cannot be accurately assessed and is comparable to ΔV for faint complexes, and that their Umin and γ dust-model parameters are functions of the observed IR fluxes. The main conclusions are the correlations plus the tentative two-regime kinematic relation.","tokens_in":15446,"tokens_out":15556,"duration_ms":153720,"significance":"The paper's catalog and multiwavelength measurements for nearly one hundred complexes in three galaxies of different morphological types are a useful resource, and the qualitative correlations are broadly consistent with established SFR-indicator behavior. The potentially novel element is the two-regime ΔV–FCO relation: if confirmed, it would provide an observational constraint on how molecular gas kinematics respond to feedback in the regime where CO emission fades while internal motions grow, complementing galaxy-scale studies of gas velocity dispersion. The manuscript has genuine strengths: it uses homogeneous archival data reduced to a common resolution, the methodology is transparent, and the authors themselves flag the principal weaknesses, including the impossibility of accurately assessing ΔV uncertainties (Section III.B) and the fitted-flux dependence of Umin and γ (Section IV). Because the paper provides no quantitative significance measures and its headline kinematic claim rests on a handful of faint complexes with errors comparable to the measured quantities, the significance of the results is provisional until the robustness tests described below are supplied.","major_comments":[{"comment":"The central kinematic claim—that FCO grows with ΔV for ΔV ≲ 70 km/s and decreases with ΔV above that—is not statistically supported as presented. Section III.B states that the authors 'cannot accurately assess the uncertainty in ΔV' and assume it is 'comparable to the ΔV values for SFCs with low FCO fluxes.' The decreasing branch in Fig. 7 rests on about five complexes (NGC 628 #3 and #61; NGC 3351 #8, #19, and #23 in Table 1), all of them faint CO sources for which the 3σ-endpoint definition is most fragile, and the 70 km/s division is a visual choice with no significance test attached to either branch. Because the abscissa uncertainties are comparable to the position of the break itself, the data as presented cannot distinguish a two-regime relation from a scatter-dominated distribution. To make the claim load-bearing, the authors should (i) quantify ΔV uncertainties by Monte Carlo resampling of the observed noise and re-derivation of the 3σ endpoints; (ii) report correlation statistics (e.g., Spearman ρ with uncertainties) separately for ΔV ≤ 70 km/s and ΔV > 70 km/s; (iii) test the stability of the decreasing branch under removal of any single high-ΔV complex; and (iv) align the Abstract with Section V, where the two-regime result is already qualified as preliminary and as requiring 'further verification due to uncertainties associated with estimation of the parameter ΔV.' The Abstract's flat statement that the increase in velocity scatter is 'accompanied by a decrease in the CO line luminosity' is stronger than the acknowledged uncertainties permit.","section":"§III.B, Fig. 7, §V, Abstract"},{"comment":"Selection and measurement effects may produce the apparent inverse branch without any real decrease in CO luminosity. The SFC sample was chosen to include emission in at least one band (Section II), the 3σ detection threshold removes faint complexes with low peak flux regardless of their intrinsic ΔV, and the cut that excludes SFCs with ΔV exceeding half the global velocity range suppresses the high-ΔV end in galaxies with small rotation ranges such as NGC 2976. A faint, broadened complex is retained only if its peak exceeds 3σ, while a bright complex with the same ΔV has a higher peak and is more likely to be kept; this asymmetry can create a spurious anticorrelation between FCO and ΔV at high ΔV even if no physical trend exists. The authors should test this with injection experiments: embed artificial CO spectra of known FCO and ΔV into the line-free channels of the data cubes, run them through the same 3σ-endpoint and half-global-velocity selection, and show that the recovery of ΔV is unbiased as a function of FCO at high ΔV.","section":"§II, Fig. 7"},{"comment":"The interpretation of the high-ΔV branch as internal kinematics disturbed by feedback is not yet supported by the definition of ΔV, which is the full 3σ velocity extent of the spectrum rather than the width of a single component. This quantity includes the relative motion of multiple kinematically distinct clouds along the line of sight and, for the largest apertures (several kpc at the distances of these galaxies), the gradient of galactic rotation across the aperture. Fig. 6 illustrates the ambiguity: complex 3 of NGC 628 shows 'several lines slightly above the background' separated by almost 300 km/s, a pattern that could equally be a superposition of unrelated molecular clouds. The manuscript should estimate the rotation-shear contribution to ΔV from the HERACLES velocity fields or a rotation-curve model and quantify how much of the reported ΔV is attributable to shear and projection; without such an assessment, the feedback-dominated interpretation in Section IV is an interpretation rather than a measurement.","section":"§III.B, §IV"},{"comment":"The paper's correlational findings—Hα with UV, 8 μm, 24 μm, HI, and CO fluxes and surface brightnesses, and the F8/F24 anticorrelation with Hα—are asserted from visual inspection of figures, with no correlation coefficients, significance levels, or scatter estimates anywhere in Section III. This matters because per-galaxy samples are small (7 SFCs in NGC 2976), several flagged outliers exist (NGC 3351 #20 and #21), and some relations are non-linear or dominated by the dynamic range of the fluxes. The authors should report rank correlation coefficients with uncertainties for each claimed relation, for the full sample and per galaxy, and state explicitly how the identified outliers are treated in these statistics.","section":"§III.A, Figs. 2–4"},{"comment":"The interpretation of the F8/F24–Hα anticorrelation as evidence for a change in excitation conditions (an increase in the model parameter γ) is partly circular. As the authors acknowledge, the values of Umin and γ adopted from [14] 'are obtained by fitting spectra, and are therefore functions of the observed IR fluxes'; since F8/F24 is itself one of the fitted flux ratios, the Draine & Li (2007) model can reproduce the observed trend by construction, and the degeneracy between a reduced PAH mass fraction qPAH and an increased γ is not broken by the present data. The hedge in Section IV ('this correspondence should not be overemphasized') is appropriate, but Conclusion 3 still presents the excitation-condition explanation as a substantive result. The authors should refit the SEDs with the 8 μm band excluded (or use an independent PAH indicator), or explicitly recast Section IV as a consistency check and state in the conclusions that the mechanism is underdetermined by these data.","section":"§IV, Conclusion 3"}],"minor_comments":[{"comment":"Fig. 3's caption and the corresponding text describe both lower panels as 'the ratio of the 8 µm flux and far-IR flux'; the right panel should be the ratio of the 24 µm flux to the far-IR flux.","section":"Fig. 3 caption, §III.A"},{"comment":"Section III.B, the phrase 'for SFRs for which ΔV is comparable to the total velocity range considered' should read 'for SFCs' (star-forming complexes), not 'SFRs' (star-forming regions).","section":"§III.B"},{"comment":"Section II, the KINGFISH survey sentence contains a typographical double bracket, '[11]]', and the survey footnote markers are inconsistently placed.","section":"§II"},{"comment":"Section I describes NGC 628 as 'a classic spiral galaxy viewed almost side-on'; NGC 628 (M74) is nearly face-on, so the sentence should be corrected.","section":"§I"},{"comment":"Table 1 as printed contains only coordinates, FHa, ΔV, FUV, and NUV; the CO, HI, 8 μm, 24 μm, and far-IR fluxes that underlie Figs. 2–8 are not listed, so the quantitative claims cannot be checked from the table; the full table, or a machine-readable version, should be provided.","section":"Table 1"},{"comment":"Several rows of Table 1 have no FHa entry (e.g., NGC 628 #1, #2, #6, #7, #33, #38, #40–#43, #54, #56, #58, #62); the authors should state whether these are non-detections or formatting truncation.","section":"Table 1"},{"comment":"The units of the CO and HI fluxes plotted in Figs. 4 and 7 are never defined; the text should state that these are integrated line fluxes (presumably Jy km/s) from the THINGS and HERACLES surveys.","section":"Figs. 4 and 7"},{"comment":"Section IV, 'the variable Umin, defined in [14], correlates well with FHα, making these variables interchangeable' should be rephrased, since Umin is a model parameter; it is the value of Umin inferred from the SED fit that correlates with FHα.","section":"§IV"}],"recommendation":"major_revision","confidential_remarks":"The authors are unusually frank about the weak points of their analysis—the unassessed ΔV uncertainties (Section III.B), the dependence of Umin and γ on the fitted fluxes (Section IV), and the preliminary status of the two-regime kinematic result (Section V)—and the body of the manuscript is honest rather than overclaimed. The gap is between this honest body and the Abstract, which states the CO-luminosity decrease as a flat accompaniment. I see no grounds for rejection: the required fixes (noise-resampled ΔV errors, rank correlations, jackknife or injection robustness tests, and a reframed Section IV or a reworded Abstract) are within the scope of a revision. The main editorial risk is that a reader relies on the Abstract without reading the caveats; a revision that either supplies the robustness tests or demotes the kinematic finding to an explicitly exploratory statement would resolve this. The work fits the journal's scope, and I have no concerns about the citation pattern."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nQuick take: this is a careful, modest archival study that mostly re-derives known correlations between star-formation tracers and floats one genuinely new kinematic claim. The new claim—a two-regime CO luminosity versus velocity-scatter relation—is real but statistically fragile. The decreasing branch at high ΔV rests on about five faint complexes, and the authors themselves say they cannot accurately assess the uncertainty in ΔV and that it may be comparable to ΔV for low-CO-flux objects. So the two-regime result is not established; it is a reasonable preliminary observation that needs proper noise-based error bars before it can carry weight.\n\nThe paper does several things well. It uses archival data from multiple surveys and reduces everything to a common resolution without overclaiming. The authors are explicit that ΔV is not a dispersion, and they discuss the fragility of the 3σ-endpoint definition for weak, multi-peak emission. They also openly admit that their F8/F24 interpretation uses Umin and γ from a prior SED fit to the same IR fluxes, which makes the agreement with the Draine & Li model partly circular. That is the right level of honesty, and it makes the paper trustworthy even where it is not convincing.\n\nThe soft spots are real but proportionate. No correlation coefficients are given for the photometric relations, so “correlates well” is visually supported but not quantified. The F8/F24 anticorrelation is consistent with PAH destruction or excitation changes, but the data do not distinguish the two. The high-ΔV, low-CO complexes are exactly the ones where the endpoint criterion is most likely to be fooled by a noise excursion, and the inverse branch depends on a handful of such points. The abstract's phrasing—that increasing velocity scatter is accompanied by decreasing CO luminosity—goes a bit beyond what the measurements can bear, though the main text is more careful. These are fixable with more quantitative treatment or a more circumspect abstract.\n\nWho is this for? Someone working on ISM kinematics and star-formation tracers in nearby galaxies will find it a useful, honest data point, especially as motivation for better measurements of CO line wings in faint complexes. It is not a major advance, but it deserves a serious referee rather than a desk reject. I would send it out, with the expectation that the kinematic claim needs substantial qualification and ideally a re-analysis with Monte Carlo or bootstrap uncertainties.\n\nMy bottom line: worth refereeing, worth citing cautiously, not worth treating as a settled result.","headline":"A careful, modest archival study whose one genuinely new kinematic result is real but statistically fragile, because the decreasing CO branch rests on a handful of faint complexes with unquantified velocity-scatter errors.","tokens_in":16006,"tokens_out":1571,"would_cite":false,"duration_ms":20519,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Star-forming complexes show two CO kinematic regimes split near 70 km/s.","keywords":["star-forming complexes","velocity scatter","CO kinematics","NGC 628","NGC 2976","NGC 3351","PAH abundance","multi-wavelength photometry"],"falsifier":"Re-observe the apparently high-ΔV complexes in CO at higher sensitivity and velocity resolution: if the inverse branch is genuine, deeper data should show broad, low-column-density CO wings and an absence of bright complexes with ΔV > 70 km/s; if it is an artifact, deeper data should reveal either bright CO emission at high scatter or show that the large endpoint spreads are produced by two unrelated velocity components along the line of sight.","tokens_in":14939,"feed_emoji":"🌌","tokens_out":7672,"duration_ms":74414,"temperature":0.7,"pith_summary":"This paper analyzes star-forming complexes in three nearby galaxies of different morphological types, combining Hα, ultraviolet, infrared, atomic hydrogen, and carbon-monoxide observations to see how gas, dust, and star-formation tracers behave together. The main new claim is that the relation between CO luminosity and internal velocity spread is two-valued: below about 70 km/s, brighter CO emission accompanies larger velocity scatter, but complexes with larger scatter are systematically fainter in CO. The authors interpret this tentative second branch as molecular gas that has been stirred up and partially dissipated by feedback, most plausibly supernova shocks. If confirmed, the result would make broad, faint CO emission a stage indicator in the lives of star-forming complexes rather than simply a tracer of gas mass.","feed_headline":"CO emission fades once gas motions top 70 km/s","feed_subtitle":"Preliminary two-regime signal links fading molecular gas to supernova feedback.","key_machinery":"The central object is the velocity scatter ΔV, computed as the difference between the 3σ endpoints of the CO spectrum for each complex rather than as a fitted dispersion, which lets the authors include faint and multi-peaked sources that previous CO studies excluded. ΔV carries the argument by splitting the sample into two kinematic regimes around 70 km/s. Supporting machinery is the multi-wavelength aperture photometry of the same complexes, reduced to a common spatial resolution, which provides the Hα, ultraviolet, infrared, and atomic-gas fluxes that establish the star-formation context and the CO-brightness comparisons.","core_discovery":"The authors define a velocity scatter ΔV for each star-forming complex as the difference between the velocities at the 3σ endpoints of its CO spectrum, deliberately avoiding Gaussian fits so that weak and multi-peaked emission can be included. In their sample of 95 complexes across three galaxies, the CO flux rises with ΔV up to about 70 km/s and then falls in complexes with larger scatter; the same two-group pattern appears in infrared surface brightness. They propose that the rising branch traces undisturbed molecular gas whose CO luminosity scales with mass, while the falling branch contains complexes whose gas has been disrupted by ongoing star formation, likely through supernova-driven shocks, and is fading in molecular emission. The paper presents this as preliminary, emphasizing that the uncertainty in ΔV cannot be accurately assessed and that weak, multi-peaked spectra make the endpoint criterion fragile.","pith_inferences":["A direct test would be to stack the CO spectra of the high-ΔV complexes: if the branch is real, the average profile should show broad, faint wings, whereas two unrelated velocity components along the line of sight would produce a double-peaked stack instead.","The same two-regime behavior may be visible in atomic hydrogen, and the authors themselves call for HI and Hα kinematic studies; if HI shows the same fading at high scatter, the feedback interpretation would be strengthened independently of CO excitation effects.","If supernova feedback drives the second branch, high-ΔV, low-CO complexes should also show elevated radio continuum or mid-infrared evidence of recent supernova remnants, a correlation the current data do not yet test."],"forward_implications":["If the two-regime relation is real, velocity-broadened molecular complexes in external galaxies can be read as objects in a later, feedback-dominated stage of evolution rather than simply as high-mass gas reservoirs.","Surveys that select complexes by bright CO emission would systematically miss the high-ΔV branch, so complete censuses of star-forming complexes must include faint, extended molecular sources.","The ~70 km/s break provides a concrete threshold that numerical simulations of supernova feedback in giant molecular clouds can be tested against.","A larger sample of galaxies spanning different metallicities and morphologies could show whether the break velocity and the slope of the fading branch vary with environment."],"supporting_citations":[{"why":"Supplies the CO line data cubes from which the velocity scatter and CO fluxes are measured.","marker":"[13]"},{"why":"Defines the star-forming complex sample and the common-resolution photometry used across all bands.","marker":"[14]"},{"why":"Reports a thick molecular disk in a nearby galaxy, supporting the plausibility of large CO velocity spreads.","marker":"[16]"},{"why":"Provides the comparison study of CO velocity dispersions that the new high-scatter complexes exceed.","marker":"[17]"},{"why":"Gives the canonical line-width–size relation used as the reference envelope for the scatter measurements.","marker":"[27]"},{"why":"Underlies the interpretation of the rising CO branch as a mass–luminosity relation.","marker":"[41]"}],"fun_headline_variants":["CO fades when gas velocities top 70 km/s","Star-forming gas dims above 70 km/s scatter","Two-regime CO: bright then fading past 70 km/s","Molecular gas luminosity drops past 70 km/s"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the velocity scatter ΔV, defined through 3σ endpoints of CO spectra, is a meaningful measure of internal kinematics whose uncertainties are small enough to support a clean two-regime split; if the measurement errors are large, or if faint high-ΔV complexes are preferentially kept while bright high-ΔV complexes are removed by the selection cuts, the declining CO branch could be an artifact.","fun_headline_variants_meta":{"raw":{"variants":["CO fades when gas velocities top 70 km/s","Star-forming gas dims above 70 km/s scatter","Two-regime CO: bright then fading past 70 km/s","Molecular gas luminosity drops past 70 km/s"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000763,"raw_usage":{"total_tokens":3385,"prompt_tokens":943,"completion_tokens":2442,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":559,"completion_tokens_details":{"reasoning_tokens":2375}},"tokens_in":559,"tokens_out":2442,"duration_ms":17635,"temperature":1.0,"reasoning_tokens":2375,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:02:31.356506+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-observe the apparently high-ΔV complexes in CO at higher sensitivity and velocity resolution: if the inverse branch is genuine, deeper data should show broad, low-column-density CO wings and an absence of bright complexes with ΔV > 70 km/s; if it is an artifact, deeper data should reveal either bright CO emission at high scatter or show that the large endpoint spreads are produced by two unrelated velocity components along the line of sight.","supporting_citations":[{"cited_title":"velocity scatter","cited_arxiv_id":null,"evidence_quote":"Supplies the CO line data cubes from which the velocity scatter and CO fluxes are measured."},{"cited_title":"J., Miura, R","cited_arxiv_id":null,"evidence_quote":"Defines the star-forming complex sample and the common-resolution photometry used across all bands."},{"cited_title":"2010, Astrophys","cited_arxiv_id":null,"evidence_quote":"Reports a thick molecular disk in a nearby galaxy, supporting the plausibility of large CO velocity spreads."},{"cited_title":"Walter et al., 2008, Astron","cited_arxiv_id":null,"evidence_quote":"Provides the comparison study of CO velocity dispersions that the new high-scatter complexes exceed."},{"cited_title":"S., & Efremov, Y","cited_arxiv_id":null,"evidence_quote":"Gives the canonical line-width–size relation used as the reference envelope for the scatter measurements."},{"cited_title":"G., Dwek, E., Blair, W","cited_arxiv_id":null,"evidence_quote":"Underlies the interpretation of the rising CO branch as a mass–luminosity relation."}],"review_version":1}