{"id":"1e81129f-eff7-4956-8062-5d62633fbdf6","arxiv_id":"2506.06149","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"A deep OH absorption survey toward four continuum sources confirms OH as a widespread molecular gas tracer, yields median abundances of about 1.2e-7 relative to H2, and finds only one CO-dark component in 23.","lead":"Deep radio observations of hydroxyl absorption toward four bright Milky Way sources show that the molecule appears in both diffuse and dense gas, with only one of 23 components lacking a carbon monoxide counterpart. The study refines OH as a molecular gas tracer and suggests that finding CO-dark gas with OH will require even deeper observations.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The linear N_OH-N_H2 slope and OH abundances rest on fixed alpha_CO and R21; environment-dependent variation, especially at low N_H2, could shift m_H2 by more than its quoted 0.1 uncertainty.","rationale":"The paper's headline results—N_OH ∝ N_H2^1.0±0.1 and X_OH,H2 ≈ 1.2e-7—are not direct measurements against an observed H2 column; they are measurements of N_OH against 12CO intensity multiplied by fixed constants. A constant offset in these constants would only change the abundance zero-point and is already inside the factor-of-2 error bars, so that alone is not fatal. The load-bearing risk is a gradient: alpha_CO is expected to rise in the diffuse, low-column gas this paper intentionally targets, and the authors concede this in Sect. 5. Because the fitted range spans ~2.5 dex in N_H2, even a factor-of-2 environmental variation can move the slope by an amount comparable to the quoted 0.1 uncertainty, and larger variations would dominate. The same argument applies to R21, which is checked directly for only two of the four sightlines. This is exactly the reader's weakest_assumption, and the present pass sharpens it from 'alpha_CO may vary' to 'the slope and abundance are conditional on the absence of a systematic alpha_CO/R21 gradient.' Other issues—the small sample, the 'never exclusive' phrasing, and HI saturation—affect secondary or interpretive statements rather than the core abundance and correlation results, which is why the reader's CONDITIONAL verdict remains appropriate pending the proposed sensitivity check.","tokens_in":26096,"tokens_out":16893,"duration_ms":170737,"concrete_test":"Extract FUGIN 12CO(1-0) spectra at the positions and velocity intervals of Table A.1 for the two sources with APEX 12CO(2-1) data (G29.935−0.053, G31.388−0.383) and measure R21 = ∫Tmb(2-1)dv / ∫Tmb(1-0)dv for each OH absorption feature. If R21 deviates from the adopted 0.7 systematically with CO intensity (e.g., lower in fainter, more diffuse features), recompute N_H2 with the measured R21 and refit the power-law slope m_H2 and median abundance X_OH,H2. If the slope shifts by more than 0.1 or the abundance by more than 0.15 dex, the fixed R21 (and by extension fixed alpha_CO) is load-bearing for the central claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that N_OH is linearly proportional to N_H2 with slope 1.0±0.1 and median abundance X_OH,H2=1.2e-7. These values are derived by converting 12CO emission to H2 using fixed alpha_CO=2e20 cm^-2 (K km/s)^-1 and R21/10=0.7 (Sect. 3.2). A constant miscalibration is absorbed by the quoted factor-of-2 error and is not the issue. The load-bearing risk is that alpha_CO or R21 varies systematically with environment, particularly in the diffuse gas this paper targets. In the low-column regime (N_H2 ~ 8e19-1e21 cm^-2), CO is expected to be subthermally excited and photodissociated, raising alpha_CO; the authors themselves state alpha_CO 'may be significantly underestimated at N_H2=0.8e20 cm^-2' and that undetected CO-dark gas 'would imply slightly steeper relations' than m_H2=1.0 (Sect. 5). Since the fit spans ~2.5 dex in N_H2, a factor-of-2 gradient in alpha_CO across this range changes the fitted slope by ~0.1, equal to the quoted uncertainty, and a larger or steeper gradient would shift both the slope and the abundances beyond the quoted errors. Thus the reported linearity is partly an assumption about the constancy of the CO conversion, not a direct measurement of N_OH versus true H2 column.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents new, high-sensitivity VLA observations of the four 18 cm ground-state hyperfine-splitting transitions of OH in absorption toward four continuum sources, complemented by VLA HI 21 cm absorption, APEX CO (2–1) emission, and SOFIA upGREAT [CII] 158 μm and [OI] 63 μm spectra. The authors derive OH column densities from the 1667 MHz line (assuming T_ex = 5 K), HI column densities from 21 cm absorption (T_spin = 100 K), and H2 column densities primarily from 12CO emission using a fixed alpha_CO = 2e20 cm^-2 (K km/s)^-1 and R_21/10 = 0.7. Comparing N_OH with N_H2 and N_H for their new sample combined with THOR data, they fit power-law relations and report N_OH ∝ N_H2^1.0±0.1 and N_OH ∝ N_H^1.2±0.1, with median abundances X_OH,H2 = 1.2e-7 and X_OH,H = 4.8e-8. They detect only one OH absorption component out of 23 without a CO counterpart, and report a tentative [CII] association toward one sightline plus excitation anomalies in W43-South.","tokens_in":26477,"tokens_out":8701,"duration_ms":80066,"significance":"The observations are a clear improvement over the THOR survey in both sensitivity (factor of 5) and spectral resolution (0.1–0.2 km/s), allowing the detection of narrow, faint OH absorption features down to N_OH = 3.7e13 cm^-2. If the reported linear N_OH–N_H2 relation and abundances are robust, they provide useful constraints on OH chemistry and on the utility of OH absorption as a molecular-gas tracer across diffuse and dense environments. The paper is commendably transparent about its assumptions, and Table A.1 provides a complete set of measured integrals that will enable future modeling. However, the central quantitative claims—especially the slope m_H2 = 1.0±0.1—rest on the assumed constancy of the CO-to-H2 conversion, and the paper does not quantify how plausible environmental variations of alpha_CO would affect that slope.","major_comments":[{"comment":"The central result that N_OH is linearly proportional to N_H2 with slope m_H2 = 1.0±0.1 is derived from N_H2 values obtained with a fixed alpha_CO = 2e20 cm^-2 (K km/s)^-1 and a fixed R_21/10 = 0.7. The quoted ±0.1 uncertainty is the statistical fit uncertainty only; it does not include the systematic effect of a possible environmental dependence of alpha_CO or R_21/10. In the diffuse, low-N_H2 regime specifically targeted by this paper (N_H2 ~ 8e19–1e21 cm^-2), alpha_CO is expected to be larger than the adopted value, and the authors themselves note in Sect. 5 that alpha_CO may be significantly underestimated at N_H2 = 0.8e20 cm^-2 and that CO-dark gas would imply steeper relations. A factor-of-two gradient in alpha_CO across the ~2.5 dex range of the fit changes the fitted slope by ~0.1, equal to the quoted uncertainty, and a larger gradient would move both the slope and the abundances beyond the quoted errors. The authors' counterargument in Sect. 5—that no trend of abundance with Galactocentric radius is seen—does not directly constrain a trend with N_H2. I therefore recommend that the paper either quantify the systematic uncertainty on m_H2 arising from plausible alpha_CO and R_21/10 variations (e.g., by refitting with an alpha_CO that increases at low N_H2 or by using an independent H2 tracer such as dust), or explicitly state that the linear slope is conditional on a constant conversion factor.","section":"Sect. 3.2 and 4.2, Fig. 3"},{"comment":"The conclusion that N_OH is linearly correlated with the total hydrogen column density N_H with exponent m_H = 1.2±0.1 is conditional on an assumption about saturated HI absorption. For the 12 non-saturated data points alone, the Spearman rank correlation has r_s = 0.6 with p = 0.08, meaning the null hypothesis of no correlation cannot be rejected at the current sample size. The m_H = 1.2±0.1 value is obtained only after assuming that N_H2 dominates in the saturated cases, and this important caveat is not carried into the conclusion in Sect. 6, where the relation is stated without qualification. The authors should either rephrase this as a model-dependent inference supported by the saturation assumption, or add a clear statement of the significance of the correlation on the unsaturated subsample.","section":"Sect. 4.2, Fig. 3, Sect. 6"},{"comment":"The combined sample used for the power-law fits does not rely on a single, homogeneous N_H2 estimator. For most line-of-sight features, N_H2 is derived from 12CO emission with a constant alpha_CO, whereas for features associated with HII regions (e.g., the 97.8 and 102.8 km/s components toward G29.957−0.018, and the 99.0 km/s component toward G29.935−0.053) N_H2 is adopted from Rugel et al. (2018), who used 13CO(1–0) with an assumed excitation temperature, an optical-depth correction, and an isotope-ratio rescaling. The authors themselves find in Sect. 3.2 that 13CO-based N_H2 is systematically lower by a factor of two or more compared to 12CO-based values for most columns. Since the HII-region points occupy the high-N_H2 end of the fitted range, a systematic offset or nonlinearity in the 13CO-based calibration could bias the fitted slope. I ask the authors to test the robustness of m_H2 by fitting only the 12CO-based sample, or by demonstrating that the slope is unchanged when the 13CO-based points are excluded.","section":"Sect. 3.2 and Table A.1"}],"minor_comments":[{"comment":"There is a duplicated word in the sentence describing the upGREAT instrument: \"which observes simultaneously and and comprises\" should read \"which observes simultaneously and comprises\".","section":"Sect. 2.2"},{"comment":"The phrase \"This is holds true if we assume a lower spin temperature for N_Hi\" is grammatically incorrect; it should be \"This holds true if we assume...\".","section":"Sect. 4.2"},{"comment":"The word \"extra-galactic\" is used several times (e.g., in the abstract and in Sect. 2.1); the standard astronomical spelling is \"extragalactic\".","section":"Throughout"},{"comment":"In the sentence \"...it is over-posed by potential absorption, baseline ripples, and Galactic emission,\" the word \"over-posed\" appears to be a typo; \"overlapped\" or \"dominated\" would convey the intended meaning.","section":"Sect. 4.4"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid observational contribution with well-documented data and a transparent discussion of assumptions. The main issue is that the headline slope m_H2 = 1.0±0.1 is more sensitive to the CO-to-H2 conversion than the quoted precision suggests, and the authors' existing discussion acknowledges the bias without quantifying its effect on the central claim. I would ask the authors to run the robustness tests outlined in the major comments and to soften the language of the conclusions where the supporting evidence is conditional. The number of typos and minor grammatical errors is also higher than typical for A&A; a careful language edit is recommended."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Solid, careful follow-up paper. The genuinely new content is the data: factor-of-5 deeper VLA OH spectra at 0.1 km/s toward four THOR continuum sources, plus new SOFIA [CII] and APEX CO (2-1) for two of them. That lets the authors resolve narrow, weak OH components down to N_OH ~ 3.7e13 cm^-2 and extend the N_OH-N_H2 relation about an order of magnitude below the THOR sample. The one CO-dark component out of 23 and the marginal [CII] association toward G31.388-0.383 are modest but honestly presented. The W43-South satellite/main-line offset is a nice resolved observation, even if the interpretation leans on Hafner et al.\n\nThe analysis is standard and transparent: line-to-continuum optical depths, fixed T_ex=5 K and T_spin=100 K with stated factor-of-two uncertainties, orthogonal regression that accounts for errors in both variables. The derived OH abundances agree with Liszt & Lucas and other prior work. Credit where due: the paper states its caveats rather than burying them, and the choice to use 12CO rather than 13CO for N_H2 is explicitly motivated.\n\nThe soft spot is exactly what the stress-test flags. The headline slope m_H2=1.0±0.1 and the abundances rest on constant alpha_CO=2e20 and R21/10=0.7. The authors acknowledge a factor-two uncertainty, and they even note that alpha_CO may be underestimated at N_H2=0.8e20 and that CO-dark gas would steepen the relations. That admission is the load-bearing issue: a factor-of-two gradient in alpha_CO across the 2.5-dex range would shift the fitted slope by about 0.1, comparable to the quoted error, and a larger gradient would move both slope and abundances beyond the formal errors. So the linearity is partly an assumption about conversion constancy, not a direct measurement of N_OH against true H2 column. I do not think this kills the paper; the abundance range and the qualitative conclusion that OH traces molecular gas across environments survive. But the precision on the slope is softer than the quoted ±0.1. Also, \"never be an exclusive tracer\" is a strong sentence for four sightlines; the 1/23 CO-dark fraction is consistent with prior work, but target selection toward weak-CO sources deserves more quantification.\n\nThe [CII] column estimate depends on assumed T_kin and n, and the 18 km/s association is plausible but not unambiguous. Minor relative to the above.\n\nBottom line: useful, honest dataset with carefully caveated conclusions. It deserves a serious referee. The main revision is to reframe the slope as conditional on constant alpha_CO and R21/10, and to soften or better support the exclusivity claim. I would cite this for the OH abundance values and the new low-column detections. Send it to review; it is a conditional-accept paper, not a desk reject.","headline":"A solid, transparent observational follow-up that pushes OH absorption to lower column densities with genuinely new data, but whose central slope and abundance claims are partly conditional on a constant CO-to-H2 conversion factor.","tokens_in":860,"tokens_out":2057,"would_cite":true,"duration_ms":36736,"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":"New deep 18-cm OH observations show hydroxyl column density scales linearly with molecular hydrogen across diffuse and dense gas.","keywords":["OH 18 cm lines","interstellar medium","CO-dark gas","molecular gas tracers","Galactic disk","hydroxyl abundance","absorption spectroscopy","W43-South"],"falsifier":"Measure $N_{\\rm H_2}$ toward these sightlines independently of CO, e.g., from dust optical depth or HI self-absorption, and re-fit $N_{\\rm OH}$ vs $N_{\\rm H_2}$; if the slope departs from $1.0$ by more than the quoted uncertainties, the linear-abundance claim fails.","tokens_in":25915,"feed_emoji":"📡","tokens_out":4433,"duration_ms":38629,"temperature":0.7,"pith_summary":"The paper sets out to determine whether the 18-cm ground-state hyperfine transitions of hydroxyl (OH) can serve as a reliable tracer of molecular hydrogen, especially in gas where carbon monoxide emission is weak or absent. Using deep VLA absorption spectra toward four bright continuum sources, combined with [CII] 158 $\\mu$m, HI 21 cm, and CO data, the authors measure OH column densities down to $3.7\\times10^{13}\\,\\mathrm{cm^{-2}}$ and compare them with H$_2$ and total hydrogen column densities. They find $N_{\\rm OH}$ is linearly proportional to $N_{\\rm H_2}$ (power-law exponent $1.0\\pm0.1$) and to total hydrogen column density $N_{\\rm H}$ ($1.2\\pm0.1$), with median abundances $X_{\\rm OH,H_2} = 1.2\\times10^{-7}$ and $X_{\\rm OH,H} = 4.8\\times10^{-8}$. The result matters because it would make OH absorption a quantitative tracer of molecular gas across diffuse and dense environments, while the detection of only one CO-dark feature out of 23 shows OH is not an exclusive tracer of CO-dark gas.","feed_headline":"OH absorption tracks molecular gas 1-to-1 across the Milky Way disk","feed_subtitle":"Deep 18-cm data show hydroxyl column density scales linearly with molecular hydrogen.","key_machinery":"The load-bearing object is the ground-state hyperfine-splitting quartet of OH at 18 cm: the 1612, 1665, 1667, and 1720 MHz transitions. The 1667 MHz line is used for column densities through $N_{\\rm OH}/T_{\\rm ex} = C_0/f \\int \\tau\\,dv$ (with $C_0 = 2.24\\times10^{14}\\,\\mathrm{cm^{-2}\\,K^{-1}\\,km^{-1}\\,s}$) assuming $T_{\\rm ex} = 5\\,\\mathrm{K}$ and filling factor $f = 1$. Molecular hydrogen columns are derived from $^{12}$CO emission with a fixed conversion factor $\\alpha_{\\rm CO} = 2\\times10^{20}\\,\\mathrm{cm^{-2}}\\,(\\mathrm{K\\,km\\,s^{-1}})^{-1}$ and $R_{21/10} = 0.7$. The linear fits are made in log space with the York (1966) method, accounting for errors in both coordinates.","core_discovery":"The central claim is that OH 18 cm absorption traces molecular gas linearly across environments spanning column densities $N_{\\rm H_2} \\sim 7.9\\times10^{19}$ to $4.7\\times10^{22}\\,\\mathrm{cm^{-2}}$. Combining new observations with THOR data, the authors fit a power law $N_{\\rm OH} \\propto N_{\\rm H_2}^{1.0\\pm0.1}$ and $N_{\\rm OH} \\propto N_{\\rm H}^{1.2\\pm0.1}$, and quote median abundances $X_{\\rm OH,H_2} = 1.2^{+0.3}_{-0.2}\\times10^{-7}$ and $X_{\\rm OH,H} = 4.8^{+0.9}_{-0.8}\\times10^{-8}$. They interpret the single CO-free absorption component and several intermediate molecular gas fractions as evidence that OH is present in both diffuse and dense gas; they conclude OH can trace molecular gas but never exclusively CO-dark gas.","pith_inferences":["If the linear relation holds at still lower columns, OH absorption could probe the atomic-to-molecular transition itself, where CO fails.","Testing $N_{\\rm H_2}$ via dust optical depth or HCO$^+$ on the same sightlines would show whether the fixed $\\alpha_{\\rm CO}$ assumption is responsible for the scatter.","The apparent lack of Galactocentric abundance trend suggests OH formation is robust to variations in radiation field, which could make OH a stable tracer across the disk."],"forward_implications":["OH 18 cm absorption can be used as a quantitative column-density tracer for molecular gas across diffuse and dense ISM environments.","The single CO-dark detection out of 23 means future surveys need deeper integrations or brighter background sources to find CO-dark gas in absorption.","The linear $N_{\\rm OH}$--$N_{\\rm H_2}$ relation supports using OH as a complementary cross-check on CO-based H$_2$ masses.","The abundance values provide observational anchors for chemical models of diffuse molecular clouds.","The satellite-line offsets toward W43-South support the expanding-shell scenario for OH inversion features around HII regions."],"supporting_citations":[{"why":"Supplies the $\\alpha_{\\rm CO}$ conversion factor used to derive molecular hydrogen column densities.","marker":"Bolatto et al. 2013"},{"why":"Provides the THOR survey OH absorption sample that this paper extends with deeper and higher-resolution data.","marker":"Rugel et al. 2018"},{"why":"Establishes that OH absorption tightly correlates with HCO$^+$, supporting OH as a molecular gas tracer.","marker":"Liszt & Lucas 1996"},{"why":"Found a linear relation between OH emission and dust extinction in Taurus, prior evidence for OH tracing molecular gas.","marker":"Xu et al. 2016"},{"why":"Recent OH absorption analysis that motivates the assumed excitation temperature of $T_{\\rm ex}=5$ K.","marker":"Hafner et al. 2023"},{"why":"Provides the least-squares fitting method that accounts for errors in both variables, used for the power-law slopes.","marker":"York 1966"}],"fun_headline_variants":["OH absorption tracks molecular gas linearly in the Galactic disk","OH 18-cm lines reveal linear molecular gas scaling","Hydroxyl absorption is a linear molecular gas probe","OH absorption: linear correlation with molecular hydrogen","Milky Way OH absorption linearly traces molecular gas"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The H$_2$ column densities rest on a single CO-to-H$_2$ conversion factor ($\\alpha_{\\rm CO} = 2\\times10^{20}\\,\\mathrm{cm^{-2}}\\,(\\mathrm{K\\,km\\,s^{-1}})^{-1}$ and $R_{21/10} = 0.7$); if that factor varies with environment, especially in diffuse gas, the quoted slopes and abundances would shift.","fun_headline_variants_meta":{"raw":{"variants":["OH absorption tracks molecular gas linearly in the Galactic disk","OH 18-cm lines reveal linear molecular gas scaling","Hydroxyl absorption is a linear molecular gas probe","OH absorption: linear correlation with molecular hydrogen","Milky Way OH absorption linearly traces molecular gas"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000703,"raw_usage":{"total_tokens":3251,"prompt_tokens":1101,"completion_tokens":2150,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":717,"completion_tokens_details":{"reasoning_tokens":2076}},"tokens_in":717,"tokens_out":2150,"duration_ms":17223,"temperature":1.0,"reasoning_tokens":2076,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:59:29.518298+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure $N_{\\rm H_2}$ toward these sightlines independently of CO, e.g., from dust optical depth or HI self-absorption, and re-fit $N_{\\rm OH}$ vs $N_{\\rm H_2}$; if the slope departs from $1.0$ by more than the quoted uncertainties, the linear-abundance claim fails.","supporting_citations":[{"cited_title":"R., Beuther , H., Bihr , S., et al","cited_arxiv_id":null,"evidence_quote":"Provides the THOR survey OH absorption sample that this paper extends with deeper and higher-resolution data."},{"cited_title":"& Lucas , R","cited_arxiv_id":null,"evidence_quote":"Establishes that OH absorption tightly correlates with HCO$^+$, supporting OH as a molecular gas tracer."},{"cited_title":"R., Nguyen , H., et al","cited_arxiv_id":null,"evidence_quote":"Recent OH absorption analysis that motivates the assumed excitation temperature of $T_{\\rm ex}=5$ K."},{"cited_title":"1966, Canadian Journal of Physics, 44, 1079","cited_arxiv_id":null,"evidence_quote":"Provides the least-squares fitting method that accounts for errors in both variables, used for the power-law slopes."}],"review_version":1}