{"id":"47157afd-83bc-4d4a-b153-f88834aedc1e","arxiv_id":"2607.24544","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.5,"correctness_risk":"low","formal_verification":"none","parameter_count":3,"one_line_summary":"Fifteen plus one serendipitous clear H-alpha bow shocks are detected among 78 IR candidates; several are consistent with radiation-supported regimes while others match classical wind-supported shocks.","lead":"A targeted H-alpha imaging survey of 78 infrared bow-shock candidates around OB stars finds clear ionized arcs in 16 systems. It shows that both stellar winds and radiation pressure can shape these structures and that optical and mid-IR morphologies differ systematically.","discovery_kind":"new_application","skeptic_critique":{"model":"moonshotai/kimi-k3","headline":"The WBS/RBS/RBW/DW classifications in Table 4 — on which the claim that \"radiation pressure can contribute\" rests — are weakly constrained: the adopted density bins span factors of 50–100, and the fallback rule of assigning \"the nearest higher theoretical value\" biases ambiguous cases toward the RBS","rationale":"The reader identified the right pressure point — the hand-chosen density bins and the known Ṁ overestimate — and correctly kept the verdict at CONDITIONAL with high confidence. My analysis sharpens the same concern: it is not only that the inputs are uncertain, but that (i) the wide bins make the regime comparison nearly unconstraining and (ii) the stated fallback assignment rule interacts with the Ṁ overestimate in a direction that systematically favours radiation-supported classes. This is a correctness-risk issue internal to the paper's own procedure, not a disagreement with consensus, and it is directly checkable by recomputation from Table 4 inputs. It does not undermine the survey's primary empirical results (15 clear Hα detections, Hα–IR offsets, non-detection statistics), which are the genuinely novel contribution and are presented with transparent methods, a full observing log, and honest caveats about reflection-nebula contamination. Because the paper already hedges the classification as \"consistent with\" and calls for spectroscopic follow-up, the appropriate response is to keep the CONDITIONAL verdict, with the condition that Table 4 classes be regarded as provisional and stress-tested as described. No verdict adjustment is warranted; the recommended Monte Carlo reclassification would settle whether the radiation-support claim lands for individual objects or only for the sample as a whole.","tokens_in":38512,"tokens_out":1736,"duration_ms":67713,"concrete_test":"Recompute the Table 4 classifications with a stability analysis, using only published inputs: (a) draw n_ISM uniformly in log within each object's adopted bin (and, for the ten objects in Table 3, substitute the EM-derived n_ISM with its uncertainty instead of the bin); (b) scale the Vink Ṁ down by the weak-wind factor appropriate to spectral type (10–100× for late-O/B dwarfs, per Puls et al. 2008 / Marcolino et al. 2009); (c) for each of ~10^3 Monte Carlo draws, re-solve Eqs. 5–8 and re-assign the regime without the \"nearest higher value\" fallback (i.e., flag objects whose R0 matches no regime). Report the fraction of draws in which each object keeps its Table 4 class. If objects classified RBS/RBW/DW retain that class in <~50% of draws, the radiation-support claim should be downgraded to a statement about the ensemble rather than individual systems.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The survey's detection statistics and Hα–IR morphology comparison are direct observables and appear sound; the authors are also commendably transparent about continuum/reflection-nebula ambiguities (five of the sixteen detections, incl. K065, K692, BB1, N1, EB23, remain unconfirmed against DSS2 blue). The load-bearing soft spot is in §4.3/Table 4, which supports the second half of the central claim (radiation vs. wind support). Three issues compound: (1) The adopted n_ISM bins (0.1–5 vs 1–100 cm⁻³, chosen from the local Hα appearance) are each a factor of 50–100 wide. Since predicted stand-off distances scale as R0 ∝ n_ISM^(-1/2) in both the R1 (Eq. 5) and WBS limits, each regime's predicted R0 range spans a factor of ~7–10, so the three regime bands overlap or nearly tile the plausible R0 axis — the comparison has little discriminatory power; almost any measured R0 falls inside some regime's range. (2) When R0 falls outside all ranges, the authors \"adopt the nearest higher theoretical value\" to assign the class. Because predicted R0 values are upper envelopes and observed R0 is a projected lower limit, this fallback preferentially assigns the regime with the smallest prediction exceeding R0 — which, given that WBS predictions are inflated by Vink et al. (2001) Ṁ values (up to 100× too high for late-O/B dwarfs, as the authors themselves note for ζ Oph/EB21, where the WBS prediction drops from 4.10–0.48 pc to 0.33–0.05 pc), will systematically be an RBS/RBW/DW class rather than WBS. So borderline objects are pushed toward radiation support by construction, not by evidence. (3) The EM-derived densities in Table 3 (~4–62 cm⁻³) sit inside the adopted bins, so the bins are not independently validated. The claim survives only in its weakest form (\"several systems are consistent with...\"), and the per-object classes in Table 4 should be treated as illustrative rather than established.","agreement_with_reader":"agree"},"referee_report":{"model":"moonshotai/kimi-k3","summary":"The authors present the first targeted narrow-band Hα imaging survey of 78 infrared bow-shock candidates around OB-type stars, selected from the van Buren, Brown & Bomans, E-BOSS, and Kobulnicky catalogues and observed with five ground-based facilities, complemented by archival Spitzer/WISE imaging. They report clear arc-shaped Hα nebulae in 15 objects plus one serendipitous detection (N1), diffuse background emission in 35, and non-detections in 28. For the 16 detections they measure arc geometry (stand-off distance R0, width, length, eccentricity) in both Hα and MIR, finding IR arcs systematically closer to the star and broader than the Hα arcs, only partly attributable to resolution differences. For ten objects they derive SHASSA-calibrated Hα surface brightnesses, emission measures, EM-based ISM density estimates, and predicted 6 GHz free-free brightnesses. Finally, they classify the bows into the Henney & Arthur (2019) regimes (WBS/RBS/RBW/DW) by comparing measured stand-off distances to theoretical predictions over adopted ISM density ranges, concluding that both wind momentum and radiation pressure can support these structures.","tokens_in":38925,"tokens_out":3999,"duration_ms":145813,"significance":"This is the first targeted Hα imaging survey of infrared bow-shock candidates around OB stars, and it fills a genuine observational gap: prior optical work rested on shallow all-sky data (SHASSA/VTSS). The paper's strengths are concrete and useful to the community: a homogeneous 78-target sample drawn from the major IR catalogues; measured arc geometries in both Hα and MIR with repeated-fit uncertainties (Table 2); SHASSA-cross-calibrated surface brightnesses with a fully documented procedure and conservative error budget (Appendix C); EM-based density estimates; and predicted 6 GHz free-free brightnesses that provide directly falsifiable targets for radio follow-up (the ζ Oph prediction of ~10 kJy/sr is consistent with the published non-detection, a nice sanity check). The authors are commendably transparent about the reflection-nebula ambiguity for five detections and about the Vink mass-loss overestimate. If the regime classifications were made robust, the conclusion that radiation pressure can contribute to supporting some of these structures would be a meaningful addition to the Henney & Arthur framework's observational tests.","major_comments":[{"comment":"The WBS/RBS/RBW/DW assignments rest on adopted n_ISM bins of 0.1–5 and 1–100 cm^-3. Since both the R1 (Eq. 5) and WBS limits scale as R0 ∝ n_ISM^(-1/2), each regime's predicted R0 band spans a factor of ~7–10, and adjacent bands overlap or nearly tile the plausible R0 axis (e.g., EB01: RBS 13.98–1.98 pc vs WBS 3.19–0.45 pc). The comparison therefore has limited discriminatory power: most measured R0 values fall inside at least one regime's range, and several rows receive dual labels (RBS/WBS, DW/WBS). Since the second half of the central claim (radiation pressure contributing to support) rests on these assignments, the authors should quantify the robustness: e.g., report for how many objects the class is invariant across the full density bin, and state explicitly how many of the 'radiation-supported' assignments would change if the bins were narrowed using the EM-derived densities of Tab","section":"§4.3 and Table 4, cols. 10–15"},{"comment":"When measured R0 falls outside all predicted ranges, the class is assigned by 'the nearest higher theoretical value.' Two asymmetries make this rule non-neutral: (i) observed R0 is a projected lower limit (Tarango-Yong & Henney 2018), so a mismatch preferentially indicates a too-small prediction; (ii) WBS predictions are computed with Vink et al. (2001) mass-loss rates, which the authors themselves note can overestimate Ṁ by up to two orders of magnitude for late-O/B dwarfs (their own ζ Oph/EB21 example drops the WBS prediction from 4.10–0.48 to 0.33–0.05 pc). Because WBS predictions are thus upper envelopes while RBS/DW predictions are largely Ṁ-independent, the fallback rule can systematically funnel ambiguous cases into radiation-supported classes. The authors should either justify the rule with a worked sensitivity test (recompute Table 4 with clumping-corrected Ṁ for the late-O/B dw","section":"§4.3, paragraph following the adopted density ranges"},{"comment":"Five of the sixteen detections (K065, K692, BB1, N1, EB23; plus EB21 unassessable due to saturation) cannot be distinguished from reflection nebulae or strong [O III] emission with the present data, as the authors candidly state. However, the abstract and §6 report 'clear arc-shaped Hα nebulae are detected in 15 objects' without this caveat, and Table 2/Table 4 do not flag the unconfirmed subset. Since the detection fraction is a headline result, the unconfirmed objects should be marked in the tables and the abstract/conclusions should state the confirmed fraction (10/78) separately from the candidate fraction (16/78).","section":"§3.1, §5 (DSS2 paragraph), abstract, and Table 2/Table 4"}],"minor_comments":[{"comment":"The text following Eq. (8) contains a sentence fragment: '...and needed to cover all cases from optically thin to optically thick shells.' Please rephrase.","section":"§4.3, after Eq. (8)"},{"comment":"R0 uncertainties are quoted as one pixel, yet several entries imply much higher precision (e.g., EB21 Hα R0 = 290.567″ on SBT images with 3.14″/pixel; EB06 R0 = 790.30″). Values should be rounded consistently with the stated uncertainty.","section":"Table 2"},{"comment":"For ζ Oph the text quotes the Gvaramadze et al. (2012) distance of 112 pc, while Table 1 adopts 0.13 ± 0.01 kpc. Please reconcile or note the difference, since derived physical sizes are quoted to two decimals.","section":"§5 (ζ Oph paragraph) vs Table 1"},{"comment":"The notation ν^(-0.1)_GHz is ambiguous; please write the frequency dependence explicitly (e.g., (ν/GHz)^(-0.1)).","section":"§4.2, Eq. (4)"},{"comment":"Typos: 'bow-schock' (§5, λ Cep paragraph); 'allsky' (§2); 'comprises of' (§2); 'discernable' (§3.1); 'laying' (§4.2).","section":"various"},{"comment":"Binary fractions (~58% of 13 objects; ~26% of 19) are quoted without uncertainties. Given the small samples, these are consistent with a wide range; either add binomial errors or soften the comparison between subsamples.","section":"§5, Fig. 5 discussion"},{"comment":"κ = 600 cm² g⁻¹ is described as 'total opacity (gas plus dust)'; since τ1 enters Eq. (8) and the regime boundaries, please state explicitly that this is a UV-band opacity and comment on its uncertainty (~factor 2), which translates directly into τ1.","section":"§4.3, definition of κ"},{"comment":"The two 'Van den Eijnden et al. 2022' papers (MNRAS 510, 515 and MNRAS 512, 5374) are cited with different capitalization ('Van den' vs 'van den'); please make the in-text citation style consistent so they are distinguishable (2022a/b).","section":"§5 and references"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is well within the journal's scope and the survey aspect is solid, publishable work. The one part I find weak is the regime-classification analysis of §4.3, which currently carries a headline claim in the abstract that the data do not strongly support. I do not believe new observations are needed; a robustness analysis and more careful language should suffice, which is why I recommend major rather than minor revision — the load-bearing second aim of the paper is affected."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The real news here is the survey itself. They took 78 IR-selected bow-shock candidates, got multi-facility Hα imaging, and came back with 15 clear arcs plus one serendipitous detection, plus clean non-detections and a large diffuse category. That is the first sizable optical sample after Brown & Bomans 2005 and a handful of single-object papers. The Hα–IR comparison is useful: same large-scale shapes, but IR arcs usually sit slightly closer in and look broader, and they show the resolution convolution does not fully erase the offset. SHASSA-calibrated surface brightnesses, EMs, and rough 6 GHz predictions are carefully done and will help radio people.\n\nThey apply the Henney & Arthur regimes honestly and flag the two biggest problems themselves: Vink mass-loss rates can be orders of magnitude high for late-O/B dwarfs (they walk through ζ Oph explicitly), and ambient densities are only known in broad bins. The stress-test note is right that those bins are so wide the predicted R0 ranges nearly tile the axis, and the “nearest higher theoretical value” fallback plus inflated WBS predictions will push borderline cases toward radiation support. So Table 4 classes should be read as “consistent with,” not as firm IDs. The abstract and conclusions already use that weaker language, which keeps the paper honest.\n\nMinor but real: five of the sixteen Hα arcs still have possible reflection or [O III] contamination on DSS2 blue; they say so. No correlation with stellar parameters is a null result worth reporting. Overlap with radio is too small for strong claims, and they do not overclaim it.\n\nThis is for people working on runaway-star feedback, bow-shock physics, or multi-wavelength follow-up. The observational core is reproducible from the figures, tables, and log. I would send it to referees without hesitation; the classification section just needs the provisional language kept or tightened. I would cite the detection statistics and morphology offsets; I would treat the individual regime labels as provisional pending spectroscopy and better Ṁ/nISM.","headline":"Solid first targeted Hα survey of IR bow shocks; detection stats and morphology offsets hold up, while the radiation-vs-wind classifications are only illustrative.","tokens_in":39726,"tokens_out":529,"would_cite":true,"duration_ms":18308,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"About one in five infrared bow shocks around massive stars also lights up in H-alpha, and radiation pressure can help hold them up.","keywords":["bow shocks","OB stars","H-alpha imaging","stellar winds","radiation pressure","interstellar medium","runaway stars","mid-infrared nebulae"],"falsifier":"Optical spectroscopy that measures density, temperature, and kinematics of the H-alpha arcs, or independent ambient-density and mass-loss measurements that move the predicted stand-off distances out of the assigned regime.","tokens_in":39473,"feed_emoji":"☄️","tokens_out":978,"duration_ms":18058,"temperature":0.7,"pith_summary":"Massive stars ploughing through space often leave arc-shaped dust glows that show up in mid-infrared surveys. Whether those arcs also contain glowing ionized gas has been almost untested. This paper images 78 infrared bow-shock candidates in the hydrogen-alpha line and finds clear arc-shaped nebulae in 15 of them, plus one new discovery by chance. Another 35 sit in messy background glow, and 28 stay dark. The optical and infrared arcs usually follow the same overall shape, but the dust arcs tend to sit a bit closer to the star and look broader. Comparing measured stand-off distances with stellar wind and radiation models shows that some systems fit classical wind-supported shocks while others look more like radiation-supported bow waves or dust waves. The result matters because it shows both wind momentum and starlight can shape these structures, and it supplies concrete targets and brightness estimates for follow-up spectroscopy and radio work.","feed_headline":"One in five infrared star bow shocks also glow in H-alpha","feed_subtitle":"Survey of 78 massive-star arcs finds ionized gas and a role for radiation pressure, not only winds","key_machinery":"Morphological classification via stand-off distance: observed arc–star separations are compared with theoretical R0 values for wind-supported bow shocks, radiation-supported bow shocks, radiation-supported bow waves, and dust waves, using stellar luminosity, wind momentum efficiency, and adopted ambient ISM density ranges.","core_discovery":"A targeted H-alpha survey of 78 infrared bow-shock candidates around OB stars detects clear arc-shaped ionized nebulae in 15 objects (plus one serendipitous find). The H-alpha and mid-infrared morphologies trace the same large-scale structures, yet infrared arcs usually lie slightly closer to the star and appear broader. Several systems match radiation-supported bow shocks, bow waves, or dust waves; others remain consistent with classical wind-supported bow shocks. No clear link is found between H-alpha detection and stellar parameters.","pith_inferences":["If Vink mass-loss rates are systematically high, more of the sample will shift into radiation-supported regimes, raising the true fraction of radiation-supported bows.","The systematic offset of infrared arcs inward of H-alpha arcs is a ready test of grain–gas decoupling models once multi-wavelength grain-size diagnostics exist.","Weather-vane (externally driven) cases flagged by misaligned proper-motion vectors may dominate the HII-region subsample and should be separated before runaway statistics are drawn.","A uniform deeper H-alpha campaign on the 28 non-detections could reveal whether the true detection rate is set by sensitivity or by genuinely dust-only waves."],"forward_implications":["Ionized-gas counterparts exist for a non-negligible fraction of infrared bow-shock candidates and can be imaged with modest ground-based facilities.","Radiation pressure must be included alongside wind ram pressure when interpreting stand-off distances and deriving mass-loss rates from bow arcs.","Objects dark in H-alpha are also generally dark at radio wavelengths in the small overlapping sample, suggesting a joint non-detection trend worth testing.","Surface-brightness and emission-measure values supplied for ten detections give lower limits on expected thermal free–free emission at 6 GHz for radio follow-up.","Some arcs may still be reflection nebulae; continuum or multi-line imaging is needed to confirm pure recombination emission."],"fun_headline_variants":["15 of 78 IR bow shocks show clear H-alpha arcs","H-alpha arcs found in one-fifth of infrared OB-star bow shocks","Ionized gas traces many IR bow shocks around massive stars","H-alpha survey detects 16 arc nebulae at OB-star bow shocks","Radiation and winds both shape H-alpha bow shocks at OB stars"],"cache_read_input_tokens":32896,"weakest_assumption_plain":"The regime assigned to each bow rests on rough ambient-density ranges chosen from how the local H-alpha sky looks, plus mass-loss rates that can be badly overestimated for weaker-wind stars.","fun_headline_variants_meta":{"raw":{"variants":["15 of 78 IR bow shocks show clear H-alpha arcs","H-alpha arcs found in one-fifth of infrared OB-star bow shocks","Ionized gas traces many IR bow shocks around massive stars","H-alpha survey detects 16 arc nebulae at OB-star bow shocks","Radiation and winds both shape H-alpha bow shocks at OB stars"]},"model":"grok-4.5","effort":"low","cost_usd":0.003023,"raw_usage":{"total_tokens":1181,"prompt_tokens":907,"num_sources_used":0,"completion_tokens":78,"cost_in_usd_ticks":30228000,"prompt_tokens_details":{"text_tokens":907,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":196,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":907,"tokens_out":78,"duration_ms":4261,"temperature":1.0,"reasoning_tokens":196,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-31T12:00:02.508359+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Optical spectroscopy that measures density, temperature, and kinematics of the H-alpha arcs, or independent ambient-density and mass-loss measurements that move the predicted stand-off distances out of the assigned regime.","supporting_citations":[],"review_version":1}