{"id":"7ac99dc3-b5aa-4fae-94ed-c85673a95d63","arxiv_id":"2501.10574","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Comparing shifted optical hydrogen lines with ultraviolet lines from 13 quasar candidates supports a binary supermassive black hole interpretation for three of them and tentatively for five more.","lead":"Using Hubble Space Telescope ultraviolet spectra of 13 quasars with offset broad hydrogen lines, the authors test whether the offsets come from a pair of supermassive black holes or from a lopsided disk around a single black hole. Three candidates pass the test strongly, five tentatively, one is disfavored, and four spectra are too absorbed to judge.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'strong support' classifications rest on an unproven premise: that a single SMBH cannot produce UV lines sharing the H-beta offset; the paper itself concedes matching offsets can arise from both single and binary black holes.","rationale":"The reader identified the same load-bearing concern: the discriminatory power of the UV test rests on the DE-derived premise that a single SMBH produces UV resonance lines near the rest frame, while shared offsets require bulk orbital motion. This is indeed the weakest link. The paper's own caveat that matching offsets can arise in both scenarios weakens the strength of the 'strong support' classification, and the ILR exclusion is essential to the three strong cases. The count inconsistency between Section 3.3 (four strong, four tentative) and Table 2/abstract (three strong, five tentative) is a real but secondary error that does not affect the central three objects. No machine-checked proof or quantitative model comparison is provided for the likelihood ratio, so 'much more likely' is an interpretation rather than a measured outcome. The concern does not overturn the paper's contribution: the data are new, the test is physically motivated, and the qualitative classifications are transparent. It does mean the claim should be presented as conditional on the DE-wind premise and on the ILR assumption, which is exactly the reader's CONDITIONAL verdict. Therefore no change to the verdict is needed.","tokens_in":22402,"tokens_out":11400,"duration_ms":122762,"concrete_test":"Run the Matthews et al. (2020, 2023) disk-wind radiative transfer models with a non-axisymmetric disk perturbation (e.g., a spiral arm or eccentric disk) chosen to reproduce the observed single-peaked offset H-beta profile, and compute the Ly-alpha and C IV line profiles. If the resulting UV lines have peaks offset by the same velocity as H-beta, the premise that only bulk orbital motion produces shared offsets is false and the 'strong support' classifications are not decisive. Alternatively, compile a control sample of known double-peaked emitters (single SMBH) with archival HST UV spectra and measure their UV line centroids; if any show a UV offset matching the Balmer offset, the DE-based premise fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that J001224, J115449, and J125142 are 'much more likely' to host SBHBs depends on the diagnostic premise stated in Section 1.2 and applied in Section 3.3: that for a single SMBH with a non-axisymmetric BLR, the UV resonance lines (Ly-alpha, C IV) should be single-peaked and centered near the narrow-line rest frame, so that only bulk orbital motion produces matching UV and H-beta velocity offsets. This premise is imported from observations of double-peaked disk emitters (DEs), but the SBHB candidates have single-peaked, offset H-beta profiles; the extension of the DE result to this regime is not established. The paper itself concedes in Section 1.2 that 'It is more difficult to interpret the scenario where the UV and optical lines have matching velocity offsets because this can arise from both single black holes and SBHBs,' and that passing the test 'does not conclusively establish a candidate as a true SBHB.' Yet Section 3.3 classifies matching-offset objects as 'strong support.' The adopted ILR interpretation (Section 3.3) also enters the strong classifications: for J001224 and J125142, a symmetric zero-velocity core is excluded as ILR, and for J115449 the Ly-alpha match is only 'under the ILR description.' If the ILR instead partakes of BLR motion, or if a single BH with a non-axisymmetric disk wind can produce a UV peak matching the H-beta offset, the three strong classifications lose their force. No quantitative likelihood ratio for the two scenarios is provided, so 'much more likely' is not supported by the analysis.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents new Hubble Space Telescope ultraviolet spectroscopy (COS and STIS) of 13 quasars at z < 0.7 that were previously identified as supermassive black hole binary (SBHB) candidates on the basis of velocity-offset broad H-beta emission lines. The authors also obtained contemporaneous ground-based optical spectra covering H-beta, H-alpha, and in some cases Mg II. The central test compares the broad UV line profiles (Ly-alpha, C IV, Mg II) with the optical Balmer profiles: under the SBHB hypothesis, all broad lines should share the same velocity offset as bulk orbital motion, whereas under the alternative single-black-hole perturbed-disk scenario, the UV resonance lines are expected to be single-peaked and centered near the narrow-line rest frame. The authors classify the 13 objects into four categories: strong support (3), tentative support (5), disfavored (1), and inconclusive owing to absorption (4). The three 'strong support' objects are J001224, J115449, and J125142. The paper concludes that these three are much more likely to host close SBHBs, although it acknowledges that passing the test does not conclusively establish a candidate as a true SBHB.","tokens_in":22671,"tokens_out":5703,"duration_ms":56963,"significance":"If the classification scheme is accepted, this paper provides an important, physically motivated discriminator for a small but heavily studied class of SBHB candidates. The new HST spectra are a valuable public resource, the contemporaneous optical spectroscopy strengthens the comparison, and the per-object notes are detailed and transparent. The test has a clear falsifiable logic: a single SMBH whose broad Balmer profile is distorted by a non-axisymmetric disk would not naturally produce UV resonance lines sharing the same velocity offset as H-beta. The paper also explicitly states limitations, including the subjectivity of the classifications and the frequent impact of absorption. The main significance, if the conclusions hold, is to elevate three objects as the most promising close-SBHB candidates from this sample. However, the strength of that conclusion is moderated by the qualitative nature of the comparisons and by the fact that the discriminatory premise is imported from observations of double-peaked disk emitters rather than demonstrated for the present single-peaked, offset-line sample.","major_comments":[{"comment":"The classification counts are internally inconsistent. Section 3.3 states 'Four objects show strong support for the SBHB hypothesis, four show tentative support...', whereas Table 2 lists three strong (J001224, J115449, J125142) and five tentative (J015530, J094603, J111916, J134617, J171448), and the Abstract and Section 4 also report three strong and five tentative. This is a factual error in the paper's headline result; the text in Section 3.3 must be corrected to match the table and the abstract.","section":"Section 3.3 and Table 2"},{"comment":"The diagnostic premise that a single SMBH produces single-peaked UV resonance lines centered near the narrow-line rest frame is not established for the present sample. The premise is imported from observations of double-peaked disk emitters, but the SBHB candidates studied here have single-peaked, velocity-offset H-beta profiles. The paper itself concedes in Section 1.2 that matching UV and optical velocity offsets 'can arise from both single black holes and SBHBs' and that passing the test 'does not conclusively establish a candidate as a true SBHB.' Given this concession, classifying matching-offset objects as 'strong support' for the SBHB hypothesis overstates the discriminating power of the test. I recommend either softening the classification language (e.g., 'consistent with') or providing a quantitative model-based likelihood ratio that quantifies how much more likely the matched-offset observation is under the SBHB scenario than under the single-BH scenario.","section":"Sections 1.2 and 3.3"},{"comment":"The three 'strong support' classifications depend critically on the adopted interpretation that the intermediate-width UV line core is an ILR that does not participate in BLR orbital motion. For J115449, the Ly-alpha match is accepted only 'under the ILR description'; for J125142, the match is accepted 'modulo an ILR core'; and for J001224, a symmetric core is excluded as ILR. If the ILR instead partakes of BLR motion, or if a single BH with a non-axisymmetric disk wind can produce a UV line peak at the H-beta offset, these classifications lose their force. This is a load-bearing assumption that needs explicit empirical or modeling support; at minimum, the classifications should be stated as conditional on the ILR interpretation, and the paper should discuss the consequences if that interpretation is wrong.","section":"Sections 3.3 and 3.4"},{"comment":"The scale-factor procedure uses iterative sigma clipping that explicitly rejects 'pixels in the core of the UV line profile' and 'pixels impacted by absorption.' This means the quantitative comparison of Ly-alpha and H-beta is driven by the line wings, while the offset of the line peak, which is the key diagnostic, is evaluated only by visual inspection of the residuals. There is no reported measurement of the velocity offset of each line peak with an associated uncertainty, nor any statistical test of whether the UV and optical offsets agree. Without such quantitative metrics, the distinction between 'strong support' and 'tentative support' is not reproducible, and the central claim that three objects show matching offsets is not supported by a formal estimate. Please add measured peak velocities (or equivalent profile centroids) with uncertainties, and a quantitative comparison significance for the strong-support objects.","section":"Section 3.2"}],"minor_comments":[{"comment":"The sentence beginning 'Four objects show strong support...' should be corrected to 'Three objects show strong support... five show tentative support...' to match Table 2 and the Abstract.","section":"Section 3.3"},{"comment":"The phrase 'We present Hubble Space Telescope ultraviolet (UV) of 13 quasars' is missing the word 'spectra'; it should read 'ultraviolet (UV) spectra of 13 quasars.'","section":"Abstract"},{"comment":"In the J001224 note, 'There is a very compelling blueshift of comparable magnitude is observed' is grammatically broken; it should read 'A very compelling blueshift of comparable magnitude is observed' or 'There is a very compelling blueshift of comparable magnitude.'","section":"Section 3.4"},{"comment":"In the J094603 note, 'not obviosuly as distinct' is a typo for 'not obviously as distinct.'","section":"Section 3.4"},{"comment":"In the J154340 note, 'notwirthstanding' is a typo for 'notwithstanding.'","section":"Section 3.4"},{"comment":"The sentence 'This work has also revealed two limitations the approach of comparing the UV and optical line profiles' is missing a word; it should read 'two limitations of the approach' or 'two limitations to the approach.'","section":"Section 4"},{"comment":"The caption refers to 'J13417' but the object is J134617; the abbreviated name should be consistent with the rest of the text (J134617).","section":"Figure 2 caption"}],"recommendation":"major_revision","confidential_remarks":"This is the fourth paper in a series with a well-established methodology, and the new HST data are a genuine contribution. The central result is plausible but the 'strong support' language is not fully justified by the analysis as presented; the paper itself concedes the key discriminating premise is not decisive. The internal count inconsistency and the absence of quantitative offset measurements are fixable within the scope of a revision. I do not see grounds for rejection, but the revision should address the four major comments substantively before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this paper is a careful observational data paper with real value, and the three objects it classifies as strong SBHB candidates deserve attention. But the classification language is stronger than the underlying argument supports, and there is a small count error that needs fixing.\n\nWhat's genuinely new: 13 targets with new HST/COS and STIS ultraviolet spectra, contemporaneous optical spectra, full spectral decomposition, and detailed per-object notes. The individual notes are honest, and the paper openly flags cases where absorption or the intermediate-line-region (ILR) interpretation makes the comparison uncertain. The data are public, the pipelines are standard, and the classifications are explicit enough to be re-examined.\n\nThe soft spots are in proportion. The main one is the interpretive leap. The test's core premise, imported from double-peaked disk emitters, is that a single black hole with a non-axisymmetric BLR will produce UV resonance lines that are single-peaked and close to the narrow-line frame, so only true bulk orbital motion would make the UV lines share the Hβ velocity offset. That premise is not demonstrated for single-peaked, velocity-offset Hβ candidates. The paper itself says in Section 1.2 that matching offsets 'can arise from both single black holes and SBHBs' and in Section 4 that passing the test does not 'conclusively establish a candidate as a true SBHB.' Yet Section 3.3 labels three objects as 'strong support' without a likelihood ratio or a quantitative model for the single-BH alternative. If the ILR core is actually emitted from the BLR, or if a non-axisymmetric disk wind can produce a UV peak at the Hβ offset, those three classifications lose much of their force. This is a real caveat, but it does not sink the paper; it argues for softer language or a quantitative comparison.\n\nSecond, the scale-factor procedure uses sigma clipping that removes the UV line core and absorbed pixels before comparison. That is a reasonable conservative choice, but it makes the match easier to achieve and should be stated more prominently.\n\nThird, there is a count inconsistency: Section 3.3 says four objects show strong support and four tentative, while Table 2, the abstract, and Section 4 all say three strong and five tentative. Minor, but it should be fixed.\n\nWho this is for: anyone working on individual SBHB candidates, pulsar timing array source searches, or broad-line region physics. The three strong candidates, J001224, J115449, and J125142, become high-priority targets for continued radial velocity monitoring.\n\nRecommendation: send it to peer review. The data and analysis deserve referee time. A good referee should ask the authors to either soften 'strong support' to 'consistent with the SBHB hypothesis' or provide a quantitative estimate of how strongly matching offsets favor a binary over a single black hole with a non-axisymmetric disk or wind. Fix the count, and the paper is close.","headline":"New HST spectra give three SBHB candidates a real boost, but the 'strong support' label overreaches what the physics can defend.","tokens_in":23269,"tokens_out":3419,"would_cite":true,"duration_ms":35957,"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":"This paper tests whether 13 quasars with velocity-shifted broad Hβ lines are close supermassive black hole binaries, finding three with strong support from matched ultraviolet and optical line shifts.","keywords":["supermassive black hole binaries","active galactic nuclei","broad-line region","ultraviolet spectroscopy","Hubble Space Telescope","double-peaked emitters","quasar emission lines","radial velocity offsets"],"falsifier":"Observe a well-studied double-peaked emitter known to be a single black hole, such as 3C 390.3, at an epoch when its Hβ peak is displaced by roughly 1000 km/s, and measure whether Lyα and C IV share that displacement. If they do, the premise that UV resonance lines stay at rest for a single black hole fails and the test's classifications lose their basis.","tokens_in":22170,"feed_emoji":"🕳️","tokens_out":12056,"duration_ms":98345,"temperature":0.7,"pith_summary":"This paper reports a test of whether 13 quasars whose broad Hβ lines are offset by roughly a thousand kilometers per second from their host galaxies' rest frames actually contain closely orbiting supermassive black hole pairs. The test compares the velocity shifts of ultraviolet resonance lines (Lyα and C IV) with those of the optical Balmer lines using new Hubble Space Telescope spectra taken nearly simultaneously with ground-based optical observations. The reasoning is that bulk orbital motion would shift all lines together, whereas an alternative explanation involving a single black hole with a lopsided accretion disk would leave the UV lines single-peaked and near the rest frame. Three objects (J001224, J115449, J125142) show matching shifts in all lines, strongly supporting the binary interpretation; five give tentative support, one is disfavored, and four are too absorbed to judge. If the test is sound, these three become the most credible close-binary candidates from the radial-velocity search method.","feed_headline":"Three quasars pass a UV test for binary black holes","feed_subtitle":"HST spectra show Lyα and Hβ share the same velocity shift in three objects, boosting the case for close binaries.","key_machinery":"The central object is the shared-shift comparison between the broad optical Balmer lines and the ultraviolet resonance lines. For a binary, bulk orbital motion of the active black hole and its bound broad-line region Doppler-shifts every line by the same velocity; for a single black hole with a non-axisymmetric disk, the Hβ offset is one peak of a double-peaked profile and the Lyα/C IV lines, formed in an outflowing wind above the disk, should be single-peaked and lie near the narrow-line rest frame. The implementation requires isolating broad profiles by spectral decomposition, rescaling one profile onto the other by least-squares with sigma clipping, and adopting the interpretation that the symmetric intermediate-width UV core is not part of the broad-line region, based on reverberation mapping results.","core_discovery":"We present Hubble Space Telescope ultraviolet spectra of 13 supermassive black hole binary candidates selected from a parent sample of quasars with broad Hβ lines offset from the [O III] rest frame by 700–4000 km/s. After decomposing the spectra to isolate the broad line profiles, we compare the UV lines (Lyα and C IV, plus Mg II where available) with contemporaneous Hβ and Hα. The comparison is governed by the expectation, taken from studies of double-peaked disk emitters, that for a single black hole the UV resonance lines form in an outflowing wind and are single-peaked and centered at the narrow-line rest frame, while in a true binary every line shares the same bulk Doppler offset. On this basis, three candidates (J001224, J115449, J125142) show strong support for the binary hypothesis because their UV lines reproduce the Hβ offset; five show tentative support complicated by an intermediate-width core or weak absorption; one (J154340) is disfavored because its UV lines sit at rest while Hα matches the shifted Hβ; and four are inconclusive because of severe absorption in the UV.","pith_inferences":["Inference: If J001224, J115449, and J125142 are genuine binaries with the masses and separations estimated in earlier papers, their orbital periods of decades to centuries are within reach of monitoring programs; a measured period would be a decisive confirmation.","Inference: The same test could be run on candidates selected by other routes, such as periodic photometric variability or double-peaked lines, but the rest-frame expectation for UV lines would need re-calibration for those selection methods.","Inference: For the five tentative cases, the outcome hinges on the intermediate-width core interpretation; a dedicated reverberation-mapping campaign on the UV cores of these quasars could settle which side each classification falls on.","Inference: Observing Mg II rather than Lyα/C IV may rescue some of the inconclusive cases, since the paper notes Mg II is unabsorbed in at least one object whose Lyα and C IV are heavily absorbed."],"forward_implications":["If the classifications are right, J001224, J115449, and J125142 are the strongest close-binary candidates from the radial-velocity method and merit continued optical monitoring to detect the orbital acceleration predicted by the binary interpretation.","The UV-versus-optical comparison works as a discriminator even though radial-velocity curves can be mimicked by a perturbed disk, so it is a complementary test for other SBHB candidates.","A substantial fraction of candidates (about 30%) cannot be tested this way because of severe absorption, so future UV work should budget for this loss or use lines less affected by absorption.","The presence of both red- and blueshifted offsets among the supported candidates, while outflows would produce only blueshifts, is consistent with an orbital origin rather than an outflow origin."],"supporting_citations":[{"why":"Paper I of this series; defined the parent sample of 88 SBHB candidates with velocity-offset broad Hβ lines from SDSS DR7, from which the 13 HST targets were drawn.","marker":"Eracleous et al. 2012"},{"why":"Provided the optical light curves and many of the contemporaneous optical spectra used here to isolate Hβ and Hα profiles.","marker":"Runnoe et al. 2015"},{"why":"Measured the radial velocity curves of the candidates, the complementary monitoring constraint the UV test extends and strengthens.","marker":"Runnoe et al. 2017"},{"why":"First comparison showing UV resonance lines of double-peaked emitters are single-peaked and centered near rest, establishing the test's expectation for single black holes.","marker":"Halpern et al. 1996"},{"why":"Summarized the double-peaked emitter results and the wind/destruction mechanisms that strip the double-peaked shape from Lyα and C IV, motivating the UV test.","marker":"Eracleous et al. 2009"},{"why":"Disk-wind model in which UV resonance lines form in an accelerating outflow, the physical basis for single-peaked, unshifted UV lines.","marker":"Murray & Chiang 1997"},{"why":"Modern numerical wind models that produce single- or double-peaked UV lines depending on wind velocity law, extending the theoretical grounding.","marker":"Matthews et al. 2020"},{"why":"Reverberation mapping showing the UV line core does not respond to continuum changes, supporting the treatment of the intermediate-width core as outside the broad-line region.","marker":"Denney 2012"}],"fun_headline_variants":["Three quasars pass UV binary test","UV spectra support binary black holes in 3 quasars","Hubble UV data strengthens case for 3 binary quasars","3 quasars show matched UV and optical Doppler shifts","Survey finds 3 strong binary black hole candidates in UV"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The test assumes that a single supermassive black hole produces ultraviolet resonance lines that are single-peaked and close to the galaxy's rest frame, so only true bulk orbital motion would make them share the velocity offset of Hβ; if that premise is wrong, the three 'strong support' classifications collapse. It also assumes the intermediate-width UV core is not part of the broad-line region, which affects several tentative classifications.","fun_headline_variants_meta":{"raw":{"variants":["Three quasars pass UV binary test","UV spectra support binary black holes in 3 quasars","Hubble UV data strengthens case for 3 binary quasars","3 quasars show matched UV and optical Doppler shifts","Survey finds 3 strong binary black hole candidates in UV"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00028,"raw_usage":{"total_tokens":1677,"prompt_tokens":979,"completion_tokens":698,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":595,"completion_tokens_details":{"reasoning_tokens":622}},"tokens_in":595,"tokens_out":698,"duration_ms":7163,"temperature":1.0,"reasoning_tokens":622,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T19:06:57.207835+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Observe a well-studied double-peaked emitter known to be a single black hole, such as 3C 390.3, at an epoch when its Hβ peak is displaced by roughly 1000 km/s, and measure whether Lyα and C IV share that displacement. If they do, the premise that UV resonance lines stay at rest for a single black hole fails and the test's classifications lose their basis.","supporting_citations":[],"review_version":1}