{"id":"40f846e0-2ed8-4879-b4fd-eb2a02e832c3","arxiv_id":"1909.02646","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"The GYT line list provides empirically refined, UV-to-infrared rotational, rovibrational, and rovibronic transitions for SH and its isotopologues, extending coverage to 0.256 μm.","lead":"This paper presents the GYT line list, a new catalog of 572,145 spectral lines for the SH radical, covering its ground-state infrared transitions plus a new ultraviolet band system. It was created to test whether SH absorption explains a mysterious UV opacity rise in the ultra-hot Jupiter WASP-121b and to support high-resolution exoplanet spectroscopy.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The A-state extrapolation beyond the fitted v'≤2, J≤12.5 data is the load-bearing weak point; all v'≥3 energies in the GYT list are unvalidated and the paper supplies the data (Schnieder et al. 1990) needed to test them.","rationale":"The reader's conditional verdict is appropriate. The paper has real strengths: well-documented fit residuals (0.06 cm-1 for X–X, 0.3 cm-1 for A–X), a transparent Duo model in supplementary material, and realistic handling of the X-state. The soft spot is exactly the A-state extrapolation; the paper itself flags the v'≤2 limitation in Section 5 but then states completeness in the abstract and conclusions without quantifying the extrapolation uncertainty. The proposed Schnieder comparison is a direct, feasible falsification: those term values span v'=0–4 and N up to 40 and are the only empirical check on v'≥3. A secondary issue is the unvalidated ab initio A–X transition dipole moment, with literature f00 values spanning 0.0017–0.0029; however, the reliability of the deep-UV line positions is the more fundamental limit on the 'accurate' claim. The predissociation omission is explicitly acknowledged and is secondary for line positions, though it makes the lifetimes in Table 4 unphysical for v'≥1. I would keep the CONDITIONAL verdict, requiring a caveat or validation before the completeness claim can be accepted.","tokens_in":19551,"tokens_out":11260,"duration_ms":120709,"concrete_test":"Use the published Duo input file to recompute A 2Σ+ term values for v'=3 and 4, N=0–40, and compare them with the empirical term values of Schnieder et al. (1990). Report the rms and maximum deviation and the resulting (3,0)/(4,0) band origins. Given that the paper quotes Schnieder et al.'s own discrepancies as roughly 1–66 cm-1, a correct extrapolation should not show systematic, growing residuals beyond that envelope; if the v'=3,4 residuals exceed the quoted uncertainty or shift the band origins by tens of cm-1, the deep-UV portion of the GYT list is not empirically validated and the completeness claim should be restricted to v'≤2.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central 'accurate and complete' claim for the A–X system rests on an extrapolation that the paper's own stated data coverage cannot support. Table 1 shows the A–X fit uses only the (0,0), (1,0) and (2,0) bands, Jmax=12.5, from four sources. Section 3 nevertheless truncates upper states at the A-state asymptote (~39 000 cm-1), and Table 4 contains A-state levels with v'=3 and 4 (e.g., 35 601.52 and 36 758.87 cm-1) that are pure EMO extrapolations. Section 5 explicitly concedes 'we have been limited to v' ⩽ 2.' Consequently all A–X lines above ~34 300 cm-1, and all A-state levels above v'=2, have energies determined entirely by the unconstrained shape of the fitted EMO potential at large r. The paper dismisses the only existing v'=3,4 data (Schnieder et al. 1990) on accuracy grounds, but those data were not used even as a validation set. In addition, the 0.014 μm offset between the synthetic and CRDS spectra (Fig. 11) is attributed to 'a calibration problem' without a quantitative argument, so the absolute wavelength scale of the A–X (0,0) band is not independently confirmed. These issues do not undermine the high-quality fit to v'≤2, but they invalidate the unqualified 'complete to 0.256 μm / up to 5000 K' phrasing.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a new ExoMol line list (GYT) for the mercapto radical SH and its isotopologues, extending the previous SNaSH line list to include the A 2Σ+–X 2Π ultraviolet band system in addition to the ground-state X 2Π–X 2Π transitions. The authors carry out MRCI/aug-cc-pV5Z-DK ab initio calculations for the potential energy, spin–orbit, electronic angular momentum, and (transition) dipole moment curves, then refine the PECs, SOCs, EAMCs, BOBCs, and SRCs in a Duo fit to experimental line positions from the literature (Table 1). The fit reaches RMS residuals of 0.06 cm−1 for X–X and 0.3 cm−1 for A–X. The resulting line list contains 7686 states and 572,145 transitions for 32SH and analogous lists for 33SH, 34SH, 36SH, and 32SD, covering wavelengths down to 0.256 μm. The authors compare simulated spectra with several experimental spectra and report a radiative lifetime and f00 = 0.0017 for the A–X (0,0) band.","tokens_in":19876,"tokens_out":7917,"duration_ms":81622,"significance":"If the line list is reliable, it is a valuable community resource for interpreting UV observations of SH in hot-Jupiter transmission spectra (notably WASP-121b), cool stars, and the ISM. The paper is transparent about its method, provides the Duo input files as supplementary data, and compares intensities with external theory and experiment. The principal value is the full rovibronic A–X band coverage with empirically tuned line positions and ab initio intensities. However, the empirical tuning is confined to v′=0–2 and J≤12.5 for A–X, so the ultra-violet part of the line list beyond the fitted range is an extrapolation. Since the stated completeness (to 0.256 μm and 5000 K) is one of the paper's main selling points, the significance is contingent on validating that extrapolation.","major_comments":[{"comment":"The A 2Σ+ potential is fitted to only the (0,0), (1,0), and (2,0) A–X vibronic bands with Jmax = 12.5 (Table 1), yet the line list includes A-state levels with v′=3 and 4 (Table 4, e.g., 35,601.52 and 36,758.87 cm−1) and transitions extending to the A-state asymptote near 39,000 cm−1. These energies are unconstrained extrapolations of the empirical EMO potential, and the paper itself states in Section 5 that 'we have been limited to v′ ≤ 2'. The abstract's claim that this is an 'accurate and complete' line list covering to 0.256 μm is therefore unsupported for all A–X lines above the highest fitted bands (roughly 34,300 cm−1). The existing Schnieder et al. (1990) A-state term values for v = 0–4 and N = 0–40 should be used as an external validation set even if they were not sufficiently accurate for the fit; if they are not used, the completeness and accuracy claims must be restricted to v′≤2 and the associated J range.","section":"Section 3; Table 4; Section 5"},{"comment":"Predissociation of the A 2Σ+ state via the repulsive 2Σ−, 4Σ−, and 4Π states is explicitly identified in the introduction and Section 4, but these states are omitted from the spectroscopic model. The lifetimes reported in the states file (Table 4) are therefore radiative lifetimes only, and Section 4 acknowledges that the computed lifetimes (449 ns for v=0) are not expected to match the measured predissociative lifetimes (which can be orders of magnitude shorter). This limitation is disclosed for lifetimes, but the same omitted crossings can also perturb the A-state energies and couplings for the high-v and high-J levels that the line list includes, and no estimate of the magnitude of these energy shifts is provided. Please either add the predissociative channels to the model, or provide a quantitative argument that they do not affect the computed line positions above the fitted range, and in any case label the Table 4 lifetimes as radiative-only.","section":"Section 4; Table 4"}],"minor_comments":[{"comment":"The sentence 'The rms error of this fit to the A–X line positions if SD is 0.4 cm−1' contains a typo: 'if' should read 'for'.","section":"Section 5"},{"comment":"The caption says 'τ: Lifetime (s−1)', but the numeric entries such as 5.1332E-07 are in seconds; the unit should be 's' rather than 's−1'.","section":"Table 4 caption"},{"comment":"The 0.014 μm offset between the synthetic and CRDS spectra is attributed to a calibration problem without a quantitative argument; please state the implied wavenumber shift, compare it with the reported fit residuals, and consider overlaying a shifted synthetic spectrum.","section":"Figure 11"},{"comment":"The phrase 'the relative property of completeness (up to ∼ 5000 K)' is vague; given the limited A-state vibrational coverage, provide a population-based justification of the temperature limit.","section":"Section 5"}],"recommendation":"major_revision","confidential_remarks":"This is a solid ExoMol data paper that fits the journal's scope. The main issue is the unqualified 'accurate and complete' claim for the A–X system when the empirical fit only covers v'=0–2, J≤12.5. The path to revision is clear: validate or restrict the extrapolation claims, and clearly label radiative-only lifetimes. I do not see a need for rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this is a genuinely useful data product, but the headline claim is too strong. The GYT line list adds the A 2Σ+–X 2Π system to the SNaSH ground-state list, covering 572,145 transitions for 32SH plus isotopologues, with fits that reach the stated experimental accuracy (RMS 0.06 cm-1 for X–X, 0.3 cm-1 for A–X). The Duo input files are supplied, so the work is reproducible, and the ab initio transition dipoles are compared with earlier calculations. That is real value, and the paper deserves a proper referee.\n\nThe soft spots are concentrated in the A-state extrapolation. The fit uses only the (0,0), (1,0) and (2,0) A–X bands up to J=12.5, so every A-state level with v'≥3 — including entries in the published states file — is a pure EMO extrapolation. The paper acknowledges being limited to v'≤2, but then presents the list as complete to 0.256 μm and 5000 K. Those two statements do not sit together. Instead of just dismissing Schnieder et al.'s v'=3,4 term values as insufficiently accurate, the authors should have used them as a validation set. Even a rough comparison would tell the reader how far the extrapolation can be trusted.\n\nTwo smaller issues. First, predissociation is omitted; the authors are transparent about this and correctly warn that their radiative lifetimes are upper limits, but the same coupling can shift line positions and intensities, so the caveat should be carried through to the completeness claim. Second, the 0.014 μm shift between the synthetic and CRDS spectra is written off as an experimental calibration problem without a quantitative argument. It might well be, but the reader needs a reason beyond assertion.\n\nNone of this undermines the well-fitted low-v' bands, and for users needing SH opacity for WASP-121b the list is still the best available. But the 'accurate and complete' phrasing needs to be tempered, and the CRDS discrepancy should be resolved or at least quantified. I'd recommend: accept after minor-to-moderate revision, not desk-reject and not accept as-is.","headline":"Solid, well-documented A–X extension for SH with good fits to the measured low-v' bands, but the 'complete' UV claim leans on unvalidated extrapolation and the paper should qualify it.","tokens_in":20536,"tokens_out":2302,"would_cite":true,"duration_ms":24925,"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 presents the GYT line list, an accurate and complete set of 572,145 transitions for the SH radical spanning infrared to ultraviolet wavelengths down to 0.256 μm.","keywords":["SH radical","mercapto radical","molecular line list","A 2Σ+ – X 2Π transitions","exoplanet atmospheres","ultraviolet opacity","isotopologues","opacity"],"falsifier":"Measure the (3,0) A–X band of ³²SH under high resolution and compare band-head positions and relative intensities with the GYT predictions around 34,000–36,000 cm⁻¹; a systematic offset beyond the claimed 0.3 cm⁻¹ rms, or a band intensity pattern that disagrees with the computed oscillator strength f00 = 0.0017, would indicate the fitted A-state curves do not extend correctly.","tokens_in":19219,"feed_emoji":"🔭","tokens_out":7532,"duration_ms":69517,"temperature":0.7,"pith_summary":"The paper sets out to give astronomers a single accurate, complete line list for the mercapto radical SH, covering both rotation-vibration transitions within the ground X ²Π state and electronic A ²Σ⁺–X ²Π transitions in the ultraviolet down to 0.256 μm. The motivation is concrete: SH is a suspected cause of the steep ultraviolet opacity rise in the ultra-hot Jupiter WASP-121b, and SH is also observed in the Sun, interstellar clouds, comets, and cool stellar atmospheres. The authors construct the list by fitting potential energy, spin-orbit, electronic angular momentum, spin-rotation, and Born-Oppenheimer breakdown curves directly to measured line positions, using ab initio curves as a starting point, then solving the coupled rovibronic problem for 7686 states and 572,145 transitions in ³²SH plus four isotopologues. If correct, the list replaces earlier ground-state-only or sparse experimental lists and supports high-resolution searches for SH in exoplanet and stellar spectra.","feed_headline":"SH line list spans infrared to UV 0.256 μm","feed_subtitle":"572,145 transitions for 32SH, fitted to experiment, supersede ground-state-only lists.","key_machinery":"The load-bearing machinery is a coupled-channels rovibronic model built for the Duo program, which solves the coupled rovibronic Schrödinger equation for open-shell diatomic molecules. The model comprises three fitted potential energy curves (X ²Π, A ²Σ⁺, and a repulsive B ²Π), plus spin-orbit curves, electronic angular momentum curves, spin-rotation curves, and Born-Oppenheimer breakdown curves, all refined against measured transition frequencies, while intensities come from ab initio dipole moment and transition dipole moment curves. Fitting the curves directly to measured lines rather than to spectroscopic constants lets the model absorb perturbations from nearby electronic states, which is what makes the extrapolation to unmeasured higher vibrational levels plausible.","core_discovery":"The central discovery claimed is that the GYT line list is an accurate and complete SH line list spanning the infrared to the ultraviolet: it reproduces measured X–X positions with rms error 0.06 cm⁻¹ and measured A–X positions with rms error 0.3 cm⁻¹, matching the quoted experimental uncertainties of ~0.03–0.3 cm⁻¹. The list contains 7686 rovibronic states and 572,145 transitions for ³²SH, extends to ~39,000 cm⁻¹ (0.256 μm) and to temperatures near 5000 K, and also provides lists for ³³SH, ³⁴SH, ³⁶SH, and ³²SD, the last incorporating non-Born-Oppenheimer corrections. The paper states that this list supersedes the SNaSH ground-state list and the experimentally limited Zahnle et al. (2009) absorption line list.","pith_inferences":["Because the A-state model is fitted only to the (0,0), (1,0), and (2,0) A–X bands up to J = 12.5, a direct test of the model is to measure the (3,0) and higher bands; if predicted positions drift by more than the claimed ~0.3 cm⁻¹ rms, the A-state potential needs further refinement.","The computed oscillator strength f00 = 0.0017 lies between the solar-derived value 0.0022 and the ab initio value 0.0029; high-resolution laboratory measurement of the (0,0) band intensity could discriminate which intensity curve is closer to reality.","Because predissociation is omitted, the line list's A-state radiative lifetimes (449 ns for v = 0, 513 ns for v = 1) are far longer than observed collision-free lifetimes; users modelling fluorescence or photochemistry will likely need to add predissociation as a separate loss channel."],"forward_implications":["The A–X band lines at ~0.3 μm can now be included in opacity models of WASP-121b and other ultra-hot Jupiters, testing the proposed SH explanation of the observed ultraviolet rise.","The line list provides computed Einstein A coefficients, g-factors, and lifetimes for every upper state, so radiative lifetimes from the model can be compared directly with experiments.","Coverage to 39,000 cm⁻¹ and temperatures near 5000 K means SH opacities can be computed for brown dwarfs and cool stars, not only hot Jupiters.","The separately fitted ³²SD model includes non-Born-Oppenheimer effects, enabling evaluation of deuterium fractionation in astrophysical SH observations.","The X–X portion agrees with the earlier SNaSH list below 10,000 cm⁻¹, so existing ground-state analyses remain valid while the ultraviolet predictions are new."],"supporting_citations":[{"why":"Provided the SNaSH ground-state line list and the X-state spectroscopic model that this work extends and re-fits.","marker":"Yurchenko et al. 2018b"},{"why":"Supplied the previous A–X absorption line list, limited to the first three vibrational states, which GYT supersedes.","marker":"Zahnle et al. (2009)"},{"why":"Reported the WASP-121b transmission spectrum whose steep ultraviolet opacity rise motivates the A–X extension.","marker":"Evans et al. (2018)"},{"why":"Measured most of the (0,0) and (1,0) A–X bands used in the fit, providing 170 transitions.","marker":"Ramsay (1952)"},{"why":"Added (1,0) and (2,0) A–X band measurements used to fit the A-state curves.","marker":"Johns & Ramsay (1961)"},{"why":"Provided early (0,0) A–X band measurements used in the fit.","marker":"Lewis & White (1939)"},{"why":"Contributed lifetime-sensitive (0,0) A–X transitions and LIF data used for comparison.","marker":"Loge & Tiee (1988)"},{"why":"Computed a high-level ab initio A–X transition dipole moment curve used to benchmark the intensity model.","marker":"Resende & Ornellas (2001)"},{"why":"The Duo program solves the coupled rovibronic Schrödinger equation that produces the energies and Einstein A coefficients.","marker":"Yurchenko et al. (2016)"},{"why":"Characterized the repulsive curves that predissociate the A state; the model omits them, making this the key comparison for lifetimes.","marker":"Wheeler et al. (1997b)"}],"fun_headline_variants":["SH line list now spans IR to UV, 572k transitions","GYT line list extends SH coverage to UV for hot Jupiters","Mercapto radical: new IR-to-UV line list for exoplanets","New SH line list: 572,145 transitions from IR to UV"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The A-state model is fitted only to the (0,0), (1,0), and (2,0) A–X bands with rotational coverage up to J = 12.5, so the line list's predictions at higher vibrational levels, higher rotation, and near the A-state dissociation limit rest on those fitted curves extrapolating reliably, while predissociating crossings are neglected.","fun_headline_variants_meta":{"raw":{"variants":["SH line list now spans IR to UV, 572k transitions","GYT line list extends SH coverage to UV for hot Jupiters","Mercapto radical: new IR-to-UV line list for exoplanets","New SH line list: 572,145 transitions from IR to UV"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00094,"raw_usage":{"total_tokens":4103,"prompt_tokens":1113,"completion_tokens":2990,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":729,"completion_tokens_details":{"reasoning_tokens":2909}},"tokens_in":729,"tokens_out":2990,"duration_ms":23543,"temperature":1.0,"reasoning_tokens":2909,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:43:46.083038+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the (3,0) A–X band of ³²SH under high resolution and compare band-head positions and relative intensities with the GYT predictions around 34,000–36,000 cm⁻¹; a systematic offset beyond the claimed 0.3 cm⁻¹ rms, or a band intensity pattern that disagrees with the computed oscillator strength f00 = 0.0017, would indicate the fitted A-state curves do not extend correctly.","supporting_citations":[{"cited_title":"A., 1952, @doi [J","cited_arxiv_id":null,"evidence_quote":"Measured most of the (0,0) and (1,0) A–X bands used in the fit, providing 170 transitions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Added (1,0) and (2,0) A–X band measurements used to fit the A-state curves."},{"cited_title":"N., White J","cited_arxiv_id":null,"evidence_quote":"Provided early (0,0) A–X band measurements used in the fit."},{"cited_title":"W., Tiee J","cited_arxiv_id":null,"evidence_quote":"Contributed lifetime-sensitive (0,0) A–X transitions and LIF data used for comparison."},{"cited_title":"M., Ornellas F","cited_arxiv_id":null,"evidence_quote":"Computed a high-level ab initio A–X transition dipole moment curve used to benchmark the intensity model."},{"cited_title":"N., Lodi L., Tennyson J., Stolyarov A","cited_arxiv_id":null,"evidence_quote":"The Duo program solves the coupled rovibronic Schrödinger equation that produces the energies and Einstein A coefficients."}],"review_version":1}