{"id":"6925b00e-f252-46ae-a282-f8a486012eaf","arxiv_id":"2508.20355","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Single-transition radiative transfer fits to high-sensitivity L1544 spectra are degenerate, and the fitted sulphur-bearing abundances exceed chemical model predictions by factors of 2 to 20.","lead":"New high-sensitivity IRAM 30m spectra of seven molecules toward the pre-stellar core L1544 show line shapes that standard 1D radiative transfer models cannot uniquely reproduce with single transitions. The fits imply molecular abundances, especially for sulphur-bearing species, that are higher than astrochemical model predictions, signaling incomplete sulphur chemistry in dense cores.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Sulphur-chemistry challenge rests on MCMC step-abundance profiles that the authors themselves flag as non-unique; single-line degeneracy makes the pyRate comparison an unreliable benchmark.","rationale":"The reader's weakest assumption correctly identifies the reliability of the MCMC-fitted step-abundance profiles as the load-bearing issue. My independent reading of the text supports this: the authors explicitly warn in Appendix A that the LOC+MCMC result may not be physically accurate, and the corner plots and Table 2 show multiple unconstrained or boundary-hitting parameters. Because the paper's own central methodological finding is that single lines cannot yield a unique solution, any comparison of those non-unique profiles to pyRate is not a solid basis for concluding that sulphur chemistry is poorly reproduced. I also note that the discrepancy pattern is not uniform across the sulphur species, which further weakens the strongest formulation of the claim. The CONDITIONAL verdict already reflects this risk, so no verdict adjustment is needed. The proposed concrete test would settle whether the S-chemistry offsets are physical by adding the multi-line constraints the paper itself identifies as necessary.","tokens_in":24725,"tokens_out":3453,"duration_ms":37641,"concrete_test":"Fit at least one additional transition per sulphur species with the same LOC+MCMC setup and Keto et al. (2015) physical model (e.g., CS 3-2 plus C34S 3-2 or C34S 1-0, SO 3-2 plus 34SO 3-2), and compare the joint posterior of (ain, aout, r) to pyRate. If the abundance offsets and outward-shifted depletion radii shrink to within ~2x or move into agreement with pyRate under multi-line fitting, the claimed chemical-model failure is a degeneracy artifact rather than a real sulphur-chemistry discrepancy. A minimal variant: rerun the CS-only fit with fv prior extended to [0,3] and sigma_turb lower limit 0.02 km/s; if the CS aout posterior shifts materially, the boundary-hitting fit is not a reliable benchmark.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that current chemical models poorly reproduce sulphur chemistry depends on using the LOC+MCMC step-abundance fits as the empirical benchmark. That benchmark is demonstrably insecure. The paper's own Appendix A concedes that 'with the lack of enough constraints, the result found by LOC + MCMC may not be the \"correct\" or a physically accurate result.' The posteriors confirm this: 34SO has a flat ain and weakly constrained r, fv and sigma_turb (Fig. 18); c-C3H2 is largely unconstrained; H2CO is bimodal; and CS sits at the prior boundaries (fv = 1.99, sigma_turb = 0.01) while reaching tau ~ 105, which the authors call 'unexpectedly high.' Moreover, the claimed factor 2-20 abundance excess is not uniform: SO's fitted aout is actually ~2x lower than pyRate, and 34SO agrees within errors. The 'poor reproduction' claim thus rests primarily on CS and C34S, exactly the fits that hit prior edges or produce unphysically high optical depths. Since the paper itself shows that single-transition fits cannot find a unique global minimum, the derived abundance profiles are not robust enough to conclude that pyRate's sulphur chemistry is wrong. The conclusion would require multi-line constraints to break the degeneracy.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents new high-sensitivity, high-spectral-resolution IRAM 30m observations of HCO+, CS, C34S, H2CO, c-C3H2, SO, and 34SO toward the L1544 pre-stellar core, together with the HC17O+ line from an earlier work. The authors couple the LOC non-LTE radiative transfer code with an MCMC sampler and fit simple two-step abundance profiles (ain, aout, r, fv, sigma_turb) to each transition. They then compare the fitted abundance profiles with the pyRate chemical model and conclude that current chemical models, particularly for sulphur chemistry, poorly reproduce the observations. They also argue that single-transition fits cannot find a unique global minimum, that the lines trace distinct layers with different kinematics, and that an extended envelope is needed for HCO+ and H2CO. The paper includes optical depth estimates, corner plots, and combined isotopologue fits in the appendices.","tokens_in":25134,"tokens_out":3832,"duration_ms":43020,"significance":"If the fitted abundance profiles were robust, the reported factor-of-2 to factor-of-20 discrepancies for CS and C34S and the outward-shifted depletion radii would be an important challenge to sulphur chemical networks and would motivate a more consistent treatment of sulphur depletion. The observational dataset is genuinely valuable, and the paper is commendably transparent about degeneracies: it explicitly states that single-line fits cannot locate a global minimum, shows flat and multi-peaked posteriors, and warns in Appendix A that the LOC+MCMC result 'may not be the correct or a physically accurate result.' This honesty is a strength. However, for the reasons developed in the major comments, the empirical benchmark used to evaluate pyRate is not robust enough to carry the central sulphur-chemistry conclusion as stated. The paper is best viewed as a careful presentation of new data and a demonstration of the limitations of single-line abundance fitting, rather than as a definitive falsification of current sulphur chemistry models.","major_comments":[{"comment":"The central claim that pyRate poorly reproduces sulphur chemistry depends on using the LOC+MCMC step-abundance fits as the empirical benchmark, but the paper itself establishes that these fits are non-unique. Section 5 states that high-resolution single lines do not provide enough information to converge to one solution, and Appendix A concedes that the fit may not be physically accurate. This is the load-bearing logical gap: the abundance profiles are then compared quantitatively with pyRate (e.g., the factor-of-9 and factor-of-20 excesses for C34S and CS), despite the fact that the profiles themselves are admitted to be underconstrained. The conclusion 'poor reproduction of sulphur chemistry' therefore needs either multi-line constraints for the relevant species or a substantial softening of the claim.","section":"Section 5 and Appendix A"},{"comment":"The CS fit, which provides the largest abundance discrepancy (a factor of about 20) and the highest optical depth (about 10^5), is the least trustworthy of the fits. The MCMC drives fv to 1.99 and sigma_turb to 0.01, both at the edges of the priors, and the authors themselves call the resulting optical depth 'unexpectedly high.' A fit that only succeeds by pinning two parameters to the prior boundaries and producing a pathological optical depth cannot alone support the conclusion that pyRate underestimates the outer CS abundance by two orders of magnitude. This point needs to be addressed with additional transitions or with an explicit demonstration that the inferred aout is insensitive to the prior boundaries.","section":"Section 4.2.1 and Appendix A, Figs. 16 and 20"},{"comment":"The 34SO fit has a flat inner-abundance posterior and only weak constraints on r, fv, and sigma_turb. Despite these broad uncertainties, Section 5.1 describes the depletion radius as 'significantly displaced' from the pyRate prediction. The paper's own criterion in Section 3.1 is that flat histograms indicate that the parameter is not constrained; under that criterion, the 34SO depletion-radius shift is not statistically established. The qualitative discussion of a sulphur-chemistry problem should be separated from the parts of the analysis that are actually constrained, such as the better-determined aout value.","section":"Section 4.1.2 and Fig. 18"},{"comment":"The abundance comparison is not uniformly in one direction: for SO the fitted aout is a factor of about 2 lower than the pyRate maximum, while for CS and C34S the fitted values are higher than pyRate, and for 34SO the values agree within errors. The abstract and conclusions frame the result as 'a poor reproduction of sulphur chemistry,' but the heterogeneous pattern of over- and underproduction does not point to a single missing depletion process without a more specific mechanistic test. The discussion of SO + C+ destruction and the initial sulphur abundance assumption in Section 5.2 is plausible, but it is not quantitatively connected to the fitted profiles. The conclusions should be reframed as a demonstration that current models fail to match the shape and normalization of the sulphur-bearing profiles in a species-dependent way, rather than as a global sulphur-chemistry failure.","section":"Sections 4.2.2 and 5.2, Fig. 8"}],"minor_comments":[{"comment":"The rest frequency of 34SO is given as a private communication from the CAS laboratory; for reproducibility, this measurement should be published in a refereed context or documented in an accessible form before the paper is final.","section":"Section 2, Table 1"},{"comment":"There is a typo in the introduction: 'the the ionization fraction' should read 'the ionization fraction.'","section":"Section 1"},{"comment":"The layered schematic of L1544 is based on line-profile morphology and fits with several poorly constrained radii; it would be helpful to label the figure explicitly as an illustrative stack rather than a quantitatively determined radial stratification.","section":"Section 5.3, Figure 14"},{"comment":"The combined isotopologue fits report sigma_turb = 0.00 with zero uncertainty in some cases, which likely indicates a poor exploration or a prior-boundary issue rather than a precise constraint; a brief comment on this would prevent readers from misinterpreting the reported precision.","section":"Appendix C, Table 3"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is from a highly competent group and the observational material is valuable, but the interpretive claim in the title and abstract goes beyond what the presented fits can support. I would encourage the editor to allow a revision in which the sulphur-chemistry conclusion is either backed by multi-line constraints or explicitly softened to a demonstration of non-uniqueness and species-dependent disagreement. The paper could also be strengthened by presenting the pyRate comparison as a motivating test case rather than as a definitive challenge."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this paper brings genuinely new observations—high-sensitivity IRAM 30m spectra of seven lines toward L1544, the first clear C34S self-absorption detection in that source, and a laboratory-measured 34SO rest frequency. The radiative transfer work is careful, and the authors deserve credit for being unusually transparent about their own uncertainties. The corner plots are there for everyone to see, and Appendix A explicitly says the LOC+MCMC result may not be physically accurate. That kind of honesty is real.\n\nThe strongest and best-supported result is the methodological one: with a single transition per species, the MCMC fits cannot find a unique global minimum. The flat posteriors, the bimodal H2CO distribution, and the CS fit sitting at the prior boundaries (fv=1.99, sigma_turb=0.01) all point the same way. If you read this paper as a demonstration that single-line, 1D radiative transfer fits to high-sensitivity spectra are degenerate, it holds up well. The combined isotopologue fits they tried do not break that degeneracy, which is a useful negative result.\n\nThe soft spot is the central claim in the title. The 'poor reproduction of sulphur chemistry' conclusion uses those same non-unique fits as the empirical benchmark, and that benchmark is insecure. CS hits the prior edge and reaches tau ~ 105, which the authors themselves call unexpectedly high. C34S is almost an order of magnitude above pyRate, but with broad asymmetric errors. Meanwhile SO's fitted aout is actually about a factor of two lower than pyRate, and 34SO agrees within errors. So the factor 2-20 excess really comes down to CS and C34S—exactly the fits that are least trustworthy. There is also a structural mismatch: a step abundance profile with constant aout is being compared to a chemical model profile with a peak, so part of the 'discrepancy' is built into the functional form.\n\nI do not think this is a fatal flaw. The data and the degeneracy argument are worth publishing. But the sulphur-chemistry challenge should be softened, or backed with multi-line constraints. Releasing the reduced spectra and best-fit abundance profiles would also make the comparison reproducible and would help the field move past exactly this kind of degeneracy.\n\nWho is this for? People working on pre-stellar core chemistry and radiative transfer, especially anyone trying to derive abundances from single-dish spectra. It deserves a serious referee. I would send it out, and I would ask the authors to reframe the conclusion: the robust finding is the degeneracy, not the chemical model failure.","headline":"Valuable data and an honest account of fitting degeneracy, but the sulphur-chemistry 'challenge' rests on abundance profiles the authors themselves show are not unique.","tokens_in":25571,"tokens_out":1938,"would_cite":false,"duration_ms":21884,"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":"High-sensitivity spectra of L1544 show sulphur abundances that differ from chemical models by up to a factor of 20, and single-line fits are degenerate.","keywords":["ISM: molecules","ISM: clouds","radio lines: ISM","stars: formation","radiative transfer","pre-stellar cores","sulphur chemistry","L1544"],"falsifier":"Fit several rotational lines of CS, C$^{34}$S, SO and $^{34}$SO (for example J = 1-0, 2-1 and 3-2 for CS and SO) jointly toward the L1544 dust peak with the same LOC + MCMC machinery. If the joint fit converges to outer abundances within a factor of 2 of the pyRate predictions, the single-line sulphur discrepancy is an artefact of degeneracy rather than a chemical-network failure.","tokens_in":2060,"feed_emoji":"🔭","tokens_out":9432,"duration_ms":152262,"temperature":0.7,"pith_summary":"This paper asks whether current chemical models can reproduce the molecular emission from the pre-stellar core L1544 when confronted with unusually sensitive, high-resolution spectra. Using a non-local-thermodynamic-equilibrium radiative transfer code coupled to a Markov Chain Monte Carlo fit of simple step abundance profiles, the authors find that no single-transition fit converges to a unique solution. They also find that the best-fitting abundances for the sulphur-bearing species CS, C$^{34}$S, SO and $^{34}$SO deviate from the pyRate chemical network predictions by factors of 2 to 20, with the depletion radius shifted outward for CS, C$^{34}$S and $^{34}$SO, while SO is overproduced by the model by about a factor of 2. Because L1544 is a benchmark pre-stellar core, this matters: it suggests current sulphur networks are missing an important depletion process, and that high-sensitivity single lines alone cannot pin down the core's abundance structure.","feed_headline":"Sulphur chemistry mismatches L1544 gas by up to 20x","feed_subtitle":"High-sensitivity spectra show sulphur abundances that differ from chemical models by up to 20x, pointing to missing freeze-out physics.","key_machinery":"The machinery is the combination of a non-LTE radiative transfer solver and a Monte Carlo sampler over a deliberately simple abundance model. The LOC code solves the statistical equilibrium equations with an accelerated $\\lambda$ iteration and computes 1D line radiative transfer through the contracting Bonnor-Ebert sphere model of L1544; the MCMC sampler then explores five parameters: inner abundance $a_\\mathrm{in}$, outer abundance $a_\\mathrm{out}$, depletion radius $r$, velocity scaling $f_v$, and turbulent velocity dispersion $\\sigma_\\mathrm{turb}$, with the prior constraint $a_\\mathrm{in} < a_\\mathrm{out}$. The 'extended' variant adds a constant-abundance envelope between 0.32 and 1 pc. The abundance profiles that emerge are compared with pyRate, a pseudo-time-dependent chemical model run in concentric shells of the same physical model; this comparison is what carries the paper's challenge to sulphur chemistry.","core_discovery":"On the authors' own terms, the central discovery is a mismatch between what the observations require and what the chemistry predicts. The observed C$^{34}$S and CS lines toward the L1544 dust peak are best matched by step abundance profiles whose outer abundances are roughly 9 and 20 times higher, respectively, than the peak abundances produced by the pyRate chemical model; for $^{34}$SO the fitted outer abundance is about 3 times higher, though within errors, and for SO it is about 2 times lower. In all sulphur-bearing species the depletion radius is shifted outward relative to the model's abundance drop, though with large uncertainties for the SO species. For C$^{34}$S and CS, the fitted abundances are so high that the inferred optical depths reach $\\tau \\sim 4$ and $\\tau \\sim 10^5$, unusually large for a rare isotopologue and for a supposedly optically thinner line. The paper also establishes that fitting a single transition per isotopologue leaves most parameters unconstrained, with flat or multi-peaked posteriors, so the fitted profiles are degenerate; Appendix A explicitly cautions that the result 'may not be the correct or a physically accurate result.' These findings are interpreted as a challenge to current chemical networks, specifically their sulphur depletion treatment, which typically begins from a reduced elemental sulphur abundance to solve the so-called sulphur depletion problem.","pith_inferences":["Editorial inference: if the pyRate under-prediction is caused by the 'depleted' initial sulphur abundance assumption, then a testable fix is to rerun the same radiative transfer comparison with a network that begins from cosmic sulphur and follows freeze-out self-consistently; that would distinguish a missing reaction set from a wrong boundary condition.","Editorial inference: the degeneracy seen here implies that published abundance profiles of pre-stellar cores derived from single molecular transitions may carry systematic uncertainties of order the factors quoted, and comparisons between species should treat them as correlated rather than independent constraints.","Editorial inference: the newly measured $^{34}$SO rest frequency (97715.395 MHz) shifts the line velocity by 0.23 km/s relative to the old value; re-observing the 6.9 km/s blue shoulder at higher angular resolution could reveal whether it is an unresolved second component or a genuinely new kinematic feature.","Editorial inference: if sulphur is indeed more abundant in the outer core than chemical models predict, the standard 'sulphur depletion problem' may be at least partly a network completeness problem; the same comparison could be applied to other cores such as L183, where SO emission is extended, to see whether the mismatch is generic."],"forward_implications":["Sulphur chemical networks for dense cores need to be revisited: the paper argues that starting from the elemental cosmic abundance of sulphur with a consistent S-depletion process would reproduce the observed CS, C$^{34}$S, SO and $^{34}$SO abundances better than the currently 'depleted' initial-S assumption.","Single-line fits to pre-stellar core spectra should not be used to claim a unique abundance structure; multi-line fits per species are necessary to break the degeneracies the corner plots reveal, and combined fits of isotopologue pairs do not by themselves add enough constraints.","H$_{2}$CO, like HCO$^{+}$, must be present in the diffuse envelope beyond 0.32 pc to reproduce deep self-absorption, so radiative transfer models of such lines cannot stop at the traditional core boundary.","The layering inferred from line morphologies — HCO$^{+}$/H$_{2}$CO outermost, CS and c-C$_{3}$H$_{2}$ in a static middle layer, SO in a contracting intermediate layer, C$^{34}$S and HC$^{17}$O$^{+}$ deepest — is a concrete prediction that mapping the core in these lines can test.","The high fitted optical depths for C$^{34}$S ($\\tau\\sim4$) and CS ($\\tau\\sim10^5$) imply that even rare-isotopologue lines toward L1544 may be optically thick, complicating simple column-density estimates."],"supporting_citations":[{"why":"Supplies the pyRate chemical network whose fractional abundance profiles are the benchmark that the fitted step abundances are compared against.","marker":"Sipilä et al. 2015"},{"why":"Provides the 1D Bonnor-Ebert pre-stellar core physical model, including density, temperature and velocity profiles, used for every radiative transfer run.","marker":"Keto et al. 2015"},{"why":"Provides the LOC non-LTE radiative transfer code that computes the modelled spectra and the optical depth estimates.","marker":"Juvela 2020"},{"why":"Provides the MCMC sampler used to explore the abundance, radius, velocity-scaling and turbulence parameters.","marker":"Foreman-Mackey et al. 2013"},{"why":"Contributes the HC$^{17}$O$^{+}$ (1-0) spectrum and its earlier modelling, reanalysed here as the innermost tracer.","marker":"Ferrer Asensio et al. 2022"},{"why":"Established the need for an extended envelope beyond 0.32 pc for HCO$^{+}$, an approach this paper adopts and extends to H$_{2}$CO.","marker":"Redaelli et al. 2022"},{"why":"Supplies the isotope ratios ([$^{32}$S/$^{34}$S] = 22 and [$^{16}$O/$^{17}$O] = 2044) used to scale main-isotopologue pyRate abundances to the rare isotopologues.","marker":"Wilson & Rood 1994"},{"why":"Sets the fixed external-layer abundances for H$_{2}$CO and c-C$_{3}$H$_{2}$ in the extended models.","marker":"Snow & McCall 2006"}],"fun_headline_variants":["Sulphur lines 20x above model in L1544 core","L1544 observations challenge sulphur depletion models","C34S optical depth ~1e5 hints at missing freeze-out","Single-line fits leave L1544 abundances unconstrained","Sulphur chemistry off by 20x in pre-stellar core"],"cache_read_input_tokens":27648,"weakest_assumption_plain":"The load-bearing premise is that the MCMC-fitted step abundance profiles are a faithful empirical benchmark; the paper itself notes the fits can be degenerate, with flat or multi-peaked corner plots and parameters sitting at prior boundaries, and that the result may not be physically accurate.","fun_headline_variants_meta":{"raw":{"variants":["Sulphur lines 20x above model in L1544 core","L1544 observations challenge sulphur depletion models","C34S optical depth ~1e5 hints at missing freeze-out","Single-line fits leave L1544 abundances unconstrained","Sulphur chemistry off by 20x in pre-stellar core"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.0003,"raw_usage":{"total_tokens":1825,"prompt_tokens":1133,"completion_tokens":692,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":749,"completion_tokens_details":{"reasoning_tokens":606}},"tokens_in":749,"tokens_out":692,"duration_ms":6548,"temperature":1.0,"reasoning_tokens":606,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:45:57.116053+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fit several rotational lines of CS, C$^{34}$S, SO and $^{34}$SO (for example J = 1-0, 2-1 and 3-2 for CS and SO) jointly toward the L1544 dust peak with the same LOC + MCMC machinery. If the joint fit converges to outer abundances within a factor of 2 of the pyRate predictions, the single-line sulphur discrepancy is an artefact of degeneracy rather than a chemical-network failure.","supporting_citations":[{"cited_title":"2015, MNRAS, 446, 3713, 10.1093/mnras/stu2247","cited_arxiv_id":null,"evidence_quote":"Provides the 1D Bonnor-Ebert pre-stellar core physical model, including density, temperature and velocity profiles, used for every radiative transfer run."},{"cited_title":"2022, , 667, A119, 10.1051/0004-6361/202243927","cited_arxiv_id":null,"evidence_quote":"Contributes the HC$^{17}$O$^{+}$ (1-0) spectrum and its earlier modelling, reanalysed here as the innermost tracer."}],"review_version":2}