{"id":"1e6ed076-500f-487c-9f6c-12c451515438","arxiv_id":"2608.09064","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Multiphase abundance mapping of M83 reveals a persistent roughly 1.5 dex nitrogen excess in ionized gas relative to neutral gas around young clusters, indicating slow metal mixing in a massive spiral.","lead":"Using UV absorption and optical emission spectroscopy, this paper maps oxygen, sulphur, nitrogen, and iron in both neutral and ionized gas around 18 young star clusters in the spiral galaxy M83. It finds nitrogen is strongly overabundant in the hot ionized gas and stays there for several million years, suggesting freshly made metals mix slowly into the cold gas in massive spirals.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Te calibration partly trained on M83 itself; hold-out test needed to confirm nitrogen offset magnitude.","rationale":"We read the paper in good faith. The central claim is well defined and the multiphase dataset is unique. We agree with the reader's weakest-assumption analysis: the ionized-phase abundances are the most model-dependent part of the analysis, and the Appendix G temperature calibration is the single largest modeling bridge. The calibration is partly trained on M83 itself, and the paper's validation is in-sample, so it does not independently establish the accuracy of predicted Te for the target sightlines. The proposed hold-out test would directly settle whether the nitrogen offset is inflated by the calibration. We do not see an internal inconsistency in the line-flux or column-density measurements that would overturn the result; the issue is a quantitative systematic that could shift the offset by a few tenths of a dex. The reader's CONDITIONAL verdict is therefore appropriate; our concern does not change it, but the test should be required before the quantitative magnitude of the nitrogen offset is accepted as final.","tokens_in":65713,"tokens_out":15845,"duration_ms":147063,"concrete_test":"Recompute the Appendix G Te calibrations excluding all Bresolin et al. (2005) M83 H II regions from the fitting sample; then recompute T_e(NII) (via Eq. G6) and N/H for the 18 YSC sightlines. If the resulting Delta N/H offsets drop by more than ~0.3 dex, the reported 1.3-1.6 dex offset is not robust to the in-sample M83 calibration. As a cross-check, compare the directly measured T_e(NII) in the five [NII]5755-detected regions (M83-3, R1, R2, R7, R12) with the values predicted by Eq. G6; disagreement >20% would indicate a systematic T_e bias.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Delta N/H ~1.5 dex, Delta N/O >1.5 dex) rests on the ionized-phase N/H, which is computed with pyneb using electron temperatures predicted by the Appendix G strong-line calibration. Equations G3-G6 are linear fits of Te(OII), Te(OIII), and Te(SIII) to strong-line indices, calibrated on a literature sample that includes Bresolin et al. (2005) H II regions in M83 itself. Applying this same calibration to the M83 YSC sightlines is therefore partially circular: the reported in-sample WRMS ~0.02 dex (Figure 28) quantifies scatter on the calibration set, not predictive accuracy for M83's metal-rich, relatively high-density HII regions. The abundance sensitivity is significant: for [NII], d log(N/H) ~ (T_N/T_e - 0.5) d log T_e ~ 3 d log T_e at T_e ~ 7000 K, so a 20% Te bias shifts N/H by ~0.25 dex. That will not erase a 1.5 dex offset, but it could reduce it to ~1.2 dex, weakening the quantitative contrast with NGC 5253 (offsets ~0.8 dex shrinking on ~8 Myr). The paper even lists five regions with direct [NII]5755 detections (M83-3, R1, R2, R7, R12) from which the predicted Te(NII) could be checked, but no such comparison is shown.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper combines HST/COS far-UV absorption spectroscopy with cospatial VLT/MUSE and LBT/MODS optical emission-line spectroscopy for 18 young star clusters in M83 (ages ~1–6 Myr) to derive multiphase (neutral H I and ionized H II) abundances of O, S, N, and Fe. Because auroral lines are mostly undetected, the authors build and apply an empirical strong-line electron-temperature calibration (Appendix G) to compute ionized-phase abundances with pyneb. The paper's central result is a large and persistent ionized-to-neutral nitrogen offset, typically ΔN/H ≈ 1.3–1.6 dex and ΔN/O > 1.5 dex, which is interpreted as evidence that freshly produced nitrogen from massive-star winds remains in the ionized gas and does not mix into the neutral reservoir on Myr timescales. This is contrasted with NGC 5253, where such offsets shrink within ~8 Myr. The paper also reports weaker phase offsets for α-elements, an iron excess attributed to dust depletion, and radial trends that differ between the phases.","tokens_in":66026,"tokens_out":9050,"duration_ms":84659,"significance":"If the central nitrogen-offset claim holds, the paper provides a rare, cluster-scale empirical constraint on multiphase metal mixing in a massive, metal-rich spiral, complementing the dwarf-galaxy picture from NGC 5253 and the CLASSY survey. The study is notable for its genuinely cospatial, multi-element (O, S, N, Fe) design, for deriving stellar ages/metallicities from UV spectral fitting, and for being transparent about key limitations: auroral non-detections, the absence of depletion corrections, the exclusion of the galactic center, and the use of ionization corrections calibrated on previous M83 work. The main claim is also falsifiable in a concrete way: direct auroral measurements in five regions and older clusters can test whether the offset is real and whether it persists. These strengths outweigh, for now, the validation gaps discussed below, but those gaps need to be closed before the quantitative claim can be considered secure.","major_comments":[{"comment":"The T_e calibration is fit on literature H II regions that include Bresolin et al. (2005) M83 regions, so the quoted WRMS of 0.018–0.026 dex is an in-sample scatter, not a predictive accuracy for the metal-rich, relatively high-density M83 sightlines studied here. Because every ionized-phase abundance in Table 6 is computed from these predicted temperatures, the central ΔN/H claim in Section 5.2.1 inherits this systematic uncertainty. Since the [NII] abundance scales roughly as T^{0.5} exp(25000/T), a 20% bias in T_e(NII) at ~7000 K shifts log(N/H) by about 0.25 dex; this would not erase a 1.5 dex offset but could reduce it to ~1.2 dex and weaken the quantitative contrast with NGC 5253. Please add a leave-one-out or hold-out validation (in particular excluding all M83 calibration points) and propagate the resulting T_e systematics into the reported N/H and N/O values.","section":"Appendix G, Eqs. (G3)–(G6); Fig. 28"},{"comment":"The paper states that [NII]λ5755 is detected in five regions (M83-3, R1, R2, R7, R12), but Table 9 lists upper limits for R1, R2, and R12 in that line, and Table 5 is captioned as derived entirely from the strong-line calibration. No direct T_e(NII) from the claimed detections is compared with the G6-based predicted T_e(NII). Please reconcile the detection list, mark direct versus predicted temperatures in Table 5, and present a direct-versus-predicted T_e(NII) comparison for every region with a meaningful auroral measurement.","section":"Section 3.3.1, Tables 5 and 9"},{"comment":"The claim that the nitrogen offset 'does not diminish with cluster age' rests on a weak, non-significant correlation (r = 0.17, p = 0.65 for ΔN/H; the N/O trends in Fig. 18 are also non-significant). With only a 1–6 Myr age baseline, the data support 'no detected decline', not 'no decline'. Please state this as a non-detection with a bounded slope or credible interval, and adjust the abstract/summary wording from 'persistent' to 'no significant evolution over the sampled ages' where that is all the data can support.","section":"Section 5.2.1, Fig. 18"}],"minor_comments":[{"comment":"The abstract quotes 'up to ΔN/H ~ 1.5 dex', but Tables 4 and 6 imply a maximum disk offset of about 2.0 dex for M83-7 (8.07 − 6.07); please reconcile the quoted range with the tabulated values.","section":"Abstract and Section 5.2.1"},{"comment":"The printed equations contain malformed bracket/comma notation (e.g., '[Te(O,III) = (0.68±0.07),O3N2 + (9.73±0.62),]'); these should be written as proper equations, and the units and axes of Figure 27 should be clarified, since the caption appears to reverse the roles of the abundance and temperature axes.","section":"Appendix G, Eqs. (G3)–(G6)"},{"comment":"The text refers to 'M83-POS-15', which is not a target in Table 2; this appears to be a typo for M83-15 or M83-POS-2.","section":"Section F.2"},{"comment":"The caption reads 'follow the same format Figure. 7' and should read 'follow the same format as Figure 7'; similar minor grammatical issues appear in a few other captions.","section":"Figure 18 caption"}],"recommendation":"major_revision","confidential_remarks":"This is a careful observational paper with a genuinely interesting central claim, and I do not see grounds for rejection. The main issue is a validation gap rather than a demonstrated error: the strong-line T_e calibration is partly trained on M83 itself, and the quoted scatter is in-sample. The offset is large enough that plausible T_e biases are unlikely to erase it, but the exact magnitude and the persistence claim need to be made robust. The requested hold-out test and direct auroral-line comparison are cheap, concrete, and within the scope of a revision. I would also ask the authors to soften 'persistent' to 'no detected evolution over 1–6 Myr', since the age trend is non-significant and the baseline is short."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the short version. The genuinely new item is the persistent ~1.5 dex nitrogen phase offset in M83: ionized gas around clusters aged 1–6 Myr shows N/H well above the neutral gas, with no sign of the convergence seen in NGC 5253 on ~8 Myr timescales. If that survives scrutiny, it is a real constraint on metal mixing in massive spirals.\n\nWhat the paper does well: it is the first cluster-scale multiphase comparison in a metal-rich grand-design spiral covering O, S, N, and Fe rather than one or two elements. The UV absorption work is careful—multi-component Voigt fits, saturation checks, ICFs from cloudy models, and a sensible decision to leave depletion corrections out of the main analysis rather than apply Milky Way prescriptions to M83. The authors are transparent about the weak auroral lines and about the need for a strong-line Te calibration.\n\nThe soft spots are real but not fatal. The electron temperatures that feed every ionized abundance come from linear fits (Appendix G) calibrated on literature H ii regions that include M83 itself (Bresolin et al. 2005). Applying that fit to M83 YSC sightlines is partially circular: the reported ~0.02 dex scatter is on the calibration set, not independent predictive accuracy for these dense, metal-rich regions. A 20% Te bias shifts N/H by ~0.25 dex; it will not erase a 1.5 dex offset, but it could reduce it to ~1.2 dex, which matters for the quantitative contrast with NGC 5253. There are five regions with direct [NII]5755 detections where predicted Te could be checked against measurement; the paper does not show that comparison. It should.\n\nOther concerns are minor: quoted errors are random only, systematic Te/ICF uncertainties are not propagated; the galactic center exclusion is post hoc but follows Hernandez et al. (2021) and the extreme N(HI) values there. The age trends are based on small samples and should be read as suggestive.\n\nWho is this for: anyone working on chemical enrichment, ISM mixing, or young clusters. It deserves a serious referee—the dataset is valuable, the main claim is testable, and a good referee can ask for the hold-out check and better error propagation. Desk rejection would waste the effort already invested.","headline":"Persistent ~1.5 dex nitrogen phase offset in M83 is a genuinely new result, but it rests on a strong-line Te calibration partly trained on M83 itself; worth refereeing with a required hold-out or direct Te(NII) check.","tokens_in":66583,"tokens_out":2827,"would_cite":true,"duration_ms":27792,"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":"In the spiral M83, nitrogen freshly released by massive stars stays locked in the ionized gas around young clusters, showing a persistent ~1.5 dex offset from the neutral phase that does not close within 6 Myr, unlike the dwarf NGC 5253.","keywords":["galaxy chemical evolution","metal mixing","multiphase interstellar medium","abundance gradients","H II region abundances","neutral gas absorption spectroscopy","young star clusters","M83"],"falsifier":"Take deep optical spectra of a handful of these 18 H II regions to detect the auroral lines directly, [O III] λ4363, [N II] λ5755, and [S III] λ6312, and recompute the ionized-phase abundances with measured rather than predicted temperatures. If the direct-method temperatures are systematically higher than the empirical calibration predicts, the ionized N/H values drop and the nitrogen phase offset would fall well below 1 dex; alternatively, if deeper UV spectra reveal that the N I absorption used for neutral nitrogen is partially saturated, the neutral nitrogen abundance would rise and close the offset.","tokens_in":65516,"feed_emoji":"🔭","tokens_out":7227,"duration_ms":68235,"temperature":0.7,"pith_summary":"This paper tries to establish how freshly produced metals move between the two gas phases that surround young star clusters: the ionized gas lit by the stars and the larger reservoir of cold neutral gas. Using 18 clusters in the nearby spiral M83, the authors combine ultraviolet absorption lines, which see neutral gas, with aperture-matched optical emission lines, which see ionized gas, to compare oxygen, sulphur, nitrogen, and iron in both phases. Their central result is that nitrogen shows a large and persistent phase offset, roughly 1.5 dex in N/H and more than 1.5 dex in N/O, with freshly made nitrogen staying in the ionized gas and failing to mix into the neutral reservoir over the ~1–6 Myr ages of the clusters. This contrasts with the low-mass dwarf NGC 5253, where the same offset closes within about 8 Myr. If right, it means mixing efficiency depends on environment: the deep potential well of a massive spiral confines wind-driven metals to the immediate star-forming region for at least several million years.","feed_headline":"M83's young stars trap fresh nitrogen in ionized gas","feed_subtitle":"A ~1.5-dex ionized-neutral nitrogen gap persists over 6 Myr in a massive spiral, unlike dwarf NGC 5253.","key_machinery":"The load-bearing comparison is the multiphase abundance offset $Δ$(X/H) = (X/H)$_{HII}$ − (X/H)$_{HI}$, computed element by element for cospatial ionized and neutral gas. Neutral abundances come from multi-component Voigt-profile fits to HST/COS ultraviolet absorption lines (N I, S II, O I via S II, Fe II) with CLOUDY-based ionization corrections; ionized abundances come from reddening-corrected optical emission lines reduced with pyneb. Because auroral lines are undetectable in most of M83's metal-rich H II regions, the paper builds an empirical multi-zone electron-temperature calibration: linear fits, restricted to the metal-rich regime (12+log(O/H) > 8.0) and anchored on literature direct-method H II regions, that predict T$_e$(O III) from the O3N2 diagnostic, T$_e$(S III) from S3O3, and T$_e$(O II) from a combined DESIRED metallicity estimator. These predicted temperatures are what convert raw line fluxes into the ionized-phase abundances whose offsets carry the central claim.","core_discovery":"The paper claims that in M83, nitrogen released by the current generation of massive O and B stars remains trapped in the ionized gas surrounding young clusters and does not transfer into the cold neutral gas on timescales of at least 6 Myr. The evidence is a systematic offset between neutral-phase and ionized-phase abundances: ionized gas shows N/H about 1.3–1.6 dex higher and N/O more than 1.5 dex higher than cospatial neutral gas, across all 18 clusters with no sign of closing with cluster age. Oxygen and sulphur, by contrast, show smaller and age-dependent offsets consistent with prompt core-collapse supernova enrichment on ~3–5 Myr timescales, while the ~1 dex iron offset is attributed mainly to dust depletion rather than nucleosynthesis. The authors interpret the nitrogen behaviour as a signature of feedback-regulated enrichment in a deep gravitational potential: strong stellar winds efficiently enrich the local ionized medium, but the potential well and dense interstellar medium inhibit dispersal of the enriched material into the surrounding H I reservoir, unlike in the low-mass starburst NGC 5253 where such offsets vanish within about 8 Myr.","pith_inferences":["If the offset really persists beyond 6 Myr, clusters in M83 at ages of ~10–30 Myr should still show elevated N/O in their ionized surroundings; observing such older clusters would test whether the confinement timescale scales with potential-well depth.","The nitrogen missing from the neutral reservoir may reside in warm molecular gas: following the paper's own finding that the center of M83 holds mostly warm H2, a testable extension is to search for the 'missing' nitrogen in CO-dark molecular gas using infrared H2 lines.","A corollary the authors leave implicit is that outflows from M83 should be metal-poor relative to the star-formation rate compared with dwarfs like NGC 5253, since metals stay trapped in the disk; this is testable through circumgalactic absorption-line surveys.","The empirical multi-zone temperature calibration is a portable tool: applied to other metal-rich spirals where auroral lines are undetectable, it would allow the same multiphase comparison without direct temperature measurements, with the caveat that its accuracy depends on how well the calibration sample matches the target's physical conditions."],"forward_implications":["In a massive, metal-rich spiral, freshly synthesised nitrogen from massive stars remains confined to H II regions for at least ~6 Myr, so neutral-gas abundances understate the recent enrichment of the interstellar medium.","Phase offsets of the alpha elements oxygen and sulphur grow with cluster age, consistent with core-collapse supernova enrichment appearing in the ionized phase on ~3–5 Myr timescales.","The roughly 1 dex iron excess in ionized gas relative to neutral gas is dominated by dust depletion of iron in the cold phase rather than by nucleosynthesis, so iron phase offsets should not be read as enrichment clocks.","The persistence of the nitrogen offset is environment-dependent: the same enrichment channel closes within ~8 Myr in the low-mass galaxy NGC 5253 but not in M83, linking chemical mixing efficiency to the depth of the galactic potential well.","Adopting spectroscopically derived rather than photometric cluster ages shortens the inferred mixing timescale in NGC 5253 to about 8 Myr, sharpening the contrast between the dwarf and the spiral."],"supporting_citations":[{"why":"The NGC 5253 multiphase study whose ~8 Myr convergence of nitrogen offsets the M83 result directly contradicts, supplying the comparative baseline.","marker":"V. Abril-Melgarejo et al. 2024"},{"why":"Prior multiphase study of M83 that supplies H I column densities, the CLOUDY ionization-correction methodology, and the earlier ~100 pc mixing comparison being extended.","marker":"S. Hernandez et al. 2021"},{"why":"Established the HST/COS-based neutral-gas abundance method and the CLOUDY photoionisation models for M83 sightlines used to compute ionization corrections.","marker":"B. L. James et al. 2014"},{"why":"The metal-rich H II region sample, including M83 itself, whose direct-method temperatures and abundances anchor the empirical electron-temperature calibration.","marker":"F. Bresolin et al. 2005"},{"why":"Provides the ionized-gas ionization-correction-factor prescriptions adopted for nitrogen, sulphur, and iron abundances.","marker":"Y. I. Izotov et al. 2006"},{"why":"The DESIRED strong-line calibrations used to build the combined Te(O II) estimator and to identify the best temperature diagnostics.","marker":"F. F. Rosales-Ortega et al. 2026"},{"why":"The technique for converting S II (and P II) column densities into neutral oxygen abundances, used because O I λ1302 is saturated.","marker":"B. James & A. Aloisi 2018"},{"why":"Spectroscopic ages for the NGC 5253 COS sample that motivate revising the dwarf's enrichment-mixing timescale down to ~8 Myr.","marker":"S. Hernandez et al. 2026"}],"fun_headline_variants":["M83's ionized gas keeps young-star nitrogen for >6 Myr","Nitrogen stays put in M83's ionized phase, defying mixing","M83: fresh nitrogen trapped near young stars, not mixing","In M83, nitrogen from young stars skips neutral gas for 6 Myr"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The ionized-phase nitrogen abundances rest on electron temperatures that were not measured but predicted from strong-line ratios calibrated on other H II regions; if those temperature predictions run systematically hot or cold for M83's specific metal-rich, high-density conditions, every ionized abundance shifts and the claimed ~1.5 dex nitrogen offset could shrink or vanish.","fun_headline_variants_meta":{"raw":{"variants":["M83's ionized gas keeps young-star nitrogen for >6 Myr","Nitrogen stays put in M83's ionized phase, defying mixing","M83: fresh nitrogen trapped near young stars, not mixing","In M83, nitrogen from young stars skips neutral gas for 6 Myr"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000295,"raw_usage":{"total_tokens":1810,"prompt_tokens":1135,"completion_tokens":675,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":751,"completion_tokens_details":{"reasoning_tokens":593}},"tokens_in":751,"tokens_out":675,"duration_ms":7412,"temperature":1.0,"reasoning_tokens":593,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T00:16:23.344751+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take deep optical spectra of a handful of these 18 H II regions to detect the auroral lines directly, [O III] λ4363, [N II] λ5755, and [S III] λ6312, and recompute the ionized-phase abundances with measured rather than predicted temperatures. If the direct-method temperatures are systematically higher than the empirical calibration predicts, the ionized N/H values drop and the nitrogen phase offset would fall well below 1 dex; alternatively, if deeper UV spectra reveal that the N I absorption used for neutral nitrogen is partially saturated, the neutral nitrogen abundance would rise and close the offset.","supporting_citations":[],"review_version":1}