{"id":"df6c421c-5196-407b-a03c-daf4fcaa19ee","arxiv_id":"2507.02054","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"For the z=6.81 galaxy COS-2987, [OIII] line ratios from JWST and ALMA rule out a single-component ionized gas and favor a two-phase ISM with a compact dense component and an extended diffuse component.","lead":"A JWST and ALMA study of a galaxy 12.9 billion light-years away finds that its gas cannot be described by one uniform temperature and density, requiring multiple gas phases. The result suggests that JWST-only observations miss part of the gas in early galaxies and that combining JWST with ALMA is needed to reveal the full structure.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Two-component model is underdetermined: the claimed phase properties (Te, ne, V1/V2 ~ 1/300) depend on the fixed assumption O++/H+ = 1e-4 and on hand-picked parameters; a systematic exploration of the degeneracy space is missing.","rationale":"The reader's verdict of CONDITIONAL is appropriate. My concern aligns with the reader's weakest_assumption: the quantitative two-component solution (Section 5.2) is not unique and depends on fixed O++/H+ = 1e-4 and hand-picked Te/ne values. This is a genuine, load-bearing concern because the abstract and conclusions present the two-component parameters and volume ratio as key results. However, the qualitative central claim — that a single homogeneous ionized gas with uniform Te and ne cannot explain the observed ratios, and that the ISM is density-stratified — is robust: it is validated by Figure 3 (the observed point lies outside the homogeneous model grids), by the dust-variation test in Section 5.1, and by the absence of spatial/kinematic separation in Appendices B and C (which, while consistent, do not rule out a stratified but unresolved medium). The paper's qualitative novelty is real and supported; the quantitative parameterization is a heuristic demonstration rather than a fitted solution. My proposed grid search would settle whether the specific claimed phase properties survive a systematic exploration. Given that the concern is about the strength of quantitative claims rather than the validity of the qualitative finding, CONDITIONAL is the correct verdict — the paper should be accepted with the requirement that the model degeneracy be quantified or the quantitative claims softened. I agree with the reader on the weakest assumption; my recommendation does not change the verdict, but provides a concrete test to sharpen it.","tokens_in":21867,"tokens_out":2029,"duration_ms":26624,"concrete_test":"Run a broad grid search over the two-component model: vary Te,i (2,000–40,000 K), ne,i (10–10^4 cm^-3), and O++/H+ independently (e.g., 0.3e-4 to 3e-4, perhaps tied to photoionization models), and find all parameter combinations that reproduce the observed [OIII]88/5008 and [OIII]4364/5008 ratios within 1-sigma. Report the degenerate manifold of acceptable (Te, ne, O++/H+, V1/V2) solutions. If the acceptable ranges for component properties span factors of several (which is expected), the specific phase properties and volume ratio should be reframed as illustrative; if the ranges are narrow under physically motivated priors, the quantitative claim is strengthened.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central quantitative claim — two discrete ionized phases with (Te,1, ne,1) = (26,000 K, 500 cm^-3), (Te,2, ne,2) = (8,000 K, 50 cm^-3), V1/V2 ~ 1/300 — relies on fixing O++/H+ = 1e-4 in both components (Eq. 4, Section 5.2). The paper itself acknowledges non-uniqueness ('other sets of physical conditions could reproduce the observed line ratios'). This is the load-bearing weakness: with only three observed line ratios and two components (eight free parameters: Te, ne, O++/H+, V for each), the model is severely underdetermined. Variations in O++/H+ between components, or a continuous density/temperature distribution, will shift the inferred phase properties and volume ratio, even though the qualitative rejection of a single homogeneous phase remains intact. Because the abstract and conclusions emphasize the specific 26,000 K / 8,000 K / 1:300 solution, the quantitative claims go beyond what the data uniquely constrain. The qualitative discovery — that the [OIII]88/5008 and [OIII]4364/5008 ratios cannot be jointly reproduced by a single phase, even after spatially varying dust corrections (Section 5.1) — is well supported and is the paper's real result; the two-component parameterization is a plausible but non-unique illustration.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"Usui et al. present JWST/NIRSpec IFS and ALMA observations of the z=6.81 star-forming galaxy COS-2987, detecting [OII] λλ3727,3730, [OIII] 4364, [OIII] λλ4960,5008, Hβ, Hα, and [OIII] 88 μm. From the optical lines alone they infer A_V ≈ 0.14 mag, T_e([OIII]) ≈ 2.0×10^4 K, n_e([OII]) ≈ 500 cm^-3, and 12+log(O/H) ≈ 7.67. The central diagnostic combines [OIII] 4364/5008 with [OIII] 88/5008; COS-2987 lies outside the PyNeb single-phase T_e–n_e grid, and a pixel-by-pixel dust correction does not bring it inside. The authors then introduce a two-component toy model with a hot/dense component (T_e ≈ 26,000 K, n_e ≈ 500 cm^-3) and a cool/diffuse component (T_e ≈ 8,000 K, n_e ≈ 50 cm^-3), showing that the observed ratios can be reproduced with V_1/V_2 ≈ 1/300. They explicitly note that this solution is not unique. The paper concludes that the ISM is density-stratified and that JWST-only measurements miss a diffuse low-density phase.","tokens_in":22215,"tokens_out":10299,"duration_ms":115104,"significance":"The qualitative result is important: it provides direct evidence at z ≈ 6.8 that a homogeneous, single-phase ionized-gas model fails to jointly explain optical and FIR [OIII] ratios, and it illustrates the need for ALMA to characterize high-redshift ISM. The inclusion of several literature galaxies strengthens the claim that the discrepancy is not unique to this object. The authors are transparent that the two-component solution is a toy and not unique, which is a strength. The weakness is that the specific numeric parameters in the abstract and conclusions (26,000 K / 8,000 K / 1:300) are not actually constrained by the data; the paper's robust product is the falsification of the single-phase model plus an illustrative stratified-ISM picture.","major_comments":[{"comment":"The quantitative two-component solution is underdetermined. The three [OIII] lines provide only two independent ratios, while the model in Eq. (4) has at least five free parameters (T_e,1, n_e,1, T_e,2, n_e,2, V_1/V_2) plus the fixed assumption O++/H+ = 10^-4 in both components. The paper explicitly notes that other parameter sets could reproduce the data, but the abstract and Conclusion (4) present (T_e,1, n_e,1) = (26,000 K, 500 cm^-3), (T_e,2, n_e,2) = (8,000 K, 50 cm^-3), and V_1/V_2 ~ 1/300 as the inferred result. Please add a caveat in the abstract and conclusions that these are illustrative values of a toy model, and provide at least a brief quantitative indication of the degeneracy (e.g., how V_1/V_2 changes if the two components are allowed to have different O++/H+, or if a continuous density/temperature distribution is assumed). The qualitative rejection of a homogeneous single-phase model does not depend on this issue, but the specific numeric claims do.","section":"§5.2, Eq. (4), Figure 4; abstract and Conclusion (4)"},{"comment":"The clumpy-dust test is limited by the resolution of the Balmer-decrement map. The A_V map is constructed from Voronoi-binned Hα/Hβ and sets A_V = 0 outside the 3σ regions, so unresolved dust clumps would not be captured by this test. The conclusion that 'clumpy dust distribution is not the primary cause' should therefore be stated with the caveat that the test probes dust structure on the scales resolved by the PSF-matched map; an extreme unresolved clumpy geometry is not strictly excluded, although it would require fine-tuning. This does not change the main conclusion but makes the claim precise.","section":"§5.1 and Figure 3 (right)"}],"minor_comments":[{"comment":"Typo: 'one-dimentional' should be 'one-dimensional'.","section":"§2.1"},{"comment":"The section heading contains a duplicated article: 'Interpretation of the the density-stratified ionized ISM'.","section":"§5.3"},{"comment":"The top-right panel would be easier to read if the two model curves were labeled (e.g., '88/5008' and '4364/5008') and if the intersections with the observed 1σ bands were marked to show the allowed range in V_1/V_2.","section":"Figure 4"},{"comment":"The statement that 'other sets of physical conditions could reproduce the observed line ratios, such as a higher T_e for component 1 and a higher n_e for component 2' would be more informative with a concrete example or a small grid; as written, the reader cannot judge the size of the degeneracy.","section":"§5.2"}],"recommendation":"minor_revision","confidential_remarks":"This is a solid ApJL-style discovery paper. The qualitative single-phase failure is convincing and new at z~7, and the literature compilation adds generality. The main risk is overinterpretation of the two-component parameters; the authors are transparent in the body, but the abstract and conclusions overstate the specificity. A minor revision with explicit caveats and a brief degeneracy illustration should be sufficient. The paper relies heavily on the companion paper (Mawatari et al. 2025) for data reduction, but the [OIII] ratio analysis is sufficiently self-contained for this Letter."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid, useful confirmation of the emerging two-phase ISM picture at z~7, built on a high-quality JWST IFS + ALMA dataset for one galaxy. The headline interpretation (two discrete phases with 26,000 K/500 cm^-3 and 8,000 K/50 cm^-3, V1/V2~1/300) is a demonstration, not a unique solution, and the paper is mostly honest about that.\n\nWhat's new: COS-2987 is the first z>6 galaxy with NIRSpec IFS (not MOS) plus ALMA [OIII]88, which avoids slit-loss systematic. The diagnostic diagram ([OIII]4364/5008 vs [OIII]88/5008) is a clean way to show the homogeneous model fails. The pixel-by-pixel dust test is careful, and it convincingly rules out dust as the main culprit. The paper also compiles literature points and shows the trend is general, though for that it leans on Harikane et al. (2025b), which is appropriately cited.\n\nWeak spots: as the authors admit, the two-component model is underdetermined. With only two independent line ratios, five free parameters (two Te, two ne, V1/V2) plus the assumed O++/H+ = 1e-4, the specific numbers are hand-picked. A referee should push for a systematic degeneracy scan, including varying O++/H+ and allowing continuous density/temperature distributions. The abstract and conclusions quote the specific solution without the caveats that appear in Section 5.2, so the quantitative take-home is more fragile than the qualitative finding. That qualitative finding — a single homogeneous phase cannot jointly reproduce the observed [OIII] ratios, even after dust correction — is robust and well supported.\n\nReadership: anyone working on high-z ISM diagnostics, JWST+ALMA synergies, or reionization. It is a nice case study for how optical-only measurements can miss a diffuse component.\n\nRecommendation: send it to a serious referee. The data are good, the analysis is transparent, and the qualitative result matters. Ask the authors to add a degeneracy exploration and to bring the abstract in line with the non-uniqueness acknowledged in Section 5.2.","headline":"Solid JWST+ALMA case study confirming the two-phase ISM picture at z~7; the quantitative solution is underdetermined and the abstract oversells it, but the qualitative rejection of a single-phase model is robust.","tokens_in":22898,"tokens_out":3493,"would_cite":true,"duration_ms":35990,"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":"The galaxy COS-2987 at z=6.81 has an ionized ISM that cannot be described by one temperature and density; two gas phases are required.","keywords":["high-redshift galaxies","density-stratified ISM","[OIII] 88 micron","electron temperature","electron density","JWST NIRSpec IFS","ALMA","interstellar medium inhomogeneity"],"falsifier":"Measure [OIII]88 and [OIII]5008 at sub-arcsecond resolution with ALMA and map their spatial extents: if [OIII]88 is not significantly more extended than [OIII]5008, the large-volume diffuse component that carries the model would be ruled out, and the ratio discrepancy would need a different mechanism. A second check would measure O++/H+ independently in the diffuse phase from a density-independent recombination line; if it differs from $10^{-4}$, the derived volume ratio is wrong.","tokens_in":21680,"feed_emoji":"🌌","tokens_out":8223,"duration_ms":80843,"temperature":0.7,"pith_summary":"The paper sets out to show that the interstellar medium of the star-forming galaxy COS-2987 at $z=6.81$ is not a single uniform body of ionized gas. Combining JWST/NIRSpec integral-field spectra of $[\\mathrm{O\\,III}]\\,\\lambda4364$, $\\lambda5008$, and hydrogen recombination lines with ALMA's detection of $[\\mathrm{O\\,III}]\\,88\\,\\mu$m, it finds that the observed $[\\mathrm{O\\,III}]88/5008$ and $[\\mathrm{O\\,III}]4364/5008$ ratios cannot be reproduced by a homogeneous ionized gas with one electron temperature and one electron density. It then shows that two discrete gas phases—compact gas at about 26,000 K and 500 cm$^{-3}$, plus an extended phase at about 8,000 K and 50 cm$^{-3}$ with roughly 300 times the volume—reproduce both ratios together. If this holds, JWST-only measurements of early galaxies are systematically incomplete, missing the diffuse gas that dominates far-infrared oxygen emission.","feed_headline":"A z=6.81 galaxy's gas requires two phases","feed_subtitle":"Optical lines alone miss the cool, diffuse gas that dominates the galaxy's far-infrared oxygen emission.","key_machinery":"The load-bearing tool is the combined $[\\mathrm{O\\,III}]$ line-ratio diagnostic. Because $\\lambda4364$, $\\lambda5008$, and $88\\,\\mu$m are emitted by the same doubly ionized oxygen ion, their ratios determine $T_e$ and $n_e$ without abundance assumptions; the ratio $\\lambda4364/\\lambda5008$ is temperature-sensitive, while $88\\,\\mu$m$/\\lambda5008$ is density- and temperature-sensitive. The large difference in critical densities—510 cm$^{-3}$ for $88\\,\\mu$m versus roughly $6.8\\times10^5$ and $3.0\\times10^6$ cm$^{-3}$ for $\\lambda5008$ and $\\lambda4364$—lets the three lines act as probes of different phases. The quantitative model is simple additivity of volume emissivities, $L_{\\rm line}=\\epsilon_{\\rm line,1}V_1+\\epsilon_{\\rm line,2}V_2$, with emissivities computed under an assumed $\\mathrm{O}^{++}/\\mathrm{H}^+=10^{-4}$. A two-component mix places the composite ratios on the line connecting the two single-phase grid points, and matching the observed ratios fixes the volume ratio at roughly $1{:}300$.","core_discovery":"The central claim is that the three $[\\mathrm{O\\,III}]$ lines of COS-2987—the optical auroral line $\\lambda4364$, the optical nebular line $\\lambda5008$, and the far-infrared fine-structure line $88\\,\\mu$m—cannot all originate from a single phase of ionized gas with one density and one temperature. A homogeneous model that matches the high $[\\mathrm{O\\,III}]88/5008$ ratio would predict a $[\\mathrm{O\\,III}]4364/5008$ ratio that disagrees with the measurement, and vice versa. The paper's two-component toy model, with component 1 at $(T_e,n_e)=(26{,}000~\\mathrm{K},500~\\mathrm{cm}^{-3})$ and component 2 at $(8{,}000~\\mathrm{K},50~\\mathrm{cm}^{-3})$ and volume ratio $V_1/V_2\\simeq 1/300$, reproduces both observed ratios simultaneously. The authors stress that this parameter set is not unique, and they rule out inhomogeneous dust as the primary explanation by showing that pixel-by-pixel dust corrections leave the discrepancy intact. The qualitative conclusion is that the galaxy's ionized ISM is density-stratified: compact hot regions embedded in a much larger reservoir of cool diffuse gas.","pith_inferences":["If the two-phase structure is typical at high redshift, gas-phase metallicities from direct-$T_e$ methods may be biased because optical $[\\mathrm{O\\,III}]$ and $[\\mathrm{O\\,II}]$ lines weight different phases; combining far-infrared lines would be needed to recover a mass-weighted abundance.","The exact volume ratio of 1:300 is model-dependent; the robust statement is qualitative stratification. A testable prediction is that high-resolution ALMA will find $[\\mathrm{O\\,III}]88$ more extended than $[\\mathrm{O\\,III}]5008$, something the current beam-smoothed radial profiles cannot yet rule out.","One speculative route the authors leave open is that very massive stars produce the 26,000 K component; rest-frame UV spectra showing strong He II 1640 emission or an enhanced N/O ratio would support that interpretation."],"forward_implications":["JWST-only optical measurements of high-redshift galaxies systematically miss a diffuse ionized component, so densities and temperatures derived from optical lines alone are weighted toward dense gas.","The same diagnostic applied to other high-$z$ galaxies with strong $[\\mathrm{O\\,III}]88$ shows the same offset from homogeneous grids, suggesting the two-phase pattern is common rather than unique to COS-2987.","A large-volume, cool, low-density phase is the natural site for the strong far-infrared $[\\mathrm{O\\,III}]88$ emission, implying a porous, \"picket-fence\" geometry in which ionizing photons leak from dense regions into surrounding diffuse gas.","For COS-2987 the inferred Lyman-continuum escape fraction of 4–14% is consistent with this porous geometry, connecting the ISM structure result to cosmic reionization.","Higher-angular-resolution ALMA observations can map $[\\mathrm{O\\,III}]88$ directly and test whether the diffuse component is spatially extended, as the model predicts."],"supporting_citations":[{"why":"Companion paper supplying the NIRSpec IFS observations, data reduction, stellar populations, and the Lyman-continuum escape fraction used here.","marker":"Mawatari et al. 2025"},{"why":"Initial ALMA detection and flux measurement of [OIII]88 in COS-2987, the far-infrared anchor of the line-ratio diagnostic.","marker":"Witstok et al. 2022"},{"why":"PyNeb code used to compute emissivities, model grids, electron temperature, and electron density.","marker":"Luridiana et al. 2015"},{"why":"Source of the critical densities for [OIII]4364, 5008, and 88 micron that motivate the multi-phase interpretation.","marker":"Osterbrock & Ferland 2006"},{"why":"Local LMC-N11 case showing [OIII]88 traces extended low-density gas, the precedent for density stratification.","marker":"Lebouteiller et al. 2012"},{"why":"Local dwarf galaxy study of the same optical/FIR [OIII] ratios that attributes discrepancies to clumpy dust, the alternative hypothesis tested and rejected.","marker":"Chen et al. 2024"},{"why":"Independent modeling of z=6-9 galaxies reaching a similar two-component, density-stratified conclusion.","marker":"Harikane et al. 2025b"},{"why":"Empirical relation between [OIII]88/[CII]158 and escape fraction used to infer f_esc approximately 4-14%.","marker":"Ura et al. 2023"}],"fun_headline_variants":["JWST+ALMA show z=6.81 galaxy needs two gas phases","Two gas phases explain z=6.81 galaxy's oxygen lines","Single-phase model fails: z=6.81 galaxy has stratified ISM","ALMA and JWST expose dual-density ISM in early galaxy"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The two-component solution assumes the ionized gas is exactly two internally uniform phases sharing the same oxygen abundance ratio O++/H+ = $10^{-4}$; if abundances differ between phases or the gas has continuous density and temperature gradients, the derived component properties and the 1:300 volume ratio would change.","fun_headline_variants_meta":{"raw":{"variants":["JWST+ALMA show z=6.81 galaxy needs two gas phases","Two gas phases explain z=6.81 galaxy's oxygen lines","Single-phase model fails: z=6.81 galaxy has stratified ISM","ALMA and JWST expose dual-density ISM in early galaxy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000305,"raw_usage":{"total_tokens":1883,"prompt_tokens":1211,"completion_tokens":672,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":827,"completion_tokens_details":{"reasoning_tokens":592}},"tokens_in":827,"tokens_out":672,"duration_ms":6397,"temperature":1.0,"reasoning_tokens":592,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T20:39:53.947339+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure [OIII]88 and [OIII]5008 at sub-arcsecond resolution with ALMA and map their spatial extents: if [OIII]88 is not significantly more extended than [OIII]5008, the large-volume diffuse component that carries the model would be ruled out, and the ratio discrepancy would need a different mechanism. A second check would measure O++/H+ independently in the diffuse phase from a density-independent recombination line; if it differs from $10^{-4}$, the derived volume ratio is wrong.","supporting_citations":[],"review_version":1}