{"id":"57b76cbe-4669-4cab-a889-58d70352fa48","arxiv_id":"2505.06340","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A spatially resolved ALMA survey of 39 main-sequence galaxies at z~4-6 finds diverse morphologies, an Arp 220-like [CII] deficit, and elevated [CII]/FIR ratios in a previously unexplored surface brightness range.","lead":"The ALMA-CRISTAL survey presents kiloparsec-scale ALMA observations of 39 star-forming galaxies at redshifts 4 to 6, paired with HST and JWST imaging. It reveals varied gas, dust, and stellar morphologies, including an extreme [CII] deficit at z~5 and a new high-redshift [CII]/FIR surface brightness relation.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The [CII]/FIR and ΣFIR results rest on FIR luminosities that for most CRISTAL galaxies come from a single ALMA Band 7 continuum point and assumed dust temperatures, with systematic SED uncertainties not propagated into the z≈0 comparison.","rationale":"The reader's weakest assumption correctly identifies FIR luminosity as the load-bearing input. Every quantitative ISM claim in the paper is a ratio or surface density built from LFIR: the CRISTAL-10 Arp 220-like deficit (Section 8.1), the CRISTAL-13 PDR comparison (Section 8.2), and the global trends in Figure 14 (Section 9). For the case studies the dust temperatures are stated, but for the global sample LFIR is adopted from Mitsuhashi et al. (2024) without propagating dust-SED uncertainties. At z≈5, ALMA Band 7 samples rest-frame ~158 µm, near the peak of a cold dust SED, and a single photometric point cannot break the T_dust–β degeneracy; plausible variations shift LFIR by factors of 2–3, moving points diagonally in Figure 14. This directly threatens the novelty claims: the 'previously unexplored ΣFIR range' and 'elevated [CII]/FIR ratios' could be reduced or vanish under alternative SED choices. I also considered the [CII]-selected nature of the sample as a possible bias, but that affects interpretation rather than the empirical parameter-space coverage, and the selection is described transparently. The in-prep companion papers support several descriptive claims but are not the numerical foundation of Figure 14. On the positive side, the ALMA fluxes are cross-checked against ALPINE and against independent visibility-plane measurements, and the public data release is a valuable product that stands regardless of the SED assumptions. Because the reader already identified this same concern and assigned CONDITIONAL, no adjustment is needed.","tokens_in":47908,"tokens_out":6948,"duration_ms":69648,"concrete_test":"Compute LFIR for all 24 continuum-detected CRISTAL galaxies from the Table 2 Band 7 flux densities using a grid of modified blackbody SEDs with T_dust = 30, 40, 50, 60 K and β = 1.5, 2.0, 2.5, including the CMB correction at z≈5. Combine those LFIR values with the dust sizes from Mitsuhashi et al. (2024) to recompute ΣFIR, then re-derive the [CII]/FIR versus ΣFIR relation (Eq. 2) and re-make Figure 14. If the CRISTAL points no longer occupy ΣFIR ≈ 10^9–10^11 Lsun/kpc^2 with ratios systematically above the local relation at fixed ΣFIR, the central scientific claims of Section 9 do not survive.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's quantitative ISM claims—CRISTAL-10 as an Arp 220-like [CII] deficit, CRISTAL-13 as a PDR/feedback case, and the global trends in Figure 14—all depend on accurate FIR luminosities. In Section 9, LFIR is taken from Mitsuhashi et al. (2024) SED fits; for most z≈5 CRISTAL galaxies the only far-IR constraint is a single ALMA Band 7 continuum measurement near rest-frame 158 µm. The integrated 8–1000 µm LFIR is therefore an extrapolation controlled by assumed dust temperature and emissivity index (T_dust, β). The case studies state their assumptions explicitly (T_dust = 50 K for CRISTAL-10, 40 K for CRISTAL-13) and show a ±10 K bar, but the same degeneracy is not propagated to the global points in Figure 14. A factor of ~2–3 change in LFIR—well within plausible SED variations—would shift points diagonally in both panels, potentially removing the claimed 'previously unexplored ΣFIR range' and reducing the asserted 3–10× offset from z≈0 galaxies. Moreover, Eq. (2) is presented without fit uncertainties, so the significance of the offset cannot be assessed once SED systematics are included.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents the CRISTAL survey overview, an ALMA Cycle 8 Large Program targeting 39 main-sequence star-forming galaxies at 4≲z≲6 with Band 7 [CII] 158 μm and dust continuum observations at ~kpc resolution, combined with HST/WFC3 and JWST/NIRCam imaging. The authors describe the sample selection, ALMA calibration and imaging, and the final data products, and report global line fluxes and continuum fluxes measured with three complementary methods (aperture photometry, 2D Gaussian fits, and S/N-threshold integration) that agree within ~10% except for four interacting systems. The science highlights include: (i) diverse morphologies and kinematics, with [CII] often more extended than the stellar light; (ii) CRISTAL-10, where a dust-bright region lacking a UV counterpart has [CII]/FIR ≈ 2.5×10⁻⁴, comparable to Arp 220; (iii) CRISTAL-13, where six young stellar clumps are spatially anti-correlated with the [CII] emission; and (iv) a global analysis showing that CRISTAL galaxies follow a [CII]/FIR versus FIR luminosity deficit trend shifted to higher luminosities, and occupy a Σ_FIR range of roughly 10⁹–10¹¹ L☉ kpc⁻² with higher [CII]/FIR at fixed Σ_FIR than local galaxies.","tokens_in":48213,"tokens_out":7052,"duration_ms":71756,"significance":"The paper delivers a valuable public dataset and a well-documented reduction, with multiple cross-checks: three independent flux measurement methods agree, and the fluxes are consistent with ALPINE and with visibility-plane measurements from Mitsuhashi et al. (2024). The resolved [CII]-to-stellar-light comparisons, the kinematic/morphological census, and the identification of extended [CII] beyond stellar light are important contributions to high-redshift ISM studies. However, the quantitative ISM claims—the Arp 220-like deficit in CRISTAL-10, the PDR/feedback interpretation in CRISTAL-13, and the global offset in the [CII]/FIR–Σ_FIR plane—depend on FIR luminosities that are not fully characterized: most global values come from SED fits with a single Band 7 continuum constraint, and the case-study values rely on assumed dust temperatures. The claims are plausible and interesting, but the lack of propagated uncertainties currently makes the quantitative conclusions stronger than the underlying evidence.","major_comments":[{"comment":"The manuscript does not state the wavelength range used to define L_FIR for the CRISTAL and Spilker et al. high-redshift samples, while the local comparison sample (Lutz et al. 2016; Herrera-Camus et al. 2018a) is, by standard usage, the 42.5–122.5 μm FIR band. If the high-z L_FIR values are total 8–1000 μm luminosities (as typically output by SED fitting codes such as CIGALE or MAGPHYS), the [CII]/L_FIR ratios and Σ_FIR values in Fig. 14 are systematically biased relative to the z≈0 sample. Please state the adopted passband definitions explicitly for all samples and, if they differ, recompute the comparison with matched definitions.","section":"Section 9, Eqs. (1)–(2) and Fig. 14"},{"comment":"Equation (2) is presented with fixed numerical coefficients and no quoted uncertainties, and the CRISTAL points in Fig. 14 have no error bars reflecting SED systematics. For most CRISTAL galaxies the only far-IR constraint is a single ALMA Band 7 continuum point (Mitsuhashi et al. 2024), so L_FIR is controlled by assumed dust temperature and emissivity index; a factor of ~2–3 change in L_FIR is plausible. Such variations shift points diagonally in both panels of Fig. 14 and can alter the claimed 3–10× offset and the 'previously unexplored' Σ_FIR range. I request that the authors quote the uncertainties on the coefficients of Eq. (2) and provide a sensitivity band for the CRISTAL sample, for example for ±10 K in T_dust or for a plausible range of β.","section":"Section 9, Eq. (2) and Fig. 14"},{"comment":"The quantitative case-study results rely on L_FIR derived from the Band 7 continuum assuming T_dust = 50 K (CRISTAL-10) and 40 K (CRISTAL-13). The ±10 K sensitivity bar is shown only for CRISTAL-10 in Fig. 12, and the text acknowledges that 'the dust temperature could potentially be higher'; no such sensitivity is shown for CRISTAL-13. The qualitative interpretations—Region 1 as Arp 220-like and the eastern component as a dense PDR—would be better supported by reporting the full range of [CII]/F_IR for a plausible range of (T_dust, β) and by clearly stating how the conclusions would change if the adopted temperatures are incorrect.","section":"Sections 8.1 and 8.2, Figs. 12–13"}],"minor_comments":[{"comment":"Typo: 'the initial CRISTAL sample consist of 19 galaxies' should read 'consists of 19 galaxies'.","section":"Section 3.1"},{"comment":"The sentence 'The [C II] and continuum fluxes in Table 2 agree on average within a ~10% and ~20% with the fluxes measured in Ikeda et al. (2025) and Mitsuhashi et al. (2024)' is ambiguous: it is not clear which quantity agrees to 10% and which to 20%, or with which reference.","section":"Section 7.1"},{"comment":"The caption for the right panel states 'best fit to the data at z≈0 and z≈3–6 (this work; Eq. 2)', but the fit includes both CRISTAL galaxies and Spilker et al. (2016) systems; please specify the exact sample and Σ_FIR range used for the fit.","section":"Figure 14 caption"},{"comment":"Typo: 'There results are further supported' should be 'These results are further supported'.","section":"Section 2.5"},{"comment":"The paper notes that the JvM effect is not corrected, but the released data products would benefit from an explicit statement in the data-release documentation telling users whether the correction should be applied; a sentence in Section 5.2 or the repository documentation would suffice.","section":"Section 5.2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a survey overview in a series with many companion papers; the data reduction and release appear to be of high quality. The main concern is that the quantitative ISM claims in Sections 8–9 are not yet supported by propagated FIR-luminosity uncertainties, and the fit in Eq. (2) lacks error bars. These are fixable within the manuscript's scope, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The CRISTAL survey paper does exactly what a good survey overview should: it puts a large, homogeneous, public dataset on the table. The 39 galaxies with kiloparsec-scale [CII] and dust continuum, plus HST and JWST ancillary imaging, are a genuine step forward for studying normal star-forming galaxies at z~4-6. The data reduction is documented in detail, fluxes from three independent methods agree within 10% for most sources, and the cross-checks against ALPINE and visibility-plane measurements give me confidence in the measurements themselves. That part is solid.\n\nThe science highlights are also reasonable. CRISTAL-10's spatially resolved [CII] deficit is interesting, and CRISTAL-13's offset between young stellar clumps and [CII] peaks is a nice, concrete example of feedback at work. The paper is honest about the assumptions: the case studies state their adopted dust temperatures and show the sensitivity to ±10 K. That is good practice.\n\nWhere I part ways with the paper's framing is the global [CII]/FIR analysis in Section 9. The FIR luminosities come from a single ALMA Band 7 continuum point and SED fits, and the dust-temperature degeneracy is not propagated into the z~0 comparison shown in Figure 14. Equation (2) is a new empirical relation with no fit uncertainties. A factor of 2-3 shift in LFIR - well within plausible SED systematics - would move the CRISTAL points diagonally and could weaken the claimed 'previously unexplored ΣFIR range' and the 3-10× offset from local galaxies. The stress-test note about this is fair. It is not a fatal flaw, but it is a real gap in the quantitative argument, and it is easily fixable with error bars and a caveat.\n\nA second soft spot: several interpretive claims cite in-prep companion papers as support (Lee et al., Birkin et al., Herrera-Camus et al., etc.). That is normal for a survey paper, but it makes it hard to evaluate some conclusions independently without digging into the companion papers. The core data release stands on its own, though.\n\nWho is this for? Anyone working on high-redshift ISM, galaxy structure, or ALMA surveys. The sample and data products will be used for years. It deserves peer review, but the referee should ask for quantified uncertainties on Eq. (2) and an explicit statement about the SED assumptions in the global comparison. I would take the paper with those revisions.","headline":"A well-executed, data-rich survey overview that delivers a valuable public kiloparsec-resolved [CII] sample at z~4-6; the global [CII]/FIR results are plausible but carry an unquantified dust-temperature dependence that should be addressed before publication.","tokens_in":49065,"tokens_out":2129,"would_cite":true,"duration_ms":23740,"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 paper claims that kiloparsec-resolution ALMA imaging of 39 star-forming galaxies at z≈4–6 shows [C II] gas extending beyond starlight, an Arp 220-like deficit in one region, and a [C II]/FIR deficit trend shifted toward higher…","keywords":["high-redshift galaxies","interstellar medium","[C II] 158 μm emission","main-sequence star-forming galaxies","galaxy kinematics","galaxy mergers","dust continuum","ALMA survey"],"falsifier":"Measure the dust spectral energy distribution of CRISTAL-10's Region 1 with a longer-wavelength ALMA band (for instance Band 9) or with JWST/MIRI to obtain a direct dust temperature; if the temperature is about 60–70 K rather than 50 K, the inferred far-infrared luminosity rises enough that [C II]/FIR climbs out of the Arp 220 regime and the extreme-deficit case fails. For the global relation, re-fitting the CRISTAL SEDs with additional submillimeter photometry or different dust templates would settle whether the shift of the [C II]/FIR–LFIR trend relative to local galaxies is real.","tokens_in":47700,"feed_emoji":"🌌","tokens_out":13409,"duration_ms":122642,"temperature":0.7,"pith_summary":"The CRISTAL survey maps the 158 μm [C II] line and dust continuum at roughly kiloparsec resolution in 39 main-sequence star-forming galaxies that lived when the Universe was about 1 billion years old, and combines those ALMA data with HST and JWST imaging of the same systems. The paper's central claim is that the cold interstellar medium of these early galaxies is not a simple mirror of their stars: [C II] emission commonly extends well beyond the stellar light, and the resolved morphologies range from rotating disks to mergers, tidal tails, and clumpy complexes. It further claims that the global [C II]/FIR ratio declines with FIR luminosity as it does locally, but the whole trend sits at higher luminosities, consistent with larger molecular gas reservoirs. Two case studies anchor the argument: a kiloparsec-scale dusty region in CRISTAL-10 has a [C II]/FIR ratio as low as the extreme local system Arp 220, and the young stellar clumps of CRISTAL-13 are spatially offset from the [C II] emission, interpreted as feedback dispersing or evaporating the gas. If these claims hold, the survey establishes a kiloparsec-resolution benchmark for interpreting gas, dust, and star formation in the first billion years of cosmic history.","feed_headline":"Cold gas extends far beyond stars in 39 early galaxies","feed_subtitle":"Kiloparsec-scale ALMA maps reveal extreme [CII] deficits and star-clump offsets in the Universe's first billion years.","key_machinery":"The argument is carried by the [C II]-to-far-infrared luminosity ratio, L[CII]/LFIR, a standard proxy for the photoelectric heating efficiency of neutral interstellar gas; the paper's central comparison is the well-known ‘[C II] deficit’, the decline of this ratio with increasing FIR luminosity or FIR surface brightness. The observational machinery is the combination of ALMA Band 7 observations with synthesized beams of roughly 0.1 to 0.7 arcseconds, HST/WFC3 and JWST/NIRCam imaging, and SED fitting from companion papers that supplies stellar masses, star formation rates, FIR luminosities, and dust continuum sizes. For the case studies, the FIR luminosity is derived from a single Band 7 continuum measurement assuming a dust temperature of 50 K for CRISTAL-10 and 40 K for CRISTAL-13, and the paper shows how the [C II]/FIR ratios move under a ±10 K temperature change.","core_discovery":"On its own terms, the paper establishes that typical star-forming galaxies at z ≈ 4–6 contain a cold gas reservoir that is systematically more extended and more disturbed than the rest-frame UV and optical starlight suggests. In many CRISTAL systems the [C II] emission reaches two to three times the extent of the stellar light, and roughly half the sample is classified as disk-dominated while at least a third shows clear interaction or merger signatures. The paper also reports a global trend: the [C II]/FIR ratio declines with FIR luminosity as in local galaxies but shifted to higher luminosities, which it attributes to higher molecular gas content. At FIR surface brightnesses of about $10^{9}$ to $10^{11}$ Lsun/$kpc^{2}$, a range previously unexplored at high redshift, CRISTAL galaxies show elevated [C II]/FIR ratios relative to z ≈ 0 galaxies at the same surface brightness, with lower metallicity, extraplanar gas, and shock excitation offered as likely contributors. The two case studies are CRISTAL-10, where a ~3 kpc dusty region reaches L[CII]/LFIR = 2.5 × $10^{-4}$, comparable to Arp 220, and CRISTAL-13, where six young stellar clumps are anti-correlated with [C II] peaks, consistent with feedback clearing the surrounding gas.","pith_inferences":["A direct test of the gas-fraction explanation: CO(2–1) or [C I] observations of a subset of CRISTAL galaxies should recover molecular gas fractions roughly ten times higher than local main-sequence galaxies at matched specific star formation rate; if they do not, the shifted [C II]/FIR–LFIR relation needs another driver.","The metallicity explanation predicts a spatial correlation within galaxies: regions with the highest [C II]/FIR ratios should also have the lowest gas-phase metallicities when mapped with JWST/NIRSpec, an internal test the current data cannot yet make.","The CRISTAL-13 feedback interpretation implies an age ordering: the six western clumps should be younger than the roughly 10 Myr photoevaporation timescale, while older clumps should sit closer to [C II] peaks; resolved stellar-age maps can falsify this ordering.","Extending the same method to fainter FIR surface brightness or to z > 6 would test whether the elevated [C II]/FIR branch continues or turns over as CMB effects on diffuse gas grow, sharpening the boundary between local and early-Universe ISM physics."],"forward_implications":["Low-resolution [C II] measurements at z ≈ 4–6 misclassify a substantial fraction of galaxies: once seen at kiloparsec resolution, about half of the CRISTAL systems are disk-dominated and at least a third show interaction or merger signatures.","Source-integrated [C II]/FIR ratios can hide localized extreme deficits, since CRISTAL-10's global ratio is about ten times higher than its most obscured region; resolved measurements are therefore needed to diagnose the ISM conditions in high-redshift starbursts.","If the shifted [C II]/FIR–LFIR trend is caused by higher molecular gas content, direct CO or [C I] observations of CRISTAL galaxies should find gas fractions well above local values at comparable stellar mass and star formation rate.","At fixed FIR surface brightness, high-redshift galaxies with log Σ_FIR between about 11.5 and 13.5 (in Lsun/kpc^2) have [C II]/FIR ratios three to ten times higher than local galaxies, implying elevated photoelectric heating efficiency or extra diffuse and shocked [C II] emission.","Extended [C II] envelopes around many CRISTAL galaxies mean that [C II] sizes should not be used as proxies for stellar or disk sizes in early-Universe galaxies, because they overestimate the spatial extent of star formation."],"supporting_citations":[{"why":"Parent ALPINE survey defining the [C II]-selected galaxy population and redshift range from which CRISTAL targets were drawn.","marker":"Le Fèvre et al. 2020"},{"why":"Supplied ALPINE [C II] fluxes and continuum measurements whose signal-to-noise and redshifts seeded CRISTAL target selection.","marker":"Béthermin et al. 2020"},{"why":"Provided the ALPINE photometric stellar masses and star formation rates used for main-sequence and mass selection.","marker":"Faisst et al. 2020b"},{"why":"SED fits deliver stellar masses, SFRs, FIR luminosities, and dust continuum sizes used in the global [C II]/FIR and surface-brightness analysis.","marker":"Mitsuhashi et al. 2024"},{"why":"Measured [C II] sizes and extended emission in CRISTAL galaxies, supporting the beyond-starlight claim and companion detections.","marker":"Ikeda et al. 2025"},{"why":"ALMA Band 9 SED of HZ10 gives a reference dust temperature near 47 K adopted for the CRISTAL-10 and CRISTAL-13 case-study FIR conversions.","marker":"Villanueva et al. 2024"},{"why":"JWST/NIRCam SED modeling dates the young clumps in CRISTAL-13 and CRISTAL-11, anchoring the feedback-offset interpretation.","marker":"Lines et al. 2024"},{"why":"Theoretical model linking the [C II] deficit to gas surface density and molecular gas content, used to explain the high-redshift luminosity offset.","marker":"Narayanan & Krumholz 2017"},{"why":"Provides the local [C II]/FIR versus FIR surface brightness calibration and nearby comparison sample for Equation 1.","marker":"Lutz et al. 2016"},{"why":"High-redshift galaxies at higher FIR surface brightness anchor Equation 2 and bound the elevated-ratio comparison.","marker":"Spilker et al. 2016"}],"fun_headline_variants":["Cold gas outspans starlight in 39 early galaxies","Cold gas reaches 3x stellar extent in 39 early galaxies","Extreme CII deficit like Arp 220 in early galaxy","Stellar clumps offset from gas peaks in CRISTAL-13","Survey maps gas, dust, stars in 39 cosmic dawn galaxies"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the far-infrared luminosities and surface brightnesses entering the [C II]/FIR ratios are accurate: the global values come from SED fits, while the CRISTAL-10 and CRISTAL-13 case studies convert a single ALMA Band 7 continuum measurement into a far-infrared luminosity by assuming dust temperatures of 50 K and 40 K, so if the true dust temperatures or SED assumptions differ, the extreme-deficit claims and the position of the high-redshift trend move.","fun_headline_variants_meta":{"raw":{"variants":["Cold gas outspans starlight in 39 early galaxies","Cold gas reaches 3x stellar extent in 39 early galaxies","Extreme CII deficit like Arp 220 in early galaxy","Stellar clumps offset from gas peaks in CRISTAL-13","Survey maps gas, dust, stars in 39 cosmic dawn galaxies"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000925,"raw_usage":{"total_tokens":4100,"prompt_tokens":1218,"completion_tokens":2882,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":834,"completion_tokens_details":{"reasoning_tokens":2790}},"tokens_in":834,"tokens_out":2882,"duration_ms":25057,"temperature":1.0,"reasoning_tokens":2790,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:46:10.414763+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the dust spectral energy distribution of CRISTAL-10's Region 1 with a longer-wavelength ALMA band (for instance Band 9) or with JWST/MIRI to obtain a direct dust temperature; if the temperature is about 60–70 K rather than 50 K, the inferred far-infrared luminosity rises enough that [C II]/FIR climbs out of the Arp 220 regime and the extreme-deficit case fails. For the global relation, re-fitting the CRISTAL SEDs with additional submillimeter photometry or different dust templates would settle whether the shift of the [C II]/FIR–LFIR trend relative to local galaxies is real.","supporting_citations":[{"cited_title":"2024, A&A, 691, A133","cited_arxiv_id":null,"evidence_quote":"ALMA Band 9 SED of HZ10 gives a reference dust temperature near 47 K adopted for the CRISTAL-10 and CRISTAL-13 case-study FIR conversions."},{"cited_title":"2016, A&A, 591, A136","cited_arxiv_id":null,"evidence_quote":"Provides the local [C II]/FIR versus FIR surface brightness calibration and nearby comparison sample for Equation 1."}],"review_version":1}