{"id":"e1393ff4-77ca-45c5-bdf2-966ce07a54fe","arxiv_id":"2501.05288","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A z=4.3 compact galaxy group hosts two optically dark starbursts with gas depletion times around 100 million years, possibly forming a massive cluster.","lead":"Astronomers used ALMA and other telescopes to find a compact group of galaxies at redshift 4.3 that includes two dusty, optically invisible galaxies forming stars at a furious rate. The structure may be an early stage of a massive galaxy cluster, offering a rare view of how such giants assemble.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Individual SFRs and depletion times hinge on the untested assumption that both galaxies share the integrated FIR SED shape; per-source 3mm/870 colors could falsify this.","rationale":"The paper presents a valuable spectroscopic confirmation of two optically dark, dusty star-forming galaxies at z=4.3 with multiple CO lines, [CI], and water absorption, and the integrated SED is reasonably constrained. The central physical claim—extremely short depletion times suggesting the onset of quenching—depends on individual SFRs for the two galaxies, which are derived by scaling the integrated SED to each source's 870 micron flux under the assumption of a common SED shape (Sect. 4.4). This assumption is untested and is exactly the kind of step that can bias SFRs by factors of two if the galaxies have different dust temperatures; since the depletion time is gas/SFR, the headline values (99 Myr and <63 Myr) and the quenching interpretation would be weakened. The reader's weakest_assumption identifies this same step; I agree. The paper does not provide any per-source photometric color check, and the abstract/text discrepancy for CGG-z4.b's upper limit adds minor sloppiness. A check using the existing ALMA 3mm and 870um maps to measure per-source colors would directly test the assumption; if the colors match, the concern is resolved, and if not, the individual SFRs and depletion times need revision. Given that the qualitative statement of a massive, actively star-forming group at z=4.3 is robust to this concern, CONDITIONAL acceptance remains appropriate.","tokens_in":28178,"tokens_out":11072,"duration_ms":104194,"concrete_test":"Analyze the public ALMA data to measure the 870 micron and 3 mm continuum fluxes of CGG-z4.a and CGG-z4.b individually (e.g., uv-plane fits at the positions of the two CO detections). Compute the 870/3mm flux ratio for each source. Under the same-SED assumption these ratios should agree within uncertainties. If the ratios (or the dust temperatures from two-point modified blackbody fits with beta=1.9) differ by more than ~2 sigma, the scaling in Sect. 4.4 is invalid and individual SFRs and depletion times must be recomputed from separate SED fits. If the ratios agree, the concern is resolved and the depletion-time estimates are strengthened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that CGG-z4.a and CGG-z4.b have 'extremely short gas depletion times' and are 'suggesting the onset of quenching' rests on individual SFRs derived in Sect. 4.4 by assuming both galaxies share the same SED shape as the integrated group SED, then scaling to each source's ALMA 870 micron flux. At z=4.3, observed 870 micron samples rest-frame ~164 micron, near the Rayleigh-Jeans tail, so the 870-to-L_IR conversion is sensitive to dust temperature. If CGG-z4.b is colder than the integrated SED, its SFR is overestimated and its true depletion time (gas/SFR) is longer than reported; if CGG-z4.a is warmer, its depletion time would be shorter. A factor-of-two SFR change moves the inferred depletion times from ~60-100 Myr to values comparable to field scaling relations in Fig. 6, erasing the claimed offset that motivates the quenching scenario. The paper provides no check of this assumption (e.g., per-source 3mm-to-870um colors), and the abstract's '<63 Myr' vs text's '<69 Myr' for CGG-z4.b indicates the upper limit is not precisely quoted. The [CI]-based gas masses do not depend on the FIR SED, but the SFRs used in the depletion-time ratio do, so the short depletion times are not robust to this assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents ALMA 3 mm and 870 µm observations confirming two z≈4.3 optically/NIR-dark galaxies, CGG-z4.a and CGG-z4.b, together with 11 photometric-redshift members in an overdense COSMOS field. The authors detect CO(4-3) and CO(5-4) in both galaxies, [CI](1-0) in CGG-z4.a, and H2O(11,0-10,1) absorption in CGG-z4.b; they fit the integrated FIR-to-radio photometry with MICHI2, derive total and individual SFRs, estimate gas masses with dust, CO, and [CI] tracers, and compute short gas depletion times. Using five methods, they estimate a halo mass of log(M_halo/M⊙)=12.8±0.4 and argue that the structure is likely forming a massive galaxy cluster. The central conclusions are that CGG-z4 is one of the most star-forming groups known at z>4 and that its two brightest galaxies may be undergoing the onset of quenching.","tokens_in":28498,"tokens_out":9134,"duration_ms":88697,"significance":"If the physical interpretation holds, CGG-z4 is a valuable addition to the small sample of spectroscopically confirmed z>4 groups and protocluster cores. The multi-line ALMA detections, including the rare H2O absorption feature, are a genuine observational achievement, and the paper is transparent about many of its assumptions and uncertainties. However, the headline claims about extremely short depletion times and the onset of quenching rest on an untested shared-SED assumption and on an upper-limit [CI] measurement for CGG-z4.b. The paper would be a solid discovery paper even with those claims appropriately softened; as written, the abstract and conclusions overstate the robustness of the individual physical quantities.","major_comments":[{"comment":"The individual SFRs that enter the depletion times are derived by assuming CGG-z4.a and CGG-z4.b share the exact SED shape of the integrated group and then scaling by each source's ALMA 870 µm flux. At z=4.3, 870 µm samples rest-frame ~164 µm, where the conversion from monochromatic flux to L_IR is highly sensitive to dust temperature; a factor-of-two change in SFR would move the inferred depletion times from ~60-100 Myr to values consistent with the field scaling relations shown in Fig. 6. The paper reports no per-source 3mm-to-870µm color or any other test of the shared-SED assumption. I request that the authors either provide such a test (e.g., per-source 3mm/870µm flux ratios, if the data permit) or explicitly state that the individual SFRs and depletion times are model-dependent and soften the abstract accordingly.","section":"Sect. 4.4; Fig. 3"},{"comment":"For CGG-z4.b the [CI](1-0) line is a 2σ upper limit, so the gas mass and hence the quoted \"<63 Myr\" depletion time in the abstract are upper limits, not measurements; the body and Table D.1 give <69 Myr. This discrepancy is not merely cosmetic: the \"extremely short\" depletion-time claim for CGG-z4.b is an upper limit based on an upper-limit gas mass, and the SFR uncertainty is not propagated into that limit in a statistically well-defined way. The authors should quote the body value consistently, label the CGG-z4.b value as an upper limit in the abstract and conclusions, and check whether the upper limit remains below the field relations once the SFR uncertainty is included.","section":"Sect. 4.5; Table D.1; abstract"},{"comment":"The statement that the short depletion times \"suggest the onset of quenching\" goes beyond what the data show. As the authors state in Sect. 5.3, this conclusion assumes a constant SFE and no gas replenishment; a short depletion time measures current high gas consumption, not the future gas supply. In a dense protocluster core, ongoing accretion could sustain star formation. I recommend moving the \"onset of quenching\" language to a conditional scenario in the abstract and conclusions, or adding a concrete discussion of why gas replenishment is expected to be negligible for these two galaxies.","section":"Sect. 5.3; abstract"},{"comment":"The adopted halo mass log(M_halo/M⊙)=12.8±0.4 is an unweighted average of five methods whose estimates range from a lower limit of >12.2 to 13.5±0.3. The quoted uncertainty does not encompass this spread, and the average is sensitive to the choice of methods. Since this value is used in Sect. 5.5 and Conclusion 6 to predict M>10^14 M⊙ at z=0, the protocluster claim should be presented as conditional on a range (e.g., ~12.2-13.5 dex) rather than as a single average, or the averaging should be justified with a systematic-error analysis.","section":"Sect. 4.3; Conclusion 6"}],"minor_comments":[{"comment":"The text states that \"program 2021.1.00246.S contains 20 spectral windows (SPWs), and project 2021.1.00246.S contains 12 SPWs\"; one of these program identifiers must refer to 2022.1.00884.S.","section":"Sect. 2.2"},{"comment":"\"Given the optically dark nature of CGG-z4.a and CGG-z4.a\" should read \"CGG-z4.a and CGG-z4.b\"; the same typo appears in both sections.","section":"Sects. 3.3 and 4.2"},{"comment":"In the gas-mass paragraph, the second occurrence of \"log(M_gas,thick,a[M⊙])\" with value 10.77 should refer to CGG-z4.b, not CGG-z4.a.","section":"Sect. 4.5"},{"comment":"The sentence \"with ranges between 10.08 < log(Mgas[M⊙]) < 11.35\" appears to refer to stellar masses, not gas masses; please correct the variable.","section":"Sect. 5.4"},{"comment":"The phrase \"an on-sky projected line of sight distance of 0.829 pMpc\" is ambiguous; if 0.829 pMpc is the comoving line-of-sight separation implied by Δz, it should be stated separately from the projected kiloparsec separation.","section":"Sect. 3.1"},{"comment":"The adopted \"average stellar masses with uncertainties encompassing both results\" is not defined quantitatively; Table C.1 should give the adopted value and range explicitly, along with the two individual estimates (IRAC-scaled and dust-based).","section":"Sect. 4.2; Table C.1"},{"comment":"Several instances of \"CCG-z4\" (e.g., Sects. 5.2, 5.3, and the Conclusions) should be \"CGG-z4\" for consistency.","section":"Throughout"},{"comment":"For CGG-z4.b the PDR parameters are only lower limits, but the text says \"we can determine that n_H > 3.0×10^4 cm^-3\"; consider phrasing it as \"we can only place lower limits.\"","section":"Sect. 5.1"}],"recommendation":"major_revision","confidential_remarks":"The spectroscopic redshift confirmation and the multi-line ALMA detections are solid, and I see no reason to doubt the reality of the group. My main concern is that the individual SFR and depletion-time claims are presented with more certainty than the shared-SED assumption and the upper-limit [CI] detection allow. If the authors can provide per-source 3mm/870µm colors or clearly present the affected quantities as model-dependent upper limits, the paper would be a strong contribution. The halo-mass averaging should also be revisited before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What should you know: Brinch et al. report a compact group at z=4.3 with two optically dark starbursts, secure redshifts from multiple CO lines, and a wealth of ISM tracers. The object is real and the line detections are solid; this adds a rare data point to the z>4 group census.\n\nThe paper does several things well. The spectroscopic redshifts rest on CO(4-3) and CO(5-4) with good SNR, and the detection of H2O absorption on one source is a nice addition to a small sample. The gas mass estimates are cross-checked with dust, CO, and [CI] tracers, and the paper is transparent about the conversion factors. The comparison with SPT2349-56 and GN20 gives context.\n\nThe soft spots are real but proportionate. The individual SFRs, and therefore the depletion times, assume both galaxies share the integrated FIR SED shape. That is untested. At z=4.3, 870 um samples rest-frame ~164 um, so the SFRs are temperature-sensitive; a factor-two change in SFR would bring these depletion times onto the field scaling relations and weaken the 'onset of quenching' claim. The paper mentions the assumption but does not quantify its impact. I would like to see per-source 3mm/870 flux ratios or a sensitivity test. There is also a small internal inconsistency in the abstract (<63 Myr) versus the text (<69 Myr) for CGG-z4.b. The stellar masses are poorly constrained (IRAC single band or M*/Mdust). The overdensity significance relies on an unpublished spectroscopic catalog, which is hard to check.\n\nNone of this undermines the core discovery. The quenching interpretation is the least secure part, and the paper could be revised to frame it as a possibility rather than a suggestion, or better, to test the SED-sharing assumption.\n\nVerdict: this deserves peer review. I would send it out and ask for the robustness check on the SED sharing before acceptance. The object is worth publishing; the interpretation needs a tighter leash.","headline":"Solid new ALMA discovery of a z=4.3 compact group; the quenching interpretation rests on an untested SED-sharing assumption and needs a sensitivity check.","tokens_in":29063,"tokens_out":2735,"would_cite":true,"duration_ms":26166,"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":"Two optically invisible galaxies at redshift 4.3 form a compact group producing roughly 2,000 solar masses of stars per year, and their gas is running out fast.","keywords":["galaxies: groups: individual: CGG-z4","galaxies: high-redshift","galaxies: star formation","galaxies: ISM","optically dark galaxies","submillimetre galaxies","protoclusters","gas depletion timescale"],"falsifier":"Resolved observations of each galaxy at several far-infrared or submillimetre wavelengths, such as multiband millimetre interferometry, would measure the two dust SEDs independently and test the shared-SED scaling; a direct CO(1-0) detection would independently test the assumed CO(4-3)-to-CO(1-0) conversion.","tokens_in":27990,"feed_emoji":"🌌","tokens_out":10496,"duration_ms":93876,"temperature":0.7,"pith_summary":"This paper reports the confirmation of CGG-z4, a compact galaxy group at $z=4.3$ whose two most massive members are optically and near-infrared dark. From the combined far-infrared-to-radio emission of the group, the authors derive a total star formation rate of about $2000\\,M_\\odot$ yr$^{-1}$, among the highest known for a group at $z>4$. Using the [CI](1-0) line to estimate molecular gas, they find the two dark galaxies have gas depletion times of roughly 100 Myr and under 63 Myr, which they interpret as the onset of quenching. If the picture holds, this is a direct view of massive galaxies exhausting their fuel in an overdense environment and may trace the birth of a galaxy cluster.","feed_headline":"Hidden galaxy pair at z=4.3 makes ~2,000 stars a year","feed_subtitle":"Their gas will run out in about 100 million years, hinting these starbursts are quenching.","key_machinery":"The argument is carried by millimeter line spectroscopy and a single integrated spectral energy distribution. CO(4-3) and CO(5-4) detections fix the spectroscopic redshifts and, through their high ratio, place the interstellar medium near thermalization; [CI](1-0) provides a molecular-gas mass estimate that relies less on uncertain conversion factors. The integrated far-infrared-to-radio photometry is fit by a physical SED model to obtain the total star formation rate, and the individual galaxy properties are then obtained by scaling that same SED shape to each source's ALMA 870 $\\mu$m flux. The halo-mass estimate combines stellar-mass-to-halo-mass relations, overdensity with a galaxy bias, and a radial stellar-mass density profile, and the future-cluster interpretation is anchored by comparison with simulated halo growth tracks.","core_discovery":"The central claim is that CGG-z4 is a genuine $z\\simeq4.3$ structure containing two spectroscopically confirmed, optically dark, dust-enshrouded starbursts. The two galaxies, CGG-z4.a and CGG-z4.b, are detected in CO(4-3), CO(5-4), and one also in [CI](1-0), giving secure redshifts of 4.331 and 4.324. Fitting the integrated far-infrared-to-radio photometry yields a total star formation rate of $2111\\pm98\\,M_\\odot$ yr$^{-1}$, and scaling that spectral energy distribution to each ALMA 870 $\\mu$m flux assigns $1408$ and $703\\,M_\\odot$ yr$^{-1}$ to the two sources, placing them about six and four-and-a-half times above the star-forming main sequence. The CO(5-4)/CO(4-3) ratios are close to the thermalization limit, indicating dense, high-pressure gas, and the low [CI]/CO ratios indicate high star formation efficiency. The [CI]-derived gas masses imply depletion times of about 99 Myr and under 63 Myr, so the authors argue these starbursts are likely already transitioning toward quiescence, and with an estimated halo mass of $\\log(M_{\\rm halo}/M_\\odot)\\sim12.8$ the structure is probably in the process of forming a massive galaxy cluster with the two dark galaxies as its brightest members.","pith_inferences":["A natural test of the quenching interpretation is to measure each galaxy's dust temperature separately; if the two SEDs differ from the integrated shape, the individual SFRs and gas masses would shift and the short depletion times might not survive.","If the short depletion times are confirmed, they would support a picture in which massive galaxies quench by rapid gas exhaustion in a compact starburst, without requiring an external feedback mechanism to remove the gas.","The same selection strategy of radio and submillimeter priors followed by millimeter line scans could uncover more optically dark starbursts at the cores of $z>4$ protoclusters, potentially revising current estimates of how the most massive galaxies assemble.","Deeper near-infrared photometry of these two galaxies would test the stellar-mass estimates and reveal morphologies that could connect their starburst phase to a merger-driven quenching path."],"forward_implications":["CGG-z4 becomes one of the most star-forming galaxy groups known at $z>4$, with a total SFR near $2000\\,M_\\odot$ yr$^{-1}$ dominated by two galaxies that are invisible in optical and near-infrared surveys.","The short depletion times imply that, without gas replenishment, the two galaxies will exhaust their fuel and could become quiescent by $z\\sim4$ with final stellar masses near $10^{11}\\,M_\\odot$.","The estimated halo mass of $\\log(M_{\\rm halo}/M_\\odot)\\sim12.8$ places the structure on a trajectory to become a massive cluster by $z=0$.","The non-detection of CO in stacked spectra of the optical and near-infrared detected members shows that the dark pair holds most of the group's gas and star formation, so surveys using rest-frame optical selection would miss the group's most active members."],"supporting_citations":[{"why":"Supplies the Super-deblended far-infrared photometry and radio prior that first selected CGG-z4.a and feeds the SED fit.","marker":"Jin et al. (2018)"},{"why":"The SED fitting tool used to model the integrated far-infrared-to-radio photometry and derive the total star formation rate.","marker":"Liu et al. (2021)"},{"why":"Provides the [CI]-to-gas-mass conversion factor that yields the gas masses and depletion times.","marker":"Dunne et al. (2022)"},{"why":"Gives the thermalization limit used to interpret the high CO(5-4)/CO(4-3) ratios and set the CO ladder conversion.","marker":"Narayanan & Krumholz (2014)"},{"why":"Supplies the multi-method halo-mass estimation pipeline adopted for CGG-z4.","marker":"Sillassen et al. (2024)"},{"why":"Provides the cluster-progenitor growth tracks used to argue CGG-z4 will become a massive cluster.","marker":"Chiang et al. (2013)"},{"why":"Gives the simulated brightest-cluster-galaxy progenitor masses used to compare with CGG-z4.a and b.","marker":"Montenegro-Taborda et al. (2023)"}],"fun_headline_variants":["Hidden galaxy pair makes 2000 stars a year at z=4.3","Two dark starbursts at z=4.3 run out of gas fast","Cosmic feast: z=4.3 group births massive cluster","Efficient star-forming duo at z=4.3 may quench soon"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The individual star formation rates and gas masses of the two galaxies are computed by assuming both have exactly the same far-infrared spectrum as the combined group and differ only in their 870-micron brightness; if their dust temperatures differ, the numbers shift.","fun_headline_variants_meta":{"raw":{"variants":["Hidden galaxy pair makes 2000 stars a year at z=4.3","Two dark starbursts at z=4.3 run out of gas fast","Cosmic feast: z=4.3 group births massive cluster","Efficient star-forming duo at z=4.3 may quench soon"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000272,"raw_usage":{"total_tokens":1782,"prompt_tokens":1242,"completion_tokens":540,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":858,"completion_tokens_details":{"reasoning_tokens":456}},"tokens_in":858,"tokens_out":540,"duration_ms":5352,"temperature":1.0,"reasoning_tokens":456,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:14:28.515593+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Resolved observations of each galaxy at several far-infrared or submillimetre wavelengths, such as multiband millimetre interferometry, would measure the two dust SEDs independently and test the shared-SED scaling; a direct CO(1-0) detection would independently test the assumed CO(4-3)-to-CO(1-0) conversion.","supporting_citations":[{"cited_title":"2018, ApJ, 864, 56","cited_arxiv_id":null,"evidence_quote":"Supplies the Super-deblended far-infrared photometry and radio prior that first selected CGG-z4.a and feeds the SED fit."}],"review_version":1}