{"id":"d1458891-e1d9-4ca8-9bc6-141bde34fb53","arxiv_id":"2607.28875","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"IR SED broadening in high-z galaxies is driven primarily by luminosity density, requiring ~1.6–1.7 (up to ~2) corrections to modified-blackbody total infrared luminosities.","lead":"This paper uses simulations of high-redshift dusty galaxies to estimate how much mid-infrared light is missed when galaxy luminosities are derived from single ALMA submillimeter measurements. It proposes correction factors of roughly 1.6–1.7, rising to about 2 for the most luminous z>4 sources, which matter for star-formation-rate estimates.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The z>4 'up to 2' correction rests on Haro 11's unconstrained mid-IR excess: the 18-galaxy composite and the model do not constrain rest-frame 8–50 µm, exactly the range that sets the correction.","rationale":"The reader's weakest assumption identifies the Haro 11 analog as the load-bearing premise for the z>4 correction, and my reading agrees. I sharpen the concern by noting that the 18-galaxy composite and the model never actually constrain the 8–50 µm wavelength range that drives the difference between correction factors; the model omits stochastic heating and PAHs, and the composite's shortest rest wavelengths are near 50 µm. Therefore, the 'up to 2' factor is an extrapolation from a single local galaxy's mid-IR excess, not a measured property of the high-z population. This does not invalidate the paper's main contribution—the 1.6–1.7 factor for typical z>2 galaxies is more directly supported by the z~2.5 lensed-galaxy check and by template fits—but it means the strong z>4 claim needs empirical mid-IR validation before being applied. The reader already returned CONDITIONAL, and this concern reinforces that conditionality rather than changing it, so I recommend UNCHANGED. I would not reject the paper: it is transparent about model limitations, provides a concrete recipe, and the mid-IR test is currently feasible with JWST/MIRI. The honest non-finding option is not appropriate because a specific, testable gap in the evidence exists.","tokens_in":25339,"tokens_out":10459,"duration_ms":114447,"concrete_test":"Obtain JWST/MIRI photometry or low-resolution spectroscopy at rest-frame 8–25 µm for 10–20 of the z>4, L_IR>10^12 Lsun galaxies in Table 2 (many are bright enough for MIRI). Combine with ALMA submm and available Herschel/SCUBA data to build observed SEDs; fit a single modified blackbody to rest-frame 70–1000 µm (or to the long-wavelength data actually available) and compute the 8–1000 µm correction ratio. Compare the distribution with the Haro 11 value (~2) and the Rieke log(L)=11.25 value (~1.63). If the observed median correction is <1.8, the z>4 'up to 2' claim should be revised downward; if near 2, the extrapolation is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central recipe is conditional on template choice, and this is most acute for z>4, L_IR>10^12 Lsun sources, where Table 4 and Section 6.2 recommend correction factors up to ~2 (Haro 11 row). The supporting evidence is indirect. First, the De Rossi et al. (2018) composite of 18 z=5-7 galaxies is best fit by Haro 11 (reduced chi2=1.57), but that composite is built from Herschel/SPT photometry and a few ALMA points, with the shortest rest wavelengths around ~50 µm (Fig. 5); it does not constrain the 8–50 µm region that carries the extra ~0.3 dex of luminosity in the Haro 11 template. Second, the theoretical model reproduces the far-infrared broadening but explicitly omits stochastically heated grains and PAHs, and lies far below the data at λ<15 µm (Section 3.3), so the mid-IR excess that produces the larger correction is not predicted by the claimed physical driver. Third, Table 2 is a selected, non-statistical list; Section 5 states the incidence is 'high enough' but gives no quantified fraction or error bar. Thus, the headline z>4 correction factor is a single-galaxy extrapolation (Haro 11) rather than a measured property of z>4 sources. If those sources lack Haro 11's mid-IR excess, the correction reverts to ~1.6–1.8, and the 0.3-dex claim overestimates SFR corrections by ~0.1–0.2 dex in that regime. The paper is otherwise honest about degeneracies and provides a useful z~2.5 check, so the concern is specifically about the extrapolated high-z tail of the prescription.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper addresses how to convert ALMA submm photometry of high-redshift galaxies into total 8–1000 µm infrared luminosities. It argues that the shape of the infrared SED is controlled primarily by the luminosity density of the star-forming ISM, and that luminous z>4 galaxies frequently have broad, Haro 11-like SEDs rather than narrower local templates. On this basis it recommends multiplying modified-blackbody luminosities by ~1.6–1.7 in most circumstances, rising to ~1.75–2 for z>4 galaxies with L_IR > 10^12 L_sun. The paper combines a radiative-transfer model tuned to early-Universe conditions, local high-luminosity-density analogs, a z~2.5 lensed-galaxy template check, and template-integral corrections tabulated in Table 4.","tokens_in":25733,"tokens_out":4753,"duration_ms":56834,"significance":"If the central claim holds, the paper provides a useful and quantitative prescription for an important practical problem: deriving total infrared luminosities and obscured star formation rates from sparse ALMA data. The physical argument that luminosity density drives SED broadening is plausible and supported by local examples and by the model. The z~2.5 check with lensed galaxies is a genuinely valuable addition, and the paper is unusually honest about degeneracies and the template-dependent nature of the corrections. The weakest link is the high-redshift, high-luminosity tail of the prescription, where the evidence for Haro 11-like SEDs is indirect and where the recommended correction is largest. The paper is therefore worth publishing after the high-z claim is re-scoped or strengthened with quantitative support.","major_comments":[{"comment":"The headline z>4 correction 'up to ~2' rests on the assumption that luminous z>4 galaxies have Haro 11-like SEDs. The supporting composite of 18 galaxies (De Rossi et al. 2018) and Fig. 5 do not constrain rest-frame 8–50 µm: the ALMA points extend only to ~50 µm, and the Herschel/SPT points are longward. This is exactly the wavelength range that contributes the extra ~0.3 dex of luminosity in the Haro 11 template. Moreover, the theoretical model in §3.3 explicitly omits stochastically heated grains and PAHs and falls far below the data at λ≤15 µm, so the model cannot predict the mid-IR excess that drives the larger correction. Please either present the z>4 correction as an explicitly conditional upper end with a realistic uncertainty, or provide direct evidence (e.g., MIRI or stacking) that the 8–50 µm excess is present in the z>4 population.","section":"§6.2, Table 4 (Haro 11 row) and Fig. 5"},{"comment":"The statement that the incidence of Haro 11-like behavior 'is high enough that it should be taken into account' is not quantified. Table 2 is a partial, non-statistical list of high-luminosity-density galaxies; there is no denominator, selection function, or uncertainty. Since the recommended correction factor depends directly on how common this SED shape is among ALMA-detected galaxies, the paper should either quote a measured incidence with errors from a well-defined sample, or explicitly state that the z>4 correction applies only to the subset known to have very high luminosity density.","section":"§5, Table 2"},{"comment":"The z~2.5 reference correction (1.74) is not measured directly: it is computed from a constructed hybrid template that joins the log(L)=10.85 Rieke template shortward of 50 µm to the log(L)=11.25 template, with the join point and normalization chosen to minimize χ² in the 6–11 µm range. This is a reasonable empirical procedure, but the resulting correction inherits the assumptions of that interpolation. Please report the sensitivity of the correction to the join wavelength (e.g., 40/60 µm) and to the choice of the low-luminosity template, and propagate this into the uncertainty quoted for the 'typical' 1.6–1.7 factor.","section":"§6.2, Fig. 7 and Table 4"},{"comment":"The central claim that SED behavior is dominated by a single physical parameter, luminosity density, is demonstrated mainly by varying SFR at fixed R=1.5 kpc. The top-right panel shows that varying R alone also changes the SED substantially at λ≲50 µm. Since R and SFR both enter luminosity density, a more direct test would be to show that arbitrary combinations of R, SFR, and Mgas with the same luminosity density produce approximately the same SED. Without such a one-parameter collapse test, the generality of the correction factors as functions of luminosity density alone is not fully established.","section":"§3.4, Fig. 6"}],"minor_comments":[{"comment":"Typos: 'difficult', 'sufficient', 'similiar', 'correponds', 'an Haro 11 model' should be 'a Haro 11 model'.","section":"Throughout"},{"comment":"The caption refers to 'A1385a and A2128a', while the text and Table 3 discuss A1835a and A2218a. Correct the figure caption.","section":"Fig. 7 caption"},{"comment":"The row labeled 'z = 6' cites Schreiber et al. (2018) as the template, but the text says the correction is for a modified template with T=47 K and β=1.6. Clarify whether this is the unmodified Schreiber template, a modified template, or a modified blackbody with those parameters.","section":"Table 4"},{"comment":"For clarity, add a dash or '—' in the Haro 11 row under z=2–4, as was done for the Kirkpatrick row under z=5–7, so that the table does not appear to have a missing entry.","section":"Table 1"},{"comment":"The choice β=0.57 for λ>150 µm is stated without uncertainty or discussion of how it affects the derived corrections. Since the correction factors involve integrating the SED, a sentence quantifying the sensitivity of Table 4 to β would be useful.","section":"§3.1.5, Eq. (7)"},{"comment":"Several entries give only a single value for r_eff with no error bar (e.g., ALMACAL-1, ALMACAL-2). State whether these are upper limits, characteristic values, or formal measurements.","section":"Table 2"}],"recommendation":"major_revision","confidential_remarks":"The paper's most impactful claim — the up-to-0.3-dex correction for z>4, L>10^12 L_sun galaxies — is also the least well supported part of the manuscript. The authors should be asked either to provide a quantified incidence of Haro 11-like SEDs in a well-defined sample, or to soften the claim to a conditional upper limit. The rest of the manuscript is solid and would be a useful contribution to the literature."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The useful thing here is the practical product: correction factors to convert modified-blackbody ALMA luminosities into total IR luminosities — about 1.6–1.7 for typical z > 2 galaxies, up to roughly 2 for the most luminous z > 4 sources. The qualitative driver, luminosity density, comes from the authors' own earlier work, and they say so plainly; the quantitative table is new and is the part people will actually use.\n\nThe best section is the z ~ 2.5 check using two lensed galaxies with mid-IR spectra. That gives the correction a real empirical anchor, and it supports the log(L) = 11.25 template in a regime where you have actual PAH and 24 µm data. The model side is also admirably transparent: it omits stochastically heated grains and PAHs, it falls far below the data short of 15 µm, and the authors openly discuss degeneracies. That is honest work.\n\nThe soft spot is exactly where the stress-test puts it: the z > 4, L > 10^12 Lsun correction. The Haro 11 analog comes from an 18-galaxy composite whose shortest rest wavelengths are around 50 µm, so it does not constrain the 8–50 µm region that carries the extra ~0.3 dex. The theoretical model does not predict that mid-IR excess, and Table 2 is a selected, non-statistical list; the text says the incidence is 'high enough' without quantifying it. If typical luminous z > 4 sources do not have Haro 11's mid-IR excess, the correction drops back to ~1.6–1.8. The paper's own ambiguity — 'up to ~2' — is appropriate, but the claim could be framed more clearly as an extrapolation.\n\nI am less bothered than the reader by the circularity concern: the corrections are template integrals, and the template choice is the model. That is a real limitation, but it is not hidden. The bigger practical issue is that the correction factors have no error bars, and neither code nor data are shipped, with one dust opacity a private communication. Those are moderate reproducibility issues, not fatal flaws.\n\nThe paper is honest, internally consistent, and useful. It deserves a serious referee. My recommendation: send it to review, with the requirement that the authors quantify how the correction factors vary under plausible template choices, and either soften the z > 4 claim or provide a measured incidence for the Haro 11 class.","headline":"A useful, honest calibration paper whose typical-z correction factors are solid but whose z>4 'up to 2' factor rests on a single-galaxy template extrapolation that the data do not actually constrain.","tokens_in":26309,"tokens_out":1755,"would_cite":true,"duration_ms":23498,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A single parameter—luminosity density—governs the infrared spectra of high-redshift galaxies, requiring a 1.6–2× correction to ALMA-derived luminosities.","keywords":["infrared luminosity","star formation rate","submillimeter galaxies","high redshift","spectral energy distribution","modified blackbody","luminosity density","dust emission"],"falsifier":"Measure rest-frame 20–50 µm fluxes for a statistically complete sample of z>4, >1e12 Lsun galaxies (e.g., with JWST/MIRI or ALMA Band 8/9). If the average SED matches narrower low-luminosity-density templates rather than the broad Haro 11-like shape, the required correction reverts to ~1.6 and the paper's headline z>4 result (~2) is an overestimate. Alternatively, comparing ALMA 1 mm continuum with total IR luminosity from full SED decomposition would settle the correction directly.","tokens_in":25204,"feed_emoji":"🌌","tokens_out":5930,"duration_ms":60982,"temperature":0.7,"pith_summary":"This paper aims to fix a systematic bias in how astronomers measure the total infrared luminosity, and hence the star formation rate, of dusty galaxies at redshifts above two. Because ALMA observes only the long-wavelength side of a galaxy's infrared spectrum, luminosities are usually obtained by fitting a modified blackbody, which ignores the warm-dust mid-infrared emission. The authors argue, using models tuned to high-redshift conditions and a set of local calibrator galaxies, that the relevant shape of the infrared spectrum is controlled almost entirely by one number: the luminosity density, or how much infrared light is packed into a cubic parsec of star-forming gas. If that is right, modified-blackbody luminosities for typical z>2 galaxies must be multiplied by roughly 1.6–1.7, and by up to ~2 for the most luminous z>4 sources, meaning past ALMA-based star formation rates are low by 0.2–0.3 dex in those regimes.","feed_headline":"ALMA galaxy SFRs underestimated by 0.2–0.3 dex","feed_subtitle":"Warm mid-infrared dust is missed by blackbody fits; correcting for it raises star formation rates at z>2.","key_machinery":"Luminosity density (Lsun pc^-3) is the organizing variable: the paper shows that the peak temperature and the width of the infrared SED scale with this single quantity. The quantitative tool is a correction factor, defined as the ratio of the integral of a full template SED from 8 to 1000 µm to the integral of a modified blackbody fitted to it over 70–1000 µm. The Haro 11 SED—a local galaxy with extreme luminosity density (~1.5e4 Lsun pc^-3) and a broad, warm infrared spectrum—serves as the reference shape for the most luminous high-redshift sources. A set of radiative-transfer models for early-Universe galaxies (fixed radius 1.5 kpc, varying SFR from 5 to 30 Msun/yr, optically thin dust) re","core_discovery":"The paper establishes that variations in the far-infrared spectral energy distributions of high-redshift galaxies—both the shift of the peak to warmer temperatures and the broadening of the SED—are dominated by a single physical parameter, the luminosity density of the star-forming regions. Using radiative transfer models of early-Universe galaxies with a fixed ISM radius of 1.5 kpc and varying star formation rate, the authors show that higher luminosity density produces warmer and broader SEDs, while metallicity, gas fraction, and extinction have at most secondary effects. Consequently, the missing mid-infrared luminosity not captured by modified blackbody fits can be estimated in a general","pith_inferences":["If luminosity density is the controlling parameter, JWST/ALMA high-resolution imaging that resolves star-forming clumps should predict which galaxies need the largest corrections—compact clumps imply Haro 11-like SEDs; this is testable with existing data.","The paper models only silicate+carbon dust continuum and omits PAH emission and stochastically heated very small grains; real galaxies with strong PAH features may have even more mid-IR emission, possibly requiring larger corrections than 2 in specific cases, or smaller if PAHs are weak.","The Eddington-limit argument implies an upper bound on SED broadening: the most compact, luminous starbursts should sit near the limit, so the correction factor should saturate around 2 rather than growing without bound—an observable prediction.","Since the models assume optically thin dust and a fixed ISM radius of 1.5 kpc, the correction factors may not transfer directly to lower-mass JWST-discovered galaxies with smaller effective radii; if such galaxies are detected by ALMA, their corrections could differ."],"forward_implications":["Typical dusty galaxies at z>2: multiply modified-blackbody IR luminosities by ~1.6–1.7; published SFRs are low by ~0.2 dex.","Most luminous z>4 sources (L>1e12 Lsun): correction up to ~2 (0.3 dex), so their SFRs are underestimated the most.","Because the driver is a single parameter, the same correction framework applies without needing full SED coverage—one can use luminosity density (or its proxy, SFR surface density) to pick the appropriate factor.","The general increase in dust temperature with redshift found in earlier studies is explained as a luminosity-density effect, not a metallicity effect; metallicity plays little role.","Cosmic star formation rate density and the bright end of the galaxy luminosity function at z>2 shift upward once corrections are applied."],"fun_headline_variants":["One parameter determines high-z IR SED shape and correction","Submm-only star formation rates miss warm dust: fix by 0.2 dex","Luminosity density alone predicts missing mid-IR emission","Warm dust raises high-z SFRs by factor 1.6–2","Single physical driver sets IR SED and SFR correction"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The largest corrections (up to ~2) assume that the most luminous z>4 ALMA galaxies have SEDs shaped like Haro 11, the local extreme-density reference; the paper states this incidence is 'high enough' to matter but does not quantify it, and the supporting composite rests on 18 galaxies with sparse wavelength coverage.","fun_headline_variants_meta":{"raw":{"variants":["One parameter determines high-z IR SED shape and correction","Submm-only star formation rates miss warm dust: fix by 0.2 dex","Luminosity density alone predicts missing mid-IR emission","Warm dust raises high-z SFRs by factor 1.6–2","Single physical driver sets IR SED and SFR correction"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00027,"raw_usage":{"total_tokens":1507,"prompt_tokens":835,"completion_tokens":672,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":579,"completion_tokens_details":{"reasoning_tokens":580}},"tokens_in":579,"tokens_out":672,"duration_ms":7942,"temperature":1.0,"reasoning_tokens":580,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T01:24:38.433913+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure rest-frame 20–50 µm fluxes for a statistically complete sample of z>4, >1e12 Lsun galaxies (e.g., with JWST/MIRI or ALMA Band 8/9). If the average SED matches narrower low-luminosity-density templates rather than the broad Haro 11-like shape, the required correction reverts to ~1.6 and the paper's headline z>4 result (~2) is an overestimate. Alternatively, comparing ALMA 1 mm continuum with total IR luminosity from full SED decomposition would settle the correction directly.","supporting_citations":[],"review_version":1}