{"id":"89597b15-6516-411a-84f8-933e2e8fdcdd","arxiv_id":"1908.08043","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Spitzer follow-up of Herschel ultrared galaxies finds a rare population of unlensed, extremely dusty starbursts at median redshift 3.3 that cannot supply most progenitors of massive quiescent galaxies.","lead":"Astronomers used Spitzer infrared images to identify 300 extremely red, dusty galaxies picked by Herschel and measured the properties of 41 that are not gravitationally lensed. The galaxies are more massive and more intensely star-forming than typical dusty galaxies at similar epochs, but too rare to be the main ancestors of massive dead galaxies.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The assumed 100 Myr starburst duty cycle directly scales the space density and SMD that drive the central 'cannot account for majority' conclusion; an unquantified shorter duty cycle would weaken the claim.","rationale":"I agree with the reader's weakest_assumption: the duty cycle is the least secure input. My reading of Sec 5.5 confirms that the space density and SMD are quoted after 'completeness and duty-cycle corrections' with no sensitivity analysis. Because the duty cycle enters multiplicatively, an order-of-magnitude error in it translates directly into an order-of-magnitude error in the corrected densities. The central claim is not vacuous: the catalog, lensed/unlensed classification, and SED fittings are solid, and the space density comparison from prior work already points in the same direction. But the quantitative strength of the claim ('cannot be reconciled by even increasing the stellar mass limit') rests on a factor-of-100 gap that could be reduced to a factor-of-10 if the duty cycle is 10 Myr rather than 100 Myr. The paper's own note in Sec 5.1 that selection effects are not corrected for the raw redshift distribution is a further caution that the SMD sample may not be representative, but I treat this as secondary because the duty cycle has a larger dynamic range. Verdict remains CONDITIONAL: the concern is specific and testable, not fatal.","tokens_in":38630,"tokens_out":11527,"duration_ms":110151,"concrete_test":"Recompute the Sec 5.5 space density, SFRD, and SMD as a function of the starburst duty cycle over 10-500 Myr, holding all other corrections fixed; determine at which duty cycle the corrected ultrared densities reach parity with the ZFOURGE quiescent galaxy values. If parity is not reached until below ~1 Myr, the qualitative conclusion is robust; if it is reached at or above 10 Myr, the 'cannot account for the majority' claim requires a duty-cycle qualification.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 5.5 bases the central conclusion on the H-ATLAS ultrared space density of ~6e-7 Mpc^-3 (4<z<6) from Ivison et al. (2016), which already includes a completeness and duty-cycle correction assuming a ~100 Myr starburst duty cycle. The duty cycle is not measured or justified; it enters linearly, so a 10 Myr duty cycle would raise both the corrected space density and the SMD in Figure 9 by an order of magnitude. That brings the space density to ~6e-6 Mpc^-3, within a factor of ~3 of the ZFOURGE quiescent galaxy density at 3<z<4 (~2e-5 Mpc^-3) that the paper claims cannot be matched, and the SMD gap shrinks from ~100x to ~10x. Since the conclusion 'cannot account for the majority of progenitors' depends on these corrected densities being far below the quiescent population, the unquantified duty cycle is the most load-bearing assumption in the chain. The paper also applies the Ivison et al. completeness correction to the SMD derived from only 41 unlensed sources from a high-resolution subsample whose own selection function is not characterized (the authors state in Sec 5.1 they do not attempt to correct for selection effects due to complicated selection functions), but this second effect is likely smaller than the duty-cycle uncertainty.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents Spitzer/IRAC 3.6 and 4.5 micron observations of 300 Herschel-selected ultrared (500 micron-riser) dusty star-forming galaxies, together with high-resolution ALMA/NOEMA/SMA imaging for 63 of them. The authors use the high-resolution data to identify IRAC counterparts, classify sources as lensed, unlensed, or multiple-component, and de-blend the SPIRE/SCUBA-2 photometry with XID+. They focus on 41 unlensed systems (63 individual components) and run magphys+photo-z SED fitting to derive photometric redshifts, stellar masses, SFRs, dust luminosities, dust temperatures, and extinctions, comparing them with ALESS DSFGs. They then estimate the space density, SFR density, and stellar mass density of the ultrared population and conclude that the Herschel ultrared sample cannot account for the majority of the star-forming progenitors of massive quiescent galaxies at z ~ 3.","tokens_in":38866,"tokens_out":5489,"duration_ms":50969,"significance":"The catalog is a valuable resource: it provides the first large Spitzer/IRAC sample of the rare, luminous 500 micron-riser population, with a clear lensed/unlensed classification from high-resolution interferometry and a consistent SED-fitting framework (same magphys code as the ALESS comparison) that avoids many code-to-code systematics. The paper is also transparent: it explicitly notes the prior-dominated dust temperature, the limited rest-frame UV coverage, and the lack of selection corrections in Section 5.1. The central conclusion about progenitor accounting is, however, contingent on external completeness and duty-cycle corrections that are not quantified in this work.","major_comments":[{"comment":"The central conclusion that the ultrared sample 'cannot account for the majority of the star-forming progenitors' rests on the comparison of the H-ATLAS space density of ~6e-7 Mpc^-3 (4<z<6, from Ivison et al. 2016) and the SMD shown in Figure 9 with the ZFOURGE quiescent galaxy density of ~2e-5 Mpc^-3. Both the space density and the SMD are corrected assuming a starburst duty cycle of ~100 Myr. This duty cycle is not measured or justified in the present paper, and it enters the correction linearly: a duty cycle of 10 Myr would increase the corrected space density and SMD by an order of magnitude, reducing the SMD gap from ~100x to ~10x and the space-density gap from >30x to ~3x. The authors should provide a sensitivity analysis over plausible duty cycles (and completeness corrections) and either soften the conclusion or justify the adopted value.","section":"Section 5.5"},{"comment":"The SMD estimate in Figure 9 is built from only 41 unlensed H-ATLAS sources (63 individual components) drawn from the 63 sources with high-resolution data, which in turn are a subset of the 300 Spitzer targets. As stated in Section 5.1, the authors do not attempt to correct for selection effects due to the complicated selection functions. Applying the Ivison et al. (2016) whole-sample completeness and duty-cycle correction to this small, selection-unquantified subsample may introduce a bias that is not addressed. At minimum, the number of sources per redshift bin and the selection function of the high-resolution subsample should be discussed, so the reader can judge the SMD normalization.","section":"Section 5.5 and Section 5.1"},{"comment":"The validation of magphys+photo-z against spectroscopic redshifts is based on only 5 sources. The mean and median relative offsets differ by ~0.1 between the FIR and magphys methods, but with n=5 the comparison has very low statistical power. Given that the magphys-derived redshifts are systematically lower than the FIR photo-z's (median 3.3 vs 3.7) and the physical property medians (M*, SFR, SMD) all depend on redshift, the paper should either expand the spec-z validation (e.g., by including more published redshifts from Table 3) or explicitly propagate the photo-z systematic uncertainty into the derived medians and SMD.","section":"Section 4.1 and Figure 3"}],"minor_comments":[{"comment":"The acronym is given as 'HIREOs' in the Figure 9 caption and in Section 6, while Section 5.5 uses 'HIEROs'; please standardize the spelling.","section":"Section 5.5 and Figure 9 caption"},{"comment":"The text reports a median dust temperature of 38 +/- 2 K but immediately acknowledges that Tdust is not constrained by the data and is prior-dominated. Consider removing or clearly flagging this value in the summary of results so it is not read as a measurement.","section":"Section 4.3"},{"comment":"The blue line showing the expected spectroscopic redshift inferred from the median FIR offset is difficult to interpret; adding the 5 spec-z sources with their magphys photo-z points would make the validation and the claimed systematic offset more transparent.","section":"Figure 3 bottom panel"}],"recommendation":"major_revision","confidential_remarks":"The main value of the paper is the Spitzer catalog and the lensed/unlensed classification; the density-based conclusion rests on an external completeness/duty-cycle correction from Ivison et al. (2016) that the present data do not independently constrain. If the authors add a sensitivity analysis and reframe the conclusion as provisional, the paper would be acceptable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague, read this if you want the state of play on 500-um risers at z>4. The paper delivers something genuinely new: Spitzer/IRAC photometry and counterpart identification for 300 Herschel ultrared DSFGs, plus a high-resolution (ALMA/NOEMA/SMA) classification for 63 of them. The lensed fraction (~35%) and multiplicity (~27%) are solid measurements, and the unlensed subset gets the first consistent magphys stellar-mass and stellar-mass-density estimates for this population. The catalog itself is the main product and will be the reference list for follow-up.\n\nThe SED analysis is careful in places. They use the same magphys version as the ALESS comparison sample, which sidesteps code-systematics. They disclose what is weak: Tdust is prior-dominated, rest-frame UV is missing, and they make no selection correction for the 41-source unlensed subsample. The photo-z validation, however, is thin: five spectroscopic redshifts. That would be fine for a redshift check of a template set, but here the stacked physical properties depend on magphys photo-z's with 1-sigma errors of about +1.5/-1.0. The medians could shift.\n\nThe real soft spot is the duty cycle. The space density and stellar mass density in Section 5.5 inherit the ~100 Myr starburst duty cycle assumed in Ivison et al. (2016). It is not justified here, and it enters linearly. If the duty cycle is 10 Myr, the corrected density goes up by an order of magnitude and the gap to ZFOURGE quiescent galaxies shrinks from ~100x to ~10x. The 'cannot account for the majority' claim survives in that regime, but the margin is much thinner and the '~2 orders of magnitude lower' phrasing would be misleading. A sensitivity plot over duty cycle would have cost little and would have made the conclusion honest.\n\nOne more thing: the central conclusion is not new. Ivison et al. (2016) and Duivenvoorden et al. (2018) already reached it from space densities. The new SMD estimates don't change the qualitative picture; they restate it with mass weighting. That's okay, but the abstract should not imply the conclusion is newly established here.\n\nWho is this for? Observers working on high-redshift dusty galaxies and anyone building samples for JWST or ALMA follow-up. The catalog is the value. The physical property medians are more fragile than the abstract suggests. I would send it to peer review: the data and catalog merit refereeing, and the duty-cycle sensitivity can be addressed by the authors. I would not desk-reject. Suggest a major revision asking for a duty-cycle sensitivity test and a quantitative statement of the spec-z sample size.","headline":"Useful Spitzer catalog and first consistent mass estimates for 500-um risers, but the progenitor claim rests on an assumed duty cycle and a thin spec-z calibration; worth refereeing.","tokens_in":39546,"tokens_out":3346,"would_cite":true,"duration_ms":32167,"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":"Herschel 'ultrared' dusty galaxies are extreme, rare starbursts at median redshift 3.3 that cannot be the main progenitors of massive quiescent galaxies.","keywords":["dusty star-forming galaxies","ultrared galaxies","500 micron risers","submillimeter galaxies","Spitzer IRAC","Herschel SPIRE","SED fitting","high-redshift starbursts"],"falsifier":"Measure spectroscopic redshifts and interferometric sizes for a complete, flux-limited sample of 500 micron-risers over a large area, and independently determine the starburst duty cycle from the fraction of ultrared galaxies caught in merger-induced starburst phases; if the corrected ultrared space density at z~5 approaches the roughly 2 x $10^{-5}$ $Mpc^{-3}$ density of massive quiescent galaxies at z~3, or if the duty cycle turns out to be about 1 Gyr rather than 100 Myr, the paper's central demographic conclusion would fail.","tokens_in":38417,"feed_emoji":"🌌","tokens_out":6149,"duration_ms":58084,"temperature":0.7,"pith_summary":"Using Spitzer/IRAC follow-up of 300 Herschel-selected ultrared dusty star-forming galaxies (DSFGs), the paper builds a catalog of securely identified counterparts and focuses on 41 unlensed systems (63 individual sources) with high-resolution ALMA, NOEMA, or SMA positions. It models their full spectral energy distributions to show that this population is more extreme than the z~2.5 ALESS DSFGs: median redshift 3.3, stellar mass 3.7 x $10^{11}$ Msun, SFR 730 Msun/yr, dust luminosity 9.0 x $10^{12}$ Lsun, dust mass 2.8 x $10^{9}$ Msun, and V-band extinction 4.0. The central conclusion is that these ultrared galaxies, though the most luminous and massive dusty starbursts known in the early universe, are too rare to account for most of the star-forming progenitors of the massive quiescent galaxies seen at z~3. If right, it refocuses progenitor searches on different selection methods and makes the unlensed ultrared sample a clean laboratory for studying extreme star formation.","feed_headline":"Ultrared galaxies too rare to be main quiescent-galaxy ancestors","feed_subtitle":"SPIRE colors isolate a rare z~3.3 starburst population that cannot dominate massive galaxy ancestry","key_machinery":"The machinery has two parts. The selector is the ultrared color criterion S500>S350>S250, the '500 micron-riser' cut, which uses the rising SPIRE spectrum to pick out distant, dusty, star-forming galaxies beyond the typical z~2 DSFG peak; the classifier is high-resolution interferometry from ALMA, NOEMA, and SMA, which pinpoints counterparts and splits the sample into lensed, unlensed, and multiple-component systems. The property extractor is magphys+photo-z, an energy-balance SED model that combines stellar and dust emission and runs on IRAC 3.6/4.5 micron plus SPIRE, SCUBA-2/LABOCA, and ALMA photometry; it turns the photometry into redshifts, stellar masses, SFRs, dust luminosities, dust masses, and extinctions, and its stacked posterior distributions carry the demographic comparison.","core_discovery":"On the paper's own terms, the discovery is a demographic one: the 500 micron-riser selection, S500>S350>S250, efficiently finds a tail of DSFGs at z>4, but when followed up at high resolution and modeled with magphys, the unlensed subset turns out to be a rare population of intrinsically hyper-luminous, massive, heavily obscured starbursts (median z=3.3, M*=3.7 x $10^{11}$ Msun, SFR=730 Msun/yr, Ldust=9.0 x $10^{12}$ Lsun, Mdust=2.8 x $10^{9}$ Msun, Av=4.0). Comparing space density, SFR density, and stellar mass density against the massive quiescent galaxies at z~3 from near-infrared surveys, the paper concludes that these ultrared DSFGs contribute at most about 1.7 percent of the total stellar mass density at z~3.25 and cannot supply the majority of the progenitors of massive quiescent galaxies; instead they are the rarer, intrinsically most extreme systems in the early universe.","pith_inferences":["If the assumed ~100 Myr starburst duty cycle is an overestimate, the corrected space density and stellar mass density would rise, potentially bringing the ultrared population closer to the required progenitor densities; the duty cycle is assumed, not measured, in this paper.","A direct test would be to measure the prevalence and timescales of merger-induced starburst phases among ultrared galaxies, since the demographic comparison assumes all of them are caught in a single short-lived phase; rest-frame optical spectroscopy could tie stellar ages to the duty cycle.","The paper's conclusion is framed for the Herschel flux-limited sample; deeper 500 micron surveys or longer-wavelength selections could uncover a more numerous fainter ultrared population that may dominate the progenitor budget."],"forward_implications":["The unlensed ultrared sample, with median redshift 3.3 and extreme properties, is a set of intrinsic hyper-luminous infrared galaxies that can be studied without lens-model uncertainties.","After completeness and duty-cycle corrections, the ultrared population contributes at most about 1.7 percent of the total stellar mass density at z~3.25, so searches for massive-quiescent-galaxy progenitors should target other selections such as fainter submillimeter-selected DSFGs.","About 27 percent of Herschel ultrared sources break into multiple components at high resolution and roughly 20 percent would fail the 500 micron-riser cut without blending, so flux-boosting by blending must be folded into counts of high-redshift DSFGs.","The stacked SEDs show the ultrared population peaks at longer wavelengths and is more dust-obscured than ALESS DSFGs, implying that longer-wavelength selection reaches intrinsically more extreme galaxies."],"supporting_citations":[{"why":"Supplies the H-ATLAS ultrared sample selection, photometric redshifts, completeness corrections, and SCUBA-2/LABOCA photometry.","marker":"Ivison et al. (2016)"},{"why":"Defines the HeLMS 500 micron-riser sample and its higher flux cut.","marker":"Asboth et al. (2016)"},{"why":"Provides SCUBA-2 photometry and FIR SED redshifts for the HeLMS ultrared sources.","marker":"Duivenvoorden et al. (2018)"},{"why":"Supplies high-resolution ALMA positional priors and sizes used to identify counterparts and resolve multiple components.","marker":"Oteo et al. (2017)"},{"why":"Provides NOEMA and ALMA spectroscopic redshifts for H-ATLAS ultrared sources.","marker":"Fudamoto et al. (2017)"},{"why":"Supplies the magphys SED modeling approach and the ALESS DSFG comparison sample.","marker":"da Cunha et al. (2015)"},{"why":"Supplies the ZFOURGE massive-quiescent-galaxy stellar mass density that anchors the progenitor comparison.","marker":"Straatman et al. (2014)"},{"why":"Validates the magphys+photo-z method for deriving redshifts and physical properties of infrared-selected galaxies.","marker":"Battisti et al. (2019)"}],"fun_headline_variants":["Ultrared starbursts are too rare to be main quiescent ancestors","Rare ultrared DSFGs not the primary quiescent galaxy progenitors","500-micron risers: extreme starbursts, too rare for ancestry","Ultrared galaxies too scarce to explain massive quiescent origins","Hyper-luminous dusty galaxies fail as main quiescent ancestors"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that, after the completeness corrections of Ivison et al. (2016) and an assumed starburst duty cycle of about 100 Myr, the H-ATLAS unlensed ultrared sample traces the true space density of the ultrared population; if the duty cycle is wrong, the stellar-mass-density comparison and the 'cannot account for progenitors' conclusion lose quantitative support.","fun_headline_variants_meta":{"raw":{"variants":["Ultrared starbursts are too rare to be main quiescent ancestors","Rare ultrared DSFGs not the primary quiescent galaxy progenitors","500-micron risers: extreme starbursts, too rare for ancestry","Ultrared galaxies too scarce to explain massive quiescent origins","Hyper-luminous dusty galaxies fail as main quiescent ancestors"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000199,"raw_usage":{"total_tokens":1493,"prompt_tokens":1191,"completion_tokens":302,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":807,"completion_tokens_details":{"reasoning_tokens":203}},"tokens_in":807,"tokens_out":302,"duration_ms":3839,"temperature":1.0,"reasoning_tokens":203,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:53:52.742644+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure spectroscopic redshifts and interferometric sizes for a complete, flux-limited sample of 500 micron-risers over a large area, and independently determine the starburst duty cycle from the fraction of ultrared galaxies caught in merger-induced starburst phases; if the corrected ultrared space density at z~5 approaches the roughly 2 x $10^{-5}$ $Mpc^{-3}$ density of massive quiescent galaxies at z~3, or if the duty cycle turns out to be about 1 Gyr rather than 100 Myr, the paper's central demographic conclusion would fail.","supporting_citations":[{"cited_title":"2016, MNRAS, 462, 1989 Ashby, M","cited_arxiv_id":null,"evidence_quote":"Defines the HeLMS 500 micron-riser sample and its higher flux cut."},{"cited_title":"Witnessing the birth of the red sequence: the physical scale and morphology of dust emission in hyper-luminous starbursts in the early Universe","cited_arxiv_id":"1709.04191","evidence_quote":"Supplies high-resolution ALMA positional priors and sizes used to identify counterparts and resolve multiple components."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the ZFOURGE massive-quiescent-galaxy stellar mass density that anchors the progenitor comparison."}],"review_version":1}