{"id":"1297db05-80b0-4202-88e1-12d4fa0ef0e3","arxiv_id":"2505.11919","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Gas around DESI quasars shows metal absorption that declines with redshift and radius, with an FeII/MgII ratio that grows toward low redshift, indicating progressive metal enrichment.","lead":"Astronomers stacked spectra from the Dark Energy Spectroscopic Instrument to map faint gas around quasars out to 6.5 megaparsecs. They find that metal absorption weakens with cosmic time and with distance from the quasar, and that the iron-to-magnesium ratio rises toward low redshift.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Missing random-pair control stack leaves the inferred QSO-associated W(z) and W(D) trends potentially contaminated by redshift-dependent IGM metal absorption.","rationale":"The reader's weakest_assumption correctly identifies the absence of a random-pair control stack or cosmic-mean IGM subtraction as the load-bearing issue. Reading the full text confirms that Section 2.2 describes no control sample, and the Ly α forest discussion in Appendix B addresses a different effect (background-forest mixing) rather than foreground-redshift IGM absorption. Without this baseline, the measured W in each bin is a sum of QSO-associated gas and an evolving IGM contribution, so the global slopes W ∝ (1+z)^−4.0±2.7, the radial slopes W ∝ D^−0.50±0.38, and the FeII/MgII redshift trend are all potentially biased. This is not an outside-consensus disagreement but a missing internal control, and it is directly testable. I also note that the quoted global slope in the abstract and Section 4 (−4.0±2.7) differs from the ensemble slopes of the normalized data in Section 3.1 (−1.78 to −2.97), and the reduced chi-squared values in Figure 4 are far below unity, both of which support caution but do not change the verdict. Since the suggested control experiment can settle the matter, the appropriate verdict remains CONDITIONAL as given by the reader.","tokens_in":18861,"tokens_out":10889,"duration_ms":111893,"concrete_test":"Construct a control sample by replacing each foreground QSO with a random sky position at the same redshift, then assign background QSO spectra with the same impact parameter distribution B and run the identical stacking and Gaussian-fitting pipeline of Section 2.2 to measure W_control for the same 33 lines in each redshift and D bin. Subtract W_control from the observed W and refit the slopes in Figures 2 and 3, the halo and linear fits in Figure 4, and the FeII/MgII ratio in Figure 6. Accept the central claim only if the refitted slopes and ratios remain within 1σ of the quoted values and W_control is below 10% of the observed W in a majority of bins; if the slope shifts exceed 1σ, the stated QSO-associated evolution and radial trends are not uniquely attributable to QSO halos.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claims in Section 3 and Section 4 rest on the assumption that the measured stacked absorption originates from gas physically associated with the foreground QSOs. However, the stacking procedure in Section 2.2 shifts background QSO spectra to the foreground QSO rest frame and co-adds them without subtracting a random-pair control stack or the cosmic mean IGM absorption at the foreground redshift. Consequently, W in any given redshift and impact parameter bin includes a contribution from unrelated IGM metal lines (e.g., Mg II, C IV, Fe II) at that same redshift, along any line of sight. This baseline is not constant: the cosmic mean equivalent width density of metal absorbers evolves strongly with redshift, so the unsubtracted IGM term is redshift dependent. That can bias the quoted global slope W ∝ (1+z)^−4.0±2.7 and the W_FeII/W_MgII ratio trend, exactly in the direction of apparent metal enrichment with cosmic time. For the radial analysis in Section 3.2, an unsubtracted uniform IGM term acts as an additive offset that flattens the W(D) profiles and biases the fitted amplitudes of the 1-halo and 2-halo terms, potentially mimicking or suppressing a two-halo component. Appendix B only tests the mixing of the background Ly α forest into the stacked spectral regions; it does not remove or quantify the foreground-redshift IGM absorption that is common to all sightlines. The bootstrap errors reported in Table 1 represent spectral noise, not this systematic baseline uncertainty.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper uses foreground-background QSO pairs from the DESI Early Data Release to stack background spectra at foreground QSO rest wavelengths and measure the equivalent widths of 33 absorption lines from 10 elements. The authors report that absorption equivalent width decreases with redshift as W ∝ (1+z)^-4.0±2.7, that W decreases with impact parameter as W ∝ D^-0.50±0.38, that the radial profiles are well described by a two-halo model, and that W_FeII/W_MgII increases toward low redshift. They interpret these trends as evidence for progressive metal enrichment of QSO-associated gas and for a multiphase gas distribution extending from the circumgalactic medium to several megaparsec scales.","tokens_in":19166,"tokens_out":5281,"duration_ms":61184,"significance":"If the trends hold, this would be a valuable multi-ion census of diffuse gas around QSOs across 0.3 < z < 3.5 and 0.02-6.5 Mpc, with the DESI EDR sample providing an order-of-magnitude increase in pair statistics over earlier work. The public data release and the systematic treatment of 33 lines spanning cold, cool, and warm phases are clear strengths. The paper also makes a useful methodological connection between stacked absorption profiles and two-halo clustering models. However, the quantitative significance of the headline slopes is modest, and the central astrophysical interpretation depends on a foreground-IGM contamination assumption that is not yet demonstrated.","major_comments":[{"comment":"The stacking procedure co-adds background QSO spectra at foreground QSO rest wavelengths without subtracting a random-pair control stack or the cosmic mean IGM absorption at the foreground redshift. The measured W therefore contains a contribution from unrelated intervening metal absorbers (e.g., Mg II, C IV, Fe II) at that redshift along every line of sight. Because the incidence and equivalent-width density of IGM metal absorbers evolve strongly with redshift, this baseline is redshift dependent and can bias the quoted slopes W ∝ (1+z)^-4.0±2.7 and the W_FeII/W_MgII evolution in the direction of apparent enrichment with cosmic time. For the radial analysis in Section 3.2, a uniform IGM term adds an offset that flattens W(D) and biases the amplitudes of the one- and two-halo terms. Appendix B addresses only the mixing of the background Ly-alpha forest into the stacked spectral regions, not foreground-redshift IGM absorption common to all sightlines. The authors should quantify this baseline, for example by stacking pairs with randomized foreground redshifts/positions or by subtracting the expected IGM metal contribution, and should re-derive the quoted slopes with that control.","section":"Section 2.2 and Section 3.1"},{"comment":"The headline trend W ∝ (1+z)^-4.0±2.7 is not derived explicitly in the text. Section 3.1 reports per-radial-bin slopes of -1.78±1.74, -2.31±1.50, and -2.97±1.32, while Figure 3 shows large scatter in the per-ion slopes; the abstract quotes a single global slope without stating how it is computed. The paper should specify whether -4.0±2.7 is, for example, an average of individual-ion slopes, and should test whether the redshift evolution is statistically consistent across radial bins and ion classes. As written, the reader cannot reproduce the headline number from the information given, and its uncertainty is large enough that the strength of the central claim is not transparent.","section":"Section 4 and Figures 2-3"},{"comment":"The two-halo model fit is central to the claim that the radial distribution is 'effectively characterized by a two-halo model', but the model equations, fitted parameter values, priors, and parameter uncertainties are not given in the text. Only reduced chi-squared values are shown for the halo model versus a linear fit. Without specifying the model definition and the number of free parameters, the reader cannot assess whether the two-halo model is genuinely preferred or whether the quoted radial slope W ∝ D^-0.50±0.38 is robust. The authors should provide the model equations, best-fit parameters, uncertainties, and a model-comparison statistic (e.g., AIC or BIC), and should clarify whether all redshift bins are combined in these fits.","section":"Section 3.2 and Figure 4"}],"minor_comments":[{"comment":"The abstract states that 'Different species that trace multi-phase gas exhibit distinct evolutionary patterns', but the text in Section 3.1 cautions that the differences among cold, cool, warm, alpha, and Fe-peak slopes are not statistically significant given the large uncertainties. The abstract should be tempered to match this caveat.","section":"Abstract and Section 3.1"},{"comment":"There is a typo: 'exanple' should be 'example'. In addition, the statement that the W_FeII/W_MgII trend is 'consistent with Dey et al. (2015)' is loose, since Dey et al. parameterize the ratio as a linear function of z while here it is fit as a power law in (1+z); the authors should clarify the comparison.","section":"Section 3.3"},{"comment":"The caption writes 'S iv' in one place; this should be 'Si IV' for consistency with the other ions.","section":"Figure 3 caption"},{"comment":"The table omits all error bars and refers the reader to the online dataset for the complete table with errors. Given that the paper repeatedly applies S/N cuts and asterisk flags, providing at least representative uncertainties in the printed table would make the quality cuts easier to evaluate.","section":"Table 1"},{"comment":"The phrase 'enhance the production of α-elements are produced' is ungrammatical and should be rewritten, e.g., 'enhance the production of α-elements'.","section":"Section 4"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope for an astrophysics journal and the data release is a genuine asset. The main risk is not internal inconsistency but an unquantified foreground-IGM baseline; if a control stack shows a large contribution, the QSO-association interpretation will need substantial revision. The other two major comments are fixable with clearer reporting. I would encourage the editor to request a revision rather than reject, provided the control analysis is added."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the short version. The paper gives you the most complete stacked absorption census of QSO-associated gas I know of: 33 lines, 10 elements, DESI EDR, with the W table publicly released. If you work on CGM/IGM enrichment, the data table itself is worth having. The qualitative message—low-ionization gas falls off steeply with radius, high-ionization extends further, and the FeII/MgII ratio rises toward low z—is consistent across ions and in line with earlier work, so I don't think the main structure is fabricated.\n\nThe soft spot is real and it's the one the stress-test flags. There is no random-pair control stack. The authors stack background QSO spectra at the foreground QSO rest wavelengths, and whatever metal lines exist in the IGM at that same redshift come along for the ride. That baseline is not constant in z, so the quoted W∝(1+z)^-4.0±2.7 and the FeII/MgII evolution are at least partly contaminated by global IGM metal evolution. Appendix B only deals with mixing from the background Lyα forest, which is a different effect. For the radial profiles an unsubtracted uniform component would flatten W(D) and bias the two-halo amplitudes. I don't see this as a fatal flaw—the trends are plausible and the authors are careful in other respects—but it is the load-bearing assumption and it needs a quantitative answer before the slopes can be trusted. A simple random-pair stack or a foreground-redshift IGM model would do.\n\nMinor presentation issue: the abstract's -4.0±2.7 is an average over per-ion slopes with huge scatter, and the Figure 2 per-bin slopes are -1.78 to -2.97. The paper doesn't make clear which statistic you're actually looking at. Same for D^-0.50±0.38 vs the per-ion β values. Fix that with a clear definition of the quoted global slopes.\n\nThe paper deserves a serious referee, but the referee should push for the control measurement. I'd say accept only after the authors subtract or bound the IGM contribution and make the slope definitions consistent.","headline":"A genuinely useful 33-line stacked absorption census from DESI EDR, but the missing random-pair control stack means the headline W(z) and W(D) slopes are not yet clean QSO-associated measurements.","tokens_in":19781,"tokens_out":2282,"would_cite":true,"duration_ms":25682,"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":"Stacked DESI spectra show the equivalent width of 33 absorption lines around quasars falling as $(1+z)^{-4.0\\pm 2.7}$ with redshift and $D^{-0.50\\pm 0.38}$ with impact parameter, indicating that QSO-associated gas has become progressively…","keywords":["quasar absorption lines","circumgalactic medium","multiphase gas","DESI early data release","spectral stacking","metal enrichment","two-halo model","Fe II/Mg II ratio"],"falsifier":"Stack the same background spectra around foreground objects with the same redshifts but positions unrelated to the background sight lines, forming a random-pair control stack. If the equivalent widths show the same $W \\propto (1+z)^{-4}$ decline without any foreground association, then the claimed cosmic enrichment of QSO-associated gas would instead be tracing the intergalactic absorber population rather than quasar halos.","tokens_in":18658,"feed_emoji":"🔭","tokens_out":8724,"duration_ms":85213,"temperature":0.7,"pith_summary":"This paper claims that gas associated with quasars has become progressively metal-enriched over cosmic time: the equivalent width $W$ of metal absorption lines in stacked background spectra falls with redshift as $W \\propto (1+z)^{-4.0\\pm 2.7}$ and with projected distance from the quasar as $W \\propto D^{-0.50\\pm 0.38}$. It also claims that the radial profiles follow a two-halo model, with absorption coming both from the quasar's own halo and from neighboring halos along large-scale structure, and that $W_{\\rm FeII}/W_{\\rm MgII}$ rises toward low redshift. If these trends are real, low-ionization gas clusters like galaxies while high-ionization gas extends farther, and the Fe/Mg ratio traces delayed iron enrichment or a top-heavy stellar IMF at early times.","feed_headline":"Metal absorption around quasars fades with redshift and distance","feed_subtitle":"Stacking 33 absorption lines in DESI spectra traces metal enrichment to scales of 6.5 Mpc.","key_machinery":"The machinery is median-stack absorption spectroscopy: background QSO spectra are continuum-normalized, median-filtered, co-added in the rest frame of the foreground QSO, and Gaussian-fitted to measure equivalent widths, with bootstrap resampling for uncertainties. The central interpretive object for the radial claim is the two-halo model, which splits the equivalent-width profile $W(D)$ into a one-halo term from gas in the quasar's own dark-matter halo and a two-halo term from gas in correlated neighboring halos; this is what converts the measured decline with impact parameter into a statement about gas extending from the circumgalactic medium into the large-scale structure. For the abundance claim, the ratio $W_{\\rm FeII}/W_{\\rm MgII}$ is the diagnostic that links absorption strength to nucleosynthetic delay times and possible IMF variations.","core_discovery":"On the paper's own terms, the discovery is a coherent set of scaling relations for diffuse multiphase gas around QSOs, measured from stacked DESI Early Data Release spectra. Analyzing 33 absorption lines from ten elements across $0.3<z<3.5$ and projected distances of 0.02 to 6.5 Mpc, the paper finds that absorption strength decreases with increasing redshift in all radial bins and that the decline is species-dependent: warm, high-ionization gas evolves faster than cold or cool gas, and Fe-peak elements evolve somewhat faster than $\\alpha$-elements, though the paper cautions that the statistical significance is limited. The absorption strength also decreases with impact parameter and is better described by a two-halo model than by a single power law; comparing with galaxy two-point correlation functions, low-ionization ions (Mg II, Fe II) have clustering scales similar to galaxies, while high-ionization ions (C IV, Si IV) are more extended. Finally, the paper reports that $W_{\\rm FeII}/W_{\\rm MgII}$ decreases with redshift as $(1+z)^{-2.59\\pm 1.41}$, consistent with earlier work, and interprets this as delayed iron enrichment from Type Ia supernovae relative to magnesium from core-collapse supernovae, possibly combined with a top-heavy IMF at high redshift.","pith_inferences":["A direct test the paper does not perform is a random-pair control stack: stacking background spectra around foreground objects with the same redshifts but unrelated sight lines would show how much of the measured $W(z)$ decline is due to the global intergalactic metal baseline rather than to gas physically associated with the foreground QSO.","The two-halo fit could be pushed further by fitting the one-halo term with an assumed halo mass, yielding an absorption-weighted host halo mass for QSOs that could be compared with QSO clustering masses to test whether the gas traces the same halos as the quasars themselves.","The Fe II/Mg II ratio trend could be converted into a quantitative cosmic [Fe/Mg] enrichment timeline if photoionization corrections were applied; the paper uses the equivalent-width ratio directly, so the implied abundance evolution could change once ionization-state variations with redshift are modeled.","The upturn at roughly 5 Mpc attributed to QSO proximity zones predicts a luminosity dependence: brighter quasars should show the upturn at larger projected separations, a test that could be done by splitting the stack by QSO luminosity."],"forward_implications":["If $W \\propto (1+z)^{-4.0\\pm 2.7}$ is correct, quasar-associated gas has become substantially enriched in metals from $z\\sim3$ to $z\\sim0.5$, with the fastest growth in warm, high-ionization phases.","A two-halo radial profile implies that quasar-associated gas extends beyond the virial radius to at least 6.5 Mpc, connecting the circumgalactic medium to the large-scale structure.","The similarity between low-ionization ion slopes and galaxy two-point correlation slopes indicates that cool, low-ionization gas traces the same biased dark-matter halos as galaxies, while high-ionization gas traces a more diffuse, extended phase.","The rising Fe II/Mg II ratio toward low redshift, interpreted as an abundance trend, links gas-phase absorption to galactic chemical evolution, specifically delayed iron production and possibly a top-heavy IMF at high redshift.","Because the same decline of $W$ with redshift appears for ELG and LRG foregrounds as well as QSOs, the metal-enrichment trend is a general property of galaxies and quasars rather than a QSO-specific effect."],"supporting_citations":[{"why":"Supplies the spectral-stacking and pair-cross-matching methodology that this analysis extends to QSOs and 33 absorption lines.","marker":"Chen et al. (2025)"},{"why":"Introduced the two-halo model fit for Mg II absorption profiles against which the radial profiles are modeled and compared.","marker":"Zhu et al. (2014)"},{"why":"Provides the earlier Fe II/Mg II versus redshift relation that the paper's ratio trend is checked against.","marker":"Dey et al. (2015)"},{"why":"Establishes the stacking approach as an unbiased way to measure the mean absorption strength of gas around galaxies.","marker":"Zhu & Ménard 2013a"},{"why":"Reports weak Mg II evolution, the trend the paper compares its own Mg II slope with.","marker":"Lan (2020)"},{"why":"Supplies the projected two-point correlation slope of galaxies at z 2-3.5 used to classify ion distributions as clustered versus extended.","marker":"Durkalec et al. (2018)"},{"why":"Publishes the Early Data Release spectra and catalogs that provide the entire dataset for the stacking measurement.","marker":"DESI Collaboration et al. (2024)"},{"why":"Provides the continuum-fitting code used to normalize the QSO spectra before stacking.","marker":"Guo et al. (2018)"}],"fun_headline_variants":["Quasar gas absorption fades with cosmic time and distance","High-ionization gas around quasars extends beyond galaxies","Iron enrichment lags behind magnesium in quasar gas","DESI stacks reveal two-halo structure in quasar gas"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the stacked absorption arises from gas physically associated with the foreground QSOs, yet the analysis does not subtract a random-pair control stack or the cosmic mean intergalactic absorption, so the measured $W(z)$ and $W(D)$ trends carry that association assumption.","fun_headline_variants_meta":{"raw":{"variants":["Quasar gas absorption fades with cosmic time and distance","High-ionization gas around quasars extends beyond galaxies","Iron enrichment lags behind magnesium in quasar gas","DESI stacks reveal two-halo structure in quasar gas"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000748,"raw_usage":{"total_tokens":3398,"prompt_tokens":1080,"completion_tokens":2318,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":696,"completion_tokens_details":{"reasoning_tokens":2252}},"tokens_in":696,"tokens_out":2318,"duration_ms":19651,"temperature":1.0,"reasoning_tokens":2252,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T20:45:49.641207+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Stack the same background spectra around foreground objects with the same redshifts but positions unrelated to the background sight lines, forming a random-pair control stack. If the equivalent widths show the same $W \\propto (1+z)^{-4}$ decline without any foreground association, then the claimed cosmic enrichment of QSO-associated gas would instead be tracing the intergalactic absorber population rather than quasar halos.","supporting_citations":[{"cited_title":"2014, Monthly Notices of the Royal Astronomical Society, 439, 3139","cited_arxiv_id":null,"evidence_quote":"Introduced the two-halo model fit for Mg II absorption profiles against which the radial profiles are modeled and compared."}],"review_version":1}