{"id":"82bd5640-3b4d-4557-b82d-4d80aeaf9a87","arxiv_id":"2411.13208","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Bright quasars at all redshifts up to 3 share a remarkably similar optical-to-extreme-ultraviolet spectral energy distribution, which is redder in the extreme UV than previous composites.","lead":"Analyzing 23,256 bright quasars with SDSS and GALEX photometry, this paper builds average far-ultraviolet to optical emission templates and finds them nearly identical across redshifts from z=0 to z=3. The result hints at a universal quasar disk with less ionizing radiation than standard models predict, affecting black hole growth and cosmic reionization estimates.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The EUV universality is produced by a single assumed IGM absorber model; an independent IGM correction or forward-model test is needed before accepting the 'surprisingly universal' SED.","rationale":"The paper's strongest claim is the redshift/luminosity-independent composite SED down to ~500 Å. The optical-to-FUV portion is consistent with previous composites, and the CW23 z~2 result provides independent support for luminosity independence at fixed redshift. However, the genuinely new result — that the EUV is universal across 0<z<3 — is obtained only after applying the IGM absorption correction in §2.4. The bias-free SEDs in Figure 4 show a clear redshift trend in the EUV, which the correction removes. A single assumed absorber population (Faucher-Giguère 2020) is used with no sensitivity test or empirical calibration. At high z the correction factors are large (mean T_FUV~0.2; median T_FUV<0.04), so even modest errors in the absorber model systematically shift the corrected SEDs and could create or erase the apparent universality. The median correction also uses a statistically unjustified ratio-of-medians. These issues make the central claim conditional rather than established. The proposed check — repeating with an independent IGM model and a forward-model fit for the median — would settle whether the universality is real or an artifact. The reader's CONDITIONAL verdict is appropriate; no verdict change is needed.","tokens_in":33275,"tokens_out":7893,"duration_ms":87298,"concrete_test":"Compute the IGM-corrected EUV SEDs for the z=1.5–2.9 bins using an independent absorption model calibrated to measured effective optical depths (e.g., Becker et al. 2021) and a different b-value/column-density distribution, then quantify the residual redshift trend. For the median, replace the ratio-of-medians correction with a forward model that convolves the assumed intrinsic SED distribution with the simulated transmission distribution and fits the observed GALEX flux distribution including upper limits; if the inferred intrinsic median SEDs no longer agree across redshift, the universality claim is an artifact.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 2.4 applies a Monte Carlo IGM transmission correction based solely on the Faucher-Giguère (2020) absorber population (log N_HI 12–22, b=30 km/s). The central claim of redshift-independent EUV SEDs rests entirely on this correction: without it, the bias-free SEDs in Figure 4 become progressively redder with redshift. The correction is not validated against the actual mean IGM opacity (e.g., measured effective optical depths), and no sensitivity analysis over absorber parameters is presented. Because at z ≳ 2 the mean transmission is ~0.2 and the median transmission at FUV wavelengths is <0.04, the corrected EUV points are highly sensitive to the assumed absorber distribution. Moreover, for the median SED the paper divides the observed median flux by the median transmission; for independent intrinsic flux and transmission this ratio-of-medians is not the median of the ratio, and with a large fraction of near-zero transmissions it can produce an unstable, arbitrary correction. Thus the 'striking consistency' in Figure 7 could be an artifact of the model choice rather than evidence of an intrinsic universal SED.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs mean and median rest-frame optical-to-EUV spectral energy distributions for 23,256 SDSS DR14Q quasars at 0 < z < 3, using GALEX photometry and modeling GALEX non-detections with an exponentially modified Gaussian distribution. After correcting for intergalactic-medium (IGM) absorption using a Monte Carlo simulation based on the Faucher-Giguère (2020) absorber population, the author reports that the intrinsic EUV SEDs are remarkably independent of redshift and luminosity above log L_bol ≃ 45.5, forming a universal composite SED for quasars since cosmic noon. The composite is redder in the EUV than previous composites, implying less ionizing radiation, and the paper argues that a simply truncated disk model is preferred over a standard thin disk model.","tokens_in":33493,"tokens_out":8556,"duration_ms":88694,"significance":"If the central claim holds, this is an important empirical result: a stable quasar EUV SED over 0 < z < 3 and roughly two decades in luminosity would constrain accretion-disk models, the ionizing photon budget of quasars, and the interpretation of broad emission lines. The paper uses a large, well-defined sample and improves on earlier work by explicitly treating GALEX non-detections and by propagating IGM-correction uncertainties. The data products are made publicly available, which is a strength. However, the redshift-independence claim currently rests mainly on visual consistency after a model-dependent IGM correction, and the selection-cutoff robustness test shows quantitative variation in the fitted EUV slope. The result needs stronger validation before the 'surprising universality' is accepted.","major_comments":[{"comment":"The redshift independence of the intrinsic EUV SED is introduced by the IGM correction: the uncorrected bias-free SEDs in Figure 4 become progressively redder with redshift, and the correction brings them into agreement. The correction is applied with a single assumed absorber population (Faucher-Giguère 2020, log N_HI 12-22, b = 30 km/s) and the filter-weighted transmissions are derived from that model alone. No sensitivity analysis over absorber parameters (e.g., the evolution of Lyman-limit systems, the Doppler parameter, or the column-density cutoff) and no validation against measured effective optical depths are presented. Since the mean FUV transmission at z ~ 2.9 is ~0.2 and the median FUV transmission at high z is as small as ~0.001, the corrected EUV points are highly sensitive to the assumed absorber distribution. I request an explicit robustness test: vary the absorber model within plausible bounds, or compare the filter-weighted transmissions to empirical IGM opacity measurements, and show that the redshift-independence claim survives. Without such a test, the 'strikingly consistent' EUV shape in Figure 7 could be an artifact of the adopted correction.","section":"Section 2.4, Figures 4 and 7"},{"comment":"The 'striking' agreement among the intrinsic SEDs at different redshifts is not quantified by any statistical test. The universality claim needs a formal comparison, e.g., binning the EUV points into common rest-wavelength intervals and computing a chi-square or likelihood for a single common SED versus redshift-dependent shapes. This is especially important for the median SED at z = 2.5-2.9, where the median FUV transmission is 0.006-0.001 and the corrected points involve dividing by very small, uncertain factors. The paper itself states that the intrinsic median SED at λ_rest < 500 Å is 'very uncertain' (p. 8), which is in tension with the abstract's claim of a universal median composite. Please report a quantitative test for the mean and median separately, and for the wavelength range where the correction is robust.","section":"Section 2.5, Figure 7"},{"comment":"The ±0.2 dex cutoff test is presented as confirming universality, but the best-fit parameters change substantially. For the mean composite, α_EUV goes from -2.73 ± 0.03 (reference) to -4.18 ± 0.04 (higher cutoffs), while for the median composite it changes from -6.78 ± 0.21 to -6.37 ± 0.07 (higher cutoffs) and -5.14 ± 0.08 (lower cutoffs). These shifts are far larger than the quoted 1σ uncertainties and indicate that the measured EUV slope depends on the selected luminosity range or on the accompanying changes in sample size and detection fraction. The claim of luminosity independence should be supported either by showing that these differences are within the systematic uncertainty, or by restricting the universality claim to the parameter range where the test is demonstrably stable.","section":"Section 3.1, Figure 9"},{"comment":"The preference for the truncated disk model over the standard thin disk model is based on visual comparison of the model-predicted and observed composite SEDs. No goodness-of-fit statistic or model-comparison metric is presented, and the comparison does not appear to propagate the observed SED uncertainties, including the IGM-correction uncertainties, into the model comparison. Given that the models differ mainly in the EUV, where the data corrections are most uncertain, a quantitative comparison (e.g., a chi-square over the fitted wavelength range, or a likelihood ratio) is needed to support the claim that the truncated disk is favored.","section":"Section 3.2, Figure 11"}],"minor_comments":[{"comment":"The caption contains the typo 'model-predicated' and should read 'model-predicted'.","section":"Figure 11 caption"},{"comment":"The notation f_w^NUV and f_w^FUV used in the Figure 4 legends is not defined in the text until the caption; please define these symbols at first use.","section":"Section 2.3, Figure 4"},{"comment":"The phrase 'since cosmic noon' is used without a definition; please state the redshift range (e.g., z ~ 0-3) explicitly at its first occurrence.","section":"Title and abstract"},{"comment":"The data are provided via a URL without a persistent identifier; consider registering a DOI for the released SED products to ensure long-term accessibility.","section":"Data Availability Statement"},{"comment":"There is a subject-verb agreement error: 'the universality and the smoothness ... suggests' should be 'the universality and the smoothness ... suggest'.","section":"Section 3.4"},{"comment":"The note at the end of the references about an error in Cai & Wang (2023) is an erratum to a previous paper; it would be clearer as a footnote or separate erratum rather than a remark in the reference list.","section":"Notes section"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about the ratio-of-medians is not the most serious issue: since the SEDs are shown in log flux and the medians are evidently taken in log space, dividing the median flux by the median transmission is equivalent to the median of the ratio under the usual log-space convention. The more serious concerns are the single IGM absorber model and the lack of a formal statistical test of redshift independence. The author's reliance on the previous CW23 methodology is legitimate and cited, but this manuscript needs to demonstrate that the central claim is robust to plausible variations in the IGM correction and to quantitative selection effects. The journal scope is appropriate for this empirical survey analysis; after major revision addressing the IGM sensitivity and the quantitative tests, the paper could be a valuable contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the optical-to-FUV side of this paper is a real advance; the headline EUV universality is not yet established. What is genuinely new is the extension of Cai & Wang (2023) from z~2 to 0<z<3, showing that the intrinsic mean/median quasar SED is redshift-independent from rest-frame optical down to ~500 Å, and that the composite is redder in the EUV than Telfer et al. (2002). The sample work is careful: 23,256 unique bright quasars, GALEX non-detections treated with an exponentially modified Gaussian, bootstrap errors, and a robustness check to ±0.2 dex changes in the luminosity cutoffs. The data tables are posted. The author also flags, rather than hides, that the median EUV below 500 Å is very uncertain. That is honest.\n\nThe soft spots are where the stress-test note lands. The bias-free SEDs in Figure 4 get progressively redder with redshift; applying the IGM correction makes them consistent. The IGM correction comes from one assumed absorber population (Faucher-Giguère 2020, log N_HI 12–22, b=30 km/s), with no sensitivity analysis over absorber parameters and no validation against measured IGM opacities. At z~2–3 the mean transmission is ~0.2 and the median is under 0.04 at FUV wavelengths, so the corrected EUV points are extremely sensitive to the assumed distribution. And for the median SED, dividing the observed median by the median transmission is not a well-behaved estimator when many transmissions are near zero. This is a load-bearing weakness, not a minor one: the central \"surprisingly universal\" claim currently rests on it.\n\nTwo smaller issues. First, the universality is judged by eye in Figure 7; there is no formal test that the SEDs in different redshift bins are consistent beyond their error bars. Second, the claimed independence from black hole mass and Eddington ratio is indirect: M_BH and λ_Edd evolve with redshift in this sample, so redshift and physical parameter dependencies are conflated. I would soften that wording. The disk model comparison is interpretative, not a derivation, and the author says so.\n\nWho gets value: anyone building SED templates or bolometric corrections for luminous quasars, and people working on accretion disk models. The optical-to-FUV composite is worth having even if the EUV story changes. My recommendation: send it to peer review, with a referee asked to demand an independent or parametrically varied IGM correction, a formal consistency test across redshift bins, and more cautious wording on the mass/Eddington independence. I would not desk-reject it.","headline":"The optical-to-FUV part is solid and useful; the EUV universality headline rests on a single IGM absorber model and needs an independent check before it is established.","tokens_in":34038,"tokens_out":3891,"would_cite":true,"duration_ms":41727,"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":"Bright quasars share one universal spectrum, from the optical to 500 Å, independent of redshift and luminosity.","keywords":["quasar SED","composite spectrum","extreme ultraviolet","intergalactic medium absorption","redshift independence","accretion disk models","SDSS quasars","GALEX photometry"],"falsifier":"Measure the mean and median EUV slopes of low-redshift ($z\\lesssim0.5$) quasars spectroscopically below 912 Å, where intergalactic absorption is negligible; slopes matching $\\alpha_{\\mathrm{EUV}}\\simeq-2.7$ and $-6.8$ would support the universality, while slopes near older composites would indicate the intergalactic correction produced it.","tokens_in":33057,"feed_emoji":"🔭","tokens_out":10968,"duration_ms":101613,"temperature":0.7,"pith_summary":"This paper reports that bright quasars have had one universal average spectrum, from the optical down to about 500 Å in the extreme ultraviolet, ever since cosmic noon (the epoch of peak quasar activity around z~2–3). The composite spectral energy distribution (SED) is built from 23,256 quasars in the SDSS and GALEX surveys, binned in redshift from z=0 to z=3, with GALEX non-detections explicitly included so the sample is not biased toward ultraviolet-bright objects. After correcting for absorption by the intergalactic medium, the SEDs in every redshift bin agree with each other: the shape is independent of redshift, independent of luminosity for bolometric luminosities above $10^{45}$.5 erg/s, and plausibly independent of black hole mass and Eddington ratio. A key consequence is that the extreme-ultraviolet continuum is redder than earlier composites, meaning quasars produce less ionizing radiation than previously assumed, which matters for models of broad emission lines and cosmic reionization. The paper argues that this universal SED favors a truncated accretion disk over the standard thin disk.","feed_headline":"Bright quasars share one universal spectrum down to 500 Å","feed_subtitle":"After correcting intergalactic absorption, 23,000 bright quasars at z<3 show one optical-to-EUV spectrum.","key_machinery":"The central object is the rest-frame composite SED, built by normalizing each quasar at 2200 Å and combining de-redshifted SDSS and GALEX photometry; for the ultraviolet bands, GALEX non-detections are replaced by $3\\sigma$ upper limits and an exponentially modified Gaussian is fit to the log-luminosity distribution so the mean and median are not biased by which quasars happened to be detected. The second load-bearing device is the Monte Carlo IGM transmission correction: for each redshift bin, 1000 simulated lines of sight drawn from an absorber population with neutral-hydrogen column densities $12 < \\log N_{\\mathrm{HI}} < 22$ and Doppler parameter 30 km/s produce filter-weighted mean and median transmissions, and these are applied to the rest-frame extreme-ultraviolet points before the redshift comparison is made. The resulting universal SED is summarized by a smoothly broken power law with break wavelength $\\lambda_b \\approx 1144$ Å for the mean and 906 Å for the median, and with EUV spectral indices of $\\alpha_{\\mathrm{EUV}}=-2.73\\pm0.03$ and $-6.78\\pm0.21$.","core_discovery":"The central claim is that an intrinsic mean/median composite SED exists for quasars since cosmic noon: after correcting for intergalactic absorption, the average SED of quasars with bolometric luminosity above $10^{45.5}$ erg/s is independent of redshift between $z=0$ and $z=3$, and the same shape holds for mean and median in a way that also appears independent of black hole mass and Eddington ratio. At wavelengths beyond about 1000 Å the composite matches previous spectra, but in the extreme ultraviolet it is redder, with best-fit EUV spectral indices of $\\alpha_{\\mathrm{EUV}} = -2.73 \\pm 0.03$ for the mean and $-6.78 \\pm 0.21$ for the median, corresponding to 1.6 and 3.6 times less ionizing radiation between 912 Å and 300 Å than earlier composites. The paper further concludes that this universal shape, and its indifference to the physical properties of the quasar, rules out the standard thin disk model as the sole source of the optical-to-EUV continuum and favors a simply truncated disk model, while noting that more sophisticated models are needed.","pith_inferences":["If the universality is real, earlier reports of a luminosity-dependent extreme-ultraviolet slope are selection artifacts, and the same non-detection-aware averaging could be applied to X-ray samples to test whether the X-ray/UV relation is similarly universal.","The large gap between the mean and median EUV slopes implies a strongly skewed distribution of EUV brightness at fixed optical luminosity; the full distribution, not just the average, could be used to constrain orientation, variability, or patchy host-galaxy attenuation.","A direct and cheap test is available: low-redshift ($z<0.5$) quasars observed in the rest-frame EUV with little intergalactic absorption should show the same red slopes if the paper's picture is correct, and a slope matching older composites would point to the intergalactic correction as the cause of the apparent universality.","If the EUV deficit holds up, reionization models that lean on quasars will need larger galaxy contributions or higher escape fractions, and broad-line-region photoionization models will need to produce strong lines from a weaker ionizing continuum."],"forward_implications":["A single redshift-independent SED template can replace luminosity-dependent templates for bright quasars at $z<3$ when fitting photometry from the optical to 500 Å.","The extreme-ultraviolet ionizing continuum is weaker than previously assumed by factors of 1.6 (mean) and 3.6 (median), which lowers the expected quasar contribution to hydrogen reionization and changes photoionization calculations for broad emission lines.","The standard thin disk model predicts redshift-dependent SEDs and is disfavored, while a simply truncated disk model with a maximum temperature nearly independent of black hole mass and Eddington ratio comes closer to the observations.","Average dust attenuation and hydrogen absorption along quasar lines of sight must be nearly unchanged from $z=0$ to $z=3$ for bright quasars, unless they are finely tuned to cancel redshift trends.","Handling ultraviolet non-detections rather than dropping them removes the detection bias that made earlier EUV composites look luminosity dependent."],"supporting_citations":[{"why":"established the bias-free SED construction method and the earlier z~2 result that this work extends to all z<3.","marker":"[42]"},{"why":"previous IGM-corrected EUV composite used as the baseline for the redder-EUV claim.","marker":"[31]"},{"why":"supplies the absorber population used in the Monte Carlo IGM transmission corrections.","marker":"[46]"},{"why":"SDSS composite spectrum providing the optical-to-FUV consistency check.","marker":"[29]"},{"why":"SDSS DR14 quasar catalog that provides the parent sample.","marker":"[44]"},{"why":"GALEX UV photometry, including exposure times used to define detection limits.","marker":"[45]"},{"why":"provides black hole masses and Eddington ratios used to test mass/Eddington independence and to build model-predicted composites.","marker":"[47]"},{"why":"introduces the simply truncated disk model that the paper favors.","marker":"[43]"},{"why":"standard thin disk model used as the rejected alternative.","marker":"[7]"},{"why":"shows the non-detection bias in EUV colors and supplies the exponentially modified Gaussian distribution used for mean/median estimation.","marker":"[41]"}],"fun_headline_variants":["Quasar SED universal from optical to 500 Å since cosmic noon","Bright quasars share one spectrum from optical to extreme UV","Universal quasar spectrum rules out thin disk model","One quasar spectrum fits all redshifts and black hole masses","Quasar spectra identical regardless of mass, accretion, and redshift"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on the assumption that the simulated intergalactic absorbing gas used to correct the ultraviolet measurements is representative of the real gas along the lines of sight to these quasars; if the real gas differs, the apparent sameness of the SEDs could be created by the correction rather than by the quasars.","fun_headline_variants_meta":{"raw":{"variants":["Quasar SED universal from optical to 500 Å since cosmic noon","Bright quasars share one spectrum from optical to extreme UV","Universal quasar spectrum rules out thin disk model","One quasar spectrum fits all redshifts and black hole masses","Quasar spectra identical regardless of mass, accretion, and redshift"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000712,"raw_usage":{"total_tokens":3284,"prompt_tokens":1107,"completion_tokens":2177,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":723,"completion_tokens_details":{"reasoning_tokens":2094}},"tokens_in":723,"tokens_out":2177,"duration_ms":17503,"temperature":1.0,"reasoning_tokens":2094,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:42:01.428359+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the mean and median EUV slopes of low-redshift ($z\\lesssim0.5$) quasars spectroscopically below 912 Å, where intergalactic absorption is negligible; slopes matching $\\alpha_{\\mathrm{EUV}}\\simeq-2.7$ and $-6.8$ would support the universality, while slopes near older composites would indicate the intergalactic correction produced it.","supporting_citations":[],"review_version":1}