{"id":"e516f56e-d19d-431e-a05e-88bec389db27","arxiv_id":"1909.01400","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"Quasar X-ray and UV flux data are weak but consistent with BAO and H(z) data, and the joint data mildly favor closed spatial geometry and dynamical dark energy over flat LambdaCDM.","lead":"This paper combines quasar X-ray and ultraviolet flux measurements with baryon acoustic oscillation and Hubble parameter data to constrain dark energy and spatial curvature in six cosmological models. The quasar data alone give weak constraints, but combined with BAO and H(z) data they stay consistent with the standard flat LambdaCDM model while mildly favoring closed space and evolving dark energy.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"QSO-based constraints hinge on a universal, selection-free LX-LUV relation (eq. 12); any redshift evolution or flux-limit bias would shift the joint posteriors that the central claim rests on.","rationale":"The reader's weakest_assumption identifies the same point I would stress: the QSO analysis is calibrated from the same data it is used to constrain, through a relation that is assumed to be redshift-independent and selection-free. This is the condition that must hold for the paper's main contribution, quasar distances, to be valid. I do not see an internal inconsistency in the central derivation: Eqs. (13) and (14) follow from eq. (12), the likelihood in eq. (16) is a standard conservative treatment with an intrinsic dispersion, and the six-model comparison is clearly described. The paper is appropriately hedged; the abstract's 'mildly favors' matches the 1-2.6 sigma significances in Sec. 5.3. The main risk is external rather than formal: an unrecognized redshift evolution or Malmquist-type selection bias in the quasar sample would bias the nuisance parameters beta, gamma, delta and propagate into the joint constraints. Because the QSO data have low weight, this may not overturn the H(z)+BAO-driven preference for closed geometry and dynamical dark energy, but it would undermine the claim that QSO data are a valid supplementary probe and could shift the combined best fits. I therefore agree with the conditional verdict rather than moving it. The proposed split-sample and redshift-evolution test is a feasible check using the same catalog and pipeline.","tokens_in":20912,"tokens_out":14346,"duration_ms":144677,"concrete_test":"Re-run the QSO-only and joint fits with a redshift-dependent LX-LUV relation, e.g. beta -> beta0 + beta1 log10(1+z), and also split the sample at z ~ 1 (or into quartiles) with beta and gamma free in each bin. Check whether beta1 is consistent with zero at 1 sigma and whether the marginalized posteriors of Omega_m0, Omega_k0, and omega_X (or alpha) shift by more than about 1 sigma relative to the baseline. If the relation parameters are redshift-independent and the cosmological posteriors are stable across the split, the universal-relation assumption is supported; if not, the QSO contribution to the central claim is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the QSO X-ray/UV measurements provide a valid supplementary cosmological probe depends on the LX-LUV relation in eq. (12) being universal with constant beta and gamma over 0.061 <= z <= 6.28 and on the 808-point sample being free of flux-limit selection effects that correlate apparent X-ray and UV luminosities. Sec. 5.2 concedes the analysis is 'based on the assumed validity' of this relation. If beta or gamma evolves with redshift, or if the sample selection correlates LX and LUV at fixed distance, then the distances inferred through eqs. (13)-(14) are biased. Because beta, gamma, and delta are fitted simultaneously with the cosmological parameters, part of this bias can be absorbed by the nuisance parameters and leak into the QSO+H(z)+BAO posteriors reported in Table 5 and Figs. 1-12. The fitted dispersion delta=0.32 and reduced chi-squared ~0.6-0.7 show that much of the scatter is absorbed by delta rather than independently validating the relation. Since the joint constraints, while mostly driven by H(z)+BAO, are the paper's headline, this untested assumption is the load-bearing point.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses the 808-point Risaliti & Lusso (2015) compilation of quasar X-ray and UV flux measurements, alone and combined with 11 BAO and 31 H(z) measurements, to constrain cosmological parameters in six models: spatially flat and non-flat versions of ΛCDM, the XCDM parametrization, and the φCDM scalar-field model. The analysis simultaneously fits the LX-LUV intercept β, slope γ, and intrinsic dispersion δ, uses MCMC with two Gaussian H0 priors, and reports best-fit parameters, marginalized constraints, contour plots, and AIC/BIC values. The paper finds that the QSO data alone give weak but mostly consistent constraints compared with H(z)+BAO, and that the joint data are consistent with flat ΛCDM while mildly favoring closed spatial hypersurfaces and dynamical dark energy.","tokens_in":21182,"tokens_out":7450,"duration_ms":75889,"significance":"If the QSO distance estimate is valid, the paper provides a useful demonstration that X-ray/UV quasar fluxes can act as a high-redshift supplement to BAO and H(z) data, extending earlier work by Risaliti & Lusso (2015) to six cosmological models. The paper is methodologically careful in several respects: it uses the covariance matrix for the correlated BAO points, it reports AIC and BIC, it checks two H0 priors, and it explicitly fits the QSO nuisance parameters rather than presenting them as fixed predictions. The main limitation is that the central probe is calibrated through an assumed universal LX-LUV relation whose redshift independence and selection properties are not tested; this leaves the headline joint constraints vulnerable to systematic bias. The cosmological conclusions are not decisive, as the paper acknowledges, but the consistency check is a legitimate and useful step.","major_comments":[{"comment":"The analysis assumes that the LX-LUV relation in Eq. (12) has constant β and γ over the full range 0.061 ≤ z ≤ 6.28, with all remaining scatter captured by the single fitted dispersion δ, and Sec. 5.2 states that the QSO analysis is 'based on the assumed validity' of this relation. Because Eq. (13) converts the flux-flux relation into distance information that enters the joint likelihood, any redshift evolution of γ or β, or any flux-limit selection effect that correlates FX and FUV at fixed DL, would bias the QSO-only and QSO+H(z)+BAO posteriors reported in Tables 4-5 and Figs. 1-12. Please add quantitative tests, for example splitting the sample into redshift bins, fitting a redshift-dependent slope γ(z) = γ0 + γ1 log(1+z), or comparing high- and low-redshift marginalised constraints, and discuss the selection-function literature for this sample. Without such tests, the claim that the QSO data are a valid supplementary cosmological probe is not fully established.","section":"Secs. 3, 4, and 5.2, Eq. (12)"},{"comment":"The abstract's statement that the joint data 'mildly favor' closed spatial hypersurfaces and dynamical dark energy should be reconciled with the information criteria reported in Tables 1-2. For the H0 = 68 prior, flat ΛCDM has the lowest AIC among all joint fits (507.01 in Table 1, versus 508.65-508.73 for the other models), while for the H0 = 73.24 prior, non-flat ΛCDM has the lowest AIC (509.85) but several models are within ΔAIC ≲ 3.5 of it. Since differences of only a few units in AIC/BIC are not significant, the 'favors' language is somewhat stronger than the model-selection evidence; I recommend either removing the claim or restating it as parameter-interval shifts within 1σ-2.6σ for individual models, with an explicit look-elsewhere caveat for the 12 model/prior combinations.","section":"Sec. 5.3, Tables 1-2"}],"minor_comments":[{"comment":"The text 'For a newer compilation of QSO data see ?' contains a missing citation; the reference clearly should be Risaliti & Lusso (2019), which is listed in the bibliography, and this should be completed.","section":"Sec. 3"},{"comment":"There are several typographical errors: 'Plank Collaboration' should be 'Planck Collaboration'; Fig. 8 caption has 'Left pnnel'; Fig. 12 caption has 'These plots are the for H0'; and the surname 'López-Corredoira' is missing an 'o' in Sec. 3. These should be corrected.","section":"Secs. 1, 5.2, and figure captions"},{"comment":"The prior range for α is stated as 0 ≤ α ≤ 3, with '(0 ≤ α ≤ 1.2 for QSO only)'. This prior inconsistency affects the QSO-only φCDM constraints in Table 4 and makes comparisons with the joint fits less direct; it should be justified or removed.","section":"Sec. 4"},{"comment":"Because the QSO likelihood in Eq. (16) includes the ln(2πs_i^2) term, the quantity called χ²_min = -2 ln(LFmax) in Tables 1-2 is a deviance rather than a standard chi-square statistic. The reported 'reduced χ²' values of 0.6-0.7 are therefore not directly comparable to ordinary reduced chi-squares, and this should be stated explicitly to avoid misleading readers.","section":"Secs. 5.1-5.3, Eq. (16)"},{"comment":"The QSO-only H0 posteriors are essentially identical to the priors (68 ± 2.8 and 73.24 ± 1.73 km s^-1 Mpc^-1), so the text in Sec. 5.2 that QSO constraints are insensitive to the H0 prior should be phrased as 'the QSO data do not constrain H0.'","section":"Table 4"}],"recommendation":"major_revision","confidential_remarks":"This is a competent and clearly written application of a standard pipeline, and the central inconsistency with the literature is not the analysis itself but the untested universality of the LX-LUV relation. The requested calibration tests are feasible within the scope of a revision. The model-selection language in the abstract should also be softened to match the AIC/BIC results. The paper fits the journal's scope; I would be supportive after these revisions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a competent and honest constraints paper that does exactly what it says: it takes the Risaliti & Lusso (2015) 808-object X-ray/UV quasar sample and runs it through six cosmological models, alone and jointly with H(z)+BAO, under two H0 priors. That is a genuine extension of the original non-flat ΛCDM analysis to flat and non-flat XCDM and φCDM. The methodology is standard and well described: MCMC with emcee, proper covariance treatment for the first six BAO points, and AIC/BIC reporting. The authors reproduce earlier results, which is a good sign.\n\nThe headline claim is appropriately modest: QSO data alone give weak constraints, they are consistent with H(z)+BAO, and the joint sample is consistent with flat ΛCDM while mildly preferring closed spatial hypersurfaces and dynamical dark energy. That preference is at 1–2.6σ depending on model and H0 prior, so it is not a discovery claim.\n\nWhere the paper is softest is the pillar it stands on: the assumed universality of the LX–LUV relation in eq. (12), with constant slope and intercept over 0.061 ≤ z ≤ 6.28 and no selection effects. The authors acknowledge this in Sec. 5.2, saying the analysis is “based on the assumed validity” of the relation. But the fitted scatter δ = 0.32 and reduced χ² of 0.6–0.7 show that the relation is not being sharply tested by these data; the nuisance parameters absorb a lot. If the relation evolves with redshift, or the sample is flux-selected in a way that correlates X-ray and UV fluxes, the distance estimates would be biased and the joint posteriors could shift. That matters for interpreting the mild preferences, even though H(z)+BAO do most of the constraining work.\n\nSmaller issues: there is a citation placeholder in Sec. 3 (“For a newer compilation of QSO data see ?”), which should have been caught, and no code is shipped. Neither is fatal.\n\nI would send this to peer review rather than desk-reject. It is a competent extension in an established program, and having these constraints on record helps calibrate the QSO probe. The right referee request is a deeper discussion or test of the LX–LUV assumptions, not rejection.","headline":"A careful and honest constraints paper extending the 2015 quasar sample to six cosmological models; the cosmological preferences are weak and the QSO relation's universality is the main caveat, but the analysis is sound and worth citing as a data-validation step.","tokens_in":21711,"tokens_out":3291,"would_cite":false,"duration_ms":30189,"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 joint analysis of quasar X-ray/UV fluxes, Hubble-parameter measurements, and baryon acoustic oscillations is consistent with flat $\\Lambda$CDM but mildly favors a closed universe with dynamical dark energy.","keywords":["quasar cosmology","X-ray and UV luminosity relation","dark energy","spatial curvature","baryon acoustic oscillations","Hubble parameter","LambdaCDM","Markov chain Monte Carlo"],"falsifier":"Fitting the $L_X$--$L_{UV}$ relation separately in narrow redshift bins and finding that $\\beta$ or $\\gamma$ drifts by more than the quoted uncertainties would falsify the universal-relation assumption; alternatively, re-running the joint analysis with the newer 1598-quasar compilation and finding that $\\Omega_{k0}$ and $\\omega_X$ move back to $0$ and $-1$ respectively would show that the mild preference was a statistical fluctuation.","tokens_in":20667,"feed_emoji":"🔭","tokens_out":8817,"duration_ms":70067,"temperature":0.7,"pith_summary":"This paper asks whether quasar X-ray and ultraviolet flux measurements can serve as a cosmological distance probe, and what they say about the composition and geometry of the universe when combined with more standard data. Using 808 quasars spanning redshifts $0.061$ to $6.28$ together with 31 Hubble-parameter measurements and 11 baryon acoustic oscillation measurements, the authors constrain six cosmological models: flat and curved versions of $\\Lambda$CDM, the XCDM dark-energy parametrization, and the $\\phi$CDM scalar-field model. The quasar data alone give much weaker constraints than the $H(z)$+BAO data, but they agree with them. In the joint analysis the data remain consistent with the standard flat $\\Lambda$CDM model, while mildly favoring closed spatial hypersurfaces and dark energy whose density changes with time. If right, this establishes quasars as a high-redshift distance probe and gives a concrete target for future cosmological observations.","feed_headline":"Quasar data mildly favor curved space and evolving dark energy","feed_subtitle":"Combining 808 quasar fluxes with BAO and Hubble data keeps flat LambdaCDM viable but nudges toward closed geometry.","key_machinery":"The load-bearing object is the X-ray-to-UV luminosity relation of quasars, $\\log(L_X)=\\beta+\\gamma\\log(L_{UV})$, rewritten in terms of fluxes as $\\log(F_X)=\\beta+(\\gamma-1)\\log(4\\pi)+\\gamma\\log(F_{UV})+2(\\gamma-1)\\log(D_L)$. This turns each quasar's measured UV and X-ray fluxes into a distance indicator through the luminosity distance $D_L(z,p)$, with $\\beta$, $\\gamma$, and a global intrinsic dispersion $\\delta$ fitted together with the cosmological parameters. The likelihoods for the 31 $H(z)$ and 11 BAO measurements, including the correlated BAO points via the covariance matrix, are combined with the quasar likelihood, and the parameter space is explored with a Markov chain Monte Carlo, with model comparison via AIC and BIC. The $L_X$--$L_{UV}$ relation is what lets the quasar sample, reaching $z\\simeq6.28$, act as a cosmological probe.","core_discovery":"The paper's central claim is that the combined QSO + $H(z)$ + BAO data are consistent with the currently standard spatially-flat $\\Lambda$CDM model, but mildly favor a closed universe and dynamical dark energy. On its own, the quasar sample, calibrated through the assumed relation $\\log(L_X)=\\beta+\\gamma\\log(L_{UV})$, constrains cosmological parameters only loosely: in the non-flat $\\Lambda$CDM model it yields $\\Omega_{m0}=0.24^{+0.16}_{-0.10}$ and $\\Omega_\\Lambda=0.93^{+0.18}_{-0.39}$. Adding the quasars to the $H(z)$+BAO data tightens the constraints and, in several models, pushes the best fit away from the flat-$\\Lambda$ baseline: the curvature parameter is negative (closed) in most cases, reaching $\\Omega_{k0}=-0.22^{+0.09}_{-0.13}$ in the non-flat $\\phi$CDM model with the high local Hubble-constant prior, and six of the eight dynamical-dark-energy cases prefer evolving dark energy over a cosmological constant at 1.3 to 2.6 $\\sigma$. The strength of these preferences depends on the assumed prior on $H_0$.","pith_inferences":["If the same $L_X$--$L_{UV}$ relation is allowed to evolve with redshift, the mild curvature and dynamical-dark-energy signal could weaken or vanish; fitting $\\beta$ and $\\gamma$ in redshift bins with the current 808-quasar sample would be a direct test.","The newer 1598-quasar compilation mentioned in the paper should settle whether the preference for closed geometry persists or was a statistical fluctuation.","The correlation between the $H_0$ prior and the strength of the non-flat or dynamical preference suggests that part of the signal may be a projection of the Hubble tension rather than independent evidence for new physics.","The consistency between QSO-only and $H(z)$+BAO contours across all six models is itself informative: it means any unmodeled systematics in the quasar relation are not dominating the joint result, though it does not rule them out."],"forward_implications":["If the joint preference is real, the quasar X-ray/UV method is a working distance probe that extends cosmological constraints to redshifts far beyond those reached by supernovae.","The mild preference for closed spatial hypersurfaces, if confirmed by future data, would mean the simplest flat $\\Lambda$CDM geometry is incomplete.","Adding quasar data to $H(z)$+BAO noticeably tightens constraints in models with more free parameters, so larger quasar compilations should sharpen cosmological parameter estimates.","The fitted slope $\\gamma\\simeq0.6$ and dispersion $\\delta\\simeq0.32$ are stable across all six models, indicating that the quasar calibration is not strongly model-dependent.","The results depend on the Hubble-constant prior: the higher local $H_0$ prior strengthens the preference for closed geometry and dynamical dark energy, linking the finding to the broader $H_0$ tension."],"supporting_citations":[{"why":"Supplies the 808-quasar X-ray/UV flux measurements and the $L_X$-$L_{UV}$ relation that turns the sample into a cosmological distance probe.","marker":"Risaliti & Lusso (2015)"},{"why":"Provides the 31 $H(z)$ and 11 BAO measurements, listed in its Tables 1 and 2, used for the standard-probe constraints.","marker":"Ryan et al. (2018)"},{"why":"Supplies the likelihood treatment and covariance matrix for the correlated BAO and $H(z)$ data that the joint analysis follows.","marker":"Ryan et al. (2019)"},{"why":"Gives the median-statistics $H_0$ prior of $68\\pm2.8$ km s$^{-1}$ Mpc$^{-1}$ used in one set of analyses.","marker":"Chen & Ratra (2011a)"},{"why":"Gives the local expansion-rate prior of $73.24\\pm1.74$ km s$^{-1}$ Mpc$^{-1}$ used in the other set of analyses.","marker":"Riess et al. (2016)"},{"why":"Defines the inverse-power-law scalar-field $\\phi$CDM model whose evolving dark energy is tested alongside $\\Lambda$CDM and XCDM.","marker":"Ratra & Peebles (1988)"},{"why":"Cited as evidence that the preference for closed spatial hypersurfaces is largely driven by the $H(z)$+BAO data.","marker":"Park & Ratra (2018d)"}],"fun_headline_variants":["Quasar data mildly favor closed universe and evolving dark energy","Combined quasar, BAO, and Hubble data hint at closed curvature","808 quasars nudge cosmology toward closed space and dynamic DE","Quasar constraints keep flat LCDM consistent, but favor closed models","Cosmological probes mildly prefer curvature and dynamical dark energy"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on the assumption that the $L_X$--$L_{UV}$ relation has the same slope and intercept at every redshift from 0.061 to 6.28, with all remaining scatter captured by one constant dispersion; if the relation evolves with redshift or the sample is biased by how quasars are selected in X-ray and UV flux, the quasar distances and the joint cosmological constraints would be biased.","fun_headline_variants_meta":{"raw":{"variants":["Quasar data mildly favor closed universe and evolving dark energy","Combined quasar, BAO, and Hubble data hint at closed curvature","808 quasars nudge cosmology toward closed space and dynamic DE","Quasar constraints keep flat LCDM consistent, but favor closed models","Cosmological probes mildly prefer curvature and dynamical dark energy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000927,"raw_usage":{"total_tokens":3970,"prompt_tokens":945,"completion_tokens":3025,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":561,"completion_tokens_details":{"reasoning_tokens":2937}},"tokens_in":561,"tokens_out":3025,"duration_ms":20958,"temperature":1.0,"reasoning_tokens":2937,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:19:11.270971+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fitting the $L_X$--$L_{UV}$ relation separately in narrow redshift bins and finding that $\\beta$ or $\\gamma$ drifts by more than the quoted uncertainties would falsify the universal-relation assumption; alternatively, re-running the joint analysis with the newer 1598-quasar compilation and finding that $\\Omega_{k0}$ and $\\omega_X$ move back to $0$ and $-1$ respectively would show that the mild preference was a statistical fluctuation.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the 808-quasar X-ray/UV flux measurements and the $L_X$-$L_{UV}$ relation that turns the sample into a cosmological distance probe."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the 31 $H(z)$ and 11 BAO measurements, listed in its Tables 1 and 2, used for the standard-probe constraints."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the likelihood treatment and covariance matrix for the correlated BAO and $H(z)$ data that the joint analysis follows."},{"cited_title":"G., et al., 2016, ApJ, 826, 56","cited_arxiv_id":null,"evidence_quote":"Gives the local expansion-rate prior of $73.24\\pm1.74$ km s$^{-1}$ Mpc$^{-1}$ used in the other set of analyses."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the inverse-power-law scalar-field $\\phi$CDM model whose evolving dark energy is tested alongside $\\Lambda$CDM and XCDM."}],"review_version":1}