{"id":"496e99a8-4705-44e0-bf72-f9847d284569","arxiv_id":"2504.13583","paper_version":1,"verdict":"REJECT","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":4,"one_line_summary":"A forecast that SKA 21-cm redshift drift observations could constrain CPL dark-energy parameters, but the claimed sub-cm/s constraints are not actual parameter errors because the paper reports signal derivatives instead of propagated uncertainties.","lead":"This paper forecasts how precisely the Square Kilometre Array could measure cosmic acceleration by watching redshift drift in hydrogen 21-cm lines over half a year. It compares two analysis methods and claims millimeter-per-second parameter constraints, but the quoted numbers are signal derivatives rather than true parameter uncertainties.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper equates the derivative ∂Δv/∂p with the parameter uncertainty σp; since σp requires dividing by that derivative (or a Fisher inverse), Tables 1–2 report sensitivities in cm/s, not parameter constraints.","rationale":"The reader identified the same load-bearing weakness: the paper replaces error propagation with the bare derivative. My independent check of the text confirms that Sections 2 and 4 consistently use ∂(Δv)/∂p as the 'precision' and that no covariance or division by σ_v appears anywhere. The negative range for σ(wa) in Table 2 is direct evidence that the quoted values are sensitivities, not standard deviations. This is not a matter of disagreeing with an established convention; standard Fisher forecasting is unambiguous, and the paper cites no alternative formalism. Because the central claim is built on this category error, the quantitative conclusions about sub-cm/s parameter constraints are unsupported. The signal-amplitude and derivative calculations themselves are standard and could be reframed as sensitivity forecasts, but they do not justify the stated 'constraints' in the abstract and Section 4. I therefore support the reader's REJECT verdict; no correction of the paper's internal equations would change this without a full re-analysis of the forecast.","tokens_in":14315,"tokens_out":5309,"duration_ms":46505,"concrete_test":"Compute σ_Ωm = σ_v / |∂Δv/∂Ωm| at z=0.5 for the canonical, 0.001 Hz dataset, using σ_v from the paper's own Fig. 3 range (~0.005–0.035 cm/s) and ∂Δv/∂Ωm from Eq. (8) at flat CPL fiducial h=0.7, Ωm=0.3, w0=-1, wa=0. Compare with Table 1's σm interval [0.08,0.48] cm/s. If σ_Ωm is a dimensionless parameter uncertainty and does not match that interval, the tabulated values are not parameter constraints. Also compute σ_wa for Table 2 to show its negative lower bound cannot result from standard error propagation.","verdict_should_be":"REJECT","load_bearing_attack":"The central quantitative claim—that SKA HI 21-cm redshift drift can constrain Ωm, w0, wa to mm/s-level or better—rests on the identification, stated in Sections 2 and 4, that 'the precision of the parameter is characterized by the partial derivative...' and that σp is ∂(Δv)/∂p. This identification is dimensionally and statistically wrong. For a measured velocity drift with uncertainty σ_v, a single-parameter forecast is σ_p = σ_v / |∂(Δv)/∂p|, and for joint constraints one needs the full covariance (or Fisher) matrix, including correlation among parameters, plus marginalization over h and calibration/nuisance terms. Equations (7)–(10) provide only derivatives; they never divide by σ_v nor invert a Fisher matrix. Table 2 even lists σ(wa) as [-0.01, 0.14] cm/s, which is impossible for a standard deviation and confirms the tabulated quantities are not uncertainties. The paper can be salvaged as a sensitivity analysis—showing the magnitude of ∂(Δv)/∂p in cm/s per unit parameter—but not as a parameter-constraint forecast, so the abstract's 'accuracy reaching the level of millimeter per second' for parameters does not follow.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes to estimate the precision with which the SKA HI 21-cm redshift drift signal can constrain cosmological parameters in a flat CPL model (parameters h, Ωm, w0, wa). Two observation strategies are considered: the canonical redshift drift and the differential redshift drift, using spectral resolutions of 0.001 Hz and 0.002 Hz over ΔT = 0.5 yr. The authors compute the velocity drift signal and its partial derivatives with respect to the model parameters, report these derivatives as 'parameter precision' values in cm/s, and conclude that both methods achieve mm/s-level parameter constraints, with the differential method being superior for simultaneously constraining Ωm and the dark-energy equation-of-state parameters. The paper also discusses systematic effects and target selection criteria.","tokens_in":14549,"tokens_out":4304,"duration_ms":38717,"significance":"If the central result were valid, it would be a significant step toward establishing SKA HI 21-cm redshift drift as a competitive, model-independent probe of cosmic acceleration at z<1. The paper correctly presents the standard redshift drift equations and illustrates the differential drift signal amplitude. However, the load-bearing claim—that the partial derivative ∂(Δv)/∂p constitutes the precision of the parameter p—is statistically incorrect, and this flaw invalidates essentially all quantitative results in Section 4. As a sensitivity analysis of the drift signal's parameter dependence, the paper could have some pedagogical value, but it does not deliver a parameter-constraint forecast. No code or machine-checkable derivations are provided.","major_comments":[{"comment":"The paper identifies the 'precision of the parameter' σp with the partial derivative ∂(Δv)/∂p and reports it in cm/s. This is not a parameter uncertainty under standard error propagation: for a single parameter the correct expression is σp = σv / |∂Δv/∂p|, and for joint constraints one must invert the Fisher information matrix to account for parameter correlations and marginalization. The manuscript never divides by σv and never constructs a covariance matrix. Consequently, Figures 4–6 and Tables 1–2 report signal sensitivities (in cm/s per unit parameter), not uncertainties on Ωm, w0, wa. The abstract's claim of 'accuracy reaching the level of millimeter per second' for cosmological parameters is therefore unsupported.","section":"§1 and §4, Eqs. (7)–(10), Tables 1–2"},{"comment":"Table 2 lists σ(ωa) with a negative lower bound (-0.01 cm/s for the 0.001 Hz case). A standard deviation cannot be negative, which confirms that the tabulated 'precision metrics' are not uncertainties in the usual statistical sense. This is a direct empirical indicator that the reported quantities are derivative amplitudes rather than parameter errors.","section":"Table 2, rows for 0.001 Hz and 0.002 Hz"},{"comment":"The noise model σv = σn N^{-1/2} (1+z)^λ ΔT^{-1/2} relies on an ad hoc normalization σn and a spectral-resolution index λ that are fitted to a small set of 'actual measurements' without a derivation or an error budget. Even if the statistical methodology were corrected, the forecasted constraints would scale linearly with σv, so the unvalidated noise model is a load-bearing assumption. The authors should justify σn and λ from an instrumental noise model or from established SKA forecasting literature.","section":"§2, Eq. (11) and Figure 2"}],"minor_comments":[{"comment":"The manuscript contains numerous typographical and grammatical errors, including 'indenpendent' in the abstract, 'mesurement' in Section 5, and 'T able 1/2' in the captions. A thorough language edit is needed.","section":"Throughout"},{"comment":"Equations (7)–(10) give derivatives of the dimensionless quantity Sz, not of the velocity drift Δv, but Section 4 refers to ∂(Δv)/∂p. The connection between these quantities and the quoted cm/s values is not explicitly made, which contributes to the unit confusion in the reported 'precisions'.","section":"§2, Eqs. (7)–(10) vs. §4"},{"comment":"The color bars in Figures 4–6 are labeled only with numeric values; their units (presumably cm/s) are not stated on the color bars themselves. The label 'n' for σn in the left panels is also ambiguous.","section":"Figures 4–6"},{"comment":"The references list three Cooke entries (2019, 2020a, 2020b) that appear to correspond to the same paper; the duplicated entries should be merged, and the in-text citations should be checked for consistency.","section":"References"}],"recommendation":"reject","confidential_remarks":"The fundamental issue is not a presentation weakness but a methodological error that invalidates the central quantitative claims. Recomputing the constraints with a proper Fisher or Monte Carlo forecast would likely change the conclusions qualitatively, as the derivatives are of order 0.1–1 cm/s per unit parameter while σv is of order 0.01–0.3 cm/s, so the resulting parameter uncertainties would be of order unity or larger. In my view this cannot be remedied by modest revisions within the scope of the current manuscript."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis one is a clear case of a good idea undermined by a bad statistical step. The paper takes standard redshift drift equations, applies them to SKA HI 21-cm observations, and claims mm/s-level constraints on cosmological parameters. That claim does not survive contact with the error budget.\n\nWhat the paper does well: the opening sections lay out the canonical and differential redshift drift formulas cleanly, with the CPL parameterization, and they correctly note that the differential method can give a slightly larger signal. The noise model in Eq. 11 is also a reasonable attempt to connect SKA spectral resolution to velocity uncertainty, and Figure 3 is a genuinely useful illustration of how σ_v scales with redshift, time, and sample size. If the paper were framed purely as a sensitivity analysis—showing how much the drift signal responds to each parameter—it would be a decent contribution.\n\nThe problem is that the paper steps over the line from sensitivity to constraints. In Section 2 and again in Section 4, they define the parameter precision σ_p as the partial derivative ∂(Δv)/∂p, and report it in cm/s. That is dimensionally not an uncertainty. For a single parameter, a forecast is σ_p = σ_v / |∂(Δv)/∂p|, and for joint constraints you need a covariance or Fisher matrix, which they never construct. The tables and figures therefore report derivatives, not constraints. The smoking gun is Table 2: the range for σ(wa) includes negative values, which is impossible for a standard deviation. The paper's own numbers betray the confusion.\n\nThis is not a minor flaw; it is load-bearing. The abstract's promise of mm/s-level parameter constraints is the entire point, and it is not supported. The paper could be salvaged as a sensitivity study, but as written it does not deliver what it claims.\n\nThere is also a smaller issue worth mentioning: the noise model's normalization constant σ_n is fitted from a handful of simulated points, and the paper does not discuss priors or marginalization over nuisance parameters like h, which would degrade any real forecast.\n\nWho is this for? Someone working on redshift drift forecasts might find the derivative plots a useful starting point, but they should not cite the constraint numbers. It deserves a serious referee because the error is instructive and the paper touches on an important observational target, but the referee should be told to focus on the confusion between derivative and uncertainty. With that fixed, a revised version could be a worthwhile contribution.\n\nMy recommendation: send it to peer review only if the referee is willing to work through the statistics; otherwise it would be a desk reject. As is, I would not cite the claimed constraints.","headline":"The paper's central forecast is invalid: it reports derivatives as parameter uncertainties, so the headline mm/s constraints do not follow.","tokens_in":15177,"tokens_out":1673,"would_cite":false,"duration_ms":17988,"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":"The paper claims that half a year of SKA HI 21-cm redshift-drift observations can constrain cosmological parameters to sub-cm/s precision, with the differential drift method outperforming the canonical method for dark-energy parameters.","keywords":["redshift drift","Sandage-Loeb effect","HI 21-cm line","Square Kilometre Array","dark energy equation of state","cosmological parameter constraints","CPL parametrization","differential redshift drift"],"falsifier":"Compute $\\sigma_{\\Omega_m}$, $\\sigma_{w_0}$, and $\\sigma_{w_a}$ using standard error propagation $\\sigma_p = \\sigma_v / |\\partial(\\Delta v)/\\partial p|$ with $\\sigma_v$ from the paper's equation (11); if, at the fiducial redshift and for the 0.001 Hz dataset, the resulting dimensionless uncertainties (e.g., $\\sigma_{\\Omega_m} \\approx 0.2$ rather than 0.2 cm/s, or $\\sigma_{w_0}$ exceeding 1) do not reproduce the paper's sub-cm/s claim, the headline precision numbers fail.","tokens_in":14069,"feed_emoji":"📡","tokens_out":5597,"duration_ms":44121,"temperature":0.7,"pith_summary":"This paper tries to establish that redshift-drift observations of the HI 21-cm line with the Square Kilometre Array, lasting only half a year at spectral resolutions of 0.001 Hz or 0.002 Hz, can constrain cosmological parameters to sub-cm/s precision. It compares two ways of using the drift signal: the canonical redshift drift relative to the present epoch, and the differential redshift drift between two non-zero redshifts, and it evaluates both under the Chevallier–Polarski–Linder dark-energy parametrization. The authors argue that both methods deliver parameter accuracies at the millimeter-per-second level or better, that the differential method is the stronger of the two, and that it is especially advantageous when constraining the matter density and the dark-energy equation-of-state parameters together. If correct, this would make real-time cosmology feasible and would establish HI 21-cm redshift drift as a competitive independent probe of cosmic acceleration at $z<1$.","feed_headline":"SKA redshift drift could pin cosmology to mm/s","feed_subtitle":"Two drift methods, one differential, are forecast to constrain dark energy below 1 cm/s in just 0.5 year.","key_machinery":"The central objects are the two redshift-drift observables: the canonical velocity drift $\\Delta v = (cH_0\\Delta t)\\left(1 - E(z)/(1+z)\\right)$ relative to the present epoch, and the differential drift $\\Delta v_{\\mathrm{ir}} = (cH_0\\Delta t)\\left(E(z_r)/(1+z_r) - E(z_i)/(1+z_i)\\right)$ between a reference and an intervening source. These are evaluated under the CPL Hubble parameter $E^2(z)=\\Omega_m(1+z)^3 + \\Omega_\\phi(1+z)^{3(1+w_0+w_a)}e^{-3w_a z/(1+z)}$. The workhorse of the analysis is the derivative-based precision metric $\\sigma_p = \\partial(\\Delta v)/\\partial p$, whose value in cm/s is read as the parameter's uncertainty, combined with the velocity-noise model $\\sigma_v = \\sigma_n N^{-1/2}(1+z)\\lambda \\Delta T^{-1/2}$ for SKA source counts.","core_discovery":"The central claim is that the precision of a cosmological parameter obtainable from redshift drift is characterized by the partial derivative of the velocity drift with respect to that parameter, $\\partial(\\Delta v)/\\partial p$, and that using this measure the SKA HI 21-cm data with 0.001 Hz and 0.002 Hz spectral resolution over $\\Delta T=0.5$ yr yield parameter precisions below 1 cm/s. The canonical method gives ranges such as $\\sigma_{\\Omega_m}\\approx 0.08$–$0.5$ cm/s and $\\sigma_{w_0}\\approx 0.2$–$1$ cm/s, while the differential method keeps all parameter precisions below 0.5 cm/s and gives the tightest bounds on $w_a$, with values near zero to 0.2 cm/s. The authors conclude that the differential redshift drift is the preferred technique when simultaneously constraining $\\Omega_m$ and the dark-energy equation-of-state parameters $w_0$ and $w_a$, while the canonical method remains competitive for $\\Omega_m$ alone.","pith_inferences":["If the derivative-based 'precision' is converted to a conventional parameter uncertainty via $\\sigma_p = \\sigma_v / |\\partial(\\Delta v)/\\partial p|$, the quoted cm/s numbers become sensitivities rather than final uncertainties; for parameters whose derivative is small, the actual uncertainty could be much larger than 1 cm/s, so the headline precision claim would need renormalization.","A natural next step would be a Fisher-matrix forecast that combines the drift signal across many redshift bins and includes the full covariance between parameters; that would show whether the claimed advantages of the differential method survive into a standard confidence ellipse in the ($\\Omega_m$, $w_0$, $w_a$) plane.","The same derivative machinery could be applied to other redshift-drift probes, such as the Lyman-$\\alpha$ forest at $z>2$, to compare forecast precision across facilities without changing the underlying formalism."],"forward_implications":["SKA HI 21-cm observations at 0.001 Hz spectral resolution over 0.5 yr should measure redshift-drift velocities with $\\sigma_v$ between about 0.005 and 0.15 cm/s out to $z=1$, enough to detect the drift signal.","Under the paper's precision metric, every CPL parameter ($h$, $\\Omega_m$, $w_0$, $w_a$) can be constrained below 1 cm/s, with the differential method keeping all values below 0.5 cm/s.","The differential redshift drift method is forecast to outperform the canonical method for combined constraints on $\\Omega_m$ and dark-energy equation-of-state parameters, while the canonical method suffices for $\\Omega_m$ alone.","Higher spectral resolution (0.001 Hz vs 0.002 Hz) tightens the forecast constraints for both methods."],"supporting_citations":[{"why":"Introduces the redshift drift effect as an observational probe of cosmic expansion.","marker":"Sandage 1962a"},{"why":"Establishes the Sandage–Loeb effect and its detection requirements.","marker":"Loeb 1998"},{"why":"Supplies the SKA HI 21-cm source-count and spectral-resolution assumptions, including $N=10^7$ per 0.1 redshift and the 0.001–0.01 Hz range.","marker":"Kloeckner et al. 2015"},{"why":"Defines the CPL dark-energy parametrization used for $E(z)$.","marker":"Chevallier & Polarski 2001"},{"why":"Provides the same CPL parametrization and its equation-of-state form.","marker":"Linder 2003"},{"why":"Introduces the differential redshift drift method between two non-zero redshifts.","marker":"Cooke 2020a"},{"why":"Provides the differential drift expression and the dimensionless redshift-drift formalism.","marker":"Esteves et al. 2021"},{"why":"Supplies forecasts for HI 21-cm redshift drift detectability and the $\\sigma_v$ scaling.","marker":"Alves et al. 2019"}],"fun_headline_variants":["Differential redshift drift tightens dark energy bounds","SKA's 21-cm drift could hit mm/s precision","Canonical vs differential drift: both precise, one better","Two drift techniques forecast sub-cm/s cosmology","Redshift drift with SKA: differential method excels"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis assumes that the partial derivative of the redshift-drift velocity with respect to a parameter, measured in cm/s per unit of that parameter, can be read directly as the uncertainty on that parameter, without dividing by that derivative to convert velocity error into parameter error.","fun_headline_variants_meta":{"raw":{"variants":["Differential redshift drift tightens dark energy bounds","SKA's 21-cm drift could hit mm/s precision","Canonical vs differential drift: both precise, one better","Two drift techniques forecast sub-cm/s cosmology","Redshift drift with SKA: differential method excels"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000268,"raw_usage":{"total_tokens":1670,"prompt_tokens":1049,"completion_tokens":621,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":665,"completion_tokens_details":{"reasoning_tokens":542}},"tokens_in":665,"tokens_out":621,"duration_ms":5760,"temperature":1.0,"reasoning_tokens":542,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T12:05:47.929302+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute $\\sigma_{\\Omega_m}$, $\\sigma_{w_0}$, and $\\sigma_{w_a}$ using standard error propagation $\\sigma_p = \\sigma_v / |\\partial(\\Delta v)/\\partial p|$ with $\\sigma_v$ from the paper's equation (11); if, at the fiducial redshift and for the 0.001 Hz dataset, the resulting dimensionless uncertainties (e.g., $\\sigma_{\\Omega_m} \\approx 0.2$ rather than 0.2 cm/s, or $\\sigma_{w_0}$ exceeding 1) do not reproduce the paper's sub-cm/s claim, the headline precision numbers fail.","supporting_citations":[],"review_version":1}