{"id":"0395603c-7675-4441-b74e-770e6ce9b00a","arxiv_id":"2507.21628","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"high","formal_verification":"none","parameter_count":7,"one_line_summary":"A spatio-temporal Gaussian random field framework generates time-resolved images of a rotating Hayward black hole, with the M87* ring-shift reproduction asserted but not quantitatively demonstrated.","lead":"This paper models time-dependent black hole images by adding stochastic brightness fluctuations from spatio-temporal Gaussian random fields to a rotating Hayward black hole accretion disk. It claims this is a fast surrogate for GRMHD simulations and that it reproduces the bright ring shift seen in M87*, although the quantitative support for that last claim is not actually shown.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The M87* ring-shift claim lacks a direct quantitative comparison: the paper never measures the simulated bright-ring position angle over time and never compares it to the EHT-measured 30-degree shift.","rationale":"The reader's weakest assumption is that the hand-set stochastic parameterization faithfully represents the M87* accretion flow, and I agree that the mapping from GRF parameters to physical turbulence is asserted rather than demonstrated. The most load-bearing specific consequence of that assumption is the headline M87* ring-shift claim, and it fails on a narrower, more directly checkable ground: the paper never actually performs the comparison it claims. Section 4 mentions the EHT's ~30-degree shift but gives no simulated position-angle time series, no uncertainty, and no figure or statistic showing the simulated ring's brightest sector moving by ~30 degrees. This is an internal gap in the argument, not a disagreement with consensus: the claim 'the simulation reproduces observed X' is unsupported when X is never measured in the simulation. The missing comparison is concrete and addable: extract a position-angle time series from the generated frames, quantify drift, and compare to EHT. The paper's framework (Matérn fields, anisotropic tensor, slow-light ray tracing, regular Hayward metric) is internally plausible, and the GRMHD consistency statements are qualitative but not obviously wrong, so REJECT would be too strong. CONDITIONAL accurately reflects that the central validation needs to be supplied. I find no need to move the verdict; the concern is the same as the reader's, sharpened to the specific missing quantitative comparison.","tokens_in":15771,"tokens_out":1671,"duration_ms":17864,"concrete_test":"Compute the azimuthal position angle of the brightest pixel (or a centroid of pixels above a brightness threshold) of the bright ring in each output frame of the slow-light simulation at theta0=17 deg, fit a linear (or piecewise) drift to the position-angle time series with uncertainties from multiple INOISY realizations, and compare the simulated drift against the EHT 2017-to-2018 measurement (approximately 30 degrees clockwise). If the simulated position-angle shift is not consistent with the observed shift within the quoted observational uncertainty, the M87* reproduction claim in the abstract must be withdrawn or rephrased as a qualitative consistency statement.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central validation claim in the abstract is that the simulations 'reproduce the dynamic positional shift of the bright ring structure observed in M87*'. But the paper never performs that comparison. Section 4 (Conclusion and Discussion) simply states that the 2018 EHT images showed a ~30-degree clockwise shift and asserts that the simulations provide new insights. No position-angle time series of the simulated bright ring is defined, computed, or plotted; no error bar or uncertainty estimate is attached to the simulated shift; and no comparison to the EHT 2017/2018 position-angle measurements appears anywhere. Without that quantitative step, the claim that the model reproduces the observed shift is unsupported: any stochastic model with a rotating asymmetric brightness pattern will produce some azimuthal drift, and a hand-tuned pitch angle of 20 degrees plus Keplerian rotation makes a clockwise drift essentially by construction. The load-bearing assumption is therefore not merely that the Gaussian fluctuation field is a faithful surrogate for turbulent accretion, but more specifically that the field's parameters (lambda1/r = 5, lambda2/lambda1 = 0.1, lambda0 = 2 pi/Omega_K, theta_angle = 20 deg, envelope g(r)=r^4 e^{-r^2}) produce a bright-ring position angle that tracks the observed ~30-degree shift. This is asserted in Section 3.2/4, never tested. One can accept the stochastic framework as plausible and still find the headline M87* claim unverified, exactly as the reader concluded.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper develops a spatio-temporal stochastic model of accretion-flow emission around a rotating Hayward black hole. The authors generalize Matérn Gaussian random fields to inhomogeneous, locally anisotropic fields (their 'INOISY' framework), with correlation structure tied to a Keplerian velocity field, and couple the resulting emissivity field to fast-light and slow-light ray tracing. They present synthetic images at inclinations 17°, 53°, and 75°, for magnetic charges g = 0.5 and 0.8, time-series snapshots, and claim (i) statistical consistency with GRMHD simulations and (ii) reproduction of the ~30° clockwise shift of the M87* bright ring between 2017 and 2018. The paper contains no quantitative comparison to EHT data or to GRMHD outputs, and the fast-light/slow-light implementation is described inconsistently.","tokens_in":16167,"tokens_out":8664,"duration_ms":105407,"significance":"The proposed stochastic surrogate is genuinely useful if the claims hold: it offers a computationally cheap way to generate time-resolved black-hole images with tunable statistical properties, which could help interpret EHT/ngEHT variability. The extension of anisotropic Matérn SPDEs to spatio-temporal ray tracing in a non-Kerr spacetime is a reasonable technical contribution, and the explicit construction of the anisotropy tensor from a Keplerian flow is a nice idea. However, the paper's headline results are currently asserted rather than demonstrated: the M87* ring-shift claim lacks any quantitative image analysis, the GRMHD comparison is qualitative, and the model parameters are calibrated to the target morphology. No code or data products are provided beyond a YouTube link, which further limits reproducibility.","major_comments":[{"comment":"The central claim that the simulations 'reproduce the dynamic positional shift of the bright ring structure observed in M87*' is not supported by any quantitative analysis. No position-angle time series of the simulated ring is defined or plotted, no estimator (e.g., the phase of the m=1 azimuthal brightness mode or a ring-centroid shift) is described, and no comparison to the EHT 2017/2018 position-angle measurements of Ref. [50] is made. Figure 17 merely reproduces EHT images; it does not compare them with the simulation. Because the model already contains a rotating, advected pattern with pitch angle 20° and Keplerian rotation, some clockwise drift is essentially built in; demonstrating that it matches the observed ~30° shift requires measuring the simulated shift from the images. Please add this quantitative step, with error bars, or weaken the claim in the abstract and conclusion.","section":"Abstract; Section 4"},{"comment":"The abstract and conclusion claim that the model 'maintains statistical consistency' with GRMHD simulations, but no quantitative comparison is presented. The text states that the images are 'qualitatively consistent' and 'closely matching' GRMHD outputs, yet no comparison metric, no GRMHD baseline dataset, and no uncertainty quantification are given. A meaningful test would compare, for example, image-domain correlation coefficients, visibility amplitude distributions, or power-spectrum slopes against published GRMHD snapshot libraries (such as those used in EHT modeling papers), and would report the resulting agreement and its dependence on the free parameters. Without this, the GRMHD-consistency claim is unverifiable.","section":"Section 3.2; Section 4"},{"comment":"The model parameters—lambda1/r = 5, lambda2/lambda1 = 0.1, lambda0 = 2*pi/Omega_K, theta_angle = 20 degrees, and the envelope g(r) = x^4 e^{-x^2}—are introduced as 'guided by' M87* and GRMHD morphology, and the paper then treats the resulting resemblance as validation. That is circular: if the parameters are chosen to reproduce M87-like and GRMHD-like morphology, the observation that the images look M87-like or GRMHD-like does not validate the model. Please provide a sensitivity analysis and an independent calibration of these parameters, for example by fitting the fluctuation-field statistics to actual GRMHD snapshots, and state explicitly which features are predictions rather than inputs.","section":"Section 3.2"},{"comment":"The fast-light and slow-light implementations are internally inconsistent. The text first says 'we adopt the fast-light approximation' and defines the image by Eq. (3.27), but later says the framework 'incorporate[s] the slow-light effect' and presents Figs. 11-13 as 'slow-light images,' while the formalism referenced for those results is still Eq. (3.27). Equations (3.29)-(3.31) define V_fast, V_slow, and a residual delta V, but no slow-light computation, no delta V plot, and no algorithm for the trajectory-dependent emission times t_s^(n)(x) are presented. Please clarify which figures are produced with which approximation, provide the actual slow-light calculation (or state explicitly that slow-light results are deferred), and show the quantitative difference between the two, if any.","section":"Section 3.3; Eqs. (3.27)-(3.31)"},{"comment":"There is a sign/exponent inconsistency in the definition of the anisotropy tensor. Equation (2.11) defines Lambda(x) = sum_l lambda_l^2 u_l u_l^T, which is consistent with Eqs. (3.14) and (3.17), but Eq. (3.12) writes Lambda(x_s) = sum_l lambda_l^{-2} u_l u_l^T, which is the inverse tensor. This matters because Eq. (3.16) uses Lambda^{-1} to define the Mahalanobis distance. Please correct Eq. (3.12) and verify that the numerical implementation follows the same convention; if the code follows Eq. (3.12), then the interpretation of lambda_1 and lambda_2 as correlation lengths is reversed.","section":"Section 2; Eqs. (2.11), (3.12), (3.14), (3.17)"}],"minor_comments":[{"comment":"The rotating Hayward metric is attributed to Ref. [39] (Pauls et al., 'Time-Dependent Ray-Tracing of Black Hole Accretion Flows with Gaussian Random Fields'), which is not the source of the metric; please cite the original Hayward metric and the relevant rotating-regular-black-hole paper.","section":"Section 2, first paragraph"},{"comment":"The time axis is confusing: Figs. 14-16 caption says 'total duration of 400 seconds,' while the rest of the paper uses geometrical units with M = 1 and times in units of M. Please specify the mass scaling and, if the target is M87*, give the conversion to physical time so that the 400-second duration can be compared with the dynamical timescale of M87*.","section":"Section 3.3; Figs. 14-16"},{"comment":"The Matérn power spectral density as written appears to be missing the factor kappa^{2*nu} and has an unusual normalization; please check Eq. (3.13) against the standard Matérn spectrum and correct any typographical error.","section":"Eq. (3.13)"},{"comment":"The acronym 'INOISY' is introduced without being defined. Please spell out the name or explain what it stands for in the text.","section":"Introduction; Section 3.2"},{"comment":"The radial infall rate iota is introduced in the velocity field but its value is never specified in the simulation setup. Please state whether iota = 0 in the presented runs or give the adopted radial-velocity profile.","section":"Section 3.3, Eq. (3.24)"},{"comment":"The caption of Fig. 17 reads like a press release and does not indicate the source of the displayed EHT images; please mark the figure as reproduced from Ref. [50] and state explicitly that it is observational data, not a product of this work.","section":"Figure 17"}],"recommendation":"major_revision","confidential_remarks":"The manuscript has a plausible methodological core but its advertised validation is absent. The authors should either add the missing quantitative analyses (position-angle time series, GRMHD comparison, fast-light/slow-light cross-check) or restructure the abstract and conclusion to present the M87* and GRMHD statements as qualitative demonstrations rather than verified reproductions. I would not recommend acceptance in the current form."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Take this paper as a framework paper, not as validation. The authors build a spatio-temporal anisotropic Matérn field model for accretion disk fluctuations around a rotating Hayward black hole, couple it to fast- and slow-light ray tracing, and show a set of time-lapsed synthetic images. That is a reasonable and potentially useful tool. The extension of Matérn/GMRF machinery to the spatio-temporal domain with a tensor-valued anisotropy Λ(x) is not in the cited GRF-imaging literature, and the slow-light comparison is a nice addition. The computational efficiency claim is plausible, since GMRF sampling is cheap compared to GRMHD.\n\nBut the abstract overstates the results. The statement that the simulations 'reproduce the dynamic positional shift of the bright ring structure observed in M87*' is not backed by any quantitative measurement. No position-angle time series is defined or plotted, no comparison to the EHT 2017/2018 shift is made, and no uncertainties are attached. A rotating asymmetric brightness pattern will naturally produce azimuthal drift; a 20-degree pitch angle and Keplerian rotation make clockwise drift almost by construction. The stress-test note is correct on this point.\n\nThere is also an internal inconsistency in Section 3.3. The text first says they adopt the fast-light approximation, then later says they 'augment' it with slow-light and describe slow-light results. The two paradigms need to be presented as separate, clearly labelled experiments. As written, it reads as if the same images were produced both ways.\n\nThe parameter choices are hand-tuned (lambda1/r=5, lambda2/lambda1=0.1, lambda0=2π/Ω_K, envelope g(r)=x^4e^{-x^2}) to mimic M87* and GRMHD morphology, and then the paper uses that as evidence of consistency. That is partially circular. It is not fatal if the parameters are presented as a calibration exercise, but the text should say so rather than claiming independent reproduction.\n\nThe paper deserves a serious referee, because the framework is coherent and the missing quantitative work is achievable. But it should not be accepted in its current form. A major revision needs to add: a measured position-angle time series from the simulated images, a direct comparison to EHT data, a clear split between fast-light and slow-light results, and a more modest abstract.","headline":"Useful stochastic imaging framework for regular black holes, but the M87* ring-shift claim is asserted, not demonstrated.","tokens_in":16684,"tokens_out":2935,"would_cite":false,"duration_ms":34345,"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 spatio-temporal Matérn field with anisotropic Keplerian structure can generate realistic time-dependent images of a rotating Hayward black hole, reproducing M87*'s 30-degree ring shift.","keywords":["rotating Hayward black hole","black hole shadow","accretion disk turbulence","spatio-temporal Matérn field","stochastic generative model","ray tracing","M87*","slow-light effects"],"falsifier":"Generate a long sequence with the stated parameters, measure the brightest-sector position angle of the photon ring in each frame, and compare with the EHT 2017-2018 M87* data: the claim predicts a roughly 30-degree clockwise shift, so a failure to produce that shift, or a shift that depends strongly on arbitrary parameter choices, would falsify the reproduction claim.","tokens_in":15559,"feed_emoji":"🕳️","tokens_out":5551,"duration_ms":65081,"temperature":0.7,"pith_summary":"Using a spatio-temporal generalization of Matérn random fields, this paper builds a fast generative model of the turbulent accretion flow around a rotating Hayward black hole and renders its time-dependent image by ray tracing. The locally anisotropic correlation tensor of the field is aligned with Keplerian motion, so the stochastic texture mimics GRMHD turbulence without solving magnetohydrodynamic equations. The central claim is that this synthetic image sequence reproduces the observed roughly 30-degree clockwise shift of the brightest sector of M87*'s emission ring between 2017 and 2018, with statistical consistency comparable to GRMHD simulations at far lower computational cost. If right, this provides a practical surrogate for producing time-resolved black hole images and for interpreting variability in EHT and ngEHT observations.","feed_headline":"Stochastic field model recreates M87*'s shifting bright ring","feed_subtitle":"Fast synthetic images of a rotating Hayward black hole match GRMHD statistics and the observed 2017-18 ring shift.","key_machinery":"The load-bearing object is the spatio-temporal Matérn field defined by the stochastic partial differential equation whose Green's function gives a Matérn covariance. The anisotropy tensor encodes temporal coherence and spatial correlation lengths along directions aligned with the flow: the temporal axis follows the local velocity field, while one spatial principal axis is tilted by about 20 degrees to mimic the spiral pitch seen in shearing-box simulations. The intensity is the radial envelope multiplied by the exponential of the normalized fluctuation field, and the resulting maps are passed through fast-light or slow-light ray tracing. This mechanism lets the authors generate statistically plausible turbulent disk images without solving MHD equations.","core_discovery":"The discovery the authors are trying to establish is that a stochastic generative model, an inhomogeneous, anisotropic, spatio-temporal Gaussian random field built on the Matérn covariance, can serve as a fast stand-in for GRMHD simulations when imaging a rotating Hayward black hole. The field is constructed through a stochastic partial differential equation whose anisotropy tensor encodes direction-dependent correlation lengths set by a Keplerian velocity field, and the source brightness is the radial envelope multiplied by the exponential of the normalized field. Images are obtained with both fast-light and slow-light ray tracing. The paper reports that the slow-light treatment smears strongly lensed, rapidly varying features, improving physical realism, and that the resulting time-lapse sequence reproduces the dynamic positional shift of the bright ring seen in M87*, advancing clockwise by about 30 degrees from 2017 to 2018. The authors claim this demonstrates statistical consistency with GRMHD while offering substantial computational efficiency.","pith_inferences":["The hand-set correlation parameters are not calibrated to MHD; a natural next step would be to fit them to a GRMHD snapshot and check whether the ring-shift statistic survives.","Whether the 30-degree shift is a robust prediction or a by-product of the assumed envelope and anisotropy could be tested by varying those choices and measuring the shift distribution.","The framework's visibility-space predictions could serve as a fast noise model for testing how temporal undersampling biases shadow recovery in future ngEHT reconstruction algorithms."],"forward_implications":["The same pipeline can generate long, time-resolved synthetic image sequences for any regular or non-Kerr metric, enabling parameter surveys of spin and magnetic charge.","The roughly 30-degree ring-shift reproduction, if robust, gives a concrete way to connect stochastic accretion variability to EHT time-variable images of M87*.","Because slow-light ray tracing changes morphology mainly in strongly lensed regions, fast-light images are adequate for quasi-steady disks but not for rapidly evolving hotspots.","Synthetic visibility functions from the model can be compared directly with VLBI data to constrain turbulence parameters."],"supporting_citations":[{"why":"Supplies the Hayward metric, the nonsingular rotating spacetime in which the images are rendered.","marker":"[13]"},{"why":"Provides the time-dependent GRMHD simulations whose output the stochastic model is compared against.","marker":"[32]"},{"why":"Earlier time-dependent ray tracing with Gaussian random fields that this work extends to non-Kerr spacetimes.","marker":"[39]"},{"why":"Origin of the inhomogeneous anisotropic Gaussian random field description of accretion disks.","marker":"[45]"},{"why":"Supplies the nonstationary SPDE construction with locally varying anisotropy used for the tensor field.","marker":"[46]"},{"why":"Establishes the SPDE link between Gaussian fields and Gaussian Markov random fields that makes sampling tractable.","marker":"[47]"},{"why":"The MHD turbulence locality study used to set the spiral pitch angle of the anisotropy axes.","marker":"[48]"},{"why":"Reports the EHT 2017-2018 observations of M87* showing the roughly 30-degree clockwise shift of the brightest ring sector.","marker":"[50]"}],"fun_headline_variants":["Stochastic model recreates M87*'s shifting bright ring","Fast generative model matches M87* ring shift","Time-lapse Hayward black hole images mimic M87*","Efficient random field captures M87* ring motion"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a zero-mean Gaussian fluctuation field with hand-set parameters, such as correlation scales, anisotropy ratio, temporal coherence, and a fixed radial envelope, faithfully represents the turbulent accretion flow around M87*, and this mapping is asserted rather than derived from magnetohydrodynamics.","fun_headline_variants_meta":{"raw":{"variants":["Stochastic model recreates M87*'s shifting bright ring","Fast generative model matches M87* ring shift","Time-lapse Hayward black hole images mimic M87*","Efficient random field captures M87* ring motion"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000233,"raw_usage":{"total_tokens":1474,"prompt_tokens":909,"completion_tokens":565,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":525,"completion_tokens_details":{"reasoning_tokens":498}},"tokens_in":525,"tokens_out":565,"duration_ms":6774,"temperature":1.0,"reasoning_tokens":498,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T12:32:02.156928+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Generate a long sequence with the stated parameters, measure the brightest-sector position angle of the photon ring in each frame, and compare with the EHT 2017-2018 M87* data: the claim predicts a roughly 30-degree clockwise shift, so a failure to produce that shift, or a shift that depends strongly on arbitrary parameter choices, would falsify the reproduction claim.","supporting_citations":[{"cited_title":"A.,Formation and Evaporation of Nonsingular Black Holes, Phys","cited_arxiv_id":null,"evidence_quote":"Supplies the Hayward metric, the nonsingular rotating spacetime in which the images are rendered."},{"cited_title":"J.942, 47 (2023)","cited_arxiv_id":null,"evidence_quote":"Provides the time-dependent GRMHD simulations whose output the stochastic model is compared against."},{"cited_title":"T., et al.,Time-Dependent Ray-Tracing of Black Hole Accretion Flows with Gaussian Random Fields, Mon","cited_arxiv_id":null,"evidence_quote":"Earlier time-dependent ray tracing with Gaussian random fields that this work extends to non-Kerr spacetimes."},{"cited_title":"Lee and C","cited_arxiv_id":null,"evidence_quote":"Origin of the inhomogeneous anisotropic Gaussian random field description of accretion disks."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the nonstationary SPDE construction with locally varying anisotropy used for the tensor field."},{"cited_title":"Lindgren, H","cited_arxiv_id":null,"evidence_quote":"Establishes the SPDE link between Gaussian fields and Gaussian Markov random fields that makes sampling tractable."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The MHD turbulence locality study used to set the spiral pitch angle of the anisotropy axes."},{"cited_title":"Akiyama, et al","cited_arxiv_id":null,"evidence_quote":"Reports the EHT 2017-2018 observations of M87* showing the roughly 30-degree clockwise shift of the brightest ring sector."}],"review_version":1}