{"id":"d3ece60e-bb16-433f-9ced-750f57355eaa","arxiv_id":"2507.07963","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A Fourier-series model reproduces orbit heating from fuzzy dark matter granules and predicts that subhalos above a critical size experience suppressed stochastic acceleration.","lead":"Fuzzy dark matter is made of ultralight particles whose wavelike interference creates random density clumps that nudge anything orbiting inside a dark matter halo. This paper builds a fast approximate model for those nudges, then shows that large subhalos feel them less strongly than point masses.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Missing null test: α is calibrated in one homogeneous FDM-SIMULATOR box and one realistic halo; the factor-of-two mismatch between α≈10 and αouter=5 leaves amplitude transferability untested.","rationale":"The reader's weakest assumption—that the amplitude identification in equation (13) is load-bearing—is correct, but the more specific and testable issue is that the model does not use aeff unmodified: calibrated α values differ by a factor of two between the point-mass model (αouter = 5.0) and the finite-size model's point-like limit (α ≈ 10). This means the amplitude normalization is being set empirically, and the model's transferability to other halo masses, radii, and particle masses rests on the untested assumption that α is universal. I do not see an internal inconsistency severe enough to reject the paper: the Fourier representation is plausible, Appendix B provides an independent analytic correlation calculation, and the calibration-based approach is a legitimate modeling strategy if properly validated. The concern is therefore about missing validation, not demonstrated failure. The concrete test—rerunning the calibration at different background conditions—would settle whether the factor-of-two mismatch is a genuine environment dependence or merely a benign convention difference. If α changes significantly, the conditional verdict should be strengthened to require out-of-sample validation before the model is used for predictions; if α is stable, the current conditional verdict is adequate. Hence the reader's CONDITIONAL verdict should stand unchanged.","tokens_in":26367,"tokens_out":8520,"duration_ms":102446,"concrete_test":"Re-run the FDM-SIMULATOR calibration at a second background condition, e.g. σJ = 35 km/s or ρ0 = 4 times the fiducial value, and re-fit α for the point-like limit. If α changes by more than the quoted ~7% realization-to-realization scatter, the calibration is environment-dependent and predictions outside the fiducial setup are unreliable. As a second check, recover α from a point-mass particle moving in a stratified density profile matching the Dutta Chowdhury halo at the same local ρ and σJ; if it does not reproduce the homogeneous-box α once the density-gradient explanation is applied, the factor-of-two discrepancy in Section 4.2 remains unresolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that individual orbits can be integrated quickly and that finite-size subhalos show suppressed stochastic acceleration above rhalf/deff≈0.40—depends entirely on the stochastic acceleration amplitude. Equation (13) sets the Fourier coefficient width to σa = aeff = G meff/d_eff², but the model does not actually use this amplitude unmodified. The point-mass model needs αouter = 5.0, while the FDM-SIMULATOR point-like subhalo limit gives α ≈ 10 (Fig. 6). The paper attributes this factor-of-two to the homogeneous simulation box versus the declining density profile of a realistic halo, but it provides no calculation or separate simulation showing how α should scale with local density, density gradient, or velocity dispersion. Because α absorbs any error in aeff, the agreement with the calibration data is in-sample; the predictive content is the claim that α is universal. That universality is untested: all finite-size calibrations use a single background (ρ0 = 1.628 × 10^4 M⊙ kpc^-3, σJ = 17.775 km/s, Section 4.1), and all point-mass calibrations use a single halo profile from Dutta Chowdhury et al. (2021). If α is environment-dependent, orbit integrations outside these setups are systematically biased. Additionally, the claimed critical size is not sharp: α drops from 10.0 to 9.3 already in 0.17 ≲ rhalf/deff ≲ 0.39, before the power-law regime begins at 0.40, so the point-mass threshold is an approximation, not a clean derived transition. Appendix B's analytic correlation calculation supports the slope of the suppression but not its normalization, and it shares the same acceleration-field modeling assumptions, so it cannot independently certify the amplitude.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a semi-analytic framework for modeling the stochastic gravitational perturbations induced by fuzzy dark matter (FDM) granules on orbiting bodies. The authors represent granule accelerations as truncated Paley–Wiener Fourier series with random coefficients, allowing individual orbit integrations rather than population-level diffusion models. They calibrate three dimensionless parameters (alpha_core, alpha_outer, beta) for point-mass particles against the Dutta Chowdhury et al. (2021) simulations, then extend the model to finite-size subhalos. For the extended case, they introduce FDM-SIMULATOR, a pseudo-spectral code that evolves subhalos in a homogeneous random-wave background, and fit a size-dependent suppression factor alpha(r_half/d_eff) with a claimed transition at r_half/d_eff about 0.4. Appendix B derives analytic expressions for the granule acceleration correlation and the resulting velocity dispersion, including anisotropy and finite-size convolution.","tokens_in":26720,"tokens_out":5373,"duration_ms":60855,"significance":"If the framework is sound, it offers a practical tool for evolving individual orbits in FDM halos at far lower cost than full wave simulations, and the finite-size suppression effect is physically interesting and potentially important for subhalo population modeling. The analytic correlation-function calculation in Appendix B and the validation of the convolution formula in Figs B4 and B5 are genuine strengths, as is the explicit treatment of acceleration anisotropy. The main limitation is that the predictive claims rest on calibrated parameters whose universality is not yet demonstrated: the point-mass parameters are fitted to the same simulations against which they are compared, and the finite-size alpha relation is fitted to and tested on the same FDM-SIMULATOR runs. As a result, the paper currently establishes that the model can reproduce its calibration data, but not that it predicts independent outcomes.","major_comments":[{"comment":"The amplitude identification is internally inconsistent. In Eq. (13) the authors set the coefficient width to a_eff through 2 sqrt(pi)/t_max v_x,n = a_eff, and state in Eq. (14) that the coefficients have width sigma_a = a_eff. However, in the actual acceleration series of Eq. (16), the variance of the summed cosine terms is approximately (1/(f_granule t_max)) * a_eff^2 * (m/2) = 2 a_eff^2 for m = 4 f_granule t_max, so the root-mean-square stochastic acceleration is sqrt(2) a_eff, not a_eff. Thus the model does not actually implement the quasi-particle amplitude from Eq. (3); the discrepancy is absorbed into the calibrated alpha parameters. Because all predicted heating scales linearly with alpha times a_eff, the paper should either correct the normalization so that the injected variance equals a_eff^2, or explicitly state that alpha absorbs this sqrt(2) factor and test whether alpha remains universal once the normalization is fixed.","section":"Section 2.2, Eqs. (13)-(16)"},{"comment":"The point-mass model is calibrated and validated on the same data. The parameters alpha_core = 0.3, alpha_outer = 5.0, and beta = 0.16 are obtained by minimizing the chi-squared statistic in Eq. (27) against the Dutta Chowdhury et al. (2021) simulations, and Fig. 3 then compares the model with those same simulations. This is an in-sample test, so the agreement demonstrates flexibility rather than predictive power. The predictive content of the model is the claim that these parameters are universal across halo masses, particle masses, and orbital families. To support that claim, the authors should provide out-of-sample tests, for example predicting orbital evolution for a different host halo mass, a different initial radius or eccentricity, or a different FDM particle mass; at minimum, the chi-squared surface and parameter degeneracies should be reported so the reader can judge how tightly the parameters are constrained.","section":"Section 3, Fig. 3, Eq. (27)"},{"comment":"The finite-size relation alpha(r_half/d_eff) is fitted to the same FDM-SIMULATOR data that are used to validate it, and the claimed sharp transition at r_half/d_eff = 0.40 is not well supported. The plateau value alpha = 10.0, the intermediate value alpha = 9.3, the transition boundaries, and the power-law parameters A and B are all determined from the same runs; no independent test is presented. Moreover, the drop from 10.0 to 9.3 occurs already in the range 0.17 < r_half/d_eff < 0.39, which is comparable to the quoted run-to-run variation of about 7 percent in V_rms, so the significance of the intermediate 'point-like' state needs to be quantified with error bars on alpha. The authors should test the fitted relation against simulations with different background density, velocity dispersion, or boson mass, and should report the uncertainties on the fitted parameters.","section":"Section 4.2, Fig. 6, Eq. (33)"},{"comment":"The relationship between the calibrated alpha and the analytic predictions in Appendix B is not established. The acceleration variance for NFW subhalos is found to scale approximately as (r_half/d_eff)^-0.41 (Appendix B, Fig. B4), while the calibrated alpha and the velocity dispersion ratio scale as (r_half/d_eff)^-0.229 (Eq. (33) and Fig. B5). Since alpha is defined as a multiplicative factor on the granule acceleration, the paper should show explicitly how alpha is derived from the acceleration variance or from the velocity dispersion integral in Eq. (B1); otherwise the power-law fit is purely empirical and its extrapolation to other halos is unsupported. The factor-of-two offset between alpha_outer = 5 and the FDM-SIMULATOR value alpha approximately 10 is attributed to the difference between a homogeneous box and a declining density profile, but no calculation is given; deriving alpha from the Appendix B formalism for both backgrounds would test this explanation directly.","section":"Section 4.2 and Appendix B, Eqs. (B31)-(B32), Fig. B5"}],"minor_comments":[{"comment":"The text says 'viral mass Mvir' and should say 'virial mass'.","section":"Section 3, text near Eq. (18)"},{"comment":"The factor of 4 in m = 4 f_granule t_max is described as empirical; please provide a quantitative convergence test, for example showing how the acceleration variance or orbit heating changes when the factor is 2 or 8.","section":"Section 2.2, Eq. (11)"},{"comment":"The symbol sigma is used for both the one-dimensional velocity dispersion and the total velocity dispersion in Eqs. (B19)-(B21); please clarify which quantity each equation defines.","section":"Appendix B, Eqs. (B19)-(B21)"},{"comment":"The alpha values in Fig. 6 should include error bars from the 20 realizations, and the text should state how the r_half values are computed from the concentration-mass relation for each subhalo mass.","section":"Fig. 6"},{"comment":"FDM-SIMULATOR is described as under active development and not publicly available; given the central role of this code in the validation, please provide a stable version or a detailed code release plan so that the results are reproducible.","section":"Data Availability"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for MNRAS and the basic idea is promising, but the current evidence is largely in-sample. The authors should be asked to provide out-of-sample tests and to clarify the normalization of the stochastic acceleration amplitude in Eq. (16). If they can show that alpha is stable across environments and derive the finite-size suppression from the Appendix B formalism, the paper would be suitable for publication. The lack of public code is a secondary concern but worth mentioning in the decision letter."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe one thing to know: this paper gives you a Fourier-series (Paley–Wiener) way to represent FDM granule perturbations that works for individual orbits rather than only statistical populations, and it extends the model to finite-size subhaloes with a claimed suppression above rhalf/deff ≈ 0.4. The point-mass version reproduces the Dutta Chowdhury et al. (2021) simulations, and Appendix B provides an analytic correlation calculation that independently supports the slope of the size suppression and the velocity-dependent anisotropy. That is real value.\n\nWhat is actually new and good: the Paley–Wiener representation applied to individual orbits is genuinely different from the diffusion approximation used in prior work. The finite-size extension with a critical rhalf/deff scale is also new. The model is fast (about a minute per orbit), which matters for population-scale FDM studies. The Appendix B calculation is the strongest part — it gives a first-principles handle on how the acceleration variance and velocity dispersion scale with subhalo size, and it matches the simulation trends.\n\nThe soft spots are real but not fatal. The amplitude of the stochastic acceleration is set by identifying Fourier coefficient variance with aeff = G meff/d_eff² (eq. 13), and the truncation factor 4 in the Fourier series is chosen empirically. More importantly, the calibration is largely in-sample: the point-mass parameters αcore=0.3, αouter=5.0, β=0.16 are fitted to the same Dutta Chowdhury et al. simulations against which they are compared, and the finite-size α(r_half) is fitted to FDM-SIMULATOR and then checked against that same simulator. The stress-test note is right that the factor-of-two mismatch between α≈10 (FDM-SIMULATOR, homogeneous box) and αouter=5 (point-mass, realistic halo) is attributed to differences in background density but never demonstrated. That leaves the universality of α untested. If α varies with environment, predictions from the model outside the calibrated setups will be systematically off. Also, the claimed critical threshold is soft: α drops from 10.0 to 9.3 already in 0.17 ≲ rhalf/deff ≲ 0.39, so the “transition” at 0.40 is more of a fitting convention than a sharp derived scale. No error bars are given for the fits, and the code/data are not public (available on request; simulator under development). These are addressable: run one independent FDM simulation at a different halo mass or density, release the code, and report uncertainties.\n\nBottom line: this is a useful framework for FDM subhalo dynamics, with a solid analytic appendix, but the predictive content currently rests on an unverified universality assumption. The paper deserves a serious referee — I would send it to review rather than desk reject. For an internal reading group, it would generate good discussion about calibration versus prediction, but I would not use it as-is for quantitative constraints on mb without an independent check.","headline":"Useful Fourier-series framework for FDM granule perturbations on individual orbits, but the amplitude calibration is in-sample and the claimed universality of α is untested.","tokens_in":27305,"tokens_out":3180,"would_cite":true,"duration_ms":29019,"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":"Granule-induced orbital jitter in fuzzy dark matter haloes can be represented as a truncated Fourier series with random coefficients, giving fast per-orbit integration and a size threshold near $r_{\\rm half}/d_{\\rm eff} \\approx 0.4$ for…","keywords":["fuzzy dark matter","ultralight axions","granules","stochastic acceleration","Paley–Wiener representation","finite-size subhaloes","semi-analytic model","dynamical friction"],"falsifier":"Run many realizations of point-mass test particles in a homogeneous wave background with known density and velocity dispersion, measuring the ensemble velocity-dispersion growth; disagreement beyond the roughly 7 per cent realization scatter reported in the paper would falsify the coefficient assignment in equation (13).","tokens_in":26102,"feed_emoji":"🌀","tokens_out":6439,"duration_ms":72251,"temperature":0.7,"pith_summary":"The paper proposes that the stochastic gravitational jitter from fuzzy dark matter 'granules' can be represented as a Fourier series with Gaussian random coefficients, so that an individual orbit gets its own random walk rather than only a population-averaged diffusion description. The authors calibrate this semi-analytic model against high-resolution simulations of an FDM halo and verify it with their own wave-mechanics simulator, then extend it to finite-size subhaloes. They find a critical ratio $r_{\\rm half}/d_{\\rm eff} \\approx 0.40$ below which subhaloes behave as point masses and above which the extended mass profile suppresses granule acceleration. If the model holds, orbit integrations in FDM haloes take about a minute, enabling fast population studies and eventual constraints on the axion mass.","feed_headline":"Granule jitter modeled as random Fourier noise for any orbit","feed_subtitle":"Semi-analytic model reproduces simulations and finds large subhaloes feel weaker kicks.","key_machinery":"The load-bearing object is the Paley–Wiener representation of Brownian motion, truncated to frequencies up to the granule oscillation frequency $f_{\\rm granule}$ with $m = 4 f_{\\rm granule} t_{\\rm max}$, where the factor 4 is chosen empirically. Random Fourier coefficients are drawn from a zero-mean Gaussian with width $\\sigma_a = a_{\\rm eff}$, which lets a single orbit be integrated while receiving stochastic kicks. A velocity-dependent anisotropic rescaling, with factors $A_\\parallel$ and $A_\\perp$, is fitted from the simulator and applied to the outer-halo component, capturing the reduced heating of fast-moving particles. For finite-size bodies, the granule acceleration variance is computed as the convolution of the subhalo density power spectrum with the granule acceleration power spectrum, which produces the observed suppression and the $\\propto (r_{\\rm half}/d_{\\rm eff})^{-0.229}$ velocity-dispersion scaling.","core_discovery":"The central discovery is that granule-induced orbital perturbations are statistically equivalent to a truncated Paley–Wiener Brownian-motion expansion: accelerations are sums of cosines with independent zero-mean Gaussian coefficients whose variance is set by a quasi-particle effective acceleration $a_{\\rm eff} = G m_{\\rm eff}/d_{\\rm eff}^2$. With calibrated constants for the soliton core and outer-halo granules ($\\alpha_{\\rm core}=0.3$, $\\alpha_{\\rm outer}=5.0$) and a dynamical-friction scale $\\beta=0.16$, the model reproduces the median radius and velocity-dispersion evolution of point masses across a wide range of particle masses. For subhaloes, the acceleration strength stays constant at small half-mass radii and declines as a power law $\\alpha = 7.755(r_{\\rm half}/d_{\\rm eff})^{-0.229}$ once $r_{\\rm half}/d_{\\rm eff}$ exceeds about 0.4; the same size-suppression trend appears directly in the wave-mechanics simulator and in the analytic convolution of subhalo density and granule acceleration power spectra.","pith_inferences":["A natural next step the paper leaves implicit is to apply the same truncated-Fourier machinery to stellar discs and tidal streams in FDM haloes, since those observables depend on individual stellar orbits rather than population averages.","A testable extension is to re-fit $\\alpha_{\\rm core}$, $\\alpha_{\\rm outer}$, and $\\beta$ at different FDM particle masses and host halo masses; if the constants drift significantly, the model would need mass-dependent calibration rather than a single global set.","The paper notes that gravitational cooling of the soliton is not modelled, which they connect to a late-time radius decay in the heaviest subhalo case; adding a time-dependent soliton mass could close that gap and improve long-term forecasts.","The predicted transition near $r_{\\rm half}/d_{\\rm eff} \\approx 0.4$ could be probed observationally: if subhalo disruption statistics in FDM-dominated dwarfs show a size-dependent threshold, that would provide independent evidence for the finite-size suppression mechanism."],"forward_implications":["Individual orbits, not just ensembles: the Fourier-coefficient model turns granule heating into a fast stochastic integration, making parameter-space exploration over particle masses and halo masses practical.","Finite-size suppression: subhaloes with $r_{\\rm half}/d_{\\rm eff}$ above roughly 0.4 feel progressively weaker granule kicks, so orbit-heating predictions for realistic satellites must account for internal structure.","Population forecasts: the calibrated model can generate FDM subhalo populations and compare them against CDM predictions, and combined with observations it could constrain the FDM boson mass $m_b$.","Consistency with kinetic theory: the random-walk amplitudes match diffusion coefficients derived from quasi-particle kinetic theory, giving the model an independent theoretical anchor outside the calibration."],"supporting_citations":[{"why":"Supplies the FDM halo simulation data and density profile used to calibrate $\\alpha_{\\rm core}$, $\\alpha_{\\rm outer}$, and $\\beta$ for point-mass orbits.","marker":"Dutta Chowdhury et al. (2021)"},{"why":"Provides the quasi-particle effective-mass picture and the diffusion-coefficient framework used for the kinetic-theory comparison.","marker":"Bar-Or et al. (2019)"},{"why":"Provides the soliton solution and scaling relations used to set the core radius and soliton boundary $r_{\\rm sol}=2.7r_c$.","marker":"Schive et al. (2014a)"},{"why":"Supplies the Paley–Wiener representation of Brownian motion on which the Fourier-coefficient acceleration series is built.","marker":"Higham (2015)"},{"why":"Establishes the FDM halo description, de Broglie wavelength, and granule picture that the model represents.","marker":"Hui et al. (2017)"},{"why":"Provides the dynamical-friction treatment whose Chandrasekhar form is adopted and re-scaled by $\\beta$.","marker":"Lancaster et al. (2020)"}],"fun_headline_variants":["Fuzzy dark matter granule jitter as random Fourier noise","Semi-analytic model predicts granule kicks on any orbit","Large subhaloes feel weaker fuzzy dark matter granule kicks","Subhalo size suppresses fuzzy dark matter granule perturbations","Random Fourier model matches fuzzy dark matter orbit simulations"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything scales with the quasi-particle amplitude $a_{\\rm eff} = G m_{\\rm eff}/d_{\\rm eff}^2$; if that estimate of granule acceleration amplitude is wrong, the model's heating and all calibrated constants shift proportionally.","fun_headline_variants_meta":{"raw":{"variants":["Fuzzy dark matter granule jitter as random Fourier noise","Semi-analytic model predicts granule kicks on any orbit","Large subhaloes feel weaker fuzzy dark matter granule kicks","Subhalo size suppresses fuzzy dark matter granule perturbations","Random Fourier model matches fuzzy dark matter orbit simulations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000187,"raw_usage":{"total_tokens":1316,"prompt_tokens":921,"completion_tokens":395,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":537,"completion_tokens_details":{"reasoning_tokens":313}},"tokens_in":537,"tokens_out":395,"duration_ms":5217,"temperature":1.0,"reasoning_tokens":313,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:28:29.110714+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run many realizations of point-mass test particles in a homogeneous wave background with known density and velocity dispersion, measuring the ensemble velocity-dispersion growth; disagreement beyond the roughly 7 per cent realization scatter reported in the paper would falsify the coefficient assignment in equation (13).","supporting_citations":[],"review_version":1}