A Velocity Coupled Radial Acceleration Ansatz for Disk-Galaxy Rotation Curves: Fits to SPARC, Bayesian Inference, and Parameter Identifiability

Nalin Dhiman

arxiv: 2512.21376 · v2 · submitted 2025-12-24 · 🌌 astro-ph.GA · physics.comp-ph

A Velocity Coupled Radial Acceleration Ansatz for Disk-Galaxy Rotation Curves: Fits to SPARC, Bayesian Inference, and Parameter Identifiability

Nalin Dhiman This is my paper

Pith reviewed 2026-05-16 19:48 UTC · model grok-4.3

classification 🌌 astro-ph.GA physics.comp-ph

keywords rotation curvesSPARCphenomenological modelgalaxy dynamicsdark matter halosBayesian inferencevelocity coupling

0 comments

The pith

A minimal velocity-coupled radial acceleration ansatz reproduces observed disk-galaxy rotation curves at a level competitive with standard dark-matter halo profiles.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper proposes a simple phenomenological model in which an extra inward radial acceleration is taken to be proportional to the local tangential speed, with a coupling strength that saturates at large radius. Inserting this term into the circular-orbit equation produces a quadratic whose physical root gives the predicted rotation velocity once the known baryonic contributions are subtracted. When this velocity-coupled acceleration (VCA) form is fitted to the 171 SPARC rotation curves using published baryonic decompositions and a uniform error model, it yields information-criterion scores that are typically comparable to those of NFW and Burkert halos for a 5 km/s systematic floor. The work also reports MCMC posteriors, notes a pronounced parameter degeneracy for most galaxies, and shows that hold-out radial predictions perform similarly to the halo models. A sympathetic reader cares because the construction supplies an explicit, closed-form alternative to halo parameterizations that can be tested directly against the same data.

Core claim

The velocity-coupled acceleration ansatz a_VCA(r) = gamma(r) v(r) with gamma(r) = v_inf / (r + r_zero), when added to the Newtonian acceleration from the observed gas, disk and bulge, produces a quadratic equation for v(r) whose physical branch reproduces the measured rotation curves of SPARC galaxies at a level competitive with two-parameter NFW and Burkert halo models under identical optimization and error treatment.

What carries the argument

The saturating coupling gamma(r) = v_inf / (r + r_zero) that multiplies the local tangential speed to supply the additional radial acceleration term.

If this is right

The VCA model supplies a two-parameter closed-form alternative to halo-based fits for rotation-curve data.
Parameter posteriors for v_inf and r_zero are strongly degenerate for most of the 171 galaxies, with only 47 showing well-constrained marginals.
Radial hold-out cross-validation (outer 30 percent of radii) yields predictive RMSE values comparable to those of NFW and Burkert.
The model remains competitive with NFW and performs comparably, though not uniformly better, than Burkert under information-criterion ranking for the fiducial error floor.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

If the ansatz is approximately correct, the same coupling form could be inserted into dynamical models of other systems such as galaxy clusters or high-redshift disks to test consistency with independent mass tracers.
The observed v_inf-r_zero degeneracy suggests that additional kinematic data, such as vertical velocity dispersions or HI linewidths, might break the degeneracy and tighten constraints.
A deeper theory that naturally produces a radial acceleration linear in tangential speed could be sought by examining whether the same relation appears in non-circular orbits or in the time-dependent evolution of disk galaxies.

Load-bearing premise

The assumption that the chosen saturating form for the velocity-dependent coupling is sufficient to account for the entire difference between baryonic gravity and observed dynamics.

What would settle it

A large set of galaxies in which the VCA model produces systematically higher information criteria than both NFW and Burkert fits when the same 5 km/s systematic error floor and identical baryonic decompositions are used.

Figures

Figures reproduced from arXiv: 2512.21376 by Nalin Dhiman.

**Figure 1.** Figure 1: Representative SPARC rotation curves with best-fit models and residuals (in units of σeff ; Eq. 16). Shown are a high-quality extended spiral (NGC3198), an intermediate case (NGC2403), and a dwarf (DDO154). The VCA model (red) is competitive with NFW (blue) and Burkert (green) in many cases, though performance varies galaxy-to-galaxy [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗

**Figure 2.** Figure 2: AIC comparison at σ0 = 5 km s−1 . Left: ∆AIC = AICNFW − AICvca (positive favors VCA ). Right: ∆AIC = AICBur − AICvca (positive favors VCA ). sufficient number of points in both splits are included (Ngal = 143 in this run) [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** Figure 3: Radial cross-validation using the outer 30% of radii as a test set. The RMSE distributions for VCA, NFW, and Burkert are broadly similar, indicating comparable outof-sample performance for this radial holdout protocol [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

**Figure 4.** Figure 4: Model anatomy for NGC3198. The implicit effective contribution vvca(r) (Eq. 11) rises and saturates, producing a flat outer rotation curve when combined with the baryonic component. through a velocity-dependent coupling rather than an explicit mass density profile. 5.4. Bayesian posteriors, degeneracy, and identifiability MCMC posteriors typically show a strong correlation between log v∞ and log r0, con… view at source ↗

**Figure 6.** Figure 6: Identifiability diagnostics for VCA posteriors. Left: distribution of log-widths for r0 (68% credible interval); right: correlation between log-widths of v∞ and r0. The dashed lines mark the 0.5-dex threshold used in [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗

**Figure 8.** Figure 8: Posterior predictive check for NGC2403 (68% and 95% credible intervals). Residuals are shown in units of σeff . 12.0 11.5 11.0 10.5 10.0 9.5 9.0 8.5 8.0 log(gbar [m/s 2 ]) 12.0 11.5 11.0 10.5 10.0 9.5 9.0 8.5 8.0 lo g(g o b s [m / s 2 ]) Radial Acceleration Relation: Data vs VCA Data Density 1:1 (No DM) Standard RAR (MOND) [PITH_FULL_IMAGE:figures/full_fig_p007_8.png] view at source ↗

**Figure 7.** Figure 7: Asymptotic parameter v∞ versus observed maximum speed Vmax for the constrained subset. The dashed line indicates equality. terior predictive summary for NGC2403. Aggregating over all galaxies and radii, we find global predictive coverage of 77.3% for nominal 68% intervals and 93.6% for nominal 95% intervals. The 95% coverage is close to nominal, while the 68% intervals tend to be conservative (slightly o… view at source ↗

**Figure 10.** Figure 10: Model-preference diagnostics as a function of galaxy properties. Here ∆AIC is plotted as VCA minus comparison model (negative values favor VCA). We find no strong monotonic dependence on distance and only weak dependence on Vmax; the dominant effect is increased scatter at large Vmax and large Rmax. APPENDIX A. DEPENDENCE OF MODEL PREFERENCE ON GALAXY PROPERTIES To assess whether relative model preference… view at source ↗

**Figure 11.** Figure 11: MCMC diagnostic summary across the sample: distributions of autocorrelation times, effective sample sizes, and acceptance fractions [PITH_FULL_IMAGE:figures/full_fig_p011_11.png] view at source ↗

**Figure 12.** Figure 12: Posterior predictive coverage as a function of radius. Nominal 68% and 95% predictive intervals are shown; 68% intervals are somewhat conservative under the adopted error model [PITH_FULL_IMAGE:figures/full_fig_p012_12.png] view at source ↗

**Figure 13.** Figure 13: Sensitivity of AIC differences to the systematic error floor σ0 [PITH_FULL_IMAGE:figures/full_fig_p013_13.png] view at source ↗

**Figure 14.** Figure 14: RAR residuals as a function of radius (pipeline diagnostic) [PITH_FULL_IMAGE:figures/full_fig_p014_14.png] view at source ↗

**Figure 15.** Figure 15: Schematic illustrating latent-curve and posterior-predictive uncertainty for NGC2403 [PITH_FULL_IMAGE:figures/full_fig_p015_15.png] view at source ↗

read the original abstract

Observed rotation curves of disk galaxies remain a sharp empirical probe of the relationship between baryons and dynamics. We study a minimal, explicitly \emph{phenomenological} alternative to standard halo parameterizations: an additional inward \emph{radial} acceleration proportional to the local \emph{tangential} speed, $a_{\vca}(r)=\gamma(r)\,v(r)$, with a saturating coupling $\gamma(r)=\vinf/(r+\rzero)$. Combining this ansatz with the circular-motion condition yields a quadratic equation for $v(r)$ with a closed-form physical branch.We fit this ``velocity-coupled acceleration'' (\vca) model to $N_\mathrm{gal}=171$ rotation curves from the SPARC sample using the published baryonic decompositions (gas, disk, bulge), and we compare to two commonly used two-parameter halo models (NFW and Burkert) using an identical optimization pipeline and error model.For a fiducial systematic error floor $\sigma_0=5~\mathrm{km\,s^{-1}}$, the \vca model is typically competitive with an NFW halo and performs comparably (though not uniformly better) than a Burkert halo in information-criterion comparisons.We further perform MCMC inference for \vca parameters, quantify posterior predictive coverage, and show that parameter posteriors exhibit a strong $\vinf$--$\rzero$ degeneracy for many galaxies; only $47/171$ galaxies yield well-identified posteriors under simple width-based criteria. We also perform a simple radial holdout cross-validation (outer 30\% of radii) and find predictive RMSE comparable to NFW and Burkert under this protocol.Finally..

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The VCA ansatz gives a closed-form quadratic solution that fits SPARC curves about as well as NFW or Burkert under a fixed 5 km/s error floor, but strong degeneracies and no sensitivity checks on that floor limit how far the competitiveness claim travels.

read the letter

The paper's core contribution is a two-parameter phenomenological model where the extra acceleration is proportional to local velocity with a saturating gamma(r) = v_inf / (r + r_zero). This leads directly to a quadratic equation for v(r) whose physical root they use for fitting. They apply it to the full SPARC sample of 171 galaxies with published baryonic decompositions, run identical MCMC pipelines against NFW and Burkert, and report that at sigma_0 = 5 km/s the information criteria are typically comparable, with a radial hold-out test showing similar predictive RMSE. They also flag the v_inf-r_zero degeneracy explicitly and note that only 47 galaxies meet simple identifiability criteria. That honesty about limitations is useful. The math is transparent and the optimization pipeline is consistent across models, which makes the comparison straightforward to reproduce on the public data. The soft spots are real but not fatal. The claim of competitiveness is tied to one specific systematic error floor without showing how AIC or BIC differences shift at 3 or 10 km/s; given the already strong degeneracies, that choice can move the relative weighting of inner versus outer points and change per-galaxy rankings. The ansatz itself is presented as minimal and empirical rather than derived, so it inherits the usual limits of any ad-hoc parameterization. They rely on external baryonic mass models without re-fitting or propagating their uncertainties. This work is aimed at people who build or compare empirical rotation-curve models and want a compact alternative to halo profiles. It deserves a serious referee because the data handling is public, the degeneracy diagnostics are clear, and the hold-out test is a reasonable check, even though the error-floor robustness and physical motivation would need tightening before the model sees wider use.

Referee Report

2 major / 2 minor

Summary. The manuscript introduces a phenomenological velocity-coupled acceleration (VCA) ansatz a_VCA(r) = [v_inf / (r + r_zero)] * v(r), derives a closed-form quadratic solution for the rotation velocity v(r) from the circular-motion equation, and fits the two free parameters to the SPARC sample of 171 galaxies using published baryonic decompositions. It compares VCA performance to NFW and Burkert halo models via AIC/BIC at a fixed systematic error floor of 5 km/s, performs MCMC sampling to diagnose v_inf–r_zero degeneracies (finding only 47/171 galaxies well-identified), and reports a radial hold-out cross-validation with comparable RMSE.

Significance. If the central results hold, the VCA supplies a minimal two-parameter phenomenological alternative to standard halo parameterizations that is competitive on information criteria for the fiducial error model. Strengths include the explicit closed-form derivation, identical optimization pipeline across models, explicit degeneracy diagnostics, and predictive validation protocol; these features support its utility for empirical rotation-curve studies.

major comments (2)

[§4] §4 (information-criterion comparisons): the claim that VCA is 'typically competitive' with NFW and 'comparably (though not uniformly better)' than Burkert is reported exclusively for the fiducial σ₀ = 5 km/s. Because the total uncertainty enters the likelihood as σ(r) = sqrt(σ_obs(r)^2 + σ₀^2) and directly scales χ²/AIC/BIC, the relative model rankings could shift at neighboring values (e.g., 3 or 10 km/s). A sensitivity test varying σ₀ is required to substantiate the competitiveness statement.
[MCMC section] MCMC and identifiability section: the threshold used to classify only 47/171 galaxies as 'well-identified' is described only as 'simple width-based criteria' without explicit numerical values (e.g., posterior width ratio or correlation coefficient cutoff). Stating the precise criteria would make the degeneracy assessment reproducible.

minor comments (2)

[Abstract] The abstract ends abruptly ('Finally..'); ensure the final sentence is completed in the submitted version.
[Notation] Notation: confirm uniform subscripting of v_inf and r_zero in all equations and figure labels.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and positive assessment of our manuscript. We address each major comment below and will revise the manuscript to incorporate the requested clarifications and additional analysis.

read point-by-point responses

Referee: §4 (information-criterion comparisons): the claim that VCA is 'typically competitive' with NFW and 'comparably (though not uniformly better)' than Burkert is reported exclusively for the fiducial σ₀ = 5 km/s. Because the total uncertainty enters the likelihood as σ(r) = sqrt(σ_obs(r)^2 + σ₀^2) and directly scales χ²/AIC/BIC, the relative model rankings could shift at neighboring values (e.g., 3 or 10 km/s). A sensitivity test varying σ₀ is required to substantiate the competitiveness statement.

Authors: We agree that the information-criterion results are shown only for the fiducial σ₀ = 5 km/s and that rankings could in principle change with different systematic floors. In the revision we will add a sensitivity analysis repeating the full AIC/BIC comparison pipeline for σ₀ = 3, 5, and 10 km/s. The new results will be presented in an expanded §4 (including a table or figure summarizing model rankings across the three values), and the text will be updated to state whether the competitiveness conclusion holds under these variations. revision: yes
Referee: MCMC and identifiability section: the threshold used to classify only 47/171 galaxies as 'well-identified' is described only as 'simple width-based criteria' without explicit numerical values (e.g., posterior width ratio or correlation coefficient cutoff). Stating the precise criteria would make the degeneracy assessment reproducible.

Authors: We acknowledge that the exact numerical thresholds for the width-based classification were not stated explicitly. In the revised manuscript we will define the criteria in full: a galaxy is labeled 'well-identified' only if (i) the ratio of the 84th to 16th percentile posterior widths for both v_inf and r_zero is less than 3, and (ii) the absolute value of the Pearson correlation coefficient between the two parameters is less than 0.8. These thresholds will be stated in the text of the MCMC section together with the resulting count of 47 galaxies, and we will add a short footnote or appendix entry listing the precise numerical cutoffs for reproducibility. revision: yes

Circularity Check

0 steps flagged

No circularity: explicit phenomenological ansatz solved via quadratic

full rationale

The paper states the VCA relation as an explicit ansatz a_VCA(r) = [v_inf / (r + r_zero)] * v(r) and combines it with the circular-velocity equation v^2/r = a_baryon + a_VCA to obtain a quadratic whose physical root is selected. This is a direct algebraic consequence of the stated assumption rather than a reduction of the output to the input by construction. No load-bearing self-citations, uniqueness theorems, or fitted quantities renamed as predictions appear in the derivation; parameter fitting and information-criterion comparisons are performed after the model is defined. The fiducial error floor affects numerical competitiveness but does not create a definitional loop.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

The model rests on two fitted parameters per galaxy plus the assumption that the circular-orbit equation plus the stated ansatz fully captures the dynamics. No new particles or forces are postulated; the ansatz is purely phenomenological.

free parameters (2)

v_inf
Saturation velocity scale fitted per galaxy to set the strength of the coupled acceleration.
r_zero
Characteristic radius fitted per galaxy that controls how quickly the coupling saturates.

axioms (2)

domain assumption Stars and gas move on circular orbits so that centripetal acceleration equals the total gravitational acceleration.
Invoked to close the equation relating observed v(r) to the baryonic acceleration plus the new term.
domain assumption The published SPARC baryonic mass decompositions are accurate enough to serve as fixed inputs.
Used without re-deriving uncertainties or alternative decompositions.

pith-pipeline@v0.9.0 · 5615 in / 1630 out tokens · 28875 ms · 2026-05-16T19:48:18.496154+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

11 extracted references · 11 canonical work pages

[1]

1974,IEEE Transactions on Automatic Control, 19, 716

Akaike, H. 1974,IEEE Transactions on Automatic Control, 19, 716

work page 1974
[2]

1995, ApJL, 447, L25

Burkert, A. 1995, ApJL, 447, L25

work page 1995
[3]

W., Lang, D., & Goodman, J

Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, PASP, 125, 306

work page 2013
[4]

2010,Communications in Applied Mathematics and Computational Science, 5, 65

Goodman, J., & Weare, J. 2010,Communications in Applied Mathematics and Computational Science, 5, 65

work page 2010
[5]

S., & Schombert, J

Lelli, F., McGaugh, S. S., & Schombert, J. M. 2016, AJ, 152, 157

work page 2016
[6]

S., Lelli, F., & Schombert, J

McGaugh, S. S., Lelli, F., & Schombert, J. M. 2016, Physical Review Letters, 117, 201101

work page 2016
[7]

1983, ApJ, 270, 365

Milgrom, M. 1983, ApJ, 270, 365

work page 1983
[8]

F., Frenk, C

Navarro, J. F., Frenk, C. S., & White, S. D. M. 1997, ApJ, 490, 493

work page 1997
[9]

1996, MNRAS, 281, 27

Persic, M., Salucci, P., & Stel, F. 1996, MNRAS, 281, 27

work page 1996
[10]

1978,The Annals of Statistics, 6, 461

Schwarz, G. 1978,The Annals of Statistics, 6, 461

work page 1978
[11]

2020, Nature Methods, 17, 261

Virtanen, P., et al. 2020, Nature Methods, 17, 261

work page 2020

[1] [1]

1974,IEEE Transactions on Automatic Control, 19, 716

Akaike, H. 1974,IEEE Transactions on Automatic Control, 19, 716

work page 1974

[2] [2]

1995, ApJL, 447, L25

Burkert, A. 1995, ApJL, 447, L25

work page 1995

[3] [3]

W., Lang, D., & Goodman, J

Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, PASP, 125, 306

work page 2013

[4] [4]

2010,Communications in Applied Mathematics and Computational Science, 5, 65

Goodman, J., & Weare, J. 2010,Communications in Applied Mathematics and Computational Science, 5, 65

work page 2010

[5] [5]

S., & Schombert, J

Lelli, F., McGaugh, S. S., & Schombert, J. M. 2016, AJ, 152, 157

work page 2016

[6] [6]

S., Lelli, F., & Schombert, J

McGaugh, S. S., Lelli, F., & Schombert, J. M. 2016, Physical Review Letters, 117, 201101

work page 2016

[7] [7]

1983, ApJ, 270, 365

Milgrom, M. 1983, ApJ, 270, 365

work page 1983

[8] [8]

F., Frenk, C

Navarro, J. F., Frenk, C. S., & White, S. D. M. 1997, ApJ, 490, 493

work page 1997

[9] [9]

1996, MNRAS, 281, 27

Persic, M., Salucci, P., & Stel, F. 1996, MNRAS, 281, 27

work page 1996

[10] [10]

1978,The Annals of Statistics, 6, 461

Schwarz, G. 1978,The Annals of Statistics, 6, 461

work page 1978

[11] [11]

2020, Nature Methods, 17, 261

Virtanen, P., et al. 2020, Nature Methods, 17, 261

work page 2020