Galactic and Astrophysical Dynamics

Galactic and astrophysical dynamics test how visible sources, unseen matter, environment, and a gravitational law combine to produce motion. Rotation-supported disks, dispersion-supported dwarfs, wide binaries, clusters, stellar streams, and nuclear stars occupy different accelerations and geometries, but none measures a force law without a source and selection model. This chapter develops the comparisons that can distinguish explanations and the nuisance structure that can make them look alike. Its frontier assessments are current through 2026-08-19.

Helpful background. Weak-Field Gravity supplies the metric-to-potential limit; Probability and Statistical Inference supplies normalized likelihoods and predictive tests; Classical Field Theory supplies source and field equations; and Black-Hole Astrophysics supplies the baryonic, plasma, stellar, and nuclear environments treated here as nuisance physics rather than gravity conclusions.

Enter the dynamical comparison

In the nonrelativistic limit, motion responds to a potential through $\mathbf a=-\boldsymbol\nabla\Phi$. That compact equation does not say how $\Phi$ is sourced. Under general relativity’s weak-field limit, visible matter plus a dark component can determine it. In a modified theory, the source-to-potential equation, external-field response, or relation between dynamical and lensing potentials can change. A test must therefore specify three distinct layers:

1. the measured phase-space, photometric, gas, or lensing data and their calibration;
2. the tracer distribution, baryonic sources, unseen components, environment, equilibrium, and selection model; and
3. the gravity map that predicts the observables from those sources.

Changing an assumed mass-to-light ratio is not the same as changing the force law. Nor is a good fit unique evidence when each galaxy, cluster, or stream receives enough local flexibility. Predicting held-out radii, stars, stream segments, tracers, or systems under a shared population model is more discriminating than reporting minimum residuals.

A first calculation. Suppose a nearly circular tracer moves at $100\,\mathrm{km\,s^{-1}}$ at $10\,\mathrm{kpc}$. Its centripetal acceleration is $V^2/R\simeq3.2\times10^{-11}\,\mathrm{m\,s^{-2}}$. In a spherical Newtonian model this is summarized by $M(<R)=RV^2/G\simeq2.3\times10^{10}M_\odot$. The acceleration follows from motion and geometry; the inferred mass additionally uses a source–force law and spherical geometry. A disk’s force cannot in general be obtained by placing all enclosed material at its center. This simple distinction recurs throughout the chapter.

The nearest next step depends on the tracer. A rotation curve uses ordered circular motion; a dwarf’s dispersion uses random stellar motion and projection; a cluster’s pressure profile uses support against gravity; a stream uses differential orbital evolution. Before comparing quoted “masses,” ask which of these dynamical assumptions produced each number. The lensing chapter adds a different measurement of the same system through light propagation.

Route through six regimes

The pages proceed from galaxy-scale regularities to local low-acceleration tests and then to multi-tracer and strong-potential systems:

1. Galaxy Rotation Curves and Acceleration Relations relates circular speeds to stellar and gas mass, and compares halo and modified-acceleration predictions after distance, inclination, and sample selection are propagated.
2. Dwarf Galaxies and Low-Acceleration Systems develops Jeans and distribution-function inference under anisotropy, membership, unresolved binaries, tides, and equilibrium assumptions.
3. Wide Binaries and Stellar Tests treats Gaia DR3 pairs as projected orbital snapshots and fairly presents the still-disputed Newtonian and modified-dynamics analyses.
4. Galaxy Clusters and Hydrostatic Dynamics compares X-ray, Sunyaev–Zeldovich, galaxy-dynamical, and lensing reconstructions while retaining nonthermal pressure and geometry.
5. Stellar Streams and Galactic Potentials models a stream as a stripped phase-space distribution perturbed by the bar, disk, satellites, and substructure—not as one orbit.
6. Black-Hole and Nuclear-Orbit Tests reconstructs S2’s relativistic orbit with the full mass, distance, reference-frame, astrometric, spectroscopic, and extended-mass covariance.

These routes have different evidential strengths. SPARC supplies an exceptionally broad resolved-kinematics resource but is not volume complete and its downloadable files are versionless, so retrieval dates and hashes are essential Lelli, McGaugh, and Schombert 2016, §§2–3 and Tables 1–2. Wide binaries are local and largely free of internal dark halos, yet unresolved multiples and gravity-dependent candidate selection are decisive. S2 probes a stronger potential, yet its result comes from a correlated 14-parameter orbit and reference-frame fit GRAVITY Collaboration 2020, §§3.1–3.2, Eq. (1), and Appendices C/E.

Recurring likelihood structure

Across the chapter, the useful schematic is

$$p(d\mid\mathcal M)= \int p(d\mid z,\eta,\mathcal M)\, p(z\mid\eta,\mathcal M)\, p(\eta\mid\mathcal M)\,dz\,d\eta,$$

normalized over the stated selection. Here $z$ represents unobserved three-dimensional positions, velocities, orbital phases, memberships, or stripping histories; $\eta$ represents distance, inclination, mass-to-light ratio, anisotropy, binaries, pressure support, reference frames, perturbing satellites, and population parameters. The gravity model $\mathcal M$ may also affect the latent population or selection. Freezing these quantities at catalog central values can turn shared calibration uncertainty into a coherent apparent force anomaly.

Three validation patterns recur. First, forward simulations must recover a known potential under the real survey response. Second, adversarial nuisance alternatives must fail the signal’s predicted dependence on radius, phase, observable type, or environment. Third, withheld data should be predicted: outer rotation-curve radii, repeat stellar velocities, radial-velocity subsets of wide binaries, a cluster tracer, a stream arm, or S2 epochs. A claim that improves only its fitted summary statistic has not completed the test.

What the chapter does and does not conclude

The rotation-curve and dwarf evidence robustly establishes a gravitational discrepancy relative to visible baryons under a simple Newtonian source model. It does not by itself select dark matter or modified gravity. Cluster and stream analyses show why environment, nonequilibrium, and perturbing structure cannot be compressed into one mass error. Wide-binary analyses currently disagree even when starting from Gaia DR3, demonstrating that significance values from different samples and likelihoods cannot be compared as though they were repeats. Nuclear stellar orbits support relativistic GR predictions in their fitted regime, but do not directly reconstruct a metric or establish higher multipoles.

The most informative future result would be a common, preregistered forward comparison that predicts multiple observables under the same baryonic and selection model. Until then, conclusions should remain conditional on the named data release, source map, environment, likelihood, and cutoff.

Review the chapter

1. Source–law degeneracy. A model predicts a rotation curve poorly. Name two changes that improve the prediction but have different physical meanings.

Solution

One may change the source model, for example by adjusting stellar mass-to-light ratio, gas distribution, or a dark-halo profile. Alternatively, one may change the source-to-acceleration law, such as a modified field equation or interpolation function. Because both can change the same speed prediction, the data must constrain source nuisances independently and test each complete model on held-out galaxies. Otherwise an apparent force-law preference may simply reward extra source flexibility.

2. Disequilibrium. Why is a disturbed cluster a useful control but a poor first object for a precision gravity constraint?

Solution

A disturbed cluster reveals how hydrostatic, galaxy-dynamical, SZ, and lensing estimators fail when mergers, bulk flow, geometry, and nonthermal pressure matter, so it tests the robustness of a pipeline. Those same effects introduce large, structured nuisance freedom that can imitate a dynamical–lensing mismatch. Therefore it is a poor precision target unless a validated time-dependent model predicts all tracers jointly; relaxed and disturbed samples should instead be compared explicitly.

References

• GRAVITY Collaboration (2020), “Detection of the Schwarzschild precession in the orbit of the star S2 near the Galactic centre massive black hole,” Astronomy & Astrophysics 636, L5. DOI.
• Lelli, F., McGaugh, S. S., and Schombert, J. M. (2016), “SPARC: Mass Models for 175 Disk Galaxies with Spitzer Photometry and Accurate Rotation Curves,” Astronomical Journal 152, 157. DOI.

Further reading

• SPARC Team (n.d.), “SPARC,” data portal (accessed 2026-08-19). Official data site.
• Wolf, J., et al. (2010), “Accurate masses for dispersion-supported galaxies,” Monthly Notices of the Royal Astronomical Society 406, 1220–1237. DOI.