Dwarf Galaxies and Low-Acceleration Systems

Scrollable formula. Use Left and Right arrow keys to reveal hidden content.

Scrollable table. Use Left and Right arrow keys to reveal additional columns.

Scrollable figure. Use Left and Right arrow keys to inspect hidden content.

Dispersion-supported dwarf galaxies reach very low internal accelerations with relatively few luminous tracers. They are therefore sensitive to dark matter and to modified dynamics, but also to velocity anisotropy, unresolved binaries, foreground membership, tidal disturbance, and departures from equilibrium. A line-of-sight velocity dispersion is not itself a measurement of a gravity law. The test begins only after a phase-space distribution and selection model connect the gravitational potential to the observed stellar sample. The evidence comparison is current through 2026-08-19.

Required background. Classical Field Theory supplies potentials and source equations, while Probability and Statistical Inference supplies mixture likelihoods, hierarchical priors, covariance, and coverage checks.

Helpful background. Weak-Field Gravity supplies the nonrelativistic metric limit, and Black-Hole Astrophysics supplies stellar-population and galaxy-evolution context without replacing the gravity inference.

Jeans dynamics and the mass–anisotropy degeneracy

For a spherical, stationary tracer density $\nu_\star(r)$ with radial velocity dispersion $\sigma_r(r)$ , the second velocity moment obeys

\frac{1}{\nu_\star}\frac{d(\nu_\star\sigma_r^2)}{dr} +\frac{2\beta_{\rm ani}\sigma_r^2}{r} =-\frac{d\Phi}{dr}, \qquad \beta_{\rm ani}=1-\frac{\sigma_\theta^2+\sigma_\phi^2}{2\sigma_r^2}.

Here $\beta_{\rm ani}=0$ is an isotropic velocity distribution, positive values favor radial motion, and negative values favor tangential motion. The tracer density counts the observed stellar population; it is not the total gravitating density. For projected radius $R$ , its surface density and observed second moment are

\begin{aligned} I_\star(R)&=2\int_R^\infty \frac{\nu_\star(r)r\,dr}{\sqrt{r^2-R^2}},\\ I_\star(R)\sigma_{\rm los}^2(R) &=2\int_R^\infty\left(1-\beta_{\rm ani}(r)\frac{R^2}{r^2}\right) \frac{\nu_\star(r)\sigma_r^2(r)r\,dr}{\sqrt{r^2-R^2}}. \end{aligned}

These equations assume no bulk rotation, spherical symmetry, and a finite, convergent tracer distribution. A Jeans solution also requires an outer condition, usually $\nu_\star\sigma_r^2\to0$ where the tracers end. The projection weights radial and tangential motions differently, so distinct potentials and anisotropies can produce nearly the same measured profile. Near the deprojected half-light radius $r_{1/2}$ , this sensitivity is reduced for commonly observed profiles, yielding

M(r_{1/2})\simeq \frac{3r_{1/2}\langle\sigma_{\rm los}^2\rangle}{G},

where the average is luminosity weighted over the tracer population, not an unweighted average of a selected spectroscopic sample. The reduction assumes a fairly flat projected dispersion around the half-light radius and a physically admissible anisotropy profile. With $R_e$ the projected half-light radius, $r_{1/2}\simeq4R_e/3$ for many common light profiles; using $R_e$ in place of $r_{1/2}$ with the same coefficient underestimates this mass by about one quarter. These conditions are explicit in Wolf et al. 2010, §§2–3 and Eqs. (1)–(2). It is an enclosed Newtonian dynamical mass within that domain, not a direct measurement of total halo mass or an alternative theory’s source mass. Flattening, rotation, velocity gradients, multiple populations, and tides can move the preferred radius or invalidate the spherical stationary reduction.

For an illustrative dwarf with $r_{1/2}=300\,\mathrm{pc}$ and $\langle\sigma_{\rm los}^2\rangle^{1/2}=6\,\mathrm{km\,s^{-1}}$ , using $G=4.301\times10^{-3}\,\mathrm{pc}\,(\mathrm{km\,s^{-1}})^2M_\odot^{-1}$ gives $M(r_{1/2})\simeq7.5\times10^6M_\odot$ . A $10\%$ dispersion bias produces a $21\%$ mass bias because the estimator is quadratic. Repeat spectroscopy that controls binaries can therefore matter more than a modest increase in the number of single-epoch stars.

A complete modified-gravity calculation replaces the right-hand side with the acceleration predicted from the luminous distribution, environment, and interpolation or screening law. A dark-matter analysis supplies a halo distribution and its population prior. Both must predict the same projected velocities through the same tracer and membership model.

Versioned stars, membership, and likelihood

A numerical result pins a named spectroscopic release containing positions, epochs, line-of-sight velocities, uncertainties, instrument configurations, and membership diagnostics, plus a versioned photometric surface-brightness product. If Gaia astrometry is used, this page’s evidence boundary is Gaia DR3 and its documentation release 1.3, together with stated known-issue filters European Space Agency and Gaia/DPAC 2023, official DR3 documentation, §8.3.4 and §14.2.2. Gaia DR4 is scheduled for December 2026 and therefore is not data at this page’s cutoff European Space Agency 2026, Gaia mission release status. The exact archive query, retrieval date, source identifiers, and cross-match method belong with any rerun.

For star $i$ , a useful mixture likelihood is

p(d_i\mid\Theta)=q_i\,p_{\rm mem}(v_i,\mathbf x_i,\mathbf z_i\mid\Phi,\beta_{\rm ani},B) +(1-q_i)\,p_{\rm fg}(v_i,\mathbf x_i,\mathbf z_i),

where $q_i$ is a prior component weight from the population and selection model, $\mathbf z_i$ denotes metallicity or gravity-sensitive spectral diagnostics, and $B$ is a binary-star population model convolved across repeat epochs. Posterior membership is obtained from this mixture after observing the data; it must not be inserted back as a fixed $q_i$ and multiplied by the same data likelihood again. Conditional weights based on external information are possible, provided that information is not also counted in the component densities. The normalized likelihood includes the spatial and magnitude selection function, spectroscopic targeting, signal-to-noise and quality cuts, radial coverage, repeat-observation cadence, and covariance in calibration zero points.

Instrument calibration includes wavelength solution, velocity standards, slit or fiber centering, template mismatch, epoch offsets, and astrometric reference frames. Systematics include foreground tails, unresolved binaries, perspective rotation, stellar population gradients, anisotropy, flattening, tidal debris, disequilibrium, the Milky Way external field, and halo or modified-gravity population priors. Priors, credible or confidence interval convention, sampler diagnostics, and posterior-predictive velocity distributions versus radius and stellar population must be reported.

Results and limitations through 2026-08-19

Position or evidence	Result or claim	Limitation or counterevidence
Equilibrium Jeans models often infer enclosed masses larger than the visible stellar mass	Supports substantial unseen gravitating mass under the weak-field GR map	Anisotropy, membership, binaries, tides, shape, and halo priors broaden inner-slope and total-mass conclusions
Modified-dynamics models can enhance the baryonic field at low acceleration	Some dwarf dispersions can be described without a conventional halo	External-field prescriptions, stellar mass, distance, equilibrium, and satellite environment are model-dependent
Disturbed or contaminated systems can exhibit inflated or structured dispersions	Supplies a non-gravitational explanation for some apparent anomalies	It does not generically explain dynamically regular systems and must predict spatial, chemical, and epoch signatures

The conservative position is that many well-studied dwarfs require more gravity than their visible stars provide under equilibrium Newtonian dynamics, naturally represented by dark halos. The competing modified-gravity position predicts an enhanced response to the same baryons, often with an environmental dependence. A third possibility—disequilibrium or contamination—must be tested rather than used as an unrestricted escape. Present data do not establish a unique force law, and cusp–core or low-acceleration conclusions remain unresolved and model-dependent.

Adversarial recovery tests

An independent check generates equilibrium mock dwarfs with known potential and anisotropy, forward-models membership, binaries, foregrounds, cadence, and errors, and verifies interval coverage for $M(r_{1/2})$ . A separate disturbed mock tests failure: if the equilibrium pipeline returns a sharp gravity constraint for a tidally structured system, its uncertainty is not calibrated. Held-out diagnostics include repeat-epoch velocities, outer radial bins, metallicity subpopulations, and proper-motion trends.

An adversarial analysis repeats the fit with flexible anisotropy, alternative membership and binary models, axisymmetric or distribution-function methods, and explicit tidal or external-field hypotheses. A claimed gravity difference should persist in posterior-predictive distributions of velocity, radius, and population—not only in one global dispersion. Machine-readable catalogs, query, cuts, calibration, code version, seeds, and mock-generation recipe are the reproduction boundary.

What could change the conclusion. The interpretation could change when repeat spectroscopy changes the binary or membership model, a superseding Gaia release changes proper motions and cross-matches, a validated nonequilibrium method succeeds on blind mocks, or multiple stellar populations produce a force-law discriminator robust to anisotropy and tides. A smaller statistical error alone is not decisive if the distribution-function assumptions remain unchanged.

Exercise

Why does measuring only a flat projected velocity-dispersion profile fail to determine both the mass profile and velocity anisotropy uniquely?

Solution

The Jeans equation contains both the gravitational derivative $d\Phi/dr$ and the anisotropy function $\beta_{\rm ani}(r)$ . Projection combines radial and tangential dispersions into the same line-of-sight moment. Increasing the mass can be offset by changing orbital anisotropy so that the projected profile remains nearly unchanged. Therefore one observed second-moment function cannot determine two unknown functions without additional distribution-function assumptions, higher moments, proper motions, or multiple tracer populations.

References

European Space Agency (2026), “Gaia,” mission status and data-release schedule (accessed 2026-08-19). Official mission page.
European Space Agency and Gaia Data Processing and Analysis Consortium (2023), Gaia Data Release 3 Documentation, release 1.3, 10 July 2023. Official documentation.
Wolf, J., et al. (2010), “Accurate masses for dispersion-supported galaxies,” Monthly Notices of the Royal Astronomical Society 406, 1220–1237. DOI.