Black-Hole and Nuclear-Orbit Tests

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.

The star S2 follows a highly eccentric orbit around the compact mass associated with Sgr A*, allowing astrometry and spectroscopy to resolve gravitational redshift and in-plane Schwarzschild precession. These observations probe a potential deeper than Solar-System orbits, but they are not a direct measurement of a metric component. A gravity conclusion comes from a joint orbit, photon-propagation, instrument, reference-frame, and nuclear-mass model. Extended mass, source confusion, and coordinate-frame drift can imitate pieces of the relativistic signal. The evidence status is current through 2026-08-19.

Required background. Classical Field Theory supplies central potentials and perturbing sources, while Probability and Statistical Inference supplies joint orbit likelihoods, priors, covariance, and model comparison.

Helpful background. Weak-Field Gravity supplies the post-Newtonian orbital limit, and Black-Hole Astrophysics supplies the Galactic-center stellar, gas, and extended-mass environment.

Relativistic orbit observables

For a test body on a bound Schwarzschild orbit with semimajor axis $a$ and eccentricity $e$ , the leading advance per radial period is

\Delta\omega_{\rm GR}=\frac{6\pi GM_\bullet}{a(1-e^2)}

in units with $c=1$ . Restoring SI units adds $c^2$ in the denominator. The GRAVITY analysis introduces a scale $f_{\rm SP}$ multiplying the first-post-Newtonian Schwarzschild terms, so $f_{\rm SP}=0$ represents the chosen Newtonian-plus-propagation baseline and $f_{\rm SP}=1$ the GR prediction. This is a phenomenological coordinate within that model, not a universal deformation parameter for every alternative theory.

The data combine sky-plane astrometry with line-of-sight spectroscopy. The orbit model includes Rømer delay, special-relativistic transverse Doppler effect, gravitational redshift, and relativistic dynamics. A central mass and distance are strongly correlated; the conversion between angular and physical orbit size makes the reference frame consequential. Extended stellar remnants or a second massive object can add Newtonian precession, typically with a different radial and phase dependence.

Why mass and distance must be fitted together. Let $\alpha_a$ be the measured angular semimajor axis in radians, so $a=R_0\alpha_a$ . At Newtonian order, the period $P$ gives

\frac{M_\bullet}{R_0^3} =\frac{4\pi^2\alpha_a^3}{GP^2}.

Astrometry and period therefore determine a mass–distance combination, not either separately. After fitting inclination and eccentricity, the spectroscopic velocity scale instead gives $M_\bullet/R_0$ through $v^2\sim GM_\bullet/a$ . The two kinds of data jointly resolve the physical scale. At fixed angular orbit and period, shifting $R_0$ by one percent would shift the astrometry-only mass by roughly three percent. Fixing a distance or mass from that same orbit as an independent prior would conceal this covariance.

For the next step—from this measurement problem to the scaling of spin and quadrupole terms—continue to Relativistic Orbits Around Massive Black Holes, which also gives a worked S2-like precession and redshift estimate.

The primary result and its construction are documented in GRAVITY Collaboration 2020, §§3.1–3.2, Eq. (1), Figs. 2–3, and Appendices C/E. The 14 independently fitted quantities were $f_{\rm SP}$ ; $M_\bullet$ and distance $R_0$ ; six osculating elements $a,e,i,\omega,\Omega,t_{\rm peri}$ ; and five reference-frame coordinates $x_0,y_0,v_{x0},v_{y0},v_{z0}$ . The listed period is derived rather than an additional free coordinate. Reducing this fit to “precession divided by a Kepler prediction” discards the correlations that determine the result.

Data, weighting, and systematic model

The named 2020 analysis combines SHARP and NACO adaptive-optics astrometry, NACO S2–Sgr A* flare separations, GRAVITY interferometric separations, and SINFONI plus limited Keck spectroscopy over stated epochs. A reproducible rerun pins the released or tabulated measurements, instrument pipeline versions, interferometric metrology and wavelength calibration, distortion correction, radio-to-infrared frame tie, spectral templates, excluded confusion epochs, and the priors on reference-frame offset and drift.

The likelihood is not one homogeneous Gaussian catalog. The collaboration compared different data selections and weighting schemes, used a soft outlier penalty for non-Gaussian flare-position residuals, ran posterior and Markov-chain analyses, and used bootstrapping as a robustness check. Appendix E reports the correlated 14-dimensional posterior. The relevant lesson is not a particular rounded uncertainty: reference-frame position, pericenter time, eccentricity, and semimajor axis covary with $f_{\rm SP}$ , so weights and priors must remain visible.

A schematic normalized likelihood is

p(d\mid f_{\rm SP},\boldsymbol\eta,\mathcal M) \propto p(d_{\rm ast}\mid f_{\rm SP},\boldsymbol\eta,\mathcal M) p(d_{\rm spec}\mid f_{\rm SP},\boldsymbol\eta,\mathcal M) p(d_{\rm frame}\mid\boldsymbol\eta),

with each factor using its declared error or robust-tail distribution and normalization. The nuisance vector includes the remaining orbit, mass, distance, frame, calibration, and extended-mass parameters. Selection includes the observing cadence, pericenter emphasis, flare availability, confusion cuts, and instrument-specific exclusions. It is a deliberately monitored star, not a complete sample of nuclear orbits.

Results and limitations through 2026-08-19

Position or evidence	Result or claim	Limitation or counterevidence
S2 astrometry and spectroscopy prefer the modeled relativistic precession over the corresponding nonprecessing baseline	Supports the GR first-post-Newtonian orbit in the fitted central-potential regime	Reference-frame drift, pericenter timing, weighting, and source confusion share covariance with the scale parameter
Independent Keck-centered spectroscopy and astrometry detect compatible relativistic redshift behavior	Reduces dependence on one telescope and frame realization Do et al. 2019, Figs. 2–4 and Methods	Data sets are not identical precession measurements and share the same astrophysical source
Extended mass or alternative central potentials can perturb the orbit	Multi-star, multi-period data can distinguish radial and phase patterns	Current constraints remain conditional on the nuclear cluster, perturbers, and theory-specific photon propagation

The conservative conclusion is that the full fitted S2 dataset supports the GR Schwarzschild-precession prediction over the chosen Newtonian baseline. The strongest competing explanations involve reference-frame or astrophysical perturbations, not an established alternative metric. Existing robustness checks disfavor simple versions of those explanations, but mapping $f_{\rm SP}$ to a complete modified theory remains model-dependent. Higher multipoles, frame dragging, and distributed mass are not settled by this one precession measurement.

Independent and adversarial checks

An independent check injects relativistic redshift and precession into blinded synthetic astrometry and spectroscopy, applies the real cadence, heteroscedastic errors, confusion cuts, and frame drift, and recovers $f_{\rm SP}$ with a separately implemented orbit and photon-propagation code. It then injects extended mass and verifies that its radial and phase signature is not misidentified as Schwarzschild precession.

An adversarial fit varies robust weighting, reference-frame priors, astrometric subsets, flare treatment, extended mass, and coordinate parameterization. Invariant predictions—redshift and sky position at held-out epochs, pericenter-to-apocenter phase structure, and consistency among stars—must agree even if fit coordinates change. The pipeline, inputs, masks, covariance, priors, and injection products define the reproduction boundary.

What could change the conclusion. The interpretation could change after new pericenter coverage, a materially improved infrared–radio frame tie, another short-period star supplies an independent radial lever arm, a robust extended-mass detection changes the Newtonian precession, or a blinded independent orbit code finds a persistent invariant mismatch. A tighter $f_{\rm SP}$ interval alone does not establish higher multipoles or identify an alternative metric.

Exercise

Why can an extended spherical mass distribution not simply be ignored when interpreting the fitted precession of S2?

Solution

Matter distributed inside and around the orbit changes the radial force away from a pure point-mass law and therefore produces Newtonian apsidal precession. Its sign and radial dependence generally differ from the prograde Schwarzschild term, but a finite, noisy orbit can leave covariance between them. Consequently the fit must include or bound extended mass and use phase and multi-orbit information to distinguish it; otherwise part of an astrophysical perturbation could be assigned to the relativistic scale.

References

Do, T., et al. (2019), “Relativistic redshift of the star S0-2 orbiting the Galactic center supermassive black hole,” Science 365, 664–668. DOI.
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.