Relativistic Orbits Around Massive Black Holes

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.

Stars orbiting a massive black hole turn astrometry and spectroscopy into a clock for relativistic precession and redshift. The leading Schwarzschild effects are now measurable for S2, but spin and quadrupole information is smaller and more vulnerable to extended mass, stellar perturbations, source confusion, and reference-frame drift. An orbit constrains a spacetime only through a joint trajectory and photon-propagation model. The evidence assessment is current through 2026-08-19.

Helpful background. Weak-Field Gravity supplies post-Newtonian redshift and precession; Probability and Statistical Inference supplies correlated orbit fitting; Black-Hole Geometry and Mechanics supplies Kerr mass, spin, and multipoles; and Black-Hole Astrophysics supplies the nuclear star cluster and extended-mass environment.

Invariant orbital signatures

Let a bound orbit have semimajor axis $a$ , eccentricity $e$ , and central physical mass $M$ . To leading post-Newtonian order,

\Delta\omega_{\rm Schw} =\frac{6\pi GM}{a(1-e^2)}

per radial period, in the weak-field regime $GM/[a(1-e)]\ll1$ and for a test body whose backreaction is negligible. It measures excess apsidal rotation relative to one radial cycle, with the orbit orientation tied to a distant reference frame. The osculating $a,e$ are Newtonian orbital elements at this order; higher-order definitions must be fixed consistently. For a Kerr black hole with $a_*=J/(GM^2)$ , frame-dragging and quadrupole contributions scale as

\Delta_{\rm LT}\sim a_*\left[\frac{GM}{a(1-e^2)}\right]^{3/2}, \qquad \Delta_Q\sim a_*^2\left[\frac{GM}{a(1-e^2)}\right]^2,

with coefficients depending on inclination and which nodal or apsidal angle is used. Their different powers of orbital size and angular dependence are more discriminating than a single fitted “extra precession.”

Spectroscopy adds another combination. After subtracting the systemic and observer motion, the leading redshift near pericenter contains

z\simeq v_{\rm los}+\frac{v^2}{2}+\frac{GM}{r},

in units with $c=1$ , with $v_{\rm los}$ positive for recession. The final two terms are transverse Doppler and gravitational redshift. More fundamentally, for emitter and observer four-velocities $u_e,u_o$ and the connecting photon wavevector $k$ , $1+z=(-k_\mu u_e^\mu)/(-k_\mu u_o^\mu)$ . A full prediction integrates the timelike orbit and the null ray to the observer; it does not attach this local approximation to a Newtonian ellipse. GRAVITY’s S2 analysis fits the relativistic orbit, mass, distance, reference frame, and instrument offsets together GRAVITY Collaboration 2020, §§3.1–3.2, Eq. (1), and Appendices C and E.

Worked scale estimate. Take illustrative S2-like values $M=4.0\times10^6M_\odot$ , $a=1000\,\mathrm{AU}$ , and $e=0.88$ , rather than a new fitted dataset. With $GM/c^2\simeq0.0395\,\mathrm{AU}$ , the leading advance is $0.00330$ rad, or about $11.3$ arcminutes per orbit. At pericenter $r_p=120\,\mathrm{AU}$ , the vis-viva relation gives $v_p^2=GM(2/r_p-1/a)$ . Hence the combined gravitational-plus-transverse-Doppler shift is $z_{\rm rel}\simeq6.38\times10^{-4}$ , equivalent to $cz_{\rm rel}\simeq191\,\mathrm{km\,s^{-1}}$ . These are measurable corrections while $GM/(r_pc^2)\simeq3.3\times10^{-4}\ll1$ : stellar tests probe relativistic gravity around a compact central object without placing S2 itself near the horizon.

The nuclear-orbit analysis examines the actual S2 fitting and covariance problem. Here the emphasis is how additional stars and orbital inclinations could separate mass, spin, and higher multipoles.

Measurement and likelihood layers

The dataset must name the astrometric and spectroscopic release, epochs, observatories, wavelength solution, plate scale, optical distortion, differential-atmospheric response, interferometric baseline calibration, radial-velocity zero point, and covariance. Source selection changes near pericenter because crowding and confusion are time dependent. A quality cut on acceleration or orbit coverage also selects the population used to constrain extended matter.

The forward model includes $M$ , distance $R_0$ , six orbital elements, epoch, relativistic dynamics, photon propagation, reference-frame position and drift, instrument offsets, and a model for distributed mass and perturbing stars. Schematically,

\mathcal L\propto \frac{\exp[-\tfrac12(\boldsymbol d-\boldsymbol m)^TC^{-1} (\boldsymbol d-\boldsymbol m)]}{\det(C)^{1/2}},

with declared priors on spin orientation, extended mass, frame drift, and any deformation parameters. $M$ and $R_0$ are strongly covariant; fixing them from a fit that assumed the tested metric can double-use the null. Robust software should publish time standards, reference coordinates, relativistic integrator, ray model, covariance, data exclusions, and posterior samples.

Systematic uncertainties include extended mass, which commonly drives retrograde apsidal precession while Schwarzschild curvature drives prograde precession, although an arbitrary cusp is not summarized by one sign. Resonant relaxation, stellar encounters, unresolved binarity, winds, blends, and reference-frame error can have structured time dependence. Multiple stars with different semimajor axes and inclinations help because a spacetime multipole has a shared scaling while local perturbers do not.

Results and limitations through 2026-08-19

Evidence or position	Result supported	Limitation or scope
S2 spectroscopic pericenter data	Detect relativistic gravitational plus transverse-Doppler redshift in the fitted orbit	Spectral calibration, source motion, and adopted orbit model remain coupled
S2 multi-epoch astrometry and spectroscopy	Detect Schwarzschild apsidal precession in the central point-mass regime	Extended mass, reference frame, distance, and perturbations bound the interpretation
Spin and higher-multipole forecasts or fits	Identify distinctive nodal, apsidal, and multi-orbit scalings	Present stellar baselines and environmental perturbations do not yet establish a model-independent multipole test

The conservative position is that S2 supports the leading GR orbital predictions in its fitted regime. A competing interpretation can absorb small additional precession into extended mass or frame drift; conversely, a restricted environment prior can make a deformation appear artificially precise. Null evidence for one anomalous-precession parameter does not exclude every non-Kerr metric because mappings differ and may be screened or degenerate.

The adversarial check injects a GR orbit and ray solution into the real cadence and covariance, then varies extended-mass profiles and reference-frame models. Independent orbit and photon-propagation implementations should recover the same invariant precession and redshift combinations. A withheld star or epoch should be predicted. Spin or quadrupole attribution remains unresolved and model-dependent until the shared scaling survives those tests.

What could change the conclusion. The interpretation could change when a superseding astrometric or spectroscopic release, a newly completed short-period orbit, a revised reference frame, or an injection study materially changes recovery of extended mass, spin, or multipole combinations.

Exercise

Two stars have the same eccentricity but semimajor axes $a$ and $2a$ . Compare their leading Schwarzschild precession per orbit and explain why this scaling helps reject a common angular calibration drift.

Solution

The formula is proportional to $1/a$ at fixed eccentricity, so the second star precesses by half as much per orbit. A common angular calibration drift is tied to observing time and reference-frame construction rather than this specific orbital-size scaling. Therefore a joint fit to both stars can distinguish the shared spacetime law from some frame errors. The test still requires their different cadences, orientations, extended-mass perturbations, and covariance to be modeled rather than canceled by assumption.

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.