Imaging, Timing, and Orbital Probes Together

An image measures angles, a clock measures times, a spectrum measures energy shifts and plasma state, and an orbit measures a gravitational potential over a finite radial range. These probes become powerful together because they have different degeneracy directions—but only if shared parameters are represented once, epoch-dependent source states remain separate, and calibration or plasma assumptions are not mistaken for geometry. Sagittarius A* is the best-developed example. Instrument-specific gravitational-wave likelihood methods are outside this page’s scope.

Helpful background. Black-Hole Geometry and Mechanics supplies Kerr scales and photon-orbit concepts; Relativistic Matter and Compact Objects supplies stellar-orbit and plasma observables; Black-Hole Astrophysics supplies horizon-scale emission models; Gravitational-Wave Astronomy supplies joint-likelihood and model-selection discipline; Relativistic Cosmology supplies distance and observer-frame conversions.

Why the degeneracies differ

Let $M$ be the physical black-hole mass and $D$ the distance. Horizon-scale imaging primarily measures the angular gravitational radius

$$\theta_g=\frac{GM}{c^2D}, \qquad d_{\rm ring}=\alpha_{\rm em}\theta_g ,$$

where $\alpha_{\rm em}$ maps the spacetime scale to a brightness-ring diameter. For Kerr, the characteristic critical-curve diameter varies relatively weakly with spin, but the emission-ring conversion still depends on viewing geometry, emission, and reconstruction; arbitrary metric departures need not be weak. An astrometric stellar orbit with angular semimajor axis $a_{\rm ang}$ and period $P$ gives, at Newtonian order,

$$\frac{M}{D^3}= \frac{4\pi^2a_{\rm ang}^3}{GP^2}.$$

Radial velocities add a physical velocity scale and help separate $M$ from $D$. A characteristic inner-flow frequency has the different scaling

$$\nu_{\rm obs}= \frac{c^3}{2\pi GM}\, f_\nu(a_\ast,c^2r/(GM),i,\hbox{flow}),$$

where $r$ is a physical radius and $f_\nu$ is dimensionless. This expression applies to the effectively zero-redshift Galactic source; a cosmologically distant source adds a factor $(1+z)^{-1}$ to the observed frequency. Timing can therefore constrain $M$ times a dimensionless dynamical model. Spectra and polarization respond to Doppler boosting, gravitational redshift, inclination $i$, magnetic geometry, optical depth, and the electron distribution. The complementarity is real: imaging constrains $M/D$, angular orbits constrain $M/D^3$, velocities set a physical scale, and timing scales as $1/M$. But spin is not identified by a period unless the emitting radius and pattern speed are known.

One likelihood with shared and local parameters

Write shared parameters as

$$\Theta=(M,D,i,a_\ast,\hbox{metric parameters}),$$

and give each probe and epoch nuisance variables $\phi_k$ for calibration, source state, and emissivity. The joint posterior is

$$p(\Theta,\{\phi_k\}\mid\{d_k\}) \propto \pi(\Theta)\prod_k \pi_k(\phi_k)\, \mathcal L_k[d_k\mid F_k(\Theta,\phi_k)] .$$

The stellar-orbit mass–distance posterior must enter once. Multiplying an EHT posterior that already used the orbit prior by the same orbit likelihood double counts it. Likewise, fitting 2017 EHT data and 2018 infrared flares with one immutable accretion state manufactures precision. A correct model can share $M$, $D$, and the metric while allowing density, electron heating, magnetic flux, and transient pattern to change.

Calibration terms are probe-specific. EHT visibilities require station gains, atmospheric phase handling, interstellar scattering, baseline coverage, and intrinsic intrahour variability. GRAVITY stellar orbits require the astrometric reference frame, spectroscopic zero points, source confusion, and possible extended mass. Flare astrometry adds centroid bias, polarization calibration, incomplete orbit coverage, and a distinction between material and pattern motion. Spectral fits add absorption, flux scale, electron distribution, and radiative-transfer model. Correlations within each data product belong in its likelihood, not in diagonal error bars.

The Sagittarius A* release set

Freeze the EHT calibrated product 2022-D02-01, released 2022-05-12, DOI 10.25739/m140-ct59, and the contemporaneous multiwavelength product 2022-D02-03, DOI 10.25739/26fq-k306. The public 2017 EHT analysis finds a thick ring with diameter $51.8\pm2.3\,\mu{\rm as}$ at 68% probability and directly models source variability. Crucially, its comparison of the ring to the expected shadow uses independent stellar-orbit information on $\theta_g$ Event Horizon Telescope Collaboration 2022, §§3–5 and 7, Tables 1–2 and Figures 2–5. That comparison is a consistency test, not an independent second measurement of the orbit prior.

For the large-radius anchor, GRAVITY’s S2 analysis fits 14 parameters: $M$, $D$, reference-frame position and motion, six orbit elements, and a Schwarzschild-precession amplitude. It reports $f_{\rm SP}=1.10\pm0.19$, consistent with general relativity, while limiting extended mass within S2’s orbit under the adopted model GRAVITY Collaboration 2020, §§2–3 and Appendix C, Figures 1–3. These data probe radii thousands of horizon scales away and do not directly measure spin.

GRAVITY’s 2018 infrared flares provide a more model-sensitive timing and centroid probe. Three flares showed loop-like centroid and polarization evolution consistent with compact hot spots at roughly $6$–$10\,GM/c^2$ and periods of 33–65 minutes in the circular-orbit ray-tracing formulation GRAVITY Collaboration 2018, §§2–4, Tables 1–2 and Figures 1–3. Partial loops, short spot lifetime, shearing, pattern motion, and magnetic variability prevent treating those periods as clean ISCO clocks.

The polarized EHT product 2024-D02-01, DOI 10.25739/k39c-mh27, adds magnetic-field and Faraday-transfer information. It can reduce the flow-model space, but a spiral polarization pattern is not a metric observable independent of plasma physics.

Identifiability and leave-one-probe-out tests

Synthetic recovery should precede a joint fit. Draw known $\Theta$, generate orbit astrometry and velocities, variable interferometric visibilities, spectra, and timing with realistic calibration covariance, then fit the whole chain. Examine the rank and condition number of the local information matrix and the prior-to-posterior contraction for each parameter. If spin remains correlated with inclination and electron heating, a narrow marginal produced by a hard flow prior is not identification.

The decisive empirical check is leave-one-probe-out prediction. Fit orbit plus spectrum plus timing and predict visibility amplitudes, ring scale, and polarization without EHT data; then rotate the omission. Report whether the held-out result lies in the posterior predictive distribution. Also refit with a common-state assumption removed. Apparent agreement that disappears when one shared prior is counted only once is bookkeeping, not complementarity.

Results and limitations through 2026-08-18

Results after 2026-08-18 are out of scope. The matrix distinguishes calibrated observables from formulation-dependent parameters.

Evidence or result	Shared parameter information	Scope and limitations	Independent or leave-one-out check
S2 astrometry and spectroscopy through 2019	Constrains $M$, $D$, orbit, reference frame, and Schwarzschild precession	Extended mass and frame systematics matter; essentially no direct spin information	Predict $\theta_g$ before using EHT visibilities
EHT product 2022-D02-01	Ring scale and variability test $M/D$ plus emission models	Sparse coverage, scattering, variability, and plasma mapping enter $\alpha_{\rm em}$	Reconstruct with independent imaging families and orbit-prior bookkeeping
2018 GRAVITY flare centroid and polarization loops	Period and angular scale can test inner-flow motion and orientation	Hot-spot lifetime, pattern speed, circularity, and magnetic geometry are uncertain	Fit nonorbital pattern and shearing alternatives
EHT product 2024-D02-01	Polarization restricts magnetic and electron-thermodynamic models	Faraday rotation and source variability can mimic geometry changes	Predict total-intensity data and rotation measures while holding polarization out

Kerr geometry with uncertain plasma remains consistent with this release set. Metric-deviation claims, a precise Sgr A* spin, a unique flow inclination, and identification of flare loops as material at the ISCO remain unresolved. GWTC-5.0 and IceCube releases contain other black-hole populations, not same-object imaging–orbit measurements, so combining them here would not break these degeneracies.

Competing positions and parameter regimes

The main positions are Kerr spacetime with uncertain plasma; a non-Kerr metric; mismatched source states between observing epochs; shared $M$–$D$ or inclination priors driving apparent agreement; and genuinely complementary observables breaking degeneracies. Parameter regime axes include quiescent versus flaring, optically thin versus Faraday-thick, material orbit versus pattern motion, and stationary image versus variable movie. A metric parameter is credible only if it survives plausible movement along the plasma and state axes.

Reproducing the result and testing its robustness

Archive each product code, DOI, checksum, calibration version, observing epoch, and shared-prior provenance. Publish the map from $\Theta$ to every forward model, dataset-specific nuisance priors, synthetic recovery, posterior correlations, and every leave-one-probe-out prediction. Refit with orbit information removed from the EHT stage before assessing joint information gain.

What could change the conclusion. The interpretation could change after a new stellar orbit, horizon-scale image, polarimetric or timing release; a truly simultaneous imaging–flare campaign; a plasma model that changes held-out predictions; or a statistically significant inconsistency that remains after shared priors and calibration covariance are represented once.

Formulation-choice diagnostic

Suppose imaging measures $q_1=M/D$ and a purely angular Kepler orbit measures $q_2=M/D^3$, both precisely. A timing peak is also observed, but its emitting radius is unknown. Which quantities are identifiable without using the timing peak?

Solution

The two geometric combinations determine

$$D=\left(\frac{q_1}{q_2}\right)^{1/2}, \qquad M=q_1D .$$

Thus $D$ and $M$ are separately identifiable in this idealized model. The timing peak can then test a dimensionless frequency $2\pi GM\nu/c^3$, but it does not identify spin until a physical model fixes or constrains emitting radius, pattern speed, inclination, and flow state. If the image conversion $\alpha_{\rm em}$ is also unknown, imaging actually measures $\alpha_{\rm em}M/D$ and the clean solution acquires that nuisance degeneracy. The correct formulation therefore fits $\alpha_{\rm em}$ and the timing source model rather than announcing a precise spin from three mutually conditional summaries.

References

• Event Horizon Telescope Collaboration et al. (2022), “First Sagittarius A* Event Horizon Telescope Results. I. The Shadow of the Supermassive Black Hole in the Center of the Milky Way,” The Astrophysical Journal Letters 930, L12, doi:10.3847/2041-8213/ac6674; calibrated data product 2022-D02-01, doi:10.25739/m140-ct59.
• Event Horizon Telescope Collaboration et al. (2022), “First Sagittarius A* Event Horizon Telescope Results. II. EHT and Multiwavelength Observations, Data Processing, and Calibration,” The Astrophysical Journal Letters 930, L13, doi:10.3847/2041-8213/ac6675; multiwavelength data product 2022-D02-03, doi:10.25739/26fq-k306.
• Event Horizon Telescope Collaboration et al. (2024), “First Sagittarius A* Event Horizon Telescope Results. VII. Polarization of the Ring,” The Astrophysical Journal Letters 964, L25, doi:10.3847/2041-8213/ad2df0; polarized data product 2024-D02-01, doi:10.25739/k39c-mh27.
• GRAVITY Collaboration (2018), “Detection of Orbital Motions near the Last Stable Circular Orbit of the Massive Black Hole Sgr A*,” Astronomy & Astrophysics 618, L10, doi:10.1051/0004-6361/201834294.
• 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:10.1051/0004-6361/202037813.