Black-Hole Astrophysics

Astrophysical black holes are inferred through matter and radiation. Orbits reveal a dark central mass; accreting plasma supplies spectra, lags, polarization, and jets; stellar and galactic environments determine which black holes form, grow, pair, and merge. None of those measurements is the event horizon itself. The scientific task is to connect data to a black-hole model through a chain whose geometry, source physics, instrument response, selection effects, and alternatives are all visible.

This volume develops that chain from stellar remnants and early seeds to nuclear dynamics, accretion, radiation, outflows, binaries, tidal encounters, demographics, horizon-scale imaging, and multimessenger populations. It emphasizes what is robust, what follows only under a declared source model, and what remained actively disputed at the evidence cutoff of 2026-08-18. Mutable catalog values live in their named releases; the pages teach the equations and inference structure needed to use them responsibly.

Helpful background. Black-Hole Geometry supplies horizon definitions, Kerr scales, mechanics, and photon structure. Relativistic Matter and Compact Objects supplies fluids, transport, stellar structure, collapse microphysics, and the construction of horizonless compact objects. Individual chapters link the additional probability, field theory, orbital dynamics, waves, cosmology, or computational preparation they actually use.

From a spacetime model to an astronomical claim

A useful first separation is among four objects:

1. the gravitational model, such as a Kerr exterior with physical mass $M$ and dimensionless spin $a_*=J/(GM^2)$;
2. the source model, including stars, gas, magnetic fields, electrons, ions, radiation, and their evolution;
3. the measurement model, which maps sky brightness, spectra, arrival times, or motions into calibrated detector data; and
4. the population model, which includes formation rates, host distributions, survey selection, and nondetections.

An observation constrains their composition. It does not normally invert one layer without assumptions about the others. For example, a visibility amplitude samples a Fourier component of sky brightness, while the brightness itself depends on null propagation and radiative transfer through an emitting flow. A broad iron line depends on atomic reflection, disk ionization, illumination, absorption, and relativistic transfer. A compact-binary radial-velocity curve measures a mass function before inclination and donor mass turn it into a compact-object mass. The pages retain those intermediate observables instead of jumping directly to mass, spin, or horizon language.

Three scales organize much of the subject:

$$r_g=GM, \qquad t_g=GM, \qquad L_{\rm Edd}=\frac{4\pi GM}{\kappa}.$$

The first two are the gravitational length and time in the site units $c=1$; the third is the spherical opacity-limited luminosity for flux-mean opacity $\kappa$. Restoring ordinary units gives $r_g=GM/c^2$, $t_g=GM/c^3$, and $L_{\rm Edd}=4\pi GMc/\kappa$. These are reference scales, not universal source sizes, variability periods, or luminosity ceilings. Geometry, opacity, beaming, advection, winds, and time dependence decide how a real source relates to them.

The central discipline is forward modeling. If $d$ denotes calibrated data, $\theta$ the parameters of interest, $\eta$ nuisance parameters, and $S$ the selection event, then an object-level posterior has the schematic form

$$p(\theta,\eta\mid d,S) \propto p(d\mid\theta,\eta,S)\,p(\theta,\eta\mid S).$$

For a population, the selection normalization also depends on the population hyperparameters. A fitted value is therefore inseparable from the data release, calibration, likelihood, priors, source family, and selection model that produced it. That is why a mass or spin quoted by two methods need not be two independent measurements of the same scalar.

The thirteen-chapter path

The chapter order follows the physical and inferential dependencies rather than wavelength or instrument.

1. Identifying Astrophysical Black Holes begins with dynamical mass and compactness, then distinguishes accretion phenomenology, horizon-sensitive evidence, and viable alternative hypotheses.
2. Stellar Origins and Remnant Populations connects pre-collapse structure, winds, metallicity, pair instability, fallback, natal kicks, and binary evolution to black-hole birth distributions.
3. Massive Black-Hole Seeds compares light remnants, direct collapse, and dense-cluster runaways, then asks what rapid growth, low-mass occupation, and high-redshift observations can actually discriminate.
4. Stars and Gas in Galactic Nuclei develops nuclear star clusters, two-body relaxation, mass segregation, cusps and cores, loss cones, relativistic stellar orbits, and gaseous inflow.
5. Accretion Flows and Disks organizes flows by optical depth, thickness, cooling, advection, mass loss, and magnetic flux, from thin disks through hot and magnetically arrested states.
6. Radiation, Spectra, and Variability turns plasma and transfer physics into thermal continua, coronal spectra, reflection, reverberation, quasi-periodic features, and spectral–timing constraints.
7. Magnetospheres, Jets, and Outflows separates force-free energy extraction from magnetic-flux supply, mass loading, acceleration, collimation, disk winds, radiation, and host coupling.
8. Black-Hole Binaries and Multiple Systems follows stellar-mass X-ray binaries and massive black-hole pairs through dynamical friction, stellar and gas hardening, triples, and the transition to gravitational-radiation control.
9. Tidal Disruptions and Stellar Encounters connects tidal scales, debris energies, fallback, relativistic stream intersections, partial disruptions, capture, and extreme-mass-ratio source supply.
10. Growth, Demographics, and Host Galaxies reconstructs mass functions, host correlations, integrated accretion, duty cycles, spin evolution, and feedback without confusing correlation with causation.
11. Imaging, Shadows, Polarization, and Strong-Field Lensing distinguishes geometric critical curves from emission features, then follows the signal through transfer, interferometry, reconstruction, polarization, and variability.
12. Electromagnetic Mass and Spin Inference compares stellar, gas, and maser dynamics with continuum, reflection, timing, reverberation, and polarization methods under shared systematics.
13. Multiband and Multimessenger Black-Hole Astrophysics combines bands, gravitational waves, neutrinos, cosmic rays, tidal events, imaging, timing, orbits, and selection-aware populations.

This order also makes recurring systems do more than serve as anecdotes. Sagittarius A* connects stellar orbits, nuclear dynamics, a low-radiative-efficiency flow, variability, and horizon-scale imaging. M87* connects stellar and gas dynamics, a low-Eddington accretion flow, magnetic flux, a relativistic jet, polarization, and large-scale feedback. Cygnus X-1 connects the binary mass function, accretion states, coronal radiation, timing, jets, and electromagnetic spin inference. NGC 4258 shows how a warped maser disk, distance, and orbital geometry jointly constrain a central mass.

The nuclear-dynamics route and its diffusion boundaries are organized around Alexander 2017, §§ 2–4. The accretion taxonomy follows the regime distinctions and conservation arguments in Abramowicz and Fragile 2013, §§ 5–9, while keeping numerical closures and source-specific interpretation explicit on the relevant pages.

Conventions and quantities that must not be conflated

The site uses metric signature $(-,+,+,+)$, sets $c=\hbar=k_{\rm B}=1$, and keeps $G$ explicit. Thus a physical mass $M$ is not itself a length; $GM$ is. The spin parameter is

$$a_*=\frac{J}{GM^2},$$

with its orientation declared relative to an orbit, disk, or observer when sign matters. Frequencies are reported in the appropriate source, detector, or cosmological frame. For time series using $e^{-i\omega t}$, a positive time delay and Fourier phase are related only after the cross-spectrum ordering and phase-unwrapping convention have been stated. Polarization pages declare Stokes, electric-vector position-angle, screen-basis, and Faraday conventions before interpreting a map.

Several rates recur with different meanings. The mass supplied at large radius, the inward mass flux at radius $r$, the mass crossing the horizon, and the rate retained as black-hole growth can differ because of winds and radiative efficiency. A convenient reference is

$$L=\eta\dot M_{\rm in}, \qquad \dot M_{\rm BH}=(1-\eta)\dot M_{\rm in},$$

for a steady radiatively efficient model without an additional mechanical channel. A paper that defines $\dot M_{\rm Edd}=L_{\rm Edd}$ and one that defines $\dot M_{\rm Edd}=L_{\rm Edd}/\eta$ are using different dimensionless accretion rates even though both may call them $\dot m$.

Likewise, an emission-ring diameter, a lensing critical curve, a brightness depression, and an event horizon are not synonyms. A virial single-epoch mass, a binary mass function, and a dynamical posterior are not interchangeable. An intrinsic merger population is not an observed catalog. A correlation between black-hole mass and a host property is not by itself evidence that feedback created the relation.

A hierarchy of evidence

The pages repeatedly move through the same five questions.

What was measured? The answer names the release, calibration, coordinates or frame, quality selections, response, covariance, and upper limits. In interferometry it is usually a visibility or closure quantity, not an image pixel. In timing it is a cross-spectrum or power estimate with a sampling window, not a unique physical period.

What is the forward model? A model predicts data-space quantities after propagation and instrument response. The likelihood should expose nuisance parameters such as distance, inclination, absorption, stellar anisotropy, disk atmosphere, electron heating, gain calibration, and survey completeness.

Which part is an established mechanism? Kepler’s laws imply a binary mass function; pair creation can soften a radiation-dominated stellar core; diffusion feeds a loss cone; photon trapping changes high-rate accretion; a rotating, magnetized horizon permits electromagnetic energy extraction. These mechanisms do not select a unique source history or observational interpretation by themselves.

What independent check could fail the interpretation? Examples include a distributed-mass fit to a dynamical signal, an alternative continuum beneath reflection, a synthetic interferometric recovery, a full energy and angular-momentum budget, a held-out state transition, or a selection-aware population prediction. Agreement only with the data used to tune a model is weak evidence.

What could change the interpretation? Pages on high-redshift candidates, mass gaps, horizon-scale images, spin populations, merger counterparts, or neutrino associations carry the 2026-08-18 cutoff. A new catalog, calibration, source reclassification, simulation benchmark, or discriminating observation can change their assessment without changing the underlying equations.

This hierarchy follows the distinction emphasized by Cardoso and Pani 2019, §§ 1–2 and 5–6 between compact-object phenomenology and a proof of horizon structure. For stellar remnants, Mapelli 2021, §§ 1.2–1.4 separates stellar, binary, cluster, and selection uncertainties. For early massive black holes, Inayoshi, Visbal, and Haiman 2020, §§ 2–7 supplies the durable formation and growth map while leaving channel selection open. The same principle governs observation: the Event Horizon Telescope’s M87 analysis tests image features across visibility-domain constraints, independent imaging pipelines, and synthetic data rather than treating one reconstruction as raw sky truth (EHT Collaboration 2019, §§ 3–9).

A first route through the physics

If the range of topics is unfamiliar, begin with three questions that use only introductory mechanics and calculus. The binary mass function turns an orbital period and radial-velocity amplitude into a lower bound on a dark companion’s mass. Eddington scales compare gravity with radiation force. Debris fallback uses Kepler’s law and conservation of mass to turn an energy distribution into a time-dependent return rate. Each page develops the calculation and includes a solved exercise.

Then connect the scales to a source. A $10M_\odot$ hole has $r_g\simeq15$ km and $t_g\simeq49$ microseconds; a $10^6M_\odot$ hole has $r_g\simeq1.5\times10^6$ km and $t_g\simeq4.9$ seconds. Both length and time grow linearly with mass. A process at a fixed dimensionless radius can therefore look very fast in an X-ray binary and much slower in a galactic nucleus. These are gravitational scales, not assertions that an observed variability period equals $t_g$.

The vocabulary also separates distinct steps. Accretion is inward mass transport; radiative efficiency is the fraction of supplied rest-mass energy emerging as radiation in a declared model; reprocessing changes radiation after its original emission; selection describes which sources enter a sample. Use the chapter overviews to establish these physical steps before attempting the more advanced inference pages. The mathematical level remains graduate, but the worked routes provide an entry point rather than requiring familiarity with every observing technique.

Reader routes

A first physical-source route is Chapters 1, 2, 5, 6, 7, 11, and 12: identify the object, form it, feed it, turn the flow into radiation and outflows, then infer geometry and parameters. Readers focused on formation and populations can take Chapters 2–4, 8–10, and 13. This route connects stellar and seed channels to nuclei, encounters, binaries, cosmic growth, and survey selection.

For inference and systematics, begin with Chapter 1, then read Chapters 6, 10, 11, 12, and 13. The essential habit is to keep latent source parameters, measured data products, shared nuisance parameters, and selection functions distinct. For computation-facing work, Chapters 2–5, 7–9, and 11 state continuum targets, initial and boundary data, benchmarks, budgets, convergence requirements, and failure modes; production algorithms and executable workflows are developed in Computational Relativity.

Readers can also enter through a recurring object. Follow Sagittarius A* from dynamical mass and stellar orbits to hot accretion, variability, image formation, and cross-probe consistency. Follow an X-ray binary from its mass function and natal kick to state transitions, corona, jet, and spin estimates. Follow a massive black-hole binary from galaxy merger and environmental hardening to counterpart association and selection-aware population synthesis.

Boundaries with neighboring volumes

This volume uses black-hole geometry but does not redefine event, Killing, trapping, dynamical, or isolated horizons. Exact Kerr geodesics, mechanics, and photon regions remain in Black-Hole Geometry and Relativistic Two-Body Dynamics. Here they become parts of stellar-orbit, tidal, image, timing, and source-population models.

Stellar equations of state, supernova transport, compact-star stability, and explicit boson-star or thin-shell construction remain in Relativistic Matter and Compact Objects. Finite-amplitude collapse and horizon formation remain in Gravitational Collapse and Black-Hole Dynamics. This volume begins where those calculations supply a remnant, compact alternative, or environmental initial condition.

Computational Relativity develops production evolution, ray-tracing, transfer, synthetic-observation, and convergence methods. The pages here state the astrophysical question, model inputs, observable diagnostics, benchmark, and failure modes. Likewise, Gravitational-Wave Astronomy develops detector response, searches, event likelihoods, and catalog selection. Here we study electromagnetic environments, formation channels, source associations, and joint astrophysical population interpretation.

Tests of metric theories and current deviation bounds belong in Tests of Gravity; cosmological background evolution belongs in Relativistic Cosmology. Those subjects enter here only where they are needed to interpret a black-hole source, seed population, host relation, or multimessenger measurement.

Check an early-growth claim

A $100M_\odot$ stellar-remnant seed is proposed to reach $10^9M_\odot$ by radiatively efficient accretion. Assume a constant radiative efficiency $\eta=0.10$, Eddington ratio $\lambda=1$, and duty cycle $D=0.80$. Using

$$t_E=\frac{\sigma_T}{4\pi Gm_p}\simeq0.45\ {\rm Gyr}, \qquad M(t)=M_0\exp\!\left[D\lambda\frac{1-\eta}{\eta}\frac{t}{t_E}\right],$$

find the required growth time. Repeat for a $10^5M_\odot$ seed, and state what the comparison does and does not establish.

Solution

The active-accretion e-folding time is

$$t_{\rm efold}=\frac{\eta}{1-\eta}\frac{t_E}{\lambda} =\frac{0.10}{0.90}(0.45\ {\rm Gyr}) =50\ {\rm Myr}.$$

The elapsed time includes inactive intervals, so it is $t=t_{\rm efold}\ln(M/M_0)/D$. For the light seed,

$$t=\frac{50\ {\rm Myr}}{0.80}\ln(10^7) \simeq1.01\ {\rm Gyr}.$$

For the $10^5M_\odot$ seed,

$$t=\frac{50\ {\rm Myr}}{0.80}\ln(10^4) \simeq0.576\ {\rm Gyr}.$$

The calculation exposes the growth burden and its exponential sensitivity to seed mass, efficiency, duty cycle, and Eddington ratio. It does not show that a massive seed formed, that either seed had the assumed gas supply, or that accretion stayed radiatively efficient and Eddington limited. Those are separate formation, environment, and feedback questions developed in Chapters 2–5.

References

• Abramowicz, Marek A., and P. Chris Fragile. “Foundations of Black Hole Accretion Disk Theory.” Living Reviews in Relativity 16 (2013): 1. arXiv:1104.5499.
• Alexander, Tal. “Stellar Dynamics and Stellar Phenomena Near a Massive Black Hole.” Annual Review of Astronomy and Astrophysics 55 (2017): 17–57. arXiv:1701.04762.
• Cardoso, Vitor, and Paolo Pani. “Testing the Nature of Dark Compact Objects: A Status Report.” Living Reviews in Relativity 22 (2019): 4. arXiv:1904.05363.
• Event Horizon Telescope Collaboration. “First M87 Event Horizon Telescope Results. IV. Imaging the Central Supermassive Black Hole.” Astrophysical Journal Letters 875 (2019): L4. arXiv:1906.11241.
• Inayoshi, Kohei, Eli Visbal, and Zoltán Haiman. “The Assembly of the First Massive Black Holes.” Annual Review of Astronomy and Astrophysics 58 (2020): 27–97. arXiv:1911.05791.
• Mapelli, Michela. “Formation Channels of Single and Binary Stellar-Mass Black Holes.” In Handbook of Gravitational Wave Astronomy. Singapore: Springer, 2021. arXiv:2106.00699.