An active galactic nucleus (AGN) is the compact central region of a galaxy whose emission is powered primarily by matter accreting onto a supermassive black hole, rather than by its stars. Its luminosity can range from far below that of its host galaxy to thousands of times the luminosity of the Milky Way. Radiation may emerge across the entire electromagnetic spectrum, from radio waves to gamma rays, and some nuclei launch narrow jets extending hundreds of thousands of light-years beyond their galaxies.
Active galactic nuclei include Seyfert galaxies, quasars, radio galaxies, and blazars. These names originally described different observational appearances; many of their differences are now understood in terms of luminosity, accretion state, obscuration, and viewing direction. The underlying engine combines extreme gravity, rapidly moving gas, strong magnetic fields, and conversion of accreted matter’s gravitational binding energy into radiation and mechanical power.
AGN are central to modern astrophysics because they reveal otherwise inaccessible black holes, illuminate the distant universe, and exchange energy with their host galaxies and surrounding gas. Their study links processes near a black hole’s event horizon to the evolution of galaxies and galaxy clusters, while providing tests of accretion physics, relativistic jets, and the growth of cosmic structure.
Discovery and the emergence of the black-hole interpretation
Evidence for unusual galactic nuclei accumulated before their physical nature became clear. In 1908, Edward A. Fath reported bright emission lines in the spectrum of the spiral galaxy NGC 1068. Such lines indicated excited gas rather than an ordinary mixture of stellar spectra. In 1943, Carl K. Seyfert described six galaxies with unusually bright, nearly stellar nuclei and strong emission lines. Some lines were extremely broad, implying gas motions much faster than those typical of the surrounding stars. These objects established an observational class now bearing his name.
Radio astronomy introduced a second route to recognizing nuclear activity. During the mid-twentieth century, sources such as Cygnus A were associated with galaxies whose radio output was extraordinarily powerful. Radio maps eventually revealed extended lobes, demonstrating that the energy source could affect regions much larger than the visible galaxy. Other radio sources appeared optically almost stellar, creating a puzzle: their spectra contained strong lines that initially resisted identification.
The decisive interpretation came in 1963, when Maarten Schmidt recognized the lines of 3C 273 as the familiar spectrum of hydrogen displaced by a redshift of approximately . Its cosmological distance implied an optical luminosity vastly greater than that of an ordinary star. Other “quasi-stellar” sources proved similarly distant. Their rapid brightness variations compounded the problem: a source changing substantially over a short interval must generally be compact, yet these objects emitted the energy of entire galaxies.
Accretion onto massive black holes offered a mechanism capable of combining small size with enormous power. In 1964, Edwin Salpeter and Yakov Zel’dovich independently developed influential accretion-based explanations. Donald Lynden-Bell argued in 1969 that massive black holes remaining in galactic centers could account for quasars and related phenomena. These ideas gradually displaced proposals involving unusually dense stellar systems or repeated stellar explosions as the general explanation for sustained nuclear activity.
Later observations strengthened the interpretation. X-ray astronomy revealed hot, compact emitting regions; radio interferometry traced jets on subparsec scales; stellar and gas dynamics established massive dark central objects. The water-maser disk in NGC 4258 provided especially clean evidence for a compact central mass through nearly Keplerian orbital motion. In 2019, the Event Horizon Telescope published an image of the black-hole shadow in M87, a nearby active galaxy. The approximately -solar-mass black hole connected horizon-scale observations directly to an established radio-galaxy nucleus.
Physical structure from the event horizon to galactic scales
An AGN is not a single uniform object but a hierarchy of interacting regions. At its center lies a black hole, commonly with a mass between about and solar masses, although active black holes below this range are known. A useful scale is the gravitational radius,
where is the gravitational constant, the black-hole mass, and the speed of light. For a black hole of solar masses, is about one astronomical unit. The event-horizon radius and the location of the innermost stable circular orbit depend on the black hole’s spin.
Outside the horizon, inflowing matter commonly forms an accretion disk. In luminous, radiatively efficient AGN, much of the optical and ultraviolet continuum arises in an optically thick disk extending across many gravitational radii. Above or around its inner regions lies a hot corona, whose energetic electrons scatter lower-energy photons into the X-ray band. Neither the corona’s exact geometry nor its heating mechanism is fully established: proposed configurations include compact regions near the rotation axis, patchy structures above the disk, and more extended inner flows.
Gas farther out produces characteristic emission lines. The broad-line region typically occupies light-day to light-month scales in nearby Seyfert galaxies and can extend to light-year scales in luminous quasars. Its permitted lines have widths corresponding to velocities of thousands of kilometers per second. The gas is sufficiently dense that many forbidden transitions are collisionally suppressed. The narrow-line region extends from parsec scales to hundreds or thousands of parsecs, where lower-density gas emits both permitted and forbidden lines with smaller characteristic widths.
Dust survives only where it remains cool enough to avoid sublimation, typically below roughly – kelvin depending on grain composition. This produces a luminosity-dependent inner boundary to the dusty environment, often at subparsec to parsec scales. The traditional “dusty torus” is a useful geometric shorthand, but observations indicate a more complex, clumpy distribution that can include a thick equatorial component and dusty polar outflows. Dust absorbs ultraviolet and optical radiation and reradiates the energy in the infrared.
Jets and winds provide additional structural components rather than universal layers. A relativistic jet can originate near the central engine and remain collimated over enormous distances. Winds may arise from the disk, dusty surroundings, or larger nuclear gas reservoirs. These regions overlap dynamically and radiatively: the disk supplies photons that ionize distant gas, winds can obscure the inner engine, and jets can disturb both the narrow-line region and the surrounding interstellar medium.
Accretion, radiative efficiency, and power limits
The main energy reservoir of a luminous AGN is the gravitational binding energy released as matter moves inward. Gas cannot simply fall radially from a galactic orbit into a black hole because it generally possesses substantial angular momentum. Angular momentum must be transported outward while some gas moves inward. In ionized disks, the interaction of differential rotation with a magnetic field can produce the magnetorotational instability, generating turbulence and magnetic stresses. These stresses enable accretion and help convert orbital energy into heat.
The bolometric luminosity, integrated over all wavelengths, is commonly expressed as
where is the mass inflow rate through the radiating accretion flow and is its radiative efficiency. A value near is often used for estimates. For an ideal thin disk with a stress-free inner boundary, the efficiency is about around a nonrotating black hole and can be much higher for prograde accretion onto a rapidly rotating black hole. The actual efficiency also depends on disk structure, magnetic stresses, winds, and energy transported into the hole or into a jet.
A useful reference luminosity is the Eddington luminosity. For fully ionized hydrogen with opacity dominated by electron scattering,
where is the proton mass and the Thomson scattering cross section. This expression balances outward radiative force on the ionized gas against inward gravity in a simplified spherical setting. The ratio characterizes accretion activity, but the Eddington luminosity is not an inviolable ceiling for nonspherical flows.
Accretion changes character with its rate and physical conditions. At relatively low accretion rates, the gas can form a geometrically thick, optically thin flow in which much of the dissipated energy is carried inward rather than radiated. Such radiatively inefficient flows help explain very faint nuclei containing massive black holes. At intermediate rates, an optically thick, geometrically thin disk often provides a useful approximation. At high rates, radiation can become trapped and advected inward, the disk can thicken, and substantial outflows may develop.
These regimes help explain why black-hole mass alone does not determine nuclear brightness. The approximately four-million-solar-mass black hole at the center of the Milky Way, Sagittarius A*, is extremely faint relative to its Eddington luminosity. Conversely, an efficiently supplied black hole can power a quasar whose emission overwhelms the host galaxy. The amount of available gas, its angular momentum, and the fraction of released energy that escapes as radiation all influence the observed luminosity.
Radiation, emission lines, and observational identification
AGN spectra combine several emission mechanisms. In many unobscured, radiatively efficient nuclei, the optical–ultraviolet continuum contains a broad thermal component often called the “big blue bump.” It is generally attributed to disk regions spanning a range of temperatures. Because the hottest disk emission frequently falls in the extreme ultraviolet, where interstellar absorption hampers direct observation, the full ionizing spectrum must often be inferred from models and from the response of surrounding gas.
The X-ray continuum commonly resembles a power law over a restricted energy range, with a high-energy cutoff associated with the temperature of the corona. Inverse Compton scattering occurs when hot electrons transfer energy to lower-energy photons. Some X-rays illuminate the disk or more distant matter and are scattered or absorbed and re-emitted. The resulting reflection spectrum can contain an iron K-alpha line near kiloelectronvolts for weakly ionized iron and a broad excess around tens of kiloelectronvolts. Reflection produced near the black hole may be strongly broadened by orbital motion and relativistic effects.
Emission-line spectroscopy is particularly effective for identifying nuclear activity. The hard ionizing continuum creates ionization states and line ratios different from those typically produced by hot stars alone. Diagnostic diagrams compare ratios such as with , , or . The widely used Baldwin–Phillips–Terlevich diagrams originate from work published in 1981. They distinguish major excitation regimes, although shocks, chemical abundance, and mixtures of stellar and nuclear emission complicate interpretation.
Different wavelength bands select different parts of the population. Hard X-ray surveys penetrate substantial gas columns, but even they can miss heavily buried nuclei. Mid-infrared observations detect dust heated by accretion, although vigorous star formation can produce competing emission. Compact, high-brightness-temperature radio cores or well-collimated jets can reveal activity when optical signatures are weak. Variability and broad emission lines offer further evidence, especially when a nucleus is faint relative to its host.
No individual diagnostic is universally decisive. A luminous stellar population can dilute an AGN spectrum, host-galaxy dust can obscure the nucleus, and a low-accretion-rate engine may lack conspicuous broad lines or a strong ultraviolet component. Estimates of total radiated power therefore often use bolometric corrections derived from a measured band, with systematic uncertainties arising from source diversity and obscuration. Reliable population studies combine several selection methods rather than treating optical, radio, infrared, or X-ray samples as interchangeable.
Observational classes and unified models
The historical classes of AGN reflect how objects appeared in particular surveys. Seyfert galaxies have conspicuous nuclear emission but usually retain easily visible host galaxies. Quasars are more luminous nuclei whose light can dominate the galaxy. There is no fundamental physical discontinuity between these categories, and the boundary depends partly on historical luminosity conventions. Radio galaxies are identified by prominent radio emission, often including extended jets and lobes. A blazar is a jet-dominated AGN viewed close to the jet axis, producing strongly beamed, rapidly variable radiation.
Optical spectra introduce another major division. Type 1 objects display both broad permitted lines and narrower lines. Type 2 objects show narrow lines without directly visible broad permitted components. Intermediate categories describe spectra in which some broad features are weak or selectively absent. In the simplest orientation-based unified model, both types contain similar central engines, but an obscuring dusty structure blocks the disk and broad-line region along some viewing directions while leaving the larger narrow-line region visible.
A landmark result supporting this picture came in 1985, when Robert Antonucci and Joseph Miller reported broad emission lines in the polarized light of NGC 1068. The broad-line nucleus was hidden in direct light but visible through scattering into the observer’s line of sight. Ionization cones, infrared emission, and X-ray absorption provide additional evidence for anisotropic obscuration. Orientation can therefore transform the apparent class of an otherwise similar engine.
Radio-loud unification also depends on direction. A powerful jet viewed from the side may appear as a radio galaxy, whereas the same kind of system viewed nearer its axis can appear as a radio-loud quasar or blazar. Relativistic beaming enhances emission from material moving toward the observer and diminishes emission from the receding side. Among blazars, flat-spectrum radio quasars generally have conspicuous broad emission lines, while BL Lacertae objects have weak or absent lines relative to their jet-dominated continua. Line appearance can depend on both intrinsic emission and dilution by the bright jet.
Orientation alone does not explain all differences. Some low-accretion-rate nuclei may genuinely lack a conventional broad-line region. “Changing-look” AGN can gain or lose broad optical lines over years, sometimes because of changing accretion luminosity rather than moving obscuration. Radio-loudness depends on jet production, not merely on viewing angle, and the classical Fanaroff–Riley radio morphologies distinguish edge-darkened and edge-brightened large-scale structures rather than simple orientation classes. Unified models are therefore conditional frameworks: they connect objects with comparable intrinsic engines while allowing accretion state, environment, luminosity, and time variability to alter their structure.
Relativistic jets and nonthermal emission
Some AGN channel a substantial fraction of their available power into narrow, oppositely directed relativistic jets. Their bulk speeds can approach the speed of light, with Lorentz factors of several to tens in many powerful sources. Jets begin on scales close to the black hole but can propagate over hundreds of kiloparsecs; giant radio sources can exceed a megaparsec in projected total extent. Their ability to preserve direction over such distances points to sustained energy transport rather than local emission alone.
Magnetic fields are central to leading launching models. In the Blandford–Znajek mechanism, developed in 1977, magnetic flux threading a rotating black hole enables extraction of rotational energy into an electromagnetic outflow. Other mechanisms accelerate material centrifugally along magnetic field lines anchored in an accretion disk. Numerical simulations show that strongly magnetized flows can accumulate enough magnetic flux near the hole to impede inflow intermittently, producing a magnetically arrested disk. The relative roles of spin, magnetic flux, disk thickness, and gas supply remain active research topics.
Jets shine through nonthermal particle populations. Electrons spiraling in magnetic fields produce synchrotron radiation, responsible for much of the radio emission and, in some sources, emission extending to optical or X-ray energies. These electrons can also scatter photons to high energies. The seed photons may be the jet’s own synchrotron radiation or external photons from the disk, broad-line gas, or dusty environment. Hadronic models instead assign a larger role to energetic protons and their secondary particles, but the dominant mechanism depends on the source and observed component.
Very long baseline interferometry sometimes reveals jet features with apparent speeds greater than light. This does not violate special relativity. If an emitting feature moves toward the observer at a small angle, the light-travel-time difference between successive positions compresses the observed interval. For speed at angle ,
The apparent transverse speed can exceed , even though the physical speed does not.
On larger scales, jets terminate or dissipate through shocks, entrainment, and interactions with surrounding gas. They inflate radio lobes containing magnetic fields and relativistic particles. In galaxy clusters, these lobes often coincide with cavities in the hot X-ray-emitting atmosphere. Their mechanical influence can persist after the central radio core fades. The jet’s particle composition—whether dominated by electron–positron pairs, electron–proton plasma, or a changing mixture—is difficult to determine because the most visible radiating electrons need not carry most of the energy or momentum.
Variability, reverberation, and black-hole measurements
Variability provides access to regions too small to resolve directly. An emitting region that changes coherently over an observed timescale has a characteristic size constrained approximately by
for a nonrelativistically moving source, where is its redshift. For a beamed jet, a Doppler factor must also be included. The relation is a causal scale estimate rather than a detailed geometric reconstruction. X-ray variations on hours or shorter timescales in some AGN indicate emission from compact regions near the central black hole.
AGN light curves are usually irregular rather than strictly periodic. Their variability power is distributed over a wide range of timescales, often with more power at low frequencies. Characteristic breaks in X-ray power spectra depend on black-hole mass and accretion state, suggesting connections with accreting stellar-mass black holes. Optical fluctuations may arise from changing disk conditions, reprocessing of high-energy radiation, or variations propagating through the accretion flow. Different mechanisms can dominate at different timescales.
Reverberation mapping exploits delayed responses of surrounding material to changes in the central continuum. When the ionizing source brightens, broad emission lines respond after light has traveled to the line-emitting gas. A measured lag gives a characteristic radius . Combining this with a line-width measure yields a virial mass estimate,
where accounts for geometry, inclination, and gas kinematics. The method assumes that the measured motions are substantially governed by the central gravitational potential.
Campaigns monitoring many epochs established a relation between broad-line-region size and optical luminosity, especially for hydrogen H-beta. This permits “single-epoch” mass estimates from one spectrum using luminosity as a proxy for radius. Such estimates make large quasar studies possible but have considerable scatter and can be biased by orientation, extinction, radiation forces, or outflow-dominated lines. Direct reverberation measurements remain more informative than applying a population calibration to an individual source.
Other echoes probe different structures. Infrared delays trace dust near its sublimation boundary, while short X-ray lags can reveal reflection from the inner disk. Wavelength-dependent optical and ultraviolet continuum delays constrain disk size and temperature structure; some measured disks appear larger than simple thin-disk predictions. Spatially resolved stellar dynamics, gas dynamics, maser disks, and horizon-scale observations provide complementary mass measurements. Their assumptions differ, so agreement among methods is particularly valuable when testing the black-hole interpretation and calibrating indirect techniques.
Fuel supply, duty cycles, and effects on host galaxies
The presence of a central black hole does not guarantee sustained activity. Gas must travel from galactic scales to the inner accretion flow while losing an enormous fraction of its angular momentum. Galaxy mergers can drive strong gravitational torques and funnel gas inward, especially during gas-rich interactions, but mergers are not the only route. Bars, spiral structures, interactions with smaller companions, stellar mass loss, and gravitational instabilities in nuclear disks can also contribute. Different mechanisms operate at different radii and need not deliver gas continuously.
AGN activity is episodic. A nucleus may undergo repeated bright phases separated by faint intervals as its local fuel supply and accretion structure change. Estimates of the cumulative time spent in luminous phases often fall around – years, but this does not imply a single uninterrupted episode. Ionization echoes and extended emission-line clouds can preserve evidence that a nucleus was much brighter tens of thousands of years earlier. Radio lobes likewise retain a record of earlier jet production after rapid nuclear variations have faded.
The energy and momentum returned to galactic gas are called AGN feedback. Radiation can heat gas or exert pressure through electron scattering, atomic transitions, and dust absorption. Disk winds can carry material at thousands of kilometers per second, while some highly ionized X-ray absorbers indicate velocities of order a tenth of the speed of light. Jets transfer mechanical energy through shocks, inflated cavities, sound waves, and turbulence. The coupling efficiency depends on gas density, geometry, cooling, and how easily an outflow escapes through low-density channels.
The clearest large-scale mechanical effects are observed in the hot atmospheres of galaxy groups and clusters. Radio jets inflate cavities visible as depressions in X-ray surface brightness. Their energy is often estimated from the cavity enthalpy, approximately for relativistic contents, where is the surrounding pressure and the cavity volume. Dividing by an estimated age gives a mechanical power. In many systems this power is sufficient, within uncertainties, to offset substantial radiative cooling of the central hot gas.
Effects on star formation are more varied. AGN-driven outflows can remove or heat gas, but detection of an outflow does not by itself prove that star formation has been globally suppressed. Dense gas can survive, cool again, or be compressed into star-forming structures. Observed correlations between black-hole mass and host-bulge properties, including stellar velocity dispersion, establish a statistical connection between central black holes and galaxies. They do not alone determine whether feedback, shared gas supply, repeated mergers, or several processes together produced that connection.
Cosmic demographics and the growth of black holes
AGN surveys measure how accretion activity changes with luminosity, epoch, and environment. A central tool is the luminosity function: the number of objects per comoving volume in a specified luminosity interval. Optical, infrared, radio, and X-ray luminosity functions differ because they select different radiation components and encounter different forms of incompleteness. Converting them into a census of black-hole growth requires corrections for obscuration, host-galaxy contamination, variability, and emission outside the observed band.
Bright quasars were much more common when the universe was younger. Their abundance rises strongly from the present to redshifts around –3 and falls at higher redshift. Less luminous AGN generally show a later peak in abundance, a pattern often called luminosity-dependent evolution or “downsizing.” This is a population statement, not a claim that individual black holes necessarily evolve through a universal sequence. Quasar activity broadly overlaps the era of vigorous cosmic star formation, but the histories are not identical.
A major accounting argument connects the integrated light of AGN to the mass accumulated in black holes. Often called the Soltan argument, after Andrzej Sołtan’s 1982 work, it relates radiated energy to accreted rest mass through the efficiency. In schematic form, neglecting additional mass-loss channels,
where is the cumulative emitted energy per comoving volume and is the retained black-hole mass density. Comparing this estimate with the local black-hole mass density supports the conclusion that radiatively efficient accretion accounts for a substantial part of massive-black-hole growth.
Obscured accretion is essential to this accounting. Gas columns above roughly have a Thomson optical depth of order unity and are conventionally called Compton-thick. Such sources can be faint even in relatively hard X-rays, depending on geometry and photon energy. Population models of the cosmic X-ray background require a mixture of unobscured and obscured AGN, but the abundance and evolution of the most heavily buried objects remain important uncertainties.
Distant quasars show that some black holes attained masses approaching or exceeding a billion solar masses within the universe’s first billion years. Under continuous Eddington-limited growth with efficiency near , the characteristic mass e-folding time is about years. Reaching such masses therefore places demanding constraints on seed mass, feeding continuity, and radiative efficiency. Proposed starting points include remnants of early stars and more massive seeds formed through rapid gas collapse. Quasar absorption spectra also probe the intervening intergalactic medium, making these sources tools for studying the universe’s ionization history as well as objects of study themselves.
Open problems and new observational approaches
Several basic parts of the central engine remain incompletely understood. Thin-disk theory captures many broad properties but does not fully explain observed variability, measured continuum delays, or the partition of power between disk, corona, wind, and jet. The corona’s heating process and geometry remain uncertain. Close to the black hole, magnetic turbulence, radiation transport, relativistic gravity, and plasma physics interact across scales that are difficult to simulate simultaneously. Numerical results depend on both computational resolution and assumptions about how electrons receive energy.
Black-hole spin is another important but challenging quantity. Reflection spectroscopy can infer spin by associating the inner edge of a reflecting disk with the innermost stable circular orbit. Such measurements depend on the ionization structure, illumination pattern, disk density, absorption, and whether the disk actually reaches that orbit. Spin is relevant to accretion efficiency and jet launching, yet the diversity of jet powers demonstrates that spin alone cannot determine the observed outcome. Magnetic flux and accretion geometry must also be considered.
Time-domain surveys are expanding the range of recognized nuclear behavior. Changing-look events test how quickly accretion flows can reorganize. Tidal disruption events, in which a star is torn apart by a black hole’s tidal gravity, can briefly illuminate otherwise quiet nuclei and can also occur in galaxies with pre-existing activity. Distinguishing a disruption from an unusual accretion flare may require long-term spectroscopy and observations at several wavelengths. Repeating flares and candidate periodicities are of interest, but stochastic variability can imitate periodic signals unless observations span enough independent cycles.
AGN also enter multimessenger astronomy. Energetic particles accelerated in jets or nuclear environments may produce neutrinos through interactions with radiation or gas. The 2017 high-energy neutrino event IceCube-170922A was associated with a flaring blazar, TXS 0506+056, providing evidence for a connection between at least some blazars and cosmic neutrinos. Determining which source populations dominate the diffuse neutrino flux requires population statistics, sustained monitoring, and models consistent with the accompanying electromagnetic emission.
Binary supermassive black holes are expected during galaxy mergers, but identifying close binaries electromagnetically is difficult. Double emission-line peaks can arise from ordinary gas dynamics, while apparent periodicity can result from stochastic accretion. Pulsar-timing measurements of nanohertz gravitational waves provide a complementary route to the population of massive binaries, though they do not by themselves establish an AGN counterpart. Horizon-scale radio imaging, infrared interferometry, high-resolution X-ray spectroscopy, polarization measurements, and increasingly detailed simulations probe different parts of this system, allowing physical models to be tested against spatial, spectral, and temporal information at the same time.