A nova is a transient astronomical event in which a white dwarf undergoes a thermonuclear eruption in material accumulated from a companion star. The eruption causes a rapid increase in brightness, often sufficient to make a previously inconspicuous system visible through binoculars or even to the unaided eye. Unlike a supernova, a nova normally leaves the white dwarf and its companion intact, allowing the process to occur again. The word comes from the Latin for “new,” reflecting the historical appearance of an apparently new star where none had previously been noticed.
Novae connect the physics of close binary stars with nuclear fusion, explosive mass loss, and the chemical evolution of galaxies. Their eruptions produce expanding gas, sometimes dust, and radiation extending from radio wavelengths to gamma rays. Observations reveal that novae are not simply spherical explosions: interacting outflows, shocks, and the geometry of the binary system strongly influence their appearance. The term classical nova generally denotes a system with one recorded nova eruption, whereas recurrent nova denotes one with multiple recorded eruptions; both are manifestations of the same underlying thermonuclear process.
Recognition and the development of the modern explanation
The historical category of a “new star” included several phenomena now known to have different causes. The brilliant objects observed in 1572 and 1604, associated respectively with Tycho Brahe and Johannes Kepler, were supernovae, not novae in the modern sense. Their importance to the history of astronomy nevertheless lay partly in demonstrating that the stellar heavens could change. Identifying a transient in an ancient record as a nova requires caution because descriptions of position, brightness, and duration are often incomplete.
The development of spectroscopy in the nineteenth century made it possible to distinguish explosive stellar events by their physical signatures rather than appearance alone. Nova Cygni 1876, also called Q Cygni, was among the early novae examined spectroscopically. Later bright examples, including GK Persei in 1901, Nova Aquilae 1918, and DQ Herculis in 1934, supplied increasingly detailed records of changing absorption lines, emission lines, and expanding ejecta. Nova Aquilae 1918 reached approximately visual magnitude , making it one of the brightest well-observed novae.
The status of novae as binary systems emerged from studies of their quiescent remnants. In 1954, Merle Walker established that DQ Herculis is an eclipsing binary with an orbital period of about 4.65 hours. Such observations helped show that an eruption was not the birth of a star but an event in an existing compact stellar system. They also linked novae to the broader class of cataclysmic variable stars, in which a white dwarf receives matter from a companion.
The thermonuclear interpretation developed through theoretical work during the mid-twentieth century. A particularly influential study by Robert Kraft in 1964 connected nova systems with mass transfer onto white dwarfs. Calculations by Sumner Starrfield, James Truran, Warren Sparks, and other researchers in the following decades demonstrated how hydrogen-rich material could ignite explosively under the conditions expected on a compact star. These models accounted for the large energy release without requiring the destruction of the white dwarf.
Space-based observations subsequently transformed the subject. Ultraviolet measurements traced the hot central source and expanding gas; X-ray observations revealed residual nuclear burning and shock-heated material. In 2010, the Fermi Gamma-ray Space Telescope detected high-energy gamma rays from V407 Cygni, establishing novae as particle accelerators. Subsequent detections in systems without a dense red-giant wind showed that shocks within the ejecta themselves are a widespread part of nova physics.
Binary systems and the accumulation of fuel
A nova requires a compact accretor and a supply of fresh fuel. The accretor is usually a carbon–oxygen or oxygen–neon white dwarf: a stellar remnant supported largely by electron degeneracy rather than ordinary thermal pressure. Typical white-dwarf radii are comparable to that of Earth, while their masses are substantial fractions of the mass of the Sun. Their strong surface gravity compresses even a relatively thin accreted layer to high pressure.
In many nova binaries, the companion is a low-mass star that fills its Roche lobe, the region around a component within which matter is gravitationally associated with it in the rotating binary frame. Gas flows through the inner Lagrange point and usually forms an accretion disk around the white dwarf. Frictional and magnetic stresses redistribute angular momentum, allowing gas to move inward. Gravitational energy released during this process powers much of the system’s ordinary optical, ultraviolet, and X-ray emission between eruptions.
Not all nova systems have the same architecture. Some contain evolved subgiant or red-giant donors and have orbital periods of days to years rather than hours. In systems such as RS Ophiuchi, the white dwarf receives matter from an evolved companion, and the eruption subsequently encounters a substantial circumstellar environment. Strongly magnetic white dwarfs can channel incoming gas toward magnetic poles, partially disrupting or preventing the formation of an ordinary disk. These differences affect both the accumulation of fuel and the appearance of the eruption.
The mass-transfer rate is a crucial control parameter. Rates relevant to many nova models lie roughly between and solar masses per year, although the boundaries between different burning regimes depend strongly on white-dwarf mass and composition. At sufficiently high rates, hydrogen burning can become approximately steady rather than eruptive. At lower rates, material accumulates until the base of the envelope reaches conditions suitable for unstable burning.
The ignition mass is not a universal constant. It depends on the white dwarf’s mass, its internal temperature, the accretion rate, and the composition of the incoming material. A more massive white dwarf generally requires less accreted material because its smaller radius produces stronger compression. Representative ignition-envelope masses range from about to solar masses across different nova regimes. Consequently, two systems with similar donors can have very different recurrence intervals, and a system’s behavior can change as its accretion history and thermal state evolve.
Thermonuclear ignition and envelope expansion
The immediate cause of a nova is a thermonuclear runaway in the accumulated envelope. At first, hydrogen fusion proceeds too slowly to counteract compression and cooling efficiently. As the base of the envelope becomes hotter and denser, nuclear reactions accelerate. In many cases, partial electron degeneracy initially weakens the usual stellar thermostat: adding heat does not immediately produce enough expansion to reduce the temperature and stop the acceleration.
Hydrogen burning proceeds predominantly through the carbon–nitrogen–oxygen cycle, rather than through the proton–proton chain that supplies most of the Sun’s present luminosity. Carbon, nitrogen, and oxygen nuclei participate in a sequence that converts hydrogen into helium. The reaction rate is strongly temperature-sensitive over important parts of the ignition process. As temperatures approach or exceed kelvin, proton captures can become faster than some intervening beta decays, leading to the hot CNO regime.
Convection develops as the energy-generation rate rises. It transports heat and freshly produced nuclei through the envelope, and may also contribute to mixing material from the underlying white dwarf into the accreted gas. Short-lived radioactive nuclei, including nitrogen-13 and oxygen-15, can be carried outward before they decay. Their decay energy is then deposited away from the deepest burning layer, helping redistribute energy through the expanding envelope. The amount and timing of core–envelope mixing remain important subjects of multidimensional simulations.
The energy available from hydrogen fusion is ample to lift and accelerate the envelope. Converting hydrogen to helium releases approximately ergs per gram of hydrogen, whereas the gravitational binding energy per gram at a white-dwarf surface is commonly of order to ergs. Only a fraction of the envelope’s hydrogen needs to burn to have a major dynamical effect. Nevertheless, an eruption’s energy is divided among radiation, expansion, acceleration, and the work required to escape the white dwarf’s gravitational field.
Expansion eventually lowers the density, reduces degeneracy, and restores a stronger thermal response. The eruption does not usually consume or disrupt the white dwarf because the burning is concentrated in its surface envelope, not throughout the stellar interior. Some material is expelled; some can remain gravitationally bound and continue burning. Radiation-driven winds may remove additional gas after the initial expansion. The outburst is therefore an extended sequence of ignition, expansion, mass loss, and residual burning rather than a single instantaneous detonation.
Optical light curves and spectral evolution
A nova’s visual brightening is partly a redistribution of radiation into wavelengths accessible to the human eye. Before eruption, the accreting white dwarf and inner disk can be very hot but optically faint. During eruption, an expanded, optically thick envelope forms a much larger effective radiating surface, often called the pseudophotosphere. Near visual maximum, its temperature may be only several thousand to roughly ten thousand kelvin, placing a large fraction of its radiation in the optical band.
Eruption amplitudes commonly range from about 8 to 15 magnitudes, although values outside this range occur. Because the magnitude scale is logarithmic, a change of 10 magnitudes corresponds to a factor of in flux. Amplitude depends on both peak brightness and the system’s quiescent luminosity. A recurrent nova with a luminous giant companion can therefore show a modest visual amplitude despite a powerful eruption.
Novae are traditionally characterized by decline times such as and , the intervals required to fade by two or three magnitudes from maximum in a specified photometric band. Very fast novae can decline by two magnitudes within a few days; slow novae may take months. Light curves can also display plateaus, oscillations, secondary maxima, abrupt dust-related dips, or irregular fluctuations. These structures reflect changing opacity, continued mass loss, shocks, dust formation, and the re-emergence of the underlying binary.
Early spectra frequently contain broad emission lines accompanied by blueshifted absorption, producing a P Cygni profile. The absorbing gas lies in front of the bright continuum source and moves toward the observer. Its velocity is inferred through the Doppler effect. Nova outflows often span several hundred to several thousand kilometres per second, with faster components in some events. Multiple absorption systems can reveal successive outflows rather than a single shell moving at one speed.
An influential spectroscopic classification distinguishes Fe II novae from He/N novae according to their prominent early emission features. Fe II spectra generally have strong singly ionized iron lines, while He/N spectra emphasize helium and nitrogen and often show broader profiles. Some novae change class during their evolution and are termed hybrids. These categories describe observed gas conditions and emitting regions; they are not simply labels for different white-dwarf core compositions. As the ejecta thin, forbidden lines from highly ionized species can dominate, allowing estimates of abundance, density, and ionization state, although clumping and geometry complicate such measurements.
Ejecta, shocks, and the formation of dust
Nova ejecta are rarely smooth spherical shells. Resolved images and velocity profiles reveal rings, equatorial concentrations, bipolar lobes, knots, and fragmented structures. The companion star, orbital motion, rotation, and the timing of mass loss can all influence the geometry. A relatively slow, dense equatorial outflow may be followed by faster material that escapes more readily in polar directions. Different viewing angles can consequently make otherwise similar eruptions appear quite different.
Estimated ejected masses often lie between about and solar masses, with substantial variation and systematic uncertainty. Spectroscopic estimates depend on assumptions about ionization, abundance, and the fraction of the apparent volume actually occupied by emitting gas. Radio estimates likewise depend on geometry and density structure. Some individual determinations fall outside this representative range, so ejecta mass cannot be assigned reliably from the optical decline rate alone.
Shocks form when fast gas overtakes slower ejecta or when an outflow encounters material already surrounding the system. In a strong shock, bulk kinetic energy is converted into heat and particle acceleration. A useful estimate for the temperature behind a strong, nonrelativistic shock in an ideal monatomic gas is
where is the shock speed relative to the upstream gas, is the mean mass per particle, and is the Boltzmann constant. Speeds around kilometres per second can produce temperatures of order kelvin, appropriate for thermal X-ray emission.
If the shocked gas cools rapidly, its energy can emerge mainly at lower photon energies after absorption and reprocessing. Such radiative shocks may contribute appreciably to the optical luminosity of some novae. Their compressed cooling layers also provide dense sites where molecules and dust can form, despite the intense radiation field nearby. Observed correlations between optical and gamma-ray variability in some eruptions support a substantial role for shocks, although not every nova need have the same balance of energy sources.
Dust formation can cause a dramatic optical fading accompanied by increased infrared emission. The grains absorb visible and ultraviolet light and reradiate the energy thermally. Carbon-rich grains, silicates, and other condensates have been identified in different novae, sometimes within the same event. Dust survival requires sufficiently cool, shielded, dense regions, conditions more easily achieved in clumps than in a uniformly expanding flow. Systems such as DQ Herculis and V705 Cassiopeiae illustrate conspicuous dust-related behavior, but many novae produce little detectable dust or never develop a deep optical minimum.
Radiation beyond visible light
Multiwavelength observations separate physical components that can be difficult to distinguish in an optical light curve. Ultraviolet radiation is particularly useful for tracing the hot central source, resonance lines, and the changing opacity of the ejecta. As the expanded photosphere contracts, its effective temperature rises, and the spectral peak moves from optical through ultraviolet wavelengths. The visual brightness can therefore decline while the total luminosity changes much less dramatically.
Many novae enter a supersoft X-ray phase when the ejecta become sufficiently transparent to reveal continued hydrogen burning on the white dwarf. Effective temperatures of a few hundred thousand kelvin produce emission concentrated at soft X-ray energies. The onset and duration depend on the residual envelope, the white-dwarf mass, and absorption by surrounding gas. The phase may last weeks, months, or years. A rapid appearance and short duration often indicate a massive white dwarf with a small remaining envelope, but interpretation requires atmospheric and absorption models.
Harder X-rays usually trace shock-heated gas rather than the white dwarf’s photosphere. In a system with a red-giant donor, a fast nova outflow can collide with the donor’s wind, generating bright X-ray emission. Internal collisions between successive outflows can do the same in more compact binaries. Distinguishing thermal shock emission from other components requires spectra and observations across time; a single X-ray measurement cannot by itself identify the eruption’s full geometry.
At radio wavelengths, expanding ionized ejecta produce thermal free–free emission as electrons interact with ions. A simple shell may initially be opaque to its own radio radiation and later become transparent, producing a characteristic rise and decline that occurs at different times at different frequencies. Real novae frequently depart from this simple picture. Spatially resolved radio observations reveal asymmetric structures, while nonthermal synchrotron emission identifies relativistic electrons moving through magnetic fields associated with shocks.
The detection of gamma rays with energies above 100 megaelectronvolts demonstrates acceleration of particles far beyond the thermal energies of the gas. Proposed emission channels include collisions of accelerated protons with ambient nuclei, producing neutral pions that decay into gamma rays, and radiation from accelerated electrons through inverse Compton scattering or bremsstrahlung. The balance of these mechanisms can differ among events and remains an active research question. RS Ophiuchi’s 2021 eruption was also detected at very-high-energy gamma-ray energies by ground-based observatories, extending the observed particle-acceleration phenomenon beyond the energy range first established by Fermi.
Recurrent novae and the eruption cycle
The distinction between classical and recurrent novae is observational, not a division between two fundamentally different ignition mechanisms. A system is called recurrent when at least two nova eruptions have been recorded. A classical nova may also erupt repeatedly, but its recurrence interval can exceed the duration of reliable astronomical records. Conversely, an eruption can be missed because the system lies near the Sun in the sky, suffers heavy interstellar obscuration, or has a brief and poorly observed maximum.
A first approximation to the recurrence time is
where is the accreted mass required for ignition and is the mean accretion rate. This expression neglects changes in the thermal state, composition, and accretion history, but captures why massive white dwarfs receiving matter rapidly can erupt frequently. Their required ignition envelopes are small. Systems with cooler, less massive white dwarfs or low transfer rates can wait thousands to hundreds of thousands of years between eruptions.
Well-studied recurrent novae illustrate several configurations. RS Ophiuchi contains an evolved donor and has recorded eruptions including those of 1898, 1933, 1958, 1967, 1985, 2006, and 2021. U Scorpii is a short-period system by comparison, with an orbital period of about 1.23 days and recorded eruptions in 2010 and 2022 among its modern events. T Pyxidis erupted repeatedly between 1890 and 1966, then remained without a recorded eruption until 2011, showing that observed intervals need not remain constant.
An extreme case is M31N 2008-12a in the Andromeda Galaxy, which has displayed approximately annual eruptions. Such rapid recurrence indicates an unusually small ignition mass and a high rate of fuel supply. It also permits repeated tests of eruption models on human timescales. Annual recurrence does not mean that all its physical properties can be measured directly: the white-dwarf mass, retained mass, and long-term evolution still depend on interpreting observations through models.
After an eruption, irradiation of the companion and disk can alter mass transfer and the system’s brightness. The disk itself may be disrupted or strongly modified and subsequently re-form. The hibernation hypothesis proposes that some nova binaries pass through extended low-transfer states between eruptions as irradiation fades and the binary evolves. Observations of old nova shells around apparently ordinary cataclysmic variables support links between these populations, but a single universal hibernation sequence has not been established. Long-term evolution involves both nova-driven disturbances and the mechanisms that remove orbital angular momentum.
Nuclear products and chemical enrichment
Novae are sites of nucleosynthesis, but their contribution differs from that of supernovae and the interiors of ordinary stars. Their peak temperatures are sufficient to drive rapid hydrogen burning and sequences of proton captures on light and intermediate-mass nuclei. They do not generally achieve the conditions required for extensive production of iron-group elements. The composition of the accreted envelope, mixing with white-dwarf material, and the duration of the hot phase strongly influence the final isotopic yields.
Carbon–oxygen and oxygen–neon white dwarfs provide different seed nuclei. Ejecta rich in neon can indicate an oxygen–neon substrate, whereas substantial carbon and oxygen enrichment can reflect mixing from a carbon–oxygen core. Nevertheless, observed abundance patterns are not a perfectly direct measurement of the underlying star. Ionization corrections, nonuniform ejecta, and uncertainty in the amount of mixing all affect the interpretation. Nova models consequently combine stellar structure, fluid dynamics, and networks containing many nuclear reactions.
Novae are important potential contributors to the Galactic inventories of nitrogen-15, oxygen-17, carbon-13, and lithium-7. Their total mass contribution is small compared with the Galaxy’s gas reservoir, but unusual isotope ratios can make them influential for particular nuclides. Assessing that influence requires both the yield per eruption and the nova rate over Galactic history. It also requires accounting for the delay between the formation of a binary and its eventual nova activity.
A particularly significant observational development was the identification of beryllium-7 in nova ejecta. Beryllium-7 decays by electron capture to lithium-7, with a laboratory half-life of about 53 days; its effective decay behavior can be modified by ionization. Spectroscopic detections, including those associated with V339 Delphini’s 2013 eruption, provided direct evidence that novae synthesize material that becomes lithium. The precise fraction of Galactic lithium supplied by novae remains dependent on yields, event rates, and models of Galactic chemical evolution.
Other predicted products include sodium-22 and aluminium-26, especially in models involving oxygen–neon white dwarfs. Their radioactive decay offers potential gamma-ray diagnostics: sodium-22 has a characteristic line near 1.275 megaelectronvolts, while aluminium-26 contributes at about 1.809 megaelectronvolts. A prediction of such emission is not equivalent to an established detection from an individual nova. Presolar grains in meteorites provide another possible record through unusual isotope ratios, although assigning a grain uniquely to a nova can require dilution models and must distinguish nova signatures from those of other stellar sources.
Galactic populations, distances, and stellar outcomes
Nova surveys measure both the properties of individual eruptions and the behavior of stellar populations. Only a fraction of eruptions in the Milky Way are discovered, because interstellar dust hides many systems and survey coverage is incomplete. Estimates corrected for these effects generally place the Galactic rate at a few tens of eruptions per year, with substantial uncertainty. Counts in external galaxies offer a complementary approach because many events can be surveyed within a single field, although unresolved backgrounds and seasonal visibility still limit completeness.
The Andromeda Galaxy is particularly valuable because it is nearby and contains a large monitored population. Comparisons among galactic bulges, disks, and galaxies with different stellar populations test how nova properties depend on the age and history of binary systems. Apparent differences in decline rate or spectral class must be interpreted together with detection biases: rapidly fading novae can be missed between observations, while faint eruptions are especially difficult to recover against a bright galactic center.
Historically, novae were investigated as distance indicators through a proposed relation between maximum absolute magnitude and rate of decline, commonly called the maximum-magnitude–rate-of-decline relation. The physical motivation is that white-dwarf mass influences both luminosity and eruption timescale. Modern surveys, however, have established substantial intrinsic scatter and populations of faint, fast novae that do not fit a simple universal relation. Nova distances therefore cannot generally be determined precisely from decline speed alone. Geometric parallax measurements, expansion studies, and well-constrained host-galaxy distances provide essential independent information.
A major evolutionary question is whether repeated nova cycles cause the white dwarf to gain or lose mass. Accretion adds material, but eruptions expel gas and can dredge up some of the underlying star. The retention efficiency depends on white-dwarf mass, accretion rate, envelope composition, and the treatment of winds and mixing. Short recurrence intervals do not by themselves prove sustained mass growth. Later helium-shell flashes may also alter the long-term balance even when some hydrogen-burning cycles retain material.
This uncertainty matters for the origin of Type Ia supernovae. A carbon–oxygen white dwarf that grows under suitable conditions may eventually undergo a destructive thermonuclear explosion, whereas an oxygen–neon white dwarf approaching its limiting mass is generally associated with a different possible outcome, including collapse triggered by electron captures. Recurrent novae are therefore relevant candidate laboratories for studying some routes toward stellar explosions, but they are not all inevitable supernova progenitors. Determining the fate of a particular system requires its core composition, mass-transfer history, retention efficiency, and binary evolution—not simply the observation that it has erupted more than once.