Nova: Thermonuclear Explosions in White Dwarf Binaries

• Novae are thermonuclear explosions on accreting white dwarfs, triggered by hydrogen buildup under degenerate conditions and resulting in high-velocity ejecta.
• They serve as natural laboratories for studying stellar structure, hydrodynamics, nucleosynthesis, and shock-induced gamma-ray emissions across diverse systems.
• Ongoing research uses high-cadence, multiwavelength observations and advanced simulations to clarify ejecta dynamics, recurrence behavior, and binary evolution.

A nova is a thermonuclear explosion on the surface of an accreting white dwarf in a close binary system. In the standard picture, hydrogen-rich material transferred from a companion accumulates under degenerate conditions until a thermonuclear runaway develops, lifts the degeneracy, and ejects part of the envelope at high velocity. Modern nova research treats these eruptions not only as optical transients but as coupled problems in stellar structure, reactive hydrodynamics, nucleosynthesis, shock acceleration, and binary evolution; current open questions center on the mass, composition, geometry, and dynamics of the ejecta, the role of shocks, and whether repeated eruptions grow or erode the white dwarf (Glasner et al., 2011, Ederoclite et al., 17 Dec 2025).

1. Thermonuclear runaway and system taxonomy

In classical novae, the white dwarf accretes from a low-mass companion in a compact binary; in recurrent novae, multiple eruptions are observed on timescales shorter than 100 years; and in symbiotic novae, the eruption occurs inside a symbiotic binary in which the white dwarf accretes from a cool giant companion (Tatischeff et al., 2023, Munari, 2024). The physical engine is the same across these categories: compression and heating of an accreted hydrogen-rich envelope, ignition under degenerate conditions, runaway burning, expansion, and mass ejection. The hot CNO cycle is central to the runaway, and the observed enhancement of heavy elements in nova ejecta requires mixing between the accreted layer and the underlying white-dwarf material (Denissenkov et al., 2012).

The composition of the underlying white dwarf is a major organizing variable. The literature distinguishes novae on carbon-oxygen (CO) white dwarfs from novae on oxygen-neon (ONe) white dwarfs, because the core composition affects the material mixed into the envelope, the temperature evolution, and the nucleosynthetic yields (Denissenkov et al., 2012). The same reviews emphasize that the envelope becomes unstable to convection days to weeks before peak burning, and that during the most explosive phase the envelope is fully convective, allowing processed material to be transported outward (Glasner et al., 2011).

Speed classes are usually parameterized by the light-curve decline times $t_2$ and $t_3$. One review states that fast novae fade by 2 magnitudes in less than about 25 days, while a later white paper uses $t_2<12$ days as a commonly used fast-nova criterion (Glasner et al., 2011, Ederoclite et al., 17 Dec 2025). This variation reflects a broader point: nova classification is operational and observationally useful, but not unique.

The term symbiotic nova has also required clarification. In the modern definition, a symbiotic nova is an otherwise normal thermonuclear nova that erupts within a symbiotic star; older, decades-long brightenings in symbiotic systems are treated separately as SETE (“Symbiotic stars Erupting in Thermal-Equilibrium”) rather than explosive novae (Munari, 2024). This distinction is important because the circumstellar environment in symbiotic binaries radically alters the observational phenomenology without changing the underlying runaway physics.

2. Shocks, ejecta, and high-energy emission

The standard nova model attributes most of the optical luminosity to energy released near the hot white dwarf and reprocessed through the ejecta. Gamma-ray observations have shown that this account is incomplete in at least some systems (Li et al., 2017). In ASASSN-16ma, the optical and gamma-ray light curves were reported to be “remarkably correlated,” with a nearly constant ratio $F_\gamma/F_{\rm opt}\sim 0.002$ while gamma rays were detected. That result implies that the majority of the optical light came from reprocessed shock emission rather than directly from the white dwarf, and it yielded a direct constraint on the non-thermal particle acceleration efficiency of $\sim 0.005$, favoring hadronic models for the gamma-ray emission (Li et al., 2017).

The same work argued that shocks in novae can accelerate particles to energies exceeding 100 GeV and therefore require magnetic field amplification at the shock (Li et al., 2017). This conclusion was reinforced in the recurrent symbiotic nova RS Oph, where very-high-energy gamma rays with energies larger than 100 GeV were detected in the 2021 outburst. In that case, the leptonic inverse-Compton contribution was estimated to be about two orders of magnitude too low, while the spectral curvature from 50 MeV to 250 GeV and the inferred maximum particle energies favored a hadronic origin involving accelerated protons (Tatischeff et al., 2023).

Shock formation can arise in different ways. In ASASSN-16ma and V392 Per, the relevant picture is internal interaction between multiple outflows: a slower early ejection followed by a faster flow that catches up and dissipates kinetic power in radiative shocks (Li et al., 2017, Murphy-Glaysher et al., 2022). In symbiotic novae such as RS Oph and V407 Cyg, the shocks also involve a dense pre-existing red-giant wind. The modern 3D picture proposed for symbiotic novae places hard X-rays, central radio synchrotron emission, permitted optical lines, and likely early gamma rays in an inner equatorial shock region, while faster bipolar lobes propagate toward the poles and later produce extended synchrotron emission and lower-density forbidden-line emission (Munari, 2024).

This shock-dominated interpretation has consequences well beyond gamma rays. It links the optical, radio, hard X-ray, supersoft X-ray, and line-profile evolution into a single dynamical framework, and it shifts attention from purely photospheric descriptions toward the structure and interaction of multiple ejecta components (Li et al., 2017, Munari, 2024).

3. Light curves, spectra, and geometrical diagnostics

Near maximum light, novae have historically been grouped into Fe II and He/N spectroscopic classes. A recent white paper argues that this two-class scheme is largely a by-product of sparse spectroscopic sampling rather than a clean physical dichotomy, because high-cadence and high-resolution monitoring before and after maximum often reveals continuous spectral evolution (Ederoclite et al., 17 Dec 2025). Hybrid novae make this especially clear.

Nova Vul 2024 (= V615 Vul) is presented as a genuine hybrid nova. It showed an early Fe II-type spectrum with very broad emission lines and high-velocity P-Cygni absorptions, then developed He/N features shortly after $t_3$, coincident with the onset of hard X-ray emission and pronounced photometric oscillations, and later entered a high-ionization nebular/coronal phase reaching [Fe VII] and likely Fe X. The same paper states that its ejecta show no evidence of neon overabundance and expand ballistically, as indicated by constant line widths, stable profiles, and castellation in high-resolution emission lines.

V392 Per exemplifies a different but related phenomenology. The 2018 eruption was classified as a fast super-Eddington nova with a plateau-type light curve, a He/N spectral class, large expansion velocities, and triple-peaked emission-line profiles. These profiles were interpreted as an equatorial ring plus bipolar outflow viewed nearly pole-on, with an inclination angle of about $9^\circ$; the stronger redward bump was attributed to a receding outflow with roughly 1.5 times higher density than the approaching outflow (Chochol et al., 2020). A later multiwavelength study of the same eruption argued that multiple and interacting mass ejections drove both the gamma rays and the early optical luminosity, with a slower outflow near $3000~\mathrm{km\,s^{-1}}$ followed by a faster one near $5000~\mathrm{km\,s^{-1}}$ (Murphy-Glaysher et al., 2022).

V1721 Aql demonstrates how extinction and geometry can complicate classification. It was described as a very fast, luminous, and highly extinguished nova with $A_V=11.6\pm0.2$, an H$t_3$0 FWHM of about $t_3$1 early after outburst, and an average expansion velocity of $t_3$2. Morphological modelling with XS5 favored a slightly elongated shell with axis ratio $t_3$3 and a face-on accretion disc, but the paper left unresolved whether the system is a U Sco-like recurrent nova with a sub-giant secondary or a very luminous fast classical nova with a main-sequence secondary (Hounsell et al., 2011).

AT 2017fvz in NGC 6822 illustrates another important issue: the relation between decline rate and peak luminosity. It was a very fast Fe II nova with $t_3$4 d and $t_3$5 d, yet its peak absolute magnitude was only in the range $t_3$6 mag; the authors therefore suggested that it may belong to the “faint and fast” class (Healy et al., 2019). A broader white paper identifies such objects as one reason the maximum-magnitude–rate-of-decline relation is increasingly viewed as unreliable as a distance indicator (Ederoclite et al., 17 Dec 2025).

4. Post-nova evolution and the hibernation problem

A long-standing evolutionary idea is the hibernation scenario: after the eruption, the white dwarf cools, irradiation of the donor declines, the mass-transfer rate $t_3$7 falls, and the accretion disk can enter the dwarf-nova instability regime. Several post-nova systems now provide direct support for this picture (Honeycutt et al., 2011, Kato et al., 2021).

V446 Her is described as the best example of an old nova that developed dwarf nova eruptions in the post-nova state. Over 19 years of photometry, its outburst magnitudes declined at about $t_3$8 mag yr$t_3$9 after correction for blended companions, consistent with the decline seen in other old novae decades after eruption (Honeycutt et al., 2011). The system shows a mean outburst spacing of 18 days with a large range of 13–30 days, confirmed bimodality in outburst amplitudes and widths, and a notable change after late 2003, when the brighter, wider outburst type disappeared. The authors suggested that this reflected falling $t_2<12$0 following the nova (Honeycutt et al., 2011).

V606 Aql extends the same argument to a century timescale. Using ZTF Data Release 6, four dwarf nova outbursts were found between 2018 June and 2021 April, with a typical recurrence cycle of about 270 d, amplitudes of about 1.5 mag, durations of roughly 10 d, and fading rates near $t_2<12$1 mag d$t_2<12$2 (Kato et al., 2021). The transition to a dwarf nova state more than 100 years after the 1899 nova eruption was taken as renewed support for hibernation. The paper also argued that the present low-$t_2<12$3 state is compatible with the historical eruption being a fast nova, and that the He II and Bowen lines seen in quiescence may indicate a massive white dwarf (Kato et al., 2021).

V476 Cyg may represent an even more restrictive case. It is currently a dwarf nova with a cycle length of about 24 days and short 2–3 day outbursts, and a tentative orbital period of $t_2<12$4 d would place it in the period gap (Kato, 2022). If that period is confirmed, the system would be a classical nova remnant in the period gap that is also showing dwarf-nova outbursts, and the author emphasized that the transition appears to have occurred after only about a century rather than the $t_2<12$5 years often supposed for classical novae below the period gap (Kato, 2022).

Not all post-nova systems return promptly to low states. After its 2018 eruption, V392 Per remained in a nova-like high mass-transfer configuration rather than reverting to its pre-nova dwarf-nova low mass-transfer state, and the authors suggested that irradiation of the donor by the nova eruption was driving the sustained high state (Murphy-Glaysher et al., 2022). Post-nova evolution is therefore diverse: some systems enter dwarf-nova behavior, while others remain in enhanced accretion states for years.

5. Recurrent systems, symbiotic environments, and nova super-remnants

Recurrent novae concentrate several of the field’s central questions: recurrence-time physics, white-dwarf mass evolution, and long-term feedback into the surrounding medium. T CrB is among the most intensively studied examples. It is a symbiotic recurrent nova with secure eruptions in 1866 and 1946, and Schaefer’s reconstruction identifies earlier events near December 1787 and October 1217 AD (Sello, 6 Jul 2026). The system reached a peak magnitude $t_2<12$6, has a mean quiescent magnitude of about 9.8, and shows a recurrence pattern near 80 years over the last three well-established intervals (Sello, 6 Jul 2026).

The same study of T CrB is notable for tying the eruption history to detailed binary parameters. It quotes $t_2<12$7, a high-state accretion rate of $t_2<12$8, and an ejected-to-accreted mass ratio of $t_2<12$9, supporting the conclusion that T CrB is consuming its white-dwarf reservoir rather than growing toward Chandrasekhar mass (Sello, 6 Jul 2026). Using a semi-empirical recurrence-time model based on calibrated white-dwarf mass–radius relations and Schaefer’s system parameters, the authors derived a point forecast of 26 February 2027 for the next eruption; the paper explicitly compares this with other forecasts, including Schaefer’s 2025.5 ± 1.3 estimate (Sello, 6 Jul 2026). This suggests a live controversy not about whether T CrB is recurrent, but about how recurrence should be forecast from historical and physical inputs.

On much longer timescales, repeated eruptions may generate nova super-remnants (NSRs). Around the recurrent nova LMCN 1971-08a, a 2025 study reported the first NSR in the Large Magellanic Cloud and only the second extragalactic nova shell identified (Healy-Kalesh et al., 17 Sep 2025). The shell is circular, coincident with the nova, bright in H$F_\gamma/F_{\rm opt}\sim 0.002$0 and [S II], very faint in [O III], and has a diameter of $F_\gamma/F_{\rm opt}\sim 0.002$1 pc. Hydrodynamical modelling indicated a shell mass of $F_\gamma/F_{\rm opt}\sim 0.002$2, expansion at $F_\gamma/F_{\rm opt}\sim 0.002$3, an age of $F_\gamma/F_{\rm opt}\sim 0.002$4 Myr, and roughly $F_\gamma/F_{\rm opt}\sim 0.002$5 eruptions in the model (Healy-Kalesh et al., 17 Sep 2025). The authors argued that the existence of such a structure suggests that LMCN 1971-08a may have a much shorter recurrence period than previously presumed.

Symbiotic novae occupy a special place within this landscape because the giant wind supplies both a dense target for shocks and an immediately ionized circumstellar medium. In RS Oph and V407 Cyg, the flash-ionized wind produces narrow recombination lines and prompt thermal radio emission, while later shock interaction with a density enhancement on the orbital plane generates hard X-rays, optical permitted lines, central radio synchrotron, and likely early gamma rays (Munari, 2024). These systems therefore connect recurrence, circumstellar hydrodynamics, and particle acceleration more tightly than ordinary classical novae.

6. Observation, modelling, and computational infrastructure

Nova research remains limited by incomplete temporal coverage, heterogeneous spectroscopic follow-up, uncertain distances and extinction, and poorly constrained clumpiness and ejecta mass (Ederoclite et al., 17 Dec 2025). The same white paper argues that decisive progress requires rapid-response, high-cadence, multi-wavelength observations, anchored by systematic high-resolution optical and near-infrared spectroscopy from eruption to quiescence. Among the stated requirements are spectral resolution above 50,000 over 300–2400 nm in one observation, high-resolution linear spectropolarimetry over 350–700 nm, integral-field spectroscopy with $F_\gamma/F_{\rm opt}\sim 0.002$6 milliarcsecond spatial resolution, and daily scheduling in both hemispheres (Ederoclite et al., 17 Dec 2025).

At the observationally simplest end of the field, bright novae still reward careful visual methods. During the 2013 outburst of V1369 Cen, a nova bright enough for naked-eye observation, visual estimates made with comparison stars on the same almucantarat and of similar color were reported to achieve accuracy within 0.1 magnitudes; on 9 December, multiple same-altitude comparison pairs agreed within less than 0.1 mag (Sigismondi, 2013). The methodological rationale is direct: stars at similar altitude suffer similar atmospheric extinction, so choosing comparison stars on the same altitude circle reduces differential-extinction errors even when the stars are 20–30 degrees away on the sky (Sigismondi, 2013). The same paper recommended the method not only for bright novae but also for Betelgeuse, Antares, bright Miras, and even a future Galactic supernova.

On the theoretical side, the Nova Framework integrates MESA for 1D stellar evolution and hydrodynamics with NuGrid’s multi-zone post-processing code MPPNP for detailed nucleosynthesis (Denissenkov et al., 2012). Its main innovation is a diffusive convective boundary mixing prescription in which the diffusion coefficient decreases exponentially below the bottom of the convective envelope, motivated by 3D hydrodynamic simulations (Denissenkov et al., 2012). The framework can also reproduce the traditional pre-enriched-envelope approximation, enabling direct comparison of the two approaches. For ONe novae, the diffusive convective-boundary-mixing treatment and the pre-mixed model were reported to be in good agreement for maximum-temperature evolution and nucleosynthetic yields, whereas for CO novae that equivalence did not hold (Denissenkov et al., 2012).

The same modelling environment produced several physically specific results. Under slow accretion, $F_\gamma/F_{\rm opt}\sim 0.002$7, onto cold white dwarfs with $F_\gamma/F_{\rm opt}\sim 0.002$8 K, in situ $F_\gamma/F_{\rm opt}\sim 0.002$9He production can trigger convection before the main nova outburst in both CO and ONe systems (Denissenkov et al., 2012). In one $\sim 0.005$0 ONe model with $\sim 0.005$1 MK and $\sim 0.005$2, the interplay of $\sim 0.005$3He production and destruction produced a thick radiative buffer zone between the convective envelope and the white-dwarf surface (Denissenkov et al., 2012). These results underscore a general point: nova theory is now sufficiently detailed that the field’s central uncertainties are no longer merely whether thermonuclear runaway occurs, but how multidimensional mixing, accretion history, and ejecta structure map onto the observed diversity of light curves, spectra, shocks, and yields.