Micronova: A Thermonuclear Flash Confined to a White Dwarf's Magnetic Pole

What a micronova is

A micronova is a thermonuclear flash on an accreting white dwarf that is localized rather than global. In a classical nova, hydrogen stolen from a companion star piles up over the entire white dwarf surface until it ignites and burns everywhere at once, ejecting a shell and blazing for weeks. In a micronova, a strong magnetic field herds that same hydrogen onto a small cap near one magnetic pole, and only that cap ignites. The burn is over in hours and releases roughly one millionth (~1e-6) of a nova's energy.

The phenomenon was identified in 2022 by Simone Scaringi and collaborators, who reported three such bursts in Nature and coined the term "micronova." All three occurred on magnetic cataclysmic variables - close binaries in which a magnetized white dwarf draws hydrogen-rich gas from a low-mass companion. The events were so brief and faint that no earlier survey had flagged them as eruptions at all.

• Same nuclear physics as a nova: a CNO-cycle hydrogen runaway in a degenerate surface layer - but geographically confined.
• Tiny and fast: ~1e39 erg, a few hours long, brightening by only ~1.5 magnitudes in the optical.
• Needs a magnet: the confinement demands a magnetic white dwarf to both funnel and hold the fuel in place.

Micronovae sit in a genuinely new category: thermonuclear like a nova, but localized like a starspot-sized event, bridging the gap between the surface flashes of white dwarfs and the pinpoint bursts seen on neutron stars.

The mechanism: a magnetically bottled thermonuclear runaway

The key ingredient is the white dwarf's magnetic field. In an intermediate polar (the class TV Columbae belongs to), the field is of order a few million to ten million gauss - roughly 1e6-1e7 G, millions of times Earth's field. That is strong enough to truncate the inner edge of the accretion disk and force the infalling gas to leave the disk and slide down the field lines toward the two magnetic poles, building compact accretion columns and hot polar caps.

Because the accretion is concentrated onto a cap covering perhaps only ~1% of the surface, the fuel piles up locally far faster, per unit area, than it would if it were spread over the whole star. The base of this growing layer is electron-degenerate: its pressure barely responds to temperature. So as more mass compresses the base, temperature and the CNO burning rate climb without the layer expanding to cool itself. When the base pressure reaches the critical value (of order 1e19 dyn/cm^2), hydrogen burning runs away exactly as in a nova.

What keeps it a micronova is the field's second role: magnetic tension resists lateral motion, so the ignited fuel cannot easily spread sideways across the surface. The runaway therefore stays penned inside the polar cap. Only that patch burns, only that patch brightens, and the star sheds little or no mass. Remove the confining field and the same fuel would, in principle, spread and grow into an ordinary nova.

Why the numbers come out so small

The tiny energy budget follows directly from the geometry. A classical nova accretes and burns a shell of ~1e-4 solar masses spread over the whole surface; a micronova burns only what fits on a cap a hundred times smaller in area and much thinner. The measured burst energy, ~1e39 erg, corresponds to fusing only about 1e20 grams of hydrogen - comparable to the mass of a rocky asteroid a few tens of kilometres across - whereas a classical nova accretes an envelope of ~1e29 grams (~1e-4 solar masses) and burns only a fraction of it before ejecting the rest.

• Energy: ~1e39 erg total, about 1e-6 of a classical nova's ~1e44-1e45 erg.
• Peak power: releasing ~1e39 erg over a few hours implies a luminosity of order 1e35 erg/s during the flash.
• Brightening: only ~1.5 optical magnitudes (a factor of ~4), because the cap is small and much of the system's light comes from the disk and hot spot.

The duration is short for the same reason: with so little fuel and such a small burning region, the runaway consumes its supply and the cap cools in hours, not the weeks a nova's massive envelope needs to expand and recombine. The event is, in effect, a nova scaled down by the fraction of the star that participates.

How TESS caught them

Micronovae were discoverable only because of continuous, high-cadence optical monitoring. NASA's TESS (Transiting Exoplanet Survey Satellite) stares at large patches of sky for weeks at a time, sampling brightness every few minutes. Conventional nova surveys are tuned to slow, days-to-weeks brightenings; a burst that rises and fades within a single day slips between their cadences. TESS's uninterrupted light curves revealed a fast spike - a rise over ~an hour or two, a somewhat slower decay - superposed on the flickering of a known cataclysmic variable.

• Light-curve signature: a sharp, roughly hours-long optical burst, distinct from the slow rise of a nova and the days-long plateau of a dwarf nova.
• Host identification: the bursts sat on already-known magnetic systems - TV Col is one of the first intermediate polars ever recognized - tying the events to magnetized white dwarfs.
• Follow-up: ground-based photometry and spectroscopy characterized the systems and their orbits, and continued TESS sectors and future wide-field surveys (such as Vera Rubin/LSST) should catch more and test how often they repeat.

Because the flashes are brief and unpredictable, catching one still relies on luck plus wide, staring coverage - which is why the discovery had to wait for space photometry with minute-scale cadence.

Micronova versus its look-alikes

Placing the micronova next to neighboring outbursts sharpens what is new about it:

• vs classical and recurrent novae: identical hydrogen-burning physics, but a nova ignites the whole surface, ejects a shell at thousands of km/s, and releases ~1e44-1e45 erg over weeks. A micronova ignites only a polar cap, ejects little or nothing, and releases ~1e39 erg over hours - a millionfold weaker.
• vs dwarf novae: a dwarf nova is not a nuclear event at all. It is a thermal-viscous instability that brightens the accretion disk (typically by 2-5 magnitudes for days to weeks) with no burning on the star. A micronova, though comparable in total energy, is thermonuclear and happens on the stellar surface.
• vs Type I X-ray bursts: this is the closest cousin. On a neutron star, accreted hydrogen and helium ignite in localized runaways lasting seconds to minutes and releasing ~1e39-1e40 erg. A micronova is essentially the white-dwarf analog, slower because a white dwarf's weaker gravity and larger radius stretch the timescales.

The upshot: micronovae fill a real gap in the phenomenology of accreting compact stars - a thermonuclear flash that is neither a global nova nor a non-nuclear disk brightening.

Named systems, open questions, and why it matters

The discovery rested on three magnetic cataclysmic variables. TV Columbae (TV Col) is the flagship: a classic intermediate polar a few hundred parsecs away (about a thousand light-years), with a white-dwarf spin period near 1,900 seconds and a ~5.5-hour orbit. The other two, EI Ursae Majoris (EI UMa) and ASASSN-19bh, are likewise magnetic systems, and TESS caught a fast optical burst on each.

Major questions remain open:

• Is magnetic confinement really the cause? The thermonuclear-in-a-polar-cap picture is the leading model, but some researchers ask whether accretion instabilities or magnetic "gating" of the infalling gas could mimic the light curves without a nuclear flash. Distinguishing them needs spectra and X-ray data caught during a burst.
• How often do they recur? The recurrence time on a given white dwarf is essentially unknown - it could be years or much longer - and it is unclear whether every magnetic cataclysmic variable produces them.
• Do they eject mass, and does it matter for the star's fate? If micronovae shed little material, they barely dent the white dwarf's slow growth, which is relevant to whether accreting white dwarfs can eventually reach the Chandrasekhar limit and explode as Type Ia supernovae.

Beyond the specifics, micronovae matter because they show that thermonuclear burning on a star need not be an all-or-nothing, whole-surface affair. A magnetic field can localize a nuclear flash to a single cap - a reminder that even well-studied systems like TV Columbae still hide new physics in their light curves.

Outbursts on accreting compact stars, from the whole-surface detonation of a nova to the pinprick flash of a micronova, plus two look-alikes: the non-nuclear dwarf nova and the neutron-star Type I X-ray burst that a micronova most closely resembles.

Phenomenon	Physical cause	Region involved	Energy	Duration
Classical nova	Whole-surface hydrogen thermonuclear runaway; ejects a shell	Entire white dwarf surface	~1e44-1e45 erg	Weeks to months
Recurrent nova	Same runaway on a near-Chandrasekhar white dwarf	Entire surface	~1e44 erg	Days to weeks (recurs yr-decades)
Micronova	Localized hydrogen runaway, magnetically confined	One magnetic polar cap (~1% of surface)	~1e39 erg	A few hours
Dwarf nova	Accretion-disk thermal-viscous instability (no burning)	The accretion disk, not the star	~1e39-1e40 erg	Days to weeks
Type I X-ray burst (analog)	Localized H/He runaway on a neutron star	Neutron-star surface/cap	~1e39-1e40 erg	Seconds to minutes