V1674 Herculis: Extreme Fast Galactic Nova

• V1674 Herculis is a classical nova defined by an ultra-fast optical decline (t2 ~1 day), marking it as the fastest known Galactic nova.
• High-cadence observations reveal a multi-phase rise to maximum and early emergence of binary periodicities, informing models of shock formation and ejecta structure.
• Multiwavelength studies capture prompt GeV emission, rapid coronal development, and complex ejecta geometry, challenging traditional decline-time white dwarf mass estimates.

V1674 Herculis, also designated Nova Herculis 2021 or Nova Her 2021, is a classical nova in Hercules whose 2021 eruption has been characterized in multiple studies as the fastest Galactic classical nova known to date. It combines an ultra-fast optical decline, unusually deep coverage of the rise to maximum, early re-emergence of orbital and spin periodicities, prompt GeV emission, strong supersoft X-ray pulsations, and rapid development of coronal spectroscopy. As a result, it has become a reference object for studies of shock formation, post-nova binary re-establishment, ejecta geometry, and the reliability of decline-time-based white-dwarf mass inferences in extreme fast novae (Quimby et al., 2021, Patterson et al., 2022, Sokolovsky et al., 2023).

1. Discovery, designation, and speed class

V1674 Her erupted on 2021 June 12. It was discovered in eruption on 2021 June 12.537 UT at about 8.4 mag, and later the same day it reached naked-eye brightness. Different analyses adopt closely related optical maxima: one near-infrared study used $V=6.14$ at 2021 June 12.96 UT, while a later reconstruction of the rise derived $V_{\rm max}=6.19 \pm 0.05$ around $t_{\rm ref}+0.79$ to $0.84$ d (Woodward et al., 2021, Quimby et al., 2024).

Its speed class is the primary reason it is regarded as exceptional. Published decline estimates cluster around $t_2 \sim 1$ d, with specific reported values including $t_2=0.904$ d, $t_2 \simeq 1.03$ d, $t_2=1.1$ d, and $t_2=1.2$ d. Reported $t_3$ values include $V_{\rm max}=6.19 \pm 0.05$0 d and $V_{\rm max}=6.19 \pm 0.05$1 d, and one near-infrared study also quoted $V_{\rm max}=6.19 \pm 0.05$2 d. Several papers explicitly identify it as the fastest Galactic nova known to date, or the fastest classical nova yet recorded (Habtie et al., 2023, Olbemo et al., 7 May 2026, Woodward et al., 2021).

The early optical light curve showed a plateau in the pre-maximum light curve at $V_{\rm max}=6.19 \pm 0.05$3, approximately $V_{\rm max}=6.19 \pm 0.05$4 mag below peak, that lasted for at least three hours. That detection was described as unprecedented at such a large offset below maximum (Quimby et al., 2021).

Spectroscopically, V1674 Her did not remain confined to a single classical subtype. It was initially reported as Fe II-like, but later work emphasized rapid evolution toward He/N-like, high-ionization, and neon-rich behavior. Strong [Ne III] and [Ne V] lines were identified early in the post-maximum evolution, and several studies therefore interpret the system as a neon nova on an oxygen-neon white dwarf (Habtie et al., 2023, Rawat et al., 23 Feb 2026).

2. Rise to maximum and early photometric structure

V1674 Her is unusual not only because it was fast, but because the rise itself was observed unusually early. Archival and reanalyzed survey data began more than $V_{\rm max}=6.19 \pm 0.05$5 mag below optical peak, and Evryscope recorded the crucial rise with 2-minute cadence. Over the high-cadence interval, the nova brightened by approximately $V_{\rm max}=6.19 \pm 0.05$6 mag in just $V_{\rm max}=6.19 \pm 0.05$7 hours (Quimby et al., 2024).

The early rise was not a single smooth brightening. The reconstructed light curve shows three phases separated by two clear breaks. The earliest phase was a slow rise. It included an “early peak” at $V_{\rm max}=6.19 \pm 0.05$8 d, followed by a sharp fade at $V_{\rm max}=6.19 \pm 0.05$9 over the next $t_{\rm ref}+0.79$0 min, with the minimum of the dip at $t_{\rm ref}+0.79$1 d. After that, the nova entered a fast-rise phase beginning at $t_{\rm ref}+0.79$2 d, during which the flux increased nearly linearly with time at $t_{\rm ref}+0.79$3. A second break then led to an even steeper, again nearly linear-in-flux rise at $t_{\rm ref}+0.79$4, about $t_{\rm ref}+0.79$5 higher than in the preceding phase (Quimby et al., 2024).

A blackbody toy model at approximately Eddington luminosity was able to organize part of this phenomenology in terms of photospheric expansion, but not all of it. In particular, that study concluded that the white dwarf was unlikely to have overflowed its Roche lobe prior to the launch of a fast wind. This directly challenged the specific version of the gamma-ray-nova picture in which a slow torus is built by Roche-lobe overflow before a later fast wind begins (Quimby et al., 2024).

A later full outburst calculation interpreted the same eruption with a $t_{\rm ref}+0.79$6 white dwarf model and likewise emphasized that V1674 Her is one of the first novae for which a one-dimensional calculation can be confronted with the observed rise itself, not only the decline. In that model, the early optical rise and very rapid decay are both consequences of a highly massive white dwarf and rapidly evolving envelope (Kato et al., 5 Jun 2025).

3. Binary periodicities and intermediate-polar interpretation

V1674 Her is now widely interpreted as a magnetic cataclysmic variable of the intermediate-polar class. A major reason is that a pre-eruption oscillation was already present in archival ZTF data: $t_{\rm ref}+0.79$7 s in one study and $t_{\rm ref}+0.79$8 s in another. During eruption, time-series photometry revealed two coherent clocks at $t_{\rm ref}+0.79$9 d and $0.84$0 s, interpreted respectively as the orbital period of the binary and the white-dwarf spin period. The power spectra also showed sidebands at $0.84$1 and $0.84$2, described explicitly as a classic signature of an intermediate polar (Patterson et al., 2022, Drake et al., 2021).

The orbital period places the system at about $0.84$3 h. Orbital modulation was reportedly visible by day 4, while the $0.84$4 s optical pulsation appeared around day 12 at first at only $0.84$5 mag full amplitude and then grew steadily to $0.84$6 mag by day 350. One test orbital ephemeris was given as

$0.84$7

These periodicities emerged unusually early in the eruption, which was interpreted as evidence that the ejecta became optically thin to radiation from the inner binary very quickly (Patterson et al., 2022).

The spin signal is also central to the system’s time-domain behavior after eruption. Comparing the pre-outburst $0.84$8 s period with the first secure post-eruption optical detection around day 15 at $0.84$9 s, one study inferred an abrupt increase of $t_2 \sim 1$0 ms, or $t_2 \sim 1$1, in less than 15 days. That same work then argued for a subsequent steady decrease of approximately $t_2 \sim 1$2 ms yr$t_2 \sim 1$3, about $t_2 \sim 1$4 times faster than usually seen in intermediate polars (Patterson et al., 2022).

X-ray timing reinforced the magnetic-binary interpretation while also complicating the period history. Chandra measured $t_2 \sim 1$5 s during the supersoft phase and found a strongly non-sinusoidal pulse with harmonic structure at $t_2 \sim 1$6, $t_2 \sim 1$7, $t_2 \sim 1$8, and $t_2 \sim 1$9; the pulse fraction was summarized as about $t_2=0.904$0. AstroSat detected soft-X-ray periods of $t_2=0.904$1 s on day 37, $t_2=0.904$2 s on day 38, and $t_2=0.904$3 s on day 54, but found no corresponding far-ultraviolet periodicity, with an upper limit $t_2=0.904$4 (Drake et al., 2021, Bhargava et al., 2023).

Not every timing campaign recovered the spin unambiguously at every epoch. Second-timescale infrared photometry with Palomar Gattini-IR, obtained on three nights between 3 and 6 days after discovery with resolution of $t_2=0.904$5 s, found no periodic variability down to a three sigma upper limit of $t_2=0.904$6 mag. A later optical study covering about 150 days after eruption likewise reported no unambiguous white-dwarf spin period in its own data, placing upper limits of $t_2=0.904$7 mag in $t_2=0.904$8 and $t_2=0.904$9 mag in $t_2 \simeq 1.03$0. These non-detections do not contradict the intermediate-polar interpretation, but they show that the detectability and amplitude of the $t_2 \simeq 1.03$1 s signal were strongly epoch- and band-dependent (Hansen et al., 2021, Rawat et al., 23 Feb 2026).

4. Shocks, coronal development, and the multiwavelength outflow

V1674 Her rapidly became one of the best-observed shock-powered novae. Fermi-LAT detected $t_2 \simeq 1.03$2–$t_2 \simeq 1.03$3 GeV emission beginning $t_2 \simeq 1.03$4 h after the eruption start, at a level of $t_2 \simeq 1.03$5 photons cm$t_2 \simeq 1.03$6 s$t_2 \simeq 1.03$7, and the GeV episode lasted only $t_2 \simeq 1.03$8 h. Eleven days later, simultaneous NuSTAR and Swift observations found optically thin thermal plasma shock-heated to $t_2 \simeq 1.03$9 keV; the preferred NuSTAR fit gave $t_2=1.1$0 keV and the absence of a detectable 6.7 keV Fe K$t_2=1.1$1 line suggested non-solar abundances, interpreted as super-solar CNO-group abundances. The radio behavior was consistent with thermal emission at early times and synchrotron at late times, with a late high-frequency slope $t_2=1.1$2 (Sokolovsky et al., 2023).

The spectroscopic evolution toward high ionization was unusually rapid. Near-infrared spectroscopy obtained from day 5.64 to day 68.32 showed broad and structured H and He recombination lines very early, followed by unambiguous coronal emission by day 11.51. That onset was described as the earliest yet observed for any classical nova. The coronal spectrum included, among many other species, [S IX], [Si X], [Si IX], [P VIII], [Si XI], [Si VI], [Al IX], [Ca VIII], [Si VII], [Al V], [Mg VIII], [Ca IV], [Al VI], and [Al VIII]. The authors argued that the earliest coronal emission could not yet be photoionized and must instead be shock-excited, with an estimated coronal-gas temperature

$t_2=1.1$3

on day 11.51 (Woodward et al., 2021).

The line profiles themselves provided kinematic evidence for internal shocks. In the early near-infrared spectra, common substructures were identified at approximately $t_2=1.1$4, $t_2=1.1$5, $t_2=1.1$6, $t_2=1.1$7, $t_2=1.1$8, $t_2=1.1$9, and $t_2=1.2$0 km s$t_2=1.2$1, each with uncertainty $t_2=1.2$2 km s$t_2=1.2$3. The persistence of similar structure into later coronal phases was taken to indicate multiple ejecta parcels moving at different speeds, naturally producing internal shock interfaces (Woodward et al., 2021).

Optical and X-ray spectroscopy likewise showed an exceptionally fast, structured outflow. Early optical spectra showed P Cygni absorption near $t_2=1.2$4, followed by faster components $t_2=1.2$5. During the supersoft X-ray phase, Chandra grating spectra displayed P Cygni-type O VIII, O VII, and N VII features with two dominant blue-shifted absorption components at approximately $t_2=1.2$6 and $t_2=1.2$7, and a blue wing extending to about $t_2=1.2$8. These data were interpreted as evidence for a clumpy, multi-component, highly asymmetric outflow (Drake et al., 2021).

AstroSat added a simultaneous soft-X-ray and FUV view of the supersoft phase. Soft-X-ray timing recovered the $t_2=1.2$9 s modulation after the peak of the supersoft phase, while the FUV showed strong emission lines from Si, N, and O but no corresponding periodicity. The FUV line fluxes declined steadily as the bright soft-X-ray state developed, supporting the conclusion that the soft-X-ray and FUV emission originated in distinct regions (Bhargava et al., 2023).

5. Ejecta geometry and theoretical interpretations

The ejecta geometry inferred from optical line profiles is strongly non-spherical. Morpho-kinematic analyses with SHAPE favored a bipolar morphology with a central equatorial ring; one optical study described the geometry as bipolar with polar blobs and an equatorial ring, with inclination $t_3$0 and position angle $t_3$1, while another found a symmetrical bipolar morphology with asymmetrical polar caps and a central equatorial ring at $t_3$2 and $t_3$3 deg. Both analyses adopted a density profile $t_3$4 and a Hubble-like velocity field (Rawat et al., 23 Feb 2026, Habtie et al., 2023).

Photoionization modeling with CLOUDY further suggested that the ejecta were spectroscopically heterogeneous. A single-density component did not reproduce the optical spectra satisfactorily, so the favored models used dense clumps plus a more diffuse component. Across modeled epochs, the central ionizing source temperature rose from $t_3$5 K to $t_3$6 K, and the luminosity from $t_3$7 to $t_3$8 erg s$t_3$9. The same work found helium, oxygen, nitrogen, and neon overabundant relative to solar, with Fe overabundant in nebular epochs and then diminishing toward approximately solar in the coronal phase. The reported ejecta mass range was $V_{\rm max}=6.19 \pm 0.05$00–$V_{\rm max}=6.19 \pm 0.05$01 in the abstract; the body text lists $V_{\rm max}=6.19 \pm 0.05$02 for the last epoch (Habtie et al., 2023).

A separate one-dimensional outburst calculation constructed a full theoretical light-curve model with $V_{\rm max}=6.19 \pm 0.05$03, $V_{\rm max}=6.19 \pm 0.05$04, $V_{\rm max}=6.19 \pm 0.05$05 h, $V_{\rm max}=6.19 \pm 0.05$06 s, and a quiescent mass-accretion rate $V_{\rm max}=6.19 \pm 0.05$07. In that model the photosphere expands to $V_{\rm max}=6.19 \pm 0.05$08, the companion is engulfed $V_{\rm max}=6.19 \pm 0.05$09 h after the onset of thermonuclear runaway and re-emerges $V_{\rm max}=6.19 \pm 0.05$10 d later, and the duration of the X-ray flash is only $V_{\rm max}=6.19 \pm 0.05$11 h. The authors argued that the decline phase is well approximated by a sequence of steady-state envelope solutions (Kato et al., 5 Jun 2025).

An additional theoretical interpretation was proposed to explain why V1674 Her was about one magnitude brighter at peak than typical very fast novae. In that model, a strong reverse shock arises $V_{\rm max}=6.19 \pm 0.05$12 d after outburst, producing an optically thick shocked shell expanding at approximately $V_{\rm max}=6.19 \pm 0.05$13. The shocked shell reaches $V_{\rm max}=6.19 \pm 0.05$14 when it expands to $V_{\rm max}=6.19 \pm 0.05$15 on day $V_{\rm max}=6.19 \pm 0.05$16, while the ordinary free-free component peaks earlier, at $V_{\rm max}=6.19 \pm 0.05$17 on day $V_{\rm max}=6.19 \pm 0.05$18. This shell model was used to explain both the “superbrightness” and the fact that the GeV gamma-ray peak preceded the optical $V_{\rm max}=6.19 \pm 0.05$19-band maximum (Hachisu et al., 23 Dec 2025).

6. Distance, white-dwarf mass, and unresolved issues

Distance estimates for V1674 Her span a wide range because different methods have been applied. A near-infrared spectroscopic study, adopting $V_{\rm max}=6.19 \pm 0.05$20, derived $V_{\rm max}=6.19 \pm 0.05$21 kpc. An optical photoionization study derived a $V_{\rm max}=6.19 \pm 0.05$22-band distance of approximately $V_{\rm max}=6.19 \pm 0.05$23 kpc and an average $V_{\rm max}=6.19 \pm 0.05$24 kpc from multi-band MMRD-based estimates. A time-stretching light-curve analysis instead obtained $V_{\rm max}=6.19 \pm 0.05$25 and $V_{\rm max}=6.19 \pm 0.05$26 kpc for $V_{\rm max}=6.19 \pm 0.05$27. Other discussions adopted 5 kpc, noted Gaia-based estimates of $V_{\rm max}=6.19 \pm 0.05$28 kpc, or used $V_{\rm max}=6.19 \pm 0.05$29 kpc from Sokolovsky et al. (2023) in quiescent X-ray work (Woodward et al., 2021, Habtie et al., 2023, Kato et al., 5 Jun 2025, Olbemo et al., 7 May 2026).

The white-dwarf mass is correspondingly controversial. Decline-time arguments and some outburst models favor a very high mass. One optical study inferred $V_{\rm max}=6.19 \pm 0.05$30 from $V_{\rm max}=6.19 \pm 0.05$31 d, and the full light-curve model described above adopted $V_{\rm max}=6.19 \pm 0.05$32 because lower masses were too slow and $V_{\rm max}=6.19 \pm 0.05$33 was too fast in decline (Habtie et al., 2023, Kato et al., 5 Jun 2025).

Direct quiescent X-ray constraints, however, point lower. A broadband XMM-Newton plus NuSTAR study modeled the hard X-rays with a post-shock accretion-column model and, assuming $V_{\rm max}=6.19 \pm 0.05$34, obtained

$V_{\rm max}=6.19 \pm 0.05$35

A joint spectral-plus-timing fit yielded

$V_{\rm max}=6.19 \pm 0.05$36

and a consistent NuSTAR-only hard-band fit gave $V_{\rm max}=6.19 \pm 0.05$37. The same study inferred a surface magnetic field

$V_{\rm max}=6.19 \pm 0.05$38

placing V1674 Her at the high end of the magnetic-field distribution for intermediate polars (Olbemo et al., 7 May 2026).

This discrepancy has become one of the system’s broader astrophysical implications. Using an empirical $V_{\rm max}=6.19 \pm 0.05$39-mass relation with $V_{\rm max}=6.19 \pm 0.05$40 d, the X-ray mass paper showed that decline-time inversion would imply $V_{\rm max}=6.19 \pm 0.05$41, which is unphysical because it exceeds the Chandrasekhar limit. That work therefore argued that empirical decline-time relations can overestimate white-dwarf masses in extreme fast novae, and that additional parameters such as accretion rate, magnetic field strength, and envelope properties must influence nova timescales. It also noted a caveat: because the adopted hard-X-ray model does not include cyclotron cooling, the true mass could be somewhat higher if cyclotron losses are dynamically important (Olbemo et al., 7 May 2026).

Taken together, the literature presents V1674 Herculis as an unusually constraining nova rather than merely an unusually fast one. Its eruption links pre-maximum light-curve structure, early shocks, rapid coronal development, magnetic-binary timing, and post-nova accretion physics in a single system. The main unresolved questions concern the exact white-dwarf mass, the distance scale, the origin of the earliest optical and gamma-ray features, and the degree to which its extreme speed is controlled by white-dwarf mass alone versus the coupled effects of magnetic geometry, accretion history, and ejecta structure.