Papers
Topics
Authors
Recent
Search
2000 character limit reached

The instantaneous mass accretion rate of novae in quiescence: - an archival ultraviolet optical spectral analysis

Published 18 Aug 2026 in astro-ph.SR | (2608.18037v1)

Abstract: Based on archival spectra, we derive the quiescent instantaneous mass transfer rates in novae using synthetic disk spectra generated with tlusty, Gaia parallax-derived distances, and updated color excess values. Our results for nine novae, based on ultraviolet spectra and on a number of optical spectra, yield mass accretion rates that are higher than those derived from simple integration of the UV and optical luminosity. HR Del, 20 years after its eruption, has a mass transfer rate of 4×10<sup>7M4\times 10<sup>{-7}M_\odot/yr, and is likely burning the accreting H-rich material. V842 Cen, thought to be a low mass transfer system, has a comparable mass accretion rate. For both novae, the enhanced mass transfer must be self-sustained by a feedback loop. RR Pic, with M˙3×10<sup>8M\dot{M}\approx 3\times 10<sup>{-8}M_\odot/yr, is better fitted with an accretion disk where the outer region is heated up to 12,000 K, in agreement with H and He emission lines coming from the outer disk and a large emission region on the leading side of the disk. Such a heated disk also gives a good fit to the spectra of CP Lac, and DI Lac with M˙4.5\dot{M}\sim4.5 and 9×10<sup>9M9\times 10<sup>{-9}M_\odot/yr. V1974 Cyg and V533 Her, with an accretion rate of 3×10<sup>9M\sim 3\times 10<sup>{-9}M_\odot/yr, have a rather flat spectrum. V446 Her and BK Lyn, caught in a state of low accretion, have the lowest mass accretion rates with M˙10<sup>9\dot{M} \sim 10<sup>{-9} and 10<sup>10M\sim 10<sup>{-10}M_\odot/yr respectively. The higher mass transfer rate systems, with M˙10<sup>7M\dot{M}\approx \sim 10<sup>{-7}M_\odot/yr, agree with the standard disk model; the remaining systems are better fitted when the outer disk is heated to 12,000\sim 12,000~K. We suggest that irradiation from the heated WD, inner disk, together with tidal interaction, the bright spot, and material overflowing the disk edge, can increase the temperature of the outer disk.

Summary

  • The paper measures instantaneous accretion rates for nine quiescent classical novae using NLTE disk models, Gaia distances, and ultraviolet–optical spectra, finding values spanning 10⁻¹⁰ to 4×10⁻⁷ M☉ yr⁻¹.
  • The analysis shows that HR Del and V842 Cen sustain near- or above-stable-burning rates, while corrected reddening explains why V842 Cen was previously classified as a low-transfer system.
  • Lower-rate novae require an augmented disk with a roughly 12,000 K heated outer region, suggesting irradiation or tidal effects help explain spectral slopes and the discrepancy with secular evolution models.

Overview and motivation

This paper presents an archival ultraviolet–optical spectroscopic analysis of nine classical novae in quiescence, aimed at measuring their instantaneous mass accretion rates M˙\dot{M} with realistic accretion disk models. The work addresses a long-standing discrepancy: observed quiescent mass transfer rates in novae (M˙109\dot{M} \sim 10^{-9}108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}, reaching 10710^{-7} in some systems) exceed theoretical secular predictions by one to two orders of magnitude. The authors fit synthetic disk spectra generated with tlusty/synspec/disksyn to archival IUE and HST spectra, adopting Gaia DR3 parallax-derived distances, literature white dwarf (WD) masses, and color excesses derived from the 2175 Å interstellar absorption feature and the Stilism 3D reddening map.

The sample comprises BK Lyn, HR Del, RR Pic, CP Lac, DI Lac, V533 Her, V446 Her, V1974 Cyg, and V842 Cen, spanning post-eruption timescales from a few years to roughly two millennia (BK Lyn is identified with the nova of 101 AD). All systems were selected for high-quality UV spectra obtained late enough after eruption that the accretion disk dominates the continuum.

Methodology

The disk models assume optically thick, vertically thin, steady-state Keplerian disks following the standard Shakura–Sunyaev temperature profile, with WD masses of 0.35–1.21 MM_\odot, inclinations of 18°–75°, and accretion rates sampled at 0.5 dex intervals. A methodological improvement over earlier work is that synspec was run to compute specific intensities at discrete angles rather than emergent flux, eliminating reliance on tabulated limb-darkening corrections. A key extension is the "augmented" disk model: beyond the radius where the standard profile reaches ~12,000 K, the outer disk is treated as an isothermal 12,000 K region extending to ~0.45aa (the binary separation), motivated by tidal heating, the bright spot, stream-disk overflow, and irradiation from the hot inner regions.

Dereddening exploits the 2175 Å feature: E(BV)E(B-V) is tuned until the bump vanishes under the Fitzpatrick & Massa extinction law. Fits are validated against the Gaia distance rather than ranked purely by reduced χ2\chi^2, since distance agreement and best-fit quality do not always coincide. The dominant uncertainties are inclination, WD mass, and reddening; a 0.05 error in E(BV)E(B-V) alone produces a ~45% flux error near 1500 Å.

Results for individual systems

The headline results span four orders of magnitude in M˙\dot{M}:

System M˙109\dot{M} \sim 10^{-9}0 (M˙109\dot{M} \sim 10^{-9}1) M˙109\dot{M} \sim 10^{-9}2 (M˙109\dot{M} \sim 10^{-9}3/yr) Heated outer disk
HR Del 0.68 M˙109\dot{M} \sim 10^{-9}4 No
V842 Cen 1.0 M˙109\dot{M} \sim 10^{-9}5 No
RR Pic 1.0 M˙109\dot{M} \sim 10^{-9}6 Yes
DI Lac 1.0 M˙109\dot{M} \sim 10^{-9}7 Yes
CP Lac (UV/optical) 1.1 M˙109\dot{M} \sim 10^{-9}8 Yes
V533 Her 1.0 M˙109\dot{M} \sim 10^{-9}9 Yes
V1974 Cyg 1.1 108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}0 Yes
V446 Her 1.1 108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}1 Yes
BK Lyn 1.1 108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}2 No

Several results carry particular weight. For HR Del, only two decades after its 1967 eruption, the derived rate of 108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}3 exceeds the stable hydrogen-burning limit, implying the WD must burn material as it accretes — direct evidence for an active self-sustained irradiation feedback loop. Notably, the same disk model fits both the IUE UV spectrum and the independently calibrated Lick optical spectrum, lending robustness to the result. For V842 Cen, previously characterized as a low-transfer system, the authors derive 108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}4; they show explicitly that prior low estimates stemmed from underestimated reddening (they find 108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}5 versus 0.55 used elsewhere) and incorrect distances, since extinction suppresses the UV flux by up to three orders of magnitude. This is a strong corrective claim: the system's faintness is an extinction artifact, not intrinsically low transfer.

For BK Lyn, caught in a low state resembling DN quiescence, the analysis yields the lowest rate in the sample, 108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}6, requiring a massive WD (108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}7); low-mass disk models are excluded by the Ly108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}8 profile. A combined disk-plus-WD fit suggests a ~40,000 K WD contributing ~30% of the flux. At this rate, envelope buildup to the next eruption would take 108Myr110^{-8}\,M_\odot\,\mathrm{yr}^{-1}9–10710^{-7}0 yr — upper limits that assume no hibernation.

A systematic trend emerges across the sample: systems accreting at 10710^{-7}1 (HR Del, V842 Cen, T Pyx from prior work) are well fitted by standard disk models, while lower-rate systems require the augmented 12,000 K outer disk, whose relative contribution grows as 10710^{-7}2 declines. Plotted against time since eruption, the ten systems hint at a secular decline of 10710^{-7}3 over decades to millennia, though the authors correctly caution that ten points — with two systems caught in low states — cannot establish this statistically.

Comparison with prior work

Against Selvelli & Gilmozzi's luminosity-integration analysis of six common systems, the disk-model rates are systematically higher: ~20% larger for V533 Her and V446 Her, twice as large for DI Lac and CP Lac, three times for RR Pic, and five times for HR Del. This reproduces the known pattern that tlusty-based disk modeling yields rates roughly three times those from simple flux integration. Part of the offset is attributable to the authors' adoption of systematically larger WD masses from Shara et al.'s evolutionary calculations — particularly for sub-solar estimates — which lowers the inferred 10710^{-7}4; had smaller masses been adopted, the discrepancy would widen further.

Limitations and open questions

The paper concedes several constraints on its conclusions. The fitting is acknowledged to be partly empirical ("visual inspection" supplements distance validation), and the physical origin of the 12,000 K outer disk component is proposed but not demonstrated. Extinction from nova ejecta dust cannot be corrected via the 2175 Å feature, introducing unquantified uncertainty, most severely for V842 Cen where the reddening error alone spans a factor of two in 10710^{-7}5. Several systems lack secure parameters: BK Lyn has no measured WD mass or recorded outburst characteristics; V842 Cen's orbital period and IP classification remain contested; V533 Her's flat spectrum may reflect stream overflow veiling rather than disk emission, meaning its fitted 10710^{-7}6 may not represent a clean disk. The declining-10710^{-7}7-with-time trend is explicitly flagged as requiring many more data points. Whether the instantaneous post-eruption rates bear on secular averages between eruptions — the quantity relevant to CV evolution theory and recurrence-time calculations — remains open, as does the question of whether hibernation intervenes.

Conclusion

By combining realistic NLTE disk atmospheres, Gaia distances, and careful dereddening, this study provides the most self-consistent set of instantaneous quiescent accretion rates for classical novae to date, ranging from 10710^{-7}8 to 10710^{-7}9. It confirms that at least some novae sustain near-stable-burning accretion rates for decades after eruption via irradiation feedback, corrects the misclassification of V842 Cen as a low-transfer system, and identifies heated outer disks as a likely contributor to the long-standing mismatch between observed and modeled CV spectral slopes. The central unresolved issue the paper leaves is quantitative: how rapidly MM_\odot0 decays after eruption, and whether the observed decline extrapolates to the low secular rates required by evolution theory.

Paper to Video (Beta)

No one has generated a video about this paper yet.

Whiteboard

No one has generated a whiteboard explanation for this paper yet.