The ”super-active” accretion phase of T CrB has ended

T CrB is a very famous recurrent nova (eruptions recorded in 1866 and 1946; Payne-Gaposchkin, 1964, and references therein) and is also a symbiotic binary by harboring a red giant (RG) as the donor star to the massive white dwarf (WD) companion.

The life-cycle of a symbiotic binary as outlined by Munari (2019), is characterized by long accretion phases interspersed by shorter periods during which the material accumulated on the surface of the WD is burned nuclearly. If the accreted shell is not electron degenerate, the burning proceeds in thermal equilibrium for decades/centuries until most of the hydrogen fuel in the shell is consumed, the burning finally quenches down, and a new long-lasting phase of accretion initiates the next cycle (examples are V4368 Sgr, HM Sge, and V1016 Cyg). When the accreted shell is instead electron degenerate, the nuclear burning proceeds explosively resulting in a nova outburst, with most of the shell expelled in the process and the residual nuclear burning on the WD extinguishes in a few weeks/months, after which accretion resumes and a new cycle begins. In addition to T CrB, other well known symbiotic recurrent novae are RS Oph, V3890 Sgr, and V745 Sco.

Traditionally, accretion in symbiotic stars has been treated as a smooth process relatively stable over long periods of time (eg. Kenyon, 1986). This approach has progressively changed in favor of a highly-episodic interpretation of the accretion process, characterized by brief periods of (very) high accretion rates in-between longer intervals spent at much lower mass-transfer rates (eg. Luna et al., 2020; Munari et al., 2021).

Munari et al. (2016) has called attention to the fact that starting with 2015, T CrB entered a ”super-active” accretion phase (SAP), characterized by a much brighter accretion disk as the result of a greatly enhanced mass-flow through it and then toward the central WD. The accretion level attained during SAP largely exceeded any other experienced by T CrB since the 1946 eruption. By noting that a similar event preceded the 1946 nova outburst, Munari et al. (2016) concluded that SAP is probably announcing a new and imminent eruption of T CrB, a view shared by Schaefer (2023).

We have been regularly recording fluxed spectra of T CrB for the last $\sim$35 yrs, initially with the Asiago 1.82m + B&C and since 2006 with the Asiago 1.22m + B&C telescope. For all the 1.22m spectra, we adopted a 300 ln/mm grating blazed at 5000 Å that paired with a completely UV-transparent optical train and a highly UV-sensitive CCD detector (ANDOR iDus DU440A with a back-illuminated E2V 42-10 chip, 2048$\times$512 array, and 13.5 $\mu$m pixel size), allows to efficiently record spectra down to the $\sim$3100 Å atmospheric cut-off imposed by the telescope 1000m altitude above sea level. Our 1.22m spectra of T CrB extend from 3200 to 7900 Å at 2.3 Å/pix dispersion. In addition to being fluxed thanks to nightly observations of spectrophotometric standard stars, their flux zero-point is fine-tuned against (nearly-)simultaneous $B$$V$$R$ photometry, so that the flux error anywhere in the spectra rarely exceed a few percent. This 2006-2023 set of T CrB spectra is therefore characterized by a highly stable instrumental set-up and robust IRAF calibration procedures, and constitutes an ideal sample for variability studies of spectral features over long intervals of time. A few of the spectra of T CrB here considered can be viewed in Munari et al. (2016).

To trace the evolution of T CrB along the ”super active” accretion phase, we have measured the integrated flux of a sample of emission lines on the 2006-2023 Asiago 1.22m + B&C spectra described in the previous section. The selected lines are H$\beta$, HeI 5876, and HeII 4686, which are representative of low, medium, and high excitation/ionization conditions, respectively. Their absolute fluxes are plotted in Figure 1 along with the $B$-band lightcurve of T CrB as recorded by ANS Collaboration.

Prior to 2014, both HeI 5876 and HeII 4686 were not visible in emission, and H$\beta$ has been present but always at rather feeble levels. During this period the $B$-band lightcurve is dominated by the ellipsoidal distortion of the RG with superimposed the scattering due to the large-amplitude and always present flickering (eg. Zamanov & Bruch, 1998; Dobrotka et al., 2010, and references therein).

The start of the ”super active” accretion phase in late 2014 is marked by the sudden appearance in emission of HeI 5876 and HeII 4686, a corresponding rise of H$\beta$ (compare the spectra for 2014-11-02 and 2012-09-03 in Munari et al., 2016), and a large increase in $B$-band brightness caused by the rapidly brightening accretion disk. SAP reached its maximum in April 2016, when the flux of all emission lines sharply peaked, as illustrated by Figure 1. Around this epoch strong satellite UV and thermal radio emission were also recorded (Luna et al., 2018; Linford et al., 2019).

Following the maximum in April 2016, the flux of all emission lines has gone steadily decreasing, at a faster pace for higher excitation/ionization lines, and by mid-2023 they have returned to pre-SAP values, indicating that the ”super active” accretion phase is finally over. Also the $B$-band photometric brightness has been quickly dropping during the last few months, while the flickering has returned to the usual large amplitude (Minev et al., 2023) compared to the much reduced impact it had on photometry collected around SAP maximum (Zamanov et al., 2016).

The disappearance of emission lines, the drop in $B$-band brightness, and the return to large amplitude flickering suggest that the accretion disk has emptied from the extra material that driven the ”super active” state and has completed its transfer onto the WD. The shell around the latter may possibly still takes a little to cool and shrink down to favorable conditions, but the stage for a new nova outburst appears now inevitably set.