Swift X-ray observations of the ~60 day super-soft phase of the recurrent nova RS Ophiuchi 2006 show the progress of nuclear burning on the white dwarf in exquisite detail. First seen 26 days after the optical outburst, this phase started with extreme variability likely due to variable absorption, although intrinsic white dwarf variations are not excluded. About 32 days later, a steady decline in count-rate set in. NLTE model atmosphere spectral fits during the super-soft phase show that the effective temperature of the white dwarf increases from ~65 eV to ~90 eV during the extreme variability phase, falling slowly after about day 60 and more rapidly after day 80. The bolometric luminosity is seen to be approximately constant and close to Eddington from day 45 up to day 60, the subsequent decline possibly signalling the end of extensive nuclear burning. Before the decline, a multiply-periodic, ~35 s modulation of the soft X-rays was present and may be the signature of a nuclear fusion driven instability. Our measurements are consistent with a white dwarf mass near the Chandrasekhar limit; combined with a deduced accumulation of mass transferred from its binary companion, this leads us to suggest RS Oph is a strong candidate for a future supernova explosion. The main uncertainty now is whether the WD is the CO type necessary for a SN Ia. This may be confirmed by detailed abundance analyses of spectroscopic data from the outbursts.
Novae result from the explosive thermonuclear fusion of hydrogen to helium in the surface layers of a white dwarf (WD). The hydrogen-rich fuel for this process is provided by accretion from the outer layers of a cooler binary companion (Starrfield 2008), in the case of RS Ophiuchi (RS Oph), a red giant (RG). Most novae have a single historically recorded outburst (the Classical Novae [CNe]), but a few, including RS Oph, are recurrent on timescales of ∼8 to less than 80 years -the so-called Recurrent Novae (RNe; see Schaefer 2010 for a review). The observably short recurrence intervals for RNe are thought to result from a higher accretion rate and a higher WD mass than in CNe, both leading to a more rapid release of energy in the ignition zone and a shorter time to runaway (Townsley 2008).
RS Oph shows recurrent nova outbursts at roughly 20 year intervals. The latest outburst was detected on 2006 Feb 12.8 at a magnitude of 4.5 (Narumi et al. 2006; see Hounsell et al. 2010 for a complete early light curve). Multi-frequency observations during the previous outburst in 1985 led to determinations of the distance, d = 1.6 ± 0.3 kpc (Bode 1987), and the interstellar column density, N H,ISM = (2.4 ± 0.6) × 10 21 cm -2 (Hjellming et al. 1986). X-ray observations were conducted by EXOSAT at six epochs, from days 55 to 251 after the outburst, over the 0.04-2.0 keV and 1.5-15 keV bands (Mason et al. 1987;O’Brien et al. 1992). X-ray emission during the first five epochs was consistent with shocks propagating through the pre-existing wind of the RG companion star. Modelling this process, O’ Brien et al. (1992) derived an outburst energy of 1.1 × 10 43 erg and an ejected mass of 1.1 × 10 -6 M ⊙ , but it proved difficult to model the low and high energy X-ray evolution simultaneously and the authors suggested that on-going nuclear fusion on the WD surface might be responsible. As will be shown in this paper, our results confirm this view.
Novae have been predicted to undergo a super-soft source (SSS) phase as the mass loss from the central source declines and the effective photospheric surface shrinks at constant bolometric luminosity (MacDonald et al. 1985). This phase, which produces emission from the surface of the WD with temperatures around 3 × 10 5 K, has been observed in several novae; see, e.g., Page et al. (2010); Drake et al. (2003); Ness et al. (2003); Orio et al. (2002). Previously, the outburst of V1974 Cyg in 1992 had the best temporal coverage, with 18 epochs of ROSAT observations ranging from 63 to 653 days after outburst (Krautter et al. 1996, Balman et al. 1998). This outburst was the brightest SSS observed at the time (in terms of observed flux), but our Swift data show that RS Oph peaked 2-3 times brighter still.
X-ray data from the 2006 outburst of RS Oph were obtained by Swift, the Rossi X-Ray Timing Explorer (RXTE), XMM-Newton and Chandra and have been extensively discussed in a range of papers. Bode et al. (2006) described the early hard emission, based on Swift-X-ray Telescope (XRT) observations (as well as a detection by the Burst Alert Telescope -the hard X-ray instrument onboard Swift), confirming basic models from the 1985 outburst, while Sokoloski et al. (2006) presented the RXTE data for a similar early time interval. Ness et al. (2007Ness et al. ( , 2009) ) and Nelson et al. (2008) discuss the grating spectra obtained by XMM-Newton and Chandra, both before and during the SSS phase, while Drake et al. (2009) concentrated on the early (pre-SSS) Chandra high-energy transmission grating alone, finding the ejecta may contain super-solar abundances. Luna et al. (2009) find evidence for extended soft X-ray emission using Chandra CCD data. Vaytet, O’Brien & Bode (2007) present 1-dimensional hydrodynamical models of the shocks within the interacting winds of the RS Oph system and compare them to the results of Bode et al. (2006). Their models reproduce the rise and subsequent deceleration of the shock velocities, but require a very high speed wind to achieve the high shock velocity observed. Three-dimensional modelling performed by Walder, Folini & Shore (2008) leads them to conclude that the WD mass in the RS Oph system is increasing with time. Similar conclusions were reached by Orlando et al. (2009), who estimated a total ejecta mass of 10 -6 M ⊙ based on 3-D hydrodynamical modelling constrained by Swift and Chandra X-ray observations.
Besides the X-ray observations, data on RS Oph were also collected in the radio (O’Brien et al. 2006;Kantharia et al. 2007;Eyres et al. 2009), IR (Monnier et al. 2006;Das, Banerjee & Ashok 2006;Evans et al. 2007a,b;Chesneau et al. 2007;Lane et al. 2007;Banerjee, Das & Ashok 2009;Rushton et al. 2010;Brandi et al. 2009) and optical bands (Hachisu et al. 2006;Worters et al. 2007;Bode et al. 2007;Munari et al. 2007;Brandi et al. 2009). Hachisu, Kato & Luna (2007) discussed the temporal evolution of the RS Oph SSS X-ray light curve, but did not include any detailed c
This content is AI-processed based on open access ArXiv data.
