Nova Lup 2016 during the X-ray decay phase
ATel #10722; Marina Orio (National Istitute of Astrophysics of Italy-Padova and Dept. of Astronomy, University of Wisconsin, Madison, WI, USA), Andrew Beardmore, Kim Page, and Julian Osborne (Dept. of Physics and Astronomy, University of Leicester, UK)
on 8 Sep 2017; 22:03 UT
Credential Certification: Marina Orio (orio@astro.wisc.edu)
Nova Lup 2016 (ASASSN-16kt; see ATel #9538, #9539, #8550, #9554, #9587, #9594 and #9644) has been regularly monitored with Swift since the observations published in ATel #10632 that revealed a luminous supersoft X-ray source with a peak XRT count rate of 61.1(+-)0.1 cts/s on 2017/2/22. The early evolution of the XRT light curve was described in ATel #10632. The decay noted to start around day 2017 Jun 21 (day 270 after discovery) has continued, with the X-ray source falling to a count rate of ~1 ct/s on 2017 Aug. 25 (day 335).
The nova was observed with Chandra, the HRC-S camera and the Low Energy Transmission Grating on 2017 August 30 for 35,000 s and was detected with a count rate of 0.463(+-)0.005 cts/s in the averaged first order spectra. The unabsorbed flux measured with Chandra in the 0.2-1 keV range
is of about 2.7 x 10^(-11) erg/cm^2/s, corresponding to an unabsorbed flux of about 1.9 x 10^(-10) erg /cm^2/s approximately 37 times less than the absorbed flux, and 21 times less than the unabsorbed flux, estimated for the Swift XRT data in the 0.3-1 keV range on 2017 March 11. While the Swift XRT spectra at maximum could be approximately fitted
with a static NLTE atmosphere in the 75-85 eV range and column density 1.2-1.5 x 10^(21) cm^(-2), the late Chandra spectrum is well fitted with a NLTE atmospheric model at 59 eV and superimposed emission lines of carbon, nitrogen, oxygen, argon and (possibly) sulphur, with column density 10^(21) cm^(-2). Thus, while the drop in X-ray flux in the last months can be partially attributed to the cooling white dwarf atmosphere, with Chandra we find that the effective temperature did not decrease sufficiently to explain the large flux decrease.
Moreover, the normalization constant of the NLTE atmospheric model in the best fit to the Chandra spectrum indicates that the emitting region we observed on 2017 August 30 may have extended to only 100th the surface of a 1 solar mass white dwarf at 10 kpc. This may due to an initially disrupted, and more recently reconstructed accretion disk that hides the white dwarf during most of the orbit;
however the root cause may also be a non homogeneously cooling white dwarf surface. The Chandra zero order light curve shows variable count rate by 30%, but we have not detected a clearly periodic modulation (with period longer than one hour), typical of binary orbital cycles with a high inclination accretion disk (e.g. U Sco, Ness et al. 2012, ApJ 575, 43; Orio et al. 2013, MNRAS 429, 1342).
