Swift observations of nova V407 Lup: detection of a UV period at 1.1 or 3.6 hours

ATel #10632; A. P. Beardmore, K. L. Page, J. P. Osborne (U. Leicester), M. Orio (Department of Astronomy, University of Wisconsin, USA and Padova Observatory, National Institute of Astrophysics, Italy)
on 11 Aug 2017; 13:38 UT
Credential Certification: Kim Page (kpa@star.le.ac.uk)

Subjects: Ultra-Violet, X-ray, Nova

V407 Lup (also known as Nova Lup 2016 and ASASSN-16kt) was reported to be in outburst on 2016-Sep-24.0 by Stanek et al. (ATel #9538 and ATel #9539). Although Swift observations started 2 days later, the source was too bright at the time for the UVOT to observe, and was not detectable in X-rays by the XRT, before the object came too close to the Sun on 2016-Oct-14 to be safely observed by Swift.

Swift observations resumed on 2017-Feb-21 (150 days after discovery), after the observing constraint. The nova was visible in the XRT as a bright, soft X-ray source (56.1 +/- 0.3 count s-1), while the UVOT data revealed the source at a uvw2 magnitude of 13.5.

We began a monitoring campaign with Swift, finding that the X-ray count rate was variable, first fading to around 30 count s-1, with a brief rebrightening to ~45 count s-1 on day 161 (2017-Mar-04), then fading again until day 174 (2017-Mar-16), at which point the count rate was 20 count s-1. Following this, the X-ray count rate slowly increased back to around 40 count s-1 on day 211 (2017-Apr-23), though this interval included a 2 week gap in Swift data caused by an observing constraint. The source then remained close to constant, between ~ 20-30 count s-1, until day 270 (2017-Jun-21). Since this time, the X-ray count rate has been following a slow fading trend; as of day 321 (2017-Aug-11) the count rate has fallen to ~6 count s-1.

Throughout this interval (days 150-321), the X-ray source has been super-soft, with almost all counts below 1 keV. A hardness ratio comparing 0.45-1 keV and 0.3-1 keV has shown distinct variability, though this does not appear to be periodic or correlated with the X-ray count rate or UV magnitude. Fitting the spectra with an NLTE model atmosphere grid suggests the temperature is fluctuating between ~70 and 90 eV, with the absorption remaining approximately constant, at ~1.5 x 1021 cm-2. Despite the constantly diminishing X-ray count rate after day 270, the effective temperature does not appear to be clearly decreasing, as expected during rapid cooling due to nuclear burning turn-off.

The UVOT light curve declined slowly after day 150 before flattening off at around a uvw2 magnitude of 14.2 by around day 200, with signs of a 0.1 mag variability superimposed on the overall decline. The UV data after day 300 (2017-Jul-22) have brightened by at least 0.3 mag.

A Lomb Scargle periodogram of the detrended UV light curve from day 199-270 reveals a periodic signal with a timescale of either 0.04593 day or 0.14887 day, which are aliases of each other caused by Swift's 96 min orbital period. The amplitude of the modulation when folded at either of the candidate periods is approximately 0.1 mag. Due to the relatively low cadence of the data, we cannot break the degeneracy between the periods using the UVOT data. Either of these is a candidate for the binary orbital period.

A periodogram of a 50s binned XRT light curve taken over the same time range, shows the strongest peak to be coincident with the longer of the UVOT periods, having an amplitude of 10 per cent. These UV and X-ray modulations are in phase. The XRT periodogram shows an number of other peaks at unrelated frequencies with only slight lower power; we do not claim an independent detection of an X-ray period.