CSS081007:030559+054715: Swift confirmation of 1.77 day period in X-ray and UV, and a suggestion of a ~45 day period.

ATel #1942; J P Osborne, A P Beardmore, K L Page (University of Leicester), J-U Ness (ESAC, ESA), G Schwarz (West Chester University), S Starrfield (Arizona State University), S Balman (METU), G Wynn (University of Leicester)
on 25 Feb 2009; 17:13 UT
Credential Certification: Julian P Osborne (julo@star.le.ac.uk)

Tweet

Since the initial report of the discovery of the 1.77 day period in the X-ray data of CSS081007:030559+054715 (ATEL #1873) Swift has observed the source continuously. It made two sets of high cadence observations between 12-24 Dec and 18 Jan to 9 Feb 2009, and approximately daily observations otherwise. The high cadence observations were taken at a rate of 500 sec per 96 minute Swift orbit, every day in set 1, and for 4 days per week in set 2. Daily observations are expected to continue until late March 2009.

Both the Swift XRT (0.3-10 keV) and UVOT (uvw2 filter, central wavelength = 1928 A) data show that the 1.77 day period is present throughout the observations (Beardmore et al., in prep). The following preliminary ephemeris is derived from the fits to 16 times of the minima in a 63-day heliocentrically corrected X-ray light curve: T_X,min = 54807.1245 (0.025) + E * 1.7751 (0.0013) days (68% confidence errors, epoch in MJD). The UV modulation is in phase with that seen in the X-ray, and has a peak-to-trough amplitude of 0.6 magnitudes.

The long-term X-ray light curve shows, after the large rise to 2.2 c/s reported in ATEL #1873, a decline over approximately 25 days to ~0.2 c/s. A second weaker peak reaching ~0.5 c/s is evident in mid-January. The UV flux shows a more repeatable pattern of long term evolution, with 3 minima and 2 peaks over the almost 150 days of observations. The UV data strongly suggest a period of approximately 45 days, although further monitoring will be needed to verify its periodic nature. The maxima and minima of the UV light curve coincide with similar features in the X-ray light curve.

The X-ray spectrum remains unchanged, being extremely soft and well characterized by a blackbody with kT ~ 45 eV in XRT spectral fits, as reported in ATEL #1847. (See ATEL #1901 for a report on a higher resolution X-ray spectrum.)

No other periods are apparent in the Swift X-ray or UV data. Specifically, the 1.694 and 0.6106 day periods mentioned in ATEL #1938 are not seen in our data (and we also do not see them clearly in the original V band data made available with that ATEL). The date of the photometric decline described there corresponds to a declining interval of the ~45 day Swift UV modulation. We also do not strongly detect in our Swift data the 11.386 day period reported in pre-outburst observations by Drake et al. (ATEL #1940). Finally, we note that the 14-day AAVSO V band light curve of this object also shows a strong periodic modulation at a period closely consistent with our X-ray period.

The bright super-soft X-ray spectrum suggests on-going nuclear burning on the white dwarf surface. The 1.77 day X-ray and UV photometric period is much longer than typical nova orbital periods (although similar to the 2-day P_orb of GK Per = N Per 1901), while it is in the range of orbital periods of persistent super-soft sources (i.e., those accreting at a high enough rate to allow continuous nuclear burning). CSS081007:030559+054715 is not a truly persistent X-ray source, as our analysis of archival data shows that it was not detected by the Rosat PSPC in a 14.2 ksec observation on 30 July 1990 (95% upper limit = 3.6e-4 c/s). Assuming the soft blackbody spectrum that we are currently seeing, this corresponds to an XRT upper limit of 1.6e-4 c/s; far below the minimum observed XRT count rate. While even long-standing SSS can exhibit substantial X-ray low states or declines, long-term optical variations appear limited to ~1 magnitude, much less than the 4 magnitudes reported for CSS081007:030559+054715 in ATEL #1825.

The approximately 45 day modulation, if it is really periodic, is suggestive of a precessing accretion disk; the modulation being caused by disk-related scattering and/or absorption.

We thank the Swift PI, Neil Gehrels, the Swift science team and the Swift mission operations team for their support of these observations.

Swift X-ray and UV light curves of CSS081007:030559+054715