[ Previous | Next ]
ATEL # 1992; P.C. Schmidtke (Arizona State University), A.P. Cowley (Arizona State University), A. Udalski (Warsaw University Observatory)
on 28 Mar 2009; 1:28 UT
Password Certification: Paul Schmidtke (Paul.Schmidtke@asu.edu)
Subjects: Optical, X-ray, Binaries, Neutron Stars, Pulsars
SMC X-2 is a transient source, dating back to the earliest X-ray studies of the Magellanic Clouds (Li et al. 1977, IAU Circ. 3125). During a bright phase in 2000, Corbet et al. (2001, ApJ 548, L44) discovered X-ray pulsations with P=2.37 s, but no orbital period has been proposed using X-ray data (e.g. Galache et al. 2008, ApJS 177, 189). The best position for this source coincides with a blend of two early-type stars (e.g. Murdin et al. 1979, MNRAS 186, 43P), which are partially resolved (2.5" apart) in the OGLE-III data base.
Schmidtke et al. (2006, AJ 132, 971) examined the first 5 seasons of OGLE-III photometry for each component. The brighter, northern star shows large "swooping" I-band variations of ~1 mag, typical of Be/X-ray pulsar binaries. The southern star shows much smaller variations and appears to be contaminated by light from the brighter star. No coherent signal was found in either component. Schurch et al. (2008, ATel #1670) studied more recent OGLE-III data from seasons 4-7 and reported the discovery of a periodicity of P=18.62 days and possible harmonics at 9.31 and 6.21 days. They suggested the 18-day signal is the system's orbital period.
We have revisited the entire OGLE-III data set, which includes an 8th season of photometry. Seasons 1-3 have large variations that cannot be adequately removed when searching for low-amplitude signals. Hence, these seasons were not included in our study. Data from each of the remaining seasons (4-8) were detrended individually prior to further analysis. The photometric scatter is noticeably smaller in seasons 4 and 5 than in seasons 6-8. This is confirmed by periodograms that show significant power only in the last 3 seasons. Signals are present near 18, 9, and 6 days, as well as aliases with strong power at higher frequencies. In particular, light curves folded on P=0.8592 and 0.9008 days have sinusoidal shapes, which can be attributed to nonradial pulsations of the primary star. Each of these periods shows small changes from season to season. The power at P=0.90 is approximately constant throughout seasons 6-8, while power at P=0.86 steadily increases. We suggest the 9-day and 6-day signals, which have folded light curves that are sinusoidal, are aliases of these short-period pulsations. The expected beat period between the two pulsation signals is 18.60 days, consistent with the peak present in our periodograms as well as that reported by Schurch et al. The light curve, folded on this long period, is not sinusoidal, but has a shape consistent with that expected from the beating of two short-period signals.