Swift/BAT Detection of a 5 hour Period in 4U 1954+31
ATel #797; R. Corbet (GSFC/USRA), L. Barbier (GSFC), S. Barthelmy (GSFC), J. Cummings (GSFC/NRC), E. Fenimore (LANL), N. Gehrels (GSFC), D. Hullinger (UMD), H. Krimm (GSFC/USRA), C. Markwardt (GSFC/UMD), D. Palmer (LANL), A. Parsons (GSFC), T. Sakamoto (GSFC/NRC), G. Sato (ISAS), J. Tueller (GSFC), on behalf of the Swift-Survey team
on 20 Apr 2006; 16:27 UT
Credential Certification: Robin Corbet (corbet@gsfc.nasa.gov)
Subjects: X-ray, Binary, Neutron Star, Star, Transient, Variables, Pulsar
The power spectrum of the Swift Burst Alert Telescope (BAT) light
curve of 4U 1954+31 obtained between 2004 December 11 and 2005
September 20 shows, in addition to low frequency noise, modulation at
a period of approximately 5.09 hours. The peak in the power spectrum
near 5 hours shows a complex structure. When the light curve is
examined on shorter time scales it is seen that during the course of
the BAT observations the period changed from approximately 5.19 to
5.02 hours in a near monotonic fashion. The mean count BAT count rate
from 4U 1954+31 in the range 14 - 195 keV is 12.7 (+/- 0.2
statistical) E-4 counts/s/fully illuminated detector (where the area
of a single BAT detector is 0.16 cm^2) which corresponds to 28
mCrab. The Fourier modulation amplitude at 5 hours is approximately
6.5 E-4 counts/s/detector.
The large change in the 5 hour period makes it implausible that it is
caused by orbital modulation. Masetti et al. (2006, astro-ph/0603227)
also determine the optical counterpart of 4U 1954+31 to be an M 4-5
giant which would imply a very long orbital period in order for the
star not to overfill its Roche lobe (cf. the Porb > 100 days
constraint for GX 1+4 where the donor is an M5 III star, Chakrabarty &
Roche, 1997, ApJ, 489, 254). In addition, the period change is too
large for the it to come from the rotation period of an accreting white
dwarf because of the large moment of inertia of these objects.
The 5 hour period is somewhat reminiscent of the 2.7 hour period seen
in the high-mass X-ray binary 2S 0114+650 which is believed to be a
neutron star rotation period (e.g. Farrell, Sood, & O'Neill, 2006,
MNRAS, 367, 1457 and references therein). If the periodicity in 4U
1954+31 does indeed represent the rotation period of a neutron star,
then the rapid spin-up requires an X-ray luminosity of approximately 5
E35 ergs/s (e.g. Joss & Rappaport, 1984, ARAA, 22, 537). In order to
reconcile this luminosity with the observed X-ray flux and the upper
limit on the distance to 4U 1954+319 of 1.7 kpc derived by Masetti et
al. (2006) a high degree of absorption of the X-ray spectrum is
required.
Alternatively, the 5 hour modulation might be caused by some type of
instability in an accretion disk or magnetosphere similar to
the cause of the 3.96 hour
quasi-periodic flares observed on one occasion from EXO 2030+375
(Parmar et al. 1989, ApJ, 338, 359). An explanation of the period as
a pulsation in the mass donor appears problematic as the fundamental
pulsation periods of M giants are very long and there is little
evidence for pulsations on periods less than 10 days
(Koen, Laney, & van Wyk, 2002, MNRAS, 335, 223).