Early Timing and Spectroscopy of Swift J164449.3+573451 with XMM-Newton
ATel #3447; J. M. Miller (Michigan), T. E. Strohmayer (NASA/GSFC)
on 22 Jun 2011; 16:05 UT
Credential Certification: Jon Miller (jonmm@umich.edu)
Subjects: X-ray, AGN, Black Hole, Blazar, Gamma-Ray Burst, Transient
We report on an analysis of four bi-weekly 25 ksec observations of
Swift J164449.3+573451. Following the identification of this source
as a strong tidal disruption candidate, we requested observations with
XMM-Newton in order to leverage its large collecting area and
throughput. These observations were obtained on 16 April, 30 April,
16 May, and 30 May 2011.
A detailed spectral analysis of the EPIC-pn spectra was made in the
0.5-10.0 keV band. A simple absorbed power-law function yielded a
formally acceptable fit to each spectrum. The first, third, and
fourth observations show a trend seen in black holes across the mass
scale: steeper power-law indices [Gamma = 1.88(1), 1.66(1), 1.61(1)]
are measured at higher average flux levels [F = 6.3(2), 4.9(2), and
1.30(3) E-11 erg/cm/cm/s). The third observation does not obey this
trend (Gamma = 1.97(1), F = 4.5(2) E-12 erg/cm/cm/s). When the
observations jointly constrain the equivalent neutral hydrogen column
density in the host frame, it is found to be N_H = 1.54 +/- 0.01 E+22
(1 sigma confidence).
There are no significant emission or absorption lines in the EPIC-pn
spectra. Fits to the first observation place a 90% confidence upper
limit of just 6 eV on the strength of a narrow Fe K line in the source
spectrum. Fits with a relativistic line give a 90% confidence upper
limit of 10 eV. Both limits are below typical values observed in
Seyfert AGN (e.g. Nandra et al. 2007, MNRAS, 382, 194).
Fourier transforms of light curves from all EPIC cameras were made,
and the resultant power spectra were averaged together. Fits to the
1-4 keV (observed frame) power spectrum are consistent with a simple
power-law with a slope of 2.15, and an average fractional rms (0.4 -
100 mHz) of about 8.5%. The 4-12 keV power spectrum is not as well fit
by a simple power-law, showing evidence for an excess near 5 mHz that
may be a weak QPO. The inclusion of a Lorentzian (nu = 4.7 mHz, width
= 1.1 mHz) improves the fit, and an F-test suggests the additional QPO
parameters are significant at the 99% confidence level. It is common
for black hole QPOs to show an increase in amplitude with energy, so
the detection of a QPO in the harder X-rays would not be particularly
unusual (e.g. Strohmayer 2001, ApJ, 552, L49; Miller et al. 2001, ApJ,
563, 928). The average fractional rms of the continuum (0.4 - 100 mHz)
and Lorentzian components are approximately 12% and 4%, respectively.
The present results are suggestive of a 4.7 mHz QPO in Swift
J164449.3+573451 but additional observations are needed to confirm or
reject this potential signal.
If the feature is confirmed, it would represent only the second X-ray
QPO detected in the flux of a massive black hole, and the first QPO
detected in a tidal disruption candidate. The QPOs would likely
signal the formation of an accretion disk in Swift J164449.3+573451.
Last, if the putative QPO traces the innermost stable circular orbit
around a Schwarzschild black hole, it would imply a black hole mass of
4.7 E+5 Msun, or log(M) = 5.7. This would be in line with
predictions based on simultaneous X-ray and radio observations (log M
= 5.9 +/- 1.1; Miller & Gultekin 2011, arxiv:1106.2502), and
expectations based on the luminosity of the host galaxy (Greene & Ho,
2007, ApJ, 670, 92).
A plot of the 4-12 keV power spectrum is available here:
http://www.astro.lsa.umich.edu/~jonmm/pds.ps
We thank Norbert Schartel and the XMM-Newton staff for executing
monitoring observations of Swift J164449.3+573451.