Brightening and hardening of new X-ray transient in globular cluster Terzan 5

ATel #4249; C. O. Heinke (U. of Alberta), R. Wijnands, D. Altamirano (U. of Amsterdam), D. Pooley(Sam Houston State U., Eureka Scientific), G. R. Sivakoff (U. of Alberta)
on 10 Jul 2012; 21:19 UT
Credential Certification: Craig Heinke (cheinke@virginia.edu)

Tweet

We have followed the new X-ray transient in Terzan 5 (ATEL #4242) with daily Swift/XRT observations, finding it to be continuously brightening. We name this transient as Swift J174805.3-244637, the third confirmed transient in Terzan 5. We observed Terzan 5 with the XRT in PC mode on 2012 July 7 (20:27 UT, 0.95 ks), 8 (23:13 UT, 0.97 ks), and 10 (02:30 UT, 0.76 ks), in each case identifying a single source consistent with the position reported in Atel #4242. Although the count rate of the July 10 observation reached 0.54 cts/s, we did not find any evidence of pileup when comparing the radial count distribution with the known Swift point spread function (see http://www.swift.ac.uk/analysis/xrt/pileup.php ). We also found no change in the photon index by using spectra from different annuli (2-20 and 4-20 pixel radii) of the last dataset.

We extracted the four Swift/XRT spectra so far from 20 pixel circles, grouped them to 20 cts/bin (except the July 6 observation to 10 cts/bin), and fit them with absorbed power laws (tbabs*pegpwrlw) in XSPEC. We find a chi-squared of 35.5 for 37 degrees of freedom. We tied the N_H of all spectra, finding N_H=1.8+0.4/-0.3e22 cm-2, consistent with the expected Galactic absorption to Terzan 5. We find absorbed (unabsorbed) 0.5-10 keV flux values of 4e-12 (1.3+1.0/-0.4e-11), 9e-12 (3+2/-1e-11), 1.5e-11 (4+2/-1e-11), and 4.1e-11 (8+2/-1e-11) ergs cm-2/s. For a distance of 5.9 kpc (+/-0.5 kpc, Valenti et al. 2007, AJ, 133, 1287), the inferred 0.5-10 keV unabsorbed Lx values are thus 5+4/-1e34, 1.2+0.8/-0.3e35, 1.7+1.0/-0.4e35, and 3.4+0.9/-0.5e35 respectively.

We found that the spectrum may have hardened with increasing Lx. If N_H is fixed to the best-fit, the photon index fits are respectively 2.6+/-0.6, 2.6+/-0.4, 2.4+/-0.3, and 2.0+/-0.2. An f-test on a spectral fit with or without the photon index held constant gives a 5% chance of attaining the improvement in chi-squared (from 43.4/41 dof, to 35.5/38 dof) by chance. The reverse, spectral softening as a source fades into quiescence, has been seen in several other transients in this Lx range (e.g., Corbel et al. 2006, ApJ, 636, 971; Armas Padilla et al. 2011, MNRAS, 417, 659). We note that this appears to be a rather slow rise, allowing us to obtain unusual detail on the rise at still quite low luminosities.

We have triggered a 10-ks Chandra observation, to occur around July 26, to pinpoint the source position, and are continuing to follow the outburst with Swift. It may be possible to identify the infrared counterpart with adaptive optics observations of this dense and highly reddened cluster, as done for another X-ray transient in Terzan 5 (IGR J17480-2446,see Testa et al., ATEL #3264).

We thank the Swift team for their support.