[ Previous | Next ]
ATEL # 893; S. Campana (INAF-OAB), G.L. Israel (INAF-OAR)
on 22 Sep 2006; 1:32 UT
Distributed as an Instant Email Notice (Transients)
Password Certification: Sergio Campana (email@example.com)
Subjects: X-ray, Neutron Stars, Soft Gamma-ray Repeaters, Transients
Referred to by ATEL #: 894
Swift performed a TOO observation of the anomalous X-ray pulsar (magnetar) CXO J164710.2-455216 in the cluster Westerlund 1 (Muno et al. 2006, ApJ 636 L41) following the burst detected by the BAT (Krimm et al., GCN 5581). This observation started Sep 21 2006, 14:30:02 UT, i.e. 12.4 hr after the BAT event. In the first dump of data (2.9 ks) we detect a bright source (about 0.6 c/s) at a position RA(2000) = 16 47 10.53 Dec(2000)=-45 52 14.3 with an error box of 3.6 arcsec (90% confidence, throughout the text). The source lies at 3.9 arcsec from the Chandra position of CXO J164710.2-455216 (0.5 arcsec uncertainty). The source is clearly fading. We identify the Swift source with the AXP. No additional bursts are apparent in the data. Within an extraction region of 20 pixel we collected 1598 counts. The source is heavily absorbed. A blackbody model can adequately fit the 1-10 keV data (chi2_red=1.1 with 69 dof). The column density (1.1+/-0.2)x10^22 cm^-2 is consistent with Chandra data. The black body temperature and radius however are larger T_BB=0.83+/-0.04 keV (w.r.t. 0.61+/-0.02 keV obtained by Chandra on May 22 and June 16 2005, Muno et al. 2006) and R_BB=1.5+/-0.2 km (w.r.t. 0.27+/-0.03 km) at a fiducial distance of 5 kpc. The black body unabsorbed 0.5-8 keV flux is 4.5x10^-11 erg cm^-2 s^-1, about 190 times the quiescent Chandra level. At 5 kpc the source luminosity is 10^35 erg s^-1. For completeness we note that the XRT spectrum can also be fit with a single power law with photon index Gamma=3.0+/0.2 and column density NH=(3.0+/-0.3)x10^22 cm^-2 (chi2_red=1.0). In addition to the observed spectral changes we did not detect a strong pulsed signal in the barycentric-corrected flux at variance with observations in quiescence during which a pulse period of 10.6107 s was discovered (Muno et al. 2006). Adopting a folding technique we identify a low-significance period at P=10.613+/-0.001 s. If this period will be confirmed by further observations a spin period derivative of the order of Pdot~5x10^-11 s s^-1 is expected. Further observations are encouraged.