NGC 300 ULX-1 / SN 2010da: Early Swift period measurement and modeling of spin-up

ATel #11229; J. A. Kennea (PSU)
on 25 Jan 2018; 22:05 UT
Credential Certification: Jamie A. Kennea (kennea@astro.psu.edu)

Tweet

We report on analysis of archival Swift/XRT observations on the supernova impostor SN 2010da AKA NGC 300 ULX-1, which has seen a recent flurry of reports of its rapid spin evolution (e.g. ATEL #11158, ATEL #11174, ATEL #11179, ATEL #11228). A total of 9 Swift/XRT observations of this source have been taken which have sufficient exposure and in which the source was sufficiently bright enough to detect pulsations. These observations were taken between April 25th, 2016 and January 25th, 2018.

A 2909s exposure Photon Counting (PC) mode observation taken on April 25th, 2016 is the earliest observation of this source in which we are able to detect a coherent periodicity, and is reported on for the first time here.

In order to perform a period search of these data, we corrected the photon arrival times to the solar system barycenter, and randomized the photon arrival times within the ~2.5s PC mode frame time to avoid aliasing issues. Performing a standard Z²₂ search of the data, we find a pulsar period of P = 44.18072 +/- 0.00079 s for this epoch. The quoted error is 1-sigma and estimated using the Monte-Carlo method described by Gotthelf et al. (1999, ApJ, 522, L49).

With measurements of the pulsar spin period spanning ~640 days, including the XMM-Newton measured value reported by Carpano et al. (ATEL #11158), we find that the spin-up cannot be described by a simple linear model. Therefore, we fit a standard pulsar spin-up model incorporating both first and second derivatives of the spin period. Fitting this model, we find that the global spin-up evolution of NGC 300 ULX-1 can be described by P_dot = -7.54 +/- 0.07 x 10^-7 s s^-1 and P_dot-dot = -1.04 +/- 0.03 x 10^-9 s s^-2.

We note that the fit has significant scatter in the residuals, likely due to the unfit contribution of the orbital motion and other effects, however the model well describes the bulk spin-up of the pulsar over a large range of spin periods (20 s to 44.2 s) over a 21 month period.

We also note that after removing aliasing effects of PC mode by randomizing the photon arrival times, we do not detect the 30.05s period reported by Grebenev et al. (ATEL #11174) in data taken by Swift/XRT on April 16, 2017, confirming that this was an erroneous detection caused by aliasing of the PC mode frame time, as suggested by Vasilopoulos et al. (ATEL #11179). However, we do detect a periodicity in those data at P = 26.711 +/- 0.002 s, consistent with other reported periods derived from observations taken around that time.

Continued observations to constrain both the spin-up and orbital parameters of this system are encouraged.