Detection of radio pulsations at 22 GHz from the Magnetar PSR J1745-2900 in the archival data from 2011

ATel #5076; Divya Palaniswamy (International Centre for Radio Astronomy Research: ICRAR), N. D. Ramesh Bhat (ICRAR), Steven J. Tingay (ICRAR), Simon P. Ellingsen (University of Tasmania), Randall B. Wayth (ICRAR), Cathryn M. Trott (ICRAR)
on 20 May 2013; 15:04 UT
Credential Certification: Ramesh Bhat (ramesh.bhat@curtin.edu.au)

Tweet

We report on the detection of radio pulsations at 22 GHz from the magnetar PSR J1745-2900, recently discovered by the NuSTAR X-ray telescope in the direction of the Galactic Centre (ATel #5020), using archival data from observations made with the 26 metre radio telescope at the Mount Pleasant radio observatory near Hobart, operated by the University of Tasmania.

The data were recorded on 7 November 2011 using the Australian Long Baseline Array baseband recorder for a duration of 6 hr; both the left and right circularly polarized signals were recorded over a bandwidth of 64 MHz centered at 22.235 GHz. These data were processed to generate filterbank data streams with effective time resolution of 640 μs and 128 spectral channels across the 64 MHz bandwidth.

As the dispersion delays are negligibly small at such high frequency even for DM ~ 1700 (i.e. the DM of the magnetar), we synchronously folded the data at the pulse period reported by the NuSTAR telescope. Data processing using the PRESTO pulsar software package yielded a ~ 5 sigma detection from the first 4 hr of data, while no detection resulted from the last 2 hr of observations. For the sensitivity estimated for our observations (i.e. a system equivalent flux density of ~ 2100 Jy), this implies a mean flux of ~ 4±1 mJy for the observed duty cycle of ~ 25%. This suggests that the magnetar may have been significantly brighter in the past compared to recent times (ATels #5040, #5043, #5053, #5058). The non-detection in the last 2 hr can be partly reconciled given the reduced system sensitivity recorded for that time, or perhaps it is to do with the flux variability typical to magnetars, or possibly due to scintillation.

Folding the data for multiple different time durations within the first 4 hr (e.g. 1, 2, 3 and 4 hr) consistently resulted in clear, albeit marginal, detections (~5 sigma), with similar duty cycles and the pulse appearing nearly at the same phase. There are some hints of possible pulse shape variations over the 4 hr time span, however the details are difficult to discern given our signal-to-noise limitations. A closer examination confirms that the data were not corrupted by terrestrial radio interference or instrumental effects. Our analysis suggests that the magnetar may have been radio active and significantly brighter at the time of our observations.

As there has not been any record of X-ray activity near our observing epoch (MJD 55872), this detection seems akin to that of the magnetar PSR J1622-4950 that was first discovered in radio while it was in an X-ray quiescent state. Furthermore, the implied flux variability of a factor of 10 between our observations and recent detections is comparable to what has been seen for other radio magnetars including PSR J1622-4950, whose flux was found to vary by a similar factor over long time spans.

Based on our current analysis and the observational history of other known radio magnetars, it appears that the Galactic Centre magnetar may have been radio active in the past, and perhaps escaped a detection due to X-ray quiescence.

Further investigations and follow-up observations are currently under way.

Detection plot