No Arabic abstract
This paper describes an analysis of the NuSTAR data of the fastest-rotating magnetar 1E 1547$-$5408, acquired in 2016 April for a time lapse of 151 ks. The source was detected with a 1-60 keV flux of $1.7 times 10^{-11}$ ergs s$^{-1}$ cm$^{-2}$, and its pulsation at a period of $2.086710(5)$ sec. In 8-25 keV, the pulses were phase-modulated with a period of $T=36.0 pm 2.3$ ks, and an amplitude of $sim 0.2$ sec. This reconfirms the Suzaku discovery of the same effect at $T=36.0 ^{+4.5}_{-2.5} $ ks, made in the 2009 outburst. These results strengthen the view derived from the Suzaku data, that this magnetar performs free precession as a result of its axial deformation by $sim 0.6 times 10^{-4}$, possibly caused by internal toroidal magnetic fields reaching $sim 10^{16}$ G. Like in the Suzaku case, the modulation was not detected in energies below $sim 8$ keV. Above 10 keV, the pulse-phase behaviour, including the 36 ks modulation parameters, exhibited complex energy dependences: at $sim 22$ keV, the modulation amplitude increased to $sim 0.5$ sec, and the modulation phase changed by $sim 65^circ$ over 10--27 keV, followed by a phase reversal. Although the pulse significance and pulsed fraction were originally very low in $>10$ keV, they both increased noticeably, when the arrival times of individual photons were corrected for these systematic pulse-phase variations. Possible origins of these complex phenomena are discussed, in terms of several physical processes that are specific to ultra-strong magnetic fields.
The Suzaku data of the highly variable magnetar 1E 1547.0$-$5408, obtained during the 2009 January activity, were reanalyzed. The 2.07 s pulsation of the 15--40 keV emission detected with the HXD was found to be phase modulated, with a period of $36.0^{+4.5}_{-2.5}$ ks and an amplitude of $0.52 pm 0.14$ s. The modulation waveform is suggested to be more square-wave like rather than sinusoidal. While the effect was confirmed with the 10--14 keV XIS data, the modulation amplitude decreased towards lower energies, becoming consistent with 0 below 4 keV. After the case of 4U 0142+61, this makes the 2nd example of this kind of behavior detected from magnetars. The effect can be interpreted as a manifestation of free precession of this magnetar, which is suggested to be oblately deformed under the presence of strong toroidal field of $sim 10^{16}$ G.
Archival NuSTAR data of the magnetar 4U0142+61, acquired in 2014 March for a total time span of 258 ks, were analyzed. This is to reconfirm the 55 ks modulation in the hard X-ray pulse phases of this source, found with a Suzaku observation in 2009 (Makishima et al. 2014). Indeed, the 10-70 keV X-ray pulsation, detected with NuSTAR at 8.68917s, was found to be also phase-modulated (at 98 percent confidence or higher) at the same 55 ks period, or half that value. Furthermore, a brief analysis of another Suzaku data set of 4U0142+61, acquired in 2013, reconfirmed the same 55 ks phase modulation in the 15-40 keV pulses. Thus, the hard X-ray pulse-phase modulation was detected with Suzaku (in 2009 and 2013) and NuSTAR (in 2014) at a consistent period. However, the modulation amplitude varied significantly; A=0.17 s with Suzaku (2009), A=1.2 s with Suzaku (2013), and A=0.17 s with NuSTAR. In addition, the phase modulation properties detected with NuSTAR differed considerably between the first 1/3 and the latter 2/3 of the observation. In energies below 10 keV, the pulse-phase modulation was not detected with either Suzaku or NuSTAR. These results reinforce the view of Makishima et al. (2014); the neutron star in 4U0142+61 keeps free precession, under a slight axial deformation due probably to ultra-strong toroidal magnetic fields of 1e16 G. The wobbling angle of precession should remain constant, but the pulse-phase modulation amplitude varies on time scales of months to years, presumably as asymmetry of the hard X-ray emission pattern around the stars axis changes.
A bright burst, followed by an X-ray tail lasting ~10 ks, was detected during an XMM-Newton observation of the magnetar 1E 1547.0-5408 carried out on 2009 February 3. The burst, also observed by SWIFT/BAT, had a spectrum well fit by the sum of two blackbodies with temperatures of ~4 keV and 10 keV and a fluence in the 0.3-150 keV energy range of ~1e-5 erg/cm2. The X-ray tail had a fluence of ~4e-8 erg/cm2. Thanks to the knowledge of the distances and relative optical depths of three dust clouds between us and 1E 1547.0-5408, we show that most of the X-rays in the tail can be explained by dust scattering of the burst emission, except for the first ~20-30 s. We point out that other X-ray tails observed after strong magnetar bursts may contain a non-negligible contribution due to dust scattering.
We present the evolution of the X-ray emission properties of the magnetar 1E 1547.0-5408 since February 2004 over a time period covering three outbursts. We analyzed new and archival observations taken with the Swift, NuSTAR, Chandra and XMM-Newton X-ray satellites. The source has been observed at a relatively steady soft X-ray flux of $approx$ 10$^{-11}$ erg cm$^{-2}$ s$^{-1}$ (0.3-10 keV) over the last 9 years, which is about an order of magnitude fainter than the flux at the peak of the last outburst in 2009, but a factor of $sim$ 30 larger than the level in 2006. The broad-band spectrum extracted from two recent NuSTAR observations in April 2016 and February 2019 showed a faint hard X-ray emission up to $sim$ 70 keV. Its spectrum is adequately described by a flat power law component, and its flux is $sim$ $7 times 10^{-12}$ erg cm$^{-2}$ s$^{-1}$ (10-70 keV), that is a factor of $sim$ 20 smaller than at the peak of the 2009 outburst. The hard X-ray spectral shape has flattened significantly in time, which is at variance with the overall cooling trend of the soft X-ray component. The pulse profile extracted from these NuSTAR pointings displays variability in shape and amplitude with energy (up to $approx$ 25 keV). Our analysis shows that the flux of 1E 1547.0-5408 is not yet decaying to the 2006 level and that the source has been lingering in a stable, high-intensity state for several years. This might suggest that magnetars can hop among distinct persistent states that are probably connected to outburst episodes and that their persistent thermal emission can be almost entirely powered by the dissipation of currents in the corona.
The fastest-rotating magnetar 1E 1547.0-5408 was observed in broad-band X-rays with Suzaku for 33 ks on 2009 January 28-29, 7 days after the onset of its latest bursting activity. After removing burst events, the absorption-uncorrected 2-10 keV flux of the persistent emission was measured with the XIS as 5.7e-11 ergs cm-2 s-1, which is 1-2 orders of magnitude higher than was measured in 2006 and 2007 when the source was less active. The persistent emission was also detected significantly with the HXD in >10 keV up to at least ~110 keV, with an even higher flux of 1.3e-10 ergs cm-2 s-1 in 20-100 keV. The pulsation was detected at least up to 70 keV at a period of 2.072135+/-0.00005 s, with a deeper modulation than was measured in a fainter state. The phase-averaged 0.7-114 keV spectrum was reproduced by an absorbed blackbody emission with a temperature of 0.65+/-0.02 keV, plus a hard power-law with a photon index of ~1.5. At a distance of 9 kpc, the bolometric luminosity of the blackbody and the 2-100 keV luminosity of the hard power-law are estimated as (6.2+/-1.2)e+35 ergs s-1 and 1.9e+36 ergs s-1, respectively, while the blackbody radius becomes ~5 km. Although the source had not been detected significantly in hard X-rays during the past fainter states, a comparison of the present and past spectra in energies below 10 keV suggests that the hard component is more enhanced than the soft X-ray component during the persistent activity.