51ÁÔÆæÈë¿Ú

Abstract:

Cyclotron line scattering features are detected in a few tens of X-ray pulsars (XRPs) and used as direct indicators of a strong magnetic field at the surface of accreting neutron stars (NSs). In a few cases, cyclotron lines are known to be variable with the accretion luminosity of XRPs. It is accepted that the observed variations of cyclotron line scattering features are related to variations of geometry and dynamics of accretion flow above the magnetic poles of an NS. A positive correlation between the line centroid energy and luminosity is typical for subcritical XRPs, where the accretion results in hot spots at the magnetic poles. The negative correlation was proposed to be a specific feature of bright supercritical XRPs, where radiation pressure supports accretion columns above the stellar surface. The cyclotron line in the spectra of the Be-transient XRP GRO J1008-57 is detected at energies from ~75 to 90 keV, the highest observed energy of a cyclotron line feature in XRPs. We report the peculiar relation of cyclotron line centroid energies with luminosity in GRO J1008-57 during the Type II outburst in 2017 August observed by Insight-HXMT. The cyclotron line energy was detected to be negatively correlated with the luminosity at 3.2 x 1037 erg s-1 < L < 4.2 × 1037 erg s-1 and positively correlated at L 5 x 1037 erg s-1. We speculate that the observed peculiar behavior of a cyclotron line would be due to variations of accretion channel geometry.

Abstract:

We report on a timing programme of 74 young pulsars that have been observed by the Parkes 64-m radio telescope over the past decade. Using modern Bayesian timing techniques, we have measured the properties of 124 glitches in 52 of these pulsars, of which 74 are new. We demonstrate that the glitch sample is complete to fractional increases in spin-frequency greater than $\Delta\nu^{90\%}_{g}/\nu \approx 8.1 \times 10^{-9}$. We measure values of the braking index, $n$, in 33 pulsars. In most of these pulsars, their rotational evolution is dominated by episodes of spin-down with $n > 10$, punctuated by step changes in the spin-down rate at the time of a large glitch. The step changes are such that, averaged over the glitches, the long-term $n$ is small. We find a near one-to-one relationship between the inter-glitch value of $n$ and the change in spin-down of the previous glitch divided by the inter-glitch time interval. We discuss the results in the context of a range of physical models.