2015MNRAS.449.1238W

The quiescent spectrum of neutron star low-mass X-ray binaries typically consists of two components - a thermal component associated with emission from the neutron star surface, and a non-thermal power-law component whose origin is not well understood. Spectral fitting of neutron star atmosphere models to the thermal component is one of the leading methods for measuring the neutron star radius. However, it has been known for years that the X-ray spectra of quiescent neutron stars vary between observations. While most quiescent variability is explained through a variable power-law component, the brightest and best-studied object, Cen X-4, requires a change in the thermal component and such thermal variability could be a problem for measuring neutron star radii. In this paper, we significantly increase the number of sources whose quiescent spectra have been studied for variability. We examine nine potential quiescent neutron stars with luminosities ≤ 10³⁴ erg/s over the course of multiple Chandra observations of the globular clusters NGC 6440 and Terzan 5 and find no strong evidence for variability in the effective temperature in seven of the nine sources. Two sources show a potential change in temperature, though this depends on the exact model fitted. CX1 in NGC 6440 is equally well fitted by a variable thermal component or a variable power law. Therefore, the results are inconclusive and we cannot exclude or require thermal variability in that source. CX5 in NGC 6440 shows a potential change in temperature, though this depends on whether a power law is included in the spectral fit or not. This suggests that thermal variability may not be widespread among quiescent neutron stars with luminosities ≤ 10³⁴ erg/s, and hence thermal radiation remains a promising means to constraining neutron star radii.