Pulse-phase resolved spectroscopy of continuum and reflection in SAX J1808.4-3658

TL;DR

This study conducts phase-resolved spectroscopy of SAX J1808.4-3658 during outburst decay, revealing that iron line pulsations are likely due to Doppler effects on blackbody temperature rather than direct pulsation, and identifying phase shifts between spectral components.

Contribution

It introduces a detailed analysis of spectral variations, distinguishing between iron line pulsations and blackbody temperature changes, and proposes a model explaining observed phase shifts due to Doppler effects.

Findings

Iron line pulsations are not intrinsic but due to Doppler-shifted blackbody temperature.

Blackbody temperature variations explain spectral pulsations better than iron line pulsations.

A larger phase-shift between blackbody and Comptonised components suggests Doppler effects dominate surface emission variations.

Abstract

We perform phase-resolved spectroscopy of the accreting millisecond pulsar, SAX J1808.4-3658, during the slow-decay phase of the 2002 outburst. Simple phenomenological fits to RXTE PCA data reveal a pulsation in the iron line at the spin frequency of the neutron star. However, fitting more complex spectral models reveals a degeneracy between iron-line pulsations and changes in the underlying hotspot blackbody temperature with phase. By comparing with the variations in reflection continuum, which are much weaker than the iron line variations, we infer that the iron-line is not pulsed. The observed spectral variations can be explained by variations in blackbody temperature associated with rotational Doppler shifts at the neutron star surface. By allowing blackbody temperature to vary in this way, we also find a larger phase-shift between the pulsations in the Comptonised and blackbody components than has been seen in previous work. The phase-shift between the pulsation in the blackbody temperature and normalisation is consistent with a simple model where the Doppler shift is maximised at the limb of the neutron star, ~90 degrees prior to maximisation of the hot-spot projected area.