The doubling of the superorbital period of Cyg X-1

TL;DR

This study reports a doubling of the superorbital period of Cyg X-1 to about 300 days, observed since 2005, and explores its implications for the system's accretion disc and emission properties.

Contribution

It provides the first detailed analysis of the superorbital period doubling in Cyg X-1 and links it to changes in the accretion disc's geometry and absorption without altering the intrinsic spectrum.

Findings

Superorbital period increased from ~150 d to ~300 d since 2005.

Superorbital modulation affects normalization and absorption, not spectral shape.

No correlation found between X-ray/radio properties and gamma-ray detections.

Abstract

We study properties of the superorbital modulation of the X-ray emission of Cyg X-1. We find that it has had a stable period of about 300 d in soft and hard X-rays and in radio since 2005 until at least 2010, which is about double the previously seen period. This new period, seen in the hard spectral state only, is detected not only in the light curves but also in soft X-ray hardness ratios and in the amplitude of the orbital modulation. On the other hand, the spectral slope in hard X-rays, >20 keV, averaged over superorbital bins is constant, and the soft and hard X-rays and the radio emission change in phase. This shows that the superorbital variability consists of changing the normalization of an intrinsic spectrum of a constant shape and of changes of the absorbing column density with the phase. The maximum column density is achieved at the superorbital minimum. The amplitude changes are likely to be caused by a changing viewing angle of an anisotropic emitter, most likely a precessing accretion disc. The constant shape of the intrinsic spectrum shows that this modulation is not caused by a changing accretion rate. The modulated absorbing column density shows the presence of a bulge around the disc centre, as proposed previously. We also find the change of the superorbital period from about 150 d to about 300 d to be associated with almost unchanged average X-ray fluxes, making the period change difficult to explain in the framework of disc-irradiation models. Finally, we find no correlation of the X-ray and radio properties with the reported detections in the GeV and TeV gamma-ray range.