[Previous] | [Session 112] | [Next]
W. B. Focke (GSFC / UMCP)
Shot noise has long been used to model the aperiodic variability of Cygnus X-1. Past investigations have been largely restricted to aggregate properties of the shots. We apply a peak detection algorithm to hard state light curves to select individual shots. Analyzing 86000~s of data collected by the Rossi X-ray Timing Explorer (RXTE), we find about 36000 shots. The detected shots are fit to several model templates, and the results of these fits are explored statistically.
The fit shot parameters are found to be distributed, so that a single shot profile is not sufficient to describe the data. The shot peak intensities follow an exponential distribution, while their timescales (FWHM) roughly follow a powerlaw. The distribution of asymmetry (ratio of risetime to falltime) of the shots, both for slow and fast risers, follow similar powerlaws. There are slightly more slow than fast risers, but the majority of the population is approximately symmetric. The distribution of waiting times between shots follows a broken exponential. Correlations between the FWHM and fluence of the shots is observed. This correlation, and the drop in the waiting-time distribution for short times, are both in accord with the predictions of the self-organized criticality model for the origin of the shots.
The cross spectrum of light curves in different energy bands yields information both through the phase lags and the coherence. The large, frequency-dependent time lags and high coherence previously observed in this source are confirmed. The uniformity of the transfer function implies that the observed distribution of shot widths cannot have been acquired through Compton scattering -- the widths must be intrinsic to the mechanism that produces the shots.
The author(s) of this abstract have provided an email address for comments about the abstract: Warren.B.Focke.1@gsfc.nasa.gov