Detailed observations of Cyg X-1 made using the PCA on-board the RXTE (Cui et al., 1997a,b; Belloni et al., 1996) have shown that the power density spectrum (PDS) of the source changes smoothly during the spectral transition from the hard state to the soft state and reverts back when there is a transition back to the hard state. The power-law index for the PDS below 0.3 Hz for the transition state was found to be between -0.3 to -0.7. The observations carried out by the PPCs in 1996 May pertain to the hard state of the source, where there were no PCA observations. We find that the index is still flatter in the hard state, about -0.09, and the PDS shape agrees very well with the traditional hard state PDS obtained earlier (Belloni & Hasinger 1990a). These observations reinforce the conclusion obtained by Cui et al. that the changes in the PDS smoothly follow the spectral changes.
Chakrabarti & Titarchuk (1995) have worked out a complete solution of viscous transonic equations and have identified the various X-ray emission regions in an accretion disc. Here we attempt to identify the different components in the PDS with the different regions of the accretion disk as described by Chakrabarti & Titarchuk (1985). Most of the white noise component in the PDS can originate in the post-shock region. In the soft state, the Compton cooling is very efficient resulting in bulk motions which can lead to much higher red noise in the PDS. The very low frequency component in the PDS (at Hz) identified in the present work appears to be independent of the spectral states. The origin of this component could be from the pre-shock sub-Keplerian component.
The distribution of shots in Cyg X-1 has been explained under the premises of self-organized criticality (SOC) model (Mineshige et al. 1994a). Negoro et al. (1995) have examined this model by analyzing the Ginga data for the hard state of Cyg X-1. They have also put a selection criterion for shots, namely, the shot peak count rates should be higher than the running average by a pre-determined factor, p. It was found that the shot widths follow an exponential distribution with time constants of 1.8 s, 8.4 s and 23.8 s for values of p 1.5, 2 and 2.35, respectively. For our analysis, we have taken all shots (p is 1) and find that the shot width has an exponential distribution with time constant of 1.4 s. This value appears to be consistent with 1.8 s (for p 1.8), since the time constant decreases with decreasing p. The important conclusion that can be drawn from our work is that the shot width as well as the total shot intensity has an exponential distribution in both the spectral states. The distinct change in the shot distribution for the two spectral states is a new input for the shot modeling and it probably indicates the differences in the basic model parameters like the critical mass density, diffusion coefficient etc.
The large shot detected in our work appears similar to the shape of the co-added shots obtained from the Ginga data (Negoro et al. 1994). They have found two time constants each for the rise and decay phase of values 0.1 and 1 s, respectively. The value obtained in the present work (0.4 s) lies in between these two values.
© European Southern Observatory (ESO) 1998
Online publication: January 8, 1998