Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 342, 736-744 (1999)

Previous Section Next Section Title Page Table of Contents

5. Discussion and conclusions

In this paper we have presented the temporal evolution of spectral parameters of Her X-1 during a pre-eclipse dip. Due to the large effective area of the PCA, the temporal resolution of our data is higher than that of earlier data. Using two different methods we were able to show that a partial covering model in which the column density [FORMULA] and the normalization of the unabsorbed continuum [FORMULA] are the only time variable spectral components is sufficient to explain the observed spectral and temporal variability. This result is in qualitative agreement with the previous analyses for pre-eclipse dips presented by Ushimaru et al. 1989, Choi et al. 1994, Leahy et al. 1994, and Reynolds & Parmar 1995. Note that models without an unabsorbed component were not able to result in satisfactory fits, which is also consistent with the low state observations presented by Mihara et al. 1991 where the unabsorbed component was clearly required. As proposed by Choi et al. 1994 from the analysis of the pulsed fraction of the lightcurve, the unabsorbed component is probably due to scattering of radiation in an extended hot electron corona into the line of sight. These authors also show that the interpretation of Ushimaru et al. 1989, who attributed the unabsorbed component to a leaky cold absorber, does not hold.

The anticorrelation between [FORMULA] and [FORMULA] found by Leahy et al. 1994 and Leahy 1997 in Ginga observations has been interpreted by these authors as evidence that the obscuring material partially also obscures the extended corona. Thus, the geometric covering fraction of the obscuring material is assumed to be quite high during episodes of large [FORMULA]. Our measured values of [FORMULA] exhibit some variability (Fig. 4c), but there is no systematic correlation between [FORMULA] and [FORMULA], neither has such a correlation been seen by Choi et al. 1994. A possible interpretation for this discrepancy is that Leahy et al. 1994 and Leahy 1997 did not include Thomson scattering in their fits. Indeed, when setting [FORMULA] in Eq. (1) and fitting the RXTE data with both normalizations as free parameters, [FORMULA] is much more variable and [FORMULA] appears to be correlated with [FORMULA]. In addition, it is generally difficult to distinguish between the absorbed and the unabsorbed component for low values of [FORMULA] such that [FORMULA] and [FORMULA] get easily confused by the fitting routine.

We also tried fitting the data with a spectral model in which [FORMULA] was held fixed at 0.005 ph cm-2 s-1 keV-1, the average value of the fits of Sect. 3.2. The resulting [FORMULA] values from this fit were comparable to those of the fits presented in Sect. 3.2 since the variations in [FORMULA] in the latter fit are small enough to be compensated by slight changes in [FORMULA] in the former. Furthermore, due to the lower temporal resolution of the previous observations, small variations of [FORMULA] could not be resolved in these data. It has been pointed out by Parmar et al. 1986, that this effect might result in a large uncertainty in the determination of [FORMULA]. As is shown by our fits to structures in the lightcurve (Fig. 8), we are able to resolve and identify individual structures with a temporal resolution of about one minute. Thus we are confident that the investigation presented here is unaffected by these problems.

In our analysis of Sect. 3.2 and 3.3 we assumed that [FORMULA], i.e., the value appropriate for material of solar composition. Previous investigations of dipping sources, however, hinted at non-solar abundances in most of these systems (Reynolds & Parmar 1995, White et al.1995). A direct measurement of the abundance ought to be possible by fitting the RXTE data with a partial covering model in which [FORMULA] and [FORMULA] are both free parameters. We find that the average [FORMULA]. Thus, our data could be interpreted as pointing towards a metal overabundance. In contrast to this result, a metal underabundance is favored by Reynolds & Parmar 1995 1. This discrepancy might be due to the higher temporal resolution of our data, where small variations of [FORMULA] can be traced. Note, however, that our method of determining [FORMULA] depends crucially on the assumption that the spectrum incident on the absorbing cloud is constant over time and also depends strongly on the value adopted for [FORMULA]. Although it is very probable that the source is constant (Sect. 3.2), slight variations of [FORMULA] cannot be ruled out. Therefore, we decided to use the solar value of 1.21 in our analysis and postpone the detailed study of the abundances until a larger sample of dips has been observed with RXTE.

To conclude, photoabsorption and electron scattering of photons out of the line of sight in cold material appear to be solely sufficient for explaining the temporal variability of the observed flux during this pre-eclipse dip of Her X-1. Further observations of a larger sample of pre-eclipse dips are necessary to verify this result.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: February 23, 1999