The study of transient dips in the X-ray lightcurves of X-ray binaries can help to provide insight into the accretion processes in these objects. The dips are thought to be caused by cold material with an appreciable column density of and higher passing through the line of sight to the neutron star. Since the dipping activity occurs primarily around the upper conjunction of the neutron star (i.e., shortly before the X-ray eclipse), this material is most probably due to the "splash" at the location where the accretion stream hits the accretion disk. A recent review on X-ray dip energy spectra has been published by White et al.1995.
X-ray dips of the X-ray binary Her X-1 were first described by Giacconi et al.1973. Two different kinds of dips are seen: pre-eclipse dips, which occur at orbital phases -1.0, and anomalous dips, which are observed at . During the "Main High" phase of the 35 d cycle of Her X-1, the pre-eclipse dips seem to "march" towards earlier orbital phases (Giacconi et al.1973). Interpretations of the marching behavior have been given in terms of a periodic mass transfer driven by the changing radiation pressure at the inner Lagrangian point (Crosa & Boynton 1980), and by the splash of the accretion stream onto a warped accretion disk (Schandl 1996). Note that the observational basis for the "marching behavior" has recently been challenged by Leahy 1997.
Previous observational studies dedicated to pre-eclipse dips of Her X-1 have been presented by Ushimaru et al. 1989, Choi et al. 1994, Leahy et al. 1994, and Reynolds & Parmar 1995. These authors found that the dip spectra can be modeled by a temporally variable absorption of the non-dip spectrum. In addition to this absorption component, a further weak unabsorbed component is present during the dips, as first seen in Tenma data (Ushimaru et al. 1989) and verified by later observations. This component has been interpreted as being due to Compton scattering of the primary neutron star radiation by an extended corona into our line of sight. Although the unabsorbed component always contributes to the observed X-ray spectrum, it can only be identified during dips and during the "low state" of Her X-1 (Mihara et al. 1991). Due to the comparably small effective areas of the earlier instruments, a study of the detailed temporal evolution of the spectral parameters during a dip has not been possible in these observations. Therefore, earlier investigators were forced to combine data obtained during (non-consecutive) time intervals with similar measured count-rates to obtain spectra with a signal to noise ratio sufficient for spectral analysis. This approach is not without problems, however, since a strong variation of during the time intervals used for accumulating the spectra can lead to spurious spectral features like an apparent decrease in the flux incident onto the absorbing material (Parmar et al. 1986).
To be able to resolve the structures occuring on short timescales, instruments with a larger effective area and moderate energy resolution are needed. In this paper, therefore, we present results from the analysis of an RXTE observation of Her X-1, focusing on the spectral analysis and the behavior of the lightcurve during a pre-eclipse dip. Results from a previous analysis of these data, focusing on the orbit determination, have been presented by Stelzer et al. 1997. In Sect. 2 we describe our RXTE observation and the data reduction. Sect. 3 presents our results on the temporal evolution of the column density of the absorbing material throughout the dip, using two different data analysis methods. In Sect. 4 we discuss the temporal behavior of the source in terms of a simple model for individual structures in the lightcurve. We summarize our results in Sect. 5.
© European Southern Observatory (ESO) 1999
Online publication: February 23, 1999