Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

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

Previous Section Next Section Title Page Table of Contents

4. Light curves

The irregular variations observed in the lightcurve during the dips are generally thought to be due to the cloudy structure of an absorber extending above the disk surface. The location of the material, i.e., its distance from the neutron star, however, is unknown and depends on model assumptions. According to Crosa & Boynton 1980, the obscuring matter is the temporarily thickened disk rim, or, in the modification of this model by Bochkarev 1989 and Bochkarev & Karitskaya 1989, consists of "blobs" in an extended corona above the disk rim. On the other hand, in the coronal wind model of Schandl et al. 1997 the dips are caused by a spray of matter at some inner disk radius where the accretion stream impacts on the disk.

Since all models assume that the absorbing matter exists in the form of clumps of material, it is interesting to ask whether these blobs can be seen in the data. The symmetry of the structures in the first half of the lightcurve shown in Fig. 4a suggests that this is indeed the case. We therefore tried to model this part of the dip lightcurve by fitting Gaussians to each of the visibly identified structures. The best fit to the data is displayed in Fig. 8. Information about the parameters of each Gaussian is summarized in Table 2 (available in electronic form only). Note that the structures appear to be quasi-periodic on a time-scale of about [FORMULA] d (=90 s) which could provide a hint on their nature. A similar 144 s periodicity has previously been reported by Leahy et al. 1992. A study of a larger sample of dips is necessary, however, before any claim on the existence of periodic structures in the dip data can be settled.

[FIGURE] Fig. 8a-c. Gaussian fits to the dip lightcurve with a temporal resolution of 2 s. Arrows mark the position of Gaussian line centroids.

Assuming that each structure in the lightcurve represents a single cloud of matter which is absorbing X-rays while crossing the line of sight, the width of the Gaussian represents a direct measure of the crossing time for such a cloud. The horizontal extent of the cloud could be derived if the radial distance of these clouds from the neutron star and their velocity were known. To give a rough estimate for the cloud size, we assume that the matter moves on Keplerian orbits close to the outer disk rim (as is favored by the model of Crosa & Boynton 1980), at a distance of [FORMULA]cm from the neutron star (Cheng et al. 1995). The Keplerian velocity at this radius is approximately [FORMULA] cm s-1. Assuming that the clouds are spherical, i.e., assuming that their angular size estimated from the FWHM of the fitted Gaussian corresponds to their radial size, we derive typical cloud diameters on the order of [FORMULA] cm. Making use of the observed column densities we find a proton density of about [FORMULA]. On the other hand, the apparent periodicity mentioned above might also indicate that the blobs are very close to the neutron star. Assuming Keplerian motion, the periodicity could indicate orbital radii as small as [FORMULA] cm (the inner radius of the accretion disk is at [FORMULA] cm; Horn 1992) and their densities would be accordingly higher.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: February 23, 1999