Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 338, 465-478 (1998)

Previous Section Next Section Title Page Table of Contents

6. X-ray light curves and spectra

The X-ray light curves extracted from both pointed observations are shown in Fig. 16 (upper panels). The original photon event tables were binned into 30 seconds intervals and the average countrate in these time intervals are plotted as a function of the photometric phase. Similarly, the average countrates in individual scans during the RASS were folded over the orbital period and are also shown in Fig. 16. At all occasions the X-ray flux was modulated by 100% and displayed a distinctly different shape compared to the optical. Bright flares (CR [FORMULA] 0.5-1.5 s-1) are superimposed on a fainter component which seems continuously to be present at a low countrate of [FORMULA]0.01 s-1.

[FIGURE] Fig. 16. X-ray light curves of RX J0203 obtained with the ROSAT PSPC and HRI detectors, respectively. Shown are from top to bottom the phase-folded PSPC data obtained during the RASS, and pointed HRI and PSPC observations obtained at dates indicated in the panels. In the lowest panel the hardness ratio HR1 derived using the pointed PSPC-data is shown as a function of the photometric phase. All data points are plotted twice for clarity.

This distinction is supported by the spectral variation indicated by the change of the hardness ratio HR1 (Fig. 16 lower panel). Bright flares are rather soft (HR1[FORMULA]) compared to the fainter parts of the orbit which are in general harder (HR1[FORMULA]...0) probably due to a reduction of the soft component. Some (probably hard) X-ray emission seems to be persistently present throughout the orbital cycle, apart from a short period at phase [FORMULA] in the Aug95 HRI-observation with almost zero count rate. The lack of a pronounced faint phase is supporting the suggestion from the optical light curves that the main accretion spot in RX J0203 undergoes no selfeclipses. The fact that a bright interval is observed during the HRI pointing at a phase ([FORMULA]) when it was definitively not bright during the PSPC pointing, is an additional hint that the bright flares occur randomly in phase.

However, RX J0203 is not a typical representative of this class. The X-ray light curves of systems which are thought to display a similar accretion geometry (e.g. EF Eri or MR Ser; Beuermann et al. 1987, Schwope et al. 1991) are modulated with the rotational period due to projection. The flare-dominated shape of the light curves of RX J0203 does not fit in that picture and interpretation is difficult with the incomplete phase coverage (55%/45% with the PSPC/HRI) achieved. At date we are left with the conclusion that RX J0203 is probably a close relative to V 1309 Ori where also strong flaring dominates the X-ray light curves (Walter et al. 1995). In these systems the X-ray emission seems to be mainly driven by highly instationary accretion of overdense blobs, and energy released as soft blackbody radiation (as suggested by our spectral analysis) due to reprocession hard X-rays released in buried shocks. It would be interesting to investigate if this property is intrinsic to all long period (and thus high [FORMULA]) AM Herculis objects.

Given the spectral variability of the source we extracted separate spectra for phase intervals showing typically soft ([FORMULA]) and hard spectra ([FORMULA]) and modelled these with an absorbed blackbody plus thermal bremsstrahlung spectrum. Since the spectral range and resolution of the PSPC is not sufficient to constrain either of these temperatures precisely we fixed [FORMULA] at 20 keV and computed a [FORMULA]-grid for [FORMULA] and [FORMULA]. This yields limits for the bright phase spectrum of [FORMULA]=42-48(28-52) eV and [FORMULA]=2-3(1.5-4)[FORMULA]cm-2 for the 68% (95%) confidence levels. The resulting unabsorbed, bolometric luminosities corresponding to the best fit ([FORMULA] cm-2, [FORMULA]=44 eV, [FORMULA]=0.99) are [FORMULA]erg/s and [FORMULA]erg/s assuming geometrical factors of [FORMULA] and [FORMULA] for the blackbody and bremsstrahlung emission, respectively.

In order to estimate the energy balance between the blackbody and bremsstrahlung components we followed Beuermann & Schwope (1994) who used the flux ratio in the ROSAT-band (0.1-2.4 keV) with fixed values of [FORMULA] eV and [FORMULA] keV for a comparative study of AM Her objects. We find [FORMULA] for RX J0203 , which fits together with the observed field strength nicely into the [FORMULA]-B dependence found by the above authors.

The count statistic of the hard phase is insufficient for a qualitative spectral estimate and we repeated the modelling with fixed best-fit parameters for the temperatures and the column density as determined above from the soft phase spectrum. The derived fluxes are then 1/40[FORMULA] and 1/2[FORMULA] of the bright phase flux showing that the variation of the hardness ratio is predominantly due to the reduction of the blackbody component.

We compared the mean-countrates of the PSPC and HRI pointing taking into account the different effective areas of both detectors. The conversion factors (PSPC/HRI) for the blackbody and bremsstrahlung component are [FORMULA]6.5 and [FORMULA]2.7, which implies that the X-ray flux was a factor of two higher at the time of the HRI observations. While this might indicate a variation of the mass accretion rate, also a small change of the blackbody temperature (shifting parts of the spectrum out of or into the ROSAT range) may account for what is observed.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1998

Online publication: September 14, 1998