Astron. Astrophys. 362, 628-634 (2000)

3. Comparison with the radio corona of Algol

VLBI radio techniques have sufficient resolution to allow direct "imaging" of the radio-bright component of the Algol corona, although the uncertainties in the relative astrometry do not allow for accurate positioning of the photosphere of the two stars onto the radio map. Mutel et al. (1998) have recently studied Algol with a long time series (along more than 1 yr) of VLBA observations, finding frequent large flares superimposed on a relatively constant quiescent emission level. The map of the radio corona shows the presence of two lobes, aligned with the poles of the system, and plausibly positioned on the poles of the active K star in Algol. The regions from which the quiescent emission originates are resolved in the radio images, and are larger than the star itself. No evidence for orbital modulation of the (quiescent) radio emission is seen, thus implying that the emitting plasma is largely located outside of the orbital plane of the system. The flaring emission comes from the same region as the quiescent one, but is not spatially resolved in the VLBA observations (i.e. the emitting regions are smaller than the radio beam), and is comparable to or smaller in size than the star. Mutel et al. (1998) thus model the flaring regions as co-located with the ones responsible for the quiescent emission, but significantly more compact (see their Fig. 13). No radio emission is identified from regions around the star's equator, nor from the inter-binary region. The radio emission coming from the two lobes is polarized, with opposite signs, implying the presence of magnetic fields of opposite polarity, which Mutel et al. (1998) interpret as possibly coming from a large-scale bipolar magnetic field.

How does the X-ray corona compare with the radio one? The evidence from the existing X-ray observations of Algol is that the level of orbital modulation of the quiescent X-ray emission is small, if not, at times, negligible: the ROSAT and ASCA observations show a decrease in the X-ray emission at secondary minimum (Ottmann 1994; Antunes et al. 1994), while no orbital modulation of the quiescent emission is seen in the EXOSAT observation (White et al. 1986). At the same time, the BeppoSAX flare undergoes a total eclipse, while the EXOSAT, GINGA and PSPC flares (the ones analyzed in the present paper) do not show any evidence of eclipses. The very shallow orbital modulation of the quiescent emission has been interpreted (e.g. White et al. 1986) as implying that the X-ray emitting region is so extended that the companion star can only eclipse a small fraction of it at any given time (thus as large or larger than the inter-binary separation). However, an alternative explanation (which Mutel et al. 1998 use also for the radio emission) is that the coronal plasma is confined largely to regions above the polar caps of the active star. The emitting region is thus not eclipsed (except for a small volume close to the K star's polar surface) in a high-inclination binary such as Algol. The seat of the quiescent emission must thus be more extended than the region being eclipsed by the primary during secondary eclipse, but not much more. This implies that it must extend above , perhaps to a size comparable to the size of the region postulated by Mutel et al. (1998) for the quiescent radio corona. The BeppoSAX flare has been shown to be located on the south pole of the Algol K star, and to be more compact than - thus fully compatible with the type of region postulated by Mutel et al. (1998) for the flaring radio corona. The other three X-ray flares have not been eclipsed and thus their location on the star cannot be established; however their size as derived here is also compact and thus compatible with the same picture.

The X-ray and radio data can thus be explained within the same scenario: the corona is essentially polar and located on the K star in Algol (with no influence from the B star except for the tidally induced rotation), with a compact flaring component (smaller than the star) and a more extended quiescent corona, as large or somewhat larger than the star itself. The latter shows little if any rotational modulation as the B star can only occult a small fraction of the southern lobe.

© European Southern Observatory (ESO) 2000

Online publication: October 24, 2000
helpdesk@link.springer.de