Thanks to their high X-ray luminosities, close binary systems were among the first stellar X-ray sources observed, and among the best studied to date. The RS CVn systems are detached close binaries with two late type stars, one of spectral type late F or G V/IV, the other of spectral type around K0 IV/III. Due to the tidal coupling of the rotational and orbital periods, all features of stellar activity that depend on rotation period are enhanced on these systems. The structure of the corona and upper transition region in RS CVn stars has been derived by fits to low resolution X-ray and EUV spectra. Swank et al. (1981) found that the Einstein Solid State Spectrometer (SSS) spectra could be modelled well using a two-temperature optically thin plasma. Dempsey et al. (1993) came to the same conclusion by using ROSAT PSPC data. Griffiths & Jordan (1998) showed that the two-temperature structure of coronal sources can also be inferred by an analysis of the combined EUVE and X-ray data sets, thereby casting doubt on the Majer et al. (1986) conclusion that two-temperature fits to X-ray data result from a combination of the coronal emissivities and the instrument response functions.
One of the most interesting results of recent EUV and X-ray missions (primarily EUVE and ASCA) is that some, but not all, of the most active stars appear to have coronae that are surprisingly metal poor. In fact, X-ray and EUV spectra of RS CVn and Algol binaries and of single active stars reveal metal line strengths much weaker than expected from a plasma with identical temperature/density structure but solar photospheric composition. The weak metal lines indicate a metal deficiency with respect to the solar photosphere by factors of 3 to as much as 10. Some of the sources with low coronal metal abundances also have low photospheric metal abundances (Randich et al. 1994). However, this is not a general rule, as demonstrated, e.g., by AB Dor, that is a ZAMS star with solar photospheric abundances and coronal metallicity Z0.3 (Mewe et al. 1996). Looking at the Sun, a difference between photospheric and coronal abundances could be expected, but in the solar corona the elements with a low first ionization potential (FIP 10 eV) like Fe, Mg, and Si are overabundant. Güdel et al. (1999) observed on the RS CVn binary UX Ari a solar like FIP effect in the flaring coronal plasma, while the quiescent corona resulted metal poor.
AR Lacertae (HD 210334, V=6.1, d=42 pc) consists of a G2 IV primary and a K0 IV secondary star, with mass ratio (Marino et al. 1998), that are spin-orbit coupled with a 1.98 days period. The K star seems to be metal-deficient, but this is not the case for the G star (Naftilan & Drake 1977). The photospheric and chromospheric activity of AR Lac have been monitored for a long time by optical photometric observations at Catania Observatory (Lanza et al. 1998) and by UV Spectral Imaging (Pagano et al. 1995). AR Lac is the brightest known totally eclipsing RS CVn binary. For this reason it has been one of the best observed coronal sources since the late '70s, when it was observed with the Einstein Solid State Spectrometer by Swank & White (1980). At that time its light curve in the 0.5-4 keV band did not show any evidence of eclipses. In 1980 June the Einstein IPC (0.1-4 keV) observed AR Lac during the eclipses and at the quadratures, covering 17% of the orbit (Walter et al. 1983). This X-ray light curve shows a prominent primary eclipse, as well as a shallow secondary eclipse. The rapid egress from primary minimum indicates both the small coronal scale height and the concentration of the X-ray emission toward the leading hemisphere of the G star. The shallow secondary minimum suggests an extended component around the K star. The light curve exhibits considerable structure, at the 10-20% level, outside of eclipse. In 1984 July the LEIT (0.005-2.0 keV) and ME (1.0-6.0 keV) proportional counters on EXOSAT observed a complete orbital cycle of AR Lac (White et al. 1990). Both primary and secondary eclipses were observed at low energies, but at high energies no sign of rotational modulation was found. White et al. (1990) suggested that the low temperature plasma (few times 106 K) was localized very close to the star in compact regions, whereas the high temperature plasma (few times 107 K) permeates the enviroment around the binary system and between the stars. Ottmann et al. (1993) reported on the ROSAT PSPC observations obtained in 1990 June, during the ROSAT calibration phase, which show that both the primary and the secondary eclipses were detected and that the data are compatible with a compact region ( 0.03 ) close to the G star, with solar flare-like emission, plus an extended region ( 1-2 ) linked to the K star, with emission resembling the solar active region emission. In 1993 June AR Lac was observed by ASCA for one complete orbit. White et al. (1994) analyzed these observations and found a 50% reduction in flux centered on the primary eclipse that was independent of energy in the 0.4-7 keV band, and a shallow minimum during the secondary eclipse. At energies 2 keV the ASCA light curve showed continuous low level flaring activity on a time scale of 20-60 min. White et al. (1994) found that the time-averaged X-ray spectrum in the 0.4-10 keV spectral region could be fitted by a two temperatures plasma model with metal abundances lower than solar by factors of 2-4. A similar result was found by the same authors when they analyzed a ROSAT PSPC spectrum obtained simultaneously with ASCA observations. A more detailed analysis of these ASCA/ROSAT observations was done by Kaastra et al. (1996), who confirm the results of White et al. (1994) at least qualitatively. Moreover, Siarkowski et al. (1996) used the same 1993 ASCA light curve to map the spatial structure of the AR Lac's coronae. He found that a) both stars are active, b) the X-ray emission is concentrated on the sides of the stars facing each other, c) there are both compact and extended coronal structures, and d) about 50% of the X-ray emission in unmodulated and could come from an extended halo region, from the poles of the larger K star, or from other symmetric or uneclipsed structure in the orbital plane.
As discussed by Siarkowski et al. (1996), the slow egress and long duration of the X-ray eclipses (e.g., see the SIS and GIS ASCA light curves in White et al. 1994) is one of the main clues that the coronae around both components are spatially extended. Another argument, albeit indirect, in favor of an extend corona is given by the spectral imaging analysis of the Mg II k line (Walter et al. 1987, Neff et al. 1989, Pagano 1994, Pagano et al. 1994, 1995), that revealed the existence of bright localized chromospheric regions on the K star at heights ranging from 0.3 to 1 R* above the photosphere, suggesting the presence of extended chromospheric regions. Walter (1996) interpreted the slow egress from primary eclipse in the EUVE DS photometer (70-190 Å) light curve as due to optically thick obscuring material confined within 15o of equator of the K star and extending outward by RK. Such cool, dense prominence material could also be responsible for the asymetric large eclipses previously seen in the EXOSAT LE (White et al. 1990) and the ASCA light curves (White et al. 1994).
The physical interpretation of the X-ray data arising from coronal structures is greatly enhanced by simultaneous or contemporaneous data obtained at other wavelengths, to probe other aspects of the stellar atmospheres (Rodonò 1982, Linsky 1988). We had arranged for contemporaneous observations to support our SAX observations at longer wavelengths: i) optical photometry and high dispersion spectroscopy at Catania Observatory; and ii) VLA and VLBA radio data from Oct 31 to Nov 4. In this paper we report on X-ray observations. The results of the optical and radio observations, as well as a complete description of the physical scenario that arises from the multiwavelength study, will be given in subsequent papers.
© European Southern Observatory (ESO) 1999
Online publication: June 17, 1999