The primary aim of this study was to search for evidence of regions of extended material around the ER Vul system. Northcott & Bakos (1967) were the first to suggest the presence of a gaseous cloud at the inner Lagrangian point extending to one side. McLean (1982) also suggested the presence of circumstellar material near the primary component in order to account for variations in the primary component spectral lines. Arevalo, Lazaro & Fuensalida (1988) also proposed that a high-temperature gas stream exists between the components based on their photometric variations; this would also account for their suggested IR and UV excesses. Apart from the detection of a very weak absorption feature approximately 200 km s-1 blue-ward of the secondary component the present study has not detected spectroscopically the existence of such extended regions. This confirms the results of Newmark (1990) who also found no spectroscopic signatures of extended material and concluded that the phase variations of the line strengths and widths were incompatible with an origin in extended structures. It is unclear whether the large filling factor derived for this system can account for the variations seen by McLean (1982); even if it does the photometric results would then also show a disparity. If extended material is present then it must be sufficiently distant from the star concerned that it is neither eclipsed nor affects photometric measurements with temporal variations. It is also possible that the appearance of circumstellar material is a transient phenomenon similar to solar prominences but such a conjecture is not testable in the present study. For ER Vul the photometric and spectroscopic inferences must therefore remain in disagreement.
The present observations have clearly detected excess emission in the activity sensitive lines of H , the Ca II IRT, and Mg I b as well as absorption in He I D3. Newmark (1990) also reported excess emission in H and the Ca II IRT for ER Vul. He concluded that the emission variations were not correlated with phase; their stochastic nature suggested a global origin. This is also confirmed in this study by deriving the velocities of the excess emission peaks. An important result is that the present observations clearly show that the secondary component of ER Vul is the more active. Newmark (1990) found that the opposite was true. This cannot be accounted for by the eclipse nature of these observations since the maximum obscuration of the primary disk is only 12%; insufficient to affect the factor of approximately two for the emission from the primary over that of the secondary in both H and the Ca II IRT found by Newmark (1990). Hence the activity levels of the components of ER Vul must be highly variable. This study has shown that during these observations the secondary component of ER Vul was more active than the primary in H and slightly more active in the Ca II IRT.
ER Vul consists of two dwarf stars unlike many other RS CVn systems which are dominated by the emission from a giant or sub-giant primary. Dwarf stars have higher surface gravities and electron densities so their emission regions may be significantly different than in other RS CVn's. In fact Buzasi (1989) has shown that the chromospheric activity in main-sequence stars should be mainly from optically thick plage-like regions where the ratio of excess emission in the H and H lines ( / ) can be as low as 2.0. Conversely evolved stars should have higher ratios similar to those found in solar prominences. Chester (1991) recently confirmed the low ratios associated with solar plage regions with high-resolution spectroscopic observations. Newmark (1990) found a / ratio for ER Vul of 1.09 and suggested that the majority of the emission was associated with plage regions as expected for dwarf stars. This hypothesis has been confirmed in the present study mainly by the detection of significant He I D3 absorption which is known to be only visible in plages for the solar case. In addition analysis of this line indicates a very large filling factor of 0.53 implying the chromosphere is highly active over the entire surface. The detection of a constant velocity offset for the emitting material is similar to that seen in the solar plages. Since the filling factor is very large for ER Vul the disk-integrated spectra may show such an offset although for the disk-integrated solar spectrum (where the filling factor is only a few percent) this is not seen. Thus it is possible these results represent the first observation of the equivalent of solar spicule emission in a stellar system other than the Sun.
Simple calculations have revealed that the chromospheric thickness for the ER Vul secondary may be very large ( 0.42 ). The excess emission in the H , Ca II IRT and Mg I b lines is indicative of a significant amount of non-radiative energy dissipation at all levels of the chromosphere although the lack of detection of Na I D emission is confusing.
X-ray and UV studies of RS CVn systems suggest that there is a level at which the emission becomes saturated (Vilhu & Rucinski 1983), although alternative explanations exist (Doyle 1996). If such saturation is related to very large filling factors then for these systems little phase variation of the emission should be seen. The very high levels of X-ray (Walter & Bowyer 1981), UV (Rucinski & Vilhu 1983) and radio emission (Drake, Simon & Linsky 1986) for ER Vul suggest it is a system where continuous high levels of activity (e.g. micro-flaring) may be taking place. The optical observations of Newmark (1990) showed no variability in line fluxes which supports the hypothesis that ER Vul is near the saturation limit for chromospheric activity. The present observation of a large filling factor for ER Vul adds credence to this proposition.
© European Southern Observatory (ESO) 1997
Online publication: July 8, 1998