Astron. Astrophys. 330, 1029-1040 (1998)

5. H line-profile variations

All spectra from February-March 1995 show H as a relatively normal absorption line with an average equivalent width of 0.9 Å , but plotting its equivalent width against orbital or rotational phase reveals a nearly sinusoidal variation very similar in shape and phase to the contemporaneous V -light curve (Fig. 10). Relative to a pseudo continuum the full amplitude of the variation reaches 600 50 mÅ or approximately 100%. We take this as evidence that at least part of the H flux must be linked to the fraction of the star's surface that is covered with cool spots and therefore the plages are presumably on top of these spots in agreement with solar analogy.

Fig. 10. Variation of the H equivalent width as a function of orbital phase for Feb.-March 1995 (filled circles) and a sinusoidal fit of the 1994/95 V -lightcurve (dashed line). Note that minimum absorption, i.e. maximum residual emission, coincides with a time of lightcurve minimum, i.e. when the most spotted hemisphere is viewed. This suggests a spatial relation between spots and plages on IL Hya.

To examine the H flux of the active part of the chromosphere we subtract a broadened and shifted spectrum of the inactive K0.5III star 16 Vir from each IL Hya spectrum. This reference spectrum is assumed to represent the non-magnetic part of the chromosphere of IL Hya but, of course, can not account for eventual differences in the atmospheric structure of the two stars. Fig. 11 shows the resulting nine residual H profiles revealing a strong and variable residual emission component centered at the rest wavelength of H and a small, redshifted absorption feature. The latter feature's redshift, obtained from two-Gaussian fits to the residual profiles, remains constant at approximately Å or km s^-1 (rms) throughout a rotational cycle and thus can not be explained with a plage-like feature on the stellar surface. Using  Gem (K1IIIb) instead of 16 Vir gives practically the same result with marginally different emission strengths but identical wavelength shifts as demonstrated in the upper panel of Fig. 11.

Fig. 11. Residual H line profiles of IL Hya (lower panel) after the subtraction of a spectrum of the inactive reference star 16 Vir (dashed line). The number adjacent to each profile is the orbital phase. Note the permanently redshifted absorption by 1.24 Å 57 5 km s-1 as obtained from a two-Gauss fit. The upper panel shows a representative spectrum of IL Hya (thick line) and the unbroadened spectrum of 16 Vir as well as the difference spectrum of 16 Vir to yet another reference star ( Gem, K1III, shifted by for better visibility).

Similar H -line profiles were already seen in other active stars, most notably in IN Vir with a redshifted absorption of 60 km s^-1 (Paper III) and HU Vir where a red feature appeared shifted by even 100 km s^-1 (Strassmeier (1994), Hatzes (1997)), but also in HD 17433 (Bopp et al. (1989)), HD 12545 (Bopp et al. (1993)) and the single star HD 9746 (Fekel et al. (1986)). The single giant HD 32918 = YY Men shows qualitatively a similar profile, but with blue-shifted emission (by approximately -80 km s^-1) and an absorption feature at the H rest wavelength (Vilhu et al. (1991)). Furthermore, some of the most active stars exhibit very strong, broad and structured emission profiles, e.g. FK Comae with an emission width at the continuum level of 1000 km s^-1 (Oliveira et al., 1997), or the Pleiades star HII 1883 with 700 km s^-1 (Marcy et al., 1985), or IN Comae with 800 km s^-1 (Paper IV) while others, most notably AB Doradus and V471 Tauri, show a periodically variable absorption profile due to prominence activity (e.g. Collier Cameron & Robinson (1989)). To confuse the phenomenology even further, Hatzes (1995) found that the H spectrum of the single-lined RS CVn binary DM UMa (K0III, days) consists of a narrow emission component and a broad absorption component, both unshifted with respect to the rest wavelength, but only the broad component also showed phase-dependent variations.

Due to such a large variety of the observed phenomenology it is likely that several processes, and most likely a combination of them, are the cause for the H profiles of active stars: stellar plages in direct analogy to solar plages (see, e.g., Neff (1996)); fluctuations of both the column density and temperature gradient within the chromosphere as suggested earlier by Smith & Dupree (1988) to explain the H profiles of metal-deficient red giants; local velocity fields and mass motions due to magnetic field inhomogeneities possibly coupled with a loop-like geometry as believed to have been detected on HU Virginis from pseudo 3D Doppler maps (Strassmeier, 1994); and, naturally, classical stellar winds and even antiwinds as indicated from redshifted UV-emission lines of e.g. Capella (Linsky et al., 1995). The approach of phase-resolved H spectra is certainly promising and allows to separate rotational modulation from other time-dependent processes, but more observations of more targets with higher time resolution are clearly necessary to establish a firm explanation of the H profiles of active stars.

© European Southern Observatory (ESO) 1998

Online publication: January 27, 1998
helpdesk.link@springer.de