Astron. Astrophys. 337, 757-771 (1998)

5. Variability of chromospheric emission

5.1. The H emission line

The H line was seen to vary rapidly with significant changes in both equivalent width and in line profile. The most pronounced variations occurred during the last night of observations. In Fig. 1 are shown the nightly mean profiles. The parameters of the gaussian functions that best fit these profiles are given in Table 2. Emission appears to be blueshifted by about -6 km s^-1 on the three nights. This, in principle, may indicate an overall rise of the chromosphere. However, as it will be seen later, there are numerous absorption and emission events that can also be contributing to the observed profile shape. For this reason, the symmetry/asymmetry of the mean profiles will not be discussed further. The mean level of emission is clearly lower on June 25 and considerably higher on June 27.

Fig. 1. The nightly mean H profiles for the nights of August 25 (- -), August 26 (...) and August 27 (-.-). Also shown is the overall mean profile (-) that was used in the analysis of the emission variability. The vertical line indicates the rest wavelength of H

Table 2. Parameters of the gaussian fits to the nightly mean H profiles

Proceeding as in Byrne, Eibe & Rolleston (1996), a search for systematic behaviour of the H line was first done by looking at the ratios between all individual spectra and a mean reference spectrum. In this case the reference is an overall mean, made of a total of 242 spectra that were not affected by significant emission events. This profile is also shown in Fig. 1 (continuous line) for comparison. Its FWHM, 2.33Å, agrees well with the value predicted for rotational broadening, 2.31Å, suggesting that doppler broadening dominates over other sources of spectral line broadening.

Dynamic H ratio spectra for each of the observing nights are shown in Figs. 2 - 4. Each line of a grey-scale image correspond to the ratio between an individual profile and the overall mean. Spectra are all displayed in a velocity scale relative to the stellar rest frame set at -23.76 km s^-1. This technique allows to rapidly obtain information about the characteristic features that are determining the shape of the line profile as well as the specific velocities at which largest variability takes place.

Fig. 2. Dynamic spectra of H on June 25. The grey-scale image is made of ratio profiles that result after dividing by the overall average. Rotational phases are indicated in the left edge and the velocity scale in the horizontal axis is referred to the stellar rest frame, set at -23.76 km s-1.

Fig. 3. As Fig. 2 but for the night of June 26.

Fig. 4. As Fig. 2 but for the night of June 27.

Visual inspection of the dynamic H ratio spectra indicates that in general the pattern of the line variability is different from night to night. Closer examination, however, reveals some common features. A narrow absorption transient is neatly defined in Fig. 2 between phases 0.2-0.3. This event appears to obscure completely the prominent emission feature that is seen to propagate from blue to red in the line profile between 0-0.5. Although they are less obvious, these two features recur on the following nights (see Figs. 3 and 4). Spectra taken on the night June 26 are of poorer quality, making detailed analysis of the profile more difficult. Unfortunately the phase coverage of the night June 27 was such that the early stages of those events were already missed when the series of observations began.

Signs of another absorption transient are detected at phases near 0.9. Its existence is, however, more difficult to trace. On the night June 25 it appears to occur in connection with an additional redshifted component, manifested in Fig. 2 by the intense absorption seen in the red during phases 0.9-0.1. On the night June 27, observations finished at phase 0.96 so that only the propagation of the absorption in the blue was recorded. Its presence was impossible to confirm on the night June 26 due to differences in phase coverage.

Evidence of a deep blueshifted absorption is also found at the end of the night June 25, emerging in the profile at 0.4. Absorption events at these phases are also detected on the following nights but they exhibit a generally different behaviour.

An extraordinary event occurred on the night June 27. This is the most noticeable feature in Fig. 4. It starts as obvious blue-shifted emission (0.6) that propagates rapidly toward the red. A slow and gradual decay phase follows until extra-emission fades completely at 0.9.

A more detailed study of these events will be deferred for a later section.

5.2. Other spectral lines

The dynamic spectra for the other interesting chromospheric lines which were observed have been constructed in the same way that in the case of H and are shown in Figs. 5 - 8. These are used here mainly for comparison and more thorough description will follow in future work. In the case of the He I D₃ and Na I D-lines the images are made of individual spectra instead of ratios in order to improve contrast. The captions of the figures are as follows:

Fig. 5. Dynamic ratio spectrum of Ca II 8498Å on June 25 (left) and June 27 (right).

Fig. 6. As Fig. 5 for the Ca II 8542Å line.

Fig. 7. Dynamic spectrum of He I D3. Individual spectra are shown instead of ratios in order to improve contrast.

Fig. 8. As Fig. 7 for the Na I D-lines.

© European Southern Observatory (ESO) 1998

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