SpringerLink
Forum Springer Astron. Astrophys.
Forum Whats New Search Orders


Astron. Astrophys. 341, 527-538 (1999)

Previous Section Next Section Title Page Table of Contents

4. Chromospheric lines

The computed radial velocity was used to correct all the spectra so that they were finally referred to the rest velocity of the photosphere. An overall mean spectrum in the region of the H[FORMULA], H[FORMULA] and He I [FORMULA] chromospheric lines, based on both nights' data, is shown in Fig. 2.

[FIGURE] Fig. 2. The overall mean spectrum of BD+[FORMULA] based on the entire two nights' data in the vicinity of H[FORMULA] (upper panel ), H[FORMULA] (middle panel ) and the He I D3 doublet (lower panel ). Note that the spectra were shifted to the rest wavelength of the star's photospheric spectrum before averaging, as described in the text. The vertical lines indicates the rest wavelength of each line while the horizontal dotted lines indicate the normalization level. Note the asymmetry of the emission in all three lines towards the blue

Both Balmer lines are strongly in emission, as would be expected for an active late-type star, while He I D3 is also in emission but weak. An interesting result is the apparent asymmetry of the emission features, in the sense that emission is mainly seen to the blue side of the rest wavelength. In view of previous evidence of H[FORMULA] variations in similar rapidly rotating late-type active stars we now examine evidence for the same kind of related variations in BD+[FORMULA].

4.1. H[FORMULA] variability

Even casual examination of sequences of spectra clearly shows that the H[FORMULA] line is variable. In order to quantify this we have proceeded initially as in BER96. Individual normalized line profiles were fitted with a single gaussian, so that the variability could be represented by means of the temporal evolution of three main parameters of the fits, i.e. the wavelength of the line centre, [FORMULA], the equivalent width, EW, and the full width at half maximum, FWHM. The results are shown in Fig. 3. Vertical dotted lines divide the plots in sections to distinguish the main changes that take place during each night, generally defined by an alteration of the slope of the EW evolution. Note that, following the usual (but not universal) convention, negative EW indicates that the line is in emission but in the text we will refer only to the absolute values. The RV parameter plotted in Fig. 3 is the Doppler velocity corresponding to the wavelength shift of the H[FORMULA] line centre from its wavelength in the rest frame of the star.

[FIGURE] Fig. 3. Evolution of the gaussian parameters derived from fits to the H[FORMULA] line profile during the night of August 4th (left ) and the night of August 5th (right ), as explained in Sect. 4.1. Phases have been calculated by defining the time of the first exposure on the first night as zero and using the rotational period, P=10.17hr (J94). The dotted vertical lines divide the plots into sections according to significant changes in the slope of the EW. Each interval of phases is labeled to help their identification in the text. The small arrows indicate the times at which He I [FORMULA] emission was detected

On August 4 the EW exhibits rapid variations during intervals A-F, with peak-to-peak amplitudes of the order [FORMULA]0.15 Å ([FORMULA]20%) on time scales corresponding to phase intervals, [FORMULA] 0.05-0.1. However, after [FORMULA] 0.45 the [FORMULA] reaches a plateau [FORMULA]0.84 Å. The FWHM appears to vary in a similar way although individual changes are less marked.

The temporal evolution of RV (Fig. 3, left ) is related to that of EW and FWHM in the sense that the main changes occur almost simultaneously. The detailed relationship between RV and the other two parameters is less readily discerned. For instance, whereas the evolution towards a minimum velocity at [FORMULA] = 0.2 corresponds to an increase of both EW and FWHM, the next blueshift, at [FORMULA]0.37, is accompanied by a decline in the EW and FWHM.

On August 5 the EW variability is more extreme than in the previous night and is characterized by two consecutive peaks. The first peak (interval B) is the highest ([FORMULA]0.92 Å) and lasts for [FORMULA]0.31. The second peak (intervals D-E), with [FORMULA] Å, extends over a shorter interval of phase, [FORMULA]0.18. As on the first night, FWHM generally correlates well with EW.

The RV has a sinusoidal appearance with total amplitude [FORMULA]25 km s-1. A marked positive deviation from this otherwise smooth trend takes place at B, coinciding with the peak in both EW and FWHM. There is another slight deviation at D but in a blueward direction.

We note that, during those phases where the two nights overlap, there is reasonable but not exact agreement between the behaviour of the gaussian parameters.

4.2. H[FORMULA] variability

Since H[FORMULA] is an intrinsically much weaker line, the signal-to-noise of the H[FORMULA] profiles is much less than for H[FORMULA]. In order to aid comparison with H[FORMULA] we have applied the same analysis to the H[FORMULA] profiles. The results are shown in Fig. 4. The intervals of phase used in Sect. 4.1 are retained here. Results are very scattered compared to H[FORMULA], making a detailed comparison very difficult.

[FIGURE] Fig. 4. Temporal evolution of the main gaussian parameters derived from the fits to the H[FORMULA] line profile during the night of August 4th (left ) and the night of August 5th (right ). Details of the plots are as in Fig. 3

Some differences in the H[FORMULA] behaviour on August 4 can be seen in Fig. 4 (left ). The peak of [FORMULA] in B does not exist in H[FORMULA] and only the peak at interval D is seen clearly in the two lines. In addition, [FORMULA] is almost constant as observed in H[FORMULA]. On the other hand, the general trend in RV(H[FORMULA]) is consistent with RV(H[FORMULA]), although absolute values are different.

As in H[FORMULA], the most important features of EW and FWHM variability on August 5 (Fig. 4, right ) are the two large increases seen at intervals A and D, respectively. The peak in interval D is in fact better distinguished in H[FORMULA]. Another sharp increase occurs in F, which is less marked in H[FORMULA]. On the other hand, G is an interval of relative constancy in both lines.

On this night, the H[FORMULA] line centre varies by a relatively larger amount and more symmetrically with respect to zero than H[FORMULA]. This is especially marked in [FORMULA]. The total amplitude of the variations is [FORMULA]30 km s-1. Deviations from the general trend are again seen in intervals B and perhaps D. This reproduces the behaviour observed in H[FORMULA] in a general way but there are important differences of detail. We note at this point that the lack of detailed agreement in RV between H[FORMULA] and H[FORMULA] supports an interpretation in terms of absorption rather than chromospheric mass motions as discussed later (see Sect. 8.1).

As seen in H[FORMULA] the general trends are consistent between the two nights with differences only in detail especially in the RV curves near [FORMULA]0.5.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: December 4, 1998
helpdesk.link@springer.de