Online material

Appendix A: The study of spectral variability on short timescales

Given the rather low S/N of the HARPS spectra, to study the spectral variability on short timescales, we employed the LSD technique, allowing us to achieve much higher S/N in the LSD spectra. The details of this technique can be found in the work of Donati et al. (1997). The line masks are constructed using the VALD database (e.g. Kupka et al. 2000). As mentioned in Sect. 3, the reconstruction of the intensity profile (Stokes I) does not need the generally time-consuming application of the SVD technique.

Appendix A.1: HD 101412

Fig. A.1 Top panel: comparison of the LSD Stokes I profiles of HD 101412 computed for individual subexposures recorded over six observing nights in 2014 February using the line mask of 1007 metallic lines. Eight individual subexposures were obtained during the first night, while four subexposures were obtained for each following night. Bottom panel: differences between Stokes I profiles computed for the individual subexposures and the average Stokes I profile. The dashed lines indicate the standard deviation limits.

Open with DEXTER

Seven spectropolarimetric observations were obtained with the HARPS polarimeter on the nights from 2013 February 14 to 21. Among them, two individual observations have been obtained during the first night on February 14. Each observation was split into four subexposures with an exposure time of 10–12 min, obtained with different orientations of the quarter-wave retarder plate relative to the polarization beam splitter of the circular polarimeter. In Fig. A.1, we present the comparison between the LSD Stokes I profiles computed for each individual subexposure recorded on the six different nights. The results for both observations obtained on the first night on February 14, each consisting of four subexposures, are presented in the same panel. No significant variation in the line profile or radial velocity is detected in the behaviour of the Stokes I profiles for each subexposure. Signal-to-noise values achieved in the spectra observed during the last three observing nights are in the range from 76 to 58, i.e. significantly lower than those achieved in the first three epochs. For these last three nights, we observe very tiny changes in the cores and wings of the overplotted LSD Stokes I profiles with an intensity variation of the order of the spectral noise of about 0.3%.

Appendix A.2: HD 104237

Fig. A.2 Upper row: comparison of the LSD Stokes I profiles of HD 104237 computed for the individual subexposures obtained with a time lapse of 2.2 min (left side) and differences between Stokes I profiles computed for the individual subexposures and the average Stokes I profile with standard deviation limits indicated by the dashed lines (right side). Lower row: Stokes V spectra calculated for the combination of pairs of two subexposures with the quarter-wave plate angles separated by 90 degrees, four pairs in all.

Open with DEXTER

The primary is a δ Scuti-like pulsator with frequencies ranging between 28.5 and 35.6 d^-1 (Böhm et al. 2004). The observation on 2010 May 3 was split into eight subexposures with an exposure time of 2.2 min. The line mask included 1007 metallic lines. In Fig. A.2 in the upper row, we present the impact of

pulsations on the Stokes I profiles computed for each individual subexposure. Because of the pulsation changes in the line profiles during the observation, the final Stokes V spectrum is expected to lead to a wrong value of the longitudinal magnetic field. To prove whether the primary possesses a field, we used pairs of two subexposures with the quarter-wave plate angles separated by 90 degrees, four pairs in all. However, no conclusions about the presence of a magnetic field in this component can be drawn due to the very low S/N achieved. The computed Stokes V spectra for each pair of subexposures are presented in the lower row of Fig. A.2.

Appendix A.3: HD 190073

	Fig. A.3 Comparison of the LSD Stokes I profiles of HD 190073 computed for the individual subexposures obtained on 2011 May 24 with a time lapse of about 10 min (left panel) and on 2012 August 7 with a time lapse of 30 min (right panel). The dashed lines on the lower plots indicate the standard deviation limits.
Open with DEXTER

This star was not reported in the literature to show δ Scuti-like pulsations. The line mask in our analysis included 375 metallic lines. The inspection of the behaviour of the LSD Stokes I profiles presented in Fig. A.3 calculated for each subexposure does not reveal any line profile or radial velocity variation on a timescale of about ten minutes during the first epoch in 2011 or on a timescale of 30 minutes during the second epoch in 2012.

Appendix B: The SVD profiles

Fig. B.1 Correspondence of SVD I and V profiles of HD 101412 obtained using different line masks. From left to right the results correspond to the following line samples: the sample of 650 lines belonging to various iron-peak elements assuming Teff = 8300 K, the sample of 339 Fe i lines (Teff = 8300 K), the sample of 29 Fe i lines assuming Teff = 10 000 K, and the sample of 52 Fe ii lines (Teff = 8300 K). The time runs from bottom to top.

Open with DEXTER

© ESO, 2015