Astron. Astrophys. 341, 768-783 (1999)

2. The data

2.1. Selection criteria for the T Tauri stars

For solar active regions, Schrijver et al. (1989) found a correlation between the flux of the emission core of the Ca ii K line and the magnetic field strength. Because the emission core of the Ca ii 8542 Å -line is formed in plage regions too, we expect a similar correlation for this line. To select the targets for the search for magnetic fields, we thus give preference to T Tauri stars with strong emission cores of the Ca ii 8542 Å-line. Since very late spectral types and large  values will cause blending of the spectral-lines, we prefer to work with stars of low , and earlier spectral type, excluding M-type stars. Another problem is that the spectra of cTTSs are often veiled by a featureless continuum source (Basri et al. 1990; Hartigan et al. 1989). Obviously, the signal-to-noise ratio requirements are easier to achieve if the star is bright and the strength of this veiling continuum is low.

The selection criteria for our stars are thus:

• km/s

• a large Ca ii emission line component

• spectral type earlier than M0

• relatively bright ()

• low veiling

2.2. Properties of the target stars

Using the criteria mentioned above, we selected UX Tau A, LkCa 16, LkCa 15, GW Ori and T Tauri as targets. Fig. 1 shows the Ca ii 8542 Å-line profiles of some stars of our sample. VY Ari is a RS CVn star with a magnetic field strength of  kG and a filling factor of (Bopp et al. 1989). The emission line core is quite strong. It is interesting to note that the emission core of the Ca ii lines is stronger in the T Tauri star LkCa 16 than in VY Ari. The narrow component of Ca II  8542 Å in LkCa 15 is again quite large. T Tau was selected because of its low veiling. The Ca ii emission in T Tau is just a broad emission line, the narrow Ca ii emission core is not visible.

Fig. 1. Ca ii lines of the magnetic template star VY Ari, the wTTS LkCa 16, and the classical T Tauri stars LkCa 15 and T Tau. The narrow emission component seen in VY Ari originates in plage regions on the surface of the star, and is thus related to the strength of the magnetic field. This component is clearly visible in LkCa 15 and LkCa 16, making these stars suitable targets for the search for magnetic fields. Only a broad Ca ii line is visible in T Tau.

Traditionally, T Tauri stars have been classified into classical and weak-line T Tauri stars on the basis of the EW of the line. While T Tauri stars with Å are called wTTSs, T Tauri stars with Å are called cTTSs (Appenzeller & Mundt 1996). UX Tau A, LkCa 15 and LkCa 16 are listed as weak-line T Tauri stars in the Herbig & Bell catalogue (1988), since the EW of was less than 10 Å when observed. Since the equivalent width in of LkCa 15 was larger than 10 Å when we observed it, we will call it a cTTS. Additionally, there is a general trend that T Tauri stars with weak emission lines do not have disks or have less massive disks, and thus do not accrete matter (Osterloh & Beckwith 1995). Using broad-band infrared photometry and the measurements of the fluxes at 1.3mm Osterloh & Beckwith (1995) have detected disks with masses of 0.024 and 0.0065 for LkCa 15 and UX Tau A, respectively. These values can be compared with the mass of the disk of the cTTS prototype T Tauri of 0.016 (Beckwith et al. 1990). Thus, calling LkCa 15 a cTTS is also justified from this point of view, and we also label UX Tau A a cTTS. The (rather coarse) upper limit for the disk mass of LkCa 16 is 0.069 . Thus, except for LkCa 16, all other T Tauri stars selected are cTTSs and have disks.

GW Ori is a binary with a separation of  AU (Mathieu 1994), and T Tau is a binary with a separation of 145 AU (Dyck et al. 1982). However, because GW Ori is a single-lined spectroscopic binary, and the companion of T Tau is very faint in the optical regime, we are confident that our optical spectra are dominated by the spectra of the primary stars.

2.3. Observations and data-reduction

For any reasonable assumption about the geometry of the fields, the field-strength will be stronger close to the surface of the star, than at larger distances from the star. The magnetic field measurements should thus be carried out using photospheric lines, rather than emission lines. If the magnetic field structure is very complex, the signals from regions of different polarity will be cancelled out and the Stokes V signal will be very weak. Although spectro-polarimetric techniques have successfully been applied to wTTSs (Donati et al. 1997), we decided to measure the magnetic field strength from the enhancement of the line equivalent widths. This method will work, even if the star is covered by many regions of opposite polarity (Robinson et al. 1980).

The spectra were taken using the University of Utrecht Echelle spectrograph (UES) of the William Herschel Telescope on La Palma during 5 nights between 1994 November 9 and 14. In each of the five observing nights, two spectra were taken of T Tau and of UX Tau A. The spectra of LkCa 15 and LkCa 16 were taken in one night. In addition we observed main sequence templates, one star with a known magnetic field (VY Ari), and each night a B2 V star (see Table 1) for telluric calibration. The spectra were obtained using the TEK2 (1124x1124, pixel) chip, and a two pixel slit (corresponding to 1.1 arcsec as projected on the sky), which gives a resolution of about . The spectra cover the wavelength range from 5260 to 9240 Å. The corresponding S/N-ratios are given in Table 1. Standard IRAF routines were used to subtract bias, flat-field, remove the scattered light, subtract the sky background and to extract and wavelength calibrate the spectra. We used the spectrum of the B2V star BS 179 to remove the slope from each order and to check for telluric absorption lines. Regions with telluric absorption lines were discarded from further analysis.

Table 1. Stars observed, determined from line widths

© European Southern Observatory (ESO) 1999

Online publication: December 16, 1998
helpdesk.link@springer.de