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

5. Results

To test the method, we applied it first to a solar spectrum, to VY Ari (B =  kG, , Bopp et al. 1989), and to HD 42807 (G2 V).

The solar spectrum was used as a non-magnetic reference, as the product of its filling factor and its magnetic field strength is kG (LaBonte 1981) even at sunspot maximum.

For VY Ari (Fig. 5) our method yields B = 1.5-2.8 kG, 0.55, and kG. Our is only slightly larger than the 1.3 kG found by Bopp et al. and the separate values of B and f found by Bopp et al. are both within the limits which we obtain (dotted line in Fig. 5).

Fig. 7 shows the result for HD 42807, with the Sun as template star. As expected, we did not find any evidence for a magnetic field. For any combination of 0 and no significant correlation between measured and computed EWs is found. The maximum correlation coefficient found occurs at 0.25 kG which is the smallest field strength tested. The value of of 0.15 is distinctly below the significance limit, indicating that no field was detected (in the case of HD 42807 one must have for a 95%-level detection).

Table 3 lists the results derived for all of our program stars. For all stars except GW Ori and the nonmagnetic test stars, all maxima found for give a significance of a linear correlation Y with X of better than 99%. The region of statistically indistinguishable correlation coefficients is relatively large compared to the narrow limits obtained for , however. Although the method allows us to determine lower and upper limits for the magnetic field strength, it gives only lower limits for possible inclinations of the field and for the filling factor, as a filling factor of one is consistent with the data for LkCa 15, T Tau, and VY Ari (see Table 3).

A magnetic field could be detected in the two cTTS T Tau and LkCa 15. Besides we can give limits for the magnetic field strength and lower limits for filling factor and mean inclination.

For the wTTS LkCa16 the contour plot shows a significant correlation but the result depends mostly on two individual spectral lines. Furthermore, the total number of lines is quite small, and as a consequence slightly wrongly chosen values for the effective temperature and microturbulence may have its effect on the results. On the other hand, these two lines have the highest magnetic sensitivity and at least indicate the existence of a field with kG strength.

The cTTS UX Tau A is also close to the limit of a detection. Again we have found a statistically significant correlation but we can give no lower or upper limit for the magnetic field strength. For all of the other T Tauri stars there is an significant reduction in the scatter of the empirical curve of growth after correcting the measured EWs for the expected magnetic intensification, whereas for UX Tau A the scatter increases. So we think we have only a marginal detection of a field with kG.

The cTTS GW Ori clearly shows the limitations of our method. Because of the larger of 30 km/s the number of suitable lines is too small to detect any weak magnetic field that might be present.

© European Southern Observatory (ESO) 1999

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