As single and binary stars exhibit markedly different behavior
(Schachter et al. 1996), we divided the program stars into two
subsamples, giving separate figures for the results for each
subsample. Our approach to this analysis begins with the simplest
3.1. One-temperature model
In Table 2 we present the one-temperature fit for all program stars which yield (column ) less than 3.5. columns (1) and (2) give sequence number and star name respectively. The emission measures are given in column (3). The best fit temperature and its confidence intervals are given in columns (4) and (5). The column densities derived from the fits are given in column (6). Note that for most stars a statistically significant improvement of the fit was obtained by including finite values for ; HD 128620 was the only program star for which no absorbing column was required.
Table 2. Fitting parameters for one-Temperature model
Combining the resulting temperature T above with the magnetic field strength B given in Table 1, we plot the resulting X-ray temperature versus magnetic field strength: in Fig. 1a for the single stars, and b for the binary and RS CVn stars. From Fig. 1a, we see a correlation between the coronal temperatures and the magnetic field strength for single stars. This result suggests that the corona is heated magnetically. The best quadratic-polynomial fit is for , with a reduced of 1.62. But there is no correlation for the binary stars and RS CVn stars. This can be explained as follows: firstly, all of the companion stars contribute a considerable emission to the total observed luminosity as independent components, secondly in the close systems the orbital motion may induce fast surface rotational velocities by means of tidal coupling, and hence increase the nonthermal heating of the individual stellar chromospheres and coronae (Maggio et al. 1990).
3.2. Two temperature model
Because the one-temperature model fails to fit some program stars, we use the two-temperature model to fit all sample stars. The fitting results which yield (column ) are shown in Table 3. Columns (1)-(5) in Table 3 are the same as the first five columns of Table 2 (presenting parameters for the low-temperature component of the model), while columns (6)-(8) provide analogous information for the high-temperature component.
Table 3. Fitting parameters for two-Temperature model
In Fig. 2, we show - for acceptable two-temperature fits- the resulting X-ray temperature versus the magnetic field strength. From Fig. 2a, we see a remarkable correlation between the high-temperature component and the magnetic field strength for single stars, and the best quadratic-polynomial fit is , with a reduced . Also the best quadratic-polynomial fit for low-temperature components is , with a reduced . This value is too large to be acceptable. So we suggest that there is no correlation for the low-temperature component of single stars, which agrees with the result that the corona in quiet regions is heated accoustically while the corona in active regions is heated magnetically (Mullan & Fleming 1996). And there is no correlation either for the high-temperature component or for the low-temperature component of binaries and RS CVns.
3.3. Two temperature model with intervening absorption
In order to study the influence of intervening absoption, we use the two-temperature model with intervening absorption to fit all sample stars which yields (column. ) . The results are shown in Table 4. Comparing with the Table 3, we find that, in most cases, the improvement is small when accounting for values, but for some sample stars (HD62044,HD115404 and HD165341), a significant improvement in the fit is obtained when finite values of are included. From Table 4, we see the high-temperature component for most of the single stars is lower than , while for most of the binary and RS CVn stars it is higher than . Only star HD222107 fails to fit any coronal model.
Table 4. Fitting parameters for two-Temperature model with Intervening Absorption
We plot the resulting X-ray temperature versus magnetic field strength in Fig. 3, and we can gain the same results as for model 2 from Fig. 3. The best quadratic-polynomial fit for high-temperature component is , with a reduced , and , with a reduced for low-temperature component.
© European Southern Observatory (ESO) 1998
Online publication: October 22, 1998