6 Parameters from the visible energy distribution for some stars of the sample

The parameters adopted by KCC are the average of parameters obtained with different methods. They therefore represent a statistically "most likely'' solution, but they mask the results obtained with a particular method. We wish to verify here how the parameters derived from the IUE spectra compare to those derived only from the visible energy distribution. Because there are 11 stars in our sample having visible spectrophotometry available in Philip & Hayes (1983), we fitted these observations to the new-ODF models in order to derive both $T_{\rm eff}$ and $\log\,g$ from the flux method for stars cooler than about 9000 K and only $T_{\rm eff}$ for the hotter stars. In fact, as is illustrated in Fig. 5, the visible flux method is not well suited to derive the gravity for giant stars hotter than about 9000 K. For these stars we will use the H$_{\gamma }$ profile to derive $\log\,g$.

Figure 5 is a plot of the the Balmer discontinuity as a function of $T_{\rm eff}$ for different $\log\,g$, where the Balmer jump is represented as a difference of two magnitudes. The first one is the magnitude averaged over five wavelengths in the UV ( $\lambda\lambda$ 3400, 3450, 3500, 3571, and 3636 Å) and the second one is the magnitude averaged over five wavelengths in the visible ( $\lambda\lambda$ 4036, 4167, 4255, 4464, and 4566 Å). The wavelengths are the same of the observed energy distributions listed by Philip & Hayes (1983) and used by them to estimate the errors of their scanner observations in the UV and in the blue, respectively. Figure 5 shows also the magnitude differences from Philip & Hayes (1983) for the 11 stars listed in Table 4. The error bars were obtained from the standard deviations quoted in their paper. They give an estimate of the error for the observed Balmer jump. The most uncertain data are those for BD+42 2309 and HD 14829.

Table 4 compares the parameters from the IUE spectra with those from the visible energy distribution and from H$_{\gamma }$. The observed H$_{\gamma }$ profiles are taken from KCC, while the synthetic profiles were computed with the BALMER9 code of Kurucz (1993). The observed H$_{\gamma }$ profiles were fitted to a grid of profiles computed for different $\log\,g$, while $T_{\rm eff}$ is that derived from the visible flux. Because H$_{\gamma }$ for HD 14829 was not observed, we used the flux method to obtain also $\log\,g$ for this star. The differences in the parameters are plotted in Fig. 6. This figure is very similar to Fig. 4, where the parameters from UV and from KCC are compared. The large difference between $T_{\rm eff}$ from the UV and from the visible for BD+42 2309 is probably related with the poor observations for this star both in the UV and in the visible.

The conclusion is that gravities derived only from the visible flux distribution for stars cooler than about 8700 K are, on average, systematically larger by about 0.3 dex than those derived only from the ultraviolet flux distribution. For instance, Fig. 7 shows that there is no doubt about the need of gravities differing by 0.5 dex in order to reproduce the short ultraviolet IUE spectrum and the visible energy distribution of HD 86986.

We remark that the parameters obtained by KCC only from the visible energy distribution (Table 7 in KCC) can not be compared with those listed in Cols. 7 and 8 of Table 4, owing to the different E(B-V) adopted in the two cases. In fact, $T_{\rm eff}$ and $\log\,g$ in KCC correspond to the E(B-V) yielding the best fit of the observed energy distribution to the models.

 

 

Table 4: Comparison of $T_{\rm eff}$ and $\log\,g$ from ultraviolet and visible energy distributions and from H$_{\gamma }$.

Star	E(B-V)	[M/H]	$\xi$	$T_{\rm eff}$	$\log\,g$	$T_{\rm eff}$	$\log\,g$	$\log\,g$
				UV		visible		H$_{\gamma }$
HD 117880	0.077	[-1.50a]	2.0	9350	3.3	9400		3.4
HD 130095	0.072	[-1.75a]	2.0	9100	3.2	9050		3.3
HD 14829	0.018	[-2.50a]	2.0	8900	3.1	9000	3.1
HD 74721	0.012	[-1.50a]	2.0	8800	3.2	8850	3.3
BD+42 2309	0.013	[-1.75a]	2.0	8750	3.0	9100		3.4
HD 109995	0.010	[-1.75a]	2.0	8500	3.0	8450	3.4
HD 60778	0.028	[-1.50a]	4.0	8250	2.9	8050	3.1
HD 86986	0.022	[-1.75a]	2.0	8100	2.8	8050	3.3
HD 2857	0.022	[-1.75a]	4.0	7600	2.8	7550	3.0
HD 161817	0.000	[-1.50a]	4.0	7600	2.7	7600	3.1
HD 202759	0.072	[-2.00a]	2.0	7500	2.8	7400	3.0

Figure 5: The curves represent computed magnitude differences, Mag(UV)-Mag(vis), as function of $T_{\rm eff}$ for different gravities and $\rm [M/H]=-1.5$a. Mag(UV) is the magnitude averaged over five wavelengths in the 3400-3636 Å region and Mag(vis) is the magnitude averaged over five wavelengths in the 4036-4566 Å region. The asterisks with error bars represent the magnitude differences for 11 stars obtained from the Philip & Hayes (1983) scans. The correspondence of each number to each star is: 1=HD 117880, 2=HD 130095, 3=HD 14829, 4=HD74721, 5=BD+42 2309, 6=HD 109995, 7=HD 60778, 8=HD 86986, 9=HD 2957, 10=HD 161817, 11=HD 202759.

Figure 6: Upper panel: differences between $T_{\rm eff}$ derived from visual and ultraviolet energy distributions, as a function of $T_{\rm eff}$ for stars with available spectrophotometry. Lower panel: same as upper panel, for $\log\,g$.