7 The effect of reddening, metallicity, microturbulent velocity and convection on the parameters

7.1 Reddening

Table 5 lists reddening values taken from the literature for some stars of our sample. We converted the E(b-y) from Hayes & Philip (1988) to E(B-V) by using the relation E(B-V)=E(b-y)/0.724 from Crawford & Mandwewala (1976). It is evident that E(B-V) is a rather uncertain quantity for several stars. We therefore examined the effect of the reddening on the parameters of two stars, HD 2857 and HD 86986. By assuming different values of E(B-V), we compared $T_{\rm eff}$ and $\log\,g$ derived by fitting the dereddened IUE short-wavelength spectra to the models with $T_{\rm eff}$ and $\log\,g$ derived by fitting the dereddened observed visible fluxes to the same grid of models. The results are given in Table 6. For HD 2857, E(B-V) between 0.04 mag and 0.06 mag gives $T_{\rm eff}$ from the visible flux in agreement within 50 K with $T_{\rm eff}$ from the ultraviolet flux, but there is always a difference on the order of 0.4-0.5 dex between the corresponding gravities, whichever is E(B-V).

 

 

Table 5: E(B-V) taken from Huenemoerder et al. (1984) (HBC), Hayes & Philip (1988) (HP), Gray et al. (1996) (GCP), Gratton (1998) (G), and Altman & de Boer (2000) (AB)¹.

Star	HBC	HP	GCP	G	AB
BD+00 145	0.05
BD+42 2309		0.000	0.001		0.000
HD 2857	0.02	0.019	0.044		0.050
HD 14829		0.012	0.015		0.020
HD 31943				0.015
HD 60778	0.02	0.014	0.067		0.020
HD 74721	0.02	0.014	0.000	0.000	0.000
HD 78913					0.030
HD 86986	0.03	0.027	0.034	0.035	0.035
HD 93329				0.156
HD 106304					0.040
HD 109995	0.00	0.000	0.001	0.001	0.001
HD 117880	0.02	0.014	0.067		0.015
HD 128801				0.037
HD 130095	0.10	(0.055)	0.063	0.064	0.064
HD 139961	0.10			0.107	0.107
HD 161817	0.02	0.014	0.020	0.020	0.020
HD 167105				0.057
HD 202759		0.068
HD 213468	0.02

¹ Note that Gratton (1998) (G) adopted the reddening from Gray et al.
(1996) (GCP) for the stars in common, and Altman & de Boer (2000) (AB)
adopted the reddening from Gratton (1998) for the stars in common.

 

 

Table 6: The effect of different choices of E(B-V) on model parameters derived from UV (1200-1978 Å) and from the visible (3400-7000 Å) for HD 2857 ( [M/H]=-1.75a, $\xi =4$ km s^-1) and HD 86986 ( $\rm [M/H]=-1.75$a, $\xi =2$ km s^-1).

E(B-V)	$T_{\rm eff}$	$\log\,g$	$T_{\rm eff}$	$\log\,g$
	HD 2857
	UV		visible
0.00	7600	2.3	7350	2.8
0.01	7600	2.5	7450	2.9
0.02	7650	2.5	7550	3.0
0.03	7700	2.6	7550	3.1
0.04	7700	2.7	7650	3.2
0.05	7750	2.8	7700	3.2
0.06	7750	2.9	7800	3.3
0.07	7800	3.0	7900	3.4
0.08	7800	3.1	7950	3.5
	HD 86986
	UV		visible
0.00	8050	2.5	7800	3.1
0.01	8050	2.6	7900	3.1
0.02	8100	2.7	8000	3.2
0.03	8100	2.8	8100	3.3
0.04	8150	2.9	8200	3.4
0.05	8200	3.1	8250	3.4
0.06	8250	3.1	8400	3.5
0.07	8250	3.2	8550	3.4
0.08	8300	3.4	8700	3.4

For HD 86986, E(B-V) between 0.03 mag and 0.05 mag gives $T_{\rm eff}$ from the visible flux in agreement within 50 K with $T_{\rm eff}$ from the ultraviolet flux, but also for this star, there is always a difference on the order of 0.5 dex between the gravities, whichever is E(B-V).

In conclusion, for the two stars examined, a reddening change may improve the agreement between $T_{\rm eff}$ derived from the visible and ultraviolet fluxes, but the discrepancy between the gravities does not decrease.

Figure 7: The ultraviolet energy distribution from IUE (dashed line) and the visual energy distribution from Philip & Hayes (1983) (open circles) of HD 86986 are compared with the computed flux which fits the ultraviolet observations ( $T_{\rm eff}$=8100 K, $\log\,g$=2.8, $\rm [M/H]=-1.75$a) (thick line) and with the computed flux which fits the visual observations ( $T_{\rm eff}$=8150 K, $\log\,g$=3.3, $\rm [M/H]=-1.75$a) (thin line).

7.2 The microturbulent velocity, the metallicity, and the mixing-length parameter

We investigated the effect of different microturbulent velocities $\xi$, different metallicities [M/H], and different mixing-length parameters for the convection $L/H_{\rm p}$ on the parameters of HD 2857 and HD 86986. Table 7 shows that changes in microturbulent velocity, metallicity, and mixing-length parameters modify $T_{\rm eff}$ and $\log\,g$ within the uncertainty limits of the fitting procedure, namely no more than 50 K and 0.1 dex, respectively. Therefore, the discrepancy in $\log\,g$ derived from the ultraviolet and visible fluxes does not change by changing $\xi$, [M/H], or $L/H_{\rm p}$.