Atmospheric Modelling of the Companion Star in
GRO J1655-40
Michelle Buxton , Stephane Vennes, PASA, 18 (1), in press.

Next Section: Discussion
Title/Abstract Page: Atmospheric Modelling of the
Previous Section: Fitting Code
Contents Page: Volume 18, Number 1

Results

We found during initial tests that gravity was not well constrained as it is primarily determined by H $\alpha$ . Our spectra do not cover any other Balmer lines which could be used to determine $\log g$ more accurately. Therefore, $\log g$ was fixed at 3.7 (Israelian et al. 1999). We also found that the fits were better for the F-star spectra with $\xi$ = 1 kms^-1 and for GRO J1655-40 $\xi$ = 2 kms^-1. Limb darkening for GRO J1655-40 was set at 0.5 (Allen 1963, Andersen, Clausen & Nördstrom 1984).

We tested our program on the solar spectrum convolved to our instrument resolution. The result is shown in Figure 3(a). The fit is good and our best-fit parameters are given in Table 4 with a comparison to published values. Another test was made on our F-star template spectra. We show a typical fit to one F-star (HR 5455) in Figure 3(b) and note that this also is a good fit, consistent with the literature (Table 5). Therefore, we are confident that our program is working satisfactorily.

Figure 4 and Table 6 give the fit result for GRO J1655-40. Our fitting code gives $v_{rot}\sin i$ = 80 $\pm$ 10 kms^-1 (1 $\sigma$ ).

Taking our value for $v_{rot}\sin i$ and K₂ = 215 kms^-1 (Shahbaz et al. 1999), we find using equations (1) and (2) that q = 0.31 $\pm$ 0.08. Therefore, from equation (3) and taking i = 67.20 $\pm$ 3.50^o (van der Hooft et al. 1998), M₁ = 7.91 $\pm$ 3.79 M $_\odot$ (1 $\sigma$ ). From q, we infer M₂ = 2.76 $\pm$ 1.79 M $_\odot$ (1 $\sigma$ ). These results agree well with Orosz & Bailyn (1997) and Shahbaz et al. (1999) as well as other published results (Table 1).

**Figure 3:** (a) Model fit of the solar spectrum. $\it {Top}$ : Histogram is solar spectrum, smooth line is the model fit. $\it {Bottom}$ : 1 $\sigma$ , 2 $\sigma$ and 3 $\sigma$ level contours of minimum $\chi ^{2}$ fit. (b) Model fit to HR 5455 (F V star) template spectrum. $\it {Top}$ : Histogram is HR 5455 spectrum, smooth line is the model fit. $\it {Bottom}$ : 1 $\sigma$ , 2 $\sigma$ and 3 $\sigma$ level contours of minimum $\chi ^{2}$ fit. Also see tables 4 and 5.
$\begin{figure} \centerline{ (a) \psfig{file=fig_sol.ps,height=10cm} (b) \psfig{file=fig_0052.ps,height=10cm}}\end{figure}$

Table 4: Fitting results for solar spectrum.

Parameter	Best-Fit Value	Published Values	Reference
T_eff (K)	5750 $\pm$ 25	5800 $\pm$ 15	1
[Fe/H]	-0.04 ^+0.01_-0.13	0.0	definition
$v_{rot}\sin i$ (kms^-1)	0 - 2	2	2

(1) Allen 1963 (2) Wöhl & Schmidt 2000

Table 5: Fitting results for HR 5455 (F5 V star) spectrum.

Parameter	Best-Fit Values	Published Values	Reference
T_eff (K)	6200⁺²⁰_-10	6146	1
		6313	2
[Fe/H]	-0.3 ^+0.02_-0.1	-0.22	1
		-0.19	2
$v_{rot}\sin i$ (kms^-1)	20 $\pm$ 5	15	3

(1) Boesgaard & Tripicco 1986 (2) Balachandran 1990 (3) Hoffleit 1982

**Figure 4:** Model fit to summed GRO J1655-40 spectrum. $\it {Top}$ : Histogram is GRO J1655-40 spectrum, smooth line is model fit. $\it {Bottom}$ : 1 $\sigma$ , 2 $\sigma$ and 3 $\sigma$ level contours of minimum $\chi ^{2}$ fit.
$\begin{figure} \centerline{\psfig{file=fig_redgro_limb0.5.ps,height=15cm}}\end{figure}$

Table 6: Fitting results for summed GRO J1655-40 spectrum.

Parameter	Best-Fit Values	Published Values	Reference
T_eff (K)	6500 $\pm$ 50
[Fe/H]	-0.15 ^+0.15_-0.10	0.1 $\pm$ 0.2	1
$v_{rot}\sin i$ (kms^-1)	80 $\pm$ 10	88.9 $\pm$ 6	2
		93 $\pm$ 3	1
(1) Israelian et al. 1999
(2) Shahbaz et al. 1999

Next Section: Discussion
Title/Abstract Page: Atmospheric Modelling of the
Previous Section: Fitting Code
Contents Page: Volume 18, Number 1

ASKAP

Public

Australia Telescope National Facility

Results