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

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

Radial Velocities

For each of our spectra, we measured the radial velocities of the H $\alpha$ line and metal lines, excluding H $\alpha$ , using the FXCOR routine in IRAF. Table 2 and Figure 2 show the heliocentric-corrected radial velocities in both cases. We have also shown in Figure 2 the sinusoid curve which Shahbaz et al. (1999) found by fitting their radial velocities resulting in K₂ = 215 kms^-1 and a systemic velocity, $\gamma$ = -141.9 kms^-1.

As one can see, H $\alpha$ and the metal lines give consistent velocities within the error bars. Therefore, we conclude that H $\alpha$ is not significantly affected by emission originating from a possible accretion disk or X-ray heating of the companion star.

In order for our points to coincide with the radial velocity curve, we have had to shift them in phase by -0.06 with respect to the van der Hooft et al. (1998) orbital ephemeris, while Shahbaz et al. (1999) shifted theirs by -0.05. This increase in phase-shift implies that the orbital period needs to be revised. As we have a very limited number of spectra, refining the orbital period from our data would be risky. It would be desirable to combine our data with that of Shahbaz et al.. Unfortunately, their radial velocities (and other groups') are not published in tabular form. We were unable to obtain Shahbaz et al.'s data in time to include in this paper.

An interesting feature is found near orbital phase 0.6. Both of our (shifted) points lie well below the radial velocity curve. This has also been seen by Orosz & Bailyn (1997). In particular, the points from 1995 May 2 (when the source was in outburst) are deviated significantly from the best-fit sinusoid at their orbital phase of $\sim$ 0.3. Radial velocities from 1996 February 25 (quiescent period) do not deviate as much, but nevertheless have a steeper gradient than the curve similar to 1995 May 2 data. Phillips et al. (1999) produced a corrected radial velocity curve using a model based on an X-ray heated companion star and using only radial velocities which were obtained during outburst. Their fit to data was an improvement on past efforts, however, there is still a hint of deviation away from the curve near orbital phase 0.6. Shahbaz et al. (1999) did not obtain any data at this phase. It is difficult to conclude from our data alone whether this feature is present in both outburst and quiescent data. Observers may wish to concentrate on this area in the future.

**Figure 2:** Heliocentric-corrected radial velocities for GRO J1655-40 spectra. The sinusoid curve has K₂ = 215 kms^-1 and $\gamma$ = -141.9 kms^-1 (Shahbaz et al. 1999), the phases are calculated from the van der Hooft et al. (1998) orbital ephemeris. Filled squares are radial velocities without any phase shift while open triangles are those shifted by -0.06. $\it {Upper}$ $\it {panel}$ : Radial velocities measured for H $\alpha$ line only. $\it {Lower}$ $\it {panel}$ : Radial velocities measured for metal lines, excluding H $\alpha$ .
$\begin{figure} \centerline{\psfig{file=fx.ps,height=15cm,width=15cm}}\end{figure}$

Table 2: Radial velocities for GRO J1655-40 obtained from H $\alpha$ and metal lines (excluding H $\alpha$ )

Heliocentric Julian Date	H $\alpha$	Metal Lines
(-2440000)	(kms^-1)	(kms^-1)
11282.1163	-333.014	-319.831
11282.1605	-360.025	-347.278
11283.0649	-176.143	-175.887
11283.0868	-173.467	-129.972
11283.1157	-162.771	-132.205
11283.1377	-151.136	-124.401
11283.1593	-125.556	-106.689

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

ASKAP

Public

Australia Telescope National Facility

Radial Velocities