Table 8 presents a summary of the radial velocities of interstellar features in several stars, taken from Table 1, together with values from the literature (v(lit)). The LWP and LWR velocities of Table 1, being consistent with each other, have been averaged together into v(LW).
The literature value for RR Tel is from Tackeray (1977). The value quoted for Oph measured in the optical Ca II K and Na I D lines has been taken from Barlow et al. (1995). This value is in good agreement with the velocities derived from GHRS ultraviolet spectra by Savage et al. (1992), -14.9 km s-1, and Brandt et al. (1996), -15.4 km s-1. The velocities for UMa and Cas correspond to measurements of the Ca II K and Na I D optical lines reported by Vallerga et al. (1993). The velocity quoted for Lep refers to the optical Ca II doublet (Frisch et al. 1990), which presents two components at 2 and 18 km s-1, which cannot be resolved with IUE. The spectrum of HD 93521 presents up to nine interstellar components, with the two strongest ones located at, approximately, -10 and -60 km s-1 (Spitzer & Fitzpatrick 1993). In the IUE spectra all these systems appear blended, and therefore, as in the case of Lep, we cannot compare reliably the IUE velocities with the optical values.
According to the data in Table 8, the mean difference between long and short wavelength (large aperture) velocities is 17.7 km s-1, i.e. SWP velocities are systematically more negative. The mean difference between the long wavelength and the literature values is 8 km s-1.
A similar test was made to check the consistency between the wavelength scale of spectra taken through the large and small apertures. Being the number of small aperture high resolution spectra very limited, useful data were available only for the star Oph in the SWP and LWR cameras. The wavelength scales of the small and large aperture spectra are fully consistent in the short wavelength range: v(LAP) - v(SAP) = 1.1 6.7 km s-1, while for the LWR camera a significant difference is found: v(LAP) - v(SAP) = -13.7 4.1 km s-1. The lack of a suitable data set precludes an accurate determination of the offset between the large and small aperture scales in LWP spectra, but the limited tests performed seem to indicate that small aperture velocities are systematically lower, although the actual difference cannot be quantified.
|Target||v(SWP)||v(LW)||v(lit)||v(LW) - v(SWP)||v(lit) - v(LW)|
|km s-1||km s-1||km s-1||km s-1||km s-1|
|HD 93521||-38.8||-20.1||-10, -60||18.8|
Mean difference v(LW) - v(SWP) = 17.7
3.7 km s-1.
Mean difference v(lit) - v(LW) = -8.0 4.2 km s-1.
The reason for the discrepancy between the short and long wavelength range velocity scales is not clear, while the large/small aperture discrepancy in LW spectra is most likely related to the transfer of the dispersion constants from the small to the large aperture: the dispersion relations were derived from spectra taken through the small aperture and then transferred to the large aperture on the basis of the assumed aperture separations.
In order to provide an internally consistent wavelength scale within the INES system, a velocity correction of +17.7 km s-1 has been applied to the wavelength scale of SWP high resolution spectra. The wavelength scale of LWP/LWR small aperture spectra has been corrected by +13.7 km s-1. With these corrections, the INES velocity scale is consistent with the optical determinations.
Copyright The European Southern Observatory (ESO)