1. Introduction

 

During its operational lifetime (26 January 1978 to 30 September 1996) the International Ultraviolet Explorer satellite observed over 3700 objects outside the Solar System in its high dispersion mode. Because a high percentage of these objects have been regularly observed by other satellites and from the ground, a comparison of radial velocities obtained from this instrument is of considerable value to multi-wavelength and time-monitoring investigations that utilize IUE data. However stable a spectrograph might be, its calibration is susceptible to systematic wavelengths errors, usually arising from such real-world complications as the differing paths of the stellar and emission-line calibration beams through the spectrograph, dispersion nonlinearities, and in the case of the IUE, both image placement in the aperture and a wandering of its echelle spectral format on the detector surface. In this paper one of our goals will be to examine several potential sources of systematic error as a function of various instrumental parameters by means of standard cross-correlation tools. Our purpose is to give a general evaluation of the robustness of the wavelength calibration of the ``NEWSIPS" which, in contrast to its predecessor ``IUESIPS" software, is not known to produce wavelength errors as a function of time or other relevant variables.

Unlike other wavelength comparisons of the IUE calibration, our study relies on cross-correlations of like spectra, that is, of the same star. Cross-correlation studies have a high internal accuracy and offer safeguards against systematic errors in the measurement process. However, they also have the disadvantage of referring the measurements to a reference spectrum template for which the zero-point itself must be determined. Thus, a second goal will be to evaluate the ``absolute" wavelength zero-point errors of the IUE cameras in high-dispersion. This assessment has been carried out with respect to instruments on other space-borne platforms, specifically the Goddard High Resolution Spectrograph; (GHRS; operational period 1990-1997) and the Space Telescope Imaging Spectrograph (STIS; 1997-present) on board the Hubble Space Telescope. We will also compare zero-points of the IUE/SWP wavelength ``system" with ground-based measurements of reference stars and absolute (laboratory) values compiled in the literature (Brandt et al. 1998).

A list of actual camera sequence numbers for observations used in this work are available by request from the MAST. The spectra utilized were the ``absolute calibrated" flux extensions of the ``MXHI" FITS files.