A Study of Wavelength Calibration of NEWSIPS High-Dispersion Spectra

Abstract

In this study we cross-correlate many IUE echellograms of a variety of stars to evaluate systematic error sources in the wavelength zeropoint of all three cameras. We first evaluated differences between the final archived ("NEWSIPS") and the originally processed ("IUESIPS") spectra. These show a clear time dependence in zeropoint for the SWP camera due to revisions in the IUESIPS wavelength scale. Small IUESIPS - NEWSIPS differences are also found for the LWR camera. We also examined wavelength zeropoint disparities between data obtained both through the small and large entrance apertures and for observations made by different target acquisition modes for faint and bright stars. We found that velocities resulting from these alternative observing modes are nil. For large-aperture observations the dominant error source is the target position placement in the aperture. We searched for spurious trends with time, and found only a suggestion of time trends for faint stars observed with the SWP camera. We also discovered 1-day, +/-3 km/s sinusoidsal patterns in intensive monitoring data which are ascribable to changes in telescope focus resulting from thermal drifts. In the second part of the paper, we measured mean zeropoint errors of NEWSIPS echellogram data against laboratory results by using the GHRS spectral atlas of the 10 Lac. We find that the derived apparent velocity difference for this star is -1 +/-3.5 km/s. Several less precise comparisons lead to similar results. The zeropoints of the NEWSIPS-processed LWP/LWR cameras are evaluated and are also found to be nearly zero (+/-5 km/s) relative to HST atlases of Arcturus and Procyon atlas. These results do not support result by Gonzalez-Riestra et al. that corrections should be introduced to the wavelength scales of various NEWSIPS high-dispersion data products.