FOS Instrument Science Report CAL/FOS-149 (rev. 1 : 30 Sep 1998)

----------------

Zero-Points of FOS Wavelength Scales 

----------------

Michael R. Rosa (1,2), Florian Kerber (1,3), and Charles D.(Tony) Keyes (4) 

1) Space Telescope European Coordinating Facility, European Southern Observatory, Garching, 
	Germany

2) Affiliated to the Astrophysics Division, Space Science Department, European Space Agency

3) Institute for Astronomie, University of Innsbruck, Austria

4) Space Telescope Science Institute, Baltimore, MD

-----------------------

June 1998 (REVISED 30 September 1998)

Abstract

	The internal zero-points of the FOS on-orbit wavelength calibration between 1990
(launch, OV/ SV) and 1997 (decommissioning) have been determined. The analysis is based on
cross-correlating all WAVECAL exposures for the high-resolution dispersers and for the low
resolution G650L mode (about 1200 data sets), using as templates those exposures which define
the dispersion solutions currently in use by the FOS pipeline. For FOS/BL systematic shifts of
the zero-points are present, which are very similar for all disperser and aperture
combinations, and amount to a maximum offset of 7 pixels (1.75 diodes) over the entire period.
The zero-points for FOS/RD present an apparently random distribution with a peak-to-peak range
of 7 pixels (1.75 diodes). The shifts are provisionally attributed to the differing behavior of
the two detectors in the geomagnetic field, as well as residual errors not compensated by the
on-board GIM algorithm. We describe a corrective procedure which will restore a maximum 1
pixel peak-to-peak uncertainty in the wavelength scale to all FOS/BL high-resolution and G650L
spectra. The information provided in this ISR will allow manual post-processing correction of
the wavelength scale for FOS/BL data taken at any epoch; STSDAS/IRAF, IDL, and MIDAS scripts
for correction of the wavelength scales are available for download on the STScI FOS WWW pages.
The correction will also be implemented in calfos, once the FOS/RD data have been fully
analyzed, so that routine re-calibration of FOS data will result in correct wavelengths. At
the time of writing the corrective measures for FOS/RD data have not yet been determined and
the effect of the wavelength correction for FOS/BL on instrumental sensitivities (flux
calibration) is still under investigation.