GHRS Instrument Handbook
The GHRS has a spectrum calibration lamp which is used by specifying SC2 as the aperture with WAVE as the target. It is a platinum-neon hollow cathode lamp manufactured by Westinghouse, providing a rich array of emission lines throughout the ultraviolet region that the GHRS observes. The lines are bright enough so that a 30 to 60 second exposure will yield a good comparison spectrum at almost any wavelength, although longer times are needed at some echelle settings. The lamp has its own aperture, offset from the two science apertures of the GHRS. SC2 forms its spectrum at the same y deflection as the SSA, but displaced in x (the direction of dispersion). The lamp aperture is 67 microns square and it forms Gaussian-shaped images with FWHM = 1.1 diode widths.
A wavelength calibration exposure made with the star in the SSA may be contaminated by the stellar spectrum. This is because the SSA has no shutter and the fact that the SC2 aperture is in line with the SSA. This contamination is rarely a serious problem, however, because it is possible to subtract the stellar component. Also, wavelength calibration exposures are generally used to just confirm the zero-point of the spectrum and not to obtain a full wavelength solution.
For a comprehensive listing of the platinum lines, see Reader et al. (1990).
The light from the spectrum calibration lamps does not enter the spectrograph along the same path that starlight takes. This introduces a wavelength shift that must be compensated for in determining the wavelength solution. The data reduction software incorporates corrections that were determined during pre-flight ground testing (new values are being measured in Cycle 5).
The image formed by the Digicons is affected by the thermal environment within the GHRS, which in turn is influenced by whatever electronics happen to be on or off in the GHRS and the other instruments. The temperature inside the GHRS can be monitored and the image motion calibrated. This correction is also provided for in the data reduction software. However, this correction is applied only once to a given exposure line. We recommend that the exposure times for individual exposure lines be kept shorter than about one hour as long as you do not encounter problems with using too much on-board memory (see Section on page 52).
As for overall image stability, geomagnetic effects influence wavelength stability. Long exposures should be divided into units of about 5 minutes each, the time over which the wavelength scale does not change measurably.
The precision and accuracy of the wavelength scale of a GHRS spectrum depend on:
- The carrousel position (geometric factors) and the repeatability of that position.
- Motions of the spectrum on various time scales due to thermal effects and the earth's magnetic field, among other things.
The quality of the wavelength scale is primarily an issue with ACCUM mode observations; RAPID mode observations cannot be interrupted and are usually made to look for short-time-scale variability, not wavelength shifts. Our requirement for routine wavelength calibrations is to have them good to about one diode rms. We achieve much better than this in practice. In particular, the default wavelength scale (i.e., the dispersion) is reliable to about 2%, and the zero-point can be established to within about 0.2 diode by obtaining a wavelength calibration exposure just before your ACCUM.
The GHRS wavelength monitoring program runs four times per year for both the echelles and first-order gratings to confirm that anomalies are not occurring and also to build up the database used to establish how motions depend on temperature, time, and magnetic field.