the observing sequence adjacent to STIS spectroscopic observations. This is necessary because non-repeatability of the mechanism holding the gratings and thermal flexure of the STIS bench can lead to small but significant zero-point errors in the wavelength scale. HST Guest Observers are allowed to add additional STIS wavecal observations, beyond those automatically inserted, by specifying an exposure with TARGET=WAVE (see page 232 in STIS IHB section 11.2.1). These user specified wavelength calibration observations are also referred to as "GO-WAVECALs".
Users can also turn off AUTO-WAVECALs for all exposures in a visit by specifying the optional parameter "WAVECAL=NO" on any exposure in the visit. This is an available, but unsupported option, and users must provide a strong scientific justification for turning off the AUTO-WAVECALs. In such a case, any TARGET=WAVE exposures specified by the user would be the only wavelength calibration exposures taken. In past cycles, when AUTO-WAVECALs were turned off, the pipeline calibration software could not make use of any TARGET=WAVE exposures to calibrate the data, and without a wavelength calibration, the pipeline would not perform any spectral extraction at all, because offsets in the wavelength scale can lead to significant misalignment of the sensitivity function, perhaps producing spurious spectral features. For data taken before the STIS failure in 2004 for which the AUTO-WAVECALs were omitted, but for which appropriate TARGET=WAVE exposures were taken, these TARGET=WAVE exposures have now been associated with the corresponding science exposures, and the data have been re-calibrated treating these GO-WAVECALs in the same manner as AUTO-WAVECALs. The re-association of exposures this entailed required re-naming some data sets. The mapping of old to new data set names can be found at http://www.stsci.edu/hst/stis/faqs/faq_asn_mapping.html . For future observations taken after STIS is repaired in SM4, if the AUTO-WAVECALs are turned off, any appropriate TARGET=WAVE exposures will now automatically be used by the pipeline calibration in place of the missing AUTO-WAVECALs. This will allow fully calibrated data to be produced by the pipeline. When used in place of AUTO-WAVECALs, the TARGET=WAVE exposures will now be put into the _wav extension of the science exposure associations, rather than being delivered to the user as separate exposure association files. TARGET=WAVE exposures which are done in addition to, rather than in place of, the AUTO-WAVECALs will continue to be delivered as separate exposure files, and will not be automatically used by the pipeline calibration. Note that TARGET=WAVE exposures must be inserted immediately adjacent to the corresponding science exposures and must specify the same grating and central wavelength as the science observation, although they may use a different aperture. Previous versions of APT had allowed the user to select almost any aperture for use with TARGET=WAVE observations. However, the ground system only allows certain grating-aperture combinations to be used for lamp exposures, and for cases where an inappropriate aperture was selected in APT, the ground system would replace the aperture selection with an allowed default aperture. However, this change was not visible to the observer until after the observation was taken. This has been fixed in version 17.2 of APT, and users are now only allowed to select those apertures for TARGET=WAVE observations that the system will allow to be used on-orbit.