1.5	STIS Closeout

Given the long and possibly permanent hiatus in STIS operations, it was decided that a final closeout of existing STIS data would be done. As part of this closeout effort, a substantial number of improvements were made in calstis pipeline software and reference files. Most notable among these are:

•	Updated echelle flux calibration and blaze shift corrections. See ISR STIS 2007-01 and Aloisi (2005 HST Calibration Workshop).

•	Updated flux calibration for first-order medium resolution spectra. See ISR 2006-04.

•	Improved flux calibration for first-order spectra taken at E1 aperture positions or with narrow apertures. See Proffitt (2005 HST Calibration Workshop).

•	Improved algorithm to correct for charge transfer inefficiency (CTI) when extracting fluxes for first-order spectra. See ISR 2006-01 and Goudfrooij et al. (2006, PASP, 118, 1455).

•	Improved spectral traces for the most commonly used first-order modes, including date-dependent rotation parameters. See the Oct./Nov. 2006 STAN article and Dressel (2005 HST Calibration Workshop).

•	Recommended fringe flat exposures are now delivered with most G750L and G750M data, and the name of this recommended fringe flat is put into the FRNGFLAT keyword in the data file header.

•

Association of GO-specified wavecals. Normally wavelength calibration exposures are automatically inserted for all STIS external spectroscopic observations, and these are used by the calibration pipeline to determine the zero point offset for the dispersion solution. Without this measurement, wavelength and flux calibrated spectra cannot be produced. If observers turned off the auto-wavecals and substituted separate, user specified calibration lamp exposures (referred to as GO wavecals), the calibration pipeline did not know how to associate these with the appropriate science observations. As part of the STIS calibration closeout, those science exposures that lack auto-wavecals, but for which GO wavecals are available, were identified, and the science data and lamp exposures were combined into associations that treat the GO wavecals in the same way as auto-wavecals. Be warned that as a side effect of this procedure, the science exposures involved have new dataset names, and the lamp exposures now used as wavecals will no longer appear in the archive catalog as independent datasets. (This task is currently in progress, and completion is expected sometime during 2007).

•

For a small number of spectroscopic data sets which were taken without either automatic or GO specified wavecals, a procedure was devised that allows fixed shift values to be read from a data base during OTFR calibration. For these data sets the SHIFTA1 keyword in each sci extension header of the _raw file is set to a specified value, and the WAVECORR header keyword in the primary header of the _raw file is set to “COMPLETE”. This allows calstis to produce fully calibrated 1D and 2D spectra despite the lack of any wavecal observation. Users should remember, however, that the shifts in such cases are imposed rather than measured. Such data sets can be easily identified, as they have the WAVECAL header keyword set to “N/A” and the WAVECORR keyword set to “COMPLETE”.

The normal HST archive ingest procedures only save the packaged telemetry “pod files”. While raw and calibrated files are produced as part of the ingest procedures, these were not saved for STIS. When a user requests a dataset via the “On The Fly Reprocessing” (OTFR) procedures, the raw files are regenerated and recalibrated using the latest versions of the software and reference files.

As part of the STIS closeout, all STIS data collected prior to the Side 2 failure were reprocessed through a modified version of the ingest pipeline, and the archive databases were repopulated. This allowed any changes in the processing software to be reflected in the archive catalog. Unlike during original ingest, during this recalibration all raw and calibrated data files produced were saved. This static archive is now available to the community. This allows STIS data to be retrieved without waiting for the OTFR procedures to run.