Space Telescope Imaging Spectrograph
Updates to the STIS Calibration Software, Reference Files, and OTFR Pipeline Products

Recent updates to the STIS calibration include changes to the CALSTIS code, updates to reference files, and revisions to rules determining the default settings for some calibration flags. Many of these changes are related to the flux calibration of 1D extracted spectra, and are especially relevant for first-order spectra obtained using narrow slits at the E1 aperture positions.

CALSTIS 2.19 changes

Version 2.19 of CALSTIS was released as part of STSDAS/TABLES v3.4 on 2005-Nov-03. This new version of CALSTIS was implemented in the MAST OTFR pipeline for STScI data as part of the OPUS 2006.1 release on 2005-Dec-13. A number of important revisions and improvements were included in this release.

1. Changes in the flux calibration for 1D extracted spectra
   
   
   • The sensitivity of some STIS modes has a small dependence on temperature (see STIS ISR 2004-04). CALSTIS now includes corrections for this dependence when calculating the 1D extracted flux, using coefficients that have been added to the TDS reference table.
   
   
   • The G430L and G750L gratings have Lyot stops that are not included for other spectroscopic modes. As a result, the aperture and grating throughputs cannot be simply combined to give the net throughput; instead, an additional correction of as much as 10% is needed for some grating/aperture combinations. CALSTIS now supports a new Grating-Aperture Correction reference file table (GACTAB) that allows an additional correction vector to be specified for any grating/aperture/cenwave combination.
   
   
   • An error in the flux correction for CTI effects for subarray data was fixed. Previously, for data taken using subarrays, the correction was calculated as if the subarray occupied the rows of the detector farthest from the readout. This could lead to a significant over-estimation of the CTI correction in many circumstances. Note that the spectroscopic CTI correction formula was determined using full frame data, so even after fixing this error, the empirical CTI calibration may not be as accurate for data taken using subarrays as it is for full frame data.


2. New xoffset parameter for x1d task.
   A new parameter is available when running the STIS STSDAS x1d task under IRAF or PyRAF. This xoffset parameter specifies a displacement of a spectrum in the dispersion direction in units of un-binned detector pixels. This shift is applied when calculating the wavelength solution for the spectrum. Previously, applying such a shift required editing the SHIFTA1 parameter in the image header. This was often awkward, especially for images containing spectra of several objects, where each object might require a different offset.


3. Miscellaneous bug fixes.
   Bugs in the CALSTIS code that could cause intermittent crashes for some long slit wavecals or for images where the virtual overscan was equal to zero everywhere were fixed.

Additional details can be found in the release notes for CALSTIS 2.19.

New STIS reference files

A number of STIS reference file revisions have recently been made. Some of these changes are to support the modifications made to the CALSTIS 2.19 code.

1. A new Grating-Aperture Correction table (GACTAB reference file type, suffix _gac) was defined to allow CALSTIS to apply additional throughput corrections that depend on the combination of grating and aperture. Currently, corrections are only supplied for G230LB, G430L, and G750L observations with the 52" long apertures.


2. The TDS (time-dependent sensitivity) reference file tables have been revised to also include the coefficients needed to correct the sensitivity for detector temperature.


3. New low-order flat-field files have been delivered for the G230LB, G430L, and G750L gratings, to give a better flux calibration as a function of the target position along the length of the aperture. This delivery was coordinated with the installation of CALSTIS 2.19.


4. New throughput curves for CCD and MAMA medium resolution first-order spectroscopic modes (G*M) were delivered for pipeline use on 2005-Aug-02. These new curves are based on calibration observations of fundamental WD standards.

Revised rules for setting some calibration flags

1. CTECORR is now set by default to "OMIT" for data taken using subarrays. This was done because the CTI correction algorithm implemented in CALSTIS (see STIS ISR 2003-03) was calibrated using data taken using the full CCD image. These results may not apply to data taken using subarrays because of the different readout clocking pattern used for the parts of the CCD image that are not saved when only a subarray is saved.


2. X1DCORR and X2DCORR will now be set by default to OMIT for all slitless data, including NUV PRISM observations. For this purpose slitless data includes any spectrum taken with apertures 6" or larger in the dispersion direction. The assumption is that many slitless observations were either done without an ACQ observation, or targeted complex fields that may contain multiple sources. For any sources displaced significantly in the dispersion direction, standard pipeline x1d or x2d extractions would have significant errors in the wavelength scale, and such a misalignment of the wavelength scale will also introduce a corresponding misalignment of the sensitivity curve, potentially resulting in large errors in the extracted fluxes. For slitless data, it will generally be necessary for the observer to determine the offset of each target in the dispersion direction, and to then perform a customized extraction.


3. For data taken using the barred fiducial apertures (e.g., 52X0.2F1), X1DCORR is now set by default to OMIT. X2DCORR may be set to PERFORM if such calibration is otherwise appropriate for the observation (i.e., a supported mode with a wavecal).