STScI Logo

Space Telescope Imaging Spectrograph
STAN - September 1997

             |               ___ _____ __  ___            |
STScI        |             ,' _//_  _// /,' _/            | STAN - STIS
Analysis     |            _\ `.  / / / /_\ `.             | Number 5
News         |           /___,' /_/ /_//___,'             | September 1997
             |                                            |

       - STIS NEWS 
         + SMOV2 Completed
         + STIS ERO Data Now Available
         + Target Acquisition Accuracy - On-board Results
         + 50CCD Performance
         + Insertion of CCD Fringe Flats
         + Changes to Limiting Magnitudes for MAMA Bright Object Protection 
         + Pipeline Calibration News
         + On-Orbit Performance Pages on WWW
         + Cycle-7 Calibration Plan
         + Calibration Workshop (Sep 22-24, 1997)
         + What's New in the STIS WWW Pages

         + STIS FAQs (Frequently Asked Questions)...
         + Target Acquisition ETC
         + On-Orbit Performance WWW Pages
         + Draft Data Handbook (STIS Chapters) On-line


SMOV2 Completed; STIS GO Science Starting in Earnest

Effective Thursday, 25 September 1997, all SMOV planned activities have
been completed.  STIS has been fully commissioned for science operations.

The Long Range Plan has been rebuilt for STIS with roughly 90% of the
GO and GTO orbits placed upon it. Starting the first week in October,
STIS GO+GTO science starts in earnest for STIS. Get ready.

STIS ERO Data Now Available

With the completion of the Servicing Mission Orbital Verification, all 
STIS Early Release Observations (EROs) are now publicly available from the 
HST Data Archive.  The results have been nothing less than spectacular 
providing providing ample evidence of the excellent quality of data obtained
with STIS.  A list of the programs executed as part of the ERO suite is 
given below.  

   7139   UV Spectral Dating of Galaxies
   7121   Imaging Spectroscopy of NGC 4151
   7122   Imaging of SN1987A and its nebula
   7123   MAMA and CCD Spectroscopy of SN1987A
   7130   STIS/Spectroscopy of the Brown Dwarf Gl229B
   7124   Kinematics of the Nuclear Gas Disk in the Galaxy M84   
   7125   CCD Coronographic Imaging of the Disk of Beta Pictoris
   7137   STIS Observation of UV High-Velocity Interstellar Lines in the 
           Carina Nebula
   7126   Early Release Observations of Shocked Interstellar Gas in the Vela 
           Supernova Remnant
   7138   Imaging and spectroscopy of the bright arcs around the most luminous 
           X-ray cluster RXJ 1347.5-1145

Target Acquisition Accuracy - On-board Results

As mentioned in the previous STAN, a flight software update to the
target acquisition routines was uploaded in August.  A set of
observations were executed to verify that the flight software updates
corrected the following problems:
1) truncation of the NSSC-1 commanded offsets during ACQ/PEAKs
2) the "bright" column on the edge of the acquisition subarray was no
   longer included in the acquisition analysis
3) the minimum flux value in an ACQ/PEAK should be subtracted from all
   samples to eliminate the bright background problem
The observations were also to verify that the FSW upgrades did not
have any negative impact on target acquisition.
The results of these tests were very positive, and are as follows:
1) Point source ACQs - there were 3 point sources acquisitions (on
both a bright and faint target), and all were successful.  The
accuracy quoted in ISR 97-12 (0.5 pixels) is a good (conservative)

2) Diffuse source ACQs - there were 2 diffuse source acquisitions, one
with the recommended flux level in the target, and one with a low
level which had previously failed (the "bright" edge column was
acquired).  With the FSW upgrade, both acquisitions succeeded, and the
accuracy is consistent with that quoted in ISR 97-12 (0.2-2 pixels,
depending on object size).
3) Peakup acquisitions - there were several ACQ/PEAKs, both imaging
and in dispersed light.  The minimum flux value was subtracted for all
points, and the accuracy of the ACQ/PEAKs are consistent with the
value quoted in ISR 97-12 (5% of the slit width).  A spiral peakup was
obtained and was successful, showing that the NSSC-1 truncated problem
has been resolved.

50CCD Performance
We have recently checked the science statistics for STIS in Cycle 7 and 
found out that GO imaging with the STIS CCD are using the F28X50LP
(long pass) filter more than twice as often as the 50CCD (clear) mode. 
While this difference in usage is certainly partly due to scientific reasons,
it is also possible that part of the GO community has been concerned by the
discussion in the STIS Instrument Handbook, Version 1.0 (see page 91) which
stated that over-illumination by UV light could cause a large elevation in
residual dark current with a long decay time (which was based on laboratory
testing of STIS CCDs).  As noted in the Cycle 7 STIS CCD Phase II update (under
the `Documentation' link on the STIS home page), later ground testing of the
actual STIS CCD showed this over-illumination effect to be much less severe
than previously suspected, prompting us to remove any restriction against the
use of the 50CCD imaging mode for parallels, coronagraphy, etc. In fact, the
50CCD imaging mode has proven to be a very powerful one indeed: E.g., a
two-orbit integration reached a limit within one magnitude as deep as the
deepest single-filter image of the Hubble Deep Field (cf. the recent article by
A. Fruchter in the STScI Newsletter of July 1997 (vol. 14, no. 3, p. 3)). 
GOs who want to change their imaging exposures from F28X50LP to 50CCD can 
still do this as part of the Phase B review process which starts 10 to 6 weeks 
prior to the opening of the first plan window in a GO program, when you will 
receive a message from your Contact Scientist.  

Insertion of CCD Fringe Flats
As noted in the previous STIS STAN, spectroscopic observations longward of
~7000 Angstroms are affected by fringing resulting from interference between
the incident beam and light reflected from the substrate of the back-side
illuminated device.  The peak-to-peak amplitudes reach 25% and 32% at 9800 A
for the G750L and G750M gratings, respectively.  Although the behavior is
stable, small shifts in the wavelength mapping caused by non-repeatability of
the grating-wheel and thermal drifts result in pattern shifts between 
successive exposures.  Therefore, contemporaneous spectral flats using the
tungsten lamps are necessary.  Manual procedures have been developed for
creation of the required exposure lines which need to be incorporated into 
the relevant programs.  CSs and PCs working your program will identify 
programs requiring the fringe flats and insert them as necessary during the 
Phase B preparation.  Later this year the capability to create the necessary
fringe flats will be incorporated into the RPS2 software interface allowing
observers to modify their own programs during the Phase B process.  In the
interim, however, you should keep in mind the need for the created flats when
reviewing your program at Phase B.

Changes to Limiting Magnitudes for MAMA Bright Object Protection 

Three updates and errors to Table 4 and 5 of STIS ISR 96-024 were found
since the publication of this ISR.
1) The limiting magnitudes of OB stars for the first order gratings are too
   bright by about a magnitude. This error was introduced by a switch from the
   local to the global rate limited case when going to earlier spectral types.
   This does not affect the count rates predicted by the exposure time
   calculator, which is independent of the magnitude tables. Therefore, few
   observers should be affected by the revision of the OB-star magnitudes in
   Table 4.
2) The NUV narrow-band filter profiles used for the magnitude calculation did
   not take into account the red leaks and the reduced throughputs at center 
   wavelengths. The updated filter profiles lead to changes of the limiting 
   magnitudes by a few tenths of a magnitude for early-type stars. The effect
   is larger for late spectral types. The current version of the exposure
   time calculator uses the most recent filter profiles.
3) Inflight measurements of the echelle grating throughputs indicate higher 
   than expected count rates.  The observed rates are about factors of 2
   and 1.5 higher than the pre-launch values for E140H/M and E230H/M, 
   respectively.  These changes are due to three components; first, the inflight
   realized throughputs seem to be slightly higher than expected; second, some
   fraction of the global flux is scattered outside the extraction apertures
   for point sources (this was not taken into account in the original ground
   calibrations); lastly, counts from the order outside of the nominal free 
   spectral range fall on the edges of the detector and boost the global count 
   rates.  Neither the magnitude table nor the current exposure time
   calculator account for this.  Since the screening limits are unchanged, the
   bright object limits for stars observed with the echelle gratings will
   become fainter by the corresponding factor.  As soon as we have reached a 
   full understanding of the increased count rates, we will contact those 
   observers individually whose programs require revision.  Updates will be 
   posted on the web as soon as they become available.
A revised version of STIS ISR 96-024 will be published by the end of 1997. This
version will address the above points.

Target Acq ETC On The Web (NEW) And Updates To The Imaging ETC

The number of pixels used in the imaging-ETC to encircle 80% of the flux 
was recently updated to the on-orbit measured values.  These numbers are 
used for S/N calculations.  The new values are 5x5 pixels for the CCD, 
16x16 pixels for the NUV-MAMA and 24x24 pixels for the FUV-MAMA.
Since the target acq software uses a 3x3 checkbox for point sources, the 
regular ETC cannot be used for target-acq purposes.  To facilitate exposure 
time calculations for target acquisitions, a separate ETC has been developed.
You may access the new Target Acq ETC from the existing STIS "Proposal
Resources" web page.

NEW - On-Orbit Performance Pages on WWW

One of the things we have recently done, is to provide an On-Orbit Performance
page for STIS on our WWW pages.  This includes information and graphics
documenting on-board measurements of sensitivity for all the low-dispersion
modes (as well as a few medium dispersion modes), target acquisition
accuracies, CCD and MAMA detector performance data including noise, flat
fielding, geometric distortion and resolution characteristics and on-orbit PSF
analyses.  STIS thermal stability, coronagraphic performance and line spread
function data are also discussed.  This provides you with a quick look at the
to-date measured performance of STIS as we work to process the data and
information into reference files and Instrument Science Reports and
enhancements to the calstis pipeline - all of which take time.  Several
`foibles' with STIS are also noted in the On-Orbit Performance pages.  Included
among these are the fringing for CCD spectroscopic observations at wavelengths
longward of 750nm, CCD long-wavelength detector halos, hot pixels, large-scale
pattern-noise for the CCD with GAIN=4, dust motes (shadows of specks of dust on
the CCD faceplate), high NUV-MAMA dark current and spectroscopic-mode ghosts. 
Some of these items have also been discussed in depth in recent STIS STANs. 
Continuing in November and on, we will be posting more in-depth information on
the STIS web pages describing on-orbit performance accuracies to date.  Also
due in November will be the release of the new Data Handbook providing
information and instructions on the analysis of STIS data. 

Pipeline Calibration News

1. Recent changes to CALSTIS

As we have learned more about the behavior of STIS, we have evolved the
calibration software, CALSTIS, to reflect our improved understanding.  We also
find it necessary to repair defects (or "bugs") in the code--this happens
infrequently, of course. ;) 

Below is a table of the most recent versions of CALSTIS. Most data taken this
summer were processed by the operational STScI pipeline (OPUS) using version
1.3. Recently we have upgraded the pipeline to version 1.5. You can tell which
version of CALSTIS was used to process your data by reading the value of the
CAL_VER keyword in any _raw FITS file. 

        ----------------	-----------------	-------------------
	1.3 (02-July-1997)	16 Jul 97		N/A
	1.4 (18-August-1997)	N/A			5 Sep 97
	1.5 (05-September-1997)	22 Sep 97		13 Oct 97
        1.5a (01-October-1997)  expected 13 Oct 97      13 Oct 97

IMPORTANT ADVISORY:   All GOs should download version 1.5a of calstis before
reprocessing their data or working with the calstis tasks. Version 1.5a of
calstis was not incorporated in IRAF 2.11/STSDAS2.0 as the changes were made
after that release.  This latest version of calstis should be available by the
second week in October and can be downloaded at that time from the STSDAS page
on the STScI WWW; see CALSTIS 1.5a and below for more information. 

The changes implemented in each version are described below.

TABLES and STSDAS, versions 2.0. have been released on Sep. 5, 1997. This
version includes new code for accessing NICMOS and STIS datasets, as well as
new tools for manipulating the data. Note that IRAF 2.11 is required to run
TABLES/STSDAS 2.0. The latest information on the release is available off the
STScI home page by linking to STSDAS via the Science Resources and Software
pages. The CALSTIS 1.4 (18-August-1997) version is included with the
TABLES/STSDAS 2.0 and is the first publicaly available version of CALSTIS. 

In particular, several problems have been uncovered that have been corrected in
Version 1.4: 

   - CALSTIS was scaling the superdark prior to dark subtraction 
     by multiplying it by the 'dark time' and the total exposure time. 
     However, the 'dark time' was already taken into account in the pipeline 
     superbias; therefore, CALSTIS was applying the dark time correction 
     twice. CALSTIS 1.4  has been modified to scale the superdark prior to
     dark subtraction with the total exposure time only.
   - corrections have been made to the sections of the CALSTIS code that 
     perform 2-D rectification and 1-D spectral extraction. These include 
     ignoring the value of the keyword SDQFLAGS when performing background
     and spectrum extraction to include all the input pixel values in the
     calculation; checking the X2DCORR calibration switch before determining 
     the MSM offset to prevent accidentally calculate the offsets on a 
     flat-fielded file instead of the 2-D rectified file; correction to the 
     2-D rectification code that was erroneously multiplying the absolute flux
     values by the aperture throughput from the APERTAB table for the extended
     source case; finding the shift in the dispersion direction by convolving
     the template spectrum with a rectangle of the same width as the slit used
     for the WAVECAL observation; increasing the search range when finding
     the spatial shifts from the ends of a long slit or from the edges of 
     occulting bars. All these corrections are included in CALSTIS 1.4.
   - The reference files used to compute dispersion solutions (the *_dsp.fits
     files) were offset by 1 pixel giving incorrect MSM offsets from the 
     WAVECAL in the dispersion direction. New reference files correcting this 
     problem were delivered on Aug. 25. NOTE HOWEVER THAT DATA CALIBRATED 
     *_x1d.fits files).

   - The reference file used to construct rectified, linearized 2-D spectra
     (the *_sdc.fits file) for the fuv-mama has been updated. New measurements
     of the G140L, G140M midline position in the spatial direction were 
     calculated and the new reference file was delivered on Aug. 25. 

CALSTIS 1.5 and CALSTIS 1.5a
Version 1.5 of CALSTIS is currently running in the operational pipeline and the
STSDAS version will be available on or about 6 October via the STSDAS web page.

Changes that have been made to CALSTIS in version 1.5 include:

    - When doing geometric correction for imaging mode (GEOCORR), X2D
      no longer attempts to conserve flux.  This is now consistent with
      spectroscopic mode.  In addition, flat fields are made with the
      assumption that flux conservation will not be included during geometric
    - When converting count rate to flux for CCD data, X1d and X2d
      have been changed to multiply by the value of the primary header 
      keyword ATODGAIN.
    - X2D now takes into account that the "plate scale" may differ
      between AXIS1 and AXIS2 in the case of binned data. The plate 
      scale is needed for converting aperture offsets from arcseconds 
      to pixels.
    - Three new keywords have been added to the headers and are populated
      by X2D. These keywords are needed to define fully the intensity units 
      for targets that are continuum vs. emission line sources, and whether 
      or not they are spatially resolved.

The additional fix in CALSTIS version 1.5a includes:

    - CALSTIS was improperly subtracting the dark for CCD and MAMA data which 
      was binned.  BASIC2d (calstis-1) now properly subtracts the integration 
      time scaled _sum_ of the dark for the binned pixels, where it was 
      previously subtracting the average of the dark for the binned pixels.

2. Changes to Calibration Reference Files

New calibration reference files have been delivered recently to the pipeline
and are also accessible from the archive: 

   - In-flight CCD darks have now been delivered for the period 
     through 8/25/97.  Weekly darks will be routinely delivered to the 
     pipeline within at least 4 weeks of their having been taken. We
     are working to reduced this time lag to 2 weeks. 
     HEADS UP: Do not be surprised to see lots of "little spots"
     in your calibrated CCD data. If you crsplit your observations and
     are looking at the cosmic ray rejected data these are not likely to
     be cosmic rays; these are hot pixels!. The number of hot pixels you will
     see due to the fact that the dark reference file used is always a few 
     weeks out old. To remove the "spots" you will have to recalibrate when
     an appropriate dark becomes available. Stay tuned to the STIS
     web page for updates to the reference files. Generally you can expect
     new darks within two to three weeks.

   - Corrected versions of the dispersion coefficients tables (the *_dsp.fits
     files) and the 2-D spectrum distortion correction table (*_sdc.fits 
     files) were delivered on Aug. 25 to include the corrections specified

3. Current State of Calibration

The following table summarizes how well the current set of reference files and
CALSTIS are doing in terms of calibration accuracy. We expect to continually
update the accuracy levels as time goes on and the current accuracy levels will
always be posted under the calibration button of the STIS Instrument WWW page. 
The accuracies given below are at slit center. 

    Physical Quantity     Accuracy
    -----------------     --------
  Relative Wavelength     0.2-1.0 pixels (based on the dispersion solution, 
                                          quite mode dependent, see the STIS 
                                          Data Handbook for more details)
  Zero-point Wavelength
       (w/WAVECAL)        0.1-0.2 pixels (generally achieved; worse results
                                          can obtain for some slit+central 
                                          wavelength combinations) 
  Zero-point Wavelength
       (w/o WAVECAL)      <= 8 pixels    (due to MSM repeatability and
                                          thermal drifts)

  Flux                    +/- 50% (this is quite mode and aperture dependent)

Please note:  the flux calibrations will improve when we update the reference
files based on the latest sensitivity calibrations. We expect to install these
reference files at or about the turn of the year.  The goal is to attain
roughly +/- 10%-15% accuracy for the absolute flux at slit center for large
apertures observations for the majority of the modes; some of the first order M
modes and the intermediate settings are likely still to have slightly worse
accuracies at that time and the first order L modes are expected to have
considerably higher accuracies. 

Cycle-7 Calibration Plan

The Cycle 7 Calibration Program has been progressing well.  Listed below are
the programs which have partially or wholely executed to date. 
  7601      CCD Dark and Bias Monitor
  7604      Cycle 7 MAMA Dark Measurements
  7605      STIS CCD Target Acquisition Workout
  7634      CCD Flat-Field Stability and Cosmetic Defect Fraction
  7635      STIS CCD Hot Pixel Annealing
  7636      STIS Spectroscopic CCD Flatfielding
  7641      CCD External Flats, Stellar
  7642      CCD Red Light PSF halo
  7644      STIS NUV-MAMA Monitoring Flats
  7648      STIS CCD G230LB and G230MB Wavelength Calibrations
  7659      Daily Darks to Update Acquisition Bad Pixel Table
  7665      CCD Geometric Distortion
  7667      MAMA Geometric Distortion
  7672      STIS CCD Sensitivity Monitor
  7673      STIS MAMA Sensitivity and Focus Monitor
  7674      IR Standards
  7720      MAMA Full Field Sensitivity Monitor & PSF Check
  7722      Lamp Flux Measurements
  7728      STIS FUV-MAMA Monitoring Flats
  7600      STIS CCD Performance Monitor

We remind you that all Calibration data is immediately non-proprietary.  We 
ask you to be patient as we work to analyze these data and extract the relevant
information and produce reference files for the pipeline.  That will take time. 

Calibration Workshop
  The Calibration Workshop describing the current state of the new and
existing instruments was held over three days this month, Sep 22-24. 
Closeout reports for the GHRS and FOS, removed from the spacecraft during the
second servicing mission in February, were also presented.  Most of the 
information provided at the workshop for STIS has, or will be, discussed in
STAN issues, and will be available later in the proceedings. 

What's New On The STIS WWW Pages
New on the STIS web pages, under "Calibration Resources", is the On-Orbit
Performance page.  Summaries of STIS performance include sensitivity, detector
performance, PSF's and LSF's, target acquisitions, throughputs, thermal
stability and more.  Also, there are descriptions of the 'foibles' found so
far: hot pixels, detector noise, fringing and ghosts. 

Additionally, a new Exposure Time Calculator for Target Acquisitions is 
available on the "Proposal Resources" page.

And see the DRAFT version of the STIS Data Handbook; the complete HST Data
Handbook Version 3.0 is expected out in November. 

                           NEW STIS DOCUMENTATION
All new STIS related documentation noted here, along with previous items, may
be found on the "Documentation" link off the STIS home page.  New STIS
advisories may be found on the STIS home page Advisories link. 

STIS FAQs (Frequently Asked Questions)...

Phase B Reviews:
   What happens during the Science Optimization (Phase B) review?

Other New Documentation...
(Check `What's New' on the STIS WWW Page)

STIS Target Acquisition Exposure Time Calculator:
   [see description above]

STIS On-Orbit Performance:
   [see description above]

Draft Data Handbook (STIS Chapters) On-line:
   An initial **DRAFT** version of the chapters of the Data Handbook relevant
   to STIS is now available (see the "Documentation" link on the STIS home 
   page).  This is very much a draft, and you will note the absence of cross 
   references, etc.  However, we still expect this to be of use to you working 
   with your STIS data, and very much want the opportunity for you to use the
   STIS Handbook and send us back your comments/criticisms/suggestions for 
   improving it.  We expect the official version to be available November 1, 
   1997.  Please send your comments and questions to us at
   with the subject "STIS Data Handbook" in your email line.

                          APPENDIX: STIS Contacts
Any questions about the scheduling of your observations should be
addressed to your Program coordinator. Post-Observation questions can
be addressed to your Contact Scientist. If you do not know who these
persons are, you can find the information on the WWW at
Analysis, STSDAS or any other HST-related questions can also be
addressed to
To subscribe or unsubscribe send a message to with
the Subject: line blank and the following in the body:
                     [un]subscribe stis_news YOUR NAME
Comments, questions, suggestions, etc. can be e-mailed to
The Space Telescope Science Institute is operated by the Association of
Universities for Research in Astronomy, Inc., under NASA contract