Hubble Space Telescope
WFPC2 History

This file provides a chronological history of the WFPC2, including:

It is designed to provide observers with information about the state of the WFPC2 when their observations were taken. The most important change occurred on APRIL 23, 1994, when the operating temperature was changed from -76° C to -88° C in order to minimize the CTE (Charge Transfer Efficiency) problem and the growth of hot pixels. Care should be taken not to mix reference files from one side of the temperature change with data from the other side.

Each item includes a short description of how this might affect the observer. More detailed information is available in:

• WFPC2 Instrument Handbook.
• Links to Documentation, including status reports, workshop proceedings, and more.
• WFPC2 Instrument Science Reports
• WFPC2 Reference Files
• HST Data Handbook
• WFPC2 Decontamination Dates

This history file contains sections sorted by item type, then a table sorted by date.

1. DECONTAMINATIONS

The WFPC2 has been warmed approximately once a month (every 49 days since Cycle 11) to about +20° C for 6 - 12 hours to remove any buildup of contaminants. This appears to:

1) Restore essentially the full UV sensitivity of the instrument.
   Recent photometric monitoring results are kept updated in the
   WWW memo and associated figure.

2) Repair many of the "hot pixels" (see Figure 2 of the
   Holtzman et al. article).  

The calibration programs include regular photometric monitoring of both a standard star (usually GRW+70D5824; alternate is Feige 110) and a standard cluster (Omega Cen = NGC 5139; alternate is NGC 6752), usually done before and after decontamination procedures. Darks are taken at the rate of at least 5 per week (and 5 more before and after decons). More details on the calibration proposals are available in the WFPC2 Instrument Handbook.

From the observer's standpoint the decon dates may be useful in order to:

  Make small adjustments to the photometric zeropoints 
  in the UV. Enhancements to SYNPHOT have been made 
  to aid the user in taking these effects into account.
  Please refer to the ISR98-03: WFPC2 Longterm Photometric Stability and the 
  SYNPHOT section of the WFPC2 Photometry page.

The following link provides a list of the decontamination dates:

DATES OF DECONTAMINATION PROCEDURES.

2. DARKS

Typically, OTFR will have used the best available pipeline dark for calibration; OTFR can be accessed via the HST Archive or STARVIEW. To assess the pedigree of the dark, see the WFPC2 Reference Files.

Pipeline darks are normally available for each 5-10 day period. Prior to Aug. 01, 1996, the dark reference files were created from sets of 10 dark frames, 1800 sec. each, (taken over ~2 weeks time). After to Aug. 01, 1996, the darks have been generated by adding hotpixels (as identified from ~5 dark frames, 1800 sec. each, taken within 2 days) to a superdark generated from more than 100 darkframes. Please refer to the Pipeline Section in this document, or to our online Summary of Darks for more details, or to the HISTORY keywords in the headers of the dark reference files themselves.

To improve upon the pipeline dark calibration, one option is to recalibrate manually using a custom dark reference file (generated from, e.g., the shorter 1000 sec. darks that are taken daily and archived only - not used in pipeline). Scripts and cookbook examples of dark generation are available.

Another option is to use a cosmicray rejection routine that can identify the hot pixels (e.g., COSMICRAYS in IRAF), but care should be taken to set the parameters conservatively or you may also clip out the peaks of bright stars. If the data has been dithered, the Drizzle software may be used to remove both cosmic rays and hot pixels.

NOTE: At T = -76° C (i.e., before April 23) the mean dark count was about 2 DN per 1000 seconds. At T = -88° C (i.e., since April 23) the dark count has been about 0.5 DN per 1000 seconds. For an analysis of the dark current versus time, see ISR 01-05.

There are lists of hotpixels available via WWW. Note that there are several flavors: those for use with the STSDAS task "warmpix", plain hot pixel lists, and HDF hotpixel lists.These may also be retrieved in the form of compressed ASCII files via anonymous ftp:

ftp stsci.edu (username) anonymous (password)  
cd /instrument_news/WFPC2/Wfpc2_warmpix
binary 
prompt 
mget filename (include wildcards to select files for the period of interest)
quit

NOTE: As an additional aid in hot pixel identification, as of July 1997, more frequent dark frames are being taken. The new Supplemental Darks program (7621, 7712, 7713) is designed to provide more frequent, short (1000s) darks to be used primarily for the identification of hot pixels. Shorter darks are used so that observations can fit into almost any occultation period, making automatic scheduling feasible. The program is designed to complement the Standard Darks program (7620), whose longer individual observations are better suited to produce high-quality pipeline darks and superdarks. This program has become possible after the Second Servicing Mission thanks to the larger storage capacity of the Solid State Recorder. Currently, about 15 darks are taken each week as part of this program. The darks will not be used in standard STScI products, but are available to observers via the Hubble Data Archive.

We also have in the archive superdarks generated from 100 individual dark frames taken from April 23 to October 01, 1994 (NOTE: standard pipeline darks are also superdarks as of Aug. 1996; see previous section for discussion). These 1994 superdarks were not used in the pipeline but may be useful for recalibrating your data in cases where the total exposure time is relatively long (e.g., > 30 minutes) and the background noise is a limiting factor in your science (e.g., faint object detection).

        DELIVERED
1994    November  (for data taken after April 23 decon)

NOTE 1: the WFPC2 IDT has delivered alternate superdarks, superbiases, and deltadarks. The IDT Reference Files memo describes how these files were generated, the names for retrieval from starview, and how to use them.

NOTE 2: The Dec. 1995 Hubble Deep Field images were processed by the HDF team using the standard HST pipeline (the STSDAS task calwp2) but with different reference files (similar to the IDT superdarks, superbiases, and deltadarks) and a slightly different treatment of the darktime. These alternate reference files are also available from the archive; the HDF Pipeline Reduction page provides some details as well as the filenames of these products. The history records in the reference file headers provide more details as well.

3. CHANGE IN FOCUS

In general, the focus changes have been very minor and are unlikely to have an effect on your science; a short summary table follows. For more details, please refer to the online HST Focus Pages.

2000  Jun. 15    Secondary mirror moved +3.6 microns. (Day 167 19:19)

2000  Jan. 09    Secondary mirror moved +4.2 microns.

1999  Sep. 15    Secondary mirror moved +3 microns.

1998  Jun. 28    Secondary mirror moved -15.2 microns back at end of campaign.

1998  Jun. 04    Secondary mirror moved +16.6 microns for second NIC3
                 campaign.

1998  Feb. 01    Secondary mirror moved -18.6, at end of NIC3 campaign.

1998  Jan. 12    Secondary mirror moved +21, for NIC3 campaign.

1997  Mar. 18    Secondary mirror moved towards primary by 2.4 microns
                 (077/22:51).

1996  Oct. 30    Secondary moved outwards by 5 microns (day 304,
                 1700 UT). The average focus position is now ~+2.5 from
                 nominal, compared to -2.5 prior to the move.

1996  Mar. 14    Secondary moved +6 microns to correct for
                 desorption (day 074/1848Z).

1995  Aug. 28    Secondary moved by ~6.5 microns to compensate for
                 OTA desorption.  Before the move, WFPC2 ~ 3.5 microns
                 inside optimal focus; move put WFPC2 about
                 3 microns outside "best focus."  Most observers
                 should be unaffected by move.

1995  Feb. 23    Possible anomaly; if your data was taken between
    - Mar. 03    these dates, and the scientific results might be
                 sensitive to the PSF shape, please contact us.
                 See Note 1, below.

1995  Jan. 15    5 micron change to adjust for desorption.

1994  Jun. 29    5 micron change to adjust for desorption.

1994  Mar. 04    Minor change.

1993  Dec. 28

Limiting factors: Breathing worst case ~ 5-10 microns during a single 
                  orbit. Desorption ~ 1 micron/month during cycle 4.

NOTE 1: Faint Object Camera observations taken on Day 58 (Feb. 27) appear blurred. A possible cause of the blurring is a defocus and/or decenter of the secondary mirror, potentially related to the unusual thermal environment of the telescope - which had been pointed very close to the anti-sun direction for several days (ie., most of the telescope structure is in the shadow caused by the wider instrument compartment at the "bottom" of the telescope). WFPC2 observations taken shortly before and during the blurred FOC observations have been inspected for degradation of the PSF. These observations contain very few well-exposed stars in the PC, and therefore our assessment of the PSF properties in WFPC2 is necessarily limited. However, it appears that any degradation in the light distribution is very minor, and the FWHM of the observed stars is close to normal - the measured increase of less than 10% (to 1.8-1.9 PC pixels in F814W) is within the errors of the measurement. The PSF monitoring carried out on Mar. 04 shows nominal results.

However, the _shape_ of the PSF in the WFPC2 observations may be consistent with a defocusing of the secondary mirror of order of 10 micron, the value suggested by the FOC data. Such a defocusing could affect photometry based on small apertures or PSF fitting at the 5-10% level; scientific programs that depend on accurate knowledge of the PSF itself, such as observations of faint objects in the vicinity of bright point sources, would also need to take the possible defocusing into account. On the other hand, morphology and surface photometry of extended sources would be largely unaffected. Information about this event is detailed in the March 1995 STAN.

More general information about the OTA and focus changes may be obtained from the Observatory Group pages.

4. CHANGE IN TEMPERATURE

The operating temperature was changed on >>> APRIL 23, 1994 <<< to reduce the CTE problem and the number of hot pixels (see the Holtzman et. al., paper for details). This date represents a major change in the calibration of the WFPC2, since many instrument characteristics were affected by this temperature change (e.g., sensitivity, dark current, hot pixels, bias level, etc.).

1994    December 26     -76° C
        April 23        -88° C

5. CHANGES TO PIPELINE

2001 Sep.
OPUS pipeline software has been updated to fix a bug in the computation of statistics of raw (d0*) data as well as populate the FILTROT keyword (the rotation angle for partially rotated filters such as the ramps or polarizers).

2001 July 18
Alternate WFPC2 flatfields, with reduced noise, are available for manual recalibrations. Note that these are NOT in the pipeline as of Sep 2001.

2001 May 16
OTFR (On-The-Fly Reprocessing) is publically released. OTFR can be accessed via the HST Archive or STARVIEW. NOTE: uchcoord no longer needs to be run on OTFR data; for details as well as a comparison of OTFC and OTFR, see OTFR page.

2000 Fall
A minor bug affecting only short INTFLATs has been fixed in the OPUS pipeline (pre-calwp2) software. Archival exposures will be fixed automatically by OTFR (On-The-Fly Reprocessing), planned for release in early 2001. The problem was that INTFLATs with non-integer exposure times have incorrect integer exposure times in the headers. More details and a full listing are given in this problem report.

2000 Spring
Minor bug in calwp2 affects bias level calculation in AREA mode images only. Typically, the errors in the bias values are ~0.01 DN, though for a very small number of images (~20, those with large numbers of saturated pixels), the error can be ~1 DN or larger. Note however, that there are very few affected exposures: out of a total of nearly 90,000 WFPC2 images in the archive, only ~150 are AREA modes. All Cycle 7 and Cycle 8 GOs using AREA mode have been notified and CALWP2 will of course be updated as soon as possible. In the meantime, for assistance with any AREA mode images, please contact help@stsci.edu.

1999 December
OTFC (On-The-Fly Calibration) is publically released.

1999 Apr. - Jul.
Single chip data was processed as PC, regardless of the actual chip that was read out: generic conversion (the step preceding calwp2) was erroneously setting detector to 1 for all images with less then 4 groups. A software fix was installed on July 08, 1999. This mode is rarely used; no GO images were affected, only WFPC2 calibration data were corrupted (17 images total).

1999 Apr.
A new version of the OPUS software was installed in the pipeline Apr. 05, 1999, as part of switching the OPUS platform from VMS/VAX to ALPHA/VMS. The data processing steps remain the same however, there have been some minor keyword changes. The changes should be transparent to the majority of WFPC2 users, further details may be found in the software update summary.

1998 Dec.
Update to SYNPHOT contamination tables. Note however, that the pipeline does not automatically correct for contamination (e.g., the PHOTFLAM keyword). The correction must be applied manually, using either the information in Tables 2 and 4 of ISR98-03: WFPC2 Longterm Photometric Stability or via SYNPHOT (see section 4 of ISR 98-03).

1998 Sep.
Excessive noise in the F160BW flat; for alternate flatfield, see Aug. 1998 STAN or WFPC2 Reference Files memo.

1998 Jan.
STARVIEW and bestref now provide most recent GRAPHTAB and COMPTAB names, for use by calwp2 in computing the PHOTFLAM and other photometry-related keywords in the calibrated header files. (See also Note 2.)

1997 Aug. 15
A rare problem in the population of the PA_V3 keyword has been found, however, only 50 observations are affected. For more details, including a list of affected images and a fix for those images, please refer to our online PA_V3 Memo.

1997 May 16
SYNPHOT tables updated. With a few exceptions, changes to non-UV filter modes were relatively minor: generally ~1-2% or less. F785LP, F850LP, and F1042M were more: ~1-4%, ~1-6%, and ~2-15% for them, respectively, depending on the chip. The UV filters required somewhat larger changes to bring SYNPHOT into agreement with the observations, ranging from ~2% (e.g., F255W and F300W) to ~5% (F170W, F218W) to 10% or more (F160BW, F375N). Any filters not updated at this time will be checked with future observations (planned for summer of 1997) and updated if necessary. An ISR is expected out end of June. A table summarizing the photflam changes and new zero-points, as well as the README file (guide to WFPC2 SYNPHOT tables) and ASCII versions of the tables, can all be found in the Photometry and SYNPHOT Page. See also the May 1997 STAN and ISR97-10: WFPC2 Synphot Update and Note 2 below.

1997 Jan. 03
Calwp2 update (1.3.5.2): fix problem with processing of images containing less than 4 groups. Since Dec 1994 (1.3.0.7), observations that read out less than 4 chips AND whose first chip was not PC have been incorrectly calibrated as if all chips were PC. The problem has been identified, current GOs notified, the code fixed and installed. The problem did not affect 4-group images.

1996 Oct.
New flatfields for polarizers installed in pipeline. Filters include polarizers (i.e., polq, polqn33, polqn18, or polqp15) used with one of the following:

f300w, f336w, f390n, f410m, f439w, f547m, f555w, f606w, f656n, f675w

1996 Aug. 15
Fix for WFPC2 images taken as a coordinated parallel with FOS (only). From Dec 95 to Aug 22,1996, WFPC2 images taken in parallel with the FOS were inadvertantly assigned FOS instead of WFPC pointing information. Note: these images have been fixed via reprocessing and rearchived as of May 1998; for a historical record of the problem, see online WFPC2 parallels fix memo.

1996 Aug. 09
New version of calwp2 (1.3.5.1; no change to data). Software now prints message to trailer file, header, and screen, if bias jump of greater than 0.09DN is found (jumps greater than 0.5DN have these messages as well as a WARNING printed to the PDQ files). For more details, please refer to the ISR 97-04: Properties of WFPC2 Bias Frames.

1996 ~Aug. 01
Darks are now generated in a new fashion; most importantly, more pixels are being flagged in the data quality files. Please refer to the short Summary of WFPC2 darks for details; an Instrument Science Report is being prepared. A Summary of WFPC2 Darks is also available.

1996 July
SYNPHOT tables updated for FQCH4N* and FQUVN*. Affects only filter curves for 6193A and 8929A methane quad filters and the UV quad filters. Adjustments consisted of 3-15% decreases in filter throughputs. More details are available in the June 1996 STAN as well as the WFPC2 SYNPHOT Readme File.

1996 Jun./July
Improved Cycle 4 flatfields installed in pipeline. Errors at edges now < 0.5% rms. Filters include all those redward of f300w:

f300w, f336w, f343n, f375n, f380w, f390n, f410m, f437n, f439w,
f450w, f467m, f469n, f487n, f502n, f547m, f555w, f569w, f588n,
f606w, f622w, f631n, f656n, f658n, f673n, f675w, f702w, f785lp,
f791w, f814w, f850lp, f953n, f1042m

1996 Apr. 18
WFPC2 headers from dithered data updated (dithered using scans, dither-line, or dither-box). Previously, due to the relatively small number of scans being performed, the pipeline provided only minimal support for WFPC2 dwell scans: only the initial dwell point position was used for the target pointing, so each exposure had the same CRVAL data, resulting in slightly incorrect metric results on all but the initial scan image. Note: the fix was only necessary for scan data; manually dithered images, e.g. using the POS-TARG option, had headers which properly reflected the pointing.

1995 Dec.
New flats for f300w, f375n, f450w, f555w, f606w, and f814w were installed in the pipeline. These are similar to the Cycle 4 flatfields but have been improved at the CCD edges: errors are now estimated to be < 1% rms. Filenames are given in the WFPC2 Reffiles Memo or will be found by Starview's 'bestref' screen.

1995 Nov.
Minor update to calwp2 (V.1.3.0.8): now prints out reference file names to go along with PEDIGREE/DESCRIP. Does not affect data; produces new informational messages during pipeline processing.

1995 Aug. 10
New maskfile (r0h) used in pipeline; columns above traps now flagged with "256" in c1h files. See ISR 95-03 for details.

1995 July
WFPC2 related SYNPHOT tables updated. See Note 2, and the July 1995 STAN.

1995 March 24
There are 184 WFPC2 images, dating from Apr. '94 - Mar. '95, which have corrupted FILT* header keywords. Only those images taken with rotated filters (e.g., fqch4n, polqn33, polqp15, fquvn33, fr418n33, etc.) were affected. There were 2 sorts of problems: 1) when rotated filters were taken with a second (non-rotated) filter, the second name was erased from the header, and 2), when a rotated filter was used in isolation, OPUS would erroneously add some second filtername to the header. For more information, see the Rotated Filter Problem memo.

1995 March
Observations taken 4/15/94 to 3/27/95, using WF3-FIX (only) may have incorrect pointing information in their image headers. In particular, errors on the order of ~2" may be present. Note that this only impacts 2 values in the header (although it can cause derived positions to be in error); there are no problems with the image data or with the telescope pointing. Note: these images have been fixed via reprocessing and rearchived as of May 1998; for a historical record of the problem, see online WF3-FIX Memo.

1994 Dec.
Minor update to 'calwp2' (V.1.3.0.7): now checks the PEDIGREE and DESCRIP keywords within reference files. Does not affect data; produces new informational messages during pipeline processing.

1994 May 04
Switch to separate values for even and odd columns. (Version 1.3.0.6 of 'calwp2'; note that new biasfiles are required with the new 'calwp2'. See Note 1.)

1994 Apr. 14
Fix installed for parallel WFPC2 data headers. Note that some archival data may still be incorrect. (See Note 3.)

1994 March 08-09
Prelaunch flats were used until this date. These were generally good to a few percent over most of the field, with a maximum deviation of 10 %. See Holtzman et al., paper for details.

1994 March 08
Overscan region used to determine the blev correction changed from [3:14,*] to [9:14,10:790]. (Version 1.3.0.5 of 'calwp2'. See Note 1.)

NOTE 1. Change to Overscan Region used for BLEVCORR

Until March 08, all 12 columns of the overscan region were used in the pipeline to determine the bias level. On March 08, a change was made to use only 6 columns, since the first few columns are too high or even saturated, having been affected by the exposure level. This results in the removal of too much bias level; a quick check of the DEZERO group keyword will indicate the amount removed, normal levels are ~300 to 400 DN.

There is also a very slight difference in bias level for the even and odd columns, typically about 0.1 DN difference although in WF4 it can be as high as 0.4 DN. Until May 04, 1994, a single value was used in the pipeline; separate values have been used in the pipeline since that time (these values are stored in the BIASEVEN and BIASODD group keywords in the science data headers). Note: the trailer (trl) file includes the calwp2 version that was used at the time of processing.

NOTE 2. New SYNPHOT Tables Available for recomputing PHOTFLAMs

Observers using the header photometry keywords (i.e., PHOTFLAM) from data processed in the pipeline before May 1997 should recompute these keywords, using either SYNPHOT directly or calwp2 (after updated SYNPHOT tables are retrieved and installed at their local site), or the SYNPHOT PHOTFLAMs and zeropoints table.

To identify epoch of SYNPHOT tables used for your observations: check the GRAPHTAB and COMPTAB keywords in the science data headers (c0h). The graphtab/comptab names used in the pipeline are based on the time of installation; for the May 1997 update, the GRAPHTAB and COMPTAB names are h5* or later (h=year since 1980, 5 is month, d is day. If the GRAPHTAB and COMPTAB names are h5* or later, the photometry keywords are already uptodate. If they predate May 1997, then the keywords should be recomputed. For more details about changes, installing files, and recomputing the keywords, refer to the Guide to WFPC2 Synphot Tables.

NOTE 3. Incorrect Pointing Information in some WFPC1 and WFPC-2 Parallel Data Headers;

Approximately 150 WFPC-1 and WFPC2 observations, taken in parallel with the FGS between Oct. 11, 1993 and Apr. 19, 1994, may have incorrect pointing information in their headers. Not all the data prior to the Apr 19 fix is affected, since workaround tools were developed after Mar 3 to allow PODPS operations personnel to fix the data but some observations were missed. The problem was related to the FGS observations, for which PODPS received the wrong roll orientation data. Some of these roll errors were significant and would affect the validity of the CD matrix. The images will be reprocessed and rearchived (expected completed by May 1996); if a list of affected observations is desired, please contact the STScI help desk at help@stsci.edu.

6. SAFINGS

Safings are usually not a major problem for WFPC2, which is placed in HOLD mode and does not warm up for most spacecraft safing situations. Even if the instrument should warm up and require a subsequent decontamination procedure, this is not an issue (as it was with WFPC 1) since there is no UV flood to preserve (although as with all decons, the UV calibration will change, see Section 1).

7. PLATE SCALE CHANGES

The IRAF/STSDAS task "uchcoord" may be used to examine the plate scale of c0h files and/or update the CD matrices with a new plate scale; please see the online help for more details on this task.

The accuracy of the platescales, as derived from the fit, is 0.00002"/pixel for PC1 and about 0.00004 for the WF chips.

                 PC1       WF2      WF3      WF4
Nov. 1993        0.046     0.101    0.101    0.101
Apr. 11, 1994    0.0452    0.0991   0.0991   0.0993
Mar. 27, 1995    0.04553   0.09958  0.09956  0.09962
Apr. 14, 1996    0.045518  0.099548 0.099527 0.099587

NOTE 1: most images taken prior to April 11, 1994 will yield positional measurements ~3" in error on average, although it may be as much as 6". A software tool is being developed to aid in correcting the image headers for the pointing offset (and rotation) but until it is available, if you have any questions concerning the pointing of images taken in early 1994, please contact help@stsci.edu. The rotation amounts to 0.8 deg for WF2 (smaller rotations for the other chips (0.28 deg in PC1, 0.46 deg in WF3, and 0.06 deg in WF4)). Since the STSDAS task METRIC uses WF2 as the reference chip, a 0.8 degree discrepancy will be present between data obtained before and after April 11, 1994, in addition to a shift.

NOTE 2: Images taken April 11-19, 1994 have positional information in the headers which does not correctly reflect the aperture updates. Affected GOs will be notified personally and an effort will be made to have the images reprocessed and re-archived. In the meantime, if you are working with images from the April 11-19, 1994 timeperiod, and have any questions concerning the pointing of images, please contact help@stsci.edu.

8. LIST OF WFPC2 OBSERVATIONS

9. MISCELLANEOUS

- scale noise can be set separately for PC and WF
- handles images with different exposure times
- will create and/or update masks with cosmic rays flagged

Note: this section 9 (Miscellaneous) has not been updated after Sep. 1998, if updates would be of use to you, please let us know via an Email to help@stsci.edu.

10. OTFC Keyword Exception Table

The Keyword Exception Table is a listing of all observations in the HST Archive with incorrect keyword values, such as wrong filter number, shutter value, etc. When OTFC processes a dataset that is on this list, OTFC will automatically use the correct keyword value. This table is intended as a quick way for users to check if WFPC2 datasets have been added/removed from the list.

11. WFPC2 History In Table Format

This table is a record of WFPC2 history in chronological order. See above for full details.

Last updated: Feb. 10, 2000.