STScI Logo

Hubble Space Telescope
WFC3 STAN - Issue 2, December 2009

WFC3 Space Telescope Analysis Newsletter - Issue 2, December 2009 ==============================================

1. CALWF3 and the STScI Pipeline

2. IR Subarray Image Artifacts

3. Revised Image Distortion Parameters for UVIS and IR Channels

4. Persistence in WFC3/IR

5. New Documentation - Instrument Handbook and ISRs

6. New WFC3 Reference Files

1. CALWF3 and the STScI Pipeline ==============================================

The latest version of CALWF3 (1.8.1) was released to the community in November 2009 as part of STSDAS v3.11. This version of CALWF3 contains a number of important updates, including corrections and enhancements to the IR ramp fitting process (the CRCORR calibration step in CALWF3), and a fix to the amplifier-dependent gain calibration issue that existed in the FLATCORR step in previous versions. For details about these changes, see the WFC3 Pipeline Release Notes webpage at Please note that a new version of the IR Cosmic-Ray Rejection Parameters table (CRREJTAB) must be used with calwf3 v1.8.1 in order for the IRCORR calibration step to work properly. The new reference table "tag1826ki_crr.fits" can be downloaded from the WFC3 Reference Files web page.

The latest version of MultiDrizzle (v3.3.6) was also released in November 2009 as part of the stsci_python package, which contains several bug fixes for correctly drizzling WFC3 images.

The STScI OTFR pipeline, which is used to process new datasets as they are received from HST/WFC3 and to apply on-the-fly-reprocessing to datasets retrieved from the archive, is still currently running CALWF3 v1.6, without any MultiDrizzle processing. The latest version of CALWF3 is expected to go into use in OTFR sometime in January 2010. We also anticipate enabling MultiDrizzle processing of WFC3 images within OTFR at that time, at which point the empty "drz" files that users are currently receiving will be replaced by real MultiDrizzle products.

We are currently working on solutions to a couple of other known issues with CALWF3 processing. First, calibrated IR grism images contain residual gain calibration errors that show up as small (0.5-1.5%) offsets in the overall signal levels between the 4 image quadrants. Second, the error (ERR) array values in IR "flt" images are systematically low by factors of 3-5, thus underestimating the true uncertainties in the SCI array values. The "flt" file ERR array values are computed during the ramp fitting process, which is being reworked to fix this issue.

2. IR Subarray Image Artifacts ==============================================

Certain combinations of IR subarrays and sample sequences give rise to images containing an artifact. The artifact appears as a sudden jump in the overall background level of the image, with an amplitude of 3-5 DN. The jump occurs exactly at the vertical center of each quadrant of the image, such that the lower and upper quarters of the image have a different overall level than the middle of the image. So far we have seen the artifact in the following image modes:

samp_seq=SPARS10 with aperture=IRSUB256
samp_seq=SPARS25 with aperture=IRSUB512
samp_seq=STEP25 with aperture=IRSUB512

The cause of the artifact is under investigation. If an operational solution can not be found, it may be possible to develop techniques to remove it during calibration.

3. Revised Image Distortion Parameters for UVIS and IR Channels ==============================================

New versions of the UVIS and IR image distortion parameter tables (IDCTABs) will be released on the WFC3 main page, under "Late Breaking News" section. The new parameter values will produce better alignment of MultiDrizzle DRZ images than the versions that were released on the WFC3 website on September 9, 2009. We plan for the new IDCTAB files to be installed in CDBS in January 2010 and to be used thereafter by the OPUS pipeline and OTFR processing to produce DRZ files.

The September 9th versions of the IDCTABs were based on initial analysis of SMOV data. We have found that the distortion in UVIS images is accurately modeled at 0.05 pixel r.m.s. and in IR images at 0.08 pixel r.m.s. Therefore that model will not be changed in the upcoming release. However, the absolute alignment of the WFC3 FOV relative to the HST V2/V3 reference system (translation and rotation) of the apertures will be updated.

The UVIS location shifts by (V2, V3) ~ (-0.6 arcsec, 2.9 arcsec)
and the IR location shifts by (V2, V3) ~ (-1.1 arcsec, 6.2 arcsec).

The UVIS axes and IR axes are rotated by ~0.1 deg. These parameters determine the RA and Dec values assigned to the output drizzled images in the pipeline. As in the earlier release, only FORWARD transform coefficients are tabulated and the distortion in all filters is represented by the polynomials determined from UVIS F606W and IR F160W images. In Spring 2010 we plan to release filter dependent image distortion representations.

4. Persistence in WFC3/IR ==============================================

Like other HgCdTe IR devices, the IR detector in WFC3/IR exhibits persistence that show up as faint residual images, or after-images from previous exposures. Persistence was observed during ground tests and is being characterized on orbit as part of the Cycle 17 calibration program. Examples of persistence will be shown in the new instrument handbook. The amount of persistence is small, typically 6 mags less than (0.3% of) the original exposure. The persistence is not a function of source brightness directly, but rather is associated with the total charge accumulated before readout relative to the full well of the detector pixels. The persistence exhibits an initial fast decay with a time scale of minutes and then has a long tail with timescales on the order of an hour. Thus persistence is usually seen from objects within dithered frames within a single orbit and more occasionally from objects observed during previous orbits. Persistence images of stars often appear "mushy" because the amount of persistence saturates at about full well. In most cases, persistence will not affect the science obtained from the WFC3/IR channel, but observers do need to be aware that persistence does exist, particularly when observing faint objects in crowded fields, like globular clusters, or bright extended objects, such as 30 Doradus.

Analysis of the issue is ongoing at STScI. Tools are being developed which identify the regions of an image that are likely to be affected by persistence and to mitigate the effects in the data that have been obtained. Tests are being conducted at the Detector Characterization Lab at GSFC to determine whether alternative operation procedures can reduce the effect. The visits that are most likely to cause the worst persistence in another visit have been identified, and the operations staff has put procedures in place that are intended to prevent WFC3/IR from being used for a number of hours after these visits. However, this will not eliminate persistence showing up in all cases, in part because most WFC3 science targets are the fainter objects in a field.

If you suspect previous visits from another observer have produced residual images in your data, you may check whether there have been earlier WFC3/IR visits shortly before your observation using the following search form on MAST:

If there were no observations within the previous orbit or two, it is unlikely that persistence will be an issue. If there are observations in the previous orbit you will want to check further. Usually, but not always, the previous observations will have been proprietary. If this is the case, displaying the 2MASS or DSS image of the field in one of a number of on-line image servers or even looking at the confirmation images for the offending proposal will give you a good idea whether there is likely to have been an issue. If the issue looks severe in the sense that you feel your science has been compromised, you will want to have a conversation with your contact scientist as to how you might proceed. In principle, one could ask for the observation to be repeated. However, in most cases our analysis suggests that one will be able to mitigate the effects of residual images with additional processing. However, most after- images will have been created within your own observation from objects that exceeded half of the full well-depth in an earlier image. Often these after-images will not be important for your science, since they usually affect only a small number of pixels, and the ones that are affected may not be within the region of the chip where the actual science target lies. Furthermore, "drizzling" will reduce the the effect of ghosts in well-dithered images. That said, all observers should be aware of the possibility of faint ghosts in images obtained with the IR detector, and inspect the images with care.

WFC3 has set up a web page as a resource of IR persistence information:

It is not yet populated, but if you do have questions about persistence in your data, please check this address and if your question is unanswered contact your contact scientist or the help desk

5. New Documentation - Instrument Handbook and ISRs ==============================================

The New WFC3 Instrument Handbook for Cycle 18 will be published in early January 2010, at the WFC3 web page:

These new ISRs have been published since the last STAN (Sept 2009):

ISR 2009-44: Radioactivity in HgCdTe devices: potential source of "snowballs". P. McCullough 03 Dec 2009

ISR 2009-43: "Snowballs" in the WFC3-IR Channel: Characterization. B. Hilbert 03 Dec 2009

ISR 2009-38: WFC3 SMOV Programs 11436/8: UVIS On-orbit PSF Evaluation. G. F. Hartig 03 Dec 2009

ISR 2009-37: WFC3 SMOV Programs 11437/9: IR On-orbit PSF Evaluation. G. F. Hartig 03 Dec 2009

ISR 2009-36: WFC3 SMOV Proposal 11443: Alignment of the WFC3/IR Apertures to the FGS Coordinate Frame. L. Dressel, C. Cox, M. Lallo 17 Nov 2009

ISR 2009-35: WFC3 SMOV Proposal 11442: Alignment of the WFC3/UVIS Apertures to the FGS Coordinate Frame. L. Dressel, C. Cox, M. Lallo 17 Nov 2009

ISR 2009-34: WFC3 SMOV Proposal 11445 - IR Geometric Distortion Calibration. V. Kozhurina-Platais, et al. 17 Nov 2009

ISR 2009-33: WFC3 SMOV Proposal 11444 - UVIS Geometric Distortion Calibration. V. Kozhurina-Platais, et al. 17 Nov 2009

ISR 2009-32: WFC3 SMOV Proposal 11549: Image Stability. T. M. Brown 19 Nov 2009

ISR 2009-31: WFC3 SMOV Proposal 11450: The Photometric Performance and Calibration of WFC3/UVIS. J.S. Kalirai, et al. 17 Nov 2009

ISR 2009-30: WFC3 SMOV Proposal 11451: The Photometric Performance and Calibration of WFC3/IR. J.S. Kalirai, et al. 17 Nov 2009

ISR 2009-29: WFC3 SMOV Proposal 11419: UVIS Gain. S. Baggett, T. Borders 17 Nov 2009

ISR 2009-28: WFC3 SMOV Proposals 11423/ 11543: IR FSM and Lamp Checks. S. Baggett 17 Nov 2009

ISR 2009-27: WFC3 SMOV Proposal 11422/ 11529: UVIS SOFA and Lamp Checks. S. Baggett, E. Sabbi, P. McCullough 17 Nov 2009

ISR 2009-26: WFC3 SMOV Proposals 11419,11426,11431, and 11448: On-Orbit Biases. T. Borders, S. Baggett 17 Nov 2009

ISR 2009-23: WFC3 SMOV Program 11420: IR Channel Functional Tests. B.Hilbert, P.R. McCullough 17 Nov 2009

ISR 2009-22: WFC3 SMOV Proposal 11421: Channel Select Mechanism Test. H. Bushouse 17 Nov 2009

ISR 2009-20: WFC3 SMOV Program 11798: UVIS PSF Core Modulation. E. Sabbi 17 Nov 2009

ISR 2009-19: WFC3 SMOV PROGRAM 11452: UVIS Flat Field Uniformity. E.Sabbi 17 Nov 2009

ISR 2009-18: WFC3 SMOV proposal 11552: Calibration of the G102 grism. Kuntschner et al. 19 Nov 2009

ISR 2009-17: WFC3 SMOV proposal 11552: Calibration of the G141 grism. Kuntschner et al. 19 Nov 2009

ISR 2009-16: WFC3 SMOV Proposals 11419, 11426, 11431, and 11446: On-Orbit Darks. Borders & Baggett 19 Nov 2009

ISR 2009-15: WFC3 SMOV program 11454: Activation Test. L. Petro 17 Nov 2009

6. New WFC3 Reference Files ==============================================

These new reference files have been delivered since the last STAN (Sept 2009):

A new IR cosmic ray rejection table for use with the new calwf3: tag1826ki_crr.fits - delivered in October 2009

A new UVIS bad pixel table, to account for bad pixel rows on either side of the chip gap, will be delivered in late December 2009.

All reference files can be found at: