WFPC2 Polarization Calibration

-John Biretta and Matt McMaster
(Last update: 23 December 1997)

What's New:

97 Dec. 23: Much progress has been made during the last few months. We have improved the instrumental polarization model in both the simulator and calibration tools. Improvements include detailed modeling of the MgF2 coating on the pick-off mirror, and correction of an error found in one of our references on polarization theory. Comparison of data processed with the old and new versions typically shows differences in the fractional polarization of 1% or less, and 1 to 2 degrees in the polarization position angle. The new model predicts the relative counts observed on-orbit in different settings of the polarizer filter to 1.5 percent RMS accuracy.

Upgraded Simulator Tool.

Upgraded Calibrator Tool.

A new Instrument Science Report describing the WFPC2 Polarization Calibration is available in PostScript form (1.3 Mbytes, 67 pages).

An example IRAF script for calculation and display of polarization images.

97 Jan. 28: The bit about changing the sign on U when taking polarization data into AIPS turns out to be a red herring. What one really needs to do, is make sure the sign on the X pixel size (CDELT1) is negative after you get the data into AIPS. AIPS uses a convention where the RA axis has a negative pixel scale. If instead you had a positive RA pixel scale, the polarization vectors would point the wrong way. (See example below for generating and displaying polarization images.)

97 Jan. 27: Added example showing how to generate Stokes images and a vector plot for R Mon.

We did more work with the on-orbit calibrators and tried to develop some simple aperture-dependent corrections to improve the accuracy below 1%, but this was not very successful. The corrections will probably need to contain at least parallel and perpendicular terms for each aperture. We will probably need to wait for more on-orbit data to derive those corrections.

96 Nov. 20: It now appears that the accuracy in the rotated POLQ is the same as the other settings (about 1 to 2%). The strange results we got for the POLQN33 and POLQP15W settings on the polarization calibrator stars was due to a bug in CALWP2 (WFPC2 pipeline calibration program) which caused the PC1 flat to be applied to WF2 data when only one CCD is read-out. We've re-calibrated some of the POLQN33 data by hand, and it looks fine. It will probably take a few weeks to recalibrate everything. More info then.

96 Nov. 19: The large errors for the rotated positions of the POLQ have been traced to a strange error in CALWP2 (WFPC2 pipeline calibration program). Under certain circumstances the wrong flat is applied to the data. More details soon.

96 Nov. 14: Test version of new (second) calibration tool made available. General re-work / update of this page.

96 Nov. 11: WWW access to polarizer flats discontinued. They are instead available through the HST archive.

96 Oct. 10: HST archive names for polarizer flats added. Pol. WWW tool upgraded to computed (predicted) - (observed) counts.

96 Sep. 25: We've added a (somewhat crude) WWW tool for photometric calibration, and some discussion of photometric results.

96 Sep. 18: We've updated the matrices for the polarizers and POM with more accurate expressions.

Polarizer Flat Fields:

The first step in polarizer calibration is to flat field the data.

The polarizer flats are available though the HST archive. Archive file names are given below. Note that the flats have suffix R4H (i.e. file GA41233FU.R4H) and their data quality files have suffix B4H (i.e. file GA41233FU.B4H). Find the appropriate flat in the list below for your spectral filter and polarizer setting. Get the flat file from the HST archive (via STARVIEW, etc.) and *multiply* it into your data.

Filter1     Filter2     File

F300W       POLQN18    GA21514JU
F300W       POLQN33    GA21514MU
F300W       POLQP15    GA21514PU
F300W       POLQ       GA21514SU

F336W       POLQN18    GA215151U
F336W       POLQN33    GA215154U
F336W       POLQP15    GA215157U
F336W       POLQ       GA21515BU

F390N       POLQN18    GA21515EU
F390N       POLQN33    GA21515HU
F390N       POLQP15    GA41233FU
F390N       POLQ       GA41233IU

F410M       POLQN18    GA41233LU
F410M       POLQN33    GA41233OU
F410M       POLQP15    GA41233QU
F410M       POLQ       GA41233TU

F439W       POLQN18    GA412342U
F439W       POLQN33    GA412345U
F439W       POLQP15    GA412348U
F439W       POLQ       GA41234BU

F547M       POLQN18    GA711082U
F547M       POLQN33    GA711087U
F547M       POLQP15    GA71108BU
F547M       POLQ       GA71108FU

F555W       POLQN18    GA71108KU
F555W       POLQN33    GA71108OU
F555W       POLQP15    GA71108TU
F555W       POLQ       GA711093U

F606W       POLQN18    GA711097U 
F606W       POLQN33    GA71109AU
F606W       POLQP15    GA713053U
F606W       POLQ       GA713058U

F656N       POLQN18    GA71305DU
F656N       POLQN33    GA71305HU
F656N       POLQP15    GA71305MU
F656N       POLQ       GA71305RU

F675W       POLQN18    GA713061U
F675W       POLQN33    GA713065U
F675W       POLQP15    GA713069U
F675W       POLQ       GA71306DU

Theory of WFPC2 Polarizer Calibration:

A new Instrument Science Report describes our current understanding of the Photometric Calibration. It is available in PostScript form (1.3 MBytes, 67 pages). Comments are welcomed.

The older HTML version of the theory discussion from January 1997 is also available.