Anthony J Weitenbeck ajw@sal.wisc.edu
In general, when making polarimetric observations, the polarized flux is
a small fraction of the total flux. As such, the percent polarization
is quite sensitive to calibration errors and requires data with a high
signal to noise ratio. For WF/PC-1 polarization imaging it is necessary
to obtain observations with three orientations of the Polacoat analyzers
in order to unambiguously determine the polarization. Errors in the flat
field will introduce spurious polarization even where none exists.
The three separate images (in general) will be shifted by several pixels
relative to each other (requiring subsequent reregistration). This means
that a particular point in the object will be using a different flat field
correction factor in each separate image. Errors in the flat field result in
erroneous flux differences between images, which will yield spurious
polarization when that calculation is performed.
The observation discussed in this poster was analyzed by
by Capetti et al (ApJ 446,155). In their figures for which data
for this WFPC observation is plotted, they blank out an area for which they
say the flat field is unreliable (near Cloud G of Evans et al 1991). The
flat field used in the current pipeline reductions was c191643ew.r6h, made on
01/09/92; from this the flatfields for the three polaroids were made, and
given the names c2313109w, c231310gw, and cd231310mw.
This blanked out area shows on the c191643ew.r6h flat field as a roughly
circular area about 0.6" across in which the flat field value is greater
than the surrounding flat field values by as much as 60%. It is also present
at similar levels on the other flat fields made earlier in the mission
and prelaunch.
In their figures 7 and 8 of polarized flux and polarization an area coincident
with Cloud D of Evans et al. shows significant polarization. We would like to
point out that there is a similar high area of the flat field at that
position also (about 0.5" across, maximum about 20% higher than the surrounding
flat field), which suggests that polarization observed at that position is
also at least partly an artifact of the irregularity in the flat field.
[Cloud D is on top of it on one of the three polaroid images, and on the
edge of it on the other two.]
There is also a similar elevated area of the flat field (up to about 30% greater
than the surrounding flat field) almost exactly on the opposite side of Cloud B
of Evans et al as the area blocked off by Capetti et al. (but which is not on
their plots because of their chosen flux and polarization value plotting
limits), which also shows high polarization in our polarization and polarized
flux plots.
Cappetti et al also find a small region with high polarization on both their
WFPC and FOC (where it is resolved into two clouds) polarization observations
which corresponds to the small area with large polarization on our plots at about
x = 275, y = 250. Since this area shows high polarization on both WFPC and FOC
images, its reality is more certain than the other areas referred to above;
however, this area, too, corresponds to a small area of slightly elevated flat
field values (about half the diameter of those referred to above, up to about
10% greater than the surrounding flat field), so any estimate of the value
of the polarization from the WFPC data must have a large uncertaintly.
This work was supported by grant
NASA/CAL TECH P.O. PC 033638.
References:
Arthur D Code
Space Astronomy Lab
U of Wisconsin, Madison
Capetti, A., Axon, D. J., Macchetto, F.,
Sparks, W. B., and Boksenberg, A., 1995,
ApJ 446, 155
Evans, I. N., Frod, H. C., Kinney, A. L.,
Antonucci, R. R. J. 1991, ApJ 369,L27