1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
|
.help linpol Apr92 noao.nproto
.ih
NAME
linpol -- Calculate linear polarization, angle, and Stokes images
.ih
USAGE
linpol input output
.ih
PARAMETERS
.ls input
A list of input images. There must be either three or four input
images taken with the polarizer at even multiples of a 45 degree
position angle.
.le
.ls output
The output data cube which will contain as separate bands the
fractional linear polarization and angle frames, and optionally the
Stokes parameter frames.
.le
.ls degrees = yes
Report the polarization angle in degrees? If \fBdegrees\fR = no, the
polarization angle will be reported in radians.
.le
.ls stokes = yes
Output the Stokes parameter images? If \fBstokes\fR = yes, the three
linear Stokes parameters, I, Q, and U, will be included in the
\fBoutput\fR data cube as separate bands. If \fBstokes\fR = no, only
the fractional linear polarization and angle frames will appear in the
output.
.le
.ls normalize = no
Normalize the Q and U frames? This is appropriate when using a tool
such as VELVECT to plot the polarization vectors. If \fBnormalize\fR =
yes, the Q and U Stokes parameter frames will be normalized by dividing
by the I parameter frame. This parameter has no effect on either the
fractional polarization or angle frames.
.le
.ls keyword = "polangle"
This must be set to the name of a header keyword that contains the
polarizer angle for each of the \fBinput\fR images. LINPOL will only
accept polarizer angles at even 45 degree separations. Either four such
frames, at 0-45-90-135 degrees, or three frames with any one of the
0-45-90-135 degree frames omitted, may be specified.
.le
.ih
DESCRIPTION
LINPOL calculates the pixel-by-pixel fractional linear polarization and
polarization angle for a set of either three or four images taken with
polarizer set at even multiples of a 45 degree position angle. At least
three different frames with the header \fBkeyword\fR set to one of
0, 45, 90, or 135 degrees must be specified in the \fBinput\fR list.
If \fBdegrees\fR = yes, the output polarization angle band of the image
will be in units of degrees, if \fBdegrees\fR = no, the angle will be
reported as radians. If \fBstokes\fR = yes, the output image
will consist of five separate bands, one each for the pixel-by-pixel
fractional linear polarization and the corresponding polarization angle,
and one for each of the I, Q, and U pixel-by-pixel Stokes parameters.
If \fBstokes\fR = no, only the fractional polarization and the polarization
angle will be saved in the output.
The \fBnormalize\fR parameter is useful for plotting purposes.
If \fBnormalize\fR = yes, the Q and U Stokes parameter frames will be
normalized by dividing by the I parameter frame. This may be appropriate
when using a tool such as VELVECT to plot the polarization vectors.
This parameter has no effect on either the fractional polarization or
angle frames.
Each input image must contain the corresponding polarizer angle
in the header keyword specified by the parameter \fBkeyword\fR
Linpol will only accept polarizer angles at even 45 degree separations.
Either four such frames, at 0-45-90-135 degrees, or three frames with
any one of the 0-45-90-135 degree frames omitted, may be specified.
The output image header will include information describing the particular
input images that went into its generation and the particular nature of
each band of the output.
.ih
EXAMPLES
An observer obtained four exposures of a particular field through a
polarizer set at a position angle of 0-45-90-135 degrees. The first
step in producing a good map of the polarized light from (extended
or point-like) sources in the field is always to register these frames
very precisely. A slight mismatch in the positioning of each pixel
relative to the shoulders of nearby sources or extended emission will
result in large errors in the determination of the polarization quantities.
Another preprocessing step that may be desirable is to match the PSFs
(Point Spread Functions) of the various frames. Ideally, these are
stable in the raw data (i.e., the seeing at the telescope was constant),
but if not they must be matched to avoid the same errors as above. Note
that it may also be a good idea to "smooth" the raw images before
applying linpol to increase the signal-to-noise of the output.
After guaranteeing the integrity of the input images, the image header
\fBkeyword\fR must be created to contain the position angle. The hedit
task can be used to do this:
.nf
hedit im.00 polangle 0 add+
hedit im.45 polangle 45 add+
hedit im.90 polangle 90 add+
hedit im.135 polangle 135 add+
.fi
At this point, the input images are ready to be processed by linpol.
To generate an output image containing the fractional linear
polarization and polarization angle in separate bands, along with the
pixel-by-pixel Stokes parameter frames:
.nf
np> linpol im.*.imh polar
.fi
To omit the Stokes parameter frames:
.nf
np> linpol im.*.imh polar stokes-
.fi
To represent the pixel-by-pixel polarization angle in radians, rather
than degrees:
.nf
np> linpol im.*.imh polar degrees-
.fi
To normalize the Q and U Stokes frames and plot the result with velvect:
.nf
np> linpol im.*.imh polar normalize+
np> imhead polar lo+
polar[100,100,5][short]: Linear polarization image
No bad pixels, no histogram, min=unknown, max=unknown
Line storage mode, physdim [100,100,5], length of user area 2147 s.u.
Created Wed 10:15:05 29-Apr-92, Last modified Wed 10:15:05 29-Apr-92
Pixel file 'ursa!/ursa/scr3/iraf/seaman/polar.pix' [ok]
...
POL0 = 'im.00.imh'
POL45 = 'im.45.imh'
POL90 = 'im.90.imh'
POL135 = 'im.135.imh'
POLAR = 'Band 1 is the percent polarization'
ANGLE = 'Band 2 is the polarization angle'
I-STOKES= 'Band 3 is the Stokes I parameter'
Q-STOKES= 'Band 4 is the normalized Stokes Q parameter'
U-STOKES= 'Band 5 is the normalized Stokes U parameter'
np> velvect polar[*,*,4] polar[*,*,5]
.fi
Note that the current version of the velvect task is not particularly
appropriate for this use. It has no support for reducing the pixel
resolution of the output plot: each pixel will generate a plotted vector
so that to produce an uncrowded (and low "noise") plot, the input images
or output bands must be manually block averaged or otherwise smoothed.
In addition, the plotted vectors are directed (little arrows) not
undirected line segments, and the length of the vectors are not easily
adjusted.
.ih
SEE ALSO
velvect, imalign, hedit
.endhelp
|
