diff options
Diffstat (limited to 'noao/nproto/doc/linpol.hlp')
| -rw-r--r-- | noao/nproto/doc/linpol.hlp | 164 |
1 files changed, 164 insertions, 0 deletions
diff --git a/noao/nproto/doc/linpol.hlp b/noao/nproto/doc/linpol.hlp new file mode 100644 index 00000000..5cde0e60 --- /dev/null +++ b/noao/nproto/doc/linpol.hlp @@ -0,0 +1,164 @@ +.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 |