polimodel -- Produce model linear polarization and position angle images
USAGE polimodel inint simlis coincid intclip pacorr random seed polran paran modpol modpa
- inint = "" [name]
- Input total signal (Stokes I) image to model.
- simlis = "" [name]
- Name of STSDAS table file listing the parameters of the polarization models. The central polarization (in %), x and y pixel coordinates of the centre, the exponent for change of polarization with distance and the scaling radius for the exponent are required. Any number of distinct models can be listed.
- coincid = "" [integer]
- Method used to treat pixel values where the polarization from two models coincides. Options for combining the polarization and position angle are: 1=maximum; 2=minimum; 3=vector sum.
- intclip = "" [real]
- Minimum value of signal in input image to produce model polarization values.
- pacorr = "" [real]
- Additive correction for position angle values.
- random = "" [boolean]
- Whether to apply random errors to output polarization and position angle values.
- seed = "" [integer]
- Seed for random number generator.
- polran = "" [real]
- Fractional value of 1 sigma linear polarization error.
- paran = "" [real]
- Fractional value of 1 sigma polarization position angle error.
- modpol = "" [name]
- File name for output linear polarization (%) image.
- modpa = "" [name]
- File name for output position angle (degrees) image.
This task produces a model linear polarization image and corresponding position angle image from an input image assumed to represent intensity (Stokes I parameter). The input image could be a model image or real data. The linear polarization image is produced from the parameters read from a table file listing:polarization at fiducual position X,Y pixel coordinates of the fiducial position of polarization radial exponent of variation of polarization radial scale factor The ouput polarization is calculated thus: Pol. = Peak Pol. * (R/R0)**EXP where R = SQRT( (X-X0)**2 + (Y-Y0)**2 )The position angle at a given pixel is the direction perpendicular to the vector from the position of fiducial polarization to the current pixel. This is understood as the centre of a centro-symmetric polarization pattern (e.g. for single Rayleigh scattering). A threshold for the input image can be defined such that polarization and position angle are not computed for input values less than this threshold.
If the input table specifies more than one polarization model then some rules are required to deal with pixels having polarization contributions from both models. Three cases are implemented:Maximum value of polarization and position angle corresponding to this maximum value are output; Minimum value of polarization and position angle corresponding to this minimum value are output; The output polarization is formed from the vector sum of the individual Q and U Stokes parameters of the models.
An attempt to produce ouput polarization and position angle images with random errors is made. Normally distributed polarization and position angle values can be produced by specifying sigma(pol)/pol and sigma(theta)/theta respectively. The normally distributed values are formed before combination of pixel values when more than one model is input.
1. To form the linear polarization and position angle images for an image of a reflection nebula centred at pixel 256,256 of fiducial polarization 5% and increasing linearly with a doubling distance of 20 pixels plot. Values are only calculated for signal values greater than 20 and no random error are applied.
cl> polimodel inint="nebint.fits" simlis="modrefl.tab" coincid=0 intclip=20.0 pacorr=0.0 random=no seed=0 polran=0.0 paran=0.0 modpol="nebpol.fits" modpa="nebpa.fits"
2. To form the linear polarization and position angle images for an image of a galaxy cluster with individual models of the polarization and its radial dependence. Coincident values in the polarization and position angle images from the different models are combined vectorially. Output values are calculated for input values greater than 100 and random errors are applied with sigma(pol)/pol of 0.1 and sigma(theta)/theta of 0.1.
cl> polimodel inint="galclust.fits" simlis="galclust.tab" coincid=3 intclip=100.0 pacorr=-10.0 random=yes seed=99 polran=0.10 paran=0.10 modpol="galclustpol.fits" modpa="galclustpa.fits"
mkobjects, polimplot, fieldplot