"Sault weighted" image is one which is more pleasant to view (without high noise at the edges of mosaic images), it is flux correct upto a where the beam coverage becomes low and is tapered off onwards just to keep the noise from rising in the overall image(see Eq[2] from Sault, Staveley-Smith and Brouw (1996), Astron. Astrophys. Suppl, 120, 375)
<casaxml xsi:schemaLocation="http://casa.nrao.edu/schema/casa.xsd file:///opt/casa/code/tools/xml/casa.xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://casa.nrao.edu/schema/psetTypes.html">
<tool module="linearmosaic" name="linearmosaic">
<shortdescription>combining images in a weighted fashion</shortdescription>
The {\tt linearmosaic} tool ({\tt lm}) provides a toolkit for stitching images in a weighted fashion.
The default linear mosaic equation is defined by Equation (6) in Cornwell, Holdaway et al (Astronomy and Astrophysics, Vol. 271, p. 697 (1993)).
I^{lm}(\theta)={{\sum_p A_p(\theta)(I_p(\theta)A_p(\theta))w_p}\over{\sum_p A_p^2(\theta)w_p}}
where $A_p(\theta)$ is the primary beam (PB) of a given pointing $p$, $w_p$ is a sensitivity weight and the image of that pointing is $I_p(\theta)$; the linear mosaic being $I^{lm}(\theta)$
<include>linearmosaic_forward.h</include>
<include>linearmosaic_private.h</include>
<method name="linearmosaic" type="constructor">
<shortdescription>Construct a linearmosaic tool</shortdescription>
<returns type="linearmosaic"/>
Create a {\tt linearmosaic} tool.
<method name="defineoutputimage" type="function">
<shortdescription>Set the output direction image parameters and name</shortdescription>
Define the direction axes output image parameters.
The output image will get the same number of spectral and polarization planes as the input images. This function create a fresh new output image. If an image of the same name exist on disk it will be erased. The spectral and polarization part of the image will be identical to the images that are being mosaiced.
The output image will by default be flux correct and the weight image will be ${\sum_p A_p^2(\theta)}$ where the primary beam is $ A_p(\theta)$
<param name="nx" type="int">
<description>Total number of spatial pixels in x</description>
<param name="ny" type="int">