[ Basic Info | User Guide ]
Basic Information on linmos
Task: linmos
Purpose: Linear mosaicing of datacubes
Categories: map combination
LINMOS is a MIRIAD task that performs simple linear mosaicing
of input cubes to produce a single output cube. If only one
input cube is given, LINMOS essentially does primary beam
correction on it. If the input cubes overlap, LINMOS combines
the overlapping regions so as to minimize the rms error.
To determine the primary beam of the telescope, LINMOS first
checks the map header for the presence of the "pbtype" and then
"pbfwhm" parameters. If present, LINMOS assumes the primary
beam is the given type. If these parameters are missing, LINMOS
checks if the telescope is one that it knows. If so, then the
known form for the primary beam is used. See task "pbplot" to
check LINMOS's primary beam models.
Key: in
The names of the input cubes - many may be given. There is no
default. Inputs should generally be on the same grid system,
use invert with options=mosaic and the offset keyword to achieve
this. If not, linear interpolation is performed to regrid using
the first plane of the first image as the template. LINMOS's
ability to do this is inferior to task REGRID.
The intensity units of all the inputs, and the pixel size and
alignment of the third dimension are assumed to be the same.
Mosaicing fields with different resolution together will cause
errors in the fluxdensities of sources.
Key: out
The name of the output cube. No default. The center and pixel
size of the first input image is used as the grid system of the
output. The synthesized beam parameters are also taken from the
first image header - you may want to pick a better average beam.
Key: rms
RMS noise level in each of the input cubes. If not specified,
the value is taken from the 'rms' item in the input image header
if that is available, and if not, then the rms of the previous
image is used. If no value could be determined for the first
image, a warning is issued and ALL images are given equal weight
by assigning an RMS of 1.0.
Key: bw
Bandwidth of the image in GHz, default 0. If specified the beam
response will either be averaged across the frequency band before
being applied to the image or, if the input images contain a
spectral index plane (created with the mfs option of restor)
the images will be evaluated and corrected across the band.
Use this for wide band images to improve the accuracy of the
correction.
Note that doing wide band primary beam correction at low
frequency will make the effective observing frequency vary
significantly across the field.
An optional second parameter can be given to set the number
of frequencies to divide the bandwidth into, it defaults to 10.
Key: cutoff
The cutoff level to use for the primary beam, e.g., use 0.5 to
restrict the contribution of each input cube to the pixels inside
the half power beam width. Normally the built-in level for each
beam model is used (generally <0.1). This can be useful
e.g., to restrict polarization mosaics to use only the part of
the beam with low instrumental polarization. Use pbplot to see
what radius each level corresponds to. Defaults to zero
which means use the built-in level.
Key: options
Extra processing options. Several can be given, separated by
commas. Minimum match is supported. Possible values are:
taper By default, LINMOS fully corrects for primary
beam attenutation. This can cause excessive
noise amplification at the edge of the mosaiced
field. The `taper' option aims at achieving
approximately uniform noise across the image.
This prevents full primary beam correction at the
edge of the mosaic. See Eq. (2) in Sault,
Staveley-Smith and Brouw (1996), A&AS, 120, 375,
or use "options=gains" to see the form of the
tapering.
sensitivity Rather than a mosaiced image, produce an image
giving the RMS noise across the field. This is
dependent on the RMS specified, either as an
input parameter or else as a header item (see
above).
gain Rather than a mosaiced image, produce an image
giving the effective gain across the field. If
options=taper is used, this will be a smooth
function. Otherwise it will be 1 or 0 (blanked).
frequency Rather than a mosaiced image, produce an image
giving the effective frequency across the field.
alpha If an I*alpha plane is present in the input,
produce a mosaiced version of this as well.
You can feed the output image to mfspin to
produce the mosaiced spectral index image.
The spectral index mosaic becomes unreliable
when the fractional bandwidth is > 0.4 (this
is because higher orders are needed to model
the primary beam response with frequency).
Revision: 1.34, 2021/06/03 07:09:31 UTC
Generated by miriad@atnf.csiro.au on 03 Jun 2021