MOPEX Online Manual

APEX Modules: Mosaic Interpolate

PURPOSE

This module performs a projection of input images onto a 2D plane defined by the FIF table, and an interpolation of the input pixel values to an output array of pixels with the user-defined pixel size. It corrects for the optical distortion in the input images, using the WCS distortion parameters in the input FITS headers. The process is designed to accept images in units of surface brightness (MJy/sr or microJy/arcsec^2) and to yield images in the same units, but it is not restricted to this. The interpolated output mosaic is used for point source detection and aperture photomery.

PARAMETER BLOCK

&MOSAICINTIN
 INTERP_METHOD = 1,
&END

INPUTS

Interp Method: Four interpolation options are available: 1. Default, 2. Drizzle, 3. Grid and 4. Cubic. (These are defined in namelists by INTERP_METHOD(int) = number.) Each has a single parameter. Most users will choose Default or Drizzle.

1. Default: Each output pixel is a linear weighted sum of input pixels with weights equal to the area overlap with the output pixel. The optional parameter is Fine Res(int) (FINERES in namelists). The input pixel can be sub-divided with sub-pixel sizes of Input_Pixel_Size/FINERES before projecting onto the output array; a typical value might be 2. The value of each sub-pixel is a linear interpolation of the input pixels with weights determined by the area overlap with a pixel the same size as the original but centered on the sub-pixel. So it's a convolution. The default value of 0 means no sub-dividing. Users should keep Fine Res, FINERES = 0. The Fine Res option is not fully implemented in the presence of bad pixels.

2. Drizzle: Each input pixel is shrunk Drizzle Factor(float) (DRIZ_FAC in namelists) times its original size along each axis, e.g. 0.5. The value of the shrunken pixel is the same as the original pixel. The shrunken pixels are then projected onto the output pixels with their values distributed into output pixels according to area overlap. Note: When using Drizzle and creating RMasks from the data, several of the mosaicking steps are run twice. First, a normal (Default) interpolation is carried out so that MOPEX can run the outlier rejection scheme for the original pixels. Once the outlier rejection masks (RMasks) have been created, MOPEX returns to the Interpolate module and re-runs the interpolation with the Drizzle algorithm, masking out pixels flagged in the RMasks. When using this option, do not include the Mosaic Reinterpolate Module in the processing flow.

3. Grid: This method is intended to create a crude first-look mosaic quickly. Each input pixel is filled with Grid Ratio(int) squared grid points (GRID_RATIO in namelists). Each grid point is assigned the value of the pixel it belongs to. Each grid point is projected onto the output frame and the flux associated with the grid point is added to the output image pixel into which the grid point was projected. You may not run Reinterpolate with this option. The gain in speed is up to 10 times that of the Default method. The price you pay is the fidelity of the interpolated images. So in general these should not be used for science.

4. Cubic This method uses a bicubic interpolation. It is like the Default method except that each output pixel is a weighted sum of the 16 nearest input pixels, with the weights determined by bicubic polynomials. A tunable parameter Alpha(float) (ALPHA in namelists) pins the weights. The default value of -0.5 should be used. Note: this method could be useful in high S/N cases because it enforces smoothness in the interpolation function, but the major drawback is significantly more noise correlation than the Default linear interpolation.

OUTPUTS

Mosaic FIF Table (mosaic_fif.tbl): The FIF table describing the final mosaic, taking into account the user-selected pixel size. Note: This is in the top level Output directory.

Geometry Output Table (interp_ImageList.txt.tbl): The interpolated images' offsets in x- and y-direction relative to the FIF and their sizes are specified in this file.

Interp Stack (interp_*fits): The output images, interpolated, distortion-corrected, and projected onto the output FIF. A list is also made.

Coverage Stack (interp_*covg.fits): The corresponding coverage maps for each output image, taking into account the bad pixels in the input. A list is also made.

DISCUSSION

Most users will choose between Default and Drizzle. The smaller you make Drizzle Factor, the more coverage you need to avoid holes in the mosaic. A rule of thumb is that if you have a coverage of 10 or less, you will probably want to use the Default interpolation.

In general, a projection onto the reference frame of an input image with optical distortions will not be a simple rectangle. Each interpolated image occupies a part of the FIF and is, in general, of different size, with different offsets from the origin of the FIF. The interpolated images' offsets in x- and y- directions relative to the FIF, and their sizes (in integral numbers of pixels) are given in the file interpolated_ImageList.txt.tbl, and also written in the headers of the interpolated images in the keywords MINTOFFX and MINTOFFY.