pyhsiehATalma
commented
2 years ago Test of stitching TP cube (spw21, HCO+)
- Step 1: smooth data to a common beam (75").
- Step 2: regrid science and weight images to match a template image. Make sure the velocity axes match (definition of rest frequency).
- Step 2.1: trim edges of low weights.
- Step 3: stitching weight cubes (wt).
- Step 4: stitching line cubes (im) with weighted average. for example: (im0*wt0+im1+wt1+...)/(wt0+wt1+...)
I started from stitching ae and ag. The weighted average works well. Some channels show seamlines. ,
Left: ag Middle: ae Right: ag+ae
The seamline is less likely from the edge effect, but from inconsistent structure (low s/n perhaps?) in the overlapping area of two maps. The inconsistent structures could be due to errors of baseline subtraction made by the pipeline.
It looks like the discrepancy came from the spectral baseline subtraction made by the pipeline. The maps without trimming edges are shown below.
Line profiles within the pink box in the map are shown.
In this particular case, the spw bandwidth was very narrow compared with the line width and (perhaps) the pipeline wrongly fitted nonlinear baselines to the line wings. The weighted averaged map is reasonably show averaged intensity of these two data.
Possible solution
- redo baseline subtraction -> could be time consuming and doesn't help much since the available baseline for fitting is really limited
- imaging (pipeline baseline subtracted) multiple fields together using sdimaging -> could be very very time consuming.
snlongmore
This issue is created to combine 7 m and TP data. CS(2-1) line. 12m: spw 22 7m: spw 24
Please find the data in Globus, the path is,
/upload/ACA-TP-CS2-1/
feather_cube_fits: feather cube per field m0_fits: integrated intensity feather map per field m8_fits: peak intensity feather map per field mosaic_fits: mosaic feather cube