When we have an observation that consists of data from multiple epochs there is a difference in the polarisation position angle. A consequence is that you cant just stack all the data from different epochs together to make a full sensitivity QU cube.
Ideally for the 20" resolution QU cubes we would make a different cube of each epoch so they can be aligned afterwards. However, these data are fairly large - 36 GB for the PB plus non PB corrected compressed QU cubes.
Presently the aligning of the QU cubes isnt done in the post processing. Furthermore, only the PB corrected cubes are used. So suggestion from Shane is to figure out which datasets are best (can be done from rms of the stokes V images) and make QU cubes for just the best datasets. To not increase the archive size we could do the top 2 datasets and just store the PB corrected maps.
When we have an observation that consists of data from multiple epochs there is a difference in the polarisation position angle. A consequence is that you cant just stack all the data from different epochs together to make a full sensitivity QU cube.
Ideally for the 20" resolution QU cubes we would make a different cube of each epoch so they can be aligned afterwards. However, these data are fairly large - 36 GB for the PB plus non PB corrected compressed QU cubes.
Presently the aligning of the QU cubes isnt done in the post processing. Furthermore, only the PB corrected cubes are used. So suggestion from Shane is to figure out which datasets are best (can be done from rms of the stokes V images) and make QU cubes for just the best datasets. To not increase the archive size we could do the top 2 datasets and just store the PB corrected maps.
Not sure if anyone else has any thoughts on this.