Background
Data sets show short baselines at the end of E and W arms with high amps in the amp vs uvwave plots. Paul suggests the long baselines have too much power in comparison. Juergen suggests S band may also be affected.
Example plot from original CIPL run for 22A-012.sb41313612.eb41801998.59714.25389122685:
Same data set/plot using a model for 3C286 made at 33 GHz (slightly lower Amp values):
Same data set/plot using a model for 3C286 made at 26 GHz (same as original model):
Likely the same issue for 3C147 - plot from 20B-342.sb39244407.eb41801147.59713.02153945602:
Tests/New Info:
Aaron: Ran the 3C286 data through AIPS and got the same result as above. Showed the 3C286_A.im model that comes with CASA to Rick and he wasn't impressed by it. Rick gave me a fits image of 3C286 at 37GHz, A-band. I then used a script provided by Frank, written by Kumar, to turn the fits image into a model CASA image. I manually calibrated the same 3C286 data as above but used this new model image in my setjy call to get the below plot. A more official pipeline run that will use that model is going in: /lustre/aoc/sciops/alawson/test/22A-012/pipelineNewModel . I suspect that run will also show a correct Amp vs UVwave plot once it is finished. We should probably think about revisiting/verifying the all the models that come with CASA.