international baselines

[mhardcastle]

It would be interesting to see whether we could run the pipeline at a factor 2 to 3 higher resolution using the short end of the international baselines. If successful, this would give us greatly improved optical IDs for bright sources, and would also complement (not replace) the work being done to do full-resolution imaging, since it would allow the selection of sources that are compact on few-arcsec scales more easily.

Some issues:

• this would need us to process more stations with pre-factor
• we would need correspondingly less averaging, so datasets would be bigger
• the TGSS-based sky model will be even less adequate than it was
• UV coverage may not be ideal -- need to think about taper vs cut, dropping the LB stations vs trying to use the short baselines including them, etc
• images would be very big -- memory issues

At the moment we image much more of the field than we really need so one option would be to try to subtract some off-axis sources early in the process (Cyril's subsquare idea revived).

We could make a better model by starting from the output of 6-arcsec imaging rather than from TGSS.

The current LB test field P205+55 might be a good place to experiment.