[evlatests] Time-Variable Closure Corrections

[Rick Perley]

    The aliasing problem should manifest itself as a slowly variable 
'closure' error.  My previous studies of high dynamic range imaging 
utilized a single, time-invariant closure correction, and showed that 
the assumption of time-invariance is valid only for those channels 
sufficiently far from the aliased response.
    I thus tried a time-variant correction on a single channel of my 
dataset -- channel 13.  I tried an hourly correction, with the result 
that the 'north-south' closure residual bar in the resulting image was 
completely removed.  However, the resulting image, although clearly very 
much better than the time-invariant version, suffers from low level 
artifacts (in the EVLA-EVLA baselines) which are not seen in the central 
channels where the aliasing is completely absent.