Closed scottstanie closed 1 year ago
Thanks @scottstanie. I assume the discontinuity was at two frame boundaries. Were you able to confirm? Updating my comment as I see the interferogram in the issue.
@hfattahi no I don't think it's a frame boundary, see here for plots https://github.com/opera-adt/COMPASS/issues/100#issuecomment-1452437338
broken off from the previous PR with the burst DB changes are
Also minor fixes to docstrings/default args, and using keyword args on the function calls which had gotten very long (I didn't want to mess up where i was inserting the new args).
The correction addition was prompted by the first test of
sweets
that @hfattahi did. He saw a burst discontinuity for one burst in a frame:I ran it with a different processor (the stanford geocoding one) that applies no corrections, and it looks like this in that area:
The strip of blurry decorrelation is because that processor doesn't account for shifting burst footprint, but notice that there phase on either side looks continuous, while the COMPASS one looks completely different.